Welcome to the EROS User Experience webinar series, where we talk to staff at EROS to learn more about the data, tools and services coming out of the USGS Earth Resources Observation Science or EROS Center. Today's webinar is titled RCMAP Long-term and Near-Real time Rangeland Vegetation Monitoring. 00:00:27:05 - 00:00:31:05 I'm your host, Danielle Golon, the remote sensing user services 00:00:31:05 - 00:00:34:08 lead here at EROS in Sioux Falls, South Dakota. 00:00:34:09 - 00:00:36:10 The time is currently 12 p.m. central. 00:00:36:10 - 00:00:38:03 So let's get started. 00:00:38:03 - 00:00:42:00 First, a few logistics to ensure the best audio experience. 00:00:42:01 - 00:00:44:08 All participants have been muted. 00:00:44:08 - 00:00:47:07 If you have any questions or comments during the webinar, please 00:00:47:07 - 00:00:50:20 use the Q&A feature, which can be found under the more settings 00:00:50:20 - 00:00:54:18 if you do not see it currently or if this feature does not work for you. 00:00:54:19 - 00:00:58:00 Please feel free to enter your questions or feedback into the chat, 00:00:58:00 - 00:01:01:04 and we will address the questions and feedback at the end of the webinar. 00:01:01:09 - 00:01:05:09 If you cannot use the Q&A feature or the chat due to your organization 00:01:05:09 - 00:01:10:13 settings, please feel free to email your questions to custserv@USGS.gov. 00:01:10:23 - 00:01:20:21 That is custserv@usgs.gov and we will answer them there. 00:01:21:11 - 00:01:23:10 Today's webinar is being recorded. 00:01:23:10 - 00:01:24:22 The recording will be available 00:01:24:22 - 00:01:28:21 later on the USGS EROS User Experience webinar webpage. 00:01:28:22 - 00:01:32:22 We'll post a link to that webpage in the chat later on, and today's webinar 00:01:32:22 - 00:01:36:17 will consist of a presentation and then a question and answer session. 00:01:37:06 - 00:01:39:11 Today's speaker is Matthew Rigge. 00:01:39:11 - 00:01:41:11 Matthew is an ecologist with the U.S. 00:01:41:11 - 00:01:44:02 Geological Survey here at EROS. 00:01:44:02 - 00:01:46:20 His work focuses on quantifying the current 00:01:46:20 - 00:01:50:21 and historical condition of rangelands across the western United States. 00:01:51:05 - 00:01:54:15 He accomplishes this by producing and analyzing fractional 00:01:54:15 - 00:01:58:09 vegetation cover maps derived from remote sensing and AI. 00:01:59:00 - 00:02:03:09 These data sets support investigations into the drivers of change in rangeland 00:02:03:09 - 00:02:08:10 ecosystems, including long term trends, interannual weather variability, 00:02:08:14 - 00:02:11:14 and the effects of land management practices. 00:02:11:15 - 00:02:14:15 We also have several EROS staff members from User Services 00:02:14:15 - 00:02:18:11 and the RCMAP team here at EROS on the line, to help 00:02:18:11 - 00:02:22:09 answer any questions you may have after Matthew has finished his presentation. 00:02:22:10 - 00:02:24:23 Again, please feel free to add your questions or feedback 00:02:24:23 - 00:02:28:23 throughout the webinar using that Q&A feature in teams or the webinar chat. 00:02:29:09 - 00:02:32:02 We'll try to answer all questions within the time allotted, 00:02:32:02 - 00:02:35:08 but if we're not able to address your question during the final Q&A 00:02:35:08 - 00:02:38:10 portion of the webinar, we will follow up with you offline. 00:02:38:18 - 00:02:41:11 And if there's a future webinar topic you would like us to cover. 00:02:41:11 - 00:02:43:20 Please feel free to suggest that as well. 00:02:43:20 - 00:02:47:21 With that, it's my pleasure to introduce today's speaker, Matthew Rigge. 00:02:47:23 - 00:02:49:07 Take it away, Matthew. 00:02:49:07 - 00:02:52:09 Thanks, everyone for taking your lunch hour 00:02:52:09 - 00:02:57:06 or they're about to learn more about RCMAP, the Rangeland Condition 00:02:57:06 - 00:03:00:08 Monitoring Assessment and Projection project. 