 Hello, everyone. Thanks for tuning in today. As Paul mentioned, my name is Sarah. I'm a field ecologist for NEON working out of domain 7, which covers the Appalachian Mountain and Cumberland Plateau Eco Region. I was really excited to hear about NEON's use in an earlier talk today. So I'm even more excited to talk to you about NEON's activities in the Smokies. So for anybody who's not as familiar with NEON, I'm going to try to very quickly give you a brief overview of what NEON is and why a program like NEON is important. I'll highlight some of NEON's available resources. And then lastly, I'm going to try to get to the meat of it and show you some current trends we're seeing in our plant data and discuss how you, as an external researcher, can build off these trends and then lastly go through some recently published research that are using NEON data in Great Smoky Mountains. So the National Ecological Observatory Network, or NEON, is a continental-scale data platform which records observations and collect samples from all major biomes within the continental US, Hawaii, and Puerto Rico. So these ecoregions or domains represent distinct vegetation, landforms, and ecosystems that are capturing the full range of ecological diversity and climate variability. NEON provides free and open data on the drivers of and responses to ecological change, a standardized and reliable framework for research and experiments. And the data collection is structured for integration with other national and international network science projects. So NEON is truly continental-scale in that it has 81 sites in total, 47 of those are terrestrial and 37 are terrestrial and 34 are aquatic. And within Great Smoky Mountains, there's a terrestrial site that spans the Tennessee side of the park and just a little bit into North Carolina, as well as an aquatic site that's located at Laconte Creek near the Twin Creek Science Center. So samples are collected from all of the sites and are sent to a biorepository located at Arizona State University for use by the external research community. And of the 181 data products recorded during NEON sampling, 50 of those are organismal and there are 62 different organismal sample types that are archived at the biorepository. And as of February 2021, Arizona State has received over 231,000 NEON samples for archive and nearly 15,000 of those are currently in use. So NEON's breadth of sampling allows researchers to address questions across various spatial and temporal scales. NEON data sampling is designed to scale within a single field site like Great Smokies but also across multiple sites to enable researchers to ask questions at the continental scale. And NEON will be collecting data and samples for a 30-year timeframe. So the long-term operations will allow researchers to answer questions on scales from seconds all the way up to decades. By collecting data and samples from numerous physical, biotic and abiotic processes, NEON employs a consistent and standardized design in an operational infrastructure to address the complexity associated with these various scales. NEON is an ambitious program but is important as it's designed to address the need for eco forecasting as well as the growing global demands on our nation's natural resources. NEON has been designed to provide new information that can be used to challenge and advance current ecological theory and to partner with other organizations to optimize investment in environmental science. And it is the standardization of methods used for data collection which goes to the core of what makes NEON so special as an observatory designed to collect comparable data utilizing consistent methods across a huge diversity of ecosystems. So NEON standardized methods is functionally a combination of three different collection systems. There's the automated instruments that are integrated on our tower and aquatic infrastructure that are collecting a wide range of abiotic measurements. There's also the observational sampling that comes in the form of on-the-ground data collection recording both terrestrial and aquatic biotic data and airborne remote sensing is performed via scheduled flights that collect a suite of LiDAR and imagery data. The diversity of data collection methods allows for comprehensive sampling at each site. The data can be studied in conjunction with each other since they're collected in close proximity at each field site. This also allows researchers to examine linkages across aquatic terrestrial and atmospheric systems. The data are comparable across ecosystems as well as field sites which means researchers can study connections and patterns and develop models to forecast environmental trends locally, regionally and even at the continental scale. So that's really briefly a little bit of background about NEON. And so now you may be thinking, how does NEON work for me? NEON has a plethora of data from nearly all of our sites available right now on our public facing data portal. The data portal is going to be your primary source for exploring and downloading NEON data in any related documentation. You can search the data in a variety of ways. So by keyword, site location, selected dates or even by data theme, it's open access. So no login is required and the portal additionally has information related to siting and publishing with NEON data. All of the data are posted after processing and quality checks are performed and availability is dependent on the type of data product and when the site was constructed. For example, some of the tower sensors stream data every 30 seconds whereas other data products are loaded once per month or even once per year. For Great Smoky Mountains, our data spans back to 2015 when we first started sampling at the site. So NEON collects a lot of data obviously and that can come to you in complex data sets. So knowing how to get started and how to interpret the data is really important. So in addition to the actual data, NEON provides a wide range of data tutorials and workshops aimed at refining coding skills and specifically directed at how to work with NEON data sets. The tutorials cover a variety of software platforms including R, Python and Git. The self-paced data tutorials can be done independently online or easily adapted into guided activities. So the structure of NEON data support learning and practicing reproducible research methods so much so that a lot of faculty adapt them to being used in