 Hello. Welcome to Portland, Oregon. Those of you who are here in person. And for those of you who are online, my name is Bush Cunhaler, and I am a dean in the College of Earth, Ocean, and Earth Science at the Oregon State University. And we are right now in a jeep. We are right now in a building that is part of the Oregon State University campus. We call this Oregon State University's Portland Center. The building was originally a department store, Myron Frank, in the 1920s, I think, and was completely remodeled. And during the remodel, it was also represented for a seismic fix. So we are in a safe spot if the big one was to hit. But of course, before, way before then, this land was the land of tribal nations, and you might not be surprised to find out. You might not be surprised to find out that what is now Portland at the confluence of the Willamette and Columbia Rivers was a very important trading post for the Indigenous people who lived here before. And some of those folks, some of those tribes, you will see their names represented in streets and counties. So Multnomah, Wasco, Powlitz, Plachamus, Chinook, Tualatin, and Califuya, you will see all of these names all around us. So if you know that those are the people who inhabited this land for centuries, for generations. And so I do want to start out with the land acknowledgement. Oregon State University, and I'm going to read our land acknowledgement as OSU really paid attention to the wording. So OSU recognizes the impact of this land grant history on Indigenous communities in Oregon. It was the Morrill Act of 1862, which established land grant universities in the United States. Federal government sees nearly 11 million acres of land, or 250,000 tribal nations, with little or no compensation. Oregon State University in Corvallis is located within the traditional homelands of the Mary's River Ampinapu Band of Califuya. Following the Malama Valley Treaty of 1855, the Califuya people were of course leaving to reservations in western Oregon. Today, living descendants of these people are part of the confederated tribes of the Grand Broad community of Oregon and the confederated tribes of the West. Indigenous people are valued contributing members of Oregon State community and represent multiple sovereign tribes among our students, among our faculty, staff, and alumni. This is a part that really resonates with me. Oregon State University accepts its responsibility for understanding the continuing impact of that history on the communities. And we are committed in the spirit of self-reflection, learning of reconciliation, and of partnership, to ensure that this institution of higher learning will be of an enduring benefit, not only to the state of Oregon, but also to the people in our communities and your fellow community. I think this committee really takes this part, right? These are not just words. There are actions that we can take as a committee in the spirit of self-reflection, learning, reconciliation, and partnership. And in particular, over the past months or so, we've had some conversations about what it would look like to include the voices and needs of Indigenous communities, and the tribal nations here and other Indigenous communities elsewhere in the West, how to include their voices in the story, right? So if we're thinking about ocean finds priorities, we want to take into account the ocean finds priorities of tribal nations and communities. We will hear as little as that perspective here today, but primarily we as a committee decided that instead of trying to invite folks from tribal nations and Indigenous communities into the meeting and give them a few minutes to speak to us or be part of a nearly one-hour panel discussion, we have decided that instead we will go to them. And I already have some meetings set up for November and December, where I will go and some of you will join and we will go and talk to tribal caucuses, specific tribal nations who have expressed interest on their time, and of course only if they have time and willingness to engage with us. You know, they have other priorities as well. And I know there are others on the committee who are making similar contacts so that we have a better sense of nationally what the needs of tribal communities are, as well as Indigenous communities in Alaska and Hawaii, for example. Okay, so with that, I want to welcome you all here into this space both physically and virtually. It is great to see so many of our committee members here. And Kirstie, we need to make sure you have a seat at the table too. We need to space together a little bit more, rearrange the tables to make that possible. Yeah, we can do it now. We can do it at the right. Yeah, I think that's good. That's what your meeting has been one that I've been really excited about. And I will tell you I had a hard time sleeping last night so I'm caffeine depth so I feel excited. Looks like I'm a little over excited that's where that's coming from. And the reason why I've been so excited is because we've been doing a lot of work that's been focused on ocean drilling lately and I want to thank those of you who really have done a lot of sweat and tears into our interim report so far. You've been doing an amazing job. And those of us who have not contributed as much because ocean drilling is not part of our, we don't feel like it's squarely within our expertise. Please know that we have been reading and thinking about what you've written. And we want to make sure that this is the consensus reports. So even those of us who haven't written a word must must make sure that we agree to every word that report. I also promise you that we'll do a lot of heavy in the same way. Along with well coming of course you all. The reason why I was excited about this report is because we have a chance to really open our aperture a little bit more and really think about ocean science overall. Right. Number two, I will remind you and we've had this conversation many times that each one of us brings to this work is specific disciplinary length. Yes. But each one of us who that lens responsible for thinking about the ocean clients enterprise as a whole. Right. If each one of us only advocates for our piece of the pie, we will have failed. So please keep that responsibility in mind. We are here representing the entire ocean finds community. And we are here from into here from some experts about their ideas. Now, the intention that I set for this week. I thought to myself, you know, what, what is my intention this week and words that I came up with myself. Were pay attention. This week, these next 2 days, I want to pay attention. And what I mean by that is not just listen to the words. But really pay attention to the whole. And so I invite you to think about what intention you bring to these next 2 meetings. And make sure that you stick to that intention. Right. That we bring that kind of intentionality. Let's see what am I forgetting logistics. There's some coffee and pastries and fruit over just on the other side here for those of you in the room on the other side of the panel. The bathrooms are around the corner to my left. Down that way and. Oregon has is promising kind of a mixture of an October day for you, which is very difficult time of year. I love October in Oregon. So I'm really glad that we're all joined up here for the students. Let's see our agenda. We have a packed agenda today. We've got a great show tonight. We have a number of panels. We tried to really build in a few breaks here and there so we can network with the folks who are visiting with us. But mostly, I would say we have very packed agenda today. Again, try to intention, try to keep the energy level up and let's make sure we take as much out of this. We really appreciate that our panel members and visitors have joined us. So travel far away to be here in person. So let's make sure that we get it at 844. I think I'm right on time. Okay, I'm going to hand it over to Jim. Yeah, I'm Jim Yoder, my co-chair. And so I'm sort of moderating the next session, which is on ocean observing infrastructure and innovation. And so we have three people. We have a panel to talk about the ocean observing initiative, and I think we'll just let them start. I'm not sure if it's going to go first, but I hope for tonight. So I'm going to share my screen. I think we're there. Right. Hi, everyone. I'm Jim. I'm the lead PI at the program management office. We really welcome the opportunity to talk to this group to the channel. This is going to be a jam packed presentation. So I'm not going to spend too much time on this slide other than to say, Ed Dever is to my right, your left. Who's the most you and the insurance array that Kelly to my left, right. This is from the University of Washington. It was me, the regional cable director Anthony conference, maybe online. I'm not sure he's at OSU and he heads up our cyber infrastructure. And I'll put him in who may or may not be on either doctors. We'll see from what's all that he runs the global crystal science. So we were given a number of questions. I'm not going through all these we hit, I think just about everyone, maybe not in the exact order. And we went off menu, maybe a couple of times, talk about various things. The one bullet that we will not address because you will see after our science highlights number three, we hit all subjects of ocean science. And so, well, without much further ado, we do need to get through these. I'm just going to introduce the arrays themselves. What do I looks like. And what we're looking at here is 1234567 arrays, two of which are then discontinued. We might have some discussion about that at some point. But of the five active arrays, they support 900 sensors, 80 platforms, and we're hoping to keep this thing running for 30 years starting in 2016. And so we're about eight years in. And so what are these arrays. There are two global arrays in the Gulf of Alaska and you can determine your sea. There are two coastal arrays on the West Coast, the endurance Iran on the East Coast, the pioneer and there is our regional cable array operated on the sea floor off the Washington, Oregon coast. And you'll hear more about all of these arrays. I did want to point out the coastal pioneer array on the mid-Atlantic by is no longer there. We're going to have to put a little discontinued in 2012, I think, or 2013. I forget, I think it was fall 2012. The first one. Yeah. So that will be discontinued. What we are working on now and you can ask questions about this later is we're moving that mid-Atlantic by pioneer array as the name implies, it's on the move. It's in its chuck wagon and moving to North Carolina off the coast there. A lot of excitement is going to go around that. So do you want to make it very clear this is a collaboration between the three institutions that you see on the bottom that you can see on the bottom of my slide between what's all you do and urgency. So that's what we've got. So I love to work with you a physicist because you get some really cool photos. Lots more than that, but you get some really nice photos. This is a nice shot of hard to film event that depth group routinely makes movies of beautiful array. I as I mentioned, I'm a meteorologist bonafide meteorologist from Penn State. But I'm a marine meteorologist. And so I work on your seat interaction and I work with oceanographers on the near surface process. So this slide is meant to do thing to things. Oh, I operates and maintains sophisticated instrumentation and demanding work. And you can see a couple of lots. Yeah, so that is a, I imagine it's a three meter distance. So you're probably up. Well, I know you're up six meters. So six meters tall is where the animal is. And you ask questions. So in demanding locations like this, and we really, really try at each of these arrays to make measurements from the sea floor to the lower atmosphere everywhere. Right. That's our goal. We a lot went into designing the UI and anything in this room will involve and many meetings, white papers, you name it. There were some curve balls thrown at us at the end. But we have a truly remarkable observatory here are some of our science teams that came out of all of those meetings. I can confidently say that we've hit each and every one of those science things. Through the operation of. We will probably work on another one of these documents working with our facilities board, who is primarily responsible for this document with our health. I will probably do something just like this. And that's some again, any upcoming five years. So I'm going to do my best. Oh, and actually, I can really do my best. These are out of the slides that I said I would talk to him on. This is our, the previous pioneer, which is on the new England shelf. And he's on his way right now to the men at my inquire, but this place was pretty amazing place to do research. And it really did result in many insights related to physical processes in the region and you can read them in the butt in the bullets. But I want to make this point. The bottom line is that the dominant processes are not the ones we expected. And this is to look at how we get stuff across the sheltering from, you know, well, it turns out it's pretty easy to get stuff across sheltering from even from Block Island, where we saw influences about the sheltering from animal school. So it's something that we did not expect. There's some talk about perhaps putting an element back, but that's that's to be discussed. And the other thing that I wanted to make clear is that this started out mean is a physical oceanography experiment that in the last years, more and more biogeochemists have been using more and more and more. That is one of our goals is to get the broad community using certain rates. And now we're starting to get papers out of all of these sub-disciplines. Next one is this Armander Sea. It's a great place to do work if you're really interested in nasty conditions, which I am. High winds, high waves, you name it, we get it. We've gotten very good at maintaining these arrays out for a year, except if you look forward for a year. And a couple of highlights, very challenging environment. I think we know that. And really, really strong collaboration with OSMAP, right? And some of you are familiar with OSMAP. We formed sort of the western edge of OSMAP East is where we are. Right at the tip, you can see the tip of Greenland. Now, this is something that Al wanted me to mention. It's easy to forget the idea that Arming your sea was not a hotspot for, well, was not a hotspot for convection, like deep convection like AMO. It is an extremely important region for maintaining AMO, perhaps even more than the Labrador Sea that was its competitor. And remains its competitor, but it's very clear that the AMO receives vital for the maintenance of that and the weakening of that as we're finding. And then lastly for Al. So Papa Janie, it's been here forever and ever. And one of the key things about Papa and why you provided there, in addition to the long time series of atmospheric measurements, it didn't have so much to say about deep water in that region. So that is what OOIs providing. It is, it sort of connects with the other regional observatory that means RCA and endurance and Papa are sort of a triangle for things like modeling ex-validation and initialization of numerical models. And it's going to talk about that a little bit. And I think that's, I think I'm going to stop there. So one for the meteorologists. I don't know if there's any other meteorologists in the room, but as I said, I'm a green meteorologist. The OOIs, the only continuous network, measuring directly measuring momentum, the only one, right? We go on our research courses and occasionally we'll put these out. I now have five years of data from these various points. With that, you can do things like improved model for immunization, model physics, using this data. And some of you may be familiar with things like the drag coefficient or the buoyancy coefficient is the dolphin stanton number we use. But these are all improved learning forecasts of wind waves and currents. And then lastly, there is a group working to direct outside of the MIOs if you're helping, that are working to deploy their system like a directly measured CO2 exchange. Not just bullpen by direct Americans. Now, with that, I'd like to end this readover today. So one of the great things about the original K-literator is that it has a submarine bi-rocket cable that's unlimited power basically in bandwidth. It's used all the data for over 150 instruments from streaming in real time. And we have two-way communications. It spans the sea floor to five meters below the air sea interface. And this is just one example that this is not my world, I'm a geologist, but it provides unprecedented full waterfall measurements of ocean acidification. Most cruises go out maybe once a year and get a few samples. In this case, we have over 45,000 profiles. There's three of these that span blue ocean three hundred miles off the coast of coastal environments. And so, and the other thing about this is that it is, sorry. So this is one of the profiles that goes up and down. So it goes up and down nine times a day, three different locations. And one of the great things is it's incredibly multidisciplinary. There's nothing else that like it in the world oceans. They've been a real workhorse and it's connected. PHCO2 is connected to 16 other instruments. So it includes biological instruments. So with this platform, you can get a really amazing data set out of ecosystem and how it responds to different events. And this is the longest time scale, highest resolution continuous record of the ocean. On the other side, so this is more of what I looked at underwater volcanoes and hot springs. What are the incredible successes for regional cable array? This is 300 miles offshore. It's about a mile down. And there's a high candy here, but there's a full suite of instruments that are not only quarter instruments for the OI, but it's been a real hot spot for PI instruments on that again. So these are externally funded projects. Over 70% of the volcanism on Earth occurs underwater. The Mid-Ocean Ridge is the longest county on Earth. It's erupting all the time, but we're never in the right place at the right time to see it. So this long-term observatory has captured the very first time live underwater eruptions erupted in 1998, 2011, and then 2015. In 24 hours, there was a major seismic crisis with over 8,000 earthquakes and 24 hours sea floor fell about seven feet. And there's a really interesting microbial story where during these eruptions, there's huge microbial blooms with billions of micro-stream currents. So it's a super exciting place to operate. It's also a hotspot for many, many researchers because of the array. It's the best image underwater volcano in the oceans. And it's really changed. There was hypothesis for restful environments that the volcanoes have stacked magnet chambers. These folding tanks of lava. And this is because it's in water. We can image actually ourselves. Since it was first installed, we've had, there's 43 investigators from 23 institutions and over $21 million in funding that's been supporting this outside of OI that's been supporting this research. I'll touch on it later, but it's also approved for a drilling program on foot course observatories in the south surface to look at the sea floor bias. One of the coolest things I think again, not my world is this is probably one of the most exciting things I've seen for a long time. To me, it's like bringing the first telescope into the ocean. So with these submarine fiber optic cables, you can hook up a system in the short station. It sends light pulses basically and interrogates fiber basically the way that the fiber behaves. During that, pick up a real-time monitoring of earthquakes, volcanoes, eternal lands. And the cool thing about this is that it extends so far right now over 100 kilometers offshore and about every meter to 10 meters is basically a sensor. They're using this on land. They're using it to look at glaciers, but this was the first community experiment in oceans. And it was phenomenal. And one of the great things I saw, this is from William Wilcox. Currently, you do it when the cable is turned off. We had a maintenance grid and they measured over tens of thousands of whale calls in just four days. When we talk about big data, this is big data. It's terabytes of data per day. And now there's new systems that are coming on. So you don't have to turn the cable off. So this is a great opportunity and I think look in our future. We'll hand it over to Ed. One of the main reasons we have observatory science and oceans observatory is to look at long-term changes in things, including effects of ocean acidification, hypoxia, wave climate changes, etc. In the coastal and in the ocean. And, you know, endurance arrays is one example of this early science. And in particular, recently, there's been a lot of interest in green carbon dioxide removal, wave energy, offshore wind energy. There's interest in using long-term observations or backgrounds, environmental information. I think one of the more important things that observatory science gives to the community is the validation extrapolation of in-situ measurements. And time series measurements are quite expensive. And in particular, when you're starting to take a lot of different multi-disciplinary measurements, the way the ocean observatory system does is something that can't be replicated at every place. And one of the big things that people want to use to replicate or understand and extrapolate pushing observatories measurements are remote sensing. So this is some early work done by Hendricks by the Sol right here. And what it shows is in-situ chlorophyll on the horizontal axis and satellite retrievals on the vertical axis. And what it shows is that, you know, if you just look at the chlorophyll estimated from satellites, you get much higher estimates off of Washington than we do off of Oregon, in particular in the near shore. And the indication here from the comparison with in-situ measurements is that the scattering is actually affecting the satellite retrievals of chlorophyll. It gives us a way of better interpreting those satellite measurements. And that's one of the reasons that they go up to four now proposals to the NASA pace validation were asked to use Hawaii data or the world of Washington. And then similarly in-situ measurements can distinguish between surface phenomena or subsurface phenomena. In particular, we're seeing climate change play out not as a steady increase in temperature or decrease in oxygen, but in terms of events that happen rapidly. And the in-place measurements provided by pushing observatories are one of the ways to look at those measurements in real time. And so this is an example, you know, the famous and from this world law from the early part of the 20s. And just using satellite images, what's seen in 2014-2015 is a pretty impressive area of warm temperature water. That's that area is replicated in other events. But what really distinguishes the law is its persistence, but also the anomalies seen beneath the surface rather than just atmosphere. So that's what's shown at the bottom slide there in the space using Hawaii data from profilers at the station. Some of the kind of ongoing innovations in ocean observing, you know, I think one of the more important things we can start to do now is to, is to pull together ocean observations from a number of different networks. There weren't these specific as one of the most highly sensed areas of the world. And what's shown here, this is from a 2019 paper by Barthendoll. The figure is from Ocean Networks Canada, and it shows the number of observations it's even hard to see at this scale. There's a large number of observations that span from the northwest of the United States up into Canada and include not just Hawaii Ocean Networks Canada, but centers from government agencies, private and now even close to tribes are making sense. There's a lot of Oregon and Washington. So that's the part of improving our ability to work across these observing networks, combining data from them and make the whole greater than some of its parts. This is a worthy goal for the next 10 years. Other things, you know, on the previous slide we showed the Marine B wave, better mobile technologies, I think that's important, in situ and detection and identification with changes in ocean ecosystems. The Ocean Observatory's initiative of sensors that we purchased back in 2012, they include measures of ecosystem function, but we have much better sensors now. They're going to start to look directly at DNA, indicators of human composition, better indicators of ocean port activity, coral algal wounds, and so forth. And then, the validation of the observatory data is important, not just for remote sensing, but also for models including physical models, by the chemical models that the model that our program created is the live ocean model that's shown on the right-hand side there. And if you look carefully, not only is it just temperature and splinted, but also oxygen, this model has in it things like arachnid saturation, nitrate, any number of different parameters that include, again, measures of ecosystem function, and we can now start to validate those models using this data. One of the challenges in terms of this validation is that modelers have a tough enough job ahead of them just making their models run, and they're not expected to be experts in all the conventional sensors. As we start to get into sensors that go beyond things like temperature, humidity, wind, and velocity, it's important to find chemical sensors. We've heard from the community that modelers in the first century are poor data sets with quality and full data, regular time and space and calculation, and embedded data, including error bounds. And putting it together in that kind of data sets can be a big challenge when you're taking as many different types of measurements as something like OLI. One of the things that we do, I think, a positive impact, a broader impact on the community, is that as we operate so many different types of sensors, we have some leverage with manufacturers. And so these are some examples of ways that we've improved measurement technologies through, I'm not going to name specific vendors here, but you can see them on, as indicated on the images, we've improved the vendors, dissolved oxygen to flow calibration procedures. We've worked with provider manufacturers to update their firmware to lower power consumption and meet specifications we developed for OLI. And then, especially on the cable array, we've worked with manufacturers to improve groundfall testing for external requirements for this. We have annual meetings with vendors to coordinate maintenance schedules, to discuss product updates, issues, obsolescence, et cetera. Observatory is also possible because people research and workforce development and a number of different ways. I think it's a great plot. It's worked out by Briana Velasco back in 2020. It's part of a virtual ROU program that was run by Rutgers University. And so the Rutgers Ocean Data Labs are probably familiar with it. Many of you are probably familiar with it. Well worth checking out in terms of their ability to facilitate use of observatory data for student lessons and for student research. And what Briana did here, she was working with Rachel Hevella at Oberlin College, so going outside the bounds of sort of traditional oceanographic institutions. She's looking at ERC, PCO2 off of Washington, or the Cognitive Shelf, and you can see here the ORM shows the PCO2 in the ocean, the blue is the piece of air. And you can see here that the upwelling system is a significant state of carbon dioxide. So again, observatory data in conflict with other things can really enhance student research. So one of the great things about the ORM is it's a program out of the University of Washington to take undergrads. Grad students are the group folks to see with all of our pre-use. Our students are usually 45 days. Over the last few years, we've taken, that's the last decade, we've taken over 200 students to see. Some familiar ones, Deb Lixon. So these are a couple of highlights of Deb Lixon. We know that Deb Lixon's student is now a director of science and resources, National Academies. He's worked with us for a couple of years with an undergrad. He went out on numerous vision cruises. He's director, director for Water Instruction and the White House. And we just got a new NSF postdoc. So really forms of foundation for workforce development in many different areas. And the students go on to work in engineering. There's no requirements for the students. Typically they'll go for two weeks to 45 days a week. And many of them have gone on to work for industry, biotech firms, geotechnical firms, academics, coast guards. So really we're bringing up a lot of students say this is fundamental change in life. So I'm really happy and humbled that I get to do that. So we have, there was a question about our impact, our measurements in life. And this is a nearly a decade of ocean observations through the observatory. This is just one example. If you could hear that, you would hear that it's whale calls. It's not your phone data. So we can track mammals off. It turns out that thin whales, fossilizations have changed over the last decade. I'm not quite sure why. The best hypothesis I like is with the embargo on whaling, they don't have to yell as loud to scare their friends. And then I mentioned already one of the newer instruments that we have, which is also really exciting. So the DK sonar, DK 60 sonar, it's a solar that images organisms in the upper water column. But in this case, this is a great example during the clips where the zooplankton, they go to the surface at night, so they don't get eaten. And then during the day time they go down. And so they, they were thinking during the close to come back up again. And we just refreshed this. There was a question about fisheries, this DK 80 sonar. It has two different kinds of functions. And one of the functions, a broadband function, enables it to actually quantify fish in the upper water column. Now, so that's a really exciting technical look. So we also have a variety of cameras in near shore, deep, and then clear off with access C map, access C map. We have a dedicated fiber for that. This seems video of the hypothetical events, looking at the vent communities, how they change over time. It's completely the only place in the world. And that can have one also that looks at the biology, how that she means a lot over time. And there's a great average. There's lots of students that are working on that vent imagery. So we get to quantify the temporal spatial changes and vent environments. We also have a really unique instrument. I think except for the EPS sampler on by a bar. This is one of the few samplers in the world that we actually have a tube that goes into a hard to thimble vent. There's a lot of interest in viruses on the deep biosphere, how the microbes in the oceans evolve over time, especially when there's volcanic events or earthquakes. And so this is an in situ sampler that measures the samples of fluid temperature and chemistry. So you look at the change in microbial timings over time. And it's a NSF awarded recent Anderson, a little career award and she's been out there for about four years looking at the samples and looking at viruses. So these science series really provide a, one of the kind of time series of microbial viral viral management over multiple years. Okay. So in order to deliver all this data, we have to have a world class cyber infrastructure on the system. And we have just that at Oregon State University at their data center. It's really is keep in mind that the day, all the data we're collecting is freely available to any, as long as you're on the internet. Actually, it's democratization of the data. We're trying to be here. The center itself is just focusing on low risk, but cost effective ways of that allow us to do things like improve compute power for things like Jupyter notebooks. It's what they work on. Modernizing the storage facilities, both in Corvallis, but also in bed for additional safety where they're real big on cybersecurity in our data center. So you can see that provides a secure daily. They do file a fair and trust principles and we do provide a lot of extra storage. It's not infinite, but we are expecting the data to be quadrupled by the efforts program. And so we are planning ahead so that we can accommodate that. This is just some metrics that we use to show how users are near the type of usage of this data. The first one is, you know, I will say the old way does not require users to register before they use that was a decision made a long time. And there has been some discussion about that, but that's how it remains right now. So we have to find creative ways of counting people. And one is through Google analytics. You know, take it with a grain of salt, but this profile shows the accumulated number of distinct users based on IP addresses. Since the data explorer system that we developed was christened in 2020. And so there's, you can see it's well over 5,000 users accumulated because these are distinct. It's accumulated, but if you go in 10 times, you get counted. That's that's the deal. So you can see from that curve, lots of users. And of course, publications are our currents. And we can see there on the top right, that from starting in from OLI 2.0, which began in 2018 to the present to OLI earlier 2.5. And we've had a dramatic increase in the number of publications. The last column 2023, that's halfway through. We do have calendar year. So I have a whole another group of publications that will be added and then one more before Christmas. And I'm hoping to meet the record, which was just. This one's kind of cool. As I don't know, people realize it's very difficult to get information about funding, especially dollars. And so you can get it through NSF, but that's not the best search engine I've ever used. So what we've been doing is, is data mining, data mining acknowledgments of all of the publications OLI publications to find out if they reference a fund. And most people do. We're smart. We want NSF or O&R or DOE to know that we're using their money wisely. Here's the publication. And so what we're seeing down here is it's still primarily NSF, as you might expect, but you can see it's a broad. It's a diverse number of funding agencies, including our NASA know, but interestingly, Europe and the UK are two of our biggest users. I think they're, of what the usage is through those, all of the countries in the EU and the UK. Yeah. Does this break apart funding when it's like multiple funders? So like, let's say someone. That's a good question, right? And please note, this is not money. This is, these are people that are indicating they use the data. And to your question, though, yes, I will, because it's, it's a, it's a bit ad hoc how I'm doing it. But if someone says NOAA and NSF, I'll check both. If it's very clear, I'll check. So again, it's because it's counting and it's not money. That's a good way. So let's keep moving. Early on, it was. Iris is basically the data base that everybody puts their seismic data in. So we did that from day one as far as we go to look at the seismic data and store. And so this is just another population of really good population of people over 800. You need addresses. And you're looking at the seismic data. They also bring in low frequency, high performance on pressure data. And planning is underway to bring these data because they're coming in at real time. There's a shake alert system that's now been installed. And so bring that into early with quite boring system because we have very poor measurements. We can do it on a continent, but very poor measure. It's not, not as good as measurements in the ocean. And it's kind of critical for figuring out when a seismic wave is going to hit the map of condition. And I should also say that from the devote who are quite one of the big lessons that they really needed. But that's not being ensured. They really needed near shore measurements of pressure. And our system is one of the few systems that has that near shore environment. So we can have coastal warning on this in place for some. Market line in 2006 said, if you build it, we will cut. They will come. And one of the great things about RCA is it was designed and built for looking to build a seismic data base. And it was designed to build a seismic data base. And it was designed and built for looking towards expansion over 30 years. So we have a lot of bandwidth and power for people to add into the sun. We have over 78 total funded awards. That's PI and several awards from lots of different institutions, 38 institutions. Over $46 million of outside funding has come in on to OOIs for the RCA. And a really diverse portfolio, which is something OOIs proud of, NSF going on and NASA. And this doesn't count as ship time. So if you figure a ship and an ROV is $100,000 a day, it's a big investment outside of OOIs. And I'll put this one up. I like this one. There's always been questions about OOIs and who's taking advantage of all the data that's coming in. And there has been a tendency for our PI, so our colleagues to think that it's all going to benefit the MIOs. UDUB, OSU, and who. Yeah, well, they benefit from the data. These are just researchers at those institutions, of course. I'm going to do research for three of the biggest, three of the largest research institutions in the country. So we're going to have some large fraction of PI. But what this shows, it's not the dominant fraction. It's about a third versus two thirds publications by institutions. All of the yellow is all of, a lot of that was all of the other science highlights that we provided are from outside our home institutions. And the same can she said about the NSF awards, here it's, well, it's actually closer to a third there. It's a quarter of publications by institutions. So we hope this kind of puts to bed this notion that we're the only ones that, that is not true. Everyone benefits from OOIs. So maybe I can take this, this is, so this was the parade for us slide. I have to say we submitted a proposal in 2022 to renew the OOI for five more years. And we heard about three weeks ago, I would say, that we were provided that award. And we are another five years. We're extremely happy about that. So this running is sufficient to maintain and raise them operational for another five years. It does provide as, you know, if we get what we are hoping to get from NSF, we does provide for significant support for increased capacity for the cyber infrastructure. It does provide funding for small teams, data teams for QA, QC activities, and something we're calling data ambassadors where we will set a team of scientists and data specialists to various places, community colleges, research universities, you name it, to explain how to, what the use of OOI data and how to access it. So that's something we hope to be able to do. And then you could look at us up here. We're not getting any younger. We really like to bring in some younger scientists to learn at RD before we move on. And this would allow us to do that with three associate project scientists playing with the OOI team. And I don't know if you want to speak to this one. Go ahead. Okay. So this is a nice point. I think Debts made this. The OOI team has significant breadth of little depth, right? We do cover a lot of things, but we have to grab someone who's working on, you know, by due chemical sensor quality assurance to go to C and run with CTD, which we don't have people dedicated as much as we'd like to things like data, especially quality control, quality assurance. This would allow to have us these enhanced data team members. The OOIs get kind of old. It's a lot of the equipment has been out there for 10 years. And you can only refresh instrumentation so much before it just breaks and there's no fixing. So we are desperately in need of both tech refresh and physical infrastructure refresh along with cyber infrastructure. And also we have these real exciting new technologies that are coming online, like the flow side of OOI. People are really excited about the pioneer array and the things we've got. I'm sure it's going to find its way to the environment's array and not do this in the future. So we show these things because we have the question, what if we get level funded, like we had been level funded for five years at 44 million for 2.0, no inflation, no increase to the tech. We got it in 2.5. If that doesn't hold, if we don't get the anticipated amount of money, it's going to be very hard to conduct these sorts of things. And something's got to get if that happens. So we're hopeful that it's all going to work out just fine because we're really excited about the opportunities that the new budget provides. But let's keep it finished. And then. So one of the questions was about the interface of the, do that this regional table rate team interact with the seismic seduction zone for the community initiative. And we're super excited about this. William Wilcock put in a MSRI proposal. It goes through a lot of pre-proposal, full proposal. We had a reverse site visit with NSF. And we were, he was just awarded, their DIs are from the UW and Scripps. And this is a $10.6 million award to add on to the infrastructure to look at the seduction zone on the Cascadia margin. And what you show here is the two table arrays that come out from the city. The one on the bottom runs out to Axl-CMAT. And then there's another one that runs up to often Newport, Oregon. The exciting thing about this, so we do have a small suite of instruments. It's very rare in the world's ocean staff instruments on the subducting plate as well as on the margin. So it's very hard to look at the coupling between those processes that happen when the plate's been seducted. On the Cascadia margins, so the last major nine magnitude earthquake was in 1700. Again, tsunami came ashore. It's had multiple breaks. And it historically, when a magnitude kind of earthquake occurs, it basically unzips the whole seduction zone from Northern California all the way off to Vancouver Island. That plate has been locked for a long time since 1700, meaning that we don't see a lot of earthquakes. The only place, luckily or not, we do see a lot of earthquakes is right where our array is. And those are those green dots. So the interpretation is this is the one place in the Cascadia margin that it's not locked. And so the 4D community has also been in this. The next one. This is the infrastructure that we're putting in. These are one of the nice things about this is some of the instruments were funded by Princess. They have the P.I. instruments on an accesse mount. Would not be that they would be tested there. And then they could be transported to the Cascadia seduction zone. And that has happened. So we have lots of pressure sensors. We're putting in three new junction boxes. These are small substations on the C4. And these are big enough. They have enough power and bandwidth that there's fours available for other research scientists. So it's not just for this installation. It'll be about 13 days on installation in 2000. 