00:03:00:09 - 00:03:05:23 I'm going to talk today about our long term monitoring data and near real time. 00:03:05:23 - 00:03:09:23 This work is supported by both the BLM, the Bureau of Land Management 00:03:09:23 - 00:03:13:06 and the USGS National Land Imaging Program. 00:03:13:08 - 00:03:18:11 Also would like to acknowledge support from our KBR contract here at EROS. 00:03:18:11 - 00:03:23:23 Brett, Dev, Logan, Trenton, Jordan, Audra and Steve with RCMAP. 00:03:24:00 - 00:03:28:09 Our goal is to capture both gradual and abrupt 00:03:28:09 - 00:03:32:14 changes in rangeland condition over the Landsat time series. 00:03:32:22 - 00:03:35:22 So that's 1985 until present. 00:03:36:03 - 00:03:41:02 A lot of that change is related to interannual weather variation, 00:03:41:06 - 00:03:46:20 but also disturbance, land management practices, all sorts of things 00:03:46:20 - 00:03:51:21 going on in the landscape that are causing change worthy of detection. 00:03:51:22 - 00:03:56:21 The overall goal of the project is that monitoring changes to vegetation 00:03:56:21 - 00:04:00:04 composition in the past, evaluating 00:04:00:09 - 00:04:03:09 the effect of management practices 00:04:03:20 - 00:04:06:08 helps us to better target future 00:04:06:08 - 00:04:09:20 improvements and better understand how the landscape is changing, 00:04:09:23 - 00:04:13:20 ultimately improving wildlife habitat and landscape health. 00:04:14:02 - 00:04:18:11 So that's one of the real emphasis of this project, is to make the data 00:04:18:20 - 00:04:22:12 as tangible to real world applications as possible. 00:04:22:12 - 00:04:26:23 So we do this by combining a series of field observations 00:04:26:23 - 00:04:32:13 with machine learning and AI Landsat data, a bunch of other ancillary data 00:04:32:15 - 00:04:36:08 sets like topography, vegetation indices, etc. 00:04:36:15 - 00:04:40:17 to make a series of ten cover maps 00:04:41:01 - 00:04:44:23 on an annual basis across the western US. 00:04:45:06 - 00:04:50:00 So we have percent cover maps for bare ground litter or basis 00:04:50:01 - 00:04:53:13 perennial, herbaceous annual or basis shrub, 00:04:54:03 - 00:04:56:16 sagebrush, non sagebrush, 00:04:56:16 - 00:04:59:20 shrub tree and the height of shrubs. 00:04:59:21 - 00:05:04:00 So this is an example of what the data would look like for a pixel 00:05:04:01 - 00:05:08:09 on the landscape were operating at the Landsat scale, which is 30 meter 00:05:08:16 - 00:05:12:19 spatial resolution and gives us a pretty robust understanding 00:05:12:19 - 00:05:18:03 of how that pixel has changed through time and helps us to sort of understand 00:05:18:07 - 00:05:21:07 what the driving factors of that change are. 00:05:21:07 - 00:05:24:08 Provide an example here in Montana, 00:05:24:09 - 00:05:29:20 the Bridger Fire, which occurred in 2020, showing the benefit of our sea 00:05:29:21 - 00:05:34:10 map data compared to some other ways of understanding change. 00:05:34:10 - 00:05:37:20 So we have on the top left a Landsat 00:05:37:21 - 00:05:41:06 image pre fire and top right post-fire. 00:05:41:17 - 00:05:44:06 We can clearly see the burn impact. 00:05:44:06 - 00:05:49:16 These darker brown areas within the burn perimeter in red indicating burn. 00:05:49:17 - 00:05:54:16 The image does provide some context into how the landscape has changed. 