the classroom as lab or homework assignments which can also be seen in NEON's teaching modules that were developed by outside faculty who want to bring NEON into the classrooms. And these modules include all the necessary materials, their concept focus but also include important skill sets for students like statistical analysis and data visualization. A really cool one that was developed recently focuses on the impact of wildfire on bird communities using NEON data from the Smokies. NEON's coding resources include a variety of packages that have been built specifically to be used with our data sets to make it easier to work with and perform some common algorithms. Much of the available code is community contributed by external researchers using NEON. Code gets shared to GitHub so other researchers can use those resources instead of building their own from scratch. And all of that code and the packages are open access as well and available to download on GitHub. One package in particular that is incredibly useful if you needed to pick one is the NEON utilities package. It downloads NEON data directly from the API and merges it into fewer data sets to make it easier to use. So there's no downloading several Excel sheets and merging them by hand. And while NEON collects a lot of data we also provide a variety of other resources for the science community. And some of those include the thousands of samples we collect which are comprised of the biorepository samples as well as some that could be collected in excess each year. The mobile deployment platforms or mini towers carry a sensor suite that mirrors instrumentation currently present on the NEON towers. And these can be used at non-NEON sites for up to a year to support external research and also used in combination with existing NEON towers to further strengthen the size and comparability of your data sets. The airborne observation platform resources include a hyperspectral imaging spectrometer, discreet and waveform lidar and a high resolution digital camera. Researchers can request to fly over non-NEON sites during times when NEON does not collect AOP data which allows you to take advantage of the existing equipment sending you time and some material costs. Investigators may also request to add sensors to existing NEON field site infrastructure to collect their own data and can request access to NEON sampling locations as well as additional data or sample collection by NEON's professional field ecologists such as myself to support PI-Lib projects at NEON sites. So many of these options in our assignable assets program provides you with the opportunity to leverage existing systems that will save you time and resources as well as allow you to better align NEON's data with more specific measurements needed to answer your research questions. So with all of this, NEON has generated a lot of documentation to support our standardized methods. For observational sampling alone, there are over 30 sampling protocols that encompass the diversity of taxa and biogeochemical measurements. NEON sampling protocols are the result of extensive research and collaboration with experts in the community, especially through our technical working groups. We have adopted standardized methods whenever possible and completed field testing on a selection of NEON sites. At this point, most of our protocols have been implemented at multiple sites across the country for up to five years with revisions occurring annually to date to improve and clarify the protocol. So using NEON's current set of protocols means that you don't have to create and test your own methods. Since ours have already been built and tested, they're ready to use without additional work on your end. And using NEON protocols further allows you to combine your data set with NEON's for greater comparability. So that is just a little bit about NEON and all it has available to you. Now I wanna focus on one of our data products, plant diversity, and look at some interesting trends we're currently seeing in our data sets. So NEON measures plant diversity on 20 by 20 meter plots distributed outside of and within the tower area. For each plot, there are a number of subplots for which plant species are measured. And very generally, we start with the smallest one meter squared subplots and record all the plant species within that subplot area, cover it in height measurements for each species, and then get cover estimates for other biotic and abiotic variables within that subplot. So things like lichen, rocks, or woody debris. From there, we span out and check the 10 meter squared subplots and then the 100 square meter subplots to record any species that were not previously listed. And these are just presence and absence observations only. So from that data, we plotted plant species across elevations to get this graph. So this graph shows a number of unique species per plot as a measurement of species richness versus elevation for five field seasons and 165 observations in 36 plots. So those field seasons span from 2015 to 2019. The background grade dots are each an individual plot. So those are the actual data. And the trend lines are predictions from a generalized mixed model with a negative binomial distribution. The model predicts that both elevation and year have an impact on species richness, but they do not interact. It shows that species richness per plot goes down as elevation increases. Both of those fixed effects have a significant p-value and the r-square value for the fixed effects of year and elevation is 0.552. So it's no surprise that species richness is higher at lower elevations. That's pretty well documented in the literature. But a couple of interesting things that we pulled from this figure is that at lower elevations, each plot has a different species richness and that there are differences between years which suggests that there's a variety of factors that are influencing the data. Which leads us to ask questions like how does rainfall and drought factor in? Or what about the wildfire in 2016? There were also recorded high wind events in early 2017 and the images below demonstrate the effects of those high winds. So the leftmost figure is data being streamed from our tower phenology camera. And the time period I wanna focus on is this one in early May. Our camera stopped streaming at that point due to high winds. And the photo on the left is before the high winds and the photo on the