 Okay, hello everyone. Sorry for the slight delay. Welcome to the CSDMS webinar. Today I'm going to be talking about making our models fair, which is findable, accessible, interoperable, and reusable. And I'm looking forward to this webinar and hoping that you'll interact and maybe ask some questions and take part in some of the polls that I've set up. So I'm going to start my timer here to make sure I stay on time. So my name is Leslie Shee and I am at the US Geological Survey as the coordinator for the community for data integration. But this talk is not necessarily about how USGS deals with their models or anything that USGS does. I see it more as geared towards the CSDMS community, which does include federally funded federal employees and also academics. So just keep that in mind as I'm talking and if you feel like maybe it's not geared towards you, I'd really love to see some comments in the chat to make it useful to you. What else I would like to thank Nicole Gasperini, who is the co-chair of the Terrestrial Working Group and she was helpful in getting me more involved with the TWG this past year. So welcome, especially any people who are from the TWG. I think the last thing I want to say on this title slide is that we have the word models here in the title, but I just want to make sure that here I when I use model I kind of mean numerical model or computational model as opposed to the algorithm or equations behind them. However, using models here is also pretty loose where we can think about models or other data or just little scripts as well. So there'll be something for all of those things. Alright, so why don't we get going. So let's start with the intended outcome of today's presentation. So I would like at the end of this webinar for you as an attendee to feel comfortable being able to explain to somebody else what fair means for CSDMS related models. Also, for you to have a bunch of resources where you can go to find out more about what fair exactly means because it's a pretty new term but it's starting to be used more and more. And finally, for you to have some strategies for making your proposals and your publications more fair savvy. So I think that from talking to people. They're not really sure what fair means or what they should be doing with their code and their data. So just having a couple of these strategies may put you, you know, ahead of the pack. And so then you might also be thinking so why are you talking to us about fair. So just a little bit of background on on why I'm talking to you about fair. So I do have a Earth surface processes background and geomorphology. But now I spend a lot of my day to day kind of also in the data community, especially in my position here I'm in the science data management branch of the USGS. I have participated in several fair related workshops and meetings in the past year or two. So I kind of think of it as like bringing this information back to my friends at CSDMS to see what the data people are talking about and how it might affect you. Again, I just want to emphasize I'm kind of assuming that most of the people here are NSF funded, but there will be a little bit about USGS and federal agencies. And this will just help you like, you know, whoever you are you probably need to collaborate with different entities from time to time. So this will help you figure out what what they know about fair and what their requirements are. Okay, so I would like to attempt to start with a poll and I'm going to put some stuff into the zoom chat, which you can go and try to answer this poll. So somebody just chat at me or say something if this doesn't come through but in the zoom chat I just added a URL or like two ways to get to this poll. So we're using an app called Slido, which some of you may have used before but you can either go to it on a browser on your laptop, or you can go on your phone and just go to Slido.com. And if you don't use the direct URL just go to the event code 3580 and there'll be two tabs questions and polls. So if you go to polls, then you'll see a poll that says on a scale of one to five how comfortable are you now with explaining what fair means for your models and data. So I can see some people have started typing that in so that's great. So I'm going to put my response here and send it. And while you're working on that let's see if we can take a look at the results. All right, so hopefully this is showing up on your screen on the zoom share now it looks like on a scale of one to five. We have a tie between, you know, one three and four so some people are not comfortable on one end of the spectrum, and some people are kind of in the middle they feel like three or four they could talk about fair. Nobody has said five yet which is the most comfortable which is interesting. Let me just leave that open for now and then maybe we'll come revisit that at the end of the webinar. Thanks for taking part in that people and if you have questions or trouble getting to this poll. Feel free to just put something into the chat so we can try to get that fixed for you. So I'm going back to the slides. Okay. So thanks for the interaction on a note that if you are there in the Slido app and you go to the questions tab you can type a question at any time. So I'm just going to type hello and send it and it's a completely anonymous question so you can send in comments and questions. And there's a place to put a little thumbs up if you really want that question answered or you agree with that comment. So I'll try to keep my eye on that and look at that at the end of the webinar. Okay, moving on then why don't we talk about what do we mean by fair. Like that's a good place to start. And there was a paper that came out in 2016, which is the most cited and sort of it actually says in the paper. This is the first published work on fair findable accessible interoperable reusable. And I think it's important to go back to this original paper because since then fair has really proliferated. And at a recent meeting there's actually one of the authors on this original paper was really pissed off that people had not it seemed like people had not read the original paper and understood what they meant by fair there. So just a few quotes from the paper. So this article