 Greetings everyone and thank you so much for attending the Science Seminar presented by the NSF National Ecological Observatory Network or NEON which is operated by Battelle. Our goal with this monthly series of talks is to build community among researchers at the intersection of Environmental Science, Ecology and NEON. We have a fantastic lineup of speakers for you who will present their unique perspectives on addressing a variety of ecological questions across scales. So this is our first seminar in the series that's going to go through the 2022-2023 season. There'll be nine talks in total and we're so excited to have Dr. Sydney record here kicking off the the talks today. So before we turn it over to today's speaker just a few logistics I'll go through. First we have enabled the closed captioning feature for today's seminar so if you'd like to use that please find the CC button in the menu bar of Zoom. You do have to have an up-to-date version of Zoom for this to work but you can go in ahead and enable the closed captioning. The webinar will be mostly a presentation followed by a Q&A session. As questions are coming to you please add questions to the Q&A box and then the moderators will facilitate discussion at the end. There will also might be an opportunity for you to raise your hand and ask a question. We welcome contributions from everyone who shares our values of unity, creativity, collaboration, excellence, and appreciation and this is all outlined in our neon code of conduct. So this would be for neon staff as well as anyone participating in a neon event. So when you get a chance to learn our webpage please do look at the neon participant code of conduct. I can't drop it in the chat I realize because there isn't a chat but you can find this on our on our webpage. The talk will be recorded and made available after the fact. If you find it interesting tell your friends. We will post it on this seminar page and make that available for asynchronous viewing. And lastly to complement our monthly science seminars we're hosting related data skills webinars where folks can learn more about how to access and use neon data. So the registration for those is actually available on the same webpage on the neon science webpage where we have all this information about the science seminar series. If you just scroll down we've got these related data skills webinars. We'd love to see you at those if you're interested please go ahead and check those out. All right now I'd love to turn it over to Brigid Haas to introduce today's speaker. I'm pleased to introduce Dr. Sidney Rekard today who recently joined the University of Maine Department of Wildlife Fisheries and Conservation Biology. Prior to that Dr. Rekard was the Associate Provost and Associate Professor of Quantitative Ecology at Brydman College and her research focuses on two main themes which are understanding the processes that drive species distributions and determining drivers of genetic trait and species diversity across large landscapes. Sidney has incorporated neon data in her research since the early days of neon and we're excited to have her give a talk today on her work compiling data on land use histories of the neon sites which is an important component of understanding the ecological processes at these sites. So with that I'll welcome Dr. Rekard to speak. Thanks. Hi thanks so much for for having me and I'm thank you to Brigid and Sam and Eric and all the neon staff. So today what I'd like to talk about is the the notion of having a database of current and past land use of neon sites and I'll begin with a story in research. All right so first of all before I start that story though I'd like to recognize that I'm going to be presenting work that's been really a truly a team science effort. This group of principal investigators Phoebe Zarnetsky from Michigan State Angela Strecker from Western Washington University and Ben Bezer from University of Florida and I in the early days of neon in around 2015 received an eager neon award from NSF to start working with the early neon data and we've worked with a whole cast of really excellent students and and postdocs throughout that time to do this work and so I just want to recognize that this has been truly a team effort and I'll give shout outs to individuals as we go go through that. So today I'd like to talk about the need for a land use database for all of the neon sites and I'll provide some context for for thinking about why we need land use information then I'll describe how I'm working on building such a database of land use for neon sites and then also talk about how there are some different synergies between the airborne observation platform and and thinking about land use. So as you all know neon measures many different things using a standardized protocol across dozens of sites in the United States and beyond and the the types of measurements made are everything from the the small mammals scurrying around on the the forest floor all the way up to the the flux of greenhouse gases in the air above the sites and so neon is this huge rich data source that we've never had in ecology that allows us to answer these interesting macro ecological questions. Now if we go back in time and think about the early days of when they were first scoping this idea of neon there were a lot of listening sessions and working groups that the National Science Foundation funded. One of these was hosted at Harvard Forest where I've done a lot of work which happens to be a neon site and also a long-term ecological research site and so this gives you an idea of what people were thinking about when when starting to consider this observatory and here you can see things like coupled human environment system really stand out in the title of this white paper that was written to the National Science Foundation as they were thinking about neon and you see under here two broad goals understanding how land use change and climate variation affect ecological systems and then the ability to forecast future dynamics of those systems. I want you to keep those two broad goals in mind as we walk through this this research story. So then we get to 2008 and neon's becoming more of a reality more than just a vision and as they're thinking about planning out neon there was this paper