26. There's also educational programs. And all these data will flow into Shackler and the Know-It's tsunami warning system. It involves the P.I. There will be a science oversight community or science advisor committee. And this includes involvement with Crescent. There's a meeting going on this week. It's a large $18 million reward to OSU for Cascadia region. There's quite a science center. There's the 4D community. I've seen a lot of endorsement letter from one's proposals. So there's been lots of communication between those. And then there's also a large program for looking at the relationships of indigenous people hazards along coastal. But there's also a question about the IODP. So there's been a long history of IODP drilling. Both on the Wadafika plate as well as along the margin there. And there was a series of three proposals that have gone in. There's one from UNC or from Canada. It goes off of Vancouver Island. There's a full proposal 947, which is to look at the four goals that court conservatories from the north and the south. And then a particular interest is, I should say, you know, there's no place in the world that's planned, spanned on one of the pipeline right now in terms of looking at cable conservatories and putting court conservatories. You guys heard about these downhole. You make downhole measurements. So high-sphere seismic, lots of different properties. Williams, again, you, there's lots of PIs. So it's not just going honestly. There's a pre-proposal that went in to quick. So the idea is that blue circles there are the drill sites and then court those in cable. So again, we would get real-time measurements over years to look at the deformation and the hydrogeology along the Cascades. And say, I want to ask you a little quick question. And you look at your NSF stats for your budget, does that include the cost of ship time or is that not included in your money? So as of this year, starting with this year, we will no longer get the ship support, you know, to our home institution, which just has been a pass through to new halls. It'll go like every other proposal NSF to you. So it's not included in a wide budget. So it's about 220 million that is being shared amongst the MIOs. It's about 61 million. They removed that's going down the street to ship. It's not an ROV. And the other quick question, just one for all three, actually is if you had, you know, a couple of minutes with the NSF director, what would you tell them was one of your most or the most scientific discovery that you're part of the observatory that was made the last, say, five, ten years? Who wants to start? So in my, so it's sort of the lead PI, I kind of see all of these observing systems. I would say, and I've seen Al Fluteman give just amazing talks about what the pioneering is doing for science on our coast. But I've been really impressed with the armor you see. And the cooperative nature of the science being conducted with host snap. We really do help. We help them. They help us. We have gotten some really nice papers written as a result of that. And so I'll leave it for me. That's Contribution. It's contribution. Whether it's weakening, et cetera. We are making some of the data that will allow them to say definitively what's happening. From the geological sector, as they ask to see that, we had during the ridge program, there was a set of views of the stories that this is the first place, only place in the world where we're getting all the real time data. And so we have to look at the biological to physical chemical properties that happen. You know, something that forms a face of our planet. But it's also the only place where because we have a time series data set on how the volcano reforms, that it's the only place right now where we can forecast when the next eruptions can take place. So again, and I think the thing to come out of that is that we get the real time data and it impacts not only the chemical nature of the hydrochemicals, but also the biological activity. And that's been there's no place right now where when a volcano is erupting that we get there fast enough to look at the some of the major things that happen in terms of heat, chemical, biological transfer from beneath the sea floor to the water. For me, it's the short time scale and space scale variability in particular the current system. The rapid changes in each CO2 and the associated kind of ecosystem changes that occur. It's not enough to have a lot of measurements couple of times a year. The real time variability is extremely important. And could I add as the meteorologists know, I'm really proud of the fact that we are improving marine forecast of ring waves occurrence. And the fact that we're directly measuring these fluxes surface stress and heat exchange is making that awesome. And so, you know, I, I, one of my missions is to bring more meteorology more atmosphere sciences into the program. And I think you can get them excited. I'm going to remind folks in the room to both respond. So we were thinking to equalize the online and in person participation that those of you in the room, if you have a question, please put it in the fly go. And if Jim choose to use that question, then you can, you can ask it yourself if you're in the room. But that way we'll have four representations. Does that make sense? Right. We'll start with the brandy on this question. Okay, thanks. So my question is, is the previous source of both was fairly critical and raised a number of key issues and made a kind of negative recognition of the program. And so, I mean, I don't want to relive all that experience, but I'm sure that was enjoyed the time. The question really is, you know, how have you responded to that? And where does OI in your view sit relative to those critical comments? I mean, I can speak that it's not as painful for me because I've only been at this for two years, right? And a year and a half. Because what I've seen is they've addressed the two main issues, which was difficulty with the physical infrastructure and getting things turned on time and getting sensors. And also the cyber infrastructure. I think we've done a fantastic job at turning the race. I think we've figured it out. I also think that the cyber infrastructure center at OSU is now delivering data. That was the whole point, right? Is to get data out to people to archive it so that we know that we're going to have it for 30 years. Maybe I could talk about this a little bit more. But those two things were kind of key things. Things that were not working well that have been fixed since 2018 when OI 2.0 took over to the point where in fact, the matter is NSF buy this with the opportunity to renew, not recompete. Because we were doing a much better job in the middle. Yeah. If I can just chime in on that, because when I was at NSF, it was when the transition occurred. The previous OI that the previous DSOS was responding to was administered the main pride was consortium corrosion leadership. And so COL was the pride. These three institutions were the subs. That structure was very difficult and challenged. I don't want to speak for NSF now. I can speak for NSF then. Very challenging. So it was recreated to be NSF to be OI 2.0 five years ago. And this team won that competition. That was the fundamental in my opinion. Fundamental change facilitated the type of work these groups are able to do. Okay. I'm going to go down the list here, but so too bad. Follow-ups. Yeah. So you spoke to some of the things that you did fix with 2.0 to really get the data out to the people. You know, I'm drawing some parallels here with the IODP program that we really should dive into, you know, a lot of these data, but how do you make that handshake so people can really use it? And so can you speak a little bit to the remaining barriers? And I'll also add to that question. I was impressed to see that different uptick in 2018 and publications, of course, publications lag the data being a little bit, right? So it's probably coincidence that that's right. In fact, whatever you did probably is right now is going to lead to different publications. So tell me about the remaining barriers though. Where do you feel like NSF could help you, or we could help you in the work of getting the data in the hands of the people you like? I'll take the first shot at that. I think one of the main barriers is to chemical and biological data that the sensors used are significantly more sophisticated and have significantly greater issues than sensors that measure physical and physical parameters and use a lot of parameters. And so we have a significant amount of expertise in house, but it's a limited number of people. And we are working our way through what we think of as industry standard, which is the NOAA developed core talk, quality assurance of real-time oceanographic data. We're working our way through that, but in many cases, additional quality control is needed to make that data improve the utility of that data. There aren't a lot of observatories that are making, a lot of pen series measurements of ocean are still for a relatively limited number of variables. And so we're, you know, it has a larger number of types of different variables than many other different types of observatories. Probably the closest to analog might be the Bi-Jic and the Hargos. We can learn from Bi-Jic and the Hargos, but we are active participants in ocean best practices. So we're engaging with the community. Some of the things that I think can be done include, you know, facilitating subject matter experts to use and improve regulatory data is something that we've done as well. Because with the staff we have, and we've used data numbers that we've been able to hire, it's never going to be enough. I think one of the other things is maybe a generational thing where, you know, we have such a fire hose of data, one hertz data, or higher. Imagine that with 18 instruments all at once. And so I think that some of the biggest breakthroughs we're going to require new eyes on, how do you visualize those in a 20, 30-year time series that are changing over, you know, very short to long time scales. And so we've had to move one of the stomach blocks as the data sets are so large now we've had to move our QHC on in the cloud because just to visualize the data along the QHC requires that we need to move it on the cloud. And so I think the younger folks, they don't mind, they're using Excel, right? They use, you know, Jupyter Notebooks or Python. So they're getting very enjoyed, they're super excited about it. But I think there's a generation that doesn't quite know that technology yet. And so I think one of the powerful things on my line is AI. And I think we'll also need that for QHC. A couple people have asked about the relationship between an OOI and I, if there is such a relationship. How does that work out? You get a lot of work. Yeah, so I think, Additionally, there was the start of OOI, there was a statement made by NSF that the OOI was NSF's contribution to IOS. I think they later walked that back. But having said that, we are partners with IOS in many cases, the regional arrays, our partners with their regional counterparts, the regional OOI arrays are regional counterparts. We attend a lot of the same meetings. Our data can be, because our data is only available, it can be picked up and reused in these systems and it is reused in these systems. And we also, again, kind of thinking about a broader collaboration with OOI, OOI Neurological Data is going to NBBC and that enhances the use of that data by researchers as well. It provides that data to the global telecommunication system, GKS, which despite its kind of strange password number, strange name, is the kind of operational way that oceanographic and neurological data is delivered for model forecasting. But that is getting used as a weather system. And I do want to mention that the ocean sites rule. We were also starting to engage with the ocean sites. I had a chuckle in this, I remember back in the early 2000s many arguments, no brisket. This relationship, the difference between the ocean site and the ocean site and the ocean site. And this has taken pains to tell us we are not an operational system. But nonetheless, when we are operating, we make that data available to operators. We'll also ask the share of the data. Can I have a follow-up? Yeah, go ahead. I think none of these are going to go away. I'm also interested in how Oval Eye, I used this other concerning entities. How do we get a comprehensive look at what's being measured, what's not being measured? I think we look at these program by program, but not in like a comprehensive, not a great visual from the ocean Canada that's kind of what the family was posed. But it would be interesting to know what's missing, what are those signal areas that should be looking at them. I think that's something we should be thinking about and leveraging on. Yeah, I think there's a lot of examples now. Things like GliderDoc where it's a way to bring all those people that are running gliders on both coasts into a central system. We are definitely talking more and more with like I just mentioned ocean sites and Argo. How do we bring these together? Is it just going to be a web page where we have linked all our stuff? That would be a start, right? But that's the type of conversation that we're having now. We are bringing that conversation actually to the COP 28 meeting. We are organizing a session. It's actually, it's not, it's called a site event, even though it has like 200 people are invited to be site events. And it's being co-run by Olaibu and Pogo, which is the Plymouth Marine Laboratories facility with COS. So we're bringing the community together. It's a start and it's, you know, likely we are sharing data directly with NBC. Whether we can do that on a larger scale, we'll work on. I think we would all like to see that happen. Yeah, we're running out of time. I think it's easy to do that question about ocean data assimilation. Right. I was curious to know if you have any examples of modelers who are actually assimilating data real-time through weather, ocean weather. So glider data did go to the glider data assembly center and glider back. And so that data is incorporated, I think, in data assimilating models. The person that I think of most, it's most closely associated with data assimilation. I'm not sure whether he's now using photo-eye data in an assimilated way or whether he's still using it as primarily his validation. Is Alexander Korobat, who's associated with the West Coast? I guess just that's a, the intention of every kind of question is trying to see what might be involved for assimilating data. And the people that I've talked to, including Chris Edwards and other folks that assimilate data that did point it out over and over again, the, from their perspective, the barriers are knowing the, getting a data set that's interpolated on a relatively even basis, but also knowing the error bounds. And in some cases, it's not a matter of knowing the exact sense of accuracy, because we have specifications on that. But really in the end, what's the in-situ error bounds of the data delivered as a whole. And that's a bigger question than the specific accuracy of being censored by a scientist. So there was a, at the Ocean Observatories Initiative, SOTI board, there was a lot of respect in our models there. And they are very strong opinions. But they would like to basically give them curated data sets. So completely, you know, take CO2, take all the flyers out and have it in a specific format that they can just plug it into their models. And that was a, I think it won't come out as a strong recommendation. It's been something that we've shied away from to some extent, just because of the enormity of the task. But there's also a place where quality control starts to shade into scientific interpretation about what's good and what's bad. What might be good for one purpose is not necessarily good for another. In many cases, I think data that the model, modelers they want to use could have looser error bounds on it than people that are interested in very finely looking at, you know, long-term trends. And then the evolution of 1.0, 2.0. Early on, we were told, we want you to give us the raw data. We don't want to work on the data ourselves. Okay, I think we've got to move on. So our next panel is made up of a bunch of physical oceanographers. We have one of them here today with us in person, Melanie Peeing. And am I correct in that the remaining four are online? Yeah. Hello, hello. I'm now using a microphone for those of you who are on Zoom. Maybe this might improve your sound experience. Okay, I'm having some echo here. Melanie, can we invite you to be up front with us? Would you like company? Shall I sit with you or are you okay? Thank you for going alone. I'll know you're right there to my right. And do we have our other speakers online? Lauren Thompson, Craig Lee, Jack Barth, Joe Schumacher. What do you say, Zoe? Are they all with us? Okay. Excellent. Thank you. So this really, this panel is about challenges and opportunities in physical oceanography. And part of the reason why we wanted to hear specifically from physical oceanographers. Well, one is there happened to be a lot of high quality physical oceanographers in the Pacific Northwest. So we wanted to take advantage of the fact that the meeting was a little bit different. But also at some point, you know, we did hear that perhaps we didn't have a lot of physical oceanographers on our committee. Some of us feel like we know a lot about physical oceanography, but may not be card carrying physical oceanographers. But I'm inspired to hear more here from these colleagues, especially because just yesterday afternoon, we celebrated one of the great coastal physical oceanographers, John Allen, who recently passed away. And it was really great to kind of reflect on the approach that somebody like John Allen would take on combining observations, models, and really getting a deep sense of processes. So really approaching it from both in miracle modeling, theoretical and observational perspective. So, and I think this panel today will be able to give us that multifronged view of the field as well. Okay, with that, I'll pass it on to Luanne. I think you are first. Is that correct? I'm up. I assume you can hear me at this point. Very well. Take it away. Okay. So I am going to talk about dynamics, modeling and climate. I am going to give kind of a large scale perspective on this. So if you could go the next slide, please. So first of all, I wanted to say that, which I'm sure you guys have already talked about, that sustained observations are needed to detect and validate models. So things like satellite, sea surface height, sea surface salinity, sea surface temperature, et cetera. And of course, Argo, the temperature and salinity profiles. And then as we get towards longer time scales, we repeat hydrography program as well as deep Argo. So go ahead to the next slide. So I wanted to sort of set the context for why we need extra effort in this area. And to remind you that the climate model inner comparison projects that are models that are used for climate protections and projections and IPCC have about a hundred kilometer grit. And we know that that misses an awful lot of the dynamic. And in particular, I chose this figure because it shows the really large biases and sea surface temperature in Eastern boundary uphelling systems where up to 40% of global fish catch happens. So when we think about climate models and what's done in the IPCC, it's really missing and has an error. A lot of the important things that we're going to do is we're going to look at the ocean scale. So this has an error. A lot of the important regions of our ocean for people in the planet. Go ahead to the next slide. So I think the challenge for ocean modeling compared to the atmosphere is that the ocean scales are small and the motions are slow. So this is just a snapshot of sea surface temperature from August 12th, 2023. And you can see Eddie's lots of small scale variability as well as large scale imprint of climate. So can we increase the spatial resolution particularly to the sort of climate projection class of models from 100 kilometers to say 10 kilometers where we're starting to resolve Eddie's. It creates 100 times the amount of data and the time step has to be smaller by a factor of 10. And that makes those models about a thousand times slower per model day. And we have to recognize that the ocean takes centuries to spin up and that we need multiple simulations for climate projections and for climate predictions. In addition, this is big data. Really, really big data. We're talking of petabytes of data. So the analysis needs to happen where the data resides. This is not something where you're going to analyze the data and analyze it on a local computer because we're talking about petabytes of data. In addition, regional downscaling for projections can also be used, but those regional downscaling will inherit global model biases. The next slide, please. Okay, so those are sort of the big issues I think for the future for both climate projections and for climate predictions. But I wanted to bring up another issue that I've thought about lately, that workforce issues and ocean dynamics and modeling. There's been a brain drain to industry that's pretty astounding. Five scientists from the ocean section at the National Center for Atmospheric Research. So these are the people that are creating the ocean models that are sort of the community models for global climate prediction kinds of things. And one professor from Columbia University that is now working in industry and no longer working on ocean modeling and dynamics. The other thing that I've noticed recently that there's actually been little progress and maybe even backsliding in gender parity in ocean modeling and dynamics. So here's a few examples. Even before this brain drain, there were zero women on the tenure track equivalent in the ocean section at the National Center for Atmospheric Research. There are two women out of 12 in the federal workforce at NOAA GFDL. Could you go back, please? The Ocean and Cryosphere Division. And recently I went to a workshop and an invitation only and only 25% of them participants were women and it was on dynamic sub mesoscale air sea interaction workshop. Okay, next slide. If I have time. So I wanted to also put in the wild cards for thinking about physical oceanography as we go into the future. The rise in offshore wind and wave energy and thinking about what are the modeling prediction and workforce needs within that. And then also preparing for rapid sea level rise, Greenland and Arctic ice sheet melt caused by potentially regional ocean mourning and what is our role as physical oceanography as we look towards the future. Okay, I'm done. Great, thank you. Thank you very much, Luanne. I think Craig Lee from the University of Washington is next. Great, thank you. I'll show my screen here. Can you guys see that? Great. Okay. Thanks, everybody. So what I'd like to do today is just touch on four subjects briefly and then open up for discussion once we've all gone. I'd like to talk a bit about the need for sustained climate scale observing, coupled with that access to challenging environments. To me, a big part of the challenge in physical oceanography right now is it's not physical oceanography as a distinction unto itself, but doing science at the interfaces between disciplines and areas and then also sustaining expertise and I'll expand on what I mean by that. Going forward here. There we go. So first the need for climate scale observing, really sustained large-scale distributed observing through the global ocean. And I'll use a lot of Arctic examples here because it's something I've been thinking a lot about. But these are figures from one of Alexander Yan's papers. And here she's talking about climate trends, right? And Arctic freshwater balances, in particular the exports of the gateways and the storage terms run off from the river's precipitation. The different components of that climate system and you see elements of that balance in the middle panel there. And then the more interesting part of this is if you look at the right, she's looking at a long-term record from CCSL models. And in particular, looking at what you call a shift being the first year that we see a change in these fluxes or storage terms that's outside of a three-and-a-half standard deviation range from the 1800-year pre-industrial control run amine and emergence that the trend remains outside of that range. And the reason I've got that here is right, if we want to start seeing things like this, we need to start seeing other small-scale climate trends, if we want to begin to understand some of the long-time scale dynamics behind them. We really need to be looking at these time scales and these are processes that unfold over these large space and long-time scales so that the effort itself is different from the idea of monitoring an efficient exhibition or a bill that may become the duration that we're talking about exceeds the span of individual careers. But we don't really have the kind of funding models we need or we need to develop funding models and we'll facilitate these things. And culturally, we need to learn how to recognize and reward researchers for actually making these contributions. I think these are tough problems that go well beyond technologies and approaches and so that we really think about and how we execute them. One of the challenges I see going forward for physical oceanography, a couple with that is access to really difficult to get to environments. Two examples again taken from my latitudes. On the left, you see a figure that shows you the Arctic basin, right, Arctic Ocean and all the ice-tethered profiler measurements have been made from 2004 to 2016. The ice-tethered profilers have been termed Argo of the Arctic. It's a high-space instrument that extends down through the ice and profiles over the upper 800 meters or so. That looks like really good coverage until you start to look at what it means on a month-by-month basis, right? That's many years of profiles, but on the right hand side, you see a comparison between a month of profiling and the month happens to be January 1st 2013, but it could be any month, really, relative to the Arabian Sea and Argo profiling in the Arabian Sea, which is as far as we profile the area for Argo. When you look at the coverage there, it doesn't look so good for the Arctic, right? It's a few select measurements made in particular places where those instruments are drifting. We just don't have the ability to make those distributed measurements in these remote environments right now, but we're working towards it. Gliderons of floats used to sample underneath the dots and ice shawl for an annual cycle per piece are, again, ridiculously difficult places to sample despite their importance in beating down uncertainties and predictions of global sea level rise, right? This is the biggest term driving that uncertainty. To get there, we need to have a tolerance for high-risk investments, right? We need to be willing to take those risks and take those on the way there. Really kind of long-term investments in different technologies and approaches. So the other challenge I see for physical oceanography, right, when we believe in the past, had big questions like finding the missing ocean mixing and understanding what's happening. Today, I think a lot of those questions exist at the interfaces between disciplines and areas. And when I say that, I mean things like the role of physics, the modulating production and exploitive carbon or the interaction between atmosphere, sea ice and ocean or the atmosphere-ocean interactions that govern the monsoons. And I would just throw up here a couple of examples from something that Mary Jane Perry or Tercero in a bunch of us did a number of years ago, making autonomous measurements at the North Atlantic which wasn't a new problem, right? People have worked on that quite a good deal before we did it. It's one of the reasons why we chose it. But the coupling of physics in the biology, chemistry in the biology and the ability to be out there for a long time led us to some new insights. On the left, you see the idea that the slumping of horizontal gradients modulates the timing of the bloom because it modulates when stratification occurs in what just simple solar warming does. On the right, you see an actual quantification of seduction of POC along an eddy, eddy-driven seduction of POC which is where it was hard to observe and hard to get up but with a larger team we were able to do that. So I see the challenges, right, being able to communicate and work across disciplines being familiar with what other people are doing and adapting existing approaches and technologies to these tasks. But I see that as one of the real areas where we can make progress in physical oceanography. And lastly the thing I'd like to touch on is the need to really sustain technological and operational expertise. And here I don't mean PIs like those of us who are giving these talks today, but I mean are the technologists and engineers and software engineers, developers who sit as part of these teams of equal importance to the science piece, right? But who are a little more vulnerable to fluctuations in funding and other things that happen within the community. And what I've got out of here is just the glider development timeline because it's something I've been involved in for a very long time. Basically 20 years between the starter development and production use there's no Align in 2014 or so, right? So that's a very long arc you see. Right hand side you see a list of 45 people and that's probably not everybody who are involved in the development of the three production level gliders within the community today. So long development arc to produce instruments and then bring them into production. The arcs are not linear, right? A lot of the things that happen are slide projects or things that people think are going to be important down the line and so they work on it on the side over the course of several years and then it intersects with an area where the development benefits from it. There are these ideas that we should be commercializing things and sending this to industry but in reality our experience has been that the scale and the demand for the instruments that we're talking about really limit commercial viability and so it limits commercial uptake and commercial interest in producing these things. There are some exceptions and we can point to those but in general that's a tough thing to do. What isn't in debate I think is that in order to make progress here it's required really consistent long-term engagement by really highly skilled technologists, engineers software developers, right? That's what is giving us the progress and instrumentation and approaches, you know, ability to do these very involved intricate field programs over the past few decades. So the question I pose to the group is how do we nurture support and reward these individuals within our community? I think it's rather unstable at this time and a difficult thing to navigate and I'll leave it there. Thank you. Great. Thank you Craig. Thank you both, Luanne and Craig. I feel like I'm already seeing some perils here with what you're talking about. Also with the OOI session, right? There's a lot of talk about interactions with physics and biogeochemistry so physical oceanography is no longer practiced in isolation. You both have talked about challenges in the modeling how much big data we're producing how we have to do the processing on the server where the data sets also challenging environments like the Arctic, you know, we're really looking at all those places. Valuing work that we know is important that does not get valued by our current evaluation system. I think that's really important. I'd love to talk about that more. And then the workforce issues, the brain drain, the diversity, the technical staff. So thank you for those thoughts. Now we'll move on to Melanie. Melanie is here in person and I'm going to hand you this microphone so folks online can hear you well. Thank you. My name is Melanie Feuings. I'm an associate professor at Oregon State University and in this green box in the lower left are a couple of the main issues I would like to touch on. One has been touched on in all the previous presentations which is our need for long time series to understand our changing climate. And I want to talk about a solution that we've implemented in my lab but also other places are doing this historical data rescue. And then I want to talk about a problem which is what my postdoc advisor live Washburn said we have chronic data accumulation. We've talked a lot about how we don't have enough data to answer the questions we want to answer but we also have these workforce issues. We already have way more data than we are able to get what we could be getting out of it. So it's already more than we can analyze and I'm going to float an idea and I'll say what that is if you haven't already guessed. So to understand our changing climate we need long time series decades long I'm not controlling those slides. Thank you. So we want to identify long-term trends. We want to be able to calculate anomalies and know in a robust statistical sense what those anomalies are. Which means we need to have statistically robust climatologies which means we need decades of data and we need that to know what the typical conditions are and how those are changing in order to identify extreme events against those backgrounds. So for example marine heat waves including El Ninos thanks. Subarctic invasions are a thing that we see in the northern California current system when we look at unusual inputs from the north kind of the opposite of El Nino. Severe droughts and their impacts on the ocean and land and so on. And we've heard other examples in the previous presentations and I've already kind of made the second point we need robust climatologies to identify these anomalies. But we know that most geophysical spectra are red meaning that there's more and more energy as you go to longer time scales. So really no time series is ever long enough right but a community standard that NOAA and others have adopted is a convention that you should use 30 years of data for a climatology and we know how rare it is to have 30 years of data for subsurface ocean measurements and like Craig mentioned that's much greater than the length of a single typical project and so NSF is presently supporting collecting time series that approach this length through for example the long-term ecological research network and the ocean observatories initiative and other initiatives but we don't have to just wait to build up 30-year long time series at new sites we can rescue historical observations that help us fill in the past and I've listed three examples there that we're working on in my lab which I can talk more about if you want. Another one is TACA's research archiving digitizing historical tide gauge data back to the 1800s really valuable for understanding how things used to work around our coastlines and for our need to understand upcoming sea level rise and the effects of dredging on harbors and so on and so these historical oceanographic data sets including things like ADCP data collected on continental shelves or in the deep ocean by individual PIs that are retiring and passing away these are national treasures that are being lost. Next slide please so recommendations and so the first point I want to touch on something more about this issue of what is known in some communities as climate data records there are areas where Argo floats don't go and time scales that Argo floats don't sample right so if you want to know global ocean mix layer depth on daily or sub daily time resolution we would need so many Argo floats that we trip over them every time we went to sea and Argo floats don't go on continental shelves and continental slopes and the spatial scales of variability that are important there are much smaller in the open ocean so could we imagine something like a coastal Argo program and back to the historical data set rescue it's my understanding that USGS has or had an office for historical data rescue they actually had an employee whose job it was to fly around the country before COVID interviewing scientists who were retired digitizing their handwritten data sets archiving those at the USGS could the NSF support something like that for data that has been funded by NSF past and our danger of being lost because those projects have ended and also just in general with data archiving I think that if there were more support to help PIs archive the data and get that done in a consistent way it would be more likely to happen and this paradox that I already mentioned we have both not enough data and more data than we can handle I think that you know with all these data sets we've been hearing about with NSI and other efforts we could be getting a lot more scientific results from the data we already have if just as one idea there were more grad student and postdoc fellowships funded to work with those existing data and back to the LTOR idea long-term oceanographic research sites similar to the NSF's long-term ecological research sites how is this different from OOI and I don't know is this something that would be alongside OOI or would be different locations this is just the important thing that I wanted to bring up for discussion is these would be hypothesis driven and they would also involve funding for the analysis because the way the OOI works right now there has been purposefully a kind of a barrier built between the people who are collecting the data and the people who are allowed to hire grad students and postdocs you can't do it with the same money right and so I think that there's a barrier there that is slowing down the science because the people who are the most expert in collecting the data are fully occupied with collecting those data and not allowed to fund grad students and postdocs on the same grant if I understand things correctly so is there some kind of structural tweak that could be made by looking at how LTER's work that would enable us to get more science out of the OOI data faster this is me as a kind of semi-external observer because I'm at OSU but I don't work for the OOI and I'm not funded by it and on the ambassador's question I also think we need something more like data helpers I understand Ocean Networks Canada has something like this especially for the interdisciplinary work I work a lot with ecologists and fisheries people NOAA and so I work a lot in that space between physical oceanography and marine ecology biological oceanography and that it takes time and work and communication and explanation from both sides there are lots of people who want to be using OOI data but it's like a level up from the volume of data they've worked with before and so if there's someone who can help make a figure for a proposal and kind of get the effort launched that kind of thing is what I'm thinking of and I'll stop there thanks. Thank you Melanie so I want to move on to Jack Barth and Joe Schumacher who are online with us but thank you Melanie for that I think one thing you point out yet again is this whole mismatch between what we should be doing to move the science like archiving our data and what in a tenure track world were rewarded for which is not that right so we can keep thinking and talking about that okay Jack take it away Alright just Tuba just checking you can hear me alright yep. Okay good morning everybody I'm Jack Barth and I'm a oceanographer at Oregon State University and I'm just going to tell you about a particular journey working with Joe Schumacher and the Quinnell Indian Nation just off the coast here off the Washington coast as usual to do any of these things takes lots of people Craig touched on that the excellent technicians and folks that we really rely on and keep funded through our research lots of partners here at the bottom including the IUS integrated ocean observing system elements we talked about basic NOAA funding with sometimes in relation to the fisheries folks but this all really started out with an NSF single PI proposal about 20 years ago and we got involved with flying underwater gliders kind of a third of the way through that timeline that Craig Lee was showing we're going to talk a little bit about low oxygen and I'll just say right off the bat that we've kind of moved a lot past that into the multi stressors world so I'm going to be talking about oxygen but obviously the same water has low pH it's got lots of warming and of course harmful algal blooms so next slide please so we're I'm glad Luann pointed out that eastern boundary current issue they're really exciting places because of this both the closeness of the so-called oxygen minimum zones onto the continental shelf but then this very active upwelling driven by the winds which of course are in turn subject to climate change so the setting here is we're going to up well from the very top of that put the water up on the shelf the nutrients are going to cause things to bloom that material falls to the bottom and undergoes respiration so you've got this recipe for low oxygen well you know we really didn't see that much in the California current until about the last decade and a half it's really because the balance of the physics biology and chemistry is changing and we really didn't think that that low DO would come right to shore where you could almost hit a hit a baseball out into the waters that are low oxygen and the communities really care about this so let's go to the next slide so as Joe will explain more we've got a number of Native American tribes along the Pacific Northwest coasts their livelihoods and their culture are oriented around the ocean they're seeing changes they worked with us to think about how we might do a better job measuring those with the tools we had and communicating those so really the question is that they need to know where when how much these events are affecting their coastal waters next please so I don't need to dwell too much on this this is just the the slope of underwater glider it's equipped with as well as lots of bio optical and chemical sensors and I'll just throw something out here right away is what I've learned is that we don't really need ships to do CTDs anymore right we've got these gliders we can measure a 3d or a 4d volume and have ships doing much of the experimental work that we really need but the basic background view beneath the sea surface these guys can do it so next please so here's just a random section from almost 20 years ago when we were first getting started I've just colored along the track of the glider up down up down starting at Newport, Oregon on the right hand side and going out into the deep ocean with temperature salinity chlorophyll and life backscatter from top to bottom there's probably 300 profiles across this for comparison along the top side there's looks like about 8 or 10 dark ticks that were the traditional ship based measurements were made but all of a sudden you've got this amazing view under the ocean multi parameter you can see the linkages between the upwelling front and the subduction of plankton down along it you can see the difference between the live and the dead down in the life backscatter the living material in the top the settled material on the bottom that's undergoing respiration so let's go one more so what we've been doing up off washing with the quenalt is flying these gliders in patterns that map the low oxygen low oxygen zone I would having been a coastal oceanographer for 30 years now you'd think that we could say well on the sea floor here's what's happening and this is why but we really still don't have that ability to dive down get those measurements come back and report maps like that you know we should be farther along on that so this is just an example from the Nanus site the Northwest Association of Networked Ocean Observing Systems the Pacific Northwest Integrated Ocean Observing System and we just look at a line like that along the glider tracks so next please oh and I should say these are amazing things for doing community development for next generation we really ought to be doing more with that next okay so this is just a typical example comes out in near real time is available on the web and can guide decisions so on the right hand side is that low oxygen zone right near the bottom we can see it's nearly half the water column is pretty amazing this is in September at the end of the the upwelling season and then the trick is how long does it stay there how close does it get to shore but what I want to point out from a physical oceanography point of view is we get to see all those structures that my dear colleague John Allen another studied for 30 40 50 years and that is the coastal upwelling fronts the along shore propagation of signals you can even see the internal tide those waves at the bottom of the oxycline there that we see in that image so next one and this will kind of