describes for foundational principles find ability accessibility interoperability and usability that serve to guide data producers and publishers so most of the CSDMS members would would be considered as data producers and we'll see what they mean by data on the next slide. So this next highlight here is that these are meant to be a concise and measurable set of principles that we that they refer to as the fair data principles. So it's important that there's metrics and ways to measure how fair something is. Next. So an important point is that the fair principles put a specific emphasis on enhancing the ability of machines to automatically find and use the data in addition to supporting its reuse by individuals. And this has become one of those almost by modal issues that if you go to a data meeting and people are talking about fair. Some people just their background and their discipline. They're thinking more about human readability for fair while others the type of data that they are dealing with really they feel like it's just not scalable at all if you're just catering to humans. It must be machine actionable for fair. So just keep that in mind as well. And then finally the last quote I want to share here is that these principles apply not only to data in the conventional sense, but also to algorithms tools and workflows that led to that data. So that's spelled right out in the original paper about they say you'll see data most of the time but the models that we create here as part of CSDMS and that data the algorithm the other tools that is all included here. Okay, so I see a question on the Slido is data produced with a model included in there. And I would say yes seeing what they are in what they've said here at least in the original authors viewpoint that would all be included the supporting stuff for the scientific paper. Okay, so now this is a screenshot of a table from the original paper and you can see how concise this is. It's a very short table and for each of the FAIR there's kind of four sub principles or things to be concerned about. And they kind of all start with data or metadata or a protocol and just reading through this list you might think, wait, this is not. I don't really understand how I'm supposed to implement this or measure it or anything. That's fine people. That's why there's so much being written about and talked about fair right now because these are supposed to be universally applicable to many different disciplines and many different types of data models information. So they're very concise here and each community kind of has to figure out what that means for them. And there's also a lot of good resources further explaining what these mean, which I'll get to in a little bit. Okay. All right, so just to review hopefully now we're kind of on the same page as the original authors about what we mean by fair their foundational principles that are supposed to guide both producers and publishers of data. They're very concise and measurable. There is an emphasis on the machine's ability to find and use data so computers applications, etc. And data is used loosely to also mean algorithms tools and workflows. All right, so now a little bit on why are we talking about fair. I mean, I know you actually attended the webinar so you're not a fair hater or anything but just a couple of slides explaining the motivation why we thought this would be a good topic to talk to the audience about. So one of the projects that has sort of followed on this original paper and idea is a project called the enabling fair data project. And this was funded by the Sloan Foundation and about the past two years there's been activity here. The American Geophysical Union AGU convened it, but they're not the owners of this project. So what they did is they had a bunch of meetings where publishers, researchers data repository people and funders got together and talked about what do we need to do to make data and models fair because this is really helping us do our science better. And one of the outcomes here was a commitment statement, which had signatories and you may recognize the names of some of these signatories so AGU you navco open topography USGS publishers like Wiley and Elsevier they are all aware of this commitment statement and they have signed on to it. And one of the things that means is that, you know, for publishers. Everyone has their own different requirements but this project is trying to get the publishers to have a somewhat standardized or at least similar guidelines so they have author guidelines here. And they're striving to get publishers to address the same set of topics related to publication of supporting data or codes. So this is something you may very well see that if you in the near future are submitting a paper for publication you say you may see this new wording that is has been influenced by these author guidelines. And I think I can throw some of these URLs into the chat. They're also at the end of these slides which we will make available, but there's a couple links there where you can find out more about this project and what the author guidelines and commitment statement are. Okay, and then also just a quick search at the awards that NSF is funding. We're starting to see fair and the titles of these projects. Most of these were proposed within the past year, because it's pretty new but there's workshops being proposed there's frameworks being proposed there's all sorts of things being proposed that are including fair so it could be a good thing to know more about. So now I would like to move on to sort of this pre webinar poll that I distributed, and you can get an idea of where your colleagues feel they are and understanding fair and implementing fair. So the first question there was, have you ever heard of the acronym fair and this was kind of distributed on Twitter so you know it could be anyone but we'll see how many of these people were from CSDMS. So, yes, 40% of people have heard of the acronym. 