that appeared this commentary that appeared in frontiers and ecology that in the environment and this paper featured you know the vision for what types of data would be collected at neon and so in this Stommel diagram which was figure two of this paper you can see time on the x-axis spatial scale on the y-axis and then those those blue boxes indicate different data products that that people were people like Dave Schimmel and Forrest Hoffman and Michael Keller and William Hargrove were envisioning as being the data products that neon would produce and some of those have come to be and some of those have not. The fundamental instrumentation unit is all of the the sensors that neon has in the sensor data. The Fundamental Sentinel Unit is referring to the Sentinel Organisms that neon measures both in aquatic and terrestrial environments. The airborne observation platform AOP it has come to fruition but what was left out as the the network was built out was this land use analysis package and so fast forward to 2018 where we're working on our our early eager neon award from NSF and we're in this paper by Quinton Reed which was one of the first papers using the organismal data from neon sites. We were looking at intraspecific trait variation in body size size of rodents from the neon observatory and what we were seeing was that there's a lot of variation in the data that we were getting at sites and we weren't really sure how you know if you look at the data as an end user you see a lot of variability in the data and sometimes you wonder is you know what processes exactly are leading to that variability and having been working for for many years at one of the neon sites Harvard Forest where they did a lot of small mammal trapping in the early days before they they built out the whole observatory as they were trying to test the protocols I knew that there was a lot of variation in land use history at the site not influences the forest dynamics at Harvard Forest and I wondered you know if you were if you're using neon data do you actually understand that there's variability in the historic land use of the site that then plays out in the ecosystems that we see today and in the small mammals and and it really got me to thinking about you know how much of the variability that were that people might see in neon data could be from land use and people's interactions with the environment that doesn't get accounted for in our analyses and then neon became fully operational in 2019 and as I mentioned before due to budgetary constraints that land use analysis package never came to be and when neon was fully operational in 2019 the main way that they're accounting for any type of land use is by linking to information from the national land cover database in the in the data products and that's not to discount this as a useful source of information but it gives you kind of a static snapshot of what the land cover is at a given site and for a given plot and and it is useful information but it's it's based on Landsat which is a NASA satellite that is earth observing and it has a 30 meter resolution and so what this means is that any kind of finer resolution information on disturbance or land use history or even current land use are kind of missing pieces to the neon puzzle when you're trying to understand the variability in the data that you see and why is this important so we were thinking about synergies between LTER the long-term ecological research network in neon in a in a group that met ANC's around 2016 and in that group we we started talking about kind of information that could come from LTER sites it could be relevant for neon sites and this is a figure from a paper in earth futures that's first authored by Julia Jones that I co-authored with her and and in this figure what I did was I laid out kind of historic information on land use from the LTER network and then the neon tower plots which are shown by those those dark triangles and I just want to also you know acknowledge that Harvard Forest is situated on Nipmuc land and and I'm going to recognize recognize traditional peoples who have lived on these different lands at the neon sites as as a way of kind of recognizing that one of the things I'd like to do with this work is move past colonial settlement era land use but I haven't quite gone there yet but I want to at least provide that recognition of the the value for traditional ecological knowledge and so what we see though in this map is 1830s land use so that's you know just after about 50 years after European settlement and we've got this all mapped out for Harvard Forest and you can see that the different neon tower plots where a number of measurements are made on the vegetation the soil microbes soil moisture etc are all laid out and you can see that they they fall on different land uses and you may think well what does that matter you know 1830s was was 200 years ago that that shouldn't have any influence on the ecosystem but at Harvard Forest there's been a lot of work done on how land use can affect ecosystems that we see today so this is a figure from a paper by Glenn Motskin who worked for a long time at Harvard Forest that was published in ecological monographs and in in western Massachusetts this isn't at the Harvard Forest site but nearby we see very different soils and plant types that are reliant upon what the former land use history was in the 1830s so in areas where where you didn't have any sort of of tilling or pasture you which would be shown on the the left hand side of this this graphic you see a whole different suite of species in this case scrub oak and if you look at the soil horizons you can see that you have very clear organic and mineral horizons that haven't been mixed in contrast on the right in this diagram would be another area that has completely different vegetation in this case pitch pine and if you look at the soil horizons you can see that they're very mixed and so even though we had that land use in the 1830s over 200 years or around 200 years ago we can still see signatures of that that ecological legacy of land use and just to give you an idea of what that looks like these are some soil cores I actually you know have pictures of from sites where there was tilling and no tilling at at Harvard Forest and you can see that