00:05:55:01 - 00:05:59:16 But looking at the sea map tree cover from the same years 19 and 21, 00:06:00:06 - 00:06:03:13 we can understand more in depth 00:06:04:00 - 00:06:07:03 what the magnitude of that change was. 00:06:07:03 - 00:06:10:13 Also, compared to data like NDVI. 00:06:11:09 - 00:06:14:09 That same fire scar here in Montana, 00:06:14:19 - 00:06:17:19 the fire indicated with that dashed red line, 00:06:18:09 - 00:06:21:22 we can see that the year after fire in 21, 00:06:21:23 - 00:06:26:02 the NDVI was lower than pre fire, 00:06:26:06 - 00:06:28:20 but it quickly recovers and it really doesn't provide 00:06:28:20 - 00:06:33:06 a whole lot of information on how that landscape changed post-fire. 00:06:33:12 - 00:06:36:12 Compare that again to our see map 00:06:37:01 - 00:06:39:09 fire indicated in that dashed red line. 00:06:39:09 - 00:06:44:23 We can see tree cover has gone down a lot post-fire 00:06:46:01 - 00:06:49:02 and or cover has gone up. 00:06:49:07 - 00:06:55:02 So we can understand how the vegetation community has redistributed post-fire. 00:06:55:02 - 00:06:58:22 In this case, with herbicides increasing. 00:07:01:09 - 00:07:02:08 Training data 00:07:02:08 - 00:07:06:16 is really important to produce a model like this. 00:07:06:17 - 00:07:12:10 One of the unique things about our map is our series of high resolution 00:07:12:10 - 00:07:16:16 satellite image sites, where we sent out field 00:07:16:16 - 00:07:22:03 crews to the ground and observe the cover of various patches of vegetation 00:07:22:16 - 00:07:26:08 on different types of geology, all the different 00:07:26:08 - 00:07:30:03 permutations of things that could exist on these image footprints, 00:07:30:03 - 00:07:34:05 and then use those data to predict 00:07:34:05 - 00:07:37:07 a two meter scale, a high resolution scale 00:07:37:13 - 00:07:40:13 map of all the vegetation components I listed before, 00:07:41:03 - 00:07:43:23 and then scale that up to 30m 00:07:43:23 - 00:07:47:03 to serve as a key element of our training data. 00:07:47:04 - 00:07:50:15 So those sites are indicated by all of these red dots here, 00:07:50:15 - 00:07:55:05 which are approximately 15 by 15 kilometer image footprint. 00:07:55:06 - 00:08:00:11 Those would be worldview two, worldview three primarily satellite images. 00:08:00:14 - 00:08:06:15 So the benefit of that approach is that it gives us a huge N 00:08:06:18 - 00:08:12:13 to operate with nearly 69,000,030 meter scale training points to operate with. 00:08:12:21 - 00:08:18:13 With a model like this, we're using a neural net model, really data hungry. 00:08:18:13 - 00:08:21:20 So having a high end is really important here. 00:08:21:22 - 00:08:25:18 We also pull in all of the Bureau 00:08:25:18 - 00:08:30:03 of Land Management aim the analysis inventory and monitoring data 00:08:30:04 - 00:08:33:07 set and their LMF data as well. 00:08:33:21 - 00:08:36:15 So these data are the gold 00:08:36:15 - 00:08:39:22 standard monitoring data collected by the BLM. 00:08:40:03 - 00:08:45:18 Most of those plots are randomly located across the domain in the west western 00:08:45:18 - 00:08:50:23 part of the US, designed to represent essentially a single Landsat pixel 00:08:51:14 - 00:08:54:14 measured with boots on the ground as well. 00:08:54:22 - 00:08:58:20 So that's definitely a key data set we've brought in here as well, 00:08:59:08 - 00:09:03:16 pulling in data from land fire, their public reference database, other 00:09:03:19 - 00:09:09:03 sort of ad hoc additional data sets as we can get Ahold of it from the G. 00:09:09:03 - 00:09:09:10 Biff. 00:09:09:10 - 00:09:13:21 As an example, the Global Biodiversity Information Facility, 00:09:14:09 - 00:09:18:22 some UAV classified imagery 00:09:19:03 - 00:09:22:12 from Aaron Johnston in western South Dakota. 00:09:22:14 - 00:09:26:11 We will take nearly any fractional component 00:09:26:11 - 00:09:30:11 cover field observation as training data. 00:09:30:11 - 00:09:33:20 So if you do have any data available that might fit the bill, 00:09:34:03 - 00:09:35:23 please feel free to reach out altogether. 00:09:35:23 - 00:09:40:09 We've got nearly 70 million training points to choose from, covering 00:09:40:16 - 00:09:47:06 the broad temporal extent and capturing a lot of the spatial variability. 