36% of people did not. Some people weren't sure maybe they part of it. And then I always love just leaving free text for people who want to add other comments so somebody says I haven't even heard of the term interoperable. So just FYI, there is a definition of that in the Wilkerson paper, I think Wilkinson paper and that's the ability of data or tools from non cooperating resources to integrate or work together with minimal effort. At least that's what it means in the context of fair. All right, so we have like, you know, there's a good amount of knows but there's also more yeses. So that's good. The next question in the survey was on a scale of one to 10. How well do you think you make your models fair. And this FYI, the N number of responses was 20 and I think about 15 of those were CSDMS members who, you know, work with. Supposedly working models more. Most people kind of put themselves in the middle, but we have the whole range from one to 10. The average there is just a little bit over five. What's interesting is that for people not in CSDMS they had a slightly higher score they gave themselves a slightly higher score. So maybe CSDMS members just know that they're not really doing all they could to make their models fair. Probably we need more responses to draw a conclusion like that. Okay, so bottom line whole range, but kind of in the middle. And then we also asked how well do you think you make your data fair as opposed to models, and this was slightly higher so for data, the average was about seven. And again, the people who are not in CSDMS actually rated themselves a little bit higher than those that were in CSDMS. And you know, I think I know who gave themselves a one don't be so hard on yourself. I'm pretty sure that it's not a one for a one out of 10. So that's just kind of fun to see where your colleagues think they lie on the scale. Okay, a question for interoperable good getting back to the interoperable quest definition, the basic model interface part of CSDMS does that count as interoperable I'm going to say yes that that does definitely at least the way I see it count as making different codes interoperable. Thanks for those questions. Alright, so the last thing in this pre webinar survey was any other comments or questions that you would like addressed in the webinar. And we have a bunch of comments here so here's some of them. It's a moving target there are multiple unknowns to that I would say yes, I agree I think it's an evolving target. The paper came out in 2016 and it's really starting to catch on now the phrase. So I think things will probably still be changing for a while. Another comment in blue so fair for models and data could be pretty different. And also I agree that's right so there could be different considerations different tools that you'd be using. Just like it's different for the different disciplines and that's one of the reasons why the original principles are so concise because it's left open to interpretation a little bit. Also, I think one of the quotes I liked in the paper was that these are aspirational principles. So just keep that in mind. So effort makes it a barrier. We need more carrots and sticks for motivation. Another comment was that there's no reward for taking the time to make your stuff fair. And again, agreed and this is also recognized by people who are endorsing fair so just be comforted by the fact that the authors of fair do not think that research scientists should be shouldering all of the burden for making their data and models fair. But until very good tools are built or other infrastructure or other staff somehow come into being it will seem like another thing that you need to do, which is that's the way it is but I think things will get easier in the future. Okay, and last comment here is just it's intimidating now especially because some people there. Some people in the CSDMS community we were not trained in computer science or programming as students we just kind of got into it and we use it to do our science and this could be pretty intimidating. So, hey, let's try to just talk amongst ourselves and figure things out and not just be intimidated. Hopefully, you'll find out that the people who talk about fair are friendly people and they want to help you out and figure out what you need. So to help with this comment let's look at some resources for understanding fair. Okay, so going back I showed this already the enabling fair data project, but this I'm going to throw a URL into the chat. And there I think that's a really good resource overview to see what the activity has been recently. Another thing to understand that this is a global. The fair data initiative is kind of global so those in Europe may be making resources and information that would be a little bit more applicable to Europe those in the US. This would be an example of one of the efforts in the US that would be a little bit more geared towards those in the US. Another good resource is the go fair site, which I believe stands for global open fair. And this is one of the ones that is in Europe. And I like it because they have a lot of explanation and sort of more plain language plainer language than the original principles. So if you take a look at that second URL there you'll see an explanation. If you want to go into that detail you can go and see the explanation there. I'm not saying that everyone needs to go there to understand all of the principles. There's just two of many, many, many resources that are sprouting up, but I think those are two good ones to start with. All right, and now just where can you maybe some of the groups are already part of or that would be easy to join will help you to stay up to date on the latest for how to make your data and models fair and just a few of those are listed here. So this webinar is just me sharing many URLs with you. But if you aren't involved in these places, I think there's a lot of great people gathered in these places and these are just some examples there's probably many more. But GSA has a geo informatics and data science division. And sometimes their meetings at the annual meeting for GSA will have something to do with fair that's already happened in the past couple years. The one on the top right is the AGU Earth and Space Informatics focus group. So probably a lot of you have a primary group and either Earth surface processes or tectonophysics or another group, but you might want to keep an eye on the SC sessions at the fall meeting. If you are interested in, you know, what are some new resources, what