there's very clear the the top soil core would be an area where there was tilling for agriculture in the 1800s and you can see that mixture of the organic and the mineral horizons whereas on the bottom soil core would be an area where we have completely different vegetation where there was not historically tilling and this then plays out not just in the soils but also in the vegetation that we see at sites so this is a paper by Dave Orwig from Harvard Forest that was published this year and this is from data from the forest geoplot of the Smithsonian network and in the panel on the left in that map with the different shades of green it shows when when there was last cutting at different sites in the the Harvard Forest Forest geoplot and you can see that dark green olive color is an area that that has never been cut because it was permanent wood lot so it it was never really cut in that and was left intact for for logging for heating homes and you can see that in the right hand side we have maps of stem maps of the major species at the Harvard Forest Forest geoplot and you can see that there's very different species dynamics in those areas so not surprisingly in that permanent wood lot on the upper left hand side there where you can see suga canadensis is is one of the later successional species at Harvard Forest that's eastern hemlock and you can see that you can you can see that kind of U shape or horseshoe shape where that permanent wood lot was so we can see differences in the forest even today in the soils and in the vegetation from this former land use and that's why I think it's really important for us to be considering this kind of land use both past and current in thinking about the the neon sites and the variability that we might see in all of the different measurements so hopefully that convinces you of why we would need a database of land use both past and current for neon sites and I'd like to go into a little bit of information about how we've been building out such a database and I just want to give some context for this the reason why I'm so interested in building this database once again is because I have a grant that's funded looking at how different drivers could influence interest specific trait variation and then how that could influence biodiversity patterns and as you can see in in this proposal diagram that I have from the grant the NSF grant that's funding this work past land use and current kind of disturbance and disturbance regime are pieces of the puzzle when we're thinking about those different drivers and and this is really nice to look at in a neon framework because we can then look at how different scales of influence vary for these different drivers at the plot site and domain scale given the nice nested hierarchical design of neon so that's kind of what I'm using planning to use this data for but I also think that it's definitely relevant for thinking about many of the other neon data streams such as the flex tower data streams and instances where you might have the flex tower footprint on different types of land uses and then also for the AOP data as well and being able to validate information on disturbance generated from the AOP and I'll go into that a little bit later on but I'm hoping that this this database that we're putting together will will not just be useful for my own purpose of thinking about interest specific trait variation and biodiversity but will also be relevant for everyone who's using neon data and I and I'd love for people to reach out to me and I want to just reiterate that this will be made openly available as soon as we feel that it's ready and it's getting very very close so in terms of methods what we've been doing and I should say that this has been the work of a whole suite of undergraduate students who've been working with me at my former institution Burnmore College and through the Harvard Forest REU program and so I just want to recognize the the great deal of effort that's been put in by undergraduates and just how awesome undergraduate research can be I feel like that sometimes doesn't get recognized as much in the academy as it should but I'll walk you through our methods of curating land use history and current land use and it's really a a large kind of web that we cast for data collection and then there's a great deal of digitization that has to go into taking historic maps and converting them into geolocated shape files and then also standardizing the data across sites so that it will be useful for these cross-site analyses and when we first start out with our our data collection one of the things that we've had to to reconcile and think about is that the types of management varies greatly across sites and so in some sites we have kind of federal or state ownership and and like the type of land use history and land use data available are very different at these sites relative to other sites so federal and state owned lands are sometimes national parks like Yellowstone or the Great Smoky Mountains other sites we have as kind of extensions of universities and so these are oftentimes private sites and they're but they're they're kind of under this educational mission so we have the Harvard Forest or the University of Notre Dame Environmental Research Center and then we have purely private sites so we're you know such as Steiger-Walter that's Jones Environmental Research Center so we really have to kind of consider different methods for data collection about land use history for these different types of land owners and what we found is that for federal and state owned properties we're oftentimes able to query across different existing archives and databases and oftentimes these even come with geo located geo reference data for private sites we found that it's been most helpful to pick up the phone and have direct contact with some of the neon site managers who can then help us to understand who the different owners are and how to best approach them we're trying to be very mindful in this process of not kind of overtaxing land owners because we don't want them to be to kind of regret being a neon site because they're getting bombarded by by requests from people and then with extensions of universities we found that it's kind of a combination of direct interviews with people at the