00:09:47:07 - 00:09:51:23 We do note that there is higher density of training data in public lands, 00:09:51:23 - 00:09:56:02 especially in BLM country, as opposed to like the Great Plains, 00:09:56:02 - 00:09:59:02 as an example, where there's simply less public land. 00:09:59:03 - 00:10:03:16 So in RCMAP, we produce two key series of data. 00:10:04:12 - 00:10:08:17 One of them being the RCMAP set of annual products, 00:10:08:17 - 00:10:12:00 those ten cover components I mentioned before, 00:10:12:18 - 00:10:17:08 and the second being a series of RCMAP e.g. 00:10:17:15 - 00:10:21:06 that is exotic annual grass near real time 00:10:21:09 - 00:10:24:10 weekly and also annual products. 00:10:24:10 - 00:10:29:10 So the annual products span the full Landsat archive 00:10:29:16 - 00:10:34:03 going back to 1985, stretching through 2025. 00:10:34:07 - 00:10:38:03 Currently, we have plans in place to extend that through 26 00:10:38:04 - 00:10:42:16 and annual releases updating additional years at the end. 00:10:43:01 - 00:10:44:00 Moving forward. 00:10:44:00 - 00:10:48:18 Ten Cover Components I mentioned before the key input data 00:10:48:19 - 00:10:54:09 here are Landsat ARD derived percentile composite. 00:10:54:18 - 00:10:58:07 So we look across all the clear observations 00:10:58:19 - 00:11:01:17 within each calendar year. 00:11:01:17 - 00:11:04:17 Find the 10th, 50th, and 90th percentile 00:11:05:01 - 00:11:09:15 of all the clear observations in that pixel and derived composites, 00:11:09:17 - 00:11:14:10 the idea there being that those percentiles help us to capture 00:11:14:11 - 00:11:18:17 the range of phenology conditions going on within that year. 00:11:18:19 - 00:11:22:14 Also, pulling in spectral indices, topography features, 00:11:23:00 - 00:11:25:19 and some texture metrics as well. 00:11:25:19 - 00:11:31:05 Already talked about the training data for RCMAP moving over to the EAG side. 00:11:31:09 - 00:11:32:10 The spring. 00:11:32:10 - 00:11:37:16 We completed some major updates to that series of data. 00:11:37:17 - 00:11:40:17 The previous PM of that project, Steve Boyd. 00:11:40:17 - 00:11:45:02 We've been investigating updates to that release that this spring. 00:11:45:18 - 00:11:48:23 So major updates to the methodology. 00:11:49:03 - 00:11:52:13 Got a paper in the works describing those updates, 00:11:52:15 - 00:11:55:15 and those are weekly products 00:11:55:18 - 00:12:01:09 designed to be released within a week of the latest satellite 00:12:01:09 - 00:12:06:15 observation going into the model produce for 2016 through 2026 currently. 00:12:06:15 - 00:12:10:18 So we're in the midst of releasing 2026 data at the moment. 00:12:11:10 - 00:12:14:10 And this data set we produce 00:12:14:15 - 00:12:19:09 component cover similar as the RCMAP products fractional cover 00:12:19:17 - 00:12:22:22 focusing on the herbs component, 00:12:23:03 - 00:12:25:23 especially invasive grasses. 00:12:25:23 - 00:12:28:23 So we have cover of total urban 00:12:28:23 - 00:12:32:12 green and sinister cover, 00:12:32:20 - 00:12:37:11 cheatgrass and medusa head, which are two of the key 00:12:37:12 - 00:12:41:05 invasive grasses impacting a lot of the western United States. 00:12:41:21 - 00:12:43:04 Field brome. 00:12:43:04 - 00:12:47:02 Sandburg's, a native perennial grass that has similar phonology 00:12:47:04 - 00:12:49:21 to many of the exotics, which is Sandburg's bluegrass, 00:12:49:21 - 00:12:56:10 and the total cover of a list of 16 species of exotic annual grasses that are 00:12:57:10 - 00:12:58:17 problematic. 00:12:58:17 - 00:13:03:18 And with the data we're pulling in instead of Landsat data, 00:13:04:01 - 00:13:09:01 we use harmonized Landsat Sentinel data, which gives us better 00:13:09:10 - 00:13:13:18 temporal resolution, allowing us to better capture 00:13:14:05 - 00:13:19:11 the logical patterns across the landscape, which is especially important 00:13:19:11 - 00:13:24:15 for understanding the differences between a site with perennial grasses 00:13:24:15 - 00:13:29:03 versus invasive grasses, where the phenology tends to be insights 00:13:29:03 - 00:13:32:13 with a lot of invasive grasses otherwise similar 00:13:32:17 - 00:13:35:17 inputs to the annual products. 