tee up Joe for some discussion about interacting with the with the tribes on the coast but as I said students love this technology right so showing some of the conalt youth about what we do in the ocean I mean it's really cool what we do and it's not just entry point for physical oceanography like myself there's the technicians and the machinists and the field folks that Craig talked about so just absolute fascination in this and we ought to be doing as much as we can to pull people into that next please so my takeaways are oceanography shapes the marine ecosystem we actually have the tools to look at these important ecosystem stressors and then of course every time we look closer we learn more the submersive scale we've known about in the coastal ocean but big studies now by NASA and NSF to look at that on the larger scales the boundary layers both top bottom and sides continue to be more important and we can do lots of these things from a single platform I'm actually doing active multi-frequency acoustics from gliders now so we can get some sense of up into the food web for fish and zooplankton I'll just pile on to the subservice data still lacking and we do need those data heroes and ambassadors to make the best use of that data and this example of co-design co-production is just teeing up for your session later in the day and then I'll hand off to Joe, thank you Excellent, thank you Jack and thank you for making those connections yourself so I wouldn't, I didn't, I don't no longer have to do that. Joe Schumacher are you online? I am online, thanks let me see if I can keep the presentation rocking there, hold on one moment and I'm just going to get the share screen really let's try this one see if this flies for us and then share are you seeing the presentation probably not aren't you, hey is there anything good in there? just a moment here hold on just one second, thank you well just as we start up here I just want to mention that you know you've got this panel of fantastic physical oceanographers and then you've got me a fishery scientist here who depends upon these folks and it's a it's the needs that I'm going to talk about here today so we're just going to go ahead and do this, hold on one moment thank you you ready there we go fun fun what happens when I have multiple screens? I know how that goes yeah by the way we're having fun here right now let's try this over here by the way Jack's pictures did take a few, I'm going to share a few of the same ones with you here whenever we get this shared appropriately oh here we go let's try it one more time we're going to see this one try to move this over to here if we can do this screen one more time just one moment folks tactical difficulties yep we're seeing your slide now thank you my goodness we got it to work first and foremost in the agenda you'll note it it has to be associated with NANUs and I have been for 20 plus years with this great team from the northwest Jack noted earlier the NANUs team is extraordinarily good for working with the tribes out here both in Puget Sound and out here on the other coast of Washington but in fact I'm employed by the Quinalt Indian Nation and have been for 24 years out here so we work on a remote environment that has not gotten a lot of attention over the years until recently so we really appreciate the fact that we have these folks here today that have helped us in this regard we live in an isolated coast and in particular this area here north of Grace Harbor on the northwest coast of Washington state is really really tough to get to and has a lot of resources but it has not had the attention paid to it for data gathering that has been needed especially for management of resources here in fact in the past all of those data collection came primarily from personal observations from tribal members that have lived here forever the ocean provides everything that the tribes need out here for them including the salmon it's the home for the salmon that return up the rivers that are so important to them but everything out here right now is in jeopardy and has been for some time and we've seen the changes and it's been noted by everybody that lives out along these coasts they're on the front line of climate change so our data needs are more pressing than ever and that's where folks like Melanie Jack Luann and Craig come in handy I mean we really absolutely need the data produced by oceanographers to help us better understanding the physical environment in this ocean that leads to the biological changes and ecological changes that are so important for us to know so that we can at least monitor for them and potentially mitigate against them so we're we're harvesters the quinoa nation we manage many many fisheries out here in the ocean quinoa is a known manager noted manager I should say that he was recognized in the federal court case that re-established their treaty rights out here the United States versus the state of Washington and in that court case the federal judge found that two tribes in the state of Washington had a history of management and that included the quinoa Indian nation and the Yakama nation and so we are co-managers and full regulators of our fisheries out here and you can see halibut, dungeness crab, sable fish, various rock fish and other demersal species on the shoreline the Pacific razor clams and down in the lower right corner the iconic blueback sockeye salmon that run up the quinoa river and these as you saw earlier from Jack's photos there of some of our events we had the first time that any of our quinoa members had ever noted a fish kill of this magnitude occurring and we went back to the elders within the tribe we went back to everybody that we could talk to up not just at quinoa but with our tribal members to the north as well from the Ho the Quileut and the McCon tribes and there were no stories no long oral history of large fish kills washing ashore on the beaches so this is an example of where as Jack noted earlier in the September we had a lighter mission off of the quinoa coast there you saw that it came in right towards the village of Tahola where I'm speaking to you from now that we had a low oxygen zone there about taking over about half of the water column at that point well this is an example where it takes over the whole water column and when this occurs we have dead fish and we've had this reoccur numerous times since this 2006 massive event that littered the beach for miles with dead fish of all species and types we have this hand in hand relationship with OA that frightens us deeply because of the future of the ocean out here we know what we know about the carbon storage in the ocean we know that upwelling is bringing this to our shores and we know that it's causing carbonate chemistry problems that are going to probably affect a lot of our resources so we need to know more and more about all of these parameters in order to better prepare for an uncertain future and to potentially mitigate for events as they occur the oceanographic data needs for Cornell are the same as much of this many of the things that have been discussed today but in particular long term data sets and I think we've heard that from a couple of folks here today we need to define these baselines and then the changes I think that it's very important that NASCM and others continue to realize the importance of traditional knowledge were available and appropriate and how they can incorporate this data into these western data sets that are available now in these spatially remote areas so we're looking for optimized or useful data off our coast out here that we can use to better manage including ocean temperature DOC level currents of waves, carbonate chemistry long term biological data including primary and secondary productivity and as Jack alluded to the determination of multi-stress on organisms out here which we're getting more and more concerned with for obvious reasons I'd like to put a plug in too for LTERs as well as LTORs as Melanie brought up but LTERs we have the one off the Newport line that's been in place for some time now off the central coast of Oregon there it's extraordinarily valuable but however between here and there we have a thing called the Columbia River and it's usefulness for the biological data in particular for our work on salmon up here could be augmented by something north of the Columbia River maybe off of Westport with that I'd just like to note the collaboration of new tools the NANUS collaboration has been extraordinarily useful for Quinalt and has led to many collaborative efforts as well as the great tools that are available on NANUS including here from the NBS this is an Aragonite saturation model off the live ocean model that shows it at the seafloor bottom out there I had it forecast to today actually 24th of October and there we go and you saw this earlier from Jack's work when we have Jack take the time as a collaborator with us to set up the missions that occur off our coast out here and to come all the way north to work with our students that is true collaboration that works not only for helping us with our data needs to help spawn that next generation of scientists that are needed to really get in place and continue the work out here off our coast I love this slide here this image on the right this is from an earlier version of the sea glider work some earlier collaborators from the center for coastal margin ocean prediction I just like the view point this is looking south at the top of that picture is the Columbia river then a little bit lower is the Willapa Bay and then Grays Harbor and so we're looking north to south there it just kind of shows that dramatic that coastal shelf and the dramatic drop down there and some glider drops in there and obviously some low oxygen zones in there in this hypoxia index image so this collaboration is essential to filling the gaps both in our knowledge and the coverage and that's just the way it works you really have to work with the folks that are out there in those locations to get the best possible information and to plan the best possible research that not only gets to the basic science needs gets to the user groups that are out here as well collaborative oceanographic sciences with non-academic entities it's not easy so I really want to thank all of our partners on the Washington coast for their support traditional knowledge should be a part of long-term data it's hard again to work on it we have strong recommendations for collaborative science with underserved communities including tribal governments and should be supported by the NAACM and others and collaboration always requires funding and resource support for all those partners involved so just a plug out there to make sure that everybody understands it we can't we want to participate but we can't do it without the potential support all of you need, thank you thank you Joe, thank you and stop sharing thank you very much for this perspective I love that you brought up traditional ecological knowledge and how that can really help move the science forward if it's put into use equal footing with the hypothesis based more western science I have a feeling we're going to talk about that more at the first session after lunch when we talk about co-design and co-development I think Joe, you and Jack really presented a case where what engaged science can look like to really benefit in immediate need so thank you for that we do have some time for questions and those of you who are online are also in the room if there's a particular question you see on Slido that you really like give it a thumbs up so it kind of moves up in the ranking here and Brad I see that your questions is up there in regards to the brain drain that Luanne talked about and Brad there's also an anonymous question that came in just before you about the brain drain maybe you can incorporate that point of view into your question as well and I think you're going to get a microphone from Eric so that the online folks can hear you really well yeah Brad that was me I didn't know how to put my name in so this is Brad Rick we'll join together and you can correct anything I get wrong so Rick's portion of this Luanne was to ask whether these people who had shifted had shifted into an activity that may still be focused on oceanography in some sense because you can go to the private sector and there's lots of potential kind of oceanographic aspects there but I was thinking of a kind of a different question as to you know on this imbalance what do we need to do to make change to this issue and how specific is this to physical oceanography I mean is it in other areas is it broadly an oceanography so for the first one the brain drain I think the reason that I'm concerned about it is that many of those were involved in the upgrade of the community or system model for the ocean component so it's been using an ocean model that's decades old and it was moving they were trying to move to mom's sex which is a newer model out of GFDL and many of those people who left are either people who develop memorizations or who were helping with the effort to modernize that model so it's more than just them leaving you know they're not many of them did go to CDR but we're losing their modeling knowledge at the institution that builds the community model in the US okay and then the second question can you remind me how I got so passionate about that second question has two parts in a sense is this specific to physical oceanography broadly but then more importantly I think the real issue maybe for all of us is how do you change this how do you correct gender imbalance because you pointed to that in a number of different areas and how do we engage or bring more people in or sustain those who are presently I think that part of I think what's happening is this sort of coding focus and machine learning that is sort of moving towards computer science which has tended to be less friendly to women and it's quite different on the theory modeling sort of computational side the gender distribution then it is in the observational side there are actually more women I mean I don't know what the but just sort of anecdotally in my experiences there are more women proportionally in the observational world than there are in the computational analysis world so there are some follow-ups to that I see let's go to Shannon here because my microphone is nearby and then Marcia can you touch no mine was more of a comment to add to that a little bit and Luanne's anecdotal evidence I think that has been proven through NSF studies that women are less likely to be found in physical oceanography or marine technology fields and to that point NSF does fund a program for retaining women in physical oceanography it's called the empower program and I think that's been pretty successful in keeping women in the field of physical oceanography but I just wanted to point out that has been a issue that has been recognized by NSF I've been involved in empower since its inception Marcia Luanne can you elaborate on why you think AI, ML and data science is unfriendly to women that might not be what you intended to say that it's unfriendly but why women aren't fitting into that field maybe yeah I don't I don't know if I have a good answer to it it's mostly an observation and if you look in at least at University of Washington computer science department it's heavily dominated by men so I'm not really sure why there's probably papers on this out there that I could explore but I'm not an expert in that it's more of an observation that I've seen the evolution over time of what physical oceanography has looked like I mean it was when I first started there were very few women physical oceanographers and then there were more and more and now it seems to be changing very anecdotal I realize but something to watch thank you Luanne clearly this is a topic that resonates with this group and maybe Melanie wants to add to that and then I want to move on to a different topic okay thanks I I just wanted to quickly say there's the article a recent article by Ranganathan at all that looks at different geosciences disciplines and to my shock as a physical oceanographer physical oceanography actually has the lowest proportion of women out of any geoscience disciplines second I totally agree the empower has helped a lot to retain women who are in physical oceanography already in fact Luanne has been my mentor for 15 years we have talked on the phone every month for 15 years where m power peer group one still going strong and the third thing is we've been talking a lot about trying to increase the numbers of women which is great but I also want to mention that there are other underrepresented groups who are so underrepresented that we don't even have statistics to talk about and so we could maybe broaden the conversation of it there but I'm glad that you brought this up Luanne. Very good points thank you Melanie so I want to move us on to a little bit of a different discussion Lahini I see that you have a question here about LTRs and LTRs do you want to go ahead and ask that. Well it was I think two part question or two part observation I was thinking as you were especially Melanie talking about long-term observations like multi-decadal scales we think about marine heat waves or decadal scale changes in the Pacific that are natural that you know I think there are more and more data sets geochemical data sets coming out now whether it's from corals down in Riviera hey hey does or sediment data sets that are high resolution enough what about interfacing with some of those data sets to really try to you know and we've done this for you know for time immemorial I guess in this discipline but I think these higher resolution data sets and also you know now with the geochemical advances where you can make these geochemical measurements on smaller and smaller sample sizes which means you can increase the temporal resolution I think you know thinking about how to combine those is valuable Melanie do you want to address that oh thank you yes I totally agree I guess my immediate thought is that it involves working often working across disciplinary boundaries so I'd like to connect it back to the issue Tuba has raised about and others about the way that our evaluation comes work right now and that you know we there's still a lot of they're not even to the point of being only vestiges right of the lone hero scientists being the most valued and and so I think in order to join together these observing efforts on different time scales and in different fields we really need to learn better how to do that kind of team work the interdisciplinary work it just takes longer as you well I'm sure no is you always having to explain what do your words mean that don't mean the same thing in the other field and vice versa and it's it's you know it's hard to know how to convey or to assess that effort yes thank you for that it seems to be coming up in the slide questions maybe we have just enough time to cover that or by going a few minutes late if that's okay Jim you have a question about commercialization let's go there and see what Craig or Jack or others think Craig just following up on your point which I think is really valid something that I've been sort of beating my head against for a while now curious to hear if you have any thoughts on how alternate modes of making new instruments available within the community work would work you think or because the idea that industry will pick up something and there'll be a market for it really doesn't seem to be there yeah thanks we're actually struggling through a part of that right now your our experience just give a really quick summary of our experience the Seabire right this designed and built at the University of Washington provided to the community through a service center University of Washington our fabrication center for a number of years commercialized through a license originally to our robot make the rumble vacuum cleaner military robots that lasted a while they let go of the license when they failed to obtain a large Navy contract Kongsburg pick that license up again with these aspirations for a big Navy contract they sold it to pundit and Ingalls maker of minutes class aircraft carriers you can imagine how that went and we've recently taken it back and started a in-house service center with a slightly different bed on how to provide service and support to the community and you know a part of this is that you know as you alluded to the the commercial market is small relative to the aspirations of most companies that would like to pick it up our relentless focus on growth is well suited to these kinds of things right it would be good living for a few people you know over a sustained period of time but it's not a good living for continue to grow a number of people with aspirations to fail ever more every year you know the models we think about are trying to empower the community to support itself by providing support and training we talk about doing open source kinds of things for these instruments seeking small ways to commercialize if you will you know not big aspirational companies but but smaller groups yeah it's a tough problem I don't really know where to go from there thank you thank you for that perspective Craig any other thoughts from the panel about that topic Jack well Craig's absolutely right about these these high end complicated robots but there's another end of the of the challenge and that is getting many sensors out for low cost into the hands of folks and we've been making a fair bit of progress and with robust inexpensive dissolved oxygen sensors of course we've been able to do temperature forever things are coming along in the biogeochemistry lots of potential there to get you know order thousand dollar instruments out there by the thousands rather than the you know the few two hundred thousand dollar instruments so I'm pretty hopeful about that end of things great perspective thank you Jack this has been a really informative panel you all I really appreciate everybody's time Luann Craig Melanie Jack and Joe thank you very much start a session on artificial intelligence and machine learning and the future of ocean science and we have three people I think are online Patrick Einbach I'm a classic on some having trouble reading Patrick persona and Kana is everyone on all three online I think you see this is Kana here I think the list has Patrick going first is that correct is that how we're going to do it Patrick are you going to go ready to go first yes that's correct hi everyone let me share my screen here okay so you're not seeing right now you're seeing the presenter mode is that right you're not seeing the now we see it the full screen well thank you very much for giving me the opportunity to talk about this subject there was actually even a bit broader the name digital twins also appeared in the request and so what I'm trying to do here in just ten minutes to cover some aspects of weaving together this concept of digital twins and then an ocean of how we use a scientific machine learning in ocean sciences and first the question here is what actually are digital twins and this brings to mind the some of you've seen this the parable of the blind man and the elephants depending on where and from how you look at this you might get quite some different answers let me start here with the perspective of this UN ocean decades effort called Ditto the digital twins of the ocean and the quote here from Martin Visbeck ditto twins of the ocean will bring together ocean data models and digital information with those who are planning and regulating human ocean interactions they will come in dispensable elements of sustainable development of the ocean space so you see how a lot of different building blocks actually weave together so it's not just one component of it but it's basically how do we have a workflow or building rocks that actually seamlessly go between the ocean observing system that is indispensable and that's still too sparse at the present day a data space on where we can seamlessly share and discover and use information then the data analytics sort of modern and evolving tools for example machine learning on how we actually rapidly and comprehensively analyze the data prediction engines which we can think of the physics models that we are that we've actually talked about today but also surrogate models or emulators then all the way to visualizations tools and the decision making tools that allow users to actually really take advantage of all of the different data spaces and so if you want to know more about this the digital website actually provides some information but I wanted to switch gear a little bit and actually take a slightly sort of somewhat complimentary perspective from an engineering perspective because doing so will actually bring us a little bit towards what is actually under the hood needed to realize sort of the concept of these digital twins and all of their different renditions and so from engineering perspective where the concept of digital twin actually started an aerospace engineering up to 50 years ago so the idea is that the digital twin is in silico replica of a system that has a number of requirements. So first it must continuously improve as it integrates new data and then provides a dynamical digital history of the assets that we're studying it must be able to issue predictions so it must be it will go beyond sort of what it's basically the scene data into unseen conditions it must do so reliably and ideally we wanted to be able to do that with quantified uncertainty we want to be able to use such little twins to support assessments of what if scenarios like we can think of these scenarios like of course that we're used to in the climate modeling community but the scenarios there are actually still the way how we can do them and how we can afford to doing them computationally are quite limited we want to be much broader again quantifying uncertainties and then we want to incorporate sort of a synergistic two-way coupling between the physical system the data collection and then the user and the social system at the other end and again here sort of what the building blocks here what they look like on the one hand we have of course high fidelity models the physical models we have advanced systems for doing data assimilation on how we integrate and merge the data with these observations advanced tools of reduced order modeling so server modeling machine learning tools that we can either use to vastly accelerate simulations or calibrate those models much better we need advanced tools of uncertainty quantification which would present major challenges and again the question of the data cyber infrastructure that provides us with the means of analyzing this vast output that can go into the petabytes for example from the C-MIP-6 suit or some very very high resolution model simulations at one kilometer resolution also take up petabytes and soon will basically also go into the petabyte range with satellite data such as the surface water ocean topography mission that NASA has just launched a year ago and so trying to move here from how can machine learning support digital twins machine learning again so it has a number of ways on how we can what machine learning actually can do for us and it's for example classification anomaly detection or it is regression which can be in the form of parameter calibration or state estimation we can think of space or time dependent state prediction we can think of autonomous systems and active sampling techniques and how we can optimize our sampling strategy given sort of the very expensive and harsh environment that often times we operate in the ocean and then emulation for example for and certainly quantification large really number of topics that we could talk about and what I was going to focus here is on point point two on the regression and model calibration and when Thompson earlier today showed how climate models still are endowed with sort of major model biases that's true for the atmosphere it's also as Luan showed for the ocean and here can we leverage machine learning tools for example to improve such model parameter parameterization calibrations this is from a review article by Laura Zana and Bolton in 2021 where they show on how we can use equation discovery or surrogate modeling to calibrate ocean models better sub grid scale parameterization, eddy parameterization in ocean models using simulated data for example from large eddy simulation to actually be able to still run course resolution models which we will not get away with for the foreseeable future even 10 kilometer resolution in some ways this course resolution if we think at the high latitudes we're still barely resolving the barricade across the rays of deformation and so we're not able to get away from the need to parameterize these models and machine learning may give us a way to to improve such parameterizations we can go much broader and actually think about how we can seamlessly integrate scientific machine learning so what's just been shown on the previous slide but also inverse method approaches that have been around for a while to also incorporate not just big data that we get from very high resolution model simulations like but also from the sparse observational and eclectic global ocean observing system that we have to do something that's called a posterior or full model or online end-to-end learning of the models and so where we actually use state of the model itself that will calibrate not just the parameterization schemes on their own and sort of the recent sort of buzzword in this context has been it's called the differentiable programming so where we're basically bringing together networks that follow this concept of differentiable programming again but then extended to the full model that we have so for example to the physical model biogeochemical model and other types of models that we wish to calibrate and confront to the data and learn from these data through the lens of those models and to do so it's really and this is really just one example and I'm wanting to use this to show how in order to make this happen make this work really requires the intense and long-term collaboration between the computational scientists, computer scientists and then the domain scientists to really leverage some of the emerging tools hardware architecture and that we have in high performance computing that enable us to run sort of GPU enabled ocean models but then also computational tools such as differentiable programming can actually use this model and seamlessly combine machine learning tools with these PDE constraint models to actually achieve something like end-to-end model learning and here this is an example from a Julia approach that basically is being applied to Julia based ocean model that's been recently run at very, very high resolution on a large number of GPUs. Finally I want to go very briefly talk about the issue of making the scientific data more usable and so that we can really fully explore the vast amounts of data that we actually have and here the point here courtesy Ryan Abernathy that he has been making is that if we talk about the science community which is the right column and the business community which is the left column, they're really disparate tools that have been used in many of the levels. So the data makes the warehouses, the storage formats, the analysis APIs, they've basically been a complete disconnect between those two communities which is unfortunate. I think Ryan thinks that there is huge gains to be made by actually really bringing some of these capabilities in the private sector where there's huge amounts of investments have gone into developing these tools bring this into the scientific community and basically merging this and think about sort of a scientific data commons that serve on the one inside the data providers that basically produce data either simulated data or observational data then data enriches which are tech companies that work in the data science, data analytics space and then all the way to the data consumers which could be increasingly the insurance companies, financial services, all kinds of ocean services for the blue economy all the way to the defense. And so like to finish here with just basically pointing sort of this brings the notion of digital twins which really tries to integrate a number of these workflow into this seamless interaction between end to end sort of data production and data data use. There is a current National Academy studies going on on the foundational research gaps in the direction for digital twins and there which has also had a session on its uses for weather and climate simulation and certainly the ocean is an integral part of this and there's a lot to be done here and the notion here of really interdisciplinary work here with the computational and computer science I think is indispensable to really make progress in that space. Thank you. So we'll now hear from Prasanna. Thank you. You see the correct wheel, right? Yes, we can see the screen. Yeah, thanks everyone and I'm Prasanna Satyari, a principal research scientist studying research and I'll be talking about trustworthy AI and its role in AI as to decision making. Yeah, I think we I think I don't have to say too much about the possibility that AI can bring to different environmental science and ocean sciences. I think the previous presentation touched upon a little bit on this just to make the point I think we can use AI in a lot of different applications and it holds a lot of promise for example we did some work around super resolution so going from low spatial grid to more high resolution enhancement using different machine learning tools. That's one possibility and we can also think of other applications where we are doing automatic classification and there are a lot of applications which are boiled down to classification and so I think my focus on my focus in the stock would be like how and when can we trust such automated systems and what are the risks associated with such systems and these are some examples they are from different domains but you can see that these models when we train some limited amount of data they can have some sort of shortcuts which could be harmful when you deploy it in real applications so if you take for example these test x-rays we see that they learn a shortcut based on some artifacts that are there and the problem with such shortcuts is that when you go slightly away from the domain that you have trained your data it kind of fails because it's relying on these shortcuts and it's making predictions and similarly it can take some feature which is not at all important and then latch on to it and make a prediction and again when these correlations which are there in the data which are not representative of the real sort of mechanisms of how we come upon some outcomes it can have like bad outcomes right so I think these are some of the things that we need to sort of think about and to sort of more build a framework around these risk we can think about is the model fair and fairness could be so different many different dimensions right it could be based on geographical regions it could be based on a lot of different aspects right so essential aspect there is is it robust to some sort of distribution shifts or when the population underlying population changes right so and why is that important right and I think the reason I'm bringing up these risk is if we identify the risk then we can enable this AI as a decision maker in a much more powerful manner right so the key thing is we use these automated systems and the model makes a prediction and we sort of want to identify if this prediction is correct or incorrect right and so what are the things we can do to enable an expert to look at models prediction and make a judgment do I trust this prediction or not and we like to think of this decision making framework using the assistance of AI in sort of two ways right so one is where the AI is just communicating its predictions to the expert and maybe they communicate slightly more things and the other more richer way of communication is where the AI and the expert are coordinating with each other they're communicating with each other and then you come up with a decision which is taking into account their strengths and weaknesses together right so so one way to sort of enable this AI human collaborator system is through uncertainty right which the previous presentation was a touch upon so if the AI system gives us a prediction saying it's let's say in this case right so if there is a contamination or not right so it's not enough for an expert to act on it it should provide some meaningful uncertainty associated with that right and let's say it gives you some confidence scores it says I have 55 confidence in one contaminated class and then I have only 0.45 in the other class then the expert can obviously determine that this is not a very confident prediction I should probably not accept the model prediction and take some other corrective measures so this paradigm has been studied under selective predictions in the literature right and I think the key thing here is uncertainty quantification right and there are a lot of different algorithms in the literature that we can get uncertainty scores we like to think in terms of intrinsic or extrinsic so intrinsic is where the model that we are using for prediction itself is able to capture different sources of uncertainty and the common source of uncertainty are either data or model right so data is something like for example if I have noisy input and so on whereas model uncertainty is something that's because of the gaps in the models knowledge so again going back to the previous example I trained my model on only some subpopulation and there is gaps in the models knowledge because it has never seen data from other subpopulation right but in this era of foundation models right and we are seeing more and more domains where foundation models are being applied where we train big models on lots of different data and then we use it for different downstream applications it becomes quite important to use extrinsic methods as well where what we call extrinsic methods are essentially things like recalibration or conformal prediction or other methods which are essentially taking the model predictions and if it comes with uncertainty estimates recalibrate them so that it has some guarantees or in some cases if you are using point estimates which the model doesn't provide any uncertainty we sort of add a layer of uncertainty proper on top of that right so so one sort of line of work along this direction is what we call as models so you can think of this as taking a point estimate so when we mean point estimate I have a model it just gives me a prediction right so it just give me a label let's say it's safe or unsafe and then we can take some held out data and then get the models predictions on those data and go ahead and collect the actual ground truth right so now we have this validation set where I know the ground truth values I can construct a data set where the label now is if the model succeed or not so with that data set now I can train a simple model whose job is to predict model success or failure and we can use things like linear classifiers or logistic regression and so on and now we get this probability estimates where we can interpret those probabilities as the model success or failure rate and we have some extension around it but the key thing here is that we need such techniques when we work with much larger models where we cannot apply standard techniques like patient methods or ensembles and so on so other key thing to note here is how do we evaluate the uncertainty measures right so we typically look at calibration which usually is an aggregate measure so we look at the whole population and then we measure calibration which essentially means that if the model is predicting confidence P on certain instances does the accuracy of the model is also around P right so that's the expectation when we say a model is well calibrated but if you slice these calibration metrics by different subpopulations in the data you might see that the model is not that well calibrated right so I'm showing this plot here so what's happening here is I'm using this model in a selective prediction framework so I'm increasing my threshold on confidence on this X axis and the expectation is as I accept more and more highly confident instances as in my threshold on the confidence is increasing my error on the accepted instances should go down right that's expectation and we can see that that is happening on an aggregate here but if I look at this curve when I slice them by different groups I can see that for one of the groups the opposite behavior is happening as I increase my confidence threshold that is as I look at instances where the model is really confident it is making more mistakes I mean this could be a sign of underrepresentation right so maybe from this pocket of data I don't have enough instances so it gives us some indication that probably I need to add more data from these points and so on so it's really important to look at all the subpopulations that we care about to keep these metrics along these dimensions so far we looked at one way communication right so the model makes a prediction it gives additional information apart from the model prediction then the human makes a decision and often we are seeing in some cases with this larger and larger models powerful models the models can have complementary strength to a particular human expert so it's also important and we are investigating how we can learn joint human AI systems such that we can take into account the human strength right so the idea is when we train such a system we will have a rejector which can take any instance and determine should I give this to an automated classifier or should I send it to humans right and the goal here is also to reduce the workload of the human right so there's a balance between how much we send to humans and what is our error tolerance and I think so far we've relied on uncertainty right but I think one of our research direction is to look at much richer forms of communications so can we use different explainability methods to sort of facilitate this human AI collaboration right so what sort of explanations will be useful and we have studied a lot of different methods and we think like explainability method should be mapped to the right persona right user and the right model and beyond this I think one of the really interesting directions that I think is how do we use these explainability methods to onboard users right such that they can understand the behavior they can catch errors and so on and each human might have different strengths so can we use these explanations such that the human can effectively use the AI system so let me stop here by saying that we doing quite a bit of work around different dimensions I touched up on uncertainty and explainability but there are a lot of other risks as well right so which may or may not apply to things like adversary robustness, fairness and looking at from causal lens and so on so the things that are listed here are all some demos that are powered by the open source tool kits that we have released in the last few years let me stop here. Thank you very much and next speaker is Kana Thank you for having me I am going to take a slightly different tack and even though my name says Zoe Alexander I am not Zoe Alexander and I also want to show you a different way of thinking or part of what we have been calling AI this is just a gratuitous way to tell you that we've been operating embedded systems which actually make decisions on board AUVs for quite some time so that's just a way to sort of set this tone and a lot of the work that we have been doing as a way to understand how to sample in the ocean is driven by what Waldemann said some time some time ago memorably and what we are really hoping to do is that we can use intelligent robotic sampling methods coupled with decision making the decision making part is critical what you've heard from Prasanna and Patrick is essentially the notion of using data