universities and then also accessing databases and archives of research that's been performed at those universities and I just want to recognize that this process of data collection is very much grounded in the social sciences and what we've been using really is a snowball sampling approach and so especially in those instances where we're reaching out to landowners you know we contact the neon site manager first and then they give us leads and we follow those leads and sometimes the leads lead to nothing and sometimes those leads lead us down a whole other rabbit hole of people that we can talk to and this you know to a scientist who maybe doesn't do social sciences might sound very you know biased but it's important to recognize that this is kind of a well-known social science method called snowball sampling but you do have to recognize that you have these kind of social networks and you might have someone who has who's very networked that can give you a lot of resources but there will be some bias in in who you're actually getting information from so we just like to recognize that that's that's the approach and that that it's used commonly in the social sciences but it does have its inherent biases in terms of the social networks that you're able to reach through this approach. So what did the data actually look like? So in terms of the type of data that that we found this is from a site called Tree Haven and it's Potawatomi land and you know sometimes we just find very old maps that we're able to get even sometimes we've we've made trips down to the National Archives down at University of Maryland College Park and we take scans of old maps and sometimes those old maps will have information on things like harvests and just the date as you can see in the polygon. We then have to geo-reference those old maps to then make them into shape files so sometimes it's a matter of actually converting paper maps that have been hand drawn into into digitized shape files. Sometimes we'll get an old map that's maybe from a publication and and in that case we have to take the we have to link the information from the old map and geo-reference that as well as pull out data on relevant data from different data tables that accompanied the publication. So this is an example from Oak Ridge National Lab down in Tennessee which is on Cherokee land. Other times we're able to get information not necessarily on on land use but on historic land cover that that we're able to to glean from aerial photographs. So this is a nice example of a collection of aerial photographs that were made available to us from the Smithsonian Environmental Research Center in Maryland and here once again we have to take that aerial image geo-reference it and then classify do a classification of the different land cover types and in instances where we have land cover information rather than land use we're we're using the land cover designations of the national land cover database. So this is just an example of what a time series from those maps from CERC would look like going from 1938 up to 1976 and you can see that the the land cover type is shown by the legend that's on the right hand side of the slide and forest is the really dark green ranging up to a lighter green which would be open pasture or grassland space and then you can see structures on there as well and what you can see overall is that there's been a lot of reforestation in this in this footprint over time. Other times just to give you another kind of taste of the types of land use data we'll have information on grazing and this is an example from the Santa Rita Experimental Range which on the left hand map you can see all of those black lines are the pasture fences for the different pens that the cows are put into and on on that map those little red dots indicate all of the the neon sampling plots and so that cluster that you see up in the right hand top corner is are the vegetation plots that go around the tower. So the good news is that for Santa Rita the whole footprint tower falls in one pasture that's not always the case for all the sites though in terms of it falling in one area of land use but what you can see is the neon sampling plots really fall in a lot of different pastures and on the right hand side here you can see the overall grazing intensity and animal units per hectare per year for each of those pastures and the purple here you know the the dark purpley blue color means a higher intensity of number of heads of cattle per year grazing a particular area gray means that those were years without data but what you can see is that over time we've had a lot of variability in the amount of grazing at those sites and the and we're able to you know at many sites this is also just illustrating that we have data that are coming up until present so although i'm focusing a lot on the kind of historical aspect of this database there's also a lot of rich information on what's happening currently that could then feed into thinking about these kind of coupled human natural systems at these sites so in in coming up with the database we're thinking a lot about data provenance and as you for those of you who've worked with neon data you know that there's different level designations for the different types of data products that neon has so you some of it are L zero data which tend to be more true to the raw form of data that were collected L one data might or level one data might be a more of a summarized version of the data and so neon uses this level designation as do many NASA data products as well and it's just a way of kind of keeping raw data separate from more derived data so that you have an idea of how the data is changing as people do their analyses so the level zero or L zero data in this database are really just our raw shape files of those different digitized maps and so it was really those kind of shape files I was showing in the previous slides and you know a challenge with with the data that come out of those kind of raw shape files that we collect is that the variables differ by site so as you can imagine there's very different types of land use across all of the different neon sites some sites are forested and being harvested other sites have different amounts of grazers on them and so forth and so these L zero data shape