right was recorded two days later after the high wind events. So as you can see, hopefully, we started with a full green lush canopy and we're left with one that was more reminiscent of early spring with much fewer leaves. So with this, you could explore if a change of canopy structure allowed more species to be able to establish themselves in the understory and maybe that's why you're seeing some increases in species richness. So these additional questions can be researched by looking at some of Neon's other data products like wind speeds, rainfall, looking at lidar imagery or leaf area index. And this is where you as an external researcher can come in and use Neon as a platform to further build research. And this was observed with just five years worth of data. So imagine what this looks like in another five years or even in another 10 years. So to wrap up the conversation today, I'd like to highlight a couple of projects that have used Neon samples or data from Great Smoky Mountains that demonstrate how researchers are currently integrating Neon into their programs. So the first study was performed in collaboration with Neon scientists to look at changes in soil pyrogenic carbon around the wildfire event in 2016. So anybody who studies wildfire knows that it's challenging due to the rapid post-fire changes in the ecosystem and lack of robust controls. So the researchers overcame some of these limitations by leveraging Neon plots and samples already located at the park. So they used soil samples collected from Neon's distributed soil plots and within the tower soil arrays. They examined pyrogenic carbon and soils from three time at three time points from areas of just low burn severity. And then at two time points from areas burned at a range of lower to higher severity. So they found several interesting results. A couple of revolved around burning even at low severity, increased pyrogenic carbon in their organic layers and moved into some of the mineral layers. And this was further significantly increased with increasing burn severity. And one other really neat finding is that the pyrogenic carbon concentration at Great Smoky Mountains samples are similar to those published values from study areas that have historically experienced frequent fire and are currently including prescribed burns as part of their management plan, which is consistent with the hypothesis that Southern Appalachian region historically had frequent fire events and indicates that even low severity fire may be an important mechanism by which pyrogenic carbon is produced and transported into mineral soils. So really this study is a great example of how scientists can leverage Neon's current infrastructure to opportunistically ask some really interesting questions around natural disaster events. And similar opportunities have been taken with other wildfire events in our Western domains and even with other natural disaster events like hurricanes and earthquakes. So the last study I wanna discuss is also a great example of how you can answer complex ecological questions solely using Neon data that is available on Neon's data portal. This study came from a group out of the University of Tennessee in Oxville. They looked at five of Neon's sites, including Great Smoky Mountains and built a new LiDAR structural metrics based on the leaf area density at each vegetation height layer and used metrics to study how different aspects of four structural heterogeneity explain variation in bird species richness. Their goals were to test whether leaf area density based metrics are better at explaining bird species richness compared to metrics based on the top of the canopy alone. And if the different aspects of structural heterogeneity had diverse effects on bird richness. And what they found is that bird species richness increased with horizontal heterogeneity while vertical heterogeneity had negative effects which is contrary to the previous research. So in brief, their findings highlight the need for structure animal diversity studies to incorporate metrics that are able to capture different aspects of forest 3D heterogeneity. And these are just two really great examples of how researchers are currently using available Neon data and samples. There are many more examples of research done at various Neon sites that can be found on Neon's website where they have lists of publications that are either mentioned Neon or are using Neon data specifically. You can also find them through the Zotero library platform where you can search for papers using Neon and further filter by site if you like. So with that, I'll just say thanks to NSF for funding such a great ecological opportunity to Great Smoky Mountains for being awesome site host to houses for the next several years. To Marie Faust, our outreach coordinator for helping me collate all this information to share today. And of course, I can't not mention our domain field staff. There's no way I could fit everyone into one of these slides. So I picked a few photos to represent our great crew. Without them, we wouldn't have an awesome data shed to share with you today. And then here's a quick list of contacts and places to find more information. We recently just launched a getting started with Neon and data and resources webpage which is a perfect place for you to start if you have a lot more questions. But there's also some contacts there. So feel free to take a screenshot or if you wanna scan the QR code, it'll take you to the Neon event page for the colloquium which also houses this information. So thanks again. Thank you very much. We have time for one question. If anyone has a question, please either raise your hand or put it in the Q&A box. Question, is Neon any closer to display real-time data on their website? Great question. Yes, so they're working on building quicker and quicker responses to being able to upload the data. They recently just did a big upload of some of our board recently and a few years past data. So I think it's still kind of a work in progress to get all of that up to speed more quickly but we're definitely getting closer and closer to that reality. So yeah, I think it's not quite a perfect system yet. We're still building that out but we're getting closer to real-time data and I can get some more specifics for you, Jim and let you know how close we are to that. Especially for our tower sensors. I think we're aiming to get that as close as possible to real-time.