we are not talking about data here we are not talking about decisions and the notion of cognitive decision making in ways that can facilitate sampling the big ocean and in particular the idea is to use cost effective methods leveraging advances in artificial intelligence and the artificial intelligence that I have been trained on is the classical forms of methods of reasoning about the world reasoning about how hardware should perform but driven by the notion that you sense the environment you assimilate the data you have a model of some sort there are different kinds of model that is an overloaded expression you plan or you project what you want to do and then you actually act in the real world or in your synthetic world and so forth so this cyclical pattern is really at the heart of a lot of the work that I have done and I want to convince you that AI is not just data in fact AI is much more than data and AI has a lot more to offer than just methods and statistics or neural networks and the reason is that there is a whole host of issues associated with being at sea dealing with the uncertainty the fact that there is a huge amount of cognitive overload models of synthetic ocean models the skill levels is still something that we desire a lot more and the big issue for people like me has been how do we maintain the sustained presence in the ocean to observe across space and time at large spatial and temporal scales so a lot of that has to deal with the fundamental notion in ocean science of sampling where when you should give an example given finite time energy resources in the uncertainty of what you are dealing with so the flavor of this talk is more about hey look there is a lot more to AI let me just give you a few examples and I am just not going to go into I am impressed with Prasanna's set of equations there are no equations here this is simply to give you an insight into the classical forms artificial intelligence is really all about a lot of my work is driven by having flown the first AI system in space and this is again not data it is decision making this was 1999 in deep space 65 million miles away followed by command and control of the two Mars rovers in the 2003 mission and then on moving to Embarie where I was a PI on building something called the teleoreactive executive which is an organization making system the AUVs could essentially follow signals not straight lines in the ocean and collect samples in this case using statistical methods and machine learning and where this is going is really what I wanted to show you about so this is just a quick slide to say that there is an entire cognitive architecture this is the notion that we want to build machines in software that leverage existing hardware to make decisions on board like human beings so we deal with uncertainty we deal with latency we deal with faults irregular phenomena we as human beings we want our software to be able to do this we want their software to be embedded I interrupt just for a second your screen is not sharing right now we see you but not the screen okay were you not able to see my screen at all or no I'm sorry let me try this again okay hang on just a second let me just see if okay is that a mac okay hang on just a second because my mac is giving me all kinds of messages so I'm sorry if you're not able to see the screen and let me see if I can try to fix that right now okay yeah yeah okay okay okay I will say that you've been I mean it's been really easy to follow even without slides you're thinking well maybe that's the way to do it let's see maybe I should send you the slides because I don't know this is not working hang on just a second let me try again you want to send it to Zoe or Leanne do you have a mac or do you have a I think we are all PC is what I'm saying okay hang on just a second it's a but even if we can't get it up on the screen we can certainly share with the committee later and I for one have not found it hard to follow your thinking even though without slides that's good to know let's see I think it's a this is cacony if you want to just send me your file I can I can I have a mac I can probably play it you have keynote okay keynote the thing is that this is pretty large it's 70 meg so I hadn't prepared for sending it by email but let me just put in a dropbox and see if you can try to get it and okay so how about if I keep doing the slide thing and then you can see the slides later on in the interest of time how's that yes yeah that sounds good because everyone was selling well alright so alright so I wanted to sort of give you the trajectory of where things where I've been coming from but it's about decision making on on on machines hardware but the decision making actually happens in software so this is all about software not hardware so this is not the notion of building machines per se it's building machines where the core component is manipulating the hardware in ways that you can and what I was showing you or I thought I was showing you was an architecture for doing it which is built in body but it's moved on in other ways to sample for instance one application was looking at the subsurface chlorophyll max and building doing field reconstruction which is essentially trying to understand beneath the upper surface where is the subsurface chlorophyll max by sampling with sensors in situ not in straight lines but following trajectories driven by some form of statistical machine learning and the picture in front of me is essentially showing you this 3D reconstruction based on the measurements that we took in an AUV in a Norwegian Frufjord with my collaborators in Norway when I was a faculty there this was published in Science Robotics some time ago I don't even remember when the next slide that I'm sort of putting up again is an Embarie perspective which is having a way to understand the spatiotemporal aspect of what the water column was looking like as you have a Lagrangian drifter moving in the water column essentially trying to have an AUV chase that Lagrangian drifter let me send the my apologies with all of this but let me just send Kukane, let me just send you the dropbox link so the idea here is I'm on slide 8 just so you know the idea here was to understand with the drifter I think about 14 meters below I had the ESP that old version of the ESP which I'm sure ESP is the environmental sensing processor which many of you are familiar with was essentially taking samples and trying to understand what is in the water column what's the community structure what are the kinds of critters that they are able to see while the idea of the AUV was to surround it as it's moving in a Lagrangian manner to be able to see it in space and time as to what's the environment and you can't script these so you are having the AUV make decisions in situ no human being in the loop it based on the fact that we are tracking a target and the target is moving in some Lagrangian manner so all of this again is to sort of give you the notion of decision making by specific examples rather than having something very abstract the next slide I don't know Kukanya are you able to get the slides I'm working on it I think you shared the keynote and I have to turn it into a zip okay the next slide is essentially an application that was that involves again robotics and AI and the AI again is machine learning which is to track sunfish mola mola in space and time so these are all very diverse applications where decisions have to be taken by machines and what I have is on this slide is an animation of what we actually did at sea in southern Portugal and this involves aerial surface and underwater platforms including a wave glider which is acting like a communication hotspot and this has led to something that my old boss Marsha McNutt who many of you know she had this notion that we need to think about having machines become like sniffer dogs tell the machine at a very abstract level what taxa you're looking at and the machine has to go off and swim and figure it out this was in a project that I did when I was at in Norway and we built an AUV which does in situ imaging image processing machine learning supervised machine learning using some data from Heidi Sosek and also from other sources to understand the different notions of different kinds of plankton and then followed by hydrodynamic a small portable hydrodynamic model which would measure current structure and understand which way the currents are moving coupled with the decision making system that I brought to the table that I showed you that I was talking about earlier I'm sorry I can't show it to you so this is also published in oceanography 2020 as I'm seeing the notion here is the scientist is sitting on his or her desk remotely anywhere and basically somebody is deployed to say AUV and the machine is going off and looking for things in some pattern where all of this is going is where I would like to and hope to sort of connect the dots between what Patrick mentioned and what Prasanna was talking about has to put together a cohort or an ensemble of assets physical assets but driven by synthetic ocean models which can assimilate you sense with the vehicles and platforms in the ocean you quality control and then you assimilate and you predict the prediction in this case what happens on shore with synthetic ocean models of a chunk of the ocean a mesoscale and then that prediction drag with some sense of uncertainty or entropy which is reflective of what the model does not know drives where the marine robots can actually be targeted to go and sample. So think of this like a massive machine learning experiment in the sense that your end goal is to have what Patrick talked about as a digital twin to refine the digital twin over time and space what you need is a systematic approach that integrates ensemble of vehicles different methods of control of the vehicles data assimilated from multiple sources including remote sensing and having a model that over time essentially you turn the crank the model makes a prediction which is invariably not highly skilled but that whole point is to reduce the error increase the scale of the model by having this big huge system in place and I don't know where we are are you just tell me what slide you want okay slide 12 let's go to slide 13 I'm sort of wrapping up at the stage okay I think that's next yeah so slide 13 is essentially that's the next one I think or the one of two after no yeah I don't see the slide number so you just have to tell me two more okay so this is what I was talking about what we call as meteor which is the sense of being able to assemble assimilate and predict and where we think this should be going as a way in which you're again you're looking at hardware but the real focus is the software and it's the software that's actually driving decision making so that you sample and you sample smartly and you sample in ways that actually make sense so final slide if you don't mind so the ocean doesn't have straight lines so why the hell are we sampling along those and so adaptation methods in control with coupled with statistical methods in sampling are really important we should be paying attention to it sensing and robotic methods and hardware have improved yet we're still doing the same approach to sampling with ships and so forth I think we need to start thinking beyond ships I'm not saying you remove the ships but to augment ship based sampling and to be able to explore more effectively across space and time with robots decision making is important because you really want to be able to chase to understand processes in fine scale across the meso scale you want not just one ship you want multiple robots how to do that some of that technique is available and we've been using it for the last 15 years certainly in Portugal and we really want to be able to increase investment not just in data science or analytics or machine learning but also to provide new tools in hardware and software a lot of it is about software it's not buying new toys out for us to to put on the table and I think we should have NSF have incentives for computer scientists to actually get them on ships because they don't understand the average I am in the computer science community I can say close to about maybe 1% are doing anything that's field driven and 0% in terms of actually going out to see if you don't know how to go to see and don't understand what what scientists are actually working on you have 0 incentives 0 ideas 0 notion of what it takes and and that unfortunately is is where we are in computer science and specifically in artificial intelligence more isn't necessarily better smarter is decision theoretic methods like the ones that I was hoping to show you at least in slices is is what I believe we should be doing any eyes not ML ML has a place it's very important no doubt about it but it's coupled with how you sense and sample the oceans and so I think we need to have that broader perspective and this is just a small little vignette the set of slides in this presentation that I wanted to bring across so I'll stop there thank you for your attention thank you very much and we've got to move along now we have actually one more speaker in this session which is cacani and so if you could go ahead and start cacani with your talk or thank you cacani I tried alright so first off thank you for the opportunity to talk I know I didn't make it to the agenda but I'm excited to be a part of this group and also have a discussion and then you know the I think the other speakers really highlighted a number of things and I'm going to come back to them as I present some of these slides and I'm also going to the focus really heavily on like the state of biological observations in the ocean because I think it's pretty clear that's a that's a gappy area in ocean science and also delve more deeply into imaging specifically and and the methods or approaches that I'm going to talk about can equally apply to the fields of EDNA or acoustics and so maybe keep that in mind as I present some of these ideas or these slides and a lot of this is collaborative work work that we've done with a number of individuals and thank you to the NSF Convergence Accelerator for supporting some of this work and I don't think I need to spend too much time convincing members of the audience that biological observations at least the state of them in the ocean is it needs to be better this is a paper that was published by Aaron Satterthway a number of others in 2021 that showed at least the spatial extent near the ocean surface of our long-term biological observations and so these are data that are pulled from OBIS and GBIF repositories and what this particular paper found was that about 7% of the ocean surface area is covered by long-term biological observations and then of course I think this audience knows the ocean is a vast majority of the and there's also additional challenges right we know that at least estimates that NOAA has provided about less than 20% of this region full region has been explored and we expect to see that those numbers drop pretty significantly when we're talking about visual observations either by direct human eyes or by image imaging sensors and so if we want to fill observational gaps in this sense and Con I think you made a really important point you know it's not just data that we're interested in processing but we're also trying to write processes information to provide decision making or at least the necessary information to make really good decisions and that could be from the robotics perspective management perspective etc but the point is at least you know of the biological observational tools available to us imaging appears to be one of the most direct methods and you know ways in which that data are collected it's just so many different options and opportunities the video I'm showing you right now is visual data collected from remotely operated vehicle platforms you can see explosions of different types of deployment techniques using drop cameras, AUVs etc and so what that means is you know when you're collecting this visual data you have to get meaningful information out of it right convert pixel information into actual data points that can be used to you know inform decision making be it on a robot or with other individuals and so the main action or activity individuals do when they look at this kind of visual data is right they do categorization and that would be categorization of the animals, categorization of the substrate categorization of also you know the physical and chemical states of the particular region as well but that task of categorization is really really important but it's also really really challenging so for example if you collect about an hour of visual data it could take a visual taxonomy expert anywhere from three to five hours to process that information never mind that you know when I talk about visual taxonomy experts they tend to specialize in particular you know animal groups or substrate types and so as you can see with this visual data there's a lot of information but there's not that many people out there who can convert all of this information right into something meaningful and useful and so there's no surprise here but I think something that we've learned as a team or as a group is kind of the astonishing state of data already been collected in the community and I'm going to go back and see if that statistic can be shown but what we know when we queried US based exploration institutions and what we found is that more than 300,000 hours of video has been collected and less than 15% of that visual data has actually been processed and if we want to fill these observational gaps right in front of us and we're talking about biological observations we would anticipate or expect the use of imaging to increase the use of vehicles to increase and take your pick what kind of vehicles but I imagine autonomy is going to be extremely important to help us fill those gaps so it's not a surprise and it's not independent of visual data this is true of all biological observation data we are just facing a deluge also really focusing in on visual categorization because there's an entire field in computer vision called fine grained visual categorization where one might be able to take an image and convert that image into a particular category either down to a dog or cat or a stop sign but what we at least in the imaging community is trying to do is to sometimes get categorization down to a genus or species level because that is at least scientifically what a lot of researchers are looking for but there's some very, very real challenges for accomplishing that and I'm only highlighting three but the reason why I also want to highlight these is these aren't just challenges specific to the ocean science community but these are also challenges that the computer vision community is actively working on from a research perspective and so the first is distribution shifts and I think for Sana talked about this a little bit it's also known as domain shifts so the idea that when you train a model and you train a model on some domain data that could be data let's say collected on the benthose and you want to apply that same model to data collected in the oceans midwaters what that is represents a distribution shift and the requirements of your model will most likely deteriorate and so and this is true for different locations different imaging systems, different days and so we see this in our deployments of these neural networks to try and solve this visual categorization problem as distribution shifts is not a solved problem for the community. The other thing is the visual appearance of complex structures we know that at least life in the ocean looks extremely different to life in terrestrial spaces gelatinous materials mucus materials really big changes in their visual character and poses and so that's really difficult to manage as well and then also the other thing is biodiversity surveys often require classification to species or genus level and so that's really difficult for a number of reasons particularly since right in the ocean space researchers are estimating anywhere from 30 to 60% of life has yet to be described or unknown to science and so being able to identify a novel class or novel type when an AI system views it is really incredibly important and so these are not easy challenges these are things that I know the computer vision community is working on and things that we need to consider or take into keep in mind we're trying to build systems that work in the ocean the other thing I wanted to add to is that thanks to the NSF funding that we've received we've done a lot of user centered design interviews and practices to understand not just what these challenges are for using AI but also the challenges for the community for implementing it I think Patrick showed a really awesome slide that showed these different types of tools maybe in the research space versus in the consumer business space and those types of slides very useful can overwhelm a user who just really needs to implement an approach and so one of the things that at least our group has started to look more closely at is trying to figure out ways to create data pipelines but also keep in mind that humans are one of the most important pieces in order to make stuff like this work and I'm going to talk like Kana said there's a lot more to AI than just neural networks I'm going to talk about neural networks though and so if you want to train a model you might need to have labeled data you can train that algorithm but as Prasanna was saying uncertainties sometimes you need humans in a loop either one way or two way communication so how do you effectively bring people into these data pipelines and so what we've decided to focus on is really thinking about the human AI interface so how do you get these groups or individuals working really closely together while also considering the fact that there's not a lot of experts in the world that have the capacity or have the knowledge to identify things from you know just species level but there's also a lot of people out there in the world who could distinguish between a coral and maybe a jellyfish and could at least help kickstart the annotation process and then again from the user center design we've recognized that you can't just come up with one software solution that solves the widespread needs of a vast user community and so what we decided to do was chunk this up into more bite size pieces that target these particular communities Fathomnet which I think was this idea of data sets and making them available this is absolutely true and Fathomnet is one approach that we can take this is modeled off after like the Cocoa and ImageNet approaches and the computer vision community where they aggregate big, big quantities of data that can then be publicly available and used to rapidly train and iterate on new algorithms. I'm going to skip this slide because I don't have time. The second approach is thinking about data pipelines how you lower the barriers to utilize machine learning for individuals who have data don't have the time to learn data science skill sets and then also apply models that the community is developing on the particular data sets and improve them over time and last but not least thinking really creatively about how do you bring more people into this problem or this challenge and recognizing that there's expertise in all these different levels and so one of the approaches that we've taken and I know we're not the only ones there's a group in IFRMAIR that also has this approach where you're building video games to try and attract really, really broad audiences to be involved in community science activities and so the idea here is that over time this multi-platform approach you might be able to accumulate the human verification that you need when you need it to then export your data and so with that I want us to have a bunch of time for a discussion and thank you for the opportunity to share. Thank you Connie and thanks to all the speakers we have a few minutes before launch for some questions let's see I think the first one has got a number of people support it is are there Rick go ahead and ask your question Hi this is Rick Murray from Woods Hole Hi this is Rick Murray from Woods Hole my question is are there barriers inhibiting applied mathematicians and data scientists who are not trained in ocean science but from partnering with ocean scientists the biomedical field tends to do this quite well but ocean scientists do not we have no problem hiring physicists to become physical oceanographers or chemists to become chemical oceanographers but there's been real hesitation or barriers in terms of bringing in data scientists classically trained modern trained into that ocean science as well so I'm just wondering what's going on and are there any inhibiting barriers thank you the question maybe was for me I mean I like to compare this a little bit to sort of the different languages the previous speaker and the previous session has talked about talking to local communities and bringing everybody on board maybe talking about similar things and meaning different things the computer science of course community has it takes some patience to talk through between the domain scientists about the specific applications and then the solutions and the computer science space myself and I'm talking to people who develop compilers for running models on supercomputers so that's a really big gap that you have to be patient and be willing to actually to talk together there may be one issue of incentives also I'm not sure maybe in the biomedical sector there's some maybe clear clear ideas of the outcome of the system and which is maybe not so clear to explain to computer sciences the role that they could play in solving some of the big challenges in ocean sciences and maybe someone else on our panel has some other ideas but I think it's really the patience and the willingness to to explain what we need and what the other people can provide to us as a card carrying computer scientists if I may I think part of the problem is exactly what Patrick mentioned incentives don't exist and computer scientists are very happy to do their own thing publish papers and I know Pete you asked that question I couldn't figure out how a slider works so I'm sorry about that this is that response I think computer scientists have a bad habit of doing things within confined spaces they like I'm making a broad brush statement but knowing the people I know who are very prominent in computer science they understand that their incentive scheme is to produce graduate students papers and that's it and I am serious that has to be some incentive to be provided for them to go out and smell the ocean if they don't they don't understand what the issue is 95% of the time when I give a talk to my colleagues in computer science and I'm here in Europe I just gave a talk in probably the most prestigious AI robotics lab in Europe called LAS they were asking fundamental questions they're sort of interested climate change the impact on the ocean that's really important but their incentives are limited so I get what Rick is saying that oh this was done there the thing with computer scientists so pardon my being a little blunt I've always thought that NSF should add a C it's National Computer Science Foundation because they give tons of money to computer scientists I think they should reverse that and make it into an ocean science so there should be an O instead of a CNOSF so it really is there's too much money going into computer science and I get it that you can monetize it the politicians are happy the economy is doing great but they're not really looking at fundamental questions and that's important for civilization and I'm blunt about it not just because I'm with ocean scientists here but also with Peter smiling because he knows I'm blunt so that is the case you're not the only one who said that in fact I was at CBPR last year and one of their keynote speakers said the same thing so exactly I did want to add to that too the incentives are different because we're all published or perished but in their case it's published every six months a big important conference paper and either the CBPRs or NURBS of the world and that often doesn't align with the time it takes for us to accomplish projects and the other thing too is the medical community despite the sensitivity of data at least that community has made that data available for computer vision experimentation if you go to the computer vision conferences it's almost like every other talk is focused on some new data set that people can use and play with and apply their algorithms too and so being sure that we're making our data accessible to that community specifically is also a challenge which is again a reason why we focus so much on building fathomnet because we want to entice that community to use our data to answer really important questions that we can then apply the state of the art algorithms too in the ocean and if I could add a little bit more following that to respond to Rick because I think we have a generational issue also the younger people who are coming in are much more more interested much more focused on doing something about the environment the ocean included so I think we have an opportunity rather NSF does have an opportunity to attract them into the ocean sciences so it's not all gloom and doom I think our generation was a little bit different because our motives and intent are different unfortunately can you hear me honestly can you hear me can you hear me can you hear me I'm sorry we have to cut this off we've got a really packed agenda but I wanted to thank all four speakers it was very interesting and I'm sure it's given us a lot of things to think about on how to get AI better used in ocean science so our next panel is one that I am very excited about because we're going to talk about ocean solutions and in particular what it looks like to co-design and co-develop those those of you who are on the committee there is a tab 8 in your briefing book that gives you a little bit of background and maybe you've had a chance to have a look at those materials but thank you for the folks who organized the session for making those available and we have four panelists for this particular panel all four of them are going to be virtual and I'm going to wait just a little bit more for the committee members to pilot and to shuffle their way back there we go I'm going to just look if there is more I'm just doing a play-by-play for those of you who are on Zoom time to come back to all I found the rest of the committee okay and so we have Rosie Eligato, Kristen Olson, Katie Arcima and Charlie Fibon on this particular panel and I see some of those names already on the Zoom do I see Rosie? Rosie are you online with us? There you are I'm here so great to have you Rosie Rosie from the University of Hawaii it looks like you're the first one on our panel discussion and just about everyone is back here in the room so we are ready for you when you are great I have some slides to share I have shared them with the staff and just out of an awareness of time my staff there's actually more I'm going to just be much refer if I can so I'm going to share screen and hopefully this all works we'll see okay can you guys see what kind of looks like a ladder yep we can great so let's see titled this little presentation co-production and co-design I'm particularly going to be focusing on Indigenous communities but I wanted to start off with some general terms that I think my co-panelists will expand upon so the idea, the really underlying idea behind why we might want to do co-production is to engage with communities or different entities and this is Sherry Arnstein's citizen participation which was developed in 1969 and it really shows the spectrum and what we might actually call it from non-participation of citizens to varying degrees of tokenism to really giving communities agency and power and I would say that our goal or at least the goal of the research that I've been working on in my expertise is to really hit around level 7 that partnership but I encourage you to read this article it's very enlightening and the idea is when you have this the participation is partnership that really occurs when you have officials, researchers allowing citizens and communities to negotiate to have power sharing to share funding put forward requests again this is very general I just want to kind of set up for the mindset if you're going to go into doing co-production and co-design it's important but it's not always the answer it's not a checkbox, it's a means to an end and it will require significant commitment of time energy and resources no matter who you're co-production with, whether or not it's industry or indigenous communities and priorities should be given to the process of how it's going to take place no matter what that means it should be specific, it should be done in the context of a particular decision making or research question you want to establish what your standards are and recognize that it's going to be really specific to whatever whoever the parties are the entities that are involved it should support adaptive learning through both formal procedures and evaluate that and in this last column I have rights holders I use that word specifically in the case of indigenous people because the word stakeholders can have negative connotations but the idea is that there should be some focus on building meaningful connections and collaborations that harkens back to not wanting to be in the bottom part of this ladder and decision making should be shared it should include respect and values of all perspectives and there should be an understanding of reciprocity again this is also very general and I'm happy to share these slides and I believe the staff has these slides when you do kind of like a best hits literature review there are a lot of common important elements that are hallmarks of what makes a successful co-production process but I don't want to tell it or word bingo but I encourage you later to go through it but it is a wide variety there are many elements as the bottom line from critical reflection flexibility to leadership ongoing dialogue and innovation okay I wanted to before switching into more specifically about the academics and working with indigenous communities I wanted to acknowledge however this is a challenge for many academics to involve in and these are what I consider some of the major challenges to having effective academic involvement with anybody with community, anybody outside of the academy the first is let's recognize that we do not have a formalized training mechanism for facilitating co-production a lot of academics don't feel prepared to engage particularly when we look at career young researchers who are engaged in transdisciplinary research particularly those who want to stay in academia and often it's because this transdisciplinarity is seen as an add-on and what they're meant to be doing is hardcore research it's hard to measure the success of co-production if you don't have a lot of specific measures it can be hard to publish co-production research and these publications can often be seen or viewed by the mainstream in some part which is not always the case there is often on the institutional level a disconnect between what institutions say they want like they say they want to engage faculty and practices that reward faculty who engage in co-production it is also viewed not always as cutting edge in the conventional sense in terms of technologies because it is very people oriented which is very very old school but can be innovated in other ways and then conventional measures might not indicate that the co-produced science is legitimate or credible so I wanted to spend some time on this before launching into adding yet another layer of complexity which is research with indigenous communities and I'm going to spend the rest of my time I'm going to try and speak quickly so that my panelists, my co-panelists have time to really have good discussion on the specific process that I have had a role in co-producing which is called Kula Na Noi and this work I have not done by myself this work is indeed co-produced and co-developed between myself, Katie Hinson Brenda Asuncion from Kua Aina Ulu Awamo which is a non-profit Miwa Tamanaha as well as Sarah Kahanamoku and this is the watershed of Heia which is where this was really developed and I want to provide some specific examples for you the reason why we developed this process for engaging with indigenous Hawaiian communities is because this in Hawaii and writ large in other kind of communities often times where marine research takes place and ocean science can take place is that these can be sites of parachute science and scientific colonialism so we know there are several studies that have come out beginning to look at this also known as parachute science this is the practice of obtaining data or resources from other countries will not returning research outputs and it's beginning to be in recognizing a scientific colonialism and this is something that we need to be really careful of particularly in ocean sciences as we have so many observational systems that are remotely located and recent studies on parachute science particularly for example in the coral triangle shows the extent of externally driven research across Oceania and we saw this happening also in Hawaii and we wanted to understand is it happening here and how can we you know change the tide of that also want to recognize that people and places matter a lot in research it's particularly noticeable among indigenous communities I am an indigenous person and many of us have strong connections to the land as a central tenant of our culture and our cultural and spiritual practices and basically to put it simply the land is the basis for indigenous knowledge in Hawaii and relationships help us to grow and apply that knowledge to contemporary issues and just wanted to cite this federal guidance the OSTPC guidance on indigenous knowledge that further recognizes centrality of place and relationships in research with indigenous people so this really brings the need to do co-produce research which is situated in place to the fore and it's really calling our practices that allow both tribes and indigenous peoples as well as the scientific community to benefit from the research and to do that we have to take a place-based people-centric approach so again why the community researcher relationships it's because we one of the questions we asked ourselves in developing the process I'm going to share with you is how do we as university researchers build equitable relationships with Hawaiian and local communities in Hawaii what expectations do communities have for these collaborations and how do we hold ourselves accountable and responsible for the work we do in communities in place it's just simply a matter of you know ethical practice again I can't under emphasize that all of the things I'm going to talk about today were developed and co-developed with the purpose of establishing really wonderful relationships in this place called Heia this is the setting from which Kulana Nuii grew it's a very active community there's a strong presence of nonprofits and native Hawaiian stewards and it's a place of overlapping government and academic spheres of interest the University of Hawaii Marine Biology National Sentinel site the National Estuarine Research Reserve Resilient Lands and Waters it's a confluence of academia community and government and there are many potential opportunities but there were many challenges for miscommunication and when all of these entities interact and so it was a really amazing case study and also a place that we were really connected to and we were approached by the stewards of Heia Fishpond which is in the foreground from Paipaio Heia with a request to develop a set of protocols to encourage more reciprocal research practices so this is a long-standing kind of project that began in 2014 with our non-profit partner Kuo Wamo that they facilitated community meetings in partnership with Sea Grant we held several workshops with stewards the community and researchers we gathered all of that information their desires for what good researcher relationship looked like we also did of course a broader body of academic literature search because this has been done in the medical field and in the social science field but has not really been applied towards the geosciences and this was our end goal was to understand that we have actually different cultures right we have an academic researcher culture that has its individual interest needs and issues and there's community issues needs we run on grant cycles they run on intergenerational cycles often within indigenous communities but what we're looking to do is to leverage these to create that equitable and sweet spot I just wanted to do one more little side note parenthetical explanation because I don't need to use Hawaiian words without explaining what they mean Kulana which is the first word you saw means what is your station what is your stance what is your attitude in other words how do you what's your how are you walking around in the community are you kind of like ashamed of how you're acting you just want to get in and out are you proud of how you act and how you can do research so those two together really mean what are our research standards and so based on the intake that we had from the facilitated committees from the stewardship meetings we had from the broader research we distilled that down because again we wanted to apply that from social science and biomedical community-based participatory research to geoscience we distilled that down to eight Kulana And with that came best practices and I think what really makes our process interesting is it comes with guiding questions for both communities and researchers. And because it's not a checklist and it's a it's a question process that requires dialogue, it's flexible enough for broad application. Again, I want to emphasize the process that we came up with is not a compliant standard. It's not like your IRB it's not a checklist for achieving reciprocal community partnership. It's a set of ideas and values so you can kind of see that there are many elements that we just kind of organically developed that have that have similar the elements for successful co production. It's a set of ideas values and behaviors that when applied alongside hard work builds more just and generative relationship. Again, we organize those for those eight into four two groups of four. One is called building and nurturing pelina pelina means relationship. And those four, which are really the beginning relationships are they're going to seem really obvious but in practice they're quite hard respect reciprocity self awareness and capacity are what is our self awareness and capacity and communication. I can give like a whole training on this and I have, but the second part is once you've established that relationship. What is the iterative process for maintaining that, and we call that. Which means teach and learn, or a lot of my which means affection given an affection received and you can see we use a lot of our cultural language, because it's important for us to articulate these in the cultural values of Hawaii. We're maintaining a long term focus on what the community needs. It's creating community engagement and co review processes. It's having agreements around knowledge stewardship and government, and it's actually putting in a process for accountability. So, there are many several models for indigenous data sharing. I'm not going to go over all of them but that is, I think, an area that if we're going to engage in that as an ocean studies that sciences community and for NSF. We really need to begin to adopt and operationalize indigenous data sharing into our data management plans. I know that a lot of our data management involves adhering to the care prince the fair principles which is findable accessible interoperable and replicable. But the global indigenous data alliance has also developed cares principles which are complimentary