files really reflect the variables that are from an original source of data we're also generating an L one or level one data product which will be standardized disturbance and land use variables across all of the sites and so we're we're generating shape files for all the sites that correspond to the L zero data but have information on disturbance type the start and end data the disturbance the time since the disturbance the intensity and frequency and in cases where it's applicable the the national land cover databases for land cover classes for land cover in terms of the progress of the database we've been focusing our efforts mostly on the terrestrial sites with some effort that I'll talk about in just a moment on the aquatic sites being slightly different but so far we have information for 33 so about three quarters of our terrestrial sites and we're still plugging away and trying to get more information on the other sites in terms of data origins most of the data are coming from existing shape files from open source databases 22 percent are coming from site contacts and then 14 percent are coming from digitized figures or aerial photos and you'll see that that doesn't sum to 100 because some some data sets actually are coming in both can be categorized as both within a site in terms of what the disturbance types look like here's an example of the forested sites across neon and I should just note that a student in my research group at Bryn Marr who is also an REU student at Harvard Forest Sam Oliva's Maizha did an incredible amount of work on this project and you can see that in terms of disturbances at forested sites of neon most of it's due to cutting and then the next kind of disturbance that's most common is planting regeneration and we have some wildfire but this gives you a sense of kind of the types of disturbances that we're seeing Francis Romero did her senior thesis on the rangelin sites at neon and in the rangelin sites the the most common disturbance type is grazing but we're also seeing a lot of prescribed fire as well I did talk about some challenges that we've had in terms of making this database and these are some of the things that we're kind of chewing on as we as we finalize this and before making it publicly available and you know some of the challenges that we've had are that at some sites the data exists really great fine scale information on land use but it's proprietary and so an example would be the Jones Environmental Research Center has collated a great wealth of information about land use but they're just simply not willing to share it other times the the data might be proprietary because they're owned by a park and maybe sensitive if they were used for for a study endangered species so those are some of the issues that we've hit and then finally I just want to also recognize that this is a lot of the data that we've been able to collect for this project has really been post-colonial history but we'd really like to decolonize this project and have more pre-colonial history so if there's anyone out there who'd like to collaborate on that please please reach out to me. I'd like to just end with a couple ways that we've been thinking about how we could synergize information from the AOP and land use and you know one of the ways that we've been thinking about doing that is getting information on kind of finer scale features that are provided by the National Land Cover Database and one way that we can do that is using the neon lidar data so this is a hillshade map that's generated from the neon AOP lidar data and it's really cool to look at this fine resolution digital terrain map because you can actually start to see features on the landscape such as roads and rivers and stream cuts and so what we've been putting together is a linear features database for the neon sites because even though you have information like the census roads and the national hydrography data set on streams you're able to capture shown in pink a lot of undetected linear features by looking at these high resolution lidar maps and so just to give you a better sense of that here is one of those areas zoomed in and you can see where the US census road shown in brown lies and you can see kind of the line for that misses just a little bit but you can tell more or less that there's a road there and there might be edge effects that are important for thinking about your ecological system you can also see where the streams are but here you can see an unmarked road and so there's a lot of dirt roads that go through neon sites that would go undetected otherwise and roads as we all know can be dispersal corridors for invasive species you can have edge effects around a road and so it's very important to kind of know where these linear features are so we've been going through and we've done this kind of exercise of drawing additional linear features onto all of the neon sites AOP digital terrain lidar maps and we've gone through and also looked at bare earth images from Google earth which has high resolution imagery to actually confirm that we're not missing a lot of roads and we're doing a good job in terms of that validation we're also using some of our historic data sets as well to validate this database and we should have a data product soon to be released that should be coming out probably in the beginning of next year and once again that's for all terrestrial and aquatic sites with AOP data for the neon sites okay the one other way that we're thinking about synergies between remotely sensed data and our land use data is in thinking about how you can remotely sense disturbance so it kind of in tandem with this work another piece of our project that the Jasper van Donink from Phoebe Zarnetzky's lab at Michigan State University is doing is taking the AOP data and the Landsat data and trying to come up with disturbance metrics for the neon sites as well as all the places in between so really thinking about how you could use a remotely sensed data to kind of take inferences from neon sites and then scale them out and fill in all the in-between places that are in the country and remotely sensing disturbance presents one challenge in that you have to have some sort of ground validation information and so this is just an example of how we've been thinking about how