are ways to make my life easier, making my data and models fair. In the bottom left, we have the earth science information partners, which I think is a really great organization that's bringing together people from federal agencies and academia to talk about lots of things. I think their, their tagline is making data matter. So this is another place, a forum where people would be talking about resources for fair and how to meet them. And then finally, I have to put in a plug for my own organization, which is the USGS community for data integration. This is a little bit US, more USGS focus, but especially if you have collaborators in the USGS, we do have a lot of broad topics that address things like fair. In fact, we actually have a group of people who are kind of looking into how do current USGS policies mesh with the fair principles. If you have any other suggestions of where you think is a good place for people to gather and talk about the stuff. Besides CSDMS of course, feel free to put that in the questions you can add a comment. Okay. There's a question that I think I will leave for a little bit later. And we'll keep going. So the question of this presentation, I want to tell you about three decisions you can make to increase the fairness of your models or data in your proposals and publications. Because as I said, and as you saw from the comments, it's, it might be a moving target or a little bit fuzzier people might not completely understand it so just having these three things in your pocket could really help you to look like, you know, more about fair than others, which may very well be evaluated. Okay, so the things we're going to talk about are where to share your code, what license to use, and what information would you need for your code to be reusable for yourself. So some considerations where to share your code. So these could be numerical models, these could be little scripts that you used in your data collection and analysis. Some considerations I think you just think about. So it's, it's better if this is a place where your colleagues usually go in your discipline because it'll be more likely to be discovered kind of serendipitously as opposed to searching for it. Ideally, it's a place that is set up to be indexed by different searches so whether it's set up to be easily indexed by Google search or indexed by other earth science searches, which of which many are being built right now. That would be a plus in my mind of where you're going to place your code. A place that allows versioning and persistent identifiers like digital object identifiers would also be a plus. And in the future there may be more and more places that say explicitly that they subscribe to the fair principles or commitment statement, and that would be one of the things that help you make your stuff more fair. So not to be to state the obvious I think one of the places which I think is a great place to share your code if it fits into their scope is CSCMS and the resources that they have set up. Okay, so we'll talk about that a bit more but just remember these are things you can consider if you have a couple options. Maybe you don't really know where at all to put your code that would be a question that we can discuss more at the at the end of the webinar maybe. Okay, the next thing to talk about a little bit is what license to use. So there are, I can't get into all of the details today, but there are some great quick, you can skim through these other slide decks, or these other resources at some time that I just put into the chat so you can figure out about licensing and again this is not something that I was ever trained in I just hear about it all the time now, especially being in the data community so if you're not in the data community you may not hear about this very much. But it is a, as you can see from this slide from another presentation, you know they get very something about licensing, whether that be like scared or intimidated or just very opinionated. So aside from somebody else's talk about considerations when choosing a license so these are some of the things, like for me, I would be thinking like why, why do I need to choose a license and it's a little bit easier for me now as a USGS employee because we need to make all of our things completely open in the public domain so that's like an easy choice for me. If you don't have a rule like that then you have many many choices. And here are some considerations you can think about so if there's rights that you want to retain or grant to others. If there's compatibility with software under other licenses, what happens to derived works and patent considerations and also expectations of the community you want to engage. So this presentation wasn't necessarily to, you know, a group of earth scientists who write models for the earth surface but this was to sort of high performance computing group of people. And these are the things they had considerations for if you want to look at something geared towards our community then wow look at that CSDMS has a page of information on their wiki that talks about all of these things and which license should you use. Now also on this page CSDMS tells us which license they use for their products. So maybe if somebody from the CSDMS team is here on the call they can either type something into either the zoom group chat or maybe at the end of the call. Tell us a bit more about which license you chose and why that would be helpful. Okay, and now the third consideration, which will be different for all of you is what information would you need for your own code or data to be reusable to yourself in the future. And I'm sure this has happened to many of us where five or 10 years after we've written the original you know type that script that happened to work. Somebody comes back and said hey I want to know exactly what you meant right I don't understand something in the paper and you know what would make it easier for you to go back and just easily say hey here's my nicely commented and described code. Where you can look at it and you can understand or you yourself could look at it and give an answer without thinking oh my goodness. I wish I understood remembered what happened 10 years ago. So I don't have a lot more information