to fair and can be used together and care stands for collective benefit authority to control responsibility and ethics, and I would really strongly that we begin to adopt and think about how to operationalize care in my role on the ocean studies board. This is one of my key issues that I always bring up. We have done this in Sea Grant. We have recognized that we have a responsibility to ensure that the programs and projects that we support because what you see that is a grant making entity. We have been able to engage in collaboratively mutual benefit partnership with communities. And so we have made training of this cool and a no e process mandatory for PIs as well as grad students. This is actually an old slide, it says that more than 600 community members have been trained. We are now approaching 900. This is 35 workshops and we're now approaching almost 50 workshops. So we have been acting as a capacity and network builder, but I do want to say just want to end with this that funders are a key player in ensuring accountability. This is one of the key places where we can build in how well and how the metrics of how successful co-production is. They have a responsibility to invest in reciprocal research practices and avoid an investment in extractive research. What I've shared with you today is just one example of a strategy for encouraging this and we're hoping that in the future funders such as NSF can support place based and Indigenous led research. I want to say that there is more here than just for the funders. There are institutional mechanisms for accountability and this is not a new conversation. It's definitely been ramping up and I wanted to give acknowledgement that in December 2022 the White House released the first of its kind guidance on Indigenous knowledge which is meant to help federal agencies incorporate Indigenous knowledge into their work from research to environmental making to co-management of lands and waters. And it has to be implemented of course at national, state and local levels and to the unique context in which these really exist. And I also wanted to share that we really need to reframe how we can implement these strategies. Okay, so just want to leave you with my contact information as well as the contact of the other members of my Sea Grant Center which I'm really privileged to be a director of and references and I'm happy again to share all of these with the committee. Mahalo. Excellent. Mahalo to you Rosie. Thank you very much for this great presentation. I love that you included the references as well. I really appreciate some of those are papers that I have not read yet, so we will definitely get to work on that. Next is Kristen Olson from Pacific Risa. Kristen, are you online? Yes, I am here. Okay, let's see if I can share my screen. I assume, can you see? Yes, we can see. Okay. Terrific. So, hello. My name is Kirsten Olson. And for the past eight years I've been part of the PI team for the Pacific Risa program. Pacific Risa is one of 11 funded, hold on a second. Pacific Risa is one of the funded by NOAA CAP RISA program. It's one of 11 national programs. CAP programs are competitive five year multi institutional transdisciplinary research collaborations that work with regional and local decision makers, natural resource managers and communities to generate, transform and translate climate information into practical adaptation tools and plans. The coverage of the Pacific Risa is Hawaii and the US affiliated Pacific Islands, and we've been fully funded by NOAA for 13 years. Our core team includes two co-directors, the lead PI's, two project specialists, a program coordinator, a communications guru, and the sustained assessment specialist, all of whom work full time on projects under the program umbrella, as well as seven academic universities and their students. CAP programs co-develop research with partners. So research supported under the current PAC RISA includes projects on a variety of topics, water resources and policy, climate health migration, ecosystems and biodiversity. We create the regional climate assessments of use to our Pacific Island neighbors. We focus on compound extremes and vulnerability, coastal hazard assessments, my field which is natural capital and nature based solutions, and integrated vulnerability and exposure. But what binds this research we do to social change are engaged scientists who have built stable relationships with decision makers to develop meaningful policy outcomes that prepare society for climate change. These elements we don't get to science informed policy change. In all cases we strive to involve project partners who are the ultimate users of information in all aspects of the project design, development and outputs. What co-development looked like, what co-production looked like in each case has differed of course. We included things like writing the research proposals together co-authoring peer reviewed publications or white papers and reports. It always has an element of sustained communication and iterative development of objectives, outputs, methods and informing policies and plans, or serving as a critical voice. All of those are kind of the different ways that co-production can look. Co-production is not straightforward and it doesn't always look the same. It's, but it is always more than a full time job for the staff members. So we end up focusing on where co-production is most effective and meaningful and being really careful of not overburdening smaller agencies and communities in asking them to be partners in co-production. I'm going to tell you a little bit about my current experience with co-production. So I spent the last weekend on Kauai with my master's students. We joined over a thousand people in a human chain to rebuild the stone wall of the ala koko fish pond, a traditional Hawaiian fish pond. Fish ponds have been used by Native Hawaiians for centuries to grow food, but most fell into disrepair after colonization. So fish ponds play a key role in food security and their restoration could be vital to community resilience in a changing climate. So we were there to build a wall, but really to build our relationship with the community-based organization restoring fish ponds. Being present, hearing stories from the families working alongside us, observing the water, the fish, the birds, listening to the staff and the volunteers talk about their questions and hopes. We plan to build a common understanding of what questions they wanted answered and locally appropriate ways we can go about answering them as scientists. My team and the County of Kauai Planning Department are interested in understanding how different approaches to nature-based adaptation are conferring resilience, as defined by the communities themselves. We're just working where and why, and how can policies support these efforts? But we're only going to get to that broader lesson through partnerships, and we will only gain partnerships through building trust and rock walls. So I wanted to zoom out to the broader program, the PACRISA program, and talk a little bit about how using the formula of engaged scientists building genuine relationships have led to meaningful policy outcomes. So for example, we've been engaged with the County of Kauai for years, working on coastal retreat. So my colleagues organized a peer-to-peer learning exchange between the County of Kauai and Boston, which helped accelerate Kauai's setback zoning legislation. What took Boston many, many years took Kauai less than a year to accomplish, thanks to this peer-to-peer learning. The program also developed a groundwater recharge model that was used by Maui's Department of Water Supply to justify limiting extraction permits in, and actually in the area that was recently devastated by the wildfire. The model was then transferred to Guam, where it catalyzed legislative action to protect the northern aquifer from military exploitation. In all these cases, it's important to note the critical role of legal scholars in the team to achieve policy impacts. So co-production and engagement is like tending a garden. It takes patience, persistence, and presence. You have to build genuine relationships, not just throw money at communities, and it speaks to the need to have dedicated engagement staff. Researchers in the team that are champions of multiple issues can persistently advocate when they perceive a gap and be ready to step in when the policy window opens. And rather than focus on all the barriers to co-production and engagement that drive my team crazy. No mention of university fiscal processes, for instance. I'll focus instead on some of the attributes that have led to impact under the PAC RISA. First and foremost are sustained interactions. These are non-extractive in nature, no parachute science, but a genuine commitment and shared values. Transdisciplinary co-development and collaboration where natural and social scientists team up with legal scholars, agency partners, and community members throughout the process of the project. And moreover, the team needs to be really sensitive to the external political, geographic, cultural context. A really important point is the flexibility of the grant. It needs to be nimble. It has to be able to catalyze creative thinking, be open to unconventional partnerships, methods, and implementation approaches, and responsive to opportunities as they arise. And finally, we need to reward applied science and non-pure reviewed outputs. If we want to leverage social change, we need to have partnerships with agency industry and communities to accelerate uptake and value things like testimony and other impactful engagement activities. Oops, my mouth is going crazy. Of course. Okay, so coming, kind of zooming out to what we're talking about today of ocean solutions. You know, these reflections from the PAC RISA experience will need, need to take, you know, we need to kind of think deeply about them, because the co-production and engagement in the ocean space is going to require extensive levels of collaboration that are very different to what NSF is used to. Ocean issues like fishing, deep sea mining, large-scale marine protected areas are international in nature. They may also involve deeply ethical and legal issues, for instance, related to human and indigenous people's rights. We're going to need to, as well, if we're moving toward solutions, supports going to need to extend into the long term, not just doing the science, not even just doing the science to policy, but then through the morass of policy implementation. So I'll conclude with kind of five ideas. If NSF really wants its research to lead to policy and societal change, you know, if they want ocean solutions, not just science, co-production will be key. This will mean that we need meaningful and sustained engagement to build credibility and trust, non-extractive co-development of projects. Engagement has to be embedded in the project plan with dedicated personnel and budget. And we need to recognize that it's going to require time and resources to build and sustain relationships. Secondly, is this point about flexibility and nimbleness of the grants. We want to be able to jump onto opportunities when they present themselves and try new things. And the community partners have to be paid and engagement activities have to be funded. Third, we need to expand the grants to increase attention to potential for societal uptake. Allow non-academic PIs, include legal scholars, fund those community partners, partner with the applied agencies to accelerate uptake, and reward these methods of amplifying research outside of peer reviewed publications. The fourth point is that we need detailed pathways to broader impacts. I've done a lot of broader impact statements in my life and reviewed many more. But we really need fleshed out program theories of change that include co-production and engagement to reach societal impacts, not just the scientific output. And they should go beyond the generic to be really specific to the community so that activities connect to the locally and culturally specific solutions. And finally, the evaluation plans for projects. Projects should include broad evaluation plans that are based on the well articulated theory of change and include locally and culturally meaningful impact metrics. And the outcomes of co-production and engagement activities should be part of the evaluation. Thank you very much. Those are my thoughts. And I look forward to the discussion. Excellent. Thank you very much, Kirsten. Next, we have Katie Arkema from the Pacific Northwest National Lab. There you are, Katie. Hi everyone, can you hear me okay? Yes, we can. Okay, I'm going to share my slide. Are you able to see them? Yes, we can. Great. All right. So I just titled this talk, the same as the name of the session today. And I like to start talking about my research by putting forward this graphic of Pasteur's Quadrant, which I'm sure is familiar to many of you in this working group today. It is a good sort of example of what is often called transdisciplinary research, and a key element of transdisciplinary research is the co-production of knowledge, which is a central theme of our session. Transdisciplinary research includes solutions-oriented science. It's most often interdisciplinary, and it also includes research and policy outputs. This is exemplified well in Donald Stokes' book that came out in the late 1990s, where he talks about Lewis Pasteur's work on bacteria, microbiology, and links to vaccination as a good example of research that really is at both the cutting edge and very solution-oriented. He doesn't talk very much about however the process for going through co-producing that knowledge and how we can work as groups of scientists, practitioners, community members, and different kinds of decision makers to do this work. And so I wanted to offer today two examples of co-design and co-developed research that I've been engaged in. The first is around nature-based solutions to climate mitigation and adaptation, and this is work that I was involved in and am still involved in in the Caribbean, especially in Belize and the Bahamas, with many collaborators of mine at the Natural Capital Project based at Stanford University, where I used to work prior to joining the Pacific Northwest National Lab. And I'll also talk a bit about some work I'm doing now with renewable energy transitions in remote coastal and islanded communities. And for the first topic, I'll focus really on the work in Belize, but you'll probably see some graphics filtered in from the Bahamas, and in the second section I'll talk primarily about our work with the city of Bainbridge Island. So this first topic around nature-based solutions, these are sort of potentially powerful approaches that involve blue carbon ecosystems and strategies. And they're powerful for climate resilience overall because protection, restoration, and management of coastal ecosystems like salt marsh, sea grass, mangroves can both store and sequester carbon and also help to provide climate adaptation co-benefits, such as nursery and adult habitat for fisheries, tourism and recreation opportunities, coastal risk reduction, water quality, and so forth. So we're able to sort of incorporate multiple goals and outcomes when we are exploring the potential of nature-based solutions for climate resilience. However, in the context of the work I'm going to be talking about today, putting forward targets for those nature-based solutions, especially in the context of the nationally determined contributions that countries are putting forward under the climate accord in order to meet the goals under the Paris Climate Accord is very difficult to do, actually having quantitative targets that not only tackle the carbon storage and climate mitigation portion of this, but also answer questions around sort of actual implementation to achieve climate adaptation co-benefits. It's not something that is kind of universally done in those existing NDCs. I sort of aim to tackle this problem, working with our colleagues in Belize. And Belize is a leading country in coastal and ocean management. It's located on the Caribbean coast of Central America. And it is particularly progressive in planning for the sustainability of natural resources. And I've worked closely with my colleagues in the Belize Coastal Zone Management Authority and Institute in the central government for about the last 15 years. And a lot of the work that we've done involves developing and applying models for quantifying benefits of coastal ecosystems for coastal community resilience. And so our colleagues in Belize came to us and said, we want to build on this past work that we've done together to tackle this issue around developing targets for nature-based solutions. And this truly was a transdisciplinary research team that involved members from local and international NGOs, the government of Belize, funders, and a whole suite of scientists from a variety of different disciplines. And one of the first sort of key aspects of co-developed and co-designed research is really figuring out, well, what are those research questions that are both interesting from a science perspective and actually tackle the challenges that people on the ground are facing? And so these are the two that we kind of iterated on and came to you together. And those are, what are the carbon mitigation and adaptation co-benefits produced by a range of potential blue carbon targets? And where should policies and actions be prioritized to provide a rich combination of co-benefits? And so that first bullet is really around, like, what does Belize want in terms of its ambition and its NDCs towards meeting those Paris global goals? And then second question is really aware. So where do we actually implement these actions? And our approach was to use sort of an iteration of both engagement. And I like to think about this sort of onion when I think about engagement where there's a core group that are, you know, many of whom are on the previous slide that I show that are meeting and, you know, maybe remotely, weekly and in touch quite continuously. And then there's sort of like engagement with the broader community of leaders and coastal planners and sort of private sector. So what we sort of pursued in terms of our collaboration between scientists and decision makers in Belize was to iterate between that larger engagement with a broader community and using, in this case, a set of social ecological models called invests that allow us to quantify the benefits that coastal ecosystems can provide to communities in Belize and beyond. And one of the things I wanted to mention here is not only in this work and my other work in the Caribbean and Latin America and here in the US. One of the things that comes up over and over again is this idea of in situ engagement. So it's about not necessarily having new events where you're bringing people in, but instead really going and meeting people where they're at. So showing up at cafes, showing up at community centers, having just open houses in churches, for example, and having the materials where people can come in and engage with the material, but not where you're having them sort of take time out of their normal day to do that. Also, conducting different kinds of exercises we've done a lot of participatory mapping where community members are sort of drawing how they see the situation now and what they want for the future of the places where they live and they work. And then one of the things that's very important is being really clear and communicating back how community inputs were integrated into the process. So not just assuming that because you've done that. That's sufficient but that you actually need to also be communicating how you've done that and bringing it back to people. So not going to dig into the sort of research results, but I did just want to show that one of the things that we found that's particularly helpful is when co designing co produced research in key incorporates multiple different potential outcomes that that cross different sectors and cross different disciplines that are measured with different kind of metrics and that's because that brings more diverse audiences to the decision making table. We've also found that like having results that show where certain actions can be taken and not just the sort of amount of an action that should be taken is where this becomes useful from an implementation perspective. So believes is now working on developing their mangrove restoration and conservation plans based on a lot of this research. That's the follow up to the update to their nationally determined contribution so having maps showing where things can be sort of most effectively done is is helpful. And as I mentioned initially we have one of the things that I like to do in my work is think about both the science outputs for the scientists and the researchers so getting this work published and validated through the peer review process, and also ensuring that it's designed that it's actually going to be useful from a policy perspective. And so this work to help to informed believes is most recent update to their nationally determined contribution submitted to the NF triple C. And that includes two time bound targets for blue carbon strategies the first is protection of 12,000 hectares of mangroves beyond the existing protected areas by 2030. And the second is restoration of at least 4000 hectares of mangroves by 2030. So I wanted to mention this this map is actually from our work on coastal risk reduction in the Bahamas. But one of the things that we found is that going ahead and developing simple interactive web based tools for our partners in in the region to use how they would like to use it for communicating to a wider audience of their colleagues, and for capacity building can be really effective. So it's sort of a way for people to engage with the data and with the analysis and really also get under the hood. So that the work that's being done is less of a black box and more something that is informed and then can be iterated on and improved based on on people's feedback. Okay, so I'm just going to have just a couple slides on the renewable energy transitions and remote coastal and islanded communities. I wanted to talk just a little bit about some work we're doing with Bainbridge Island in Washington State. Bainbridge Island has two key goals for its renewable energy transition. One is that they want to be 100% renewable by 2040, which is ahead of the Washington State goal, and also increased energy resilience for emergencies and climate change. And Bainbridge sort of exemplifies challenges that coastal communities are facing all across the country and all across the world. And that's that places that are already sort of at risk for loss of power are becoming more at risk for that as we're seeing increases in coastal natural disasters. And a lot of these, in the case of Bainbridge is an island, in the case of the Makatrive, where I'm working in Nia Bay, it's more remote. And so these conditions are sort of leading them to think, I think in some ways much more on the cutting edge than then say many cities are doing about sort of local generation of energy using new renewable energy technologies. And it's because getting that in a lot of places sort of have to import power and it can be, sorry, import fuel and it can be very costly. And in the case of other places that at the end of really sort of fragile transmission lines. And so if they can locally generate renewable energy, it's a way to tackle both the sort of decarbonization goals and the local resilience goals. And so this is some work that I'm working on. It's funded by the Department of Energy. It's with not only the Pacific Northwest National Lab but also the Renewable Energy National Lab and other sort of community based energy partners. One of the things that's unique about this program, it's, Bainbridge is just one of many communities that's involved in it is that it's technology agnostic. And so what that means is that we're not going in and saying, hey, community we want to test out this like, you know, solar option and saying, hey, we want to test out this distributed wind option. We're hearing from the communities what kinds of technologies are you interested in and how do these interact with your coastal and ocean environment and the values of your community that you want to uphold. The other thing is that communities apply into this program. So again, it's not the Department of Energy seeking out communities or places to test technologies it's communities raising their hand and saying, we want to work with scientists to figure out sort of potential pathways for meeting our energy goals. And the latest kind of aspect of this is beginning to incorporate other non energy considerations. So in development of new renewable energy, how do we also consider outcomes in terms of sort of impacts on other benefits that people want from coastal and ocean systems like the ones that I showed for the bullies example so recreation tourism opportunities fisheries related outcomes coastal risk reduction related related outcomes. And I'm not going to go through this but this is just to say that one of the key parts of any, many of the co produced and co design work that I've been involved in is this sort of development of pathways or development of scenarios. So here I find that the science and the, and the sort of practice elements are applied elements of the work really kind of collide. And that's that is that's in this sort of like quantitative and qualitative scenario space, where we're sort of, for me as a scientist listening to community members say, these are the things that I want to explore. These are the potential solutions that we're interested in, and then figuring out okay how do we now bring some some analysis some interdisciplinary research to help answer which one of these pathways or scenarios might be most useful to move forward on. Maybe it was Rosie that mentioned this I can't remember but I, one of the things that that I've often sort of groups that I've worked with and I have often thought about and struggled with the sort of understanding what the impact is of co designed and co developed research and this is a graphic from some paper from colleagues of mine at the National Capital Project at Stanford. And this is one way of thinking about potential increases in impact. And this first pathway here on the left is the conduct research pathway. So results are produced they're published they're disseminated. This is almost often a thought of as kind of the end goal of a traditional research or academic research but really sort of co designing co developed research this is often the first step. And then how can that analysis be used to change perspectives in a place. So, for example and our work in the Bahamas people there really weren't thinking a whole lot about nature based solutions. About 10 years ago when we first started working there, and then a number of hurricanes and a number of collaborations and major efforts by all kinds of, you know, local and international NGOs and scientists have really kind of changed the conversation and the Bahamas, and and highlighted the importance of nature based solutions. And this pathway three. When is research when is analysis when is sort of the solutions oriented perspective of the sort of community part of this work when is that all come together to generate action. And so maybe that's it's informed plans or policies. Maybe it's led to new financing mechanisms, and then ultimately where we really want to get which is this sort of outcomes. So have we improved the way that in this case biodiversity and ecosystem services is considered or the health of nature based systems that are providing nature based solutions, and improved outcomes for human well being. And so I just want to end with this sort of challenges successes, and the future of co produced research with a few things that I find in particularly important. One is the importance of investing in iteration and long term relationships. And we've all really emphasize this, but but I sort of, I want to add to that by saying it happens on all different scales. So, often within just a single project over the course of the year, you're constantly scoping and re scoping and sort of reminding each other why you're working together. And that that needs like process and investment in that iteration. And then there's iteration over years and decades where we've developed these long term relationships, and then community leaders come and say we want to build on this last thing that we did and now we want to do this new thing, like I talked about in Belize. I think the sort of like scenario development space is where really like the rubber meets the road for scientists and and practitioners and communities working together, and that needs to include both qualitative and quantitative approaches. There's a lot to sort of dig into here but I think this is an area that that really is fruitful for advancing our research space for co co produced research. The importance of exploring multiple interdisciplinary outcomes that resonate with diverse audiences is really key. And then supporting capacity building for community scientists and decision makers. And that's community so they can best engage in this work. The resources are are are are made available for for communities to build capacity to actually fund positions for people to specifically engage in and co produced research to advance some of the goals that they have. And the fact that scientists are often aren't trained is something that that we talked about here. I teach a class at the University of Washington through my joint appointment there. It's specifically with professional masters students, working through, like on the ground examples of how science at science can be brought to inform some of these sort of major decision challenges. And those professional masters level students will be the ones that are in agencies, NGOs, other kinds of community based organizations that can actually use this work going forward in their professional careers. And then of course decision makers they need capacity building to understand how do we actually help to foster these kinds of teams. So that's my last slide and I'll end there. Excellent. Thank you Katie you covered a lot. Thank you very much. And last but not least, we have Charlie pie bond from Sir frider foundation. Hey, thanks for having me and I will try to share my screen. When I realized everybody was doing PowerPoints. You can be special. Well, I prefer not to look at myself it just messes with my head while I try and present so I'm just going to give you guys a slide as a picture of me. So I want to look at myself talk while I talk to you guys. Thanks for having me my name is Charlie Plyman I'm the Oregon policy manager for Sir frider foundation. I a little disclaimer I am not an academic. I don't consider myself a scientist. And I don't have a lot of letters behind my name, or have not spent a lot of time in the ivory tower. I do a bit of science and rub up against science a lot. And I've been involved in a lot of co design co developed projects here in Oregon and so I think that's why I was invited here and I'll share a little bit more on that. For folks that don't know Sir frider foundations and nonprofit environmental organization dedicated to the protection and enjoyment of our oceans waves and beaches. And we represent recreational users. We work on clean water coast and climate issues like protecting beaches, adapting to climate change sea level rise. And, you know, and protecting special places, both ecological places that are important to our members, but also recreational spots that are really important to our members. And while it sounds like we're a bunch of surfers the name does sort of imply that we do sort of represent a pretty broad constituency of ocean users people that love the beach people that like to go to the beach. And of course ocean recreation is really expanding and changing rapidly. So there's a lot of people that participate in ocean recreation and beach, beach recreation other than just surfers. So our name is a little above a misnomer. Again, I think one of the reasons why I was invited here was that our organization in our advocacy uses a lot of science and we work with a lot of scientists and depend on science. And we're ultimately to advocate for solutions, both ocean solutions coastal solutions. And in every day issues like plastic pollution I would say is we've been involved in a lot of co design work in that space as well. And I think another reason why I was invited here is that for the past 10 years in my experience and advocating and working for Sir frider in Oregon. I've had the pleasure of serving on the Oregon Sea Grant advisory council. We do a lot of grants, we see a lot of grants come in and I'm on a advisory council of a lot of different types of members from the community, not just scientists but farmers and fishermen and otherwise that actually review these grants. And a lot of that's because Sea Grant sort of demands a high level of community engagement in Oregon and they have some high expectations for that in their science and in their granting. But that's reflected in their review process and I review things under rubric specifically for on the level of community and societal relevance. So, again, like Sea Grant, sort of an opposite problem maybe then NSF in that they have a high demand for community engagement and a high demand for co design and co development in Oregon but they don't have a lot of money. And one's an interesting predicament for scientists and one that we watched in years and then being challenged with applying for funding. And so I wanted to talk a little bit about how Sea Grant sort of met the needs of scientists and trying to be better co designers and be better community engages and work better in a transdisciplinary way. The way that sort of came about was what we heard from scientists was largely like, this is a small amount of money that Oregon Sea Grant is funding each year your expectations are really high. We need more time, and we need more money. If you want us to do this type of transdisciplinary work we need time to pivot build relationships and trust in our community. And actually test and pilot out the questions that we're asking right now and see if these are the right questions. So a lot of these things that we've heard reiterated over and over again from our previous presenters. What I found was sort of built into the process that Sea Grant reflected on and created in response called seed to leaf funding in Oregon and in the way that works is that seed to leave proposals are required to approach research from at least two different disciplines. They require an engaged component and it's designed to integrate potential information users and stakeholders into the research process itself. And it requires an outreach plan to ensure that the research process and results are useful and usable to those constituencies and communities. And so the seed to leaf granting works where the first seed proposal is funded for 12 months and it's an exploration of that idea. It's an opportunity to build those relationships to understand whether or not the research questions are correct to build your community of knowledge that's going to inform hopefully a leaf project after that 12 months of initial investment. That gives scientists time to do all of that work, sort of build and maybe that's, they realize halfway through that or through that 12 month period that this isn't the right project. But hopefully what that what that leads to is a leaf project was a second round of funding. And in the second round of funding that's a more longer term engaged project. Building time in relationships and trust into a two phase process that sort of allows for co design and co development. Again, researchers were kept telling the folks at Sea Grant that this was just like too too big of an ask for such a small pot of money and such a limited amount of time. And so this is sort of how they reflected in and change their funding opportunities for that. So I think it's been a really powerful way of funding projects and culturing transdisciplinary research and co design and co development. One project that I've been involved with is called the Cascadia co pes, or it's the coastline and people's hazards project. And this has been funded now at NS NSF level. I think somewhere along 14 or $19 million, three different states. And again this is addressing sort of communities and people's hazards along the coast, particularly particularly responding to hazard events like the big one the big earthquake and the Cascadia Cascadia co pes is an ecosystem zone that we may be exposed to chronic hazards from sea level rise, as well as the cute hazards from things like earthquakes and tsunamis. And to do that it's built across five different teams, trans transdisciplinary teams that are engaging with communities and exploring how to fill the best knowledge gaps to help them adapt to these coastal hazards in the future. And in that advisory council, I've learned a lot about co design and co development. I come in to that advisory council of that project as a recreational user as a recreational interest. I mean so a lot of my advocacy to the science team is to try and steer things that steer the science and the knowledge building around things that support recreation so where a lot of that scientist in the past have been engaging local planners and maybe engaged homeowners around hazards for everything behind the shoreline, engaging recreational users really, you know, identified new beach safety and beach hazards that the, but much larger portion of the public would be exposed to, and, and unless so just and you can see how slowly this starts to peel back the layers of economy and housing and behaviors and how we act and how we buy and sell houses on the coast and a whole new line of science other than just sort of the geophysical science of the coast starts to open up. A lot of socioeconomic, a lot of decision making and how people act and behave isn't based on how much sand is moving around and how we can predict that over the future it's really based on how people act and behave. It's based on emotions and responses and so this project sort of builds into that space the opportunity for us to ask those questions and so now we have a beach safety metric. It's built into some of the exploration of run up and exploration of erosion that we classically looked at as back shore hazards as protecting the back shore but looking at it from protecting the beach as a recreational user has been a high interest of ours. Other users like tribes have looked at this from a cultural perspective, the entire coast that's eroding away underneath of our feet isn't just a hazard but it's also exposing a lot of archeological things very important cultural cultural information. There's a whole line of other science that goes with that and informing that and actually protecting that information which is really sensitive information. And so you can imagine that this, this, this is really about peeling back the layers of the questions around the hazards on the coast. And as you peel back those layers are not as there's many many more questions and a lot more solid science and knowledge that needs to be built into that space to help communities adapt to program to see level rise and other types of coastal hazards. I hope I can have more maybe to the discussion. As we move forward here I don't have a whole lot to share other than my experience in this space as I guess what would be classically be called a stakeholder or a constituent and I represent a lot of folks in that space probably, but I'm a quite engaged one. And so I've been involved with many projects in the co design and co development space, as well as the funding of those projects through my participation in the Oregon Sea Grant Advisory Council and their seed to leaf granting project program. So that's all I have. And thanks for the opportunity to speak with you guys today. Excellent. Thank you, Charlie. Thank you for offering your perspective. It's really great to hear from somebody who's really been hands on and all of these projects. So we do have plenty of time for discussion. There's already some questions on slide. Oh, those of you who are online or in the room as your questions arise do put them into slide. Oh, Mona, I see you have two questions here and I don't know what order you want to cover them. But maybe we'll start with you. Thank you all for the great presentations. I'm delighted to see Sea Grant highlighted. Obviously I'm biased I work for Georgia Sea Grant. I wanted to ask all of you, are there the idea of aligning the academic incentives, hiring promotion tenure to recognize co co production comes up very often. And yet I, I do, I'm not aware of colleges programs departments that are using innovative methods, innovative frameworks to reward faculty students professionals who do this kind of work. Are you aware of any successful programs that any departments or schools that have done that. And my second question is related to co benefits of co producing knowledge. Other communities engaged in evaluating those co benefits and I think this question is probably for Christine. Thanks. I can jump in to start with the, the second aspect, I mean in the Pacific Risa. It's a large program, you know many millions of dollars and one of the things we've had since the beginning is an external evaluator. It's a really detailed theory of change that's driven that evaluation. And so yes, the, the, the stakeholders and the rights holders are included in, in her evaluation. I think it takes a genuine interest in the parts of the PIs to want to know that information and use that information and adapting their science. So it's, but it's, it's a non trivial task, for sure. And, and it's expertise that is not necessarily, you know, true evaluation program evaluation is, is not something that many academics are trained in either. So I guess that would, that would be my summary then. Yeah, any thoughts about the question about the evaluation P&T. So I'll take a shot at that. So, I'm not fee and the, you know, through the Keck Institute future initiatives. They, I was lucky enough to participate in an entire multi year workshop on interdisciplinarity is what it was called and how to appropriately this was led by Tamara Tickton at the University of Hawaii at Manoa, as well as Bonnie Keeler at the University of Minnesota and they brought together a number of folks who are working in, in this space to develop really a working report. And I believe I sent that to someone when we had our pre committee meeting when we had our pre panelist meeting. But there were a number of suggestions of how, how in the administration and it really has to be in the tenure and promotion guidelines. We really redefine what meets the criteria for expectations of tenure around scholarship. We found in our analysis that that is really where the rubber meets the road is in the tenure and promotion guidelines. And so that's really where it starts for people who are interdisciplinary you can see I have a joint appointment Kirsten has a joint appointment. Potentially in a unionized environment to also set up memoranda of agreements that the union or bargaining unit can help to have the faculty member to protect them and to kind of balance out. I don't want to get down into the nuts and bolts but it can include things like evaluations can come from community members or can come from people who are able to judge the full range of that individual scholarship. So that can be encapsulated in a memoranda of agreement for their evaluation for tenure and promotion. Great ideas Rosie do you want to talk a little bit about you all's advanced program to that you all started out. I'm sure I happen to be on the advisory board of the NSF advanced program that is here at the University of Hawaii at Manoa. We are in the phase of moving from our data collection. We did a huge survey on in this case it's, it's gender. Not so much community engagement but so I actually can't speak to that yet because