we can use the information from our land use database to actually provide some level of validation for from remotely sensed disturbance layers and this is just an example from Bartlett experimental forest which is in the northeast domain and Bartlett is in the New England temperate forest and has has a long logging history it's a U.S. Forest Service site and this is just an example of some of our land use data for Bartlett and all of those different sections that you're seeing there are different compartments for forest service compartments where there's different management and we have we have essentially the logging histories for all of these different compartments and some of these compartments go range from being unmanaged so they haven't been logged to being very highly managed and logged repeatedly and so what we've been trying to think of are ways that we can use our land use data that that we have information on to then validate some of Yaspers disturbance layers and as I mentioned before Yaspers use using Landsat which is a NASA Earth Observing Satellite that that generates the national land cover database and also AOP data and doing essentially a time pixel by pixel time series analysis to see how the spectral signature of each pixel changes through time to then have have it a way to detect change and model change detection and he's one thing that that Yaspers really thinking about in his project is how not just looking at a pixel through time but also the importance of really thinking about the space that surrounds a pixel so the neighborhood around a particular pixel that you're modeling and he's been playing around a little bit with an REU student Chakata Hart from Michigan State University thinking about if you could actually do not just a temporal segmentation of the spectral signature through time but actually a spatial segmentation that's kind of object based and you're thinking about how that spatial object changes through time to then detect disturbance and also attribute that disturbance and this is just you know to show you that these are the kind of spatial object segmentations that that Yaspers been coming up with these all those like little boxes and what we've been doing is using those to kind of taking those and comparing them to our validation data so that's just to get you thinking I know a lot of people are using the AOP data for Neon and just to get people thinking about how this land use database can be a resource for thinking about getting disturbance information or change through time information and linking it back to on the ground measurements so this is just another in situ data resource for people so I'm getting close to the end of my talk which I'm hoping will have you know plenty of time for for questions because I always think that's the best part and you know what I'd like people to consider as the takeaway from this talk is that is really to go back to that planning document from 2006 for Neon that was a white paper that was generated by folks at Harvard Forest who are thinking about kind of this goal of understanding climate change and coupled human natural systems and then being able to make projections so that we can understand what might happen to our ecosystems and biodiversity in the future and I really think that if we're going you know if we're going to be trying to understand and forecast out into the future we really need to be thinking about how the ecosystems and biodiversity that we see may have been influenced by the past and that can actually help us to inform how we're going to model out those projections and think about the future and what the future holds for our children and I just want to give take a couple last minutes to give a little bit of plugs so this is an ESA Ecological Society of America water cooler chat that's going to be coming up in October it's graciously organized by Theresa Marad who is run seeds for the ESA society and it's part of a series that is is led by the landscape exchange network RCN that's led by Andrew Elmore just really thinking about how to get social data into thinking about neon sites and we'll be talking a little bit about this project you know just for like about 10 minutes and then we're going to talk a little bit about how environmental data science is a way to actually broaden participation in ecology and environmental sciences and there's been a lot of talk this year about how field sites can be a bit unwelcoming to people and especially people from traditionally underrepresented groups and and data science kind of provides a nice a nice avenue for for I believe bringing together more people from diverse backgrounds in ecology so we'll be talking about that and that will feature Dinesia Bill St and NSF postbaccalaureate fellow who I've been working with and also Sam who did a lot of the the work on the forested land use history and I also just want to give a quick plug that I've just moved to you main it's very nice here there's a you know it's about an hour from the coast an hour from Katahdin and there's a lot of positions open in my research group that I'd love to hear from people about we have two postdocs all of these are kind of related to neon in some way so there's a postdoc on neon land use to continue this work and think about interest specific trait variation in niches there's a phd on linking neon phenology data and herbarium data to to think about phenological sensitivity and then there's a NASA funded postdoc on scaling forest diversity and using neon AOP data along with other NASA data sets so please reach out if you have any questions about that and please help me spread the word we'd love to get some great applicants I'll start reviewing applications on September 30th and I just want to end by saying thank you to um to especially to the neon staff so I thank you Sam for organizing and um and Brigid for the invites um I'd also just like to recognize um you know many long-standing collaborators Eric Sokol, Kate Tebow um if I'm not naming you I apologize but there's just been so many people from neon who have been so helpful for various aspects of this project and so many others over the years and I just really value the insights and um and wisdom from the neon staff and I also want to thank the National Science Foundation for keeping