on this one but I think it's a great thing to think about. Okay and thank you mark so the CSDMS software engineers frequently use the MIT license which is permissive but requests attribution on reuse and I think that is a very logical and probably popular thing that people would want in the CSDMS community. So thanks for explaining that. Great. So those were three things you can think about when you're writing your proposals when you're getting ready to submit for publication. And here's another comment that came in from the survey which is can you show an ideal example of fair and compare it with the usual. That's a really good question and that might even require more time than just one or two examples in this webinar but why don't we see what what is the usual and what we see in our community. Well here's where I really want to talk about the CSDMS role in making models fair. This topic of fair kind of came up without anyone saying specifically like we should talk about what CSDMS does for making models fair but if you think about it a little bit. Almost everything that CSDMS does is helping to make your models and your products more fair. So here's just a shot of the model repository. Put that in the chat. So that's a place where things are very findable and accessible. You can go and look at terrestrial models or other categories of models and get straight to a very nice description and links to the actual code. So here on the next slide just shows you look at all this description that is here and available for you to put down. And this is just an example of the child model. You know all of this the infrastructure that CSDMS has put in place even before fair was written about they really have a lot of capability for people to make their models more fair. Here's another screenshot if you haven't seen it. Most of the models have a digital object identifier assigned to it which allows citation metrics. So all of this is helping people to find and access the data model sorry I'm using data kind of loosely here. And then if you are at some of the recent CSDMS models there can CSDMS webinars are just continuing to build out the infrastructure and build out tools for models that have been submitted to CSDMS to be interoperable and reusable. So if you attended PMT that's an example of that which uses the basic model interface which was also asked about earlier. Okay I'm going to check the questions to see if any other questions have come in. Okay so one question is for a model to be reusable does it have to run on many different operating systems. And I would say that that is not a yes no question it kind of depends on the community so think about CSDMS community. I wouldn't say that it has to run on many different operating systems but if it runs on an operating system that almost nobody in the community uses then it wouldn't be considered reusable. It would make it more reusable if it did run on many different operating systems. But you know these days we see all sorts of things with containers things running on browsers so the technology is always changing and how to define reusable is probably also going to be changing. Another question is would post processing scripts be included in what we're talking about data and models for fair and again I'm going to say yes because it's a very loose definition of data or models and post processing scripts are important to get to the scientific result so those would be included in fair. In fact it's almost hard to think of something that would not be included in the data that we can make fair. I'm doing a time check and I think I'm basically on time I hope, because I have a little bit more information like a side note about working with federal partners and then we have a little bit more polling and then hopefully we'll have time for questions more questions. So this is a note about working with federal partners because since I'm at the USGS I this is what I see day to day. So just know that if you have collaborators in the USGS or other federal agencies they will probably have different and stricter requirements for data and software sharing and documentation. So if you're publishing with them just realize it might take a little bit more time for them to meet all of these requirements. And these requirements can be a little bit arduous but they do go towards making products more fair. This screenshot is just of the code.gov website and in there, oops, information a little bit lower than that. You can see under policy info there's something referred to the federal source code policy where they have a policy and then they have suggested fields to document anything that is you know federally produced code which has a very wide definition. So that could be something to go check out at code.gov. If you are a federal employee or you have partners because to tell you the truth most of the federal employees like they don't know about these things until it is time to do them and somebody's telling them you have to follow the federal source code policy. It's good to try to be a little bit on top of those things before you have to do them in the next week. And here's just a screenshot of the lovely fundamental science practices survey manual chapter. So if you are a USGS employee, you may be familiar with these but just showing those of you who are not a part of a federal agency. There are a lot of rules sometimes and maybe you might be thinking there's no rules, nobody's telling me what to do. So this is the other side of the spectrum where there's a lot of rules and things to do. And you know if you're stuck and you're not part of the USGS but you just want to see what are some ways to approach these things of release of information products and peer review. So this might be a good place to just take a look, you know, if you want to throw in there into your proposal like I looked at this and this is what I'm going to follow that could earn you some points. So I'm putting that URL in there. And again, even though we have all of these directives it doesn't mean that USGS data encoder automatically fair. There's actually, as I mentioned, there's a group that's looking into how USGS policy and fair mesh right now. Okay. So now I just want to kind of review what we've talked