we have our first advisory meeting next week. But we are looking at ways to directly interact and interface with the tenure and promotion process in order to better recognize the kinds of scholarship that particularly women in minoritized faculty engagement, which happened to also be a lot of transdisciplinary high risk work. Thank you Rosie as you can see I'm impatient to see the results of that work. Really exciting stuff. The next question she made you want to ask yours. Yeah, thank you to all the speakers. Hi Kirsten. Hi Rosie. So we've been talking about this actually at lunch break. Are there any suggestions I guess Kirsten I you I borrowed your words on how NSF can facilitate quote unquote flexible and nimble grants to match community interests and their timelines, especially with Congress timelines on funding cycles and and whatnot. Just some ideas. Yeah, I mean I think Charlie brought up, you know, a great model, right where there's seed funding where the expectation is that you develop those partnerships and develop full proposals. And I mean NSF has some things that are akin to this like in the workshops that you can fund a workshop and within a program or. You know, the, you know, the intent of those workshops is, is about building the partnerships and coming up with the full proposal ideas. I like that seed to plant the idea. And I believe model does is a good, I think example of kind of staging the funding in a way that allows for flexibility reflection, you know, building the trust the relationships but also you know science doesn't happen on a schedule all the time right like if you're doing ocean science. You might not, you might not get a weather window in a 12 month period that presents the right at sea time to conduct the research that you want to do right that's that's flexibility that on the Oregon coast, you know, we have to have. And so, I think it's important to recognize that it's not just co development co design that demands time and flexibility but sometimes science, physical science itself and that the physical constraints of the environment, demand that as well. And so we heard that, you know, over and over again at Sea Grant when we were reviewing and hearing from scientists and I would engage with scientists a lot in that space. I think that the ability was important I think staging things out is also a way to take some risks and find out whether or not you do have a good community to co develop knowledge and co develop a project. I'm sure people have been engaged in a project where you got some funding and about halfway through that project you realize it. It really wasn't going to work now we always make it work right, but there's times when you need to reflect and maybe shift and and I think that's what's great about staging and thinking about staging funding in that that way. Yeah, and connecting the dots I mean earlier this morning we talked about the long term ecological research model that to sort of a long term investment that can assure that the relationships are sustained and, you know, remain healthy. So maybe there's some sort of the other extreme perhaps. There's a question on Slido by Peter Gurgis Peter do you want to unmute yourself and just ask it yourself. Sure. Thank you. Can you all hear me okay to. Yep. Excellent. Thank you. Well, first, thanks for the fantastic presentation folks I really appreciate it. It means a lot. And my question was about building trust. And, you know, so many of the points you all raised. I've heard from my colleagues. I have a couple of my colleagues happen to be leaders of minority and aquaculture and minorities in aquaculture and black and marine science and they have, among others have submitted proposals to NSF, they've been rejected. And, you know, it immediately it shatters trust and that they just they they feel like it's gatekeeping and they're being kind of left out so I would love to hear from your points of view and experience. So what are the things that NSF as an agency can do to really start building trust with with, you know, our, our, you know, our diverse colleagues right our colleagues and scholars from from a variety of backgrounds and institutions. Thank you. Great question Peter. So, you know, Charlie put it in the chat diversify the review team. I think that we also have to do a better job at disaggregating data. I'll share a manuscript that I was a co author on last year where we exposed over 20 years of funding inequities with the National Science Foundation. We were able to definitely see that it is a pervasive issue, not just in the geo directorate or the bio directorate but pervasive across all the directorates. And there are really strong biases. I identified the scores that, at least for the disaggregated ethnic and racial data that we have to be one of the major factors and so I do think diversifying the review team is a critical component. I just have to start getting better data, because I think a lot of folks don't still, and I can say this with confidence because we've had meetings with NSF are our, our, our author team that they are looking for more evidence but that is evidence that any NSF can do because they need to actively disaggregate their data for us to be able to understand what is the true nature of the magnitude of this disparity so I would say, before we can even implement interventions, we need to measure the disparity in order to build trust that NSF is willing to accept that there is a disparity in place. So I mean I think to me that's a very strong first step but maybe my co-panelists also have other ideas. Maybe I'll interject here to take the co-chairs prerogative I agree with you Rosie that we need to measure, because if we don't measure we won't see any improvement. But what I want to really just, let's just not deny that there is a problem. Let's just accept we have a problem we don't need any more data to accept that there is a problem. We need more data to see whether or not the interventions that we might think of are actually improving the issue. But I think we should really be past the stage of still needing to prove that there is a problem. And if there's anything this committee can do to help. That would be great. But anyway, coming back to the question I'll step off my soapbox. Any other ideas Charlie it's great that you put some thoughts into the chat. Any other ideas for many of you Charlie. But I don't know anything about NSF's review process and so I am, you know, would go out on a limb to make recommendations to you on that. But I guess my, my, my thought and I hate to keep going back to what I know, which is the Sea Grant process and what I've been engaged in over the past decade is that, you know, when I got involved with Sea Grant, I, and I was on this review team and I looked into who was reviewing the science grants that were coming in high level research work to not easy to understand and comprehend, but they were farmers and fishermen and people with all ranges of education and backgrounds and I kept asking myself why are there farmers from Eastern Oregon on Sea Grant's advisory council. You know, well, well they're part of society, and they, they are reviewing this for societal relevance they don't have to know how the science works to understand whether or not it's relevant to them. And so, I think that in the process of review there's a lot of discussion from that societal relevance group. And, and we may not think that we influence projects a lot. But I actually think we do. And I've watched that in real time happen, particularly through that through that seed to leave kind of process and so I think it demands that we take a hard look at who's using the science, who's applying for funding. And we represent those communities of interest in the review process in some way shape or form. And that mean that might mean you have many, many types of review teams, as does Sea Grant in Oregon and technical review teams have societal relevance review teams have community engagement and outreach for review teams who all have expertise in those areas but I'll stop. That's that's what I have to share. Great thoughts. Thank you. I have a question from an anonymous person on Slido and since I don't know who they are I'll just go ahead and read it out loud. And this maybe specifically to Rosie but others might have some thoughts too. What are common challenges and bringing the care principles into data collection management sharing that could be addressed with revised DMP guidelines. So, there are different ways to operationalize the cares principles, but I would say one of the biggest issues centers around the a which is authority to control and that is because the western model of data is that is an ownership model that we own the data that we as researchers who collect that information on the data. And at the same time it's quite actually at odds with the idea around fair right like that it should be freely accessible and etc. But we do that that implicit within Western society is this idea that we own the data and knowledge, but a lot of indigenous peoples and I don't want to be monolithic in scope. And I think that information is collective. That that it not only is it collective but also people have to earn the right to have that data so not everybody should necessarily have the right to know where the fishing area where the you know should have GPS coordinates to the fishing areas where fish aggregation sites are right because that could open your community up to to being poached by outside commercial fishing. So, it's it's this idea. So those I think there's some really foundational fundamental epistemological, I would say, differences that are at odds between different indigenous peoples cultures and viewpoints and the western way of thinking about data ownership and So I'm not saying that data shouldn't be accessible but we should, we really need to think hard about what are the potential harms that might come from making data completely open. I have a lot of thoughts of this with regard to microbiome data, but and DNA and other kinds of things but I think that is the number one big picture thing is that there is a fundamental conflict or variance between notions of data ownership and that's what I think about it as knowledge stewardship and data governance and not ownership. So that's what I would say is the number one sticking point. Thank you Rosie. Melanie feelings has a question here in the room. Thank you. Yeah, so my little bit of context for. Well, the question first is, what are alternatives or possible alternatives to participant support which is category within NSF proposal budgets for by which non academic partners in co production of knowledge could receive NSF funds without adding a large administrative burden for them. And the context here is, I'm a PI on a NOAA project in Joe Schumacher who was on an earlier panel is actually a collaborator. And there are quite a few other collaborators from Washington coastal treaty tribes, but they're all unfunded collaborators, partly because of the short timeline of NOAA calls but also because the research coordinator at National Marine Sanctuary we worked with was helping educate me and my team and how to respectfully work with the tribal partners who we wanted to partner with for the first time and one thing they said was, because I said can we have sub contracts to tribal partners who are interested in that and, and they said probably they won't be interested because there'll be a huge administrative paperwork burden like they're not set up with DUNS numbers and all the reporting and all that so participants support will be probably their preference and I'm just wondering your opinions on that and and those are the only two mechanisms I know so are there others could there be others, what, what have you pursued. I'll take a sort of quick crack at that so this is a huge problem. The communities that need the support the most are the ones that have the least capability of like taking that support because of that added burden. And so one of the things that that I'm hearing from some of our in particular tribal partners, who are repeatedly being approached to engage on a whole variety of issues along the West Coast is that they actually want to have funded FTE for people within their, within their staff to then engage with scientists with representatives from a whole variety of different agencies, and to do it in some of the like newer sectors so, for example in this like renewable energy space a lot of them have some, some ocean policy staff already on board, especially with the focus on fishery science but now when you're wanting people to be able to respond related to new technologies it's like it's exceedingly challenging and so I think really respecting the fact that we shouldn't just be giving honorariums we shouldn't just be giving some contracts we should actually be funding positions. Yes, I saw a lot of nodding around the zoom screen here as you were talking Katie thank you. We still have a few minutes left and I guess I want to ask you all a question. As I was listening to all four of you talk about your work. I can see how there is a thread in there that all of you are really co producing means you're bringing the knowledge of the communities of place or communities of practice, you're bringing that knowledge they bring to the table, and sort of an equal footing with the more hypothesis driven science that we've been trained in. But can you all maybe some of you speak a little bit more explicitly to what it looks like to bring traditional ecological knowledge or local ecological knowledge into this work. Can you all speak a little bit more to them. Yeah, I can. So, So in my work, I've noticed that there's key steps in this iterative process. There's that scoping step, which we talked earlier about funding, sort of in and of itself, in order to work with communities and understand what the challenging is that they're facing. Then there's sort of a data collection step. An analysis often or scenario development step where you're sort of trying to understand different options that communities are thinking about and then sort of bringing the science to bear on that. And then sort of now how do we synthesize what we've learned and make sure that it's in a way that there can actually be uptake on the policy or the decision making side. And I think it's really important to recognize that the community knowledge like comes in at each one of those phases. So the scoping often includes like framing of the research questions but those aren't just research questions they're their management or their their investment related questions and so I'm really ensuring that that that that it's sort of like the on the ground knowledge that's helping to frame those out. That data collection step it might be bringing sort of globally date available data or or sort of locally generated empirical information. It's also going to be bringing in that local knowledge where that local where those local knowledge holders want to share it so as as Rosie was saying there's there's a really important need to recognize that that we need to be careful about how we're we're we're incorporating that knowledge into this this process. And then, and some of the ways that we've done that I mentioned it briefly in my talk are through sort of. In the case of the work that I'm thinking of we, we did a whole bunch of participatory mapping we probably came up with, like, close to 500 maps from community members talk, sort of where we asked them to explicit questions. What's happening now in the places where you live, and what do you want to see happen in the future. And can you help us understand where those different things could occur and how they could influence the aspects of the social ecological system that you care about. And we took those maps and based on our conversations with community members synthesize those together into some distinct kind of alternatives that represented alternative viewpoints. And then returned those back to the communities. So we digitize them and then said, does this reflect what you are telling us. And so we kind of like integrated that that drawing into then something that you can think of maybe as a more Western science science approach that there are these digitized reflection of their ideas, and then change those things where we've gotten them wrong. And then use some interdisciplinary models to sort of analyze the outcomes they told us they cared about underneath those, those different, those different alternative sort of options. So in that synthesis phase it's all about how do you want to quantify these outcomes that you told us that you care about what metrics do we use. How do we visualize that which visualizations are most important. So I think really just understanding that that it's not just in that data collection phase that that traditional ecological knowledge might come in but it's framing the whole process. Great, Katie that's very thoughtful and useful. Thank you. I want to be respectful of your time. Thank you very much to the four of you. Okay, we're going to talk about research priorities for related to ocean acidification and de oxygenation and habs. And the first speaker is Alexis is Alexis are you ready to go. I am here. I'm, I'm ready. I will share Maria, were you going to go first or was it me first. Okay, I can go first. Okay. Let me get ready to share my sermon. Let's see. It looks like I just have the option to share my desk top so I will do that. And let me know when you're able to see it. Yeah, we can see it and I apologize it only gave me the option to share my desktop, not the, the PowerPoint so I'm trying to hide the zoom controls. So you can see, hopefully just my slides more or less, we can see fine. Okay, great. Thank you very much for inviting me to speak. My name is Alexis the Lori Orton, and I am a program officer at the ocean foundation, where I run our ocean science equity initiative. And as requested, I'm here to talk about ocean acidification research priorities and and from my perspective at the ocean foundation really on technology research and capacity development needs. And to frame when I was thinking about, okay, how would I approach this discussion topic, you know, in proposing priorities for for 10 years. I thought of these two discussion questions and so I wanted to kind of put these at the front of my presentation and I'll also kind of pose them at the end as well, and kind of put them as context for the presentation I'm giving as these two fundamental questions are the tools that we currently have to measure away the oxygenation and have sufficient for the scientific questions, we need to answer today. And what role can the US and NSF play in testing and defining new methodologies for assessments of sufficient quality to answer emerging questions. And when I say measure and hubs, I don't just mean measure, are they happening, but really the full assessment of are they happening where where is it happening why what are the effects. So these are just some discussion questions that came to my mind as I was preparing these slides and I thought we could keep in mind as I as I speak. So, I come from probably a different background from most of the people you've heard from today, and that I work at a nonprofit foundation so I run our ocean science equity initiative well what is that we believe that as our blue planet changes our community's ability to monitor and understand the ocean is inextricably linked to their well being, but we also recognize that the physical human and financial infrastructure to conduct ocean science is inextricably distributed across the world and within the United States. And so we work to ensure that all countries and communities can monitor and respond to changing ocean conditions, not just those with the most resources. And most of our work is international but we do do work within the United States, particularly in Puerto Rico. So, a lot of our work has focused on ocean acidification and trying to support increased research capacity, and we often find ourselves at this crossroads, where we sort of have two sides of the coin, where we have either ineffective or inaccessible technology at our fingertips for addressing away research questions. And I imagine this is likely the case for habs and deoxygenation and other ocean parameters of interest, where you might have something like on the left where you have a handheld instrument that's great for taking out into the field that might be rugged and robust. You can purchase it from someone there's a company that will pick up the phone and answer questions if there is an issue. You can have it serviced, but the information that it gives you might not be of sufficient quality to actually answer a scientific question that you have. There's something on the right, which is the Berkley later, which is this custom built amazing system that has actually made huge differences for the shellfish industry on the West Coast, and gets climate quality data, carbonate chemistry data, but it's custom built, it's $50,000. It takes possibly two years of training to learn how to run. It is, you know, difficult, expensive, you can call Burke and he'll tell you how to fix it but there's no right that you can purchase one of these from. And so you're sort of in this gray area of how do you address the scientific questions that you have with the technologies available when you either have ineffective or inaccessible. So some of the work that we have done to try and address this is to find that middle ground or either by compiling existing equipment into a suite of materials or designing new systems. And I think this is something that, you know, NSF already supports quite a bit with SBIR grants and things and we've worked with some recipients of those grants, including sunburst on their fighter to try and get those into the hands of more users but doesn't always require starting to scratch the go on in a box kit as an example of something that we developed with go on the global ocean acidification observing network, where it's all off the shelf components that when compiled 90 pieces of equipment that you can purchase from Fisher and other suppliers, but when you put them all together, you get weather quality carbonate chemistry data, and all you need to see water and electricity to start so we've shipped 20 of these kits around the world. And people are publishing data and what's nice about it is that it's modular, you can, I've diagnosed problems easily, there's training materials that we've built around them videos, you know, best practice guides, data templates, all sorts of things and so kind of recognizing there's a lot of different approaches towards filling that gap between ineffective and inaccessible. On the right hand side is something we're building with Burke, as sort of that answer to the burglater of is there sort of a weather quality or another type of quality of data that we can collect that's scientifically useful, but doesn't require the same cost or level of expertise to operate. And then I wanted to share something that may not feel quite as relevant immediately but I think there's relevance from a science diplomacy standpoint and as a model for regional capacity in some of the more remote areas of the US is something we've done with us support so this is funded by the State Department and in partnership with Noah, we've supported the development of the Pacific Islands Ocean acidification Center, which is run in Suva BG by the Pacific Community and the University of the South Pacific with the Ocean Foundation support, and this is a regional training hub but also a repair center and an inventory of spare parts to help ensure that ocean acidification monitoring programs can be sustained over time, so that there's a place where locally people can go for support with equipment maintenance with research challenges with if there's new personnel that they can get training. And I think that there's a model this can pose for science diplomacy and working internationally but also for more remote regions tackling some of these problems to have that in region capacity to manage the scientific process from start to finish and one thing I emphasize is that this isn't citizen science this is professional level science that can be kind of managed in a more remote place more easily. Okay, so, so where does that leave us today I think that, again when I was thinking about from our perspective trying to help people establish research programs on ocean acidification carbonate chemistry. There's a lot of movement right now in this space around and you're going to hear after myself and Maria from people working in the marine CDR space but there's a lot of new initiatives and public and private investment and new technologies. The reality is we simply not only do we not have the tools I would say to understand ocean acidification and the carbonate system, you know, as it is, but with interventions we certainly don't have the tools, we need to monitor and verify proposed technologies and interventions at scales. And I think that's true within the United States but also internationally and there may be examples where actors may focus their work outside of the US, where ocean science capacity and regulars very frameworks or weaker, or perhaps in the more remote regions of the US where there's less of that capacity, perhaps. And then, you know, so there's also new legal frameworks like the BB&J, which require environmental impact assessments at larger scales and I think we've had to do before for potentially things like MCDR where we're going to need new scientific tools to assess these things. So one possible thing that NSF could do is sponsorably the development of some best practices related to these things and there are processes and projects underway. The Aspen Institute is working on a code of conduct for MCDR, you know, ocean vision so you'll hear from in a moment has done some work on this. I do think that there's a benefit to it being led by a bit more of an authoritative and unbiased actor, perhaps, on these very fast moving time sensitive issues. And so, you know, kind of summarizing a little bit of what I would see as some research investments that the US could make. I didn't talk too much about co design but I think with these highly complex issues where we don't necessarily have all the tools we need to understand them. And the reality is changing so fast, co design, yes it's hard, yes it's slow, but it's actually faster to do it from the beginning than to do a whole research project that has no value. And then go back and learn that so encouraging co design across academic disciplines and between academic and non-academic sectors from the very inception from the project development and funding development stage. I think strengthening international collaboration is something that could be done because these are global processes that need to be studied globally and there's, I think models of that being done that could be replicated again for science diplomacy reasons or for just pure science reasons. And then the rigorous testing of new methods I think we have the infrastructure required to test emerging tools against the best in class methods for especially for carbonate chemistry. And I think that NSF and other agencies are, like I said, well positioned to provide that authoritative and unbiased perspective on best practices and codes of conduct. So again, circling back to the discussion questions that I posed at the beginning are the tools that we have sufficient for the science fit questions we need to answer today. I think that the answer is probably no, but there's a lot of work that could be done to say, how can we bundle the various tools that we have available these combinations of high tech low tech, you know, in situ satellite to rapidly assess the ways that we can use tools and new ways and create training guides for that. And what role again can the US play in testing and defining new methodologies because things are moving very quickly around us when it comes to away and some of these other issues. And so we're going to need to move very quickly to keep up with it. So with that, I think I will stop sharing my screen and say thank you and hand it back to the moderator in case there's time for questions. Thank you, Alexis. I believe we'll wait till after the second talk and then and then have questions for both those speakers so. Next on the agenda is Maria. And so Maria, why don't you go ahead and start. Okay. Let's see. I'm you myself. Sorry, I'm working remotely and everything is everywhere all at once. Can everyone see. Yeah, looks good. Okay. I'm going to go ahead and hide. Meeting controls. Okay. Let's go ahead and go back to the beginning of this presentation because it was. All right, here we are at the beginning and I'm going to go ahead and hide everybody too. So I'm Maria Kavanaugh and I'm an assistant professor at Oregon State University and I've collected some thoughts from a number of colleagues kind of near and far. And so on behalf of all of us, thank you very much for for having us here. And today I'm going to be talking about away and have some primarily coastal ecosystems and with a focus on the Northern California current just simply because that's where I'm currently working. However, I'll try to generalize a little bit. And so over the next 12 minutes, you'll be hearing from us. Okay, so I'm just kind of going over a state of the science. We know that multiple climate stressors are and will affect the ocean ecosystems. And the stressors will and do overlap and they interact. Organisms respond to the stressors at multiple scales. So looking at the center figure there. And unfortunately, we rarely co-measure organismal responses and their stressors either at multiple scales and time and space or across multiple levels of biological complexity. And in the coastal region, this is especially challenging due to the increased dynamics and increased heterogeneity and poor performance of remote sensing and model algorithms. However, it's here where humans have both the most impact, but also the capacity to positively affect conservation and management outcomes. So I'm starting kind of in the open ocean. We know that roughly a quarter of the annual CO2 emissions absorbed from the atmosphere is absorbed by the ocean. This CO2 undergoes, you know, a series of chemical reactions that makes the sea water more acidic, reducing pH and causing other changes in chemistry, including the reduction in carbonate ions. And several trends across many open ocean time series have shown that as CO2 in the atmosphere rises, pH declines and CO2 increases in the ocean in pretty much locked step. However, coastal physics add to these trends and patterns, and these include seasonal wind driven up wellings such as what we have on our coast, which brings up water that's rich in nutrients, but also rich in dissolved in organic carbon. It also includes river and glacial inputs, which tend to be low enough, but can also absorb more CO2, especially if they're cold. And so for different reasons, local physics can result in patchiness of calcite and aragonite saturation and other components of the carbonate system. Local biology, including respiratory responses to eutrophication or nutrient loading can also exacerbate regional carbonates chemistry trends. And so many marine organisms are sensitive to these changes and we've learned over the past 15 or so years that particularly coral shellfish and other marine life that make their skeletons and shells from calcium carbonate are sensitive. However, the OA research community is moving beyond the single species, the single responses, and has begun to look at the extended effects through communities and food webs. And so this could include investigating the role that calcifying organisms have as a foundation or habitat forming species such as that this Coraline algae shown on the bottom left, or modeling more complex responses. For example, diatoms are predicted to benefit from increased CO2. However, this study based in off of Antarctica found that with natural enrichment, diatoms were indeed enhanced. However, the composition and size of the phytoplankton, the diatoms shifted. And this resulted in changes in nutrient use efficiency and reduced export efficiency. So there's sometimes some unintended outcomes or not unintended, but rather unexpected outcomes. So other studies focused on complex meta-analyses to look at some of these outcomes to determine whether the type of response to CO2, whether it's consumption, growth, metabolic rate, movement, or survival, and whether it's to CO2 warming and low oxygen, whether that's predictable across responses and across a broad array of species. And so we are learning much, but we also have much to learn. So we talk a lot about non-stationary responses and non-stationary relationships indicate that the effects of environmental conditions such as temperature and OA on organismal patterns and processes can vary over time and intensity and or direction. They suggest sometimes that there is some sort of physiological threshold or tipping point that has been reached. This makes predictions and projections really tricky, but it is really important to understand these for mitigating OA or in understanding how OA signals are communicated to higher trophic levels. While best practices documents provide guidance on how to disentangle components for the carbonate system, it's also important to quantify OA in a multiple stressor context. As OA may exacerbate or induce additional stressors, including HABS, which I'll talk about in a minute. Multiple stressors can interact and be different in their dominance over time and over different life history stages, as is shown in this work by Berger, who mapped out present and future stressors for a dungeonous crab, which occupy a benthic or pelagic habitat at different life stages. And even if concurrent stressors do not have interactive effects at a given life stage, multiple stressor effects could emerge across development. Such carryover effects have not been studied much in the context of OA and could be a fruitful avenue of research. And finally, while regional biogeochemical models embedded in Earth's system models are ideal for forecasting, projection, and attribution of drivers to assess for mitigation and adaptation pathways, this will require access to high performance computing and potentially coordinated nested approaches across many institutions. All right, so to phytoplankton, phytoplankton community composition is also changing in response to multiple stressors. Harmful algal blooms occur when species rapidly grow and accumulate, and these species can be noxious, resulting in hypoxia when the blooms subsequently decay, or they can produce toxins that directly or indirectly through bioaccumulation can affect human health. There's much attention on the role of OA in promoting HAB blooms and hypoxia. HAB blooms, sorry, HAB blooms and toxicity, and this is in terms of CO2 fertilization effects, changing the availability of dissolved nutrients, and changing the behavior of grazers. This is especially true for the case of pseudonitia, which can form large blooms off of our coast of North America and the West Coast, such as what happened in response to a persistent marine heat wave in 2015. Some species, but not all, of pseudonitia produce demoic acid, which is a neurotoxin that can bioaccumulate and shellfish and making them unsafe to eat. In 2015, toxin levels were incredibly high, and they made their way through the food web to impact the number of fisheries and marine mammals. This led to long delays in the dungeness crab fisheries and a complete shutdown of the razor clam harvest, and resulting in financial and cultural devastation throughout the West Coast fishing and tribal communities. Newspapers described that surge in food bank usage in their local communities, suggesting a lack of revenue was leading to food insecurity. Now, coherent management relies heavily on monitoring with some course predictive capacity based on circulation based modeling, but little ecosystem modeling. And there's still huge uncertainty in the science. So this case study and general uncertainties have resulted in a suite of research priorities that have been set through a couple of recent synthesis by the HAB community across four broad themes. Bloom ecology and toxin production is one, a second one is toxic pathways and effects. A third is food webs, understanding the effects of HABs on food webs and ecosystems. And this includes novel interactions such as what is currently happening and being investigated off the coast of Oregon and as a potential cause for recent muscle die offs. Finally, we, one priority is understanding the public health and socioeconomic impacts. Now, while management current management mitigates most of the acute health issues as climate continues it will be important to understand the cumulative effects of low levels toxins through time. So, and also an additional priority or perhaps even an umbrella would be to integrate this social understanding into HABs within a multi stressor framework. So continued development of both high tech tools such as imaging systems and environmental DNA and training the next generation to use them will be incredibly critical. So, in terms of science support there are no easy solutions to the issues of context dependency and non stationary ecosystems. So, I would encourage us to think about our strategies for addressing these and whether we invest in a means to conduct more experiments, develop flexible and modular yet mechanistic model frameworks and or bolster our observational capacity. And I would encourage us to think about how we do that in a way that promotes access, equity and inclusion. I would also encourage us to think about what infrastructure and history exists and support activities that add capacity to the science and to the community that address ecological complexity and promote inter calibration activities with similar programs in other regions. So one example that I'm showing you here is the Newport line, whose observations of ocean conditions and lower traffic level patterns have provided a platform for collaborative science instrument testing and fisheries management and indicator development for the last 26 plus years. I would suggest that we strengthen partnerships between networks. For example, NSF long term ecological research sites can exchange mechanistic insights for the observations and technology associated with the marine biodiversity observation network. The have and global ocean acidification observation networks and oh I this could illuminate things such as how the composition, the magnitude and location and fate of blooms changes with changing circulation, chemistry and temperature that contribute to have and fisheries research, but also basic science on the carbon cycle and climate feedback. So one path for interagency cooperation is through the national ocean partnership program, which is highlighted as a means to accomplish broad transdisciplinary goals in the ocean climate action plan. And last to end, I'd like to state that the understanding the effect of multiple climate stressors on marine ecosystems will require a long term transdisciplinary vision that may require reframing our traditional structures of teaching research and collaboration. We'll need to build a climate ready workforce to address have a issues going into the future. And one great example of this is the next gen program from NETA, which provides student scholarship support, meaningful paid internships fellowships and job opportunity match ups, including in the federal sector. We'll need to train and practice transdisciplinary research. And one example I'm showing you here is from OS use NSF research traineeship in risk and uncertainty. And here students teamed upon problems under to including to understand the multiple stressor effects on Pacific cod. They brought together machine learning, geo visual methods, bio energetic modeling and vulnerability assessments, all together in a collaborative framework. And then finally, I'd like to encourage us all to as as we're going forward over the next decade to form meaningful and sustainable community collaborations. And one such pathway is something such as collaborative fisheries research, such as the pilot harmful angle bloom observatory shown below. And here commercial fishermen partnered with scientists off the Oregon coast to contribute to observations and modeling to promote and to create a have early warning system. And with that, I'll go ahead and close and say thank you. And I guess we'll start to collect questions. Okay, well, thank you. Thank you very much. Got some questions already and I'll be some more coming along. Thank you. The first one I have is, is from tuba and it's endorsed by several others it says you give, I think it's probably for Alexis you gave examples of technology to measure and assess away habs. Is there technology being developed to mitigate. Yes, and so technology what is technology, I think is a little bit of things so mitigating away and habs I think for away the first thing is carbon emissions so obviously reducing carbon emissions habs have multiple, I think things that cause them so there's definitely some cultural policies that are being looked at that also are contributing to ocean acidification and habs from technology standpoint. I think it's very early. There's some low tech, like multi trophic culture and things like that has that has been studied. And then things that you're going to hear about that I say are still very experimental on the marine carbon dioxide removal, some of which are focused more on removing atmospheric carbon. Some aren't focused on reducing ocean acidification. Some are more focused on reducing ocean acidification, but I will caution that they're very in the early phases. And I think we do not get understand those. I personally think that the number one thing to emphasize is carbon emission reduction, and then local lands management and agricultural and local sources of carbon. So I don't know if that would be technology, so much as policy and but I do think that the technology role is helping with those decisions right and those trade offs of understanding what's the attribution of local sources of carbon or other things to the coastal system and where can local policy interventions make a difference to the technology for the mitigation is understanding the attribution to inform when a pot local policy intervention will make a difference versus when it's a global process. Maria, you probably can weigh in on that as well. I'm less familiar with the mitigation side of things. It's more of the adaptation. And it's I'm in particularly with with halves part of that has to do with understanding early. Whether a bloom is in development and where it might actually end up in a couple of weeks. And so that's where some of the circulation modeling comes in. However, what doesn't happen concurrently and there's a little less organization on in terms of the states is actually looking in the shellfish. I mean, having sufficient monitoring in the shellfish to see if they're actually consuming those those hats. And that's that's kind of where they want to go because they don't want to shut down and shutting down means the loss of revenue. It means the loss of cultural activities. And quite often they're not accumulating at that at that time, even if it's it's hot right above them. So I think thinking about multi-scaled observational systems that are somewhat nimble would be helpful. I'm not entirely sure how the policy and the management framework that goes from perhaps a tribe or a municipality to the state to a region to say NOAA, how that might maintain being nimble. Yeah, I think you're right. I think it's more adaptation that is really has been the focus, like even on the OASide, the adaptation in the oyster hatchery has been on monitoring and then making adjustments to the tanks or the intake water. That's not mitigation of the problem, whereas reducing the influx of carbon into the system that would be a mitigation process, which I think there have been some efforts in Washington state. To look at that and