neon running and um for also funding our projects so at this point I'd be happy to take any questions. Thank you so much uh for that wonderful talk Sydney very very interesting work um I am not seeing any questions in the Q&A box but for all our attendees please if you have some questions do either go ahead and type them in the Q&A or you should have a feature where you can raise hands at the bottom of your zoom screen and we can unmute you and give you a chance to ask your question to Sydney if you would like um Brigid or Eric did either of you have a question you wanted to start with or I could I could throw one out there yeah Brigid um I just have a quick logistical question um so Sydney are you looking at are you providing maps of the entire flight area or are you focusing mostly on the observation? That's a great question so we've we've for the detailed land use data it's it's where the observational data are for now I'd like to kind of write another grant to to bump that out it's just we had to kind of we we we bit off a lot to chew through already um but for the um for the roads layers that's the whole AOP. Okay great and then we do have a question from the Q&A from Rachel Swanson she's asking curious how this work uses or could use the neon site management and event reporting data for current disturbance events? Yeah so that is a fantastic question and um I've been talking for years with Mike Sanclements about that data product um and and thinking about it and so I think that you know what I the the short answer is we haven't used that data resource a ton yet um but as I'm kind of pivoting from kind of getting a broad scale brush of the of the database um created that's what I want to be thinking about moving forward um and also you know I and I'm I'm also interested in kind of talking with um with the neon field crews too about like how that reporting actually happens and thinking about ways that we can make it um like more just as neon uses standardized data protocols protocols for everything else how we could maybe standardize how field crews are actually thinking through that process too so I think that's like definitely like my next step of things that I'm wanting to think about and yeah sounds like let's do it all right um yeah but that's definitely it's an untapped resource and I also think that it's a resource that with community guidance could be um maybe maybe even more impactful. Kelly go ahead and ask your question. Yeah hi hey Sydney this is great um thank you so much it's amazing to see how much you've done um and so glad you're doing it and I know you've used the watershed boundaries for the road layers um but as an aquatic ecologist I and I know this is asking a lot but like obviously the whole drainage area impacts the aquatic site and so in addition and some of our watersheds are small enough that AOB does fly them all and so it is there a future plan five ten years down the road to actually do the land use history of the watershed not just roads. Yeah that that's definitely something that we'd love to do and I feel like I so I'm a little biased because I do mostly terrestrial work although we do have an aquatic scientist Angela Schecker in our group um but because I was kind of tasked with the land use I've been a bit biased with the terrestrial goggles um but that's definitely something that we're thinking about um and we would you know we the other piece is so a few ideas here um you know in some instances we are limited a bit because of the footprint of of the AO of the AOP um and just for for aquatic sites just if I'm thinking about like roads layers and stuff like that um and so you know we're thinking about ways that we could essentially kind of potentially use like drones to get finer resolution on the um you know on a watershed area and use um artificial intelligence to draw the roads in the future now that we have a validation data set so that's one piece um the other piece that's going to help with scaling up beyond kind of the sites that we have um is at least for land use from 1984 to present Jasper's disturbance layer so he's he's been like agonizing over getting that set up for running on the whole continental united states and so that would then be able to enable us to scale even you know at different size hucks moving up um different watershed sizes designations and then for the kind of really really um like historic fine land use and stuff yeah I mean I would I'm I would love to kind of keep working on this and the nice thing about neon is that it aligns really well with some of our careers you know so this is kind of like something I'm hoping to continue working on until I'm an old lady um and but I hope to get it to you know so that would be kind of like you know a couple steps further um in terms of thinking about scaling out the land use great thank you thank you looking forward to it I thought Dave Barnett had his had his hand up a little while ago it kind of went up and down so should we call on um Dave Barnett's last yeah yes absolutely um wonderful talk as always I um involved with Rachel and others on that um site management data which is new as a formal data product and as with many of our data products we have these external advisory groups technical working groups and we don't actually have one of those for that data product yet with something we've talked about and so um getting more you know more feedback from you and others like you know to diversify would be a really helpful thing I think um and we have thought a lot about how we get that data and I think just like how more in the historical information that you collected we run into a lot of different um situations by site with respect to how much they know about what's going on and you know the scale at which they're willing to or not willing but uh capable or actually actionably can report information on so all these things I think directly are relevant to information you're collecting so I just wanted to make that that we will likely reach out to you for more on that um my second question is I've had a lot to do along with Rachel on placing our neon terrestrial plots at sites and so it was fun to see and I've looked at this kind of thing a lot um you know how the plots showed up on the I forget if it was a harvest map you showed it while