00:13:36:00 - 00:13:38:01 Similar training as well. 00:13:38:01 - 00:13:39:05 I mentioned this before. 00:13:39:05 - 00:13:45:02 Both of these products use neural network models for classification. 00:13:45:07 - 00:13:49:23 Both of these data sets fit within a broader mapping hierarchy. 00:13:51:09 - 00:13:52:13 A lot going on here. 00:13:52:13 - 00:13:54:06 I'll give the basics. 00:13:54:06 - 00:13:58:20 So in RCMAP we've got five primary components 00:13:58:21 - 00:14:02:19 or basis shrub, tree, bare ground and litter. 00:14:03:03 - 00:14:09:03 The sum of those five will equal 100% cover in each and every pixel 00:14:09:07 - 00:14:14:06 that is enforced through post-processing as needed. 00:14:14:15 - 00:14:19:15 We produce a series of secondary products for some of the primary components 00:14:20:02 - 00:14:23:19 that add relevant floristic detail. 00:14:23:19 - 00:14:27:19 For example, for sage grouse applications, understanding 00:14:28:02 - 00:14:33:06 what portion of the shrub component is sagebrush is highly relevant. 00:14:34:12 - 00:14:35:23 Those nest 00:14:35:23 - 00:14:38:23 within their respective primary component. 00:14:39:05 - 00:14:43:17 So on the total or basis side, we separate broadly into annuals and perennials, 00:14:44:12 - 00:14:48:20 and these weekly and annual map EAG products 00:14:48:20 - 00:14:52:17 all fit within the total or basis hierarchy. 00:14:52:21 - 00:14:56:21 The overall scheme here is to capture 00:14:56:23 - 00:15:02:09 the diversity of the landscape free of contradictions. 00:15:02:09 - 00:15:05:09 And in this hierarchy, 00:15:06:14 - 00:15:08:09 looking at our annual components, 00:15:08:09 - 00:15:12:17 this is our most recent data release from the spring. 00:15:12:18 - 00:15:15:18 Our version eight. 00:15:15:22 - 00:15:18:13 These animations are showing how 00:15:18:13 - 00:15:22:10 the component cover has changed from 1985 00:15:22:23 - 00:15:26:11 until 2025 on an annual time step, 00:15:28:09 - 00:15:30:14 showing that they are quite dynamic, 00:15:30:14 - 00:15:34:07 especially components like bare ground and or basis. 00:15:35:09 - 00:15:38:03 A lot to absorb at this level, 00:15:38:03 - 00:15:41:13 which is why it's nice to condense them 00:15:42:01 - 00:15:45:16 these temporal patterns into trends. 00:15:46:03 - 00:15:49:03 And to me, time is the most interesting 00:15:49:04 - 00:15:52:04 thing going on here with RCMAP. 00:15:52:06 - 00:15:55:06 Understanding how the landscape has changed through time 00:15:55:09 - 00:15:58:11 is really the take home message for all of this. 00:15:59:04 - 00:16:02:10 So what I've done here is summarized 00:16:03:00 - 00:16:07:01 the 1985 to 2025 trend per 00:16:07:01 - 00:16:10:21 component per level three ecoregion, 00:16:12:02 - 00:16:13:13 where ecoregions in 00:16:13:13 - 00:16:17:08 blue have a net decreasing trend 00:16:17:14 - 00:16:20:16 in that component through time, and the opposite 00:16:20:16 - 00:16:23:16 for ecoregions in red. 00:16:24:09 - 00:16:28:11 Some of the big stories here are related to 00:16:28:11 - 00:16:32:09 annual basis, which would include invasive grasses. 00:16:32:22 - 00:16:35:12 That problem is expanding 00:16:35:12 - 00:16:38:12 across much of the West and the Great Basin. 00:16:38:23 - 00:16:42:20 Some new areas of invasion of annual grasses 00:16:44:04 - 00:16:47:04 into Montana and Wyoming as well. 00:16:47:12 - 00:16:50:12 And that increase in annual 00:16:50:12 - 00:16:54:19 invasive grasses is driving an increase in the fire regime, 00:16:55:16 - 00:17:00:01 leading to more burned area, which results in a declining 00:17:00:02 - 00:17:04:06 trend in sagebrush since 85 across 00:17:04:06 - 00:17:07:06 most of the West. 00:17:07