about. And these are the takeaways I hope you have from today's webinar, which is fair. It's not a standard or a specification. There's a five or six page paper in 2016. And that the goal of that was to be a guide to assist discovery and reuse by third parties, because everyone's realizing that if what we produce is not reusable that is a really big waste. Fair as originally conceived put specific emphasis on machine actionable characteristics through open technologies not through proprietary technologies. So if you are able to get beyond human readable fair and get to machine actionable fair. That was the original attempt intent and that is would make your stuff more fair and everyone realizes how hard that is. So third there can you often see fair data just used as a phrase fair data but that really applies to code algorithms and workflows as well. And then finally by referring to a few considerations about fair and your proposals and procedures. I really think at this point you're going to be doing better than the status quo. Great. So this is a review of the intended outcome of today's presentation. Hopefully you are now able to explain to someone else what fair means for CSDMS related models and data or at least point to something or say I know where to find out more about that. So you know where to go to find out more and you have some strategies for making your proposals and publications more fair savvy. So now I would like to go back to our little polling app and I'm going to put that in the chat again in case anyone wasn't here when we did it at the beginning. There's a URL where I'm going to open a second poll but first let's look at the results of the first poll that's still up there. Okay so since we have left looking at this poll it looks like a lot more a few more people few more people have marked one for how comfortable they are with explaining what fair means to your models and data. And hopefully now at the end of this webinar that number has increased for you. Okay so I'm going to stop that poll and go to the next one which is if you were to attend a one and a half hour clinic on fair models and data how would you like to spend your time. Because who knows maybe one of those clinics is going to be happening in May at the CSDMS meeting so let's see I'm going to go ahead and put my answer here you can click up to three of these. So let's see here. I have sent my votes and we'll leave that open for a little bit. I'm going to go back to the questions. I don't see any other questions right now but just to note that if you go to questions at that URL, you can put another recommendation of what you would want to happen at a clinic on fair or you can keep asking more questions there. All right so currently the front runner is developing and discussing a checklist to make models and data fair for our community. Sounds good to me. I'm going to leave that open and if anyone has a question of how to get there. It's in the chat there's a URL that you can either go to on your computer or you can go to slido.com on your computer or your mobile phone browser. And put in 358 because it would be nice to see what more of you think there. Okay, so that is the end of my presentation. And if there are any other questions we can talk about those. There might be an extra poll in case there aren't any questions but I'll just wait here. And is it a real inflow or is it just a surface storage is it just these little spikes. Yeah, yeah, this is Lynn maybe you did you unmute everyone basically just like everyone just be aware that you might have been just unmuted the surface storage. Right, I'm not seeing other questions at time at this time I'm looking at the Slido and the zoom chat. I guess we can do our last poll since it looks like we have time for that. So we do a true or false poll. So this is back to what we would want to do on a clinic. And I'm going to stop that one. And if you're still on that page. Let's just talk about fair after listening to this webinar do you think fair models and open source models are mostly the same thing. I guess it's not cool to like show the results before people click them in because maybe you would be influenced by them. But for now, everyone who has answered has said false. And let's just talk for a minute or two about why that would be. I had a link because I have a link for everything. And then this is one of the resources I shared but it's what's the difference between fair data and open data. So this is on the go fair website. And you know the answer is they are different although there are similarities so the key difference is that open data should be available to everyone to access using share without licenses copyright or patents. And it's expected that it should be subject to permissive licenses. But in when you say fair data accessible means accessible by the appropriate people at an appropriate time in an appropriate way. So this means that you can have private data protected data stuff that's sensitive locations of species that's what they talk about here at the USGS or fossils so open and fair are not exactly the same you can have your code open and throw it someplace where it's completely open. But it is not well documented commented or described and it could be very low on the fair spectrum, but still be open. So that was a question that comes up commonly. Alright, I'm just going to throw up the takeaways then and show those. And then Lynn if there's not anything else that we have to do at the webinar I just want to thank everyone for attending and listening. The slides are online we're talking about putting them on the CSDMS terrestrial working group web page but I will also put in my last link of the webinar. I can find it. Maybe it will be too difficult for me to find when people are on the webinar, but maybe the best way is to have it say that we're going to put it on the terrestrial working group page. And if you stay on for another second or two I will find it and put it into the chat. I found it. Alright everyone that's the last link. It goes to these slides and please don't hesitate to get in touch with me. You can find me. I'm sure you can find my email online LHSU at USGS.gov if you want to talk about any of this stuff anymore. Okay, thanks. See you next time.