I think it was at CERC or somewhere in domain too. So Santa Rita was like little red dots with the pasture fences yeah and we've looked at this kind of thing a lot and you know early on I was talking to some folks at Kanza about it and they said well if you don't pay attention to the different because there they have very specific spatially grazing and fire various combinations of those if you don't and we just don't have enough plots to capture all those so it was interesting to see that map but I thought just if you could speak a little bit more about how you've seen those things overlay um on other you know history and how we've done it would be interesting feedback. Yeah so I mean I think that you know and my understanding from the the terrestrial the TOS design the um I'm sorry I'm speaking of neon acronyms the terrestrial observation system design is that um hopefully I got the acronym right um is that that they uh that they strata my understanding from I think it was your paper Dave um was about how you stratified by land cover type right and so um I think yeah there's definitely in instances like Kanza um also places like Ordway Swisher where you have especially where you have these like prescribed burns um sometimes even nested within like an experimental design that land cover doesn't quite capture that disturbance um and so so I mean I think I'm and this isn't me being critical at all I'm like so grateful to everyone at neon and all of the thoughtful you know work that you've put into everything since the inception of the observatory um I just think that you can't include everything right and if you had to have that kind of land use analysis package group um then you could have you know thought through all of this but I do kind of think that the NLCD classes don't always capture that kind of disturbance and um and I do think that there's opportunity moving forward you know to to also be thinking about how the AOP could be better leveraged um to capture some of these disturbances like it it is challenging to kind of have people reporting right um and I know that there's a lot of turnover in field crews from summer to summer and keeping that consistent and like what you deem a disturbance and all of that um but I do think that you know there might be a case that could be made for more frequent AOP flights that could capture maybe some of this disturbance um so yeah just something to think about and then um yeah like the the phenocam data is interesting too I remember talking with Kyla Dahl and um you know she was about to go from Michigan State she was going to go sample down at Talladega um for one of her um eager awards uh and she said like she was planning her sampling trip and she like just out of like on a whim turned on the phenocam or like checked the the phenocam information for Talladega and it was on fire and she was like oh I guess we're not gonna go there for a while so I mean there might be other ways that we could leverage different data sources um but I think a twig a technical working group on this topic would be great um and that lens RCN might be another good place to recruit people I'm on the steering committee for that so I'd be happy to help set that up okay wonderful thank you yeah I think a lot of that comes back to sample size and with so many things at NEON in terms of the plot distribution trying to do things and maybe for better or for worse we so often try to do things consistently apply things consistently across sites and so to think about all these um you know different sites are obviously exposed to different disturbances and then to stratification always adds complexity to the design so um to have some double stratification further complicated it so I appreciate your answer and I know you're not being critical so you know it's one of those things we did wrestle with some and certainly we didn't get everything right but we do have a racial now we're sort of working on a what did we miss paper and looking at other things besides nlcd to see how we're covering um and that's been pretty interesting so maybe we'll even run that past you too before we we submit yeah that sounds awesome thanks thanks Dave Chris jump in for the next one okay hi uh well thank you it's really cool to see researchers that are kind of um uh maintaining like some of the historical pieces of the observatory that might be otherwise forgotten so that's really great my question is just um do you I know you haven't released this database yet but do you have any like papers that are rearing to go or nuggets of data that that will like that will use this data to explain any variation in any of the observational data we collected before yeah so um the answer is yes there's a lot of stuff in the queue um I was I kind of like on a whim became associate provost for a year and it kind of sucked the life out of me um but we've got papers looking at mostly at organizational diversity and interest specific trait variation responses to these um and just because I I'm more of a community ecologist uh so Dinesh Bill St the post back who I mentioned in that ESA water cooler chat we've got a really interesting paper that she's just worked that that she's worked up the analysis for and is going to write up this fall on the effects of roads the direct and indirect effects of roads on the microbial diversity across sites um and we've got you know a paper in the queue from REU on plant diversity we've also got um like a couple papers on like just the the history of the neon sites um that we're hoping might get a a bite from ecological monographs um since it's they're very kind of nice long stories about the sites and then the the data products um but the answer is yes there's there's a lot in the queue um and if um and I'm also you know I'm hoping that we can get this out sooner rather than later so other people can also be using it to answer cool questions and stuff and I'm also happy to kind of like talk with people because we've got a lot of knowledge at this point about the sites um synthesized thank you so much I know there's more questions we could go on but we're at the top of the hour thank you so much the knee for a wonderful presentation thank you all for coming we'll see you again here hopefully in in a month's time for our next seminar bye everyone