look at attributes of the agricultural runoff versus atmospheric signal versus upwelling signal of OAS to understand how, but there's other people who can answer that better than me. Hey, I think Allison has a question. I wasn't expecting to read it. Thanks, ladies. My question is a little bit related to tubas, but do you believe that NSF is investing enough in terms of funding for innovative tools and defining new methodologies for assessing OAS and HABS? I think that there can be more. I think that there's some great examples of investments in concrete, like individual pieces of technology. What I'm really interested in is how do we match the different existing pieces of technology in new ways to answer questions, because I think that that's not necessarily something that I'm seeing a lot of people working, or at least it's not may talked about as much maybe because it's not funded and it's not something people get excited about to apply to. Okay, how can you utilize existing platforms or new platforms to answer some of these really tricky questions? So the multimodal, the bringing in satellite and low tech and high tech and all these things. And really incentivizing people to do the hard work of integrating across platforms and creating methodologies for that because these questions are coming really fast and we don't have time necessarily to design new technologies. But there's a lot of good technology out there that possibly could be adapted. And then that will help us better understand where the gaps really do need something new, like a new sensor or a whole new system. Maria, what do you think? Yeah, no, I completely agree. And I think while high tech sensors are always going to be useful and useful in terms of cutting edge science, I think sometimes we need that some funding to go to the kind of cross calibration communities in where you're taking the $2 million sensor and you're comparing it to the $10,000 sensor that the community is going to have in their hands. Because sometimes it's a matter of not necessarily the precision of the instruments but being able to see the processes where the process is happening. And also to empower and make those relationships with those local communities who will be your eyes, your eyes on the water. I mean, with cooperative fisheries research, for example, we have to make concessions about the quality of a given single measurement, but we're getting 300 instead of none during things being shut down. We're also having to kind of think a little more creatively about how we frame scientific questions because we may not be able to, I mean, we can't create a perfect sample grid. They're going to go where the fish are. And so, but at the same time, we can use that, for example, to help inform, to validate satellites that will fill in those gaps or to validate instruments or novel instruments. So I think having those kind of partnerships with, as you're doing those high tech, low tech intercalibration, it's going to, I think it'll be fruitful. I completely agree. And I think it's not really being done at large scales. I think it's being done in small scales on people's volunteer time. Or by groups like mine that are kind of nonprofits kind of trying to do these things with the pieces of funding that we get, but doing it at a large scale with a group of experts that have the expertise to do those, those assessments and to say, okay, it looks like, you know, if you get 100 samples of this quality, you're nearing the ability to make a judgment on this. And to be able to write those sort of decision trees of, okay, you could use this type of data to answer this type of question. If you have this type of calibration process, I think that would be very useful. Okay, Jason, you have a question. Yeah, thank you both for the talk. This is Jason link. You partly touched on the answer to the first question, but a lot of the Habs community away community focus on observations and measurements and fancy sensors and all this stuff that I couldn't even begin to name that you showed. So thank you. What I'm wondering is what about predictions what about forecasts and you alluded to mechanistic understanding and that's great. But are we there yet or if not, what do we need to get to Habs forecast and not just in Lake Erie or western Florida around the country. Thank you. Right. And, and I think in the near term I think like the short term for forecasts are where I think some of the rapid machine learning methods, even though we may lose some of that topology some of those like physical mechanisms and we can look at, you know, where we can actually investigate the shapes of the data. We may lose some of that, but we get we get an answer. And then I think what needs to happen is that there needs to be some sort of retrospective analysis to look at how those rapid models and those rapid machine learning models, how they may spatially or morally actually reproduce faithfully some of the biophysical interactions that we may be expecting. And they may also, depending on the type of model they may actually produce new hypotheses and illuminate new interactions that we that may have been obscured in the in the past. So, so I'm not sure if that if that answers are we there yet I think certainly in the West Coast, I don't think we're there in terms of something mechanistic. But I mean, and in part because we may be able to tell when a, you know, as when pseudonymia is dividing or when it even maybe when it's toxic. But we won't be able to necessarily couple, couple that with when the razor clam are feeding or and because that might be up to local, you know, something very, very local and constrained and so I think there's still always going to be a need for observations. But I think there are ways that we can be much more savvy with our model inter comparisons than we maybe happen. Yeah, on the way front I'm definitely not the right person that answer that I think like Sam said lucky who's one of the people that Maria had on her would be the right person to answer that I do think that having heard Sam, give a lot of talks and is that the only places where I think we're approaching that modeling and predictive capacity is where there's a lot of data. I think that gives you in a little bit of information about the importance of the monitoring. We're only able to produce any sort of reliable predictions in the places where we have the best, honestly data in the world, which is in the Pacific Northwest. And I think we have to recognize that now are there other ways to collect that data that might be cheaper, maybe, right. And so we don't necessarily have to do it the same way that it's been done here. But that's just the reality of I think being able to produce those types of models but I think Sam said lucky is the best person to answer that question. Okay, Peter, you have a question I think could you unmute and ask a question. Gladly. Thank you all that was fantastic. So, I'm going to sort of stumble through this so bear with me if it's not very clear but so often our approach to understanding or meeting kind of community needs is is for NSF to reach out to academics and try and make informed decisions but but sometimes I feel like we in the Academy often, you know, fail to inverse invert that model. And what I'm getting at is, do you have suggestions for us and for the NSF on, or for all of us on how communities might best represent their local needs you know in some cases, they at some communities may be more financially replete than others but they have a particular different set of concerns that others and there's a lot of granularity that I think we lose. And I'm not, I'm not sure how to how to collate and present that information I'm curious if you two have any ideas. Thank you. So, in terms of information transfer, at least with the fishing community on our local Sea Grant and Oregon Department of Fisheries and Wildlife, they've set up a science and Fisher, well it used to be called a science and fisherman exchange or safe and so it was kind of a closed door situation where the scientists and fishermen would have a have a dialogue. And that's a space I think where people felt safe enough to articulate their needs without necessarily having a shared vocabulary, because that I think is a that's that's a stumbling point and in terms of what we would call like spatial variability what that means and all yeah and so And I think that's that's really useful. It does require someone embedded in the community and who maybe is that part of a boundary organization or something that that can talk about talks and can get the people in the room who need to be in the room and because the flip side is is that it's part of our job right to go to conferences and go and listen and dialogue. It's part of it's not part of a fisherman's money making right it's it's detracting them from from their work. And so, getting them is really, really a service and so part of this. I mean, and part of the things we've discussed is like, how do we compensate people for their participation for their time and for their honest feedback and so I mean that that might be one. I completely agree on the boundary organization I was just gonna say finding the boundary organizations which is like kind of a buzzword but like basically the people who are already in those spaces who already have the trust it can take 1020 years to build that level of trust and the grant we already have that amazing institution in all of our states so I think that's an obvious one and investing in them, but working with them to identify some of those needs, because I think they're in the best position to answer some of those questions. Because it's really hard and yeah I think a lot of people are getting hit there as there's more emphasis on co design and equity, a lot of these communities are getting more more requests for their time, which is like the opposite of what I think is useful to then figuring out how to do that in a good way and then again I think the boundary organizations are key, but also they need staff time because they're also getting overwhelmed. And so, you know, if it's the kind of thing where the academies and NSF want more of that engagement then figuring out if there's a strategy for that for a national strategy for boundary organizations to get this type of feedback, you know, working with Sea Grant seeing how they're thinking about this from a national level, working in the interagency working groups to figure out how they're considering that there's more and more agencies are putting co design requirements in their funding proposals. So, you know, oftentimes that includes partnering with the Sea Grant office in order to deliver on it, for example. I think Kersi has a question and maybe if you pass the mic over. Thanks. So, turn my camera on for the people that are on the call. So, as part of the infrastructure act that was passed that trillionaire act that was passed back in 2021. It's an amendment to that out of continental shop lands act where they empowered boom and best see that oversight over carbon sequestration in the ocean. And what's transpired from that is a number of terrestrial carbon sequestration companies pursuing and exploring the concept of deep sea carbon sequestration. And so, I recognize that the research on impacts is not there yet in terms of like having a well having that well fleshed out. But if you could speak to the concerns around that, and how we balance that with the concerns for addressing climate change, especially with respect to what I've recognized from engaging with some of these groups is that there's still a misperception about interconnection of the ocean. And so there's a kind of view that, oh, we dump it in the deep sea and it's kind of gone goodbye. Thanks. You may have spelled it out there, right? We don't understand. And that's the problem. I think from the ocean foundations perspective, we are very vocal that we think it's irresponsible to move forward without understanding the risks. And we think that there need to be precautions in place to particularly protect communities that might be taken advantage of as far as experimentation of methods in their backyards, but also recognizing that the ocean is connected. And so any effects are effects for everybody. We don't understand it. And there's a lack of understanding where a lot of these decisions are being made in policy spaces that are not ocean science spaces and financial spaces that are not ocean science spaces. And so I think there's a need for the ocean science community to be vocal about what we don't understand. And about what we know and what we don't know, because no one else is going to be vocal about it if we're not vocal about it. And so, yeah, what are the implications for, you know, sinking biomass in the deep sea? Well, what does it mean to not know? What are the risks associated with not knowing? You know, does there need to be a statement about that? Who needs to make that statement? These conversations are happening. Again, you're going to hear from a few people I think next, though I think it's important to note kind of where they're coming from in terms of industry or advocating for certain solutions. Not, you know, there's a lot of science driven work that's happening. That's very good. I think there's also a lot of financially driven work that's happening right now. And I think that there's a need for more rigorous statements from the scientific community, in my opinion, and I think in the ocean foundations opinion is we just don't know. That's a great transition to the next speakers. And so I wanted to thank both of you for your contributions. Very interesting. Research priorities for. OCE Marine CDR. And we have two speakers. My agenda shows Julie first or David second, but whatever whoever wants to go first, go for it. Hi everyone, this is Dave Coak here. I think Julie and I talked about it and decided that I would go first. So I'll get my screen share set up. Thanks, David. Okay. No problem. I actually preface coach coach here prerogative with the thought about this next session because I think what Alexis said was really important. You know, these conversations around marine carbon dioxide removal are happening. Companies are being formed. And so my personal thinking is that we need to be at the table. The science needs to be at the table when these conversations are being had. Because it is important, like Lex has said, that the development of any kind of mitigation technologies happen with the understanding of the science right there from the get go. Right. And so I feel strongly that that that, you know, just, we, we can't just, we can't detach ourselves from that. So that's one of the reasons why I'm excited about this next panel, because I think I'm hoping you all will give us ideas as to how we engage in a productive way with this work. So with that, I'll hand it over to you, David, and then back to back to you. Great. Well, thank you very much. I just want to check really quickly to see that my slides are projecting. Can somebody give me a thumbs up if that's the case. Yeah, we see. Okay, great. Thank you very much for this opportunity, everyone. My name is Dave. I'm the chief scientist of ocean visions. And if you don't know ocean visions is a US based field building organization, we work to scale durable and equitable climate solutions and we've done a lot of work around accelerating ocean based carbon dioxide removal. We work as a networked organization, which means that we're a relatively small nonprofit, but we work with a whole host of institutions that have, excuse me, formal partnerships with ocean visions, they're shown on this graphic. And together we work to develop this agenda around identifying testing evaluating and ultimately scaling viable, equitable, durable climate solutions. So this is a really science based effort. And I actually want to speak to the comment about, I want to speak to the comment from the, from the last speaker, ocean visions organized a an open letter from the scientific community that now has over 400 signatories. These are PhDs in earth, ocean and climate science from around the world that is calling for responsible research for marine CDR technologies. And so if you're concerned about the environmental impacts of MCDR, you should want the science done. And if you think it's a good idea, you should want the science done. And if you're agnostic, you should want the science done because it's the science that will provide the evidence based perspectives that can really help cool the temperature on the conversation around MCDR and help us drive towards the most viable solutions the most quickly. So we're really at this place around CDR because of our relatively slow rated decarbonization and so I want to take a step back here, just for those of you who don't know, and, and give you just a very bit of brief history which is that in 2019 or 2019, the IPCC released a special report on one and a half degree of warming and this is what really set the stage and set the scientific basis for understanding and motivating the need for gigaton scale carbon dioxide removal. There's a very powerful statement in this that in order to limit warming to one and a half degrees with limited or no overshoot, we need something like 100 to 1000 gigatons of carbon dioxide removal this century. That partly depends on our rate of decarbonization and how well we can abate the hard to abate sectors. It also, yeah, so every year that we continue to not decrease our emissions means that we're going to push ourselves further and further up this ladder. So the question amongst the scientific community, the question is not do we need carbon dioxide removal, it is where does that carbon dioxide removal come from. And if you want to think about carbon dioxide removal as a means to rebalance the global carbon cycle you have to understand the various reservoirs of the global carbon cycle so this diagram is showing the various parts of the global of the global carbon cycle. The size of the dot here is proportional to the size of the reservoir and then the yellow is showing the amount of anthropogenic carbon that has been emitted and where that carbon has ended up. And what I want to drive home to you on this plot here is that the deep ocean is by far and away the largest reservoir of carbon on the planet. It's so large that if you took all of the anthropogenic carbon that's ended up in our atmosphere and somehow magically put it in the ocean, it would increase ocean deep ocean dissolved in organic carbon levels by less than 1%. And so when we think about the global carbon cycle, it's really impossible to escape the idea that we could reach gigaton scale carbon dioxide removal without the oceans playing some substantial role. Now there's a whole host of technologies that can contribute to providing ocean based carbon dioxide removal. They're put together on this infographic or the schematic here and they basically span biological and chemical pathways and fundamentally all of them are about either speeding up photosynthesis in the ocean so that the oceans can capture more carbon via photosynthesis and then sequester that carbon in the deep ocean. Or they are chemical approaches that enhance the alkalinity of the ocean in order to allow the oceans to store more carbon in the form of dissolved in organic carbon or more specifically bicarbonate. There's lots of minor details and nuances here but for the principles of this talk I think that's what's most relevant. Ocean Visions has been doing a lot of work for the last three or four years to try to accelerate the research and development and notice I'm not saying the word deployment but the research and development of marine CDR technologies. We've built this suite of online living digital road maps. You can find them on our website ocean visions.org. And they are broken down by specific technologies for ocean based carbon dioxide removal as well as some of the cross cutting social and governance challenges and opportunities that are needed to create the enabling environment for responsible research and development. Each of those maps are broken down into three domains the state of the technology that which goes into the scalability that what we know about the state of the technology currently the environmental risks the environmental co benefits the social risks and the social co benefits. Once we know the state of the technology it helps inform the development gaps and needs which are the second domain this is what we don't know and then the first order priorities which are attractable actions that can be taken to address those knowledge gaps and therefore improve the state of the technology. So, like I said, these have existed on our website for some time now. And one of the things that we have been working on very recently is bringing these into a cohesive narrative framework. And so this is a white paper that we've just produced. It is in press and should be coming out very shortly, like in the next week or two. And the basic idea here behind this white paper is to ask the question, what needs to be done between now and 2030 in order to do all of the science and engineering that's necessary to figure out whether or not these are viable climate solutions and so by viable. We are these effective at drawing down carbon dioxide removal or drawing down carbon dioxide with acceptable environmental and social impacts and I think we shouldn't kid ourselves that if we're going to have gigaton scale carbon removal that we would be able to do that in a way that has no environmental impacts or no social impact. So this is really a question about characterizing those and and society coming together and trying to determine what are acceptable and of course all of that needs to be done on a comparative risk assessment for other mitigation technologies as well as the substantial risks of doing nothing. Okay, so this white paper basically has three pillars they're all interconnected. There's science and engineering. There's policy and then their scalability and for the purposes of this briefing for the committee and thinking about NSF's gender for the next 10 years. I really want to focus on this science and engineering pillar and the science and engineering pillar has a number of needs. I think the number one area that must be supported by governments around the world the US as well as others are a series of controlled field trials for the various MCR technologies that allow you to make data driven statements about efficacy and impacts. There's a number of precursors that need to happen in order to support those controlled field trials. I think we need collaborative design processes to bring people together to agree on how those field trials should be designed and executed. And we also need the spaces in order to do them. So one concept that's been advanced are pre permitted test beds. These are sites in the ocean that have one central regulatory authority and so everybody applies and comes in under a sub permit. And it basically allows you to accelerate the pace of responsible research. There are still outstanding laboratory and mesocosm science questions. I think a really good model here was NSF OAP or ocean acidification program. I think that there could be. There's just very, very strong analogies and so seeing a very similar program set up for addressing laboratory and mesocosm science questions is really key. There are a number of needs around monitoring, reporting and verification related technologies. We need advanced capabilities and sensors. We need even higher resolution models and the associated advances in compute that allow those models to become feasible. And then we need advances in data assimilation capabilities that allow us to integrate models and data to produce the most accurate forecasts and estimates. And finally, I think we need a really coordinated social science research program. I think NSF should get involved in this OCE should be involved in this and they one possibility would be to collaborate with a directorate for social behavior on economic sciences. I think as much as possible, the social science research should be coordinated with field trials so that we're assessing, we're assessing key social science hypotheses in the context of R&D that is actually happening. And it should be coordinated with other governments around the world because social context can differ and different communities and cultures will have very different risk perceptions. So with that, I'd just like to say thank you very much for the opportunity to speak and I'm really looking forward to hearing Julie speak and I'd be happy to answer any other questions that I can. Thanks. Okay, I think we'll hold the questions until the second speaker goes and so Julie, you're on. Julie, are you ready for your presentation? Yeah. Yeah, I'm trying to share my screen. Let me see. I think I'm really close. We see it. Is it working? Yep, it is. Okay, super. Thanks. Thanks everyone. Some of you all know me from my past life as an academic. I left that five years ago to help lead a startup in the climate risk space and now I'm on the investing side. And too, but I just wanted to say appreciation for your comments about John Allen. As you know, I was one of his few students that he trained. I did my PhD with you. And I'm so sorry to have missed the tribute. I, my flight got delayed and I was flying back from Iceland at the time of the tribute that sending sending my thoughts to you all in the community on such a big laugh. On my flight, I did get to see the arm under installation flew right off the tip of Greenland and really happy to see that terrific work spotlighted propeller is a venture capital funds that was set up in order to accelerate ocean climate solutions. We have a founding core partnership with Woods Hole Oceanographic. So we look at technologies within Woods Hole and ocean science more generally to help solve the climate crisis that we find ourselves in. The, the other activities in our fund relate to investing out in the market were very, very early. So pre seed and seed stage. And we really want to catalyze the acceleration of climate solutions. And one of the motivations for setting up the fund was actually the recognition of the transformations happening in the new blue economy. So these are along the different dimensions you see over on the right. Brick's been read who is who's known to many of us has articulated so clearly. What is entailed in the new blue economy and there's a range of different estimates on the scale of the new blue economy, but it's on the order of several hundred billion dollar market value in the transformations that need to happen in the in the very, very near term. Those, those areas show up in our themes for our investments. You can see in the turquoise highlighted under ocean carbon, ocean organics and ocean industrials are the areas that overlap with the new blue economy. There are other areas that fall somewhat outside of the new blue economy. And that includes ocean CDR and the associated monitoring reporting and verification that would underpin ocean CDR as well as the blue carbon areas seen above and the turquoise. The, as David mentioned, the, the recent IPCC report really emphasizes the role that negative emissions will play and meeting the net net zero goals. And those different approaches. When you look across what what it will take to do those negative emissions represent waste management on a really massive scale. So by some estimates, you can see low and high estimates there by 2030, these just in and of themselves are hundreds of billions of dollars market growing to potentially over a trillion dollars by 2050. So this is why, you know, growing markets, large markets are accelerating interest, not just from the investing side, but, but from startups as well. So some of the areas that David mentioned are listed in a slightly different way here in this diagram that shows on the y axis the technological readiness for deployment at scale for ocean based CDR. And on the x axis is the advancement potential for the new research development and demonstration. So you can see that the, a lot of those approaches that David mentioned are captured here and then they were, they require as an as a prior national science report mentioned a significant investment in order to research them and understand their potential to scale and to to progress along the TRL and a research and development framework. I didn't want to leave this topic before pointing out that the current cost of carbon dioxide removal is quite high. And this is a report from CDR dot FYI, which shows some of the companies and approaches that are now being offered through advanced market commitments. There's a fund called frontier fund that fund that's over a billion dollars from some mainly tech companies who have come together and said that by making advanced market commitments they can help accelerate the technology, lower the cost and lead to greater scalability of some of these approaches and marine CDR is representing a growing part of the portfolio you see it here is that in the purple and the dark green the macro algae and some of the enhanced weathering approaches like project Vesta. But as you can see these the, the price of these is is per ton is still extremely high and these subsidies from frontier fund, for instance, are really designed to grow these capabilities and make them more affordable. So just a few closing thoughts on research needs and marine CDR. As I mentioned, I mean, there's a huge opportunity to on the technology side, really produce sufficient the gains and carbon removal, so leading to lower power more scalable technologies, advancing the TRL and the RD and D aspects of existing technologies is a really important area. Also innovation. I mean, as we see these waves of innovation on CDR and particularly ocean CDR, there's more and more hybrid approaches that really work with the natural systems that are coming in to the in the country defy easy categorization but are some of the more promising approaches that are emerging on the science side. It's come up before during today the importance of widespread carbonate chemistry monitoring. It was also mentioned in relation to monitoring the importance of understanding that can measuring the changes in the stability of ocean carbon sinks. For instance, that might be changed through shifts and amok or the the weakening of a mock ecosystem impacts and also research to inform and or advanced and or advanced regulatory frameworks for how to approach the, the use of these technologies. We had really a beautiful phrase climate scale observatory and I want to echo that here and really emphasize the importance of novel adaptive multi use platforms with low cost sensors and that can build on the capabilities and really hopefully expand capabilities like BCG Argo and CO2 direct flux measurements that were shown so nicely through on the LLI platform. Also, I think it's really important to improve near term high resolution decadal modeling, particularly with the advent of new generative aid eye tools like the transformer architectures and diffusion models for uncertainty that were mentioned during the the AI session earlier. I wanted to just draw attention to a few things that was part of a Clive our AC transition study group. I came out with the report in August and the whole approach of our study was really looking at strategic use cases driven by climate extremes and impacts and to really examine the ways that the measurements and locations of measurements can really tie into societal needs and the importance of identifying climate extremes and there's a lot of strategies in there for funding agencies who are looking at that are seen in our face. A couple of themes I want to draw out. One is on workforce preparation we're seeing a really strong need for people trained in data science it's these are our positions that are are growing over time. We're seeing climate tech and just general climate science and the importance of diversity here was also brought out by by Luan. In terms of orchestration and coordination. I'm going to talk about program centers and initiatives and one of the things that I was part of a national commu sciences study group on sustaining ocean observations. I'm going to go on to the concept of a collective impact organization that can act as a backbone and create the interstitial connectivity to a range of different activities that could span industry startups national labs and be very nimble and agile and yet create transcendent outcomes that go beyond the capabilities of any one of those entities. So an example that's come up through NSF and a collaboration with NOAA are these industry consortia and this new one amounts between NSF and NOAA on climate risk. As it relates to the insurance sector is the latest one there's also been one on wildfire risk in the climate space so I wanted to draw your attention to that. I think this is almost my last slide. There was just announced a dear colleague letter from NSF on interest in CO2 removal and solar radiation management strategies. And this emphasizes that the the it's essentially a drawing attention to these topics. And there isn't a dedicated funding line or a particular call for proposals that's associated with it. And yet it covers a lot of different groups within NSF. And so I think there's an opportunity here to reach also over to the tip directorate or some sort of focused research organization or potential programs that really center CDR research development and demonstration. And also climate scale modeling and observing that could really get at some of the fundamental questions we have about tipping points in the earth system. Thanks for your attention. I'd be really happy to answer questions and engage in a discussion. Thank you. Excellent. Thank you to both you David and Julie. I think you all hit hit this just the right level for us. I see some questions coming in on Slido and I see multiple questions that are really, you know, variations of the same question. So I'll call on on one of those who asked the question. It's sort of maybe that belongs into this family. Cursey, do you want to go ahead and ask your question. And I saw that a number of people also kind of ask similar questions. I'll try to incorporate their questions into my overall question, which is really related to recognizing that the industry related to carbon dioxide removal as far ahead of the science. And I know, David, you were talking about this timeline of the amount of research we need to do. And I think it was 2030 by 2030 or something to that effect if I'm remembering that correctly. And it occurred to me that right now. There are organizations that are applying for permits and getting them to do pilot projects for marine carbon sequestration. So how do we as a science community kind of balance the need to both do work for an industry that's moving faster than we have answers for in terms of impacts. And more importantly, how do we avoid being in this position in the future for other potential. So what are the potential impacts that humans might cause from innovative desires to utilize the ocean and various capacities. I'll maybe start. Yeah, why don't you start. Yeah, so your second question is very difficult to answer but let's start with the first because that's tractable and that's actually happening on the ground. I want to add that I think part of the answer is in more engagement with private industry from the science community but go ahead. Well, I think you foreshadowed my response which is that I think you're right to characterize a lot of activity in the mcdr industry that very nascent but growing mcdr industry. And I think the academic academic community is quite frankly at a crossroads where you can collectively as a body decide that you want to lean in to engagement with mcdr the private sector or you can decide that you don't want to that you want to lean out. And my recommendation is to lean in because it's only by leaning in that you can help steer the ship and help make sure that those pilot projects are being done to the best available science. I want to stress that many of the pilot projects that are being done and led by these companies are founded on technologies and often have members of their founding team that are among some of the best scientists and engineers that we have in this country. And so I'm thinking about examples like Ed Carbon that was founded on technology developed by Dr Matt Eisenman as an example. And there's many other examples like that. And so I think a challenge for the whole community right now is to think about and get comfortable with the idea of new models of partnership and collaboration that are going to cross traditional sectors so that means that academics industry and national labs. And even some nonprofits are all going to be working together and I think the question to ask is not necessarily who's leading these pilot projects, but how are they being done. Are they transparent. Are they equitable. Are they inclusive. Have they engaged local communities. Are they collecting data in a manner that facilitates outside scrutiny and transparency and if the answers to those questions are yes, then I think those are really strong indicators to lean in. Thank you, David. Julie, do you have thoughts on this. No, I have nothing more to add. Yeah, David, you did do a really good job. I guess I will just add a little bit of a story that I might have shared with some of you but certainly not all of you, a much smaller example of how this can play out. And it's sort of the experience that we've had with the wave energy industry here in Oregon where about 15 years ago, all of a sudden wave energy wanted to come to Oregon they realize we have big waves, big waves have big energy. And I think what we did choose to do with the help of organizations like Sea Grant was to lean in and and essentially one of the things that happened is that the development of the industry did slow down. And it did, because now we were, we had sort of gotten into this mode of, you know, community engagement and making sure that we take environmental impacts into account one step at a time testing first in the lab then maybe in the bay. And right now we're building it the first grid connected wave energy testing facility for the United States and so off the coast of Oregon. Again, with the idea that this is a pre permitted place, but it also comes with very strict environmental monitoring requirements so the moment something starts going in the wrong direction, we can yank this thing and you know stop that trial. And so David I'm hearing you say some of these words now the example that I gave is, you know, the potential environmental impacts are tiny compared to what we're talking about here in terms of marine CDR. But nonetheless it feels like that's the kind of approach David that you're talking about and we do have an example of that being successful in pacing the industry development so that we can do things responsibly. I'm going to respond to that very quickly with one point which is that my comments were about an accelerated research and development agenda, which includes starting with small scale field trials and I think we often have this. mindset where we think these small scale field trials won't matter at all for climate benefits but that they're likely to have the same kinds of environmental impacts as giga pun scale deployment operations. And I think we that's not true we have to be, we have to really consider that the size of the field trial is likely to dictate both any carbon sequestration benefit as well as any environmental impacts and so if we do this in a really stage gated way, we can learn and the size of these field trials can increase and we can mitigate environmental risks associated with those field trials but I think it's, it's not right to assume that very small scale field trials that may or may not commence in the next few years would have lasting irreversible substantial impacts to marine ecosystems. There's no historical precedent for that. Yeah, very good. Thank you, David. Now the one thing that you did say though is that you said you were imagining it accelerated program. So I guess maybe another way to think about the question that the question that Casey asked, thank you, is to think about, you know, if the industry is moving so fast maybe another way to approach it is to have our research accelerate. Is that sort of where you were going with that David or no. Yeah, exactly that's what I was talking about laying out that agenda, and how NSF can be really supportive of this again I think the biggest thing that NSF can be doing is really carefully well designed field trials that are designed to give information at relevant spatial and temporal scales about the efficacy and impacts of any one of these technologies. There's a whole host of ancillary other science needs that I laid out you know MRV related needs, fundamental laboratory and mesocosm science needs and social science research, but that's the single biggest ticket item and I just want to put some in here. We, we did a budgeting exercise where we designed what a field trial would look like if you did research for sinking seaweed, and you know large scale field trials are likely to run into the 10s to hundreds of millions of dollars. I don't know if it's on the duration and the size and the scope but I just, I really want to put that out there so that you know you're all when you're all thinking about NSF priorities for the next decade you have it really clear in your mind how expensive the science is going to be and that's why I think it's really critical to government step in. There is a question on Slido from one of our online participants I think, and I'll read it. CDR research makes sense. It makes sense to move. Oh, it's not one of our so I'll hand the microphone to the person who asked this question. Yeah, I saw a talk last week at scientific talk about marine CDR and the research research was widely, widely supported by the scientific audience, but the person started their talk by saying that large scale deployment of technology, large scale of CDR makes no sense. If, if we don't cut admissions first, or at least drastically reduce them. And I wonder what you think about that because it would, you know by 2030 there's not going to be a huge cut in emissions is based on everything we've seen and and yet perhaps the companies themselves are starting to move toward large scale. So, marine CDR by 2030 was that. Can you comment on that. Julie you want to take the first step I've been talking a lot. Emissions reductions is always the best solution. And it won't be enough. So, you know, the research program that David's organization has laid out ocean visions for what it will will take to understand how these technologies work to de risk them what relative to the climate risks that we face. And to do that stage skating and that's a multi year process. And quite frankly, it's in some ways open ended like we don't know when we're going to get there with any particular technology but that shouldn't hold us back from going forward with it just like, you know, based on the global stock take we are headed for a 2.6 degree C world and that's based on the latest global stock take. And yet we shouldn't stop and emissions reductions we should actually, you know, amplify what we're doing and we just need to go a lot faster. And so, I think, I mean, the urgency is clear is clear to everyone and I think that there is an ability to move forward on on multiple multiple friends but but it absolutely is incredible. And it's vital that the emissions reductions happen there will still be some sectors that will be really difficult to eliminate all emissions from and that's why you see those plots of the net zero and the expectation that some of the countries aren't going to meet their emissions targets. So for all of those reasons, the negative emissions need to need to occur. So I think that the main piece of that the ocean CDR is is the one that we're talking about today there are also terrestrial and air sides of the CDR component that you will maybe familiar with as well. So I want to expand on the question that Jim just asked because there is a question that I've heard asked. You know, there is one thought that countries that don't actually want to reduce their emissions, because it is so difficult will instead move into this mode of operation of funding climate intervention technologies to almost take the spotlight off of the fact that they're reducing their emissions, and maybe that work on climate interventions will involve yet more increased emissions. So the sort of negative feedback loop is that something you've heard about I was kind of a little taken aback by that question and not sure how to answer it. I mean that is one of the concerns that people have raised that there is sort of a moral hazard with employing something like solar radiation management because it decreases the incentives to decarbonize and to scale up durable carbon removal. I would say, look, if the, my understanding of this committees, the purpose is to develop a research agenda for NSF for the next 10 years and if that's correct, I would say that the questions around MCDR are some of the most fundamental and substantial that I think NSF ocean science division could wrap its head around and and so I think that those questions are philosophical and important, but I think they're really separate from places where NSF can make a really meaningful contribution to generating new knowledge that answers questions about whether or not these technologies are going to be viable climate solutions and that's still an outstanding question. It's possible that the answer to that is no, I want to be, I want to be super clear there but I think we all collectively have a responsibility to look, and I think NSF is really well positioned to be one of the leaders in helping us look collectively a society. Very well done David you brought us right back to our statement of task people long in a national Academy committee well done well done. So more questions are pouring in I see Jay Z that you put in a question would you like to ask that. If I got you on microphone. Yeah, I have multiple questions in but I'll just keep it to sort of one central question which has to do with, you know, for better or worse, we have to rely on earth system models to, in some ways, project the impacts of one ocean acidification in general but CDR, especially for, we're using say ocean, our alkalinity enhancement are we in it at a state where the models where there is this sort of collaboration with the modeling community. One, and are the models still not good enough to accurately project the impacts of ocean CDR. Julie you're the modeling expert. Yeah, this is an active area of research. You know there's multiple groups working very quickly to improve biogeochemistry and applying it to these kind of use cases than the not proposal or proposal call are but he have been government and Noah and another itself have been funding lines that have supported the coordinated research between the modeling academics and some startups to really tackle how these work together and how they enter play and the ability to assess the efficacy of these technologies when they might be trialed and added in very small pilot projects. And it's, I look, it's a, it's a, it's a to do this on the scale that is required it's an absolute grand challenge it's the sort of grand challenge that in the past NSF has just risen to the occasion. I was with some people the other day philanthropic funding and as they sort of thought about the what all of this entails I mean they started talking about Manhattan Project scale funding to really accomplish what needs to be done here I mean it's a breath taking ambition on the modeling on the observing side on the technology development side. And there's just so much at stake that I don't see how we can, we can fail at it I mean there'll be certain technologies that will fail. And then we move forward with new ones I mean we, we, we, we just have to go after this and figure out what will work, how when and where and the models are really key to all of it. Thank you Julie we have another question online on Slido that sort of pertains to what you just said about how we move forward. Mona, do you want to ask your question. Sure. Hi Julie. Hi David. Thanks so much for your, for your remarks. I wanted to ask you a question related to governance. So the open ocean is relatively free from regulation and so could you reflect on the kind of governance frameworks that will be required to facilitate research on MCDR. I'm also thinking in the context of deep sea mining for example, you know how do we, how do we think about frameworks in which science and technology advances on issues like this. Julie do you want to start or you want me to start. Yeah, you, you spent more time on this one get you go ahead. So I would contest the statement that the high seas are relatively free of governance, you know the UN clause the UN convention on the law of the sea the London convention and London protocol the recently signed BBNJ Treaty of biodiversity beyond national jurisdictions. I do think and it's well to be recognized within the scholarly community that studies these things that none of these governance regimes are fit for purpose for marine carbon dioxide removal, research and development, much less deployment and so one of the outstanding needs of the law of the sea is governance clarity and regulatory clarity both within national waters and in international waters. There are a number of efforts underway to try to provide that regulatory clarity within US waters I'll highlight to one is the recently announced fast track action committee through the office of science or through, excuse me, was announced through the science and technology policy. It has representatives from 12 federal agencies who are sitting there to consider the science technology permitting and governance needs for MCDR research in US waters. There's another effort that was released by Columbia University to propose a model federal law that would streamline the permitting process for doing responsible research in US waters. In the international scale though moving beyond the US. It is a lot there's a lot of confusion and lack of clarity about how people may or may not move forward with research and development projects and especially when those research and development projects may include potential interests. And so that's actually like an area of contention and consideration and deliberation that was just re energized in the last week or two with some announcements that came out of the international maritime organization and some analysis of those comments and the reality is that you can talk with multiple people right now and get multiple interpretations so there's a big need for clarity there. Thank you very much David. And thank you very much Julie we are unfortunately out of time I feel like this conversation could go on for much longer. Clearly this is an area where I am sensing some discomfort in the room in a way that I didn't earlier in the day. So these are topics that we just need to keep talking about so I really appreciate the input David and Julie. I'm going to move on and the next topic which is critical minerals, potential, potential topic for us to include in our report. And we have two speakers one is Beth or cut and Amy. I'm not sure who's supposed to who wants to speak first, but whoever is going to go first is welcome to start. Great. Thanks for the intergem I'll go first let's see. I succeeded this year. Yeah he looks good. Okay. Great. So it's still in presentation mode. I'm going to go for a minute and I lost it. Let's go ahead and start. I'll just hit the start there you got it now. Okay. Yeah, thanks again for the invitation to speak. It's my pleasure to chat about critical minerals for this NSF OCD meeting. I have some background on the topic that we're discussing, most specifically, entangling a few different terms intersecting phrases that are sometimes used to so their synonyms, but they're really not. And those phrases are marine minerals and deep sea minerals marine mineral resources and critical minerals, specifically in this sphere critical minerals relevant to the marine zone. So first, I'd like to introduce the topic of deep sea minerals. So the three deep sea minerals that are most often discussed are cobalt rich ferro manganese crusts, polymetallic nodules and polymetallic sulfides. Also known as ferro manganese crusts, ferro manganese nodules and hydrothermal sulfide minerals. The only three that have been defined by the International Seabed Authority, which is the body that's tasked with regulating seabed mining in areas outside of domestic waters. And so there are only three that have exploration contracts issued in those waters. To date, there has been no mining of any of these three mineral types, any one ocean. They occur in completely different ocean environments, crusts are predominantly found on sea mounts and hard rock surfaces. The nodules are predominantly found on on a vessel planes. I'll note that there's also sea mount nodules but there's very significantly an extent and composition compared to this little thing nodules to the best of this talk. We're going to talk about a vessel plane nodules nodules just a morphological term. And then hydrothermal minerals form anywhere you have extensional settings. The points on the bottom are showing the elements, the minerals that that these minerals are of interest for, get into that in a little bit. And then I've folded the elements that are on the current critical minerals list. And the minerals that are being dispatched from these crusts are being discussed as potential sources for mainly cobalt and cobalt copper, and then potentially rare earth elements, maybe some other minerals as byproducts. Polymotoc nodules are being discussed as potential sources for nickel copper cobalt manganese with potential byproducts for elements in lithium. Polymotoc sulphurants are considered potential resources for copper, zinc, sulfur, gold, and lead. So there's a couple of other types of deep sea minerals that are a little bit less frequently discussed. I do want to acknowledge them, but I think they're less relevant to the critical marine minerals lens. So I think we can get into this later and I'm happy to chat about them, including in the Q&A, but those are phosphate minerals, which is phosphates or sedimentary rock with a high concentration of phosphate minerals. There's a number of different settings in which marine phosphates, phosphates occur. Most relevant are those occurring on continental shafts and slopes, and then there's also rare earth element in neotrium rich mines. So these occur in deep water abyssal plains with low sedimentation rates. They often, they can co-occur with nodules. The rare earth elements in these are appetite-hosted, and they occur fairly deep into the sediment column. So let's write them before, so four, but five or one meters into the sediment column. So these maps, again, for crust nodules and hydrothermal minerals, which I'm not describing by the geologic terms that refer to them, rather than the economic geology or commodity terms, because crust nodules are both predominantly exposed to ferramanganese, whereas cobalt rich refers to cobalt being an element of interest in them. They're not made of cobalt, they're something with polymetallic. All of these minerals are polymetallic, but polymetallic refers to the fact that that's the potential commodity profile. So these maps, specifically those for crust nodules, show the areas that crust nodules are predicted to occur based on oceanographic and geologic criteria. So we, they don't show locations containing mineral deposits. Same thing with hydrothermal minerals, we've highlighted extensional settings, so mid-ocean ridges, arcs and back arcs. So these, again, we could overlay these maps with regions where we've sampled particular examples of all of these mineral types. But again, that helps refine the predictions, but it does not show us a mineral resource, right? One sample, the handful of samples does not indicate a mineral resource. So what is this distinction that I'm making between marine minerals as geologic and oceanographic occurrences, which are widespread for ferramanese. We've referred to them as some superglutus in the global ocean, versus marine minerals as potential commodities, right? Because there is a distinction here. So marine mineral resources are those that occur in such form that economic extraction is currently or potentially feasible. And since no mining of deep ocean minerals specifically has occurred to date, that's actually a bit, a little bit tricky to consider where these cutoffs may be. Thank you. The really crucial point here is that there is in fact a cutoff. There is a distinction. And when we happen to see some sea modules or some ferramanese crust on an ROV dive, it's not valid. We can't immediately refer to that as a resource or a potential resource without more information about the minerals, extent and composition. And so that's the few steps of background before we get to the topic that we're actually here for today, which is critical minerals. So critical minerals, I've got the definition here up here at the top. There's a non-fuel mineral or mineral material central to the economic or national security US and has a supply chain vulnerable to disruption. So few points specifically. This definition is from the Energy Act of 2020. It is a US specific definition. And besides that fact, the other thing that I need to point out is that this list can change through time, right? So it's based on supply risk, disruption potential, trade exposure, economic vulnerability. All of these factors are factors about they're not intrinsic to the minerals themselves. They have to do with the minerals as commodities and within the supply chain. And so I think that's a really crucial thing to keep in mind. Again, because deep ocean minerals are not currently in the supply chain. And so I think this is a really interesting contrast with the maps that we've been looking at earlier, because the maps we've been looking at before are the really broad regions in the oceans where we predict marine minerals to occur. And now we're looking at locations where marine minerals are known to occur and have quantitative resource data to go with that. So, within the Clarencle-Burton zone, which is broadly outlined here, there are a number of separate regions where resource has in fact been demonstrated potential commodities to be extracted from modules in the CCC with nickel, manganese, cobalt and copper. So nickel, manganese and cobalt are critical minerals. So we can say that those are potential deep ocean critical minerals in the near term. So for polymetallic sulfide locations, potential commodities of interest include coppers and silver gold, maybe lead. Zinc is a critical mineral. So those are potentially near term critical minerals of relevance to the sphere. Another topic that I'd like to briefly throw in there, I think this group is mainly interested in deep ocean minerals, but coastal minerals are an entirely different category. Coastal minerals are currently extracted most significantly, tin from offshore Indonesia. So that is a current critical mineral, potentially of interest to the sphere that I think we don't typically talk about because it's not a deep ocean mineral. And that's what we usually put together. But I think one of the really important things to keep in mind is that if you want to move beyond, so tin, which is entering the supply chain and in the coast and the handful of other locations where we do have deep ocean minerals that have been delineated. The uncertainty is not just about environmental setting and potential consequences of extraction, but the uncertainty extends to the minerals themselves, whether the minerals constitute a resource, whether or not in a specific location. If they have the potential to enter the supply chain in the timeframe over which that may occur. So it's kind of a big level of uncertainty. Given that, how do you decide where to prioritize the on those handful of regions. I'd suggest this is a really great opportunity for partnerships. Because as far as I'm aware, economic geology is not something that's been emphasized in OCD for quite a while, even in the terrestrial sphere, I'd suggest that it's a discipline that's been deemphasized in the US for quite a while. Excuse me, to expand beyond the couple of locations that are well measured and constrained, we need partnerships are really crucial. So if you don't want to end up with heavy investments in regions that fall under the broad umbrella, and those those really broad maps where may contain critical minerals, but when we do that work, we find that they might actually have no relevance from the perspective. Briefly, I think also challenges and priorities relate to assets to do the work with the exception of critical minerals. So most of the areas of interest that we're talking about are of this old apps that require ships and assets to do that work. So, in the US, it's pretty exclusively, you know, as global class vessels, this old up winch wire for mineral quantification for coin or dredging drilling is it is a requirement. I think, yeah, another opportunity and a broad challenge for OCD, you know, there's a number of studies that are working on potential impacts in a given location. And I think it's harder to address and a more community scale question is, is how far did those impacts extend and affect broader marine processes and what thresholds for these may be. So, over time. Sorry, though. Can everybody in the room hear me okay. Yeah, well, yeah, you had it you had it up there a minute ago. Yeah, looks good. All right. Thank you Amy for that introduction. Thank you to the committee for inviting me to speak today. We'll recognize that, you know, there's a lot of traumatic events going on in the world that might be distracting your attention so thank you for spending the time thinking about this topic. I will try to be an engaging final speaker for this day very long day for the committee. Thank you for the Q&A. I'm pretty bad that doesn't know me, but work at I'm the vice president for research at the go laboratory but I'm really here on behalf of the Cobra. NSF funded a cell net, which is really trying to think about research gaps in the space. And so if you don't know about Cobra, there's our places to find us online. So from my perspective on some science priorities related to the critical minerals right maybe give us a good background about how we use these terms. So I'll try to follow that. I like to start with an image like this to really frame how I am thinking about what might be happening in the deep sea related to interest and potentially exploiting marine minerals. In all of the environment types that Amy talked about and minerals that are found there. You can have complex animal and microbial communities living in those places. In some cases very unique animals that are dependent on the chemistry or those environments. OCE has a long history of supporting research in those places and understanding those cool ecosystems. There's a lot of research into understanding kind of those chemical processes and things and so what I'm thinking about what Cobra is thinking about is, you know, how might this new human industry impact what is happening in these environments and we'll see in this diagram that there's a part over there on the right. So I'm using a scientific drilling vessel and connecting that to some C4 carbon sequestration. So sidebar if anyone wants to talk about that I'd be happy to. That's not what I'm going to focus on here. So, as scientists we often talk about what happens in the deep sea and like very academic terms of like carbon fixation and chemo synthesis. And I have been training myself to also think about these things more from a societal context and this often gets put into kind of more of a capitalist framework of what does the ecosystem do for you. And you know, referring to things that happen in the environment kind of in that framework right so a service is that these ecosystems provide food they do climate regulation. Nutrients that support everything else. There was a bar for genetic resources on and on. And there's some great resources available from the deep ocean stewardship initiative if anybody would like to learn more about these types of services that happen in the environments that might be targeted for marine mineral exploitation. And when we think about those services and we think about what might be potential impacts of exploitation. And so here's one schematic of one of those types of impacts where you can imagine that to collect marine minerals, you would use some kind of collection device that will create sediment plume disturbances. We'll also have disturbances higher in the water column from the return plume from the after the ship has processed the materials. I think it's important to recognize and some of these really nice schematics that a key element is missing in them and that there are no animals shown here. But we clearly know that there are animals on the sea floor that will be impacted by these. activities. And I was part of a paper led by diva and mom that came out last year that did an extensive literature review to try to understand what do we know about environmental baselines in these types of habitats. And some of the key topics you would need to study to have a baseline. The intensity of the color is a reflection of that there's more and more information available. And so, maybe not surprisingly considering the investment of OCE and other programs right we have comparatively more information from the hydrothermal vent environments than we do from some of these other habitats that might be considered for mineral exploitation. And in particular for instance if you look at that bottom line life histories things like how do animals reproduce in the deep sea. We have very little information about that in some of these ecosystems. Furthermore, if you look at things like what are potential impacts how do these ecosystems respond to those impacts, being things like noise. The, the creating sediment plume disturbances metal toxicity. We have really little information and so I think that's a key thing for OCE to be considering is, you know, these are big knowledge gaps that would need to be resolved to underpin understanding if we can start this industry responsibly. I'm a microbial biogeochemist. And so, I am often thinking about this in a microbial context, and I would say that the role of microbes in these deep sea ecosystems hasn't gotten as much attention as animal things maybe for obvious reasons animals are a little bit easier to think about the microbes. But again, if we think about what microbes do in these deep sea ecosystems right they they provide services that are really essential for the ecosystem sub function things like primary production, the transformation of organic material nutrient mineral recycling. And so to give you some examples if you don't think about what microbes do right like the conversion of carbon dioxide into organic carbon fueled by chemo synthetic processes are chemo lithotropic reactions right that's a very important service right they are providing new food to this ecosystem. Okay, microbes are an incredible reservoir of genetic information that could potentially be used for things like antibiotic discovery and cancer compounds. You name it. The point I want to emphasize here is, again, we have, we have big knowledge gaps in terms of what are the values, what are the ranges of these services, what is the variability of these services. We understand how we might value them in comparison to the value we might get from exploitation of critical minerals or marine minerals. So just as one example. This is from a study that came out a few years back led by some teams based in Germany, where they had gone to an environment that had previously been experimentally dredged, and it was a place that had nodules. And went back and asked a bunch of different questions I'm just focusing on one of them. How is primary productivity in the form of creation of new organic carbon from carbon fixation. How is that rate affected if you compare pristine sediment, sediment that was dredged 26 years ago, and sediment that was just dredged on the expedition. And so to draw your attention to that graph on the right, the orange bars are the rates in the that have been impacted decades ago, right so often we think of microbial processes as being kind of very resilient. And that they can recover on much faster time scales than animals might be able to like we colonize for example, and this study shows that like you even decades later have a loss of microbial ecosystem service. So that's something that we'll want to keep in mind for evaluating what's happening in these ecosystems. So I would say some critical knowledge gaps related to this topic of marine minerals that OCE might want to think about. First one echoing Amy right we need to actually understand the marine mineral content of what is on the sea floor. These, these global maps might tell us where there's potential, but they don't actually tell us what the resource is what the content is. You know what is the mineral forms, all those things right so that's still an area that requires research. But we also then need to understand the biodiversity and ecosystem services that are in these habitats and how they can be impacted by exploitation, and not just on the sea floor but also in the water column. Likewise, we have very, very little data on resilience of ecosystem services in the deep sea to these kinds of perturbations right again I emphasize that we almost we know almost nothing about how animals reproduce in the deep sea. So it's really hard to understand how impacts might translate to organisms that you don't even know how they reproduce. But we need that kind of research we need studies on impacts and resilience. We also need, if we're this industry is going to develop right we need to figure out cost effective strategies to provide early warning related to preventing harm. So, monitoring strategies, observation strategies, shipboard analysis, all those kinds of things are going to be need to, you know, accelerate. And then we also need to think about this in the context of cumulative impacts with other stressors in the deep sea, for instance like climate change. If we're going to accelerate on ocean CDR strategies that are sinking all kinds of things the deep sea like how does that impact areas where there's other stressors going on. And I would say that a critical need that we have to address these knowledge gaps and again a place where OC could help is in relation that we need a lot more deep sea scientists to be helping with this kind of research and we need to also think about how scientists are trained to communicate their scientific findings to the stakeholders right. This was also brought up in the other sessions about CDR and deoxygenation, you know all those things. We know we're really good at talking to each other as scientists but we may be less good at explaining what we're finding and what matters. I would just like to emphasize that one of the other findings of that paper, the Amon and El paper is that you know this is like decadal scale research that is required for each of these research types to really get at these baselines the natural variability. What are the methods we should be using. So, you know I think this is appropriate for considering for this decadal survey of ocean sciences like this could be a decadal type program. There's obviously a way more I could say about this if you want to read more about these knowledge gaps or about what's going on related to deep sea mining potential. I've learned that I talked about, we've had several webinars on this topic over in the past year. So I'd invite you to see those they're recorded they're up on YouTube. In case you want to some more detail about all these things. And then just again to emphasize that part I made when I was talking about the needs right. What Cobra is doing as well as other cell nets that are in the ocean space like the deep ocean observing strategy as another example right of these kind of international network of networks are I think are really essential for trying to help identify research strategies to like accelerate filling those knowledge gaps to bring the training to for a scientist and students on how to translate to policymakers, and also to try to increase that education for the next generation. You know, in terms of thinking about OCE strategies I think it's also a broader NSF wide strategy of like are there other programs in NSF that we can look to as examples so that we're not just individual project focus, but thinking about things more collectively. And just as an example of that just a few weeks ago, the Cobra and the other, one of the other NSF cell nets for in the ocean space I do collaboratively held this workshop, specifically about like okay scientists you have all this knowledge or you're developing all this knowledge, how would you translate this for policymakers who are right now trying to come up with thresholds and guidelines for these very important topics that we have very little data on and and helping with that data translation to policymakers. And taking that a step further like actually like trying to show scientists, like you can actually go to these policy meetings and share your information and have direct impact to make sure that science is informing these new industries that are forming and the regulation of them. I know I'm going a little over time, hopefully, I'm okay here, just a couple things I want to emphasize. So I think I said this already, like we need to accelerate the development of shipboard and in situ methods for assessing ecosystem service changes in real time. So that if an industry is operating, they see they're causing harm, they can stop before it becomes serious. Right now we don't have a lot of technology to do that. We were going to need more ocean observing technology and approaches right and translating what we're learning from oh I to these types of environments, because they're going to require monitoring. And I love this graphic because it shows all the ways you can measure stuff in the surface ocean and like the deep sea is like just a little part of that right so like we've really got to push the deep sea aspect of our ocean observing technologies. As Amy said right, we're going to need deep sea assets for this kind of research, ships, ROVs, AUVs, HOVs, moorings, profilers, coring systems right like we, we need to make sure that we have these and if not like increase these capacities. If it's only like one expedition a year studying this we're never going to close that decadal gap. And I'm going to end with this, just highlighting two things from a recent advisor, the European Academy Science Advisory Council right so like the individual science advisory councils of all these different European countries in a higher level at the European Academy level. They recently came out with a report just a few months ago about this topic. And one of the things they highlight, and this is something that Amy also brought up right like the types of minerals we're considering for this new industry. It's questionable how much we can eat them from the deep sea in terms of their supply risk is one of the things they emphasize here but what I wanted to point out is one of these statements that was in this summary. One of the things is that the lack of consensus on what constitutes serious harm, and the current lack of quantitative thresholds is going to limit the ability of international waters and international seabed authority that you can translate this any country that's considering it in their waters. It's going to limit the ability to effectively protect the marine environment, until we understand these ecological consequences. And so that's really what I want to emphasize here is, again, we have an industry that's maybe going out the door and we don't have yet the baselines. We need to fill those critical knowledge gaps and I really hope OCE is interested in doing that. With that, I am done. Amy and Beth, thank you very much. And we have plenty of time for questions. We'll ask a two bush. Sounds good. Thank you. Thank you very much. So tell me, can you tell me a little bit more about the kind of seagoing capabilities you would need in order to do this research that you're thinking about as prior research areas in particular. You know, we've recently taken a deep dive into the ocean drilling program, but overall, you know, this morning we heard about the ocean observing initiative and of course we're going to hear more about the, you know, seagoing capabilities. Can you tell me a little bit more of what is required on your end. I'll go first Amy I'm curious your perspective. Right so that the research that's currently happening that I'm aware of, you know, it requires commitments to like do multiple cruises to the same place repeatedly. You know, if not within a year like several years right again understand those impacts. You want to, you know, you want to set some baseline data you want to understand the variability, you want to understand impacts you want to understand recovery from impacts right so that's a commitment to have assets in a place where there's kind of an agreement we're going to try to do this work in this place. And you're to do these studies well requires like having moorings having sediment traps having things in the water column to look at sediment blooms having ROVs that can go in the water right as impacts are happening to see like where things are moving, or other, you know, like cameras on landers. I mean, where this is all proposed to happen is in in the deep sea far away from land and so thinking about right like so many of the oh I demonstrations that have happened right based on like the regional cable to rate right there's power. Like that's going to be a huge thing we've got to solve for doing this out far away from land is like how are you going to power all the equipment to do monitoring in an effective way. So I don't know that answers your question but at least it gives you a sense of the scale that I think we're going to need. Yeah, I'll add to what Beth said. I think part of it depends on what OCE decides is within their mandate. So, you know, that's talked about monitoring is a really critical component of understanding, you know, impacts if this goes forward. You know, we need a minimum we need, you know, ships that can stay at sea for for a month at a time to do this work, because you're often in the weeks transit from shore so shorter trips and that don't don't really make sense. ROVs and AVs that can work to 6000 meters so it's a basic requirement. I mentioned drilling before and you mentioned it too so if you want to quantify crust or header for momentum so you do need, you know, lander type or ROV type drills and I'm not sure whether OCE would would consider that to know what but yeah there's pretty serious aspects depending on what you want to do. Thank you ladies, this is Allison. While I fully appreciate the need to do more scientific research on some of the examples you gave Beth the impacts how animals reproduce in the deep sea, etc, etc. Do we really need to know more to know that mineral extraction would be bad for these animals or for these ecosystems and I'd like to get your thoughts on that. Yeah, thanks for that question. It's an interesting one. So one of the case studies we have is the work that's been happening in the Clarion Clipperton zone, which Amy highlighted in her talk as a place where there's a lot of attention for nodules right. And there you that there's kind of an agreement that there's these set aside areas where no harm should happen nothing should change, and that even within a contracted area for potential exploitation that a contractor would have to have impact zones and preservation zones. To show that to kind of think about that, you know, maybe we're impacting the animals right where we're doing our work but if they're similar animals somewhere nearby, and they're not impacted maybe it's okay, you know, maybe the larva could spread or whatever they need. And what I have seen in the data so far is that often the assignments of those preservation zones and the impact zones have actually different communities in them. And so we maybe don't know are we are we preserving what we need to preserve. You know, do we have enough spatial variability here so in some ways I think, yes, we still need to understand if we're going to cause impacts like, I don't know if we're going to be able to understand everybody every organisms reproductive strategy. We at least need to know, are they in the, are they in places where they're going to be impacted or not if they're more widespread that's one thing if they're very rare. That's a different thing. And, and then I'll answer this another way to. I know that for instance, other federal agencies are also interested in this topic in terms of should we be accelerating technology development, for instance, in this space, and kind of doing some of those economic assessments of, given what we already know at a low level about ecosystems and what they provide and the potential monetary value of those services and then we compare that to the potential monetary value of exploiting minerals. Is there a justifiable economic case for doing that. And, you know, I think that's also something that we're going to need to do, you know, scientists collaborating with economic economists to think about those things and to help calibrate those kinds of estimates. I had a follow up question to that related question is, given the, what you said about the need for technology and observations on impact and so on it would. What is the responsibility of the mining companies in the open ocean away from the ocean, ocean control zones of individual countries are, are they responsible for doing an environmental impacts and assessing that. Does that depend on where those companies are based which country. Those countries are based. Right so the mining code is still being developed for work in international waters in the area. But yes, in principle, every contractor would have to do an environmental impact assessment and present a plan for how they're going to monitor environmental impacts. That's a very hot topic right now at the International Seabed Authority of what are the requirements for those types of assessments. What would be the required monitoring strategy. Is it a one size fits all or is it kind of resource location approach dependent right. So, there's not a firm answer to your question, other than, yes, in general is the contractors responsibility to do that monitoring assessment. The worry of people like myself is that if we develop the regulations before we actually know that the regulations will work to ensure the effective protection of the marine environment. We're setting ourselves up for a precedent of giving permission to do something before we know if it actually will prevent harm. Thank you, Beth. This is too but Jim just handed me the microphone for my second question. Amy, you spoke about the need for more partnerships towards the end of your remarks. Can you tell us a little bit more about what kinds of partnerships you were. Referring to is this with other agencies like USGS or is this with industry. Tell us more. Yeah, I mean, potentially both right so I think the one of the points that I was trying to make is that if we're talking about critical minerals. We're talking about a topic that is based on minerals entering the supply chain, which to date is terrestrial minerals and coastal minerals so deep sea minerals are currently not in that sphere right like this is a potential future where deep sea minerals enter that sphere. I think on all sorts of topics we need partnerships with people who understand the current landscape of critical mineral systems right so when we're talking about farm or environmental impacts of deep ocean mineral extraction. I think one of the things that people have been trying to do for a while and it's not in any way trivial to do is to compare those impacts to impacts of terrestrial mineral extraction. Can we compare those. And so that's a major ongoing challenge. I think when we're talking about, you know, designing in environmental baseline study or some experimental work to consider what impacts of extraction in a given region might be. When I talk there's a handful of regions where we have quantified resources in the deep sea for anything outside of those regions. I think you need a partnership to determine if the region is relevant for the work. So I think this is a far, far reaching need. I see that Shannon has a question. Hi, I guess the one is for Amy as well just curious about the scientific co benefits if you will. What are what other research lines could benefit from exploration of looking at critical minerals. Yeah, I think that's a really great topic. I think if you switch from the critical minerals lens briefly to marine minerals lens and all of a sudden there's many different aspects of OCE of another topics that are relevant you know for many of these projects that create really slowly from seawater. So they have significant uses as paleo oceanographic records, which is really important thing to understand right now I think, you know, in terms of hydrothermal systems. It's an ongoing question how trace elements from better thermal systems impact the wider oceans and how better thermal systems impact ocean ecosystems, once they become inactive what that intersection is my team does a lot of work on farming these minerals and how they vary in composition in different areas, which is generally as well, almost entirely as a result of changing oceanic parameters and in basin scale regions. So I think there's a wealth of information to to be gained about the ocean system from from studying marine minerals. I think that I have a question for you about the slow rate of recovery. I was curious to know if that's just because of the slow metabolic rates in deep sea, or if there's something else there, and how we should think about that in terms of impacts over long periods of time. Yeah, thanks for that question I'm not sure who asked it I can't really see the room. This is a jeep. Oh hi jeep. So the study I was referencing right is in a nodule area, those areas tend to co locate with places that are really low sedimentation rates. And so, in this case right it was experimental removal of the top layer of sediment, which would also happen with industrial sale collection of nodules right it would basically perturb the upper, let's say 10 centimeters of sediment. That's like millions of years of sediment decretion in these low, you know, 10s of thousands to millions of years, and those are relatively organic for materials and so it essentially what you're doing is you're removing the most organic rich material, and exposing the older sediment at the sea floor. And so it's going to take, you know, geologic time scales to get the sediment back to where it had been. And so you're basically just left with a much more organic poor and lower biomass microbial community that can never recover to those rates of activity is essentially how I view those types of I will say right that one thing keep in mind right that was an experimental study that doesn't actually replicate what mining will look like. And in every study that has been done. Since that study right of like actual test mining right the results end up being different because you're scaling the the impacts to like closer to what they will be. And it doesn't mean that what you did an experimental study actually scales to the industrial level scale in terms of how sediment is redistributed. You might have places. Now you've moved all the more organic rich and put it somewhere else. You might reactivate heterotrophy you might have more reminalization. So it's going to be a heterogeneous response within ecosystem is something I also think we need to keep in mind. I'm out of time for this session and I want to thank both Beth and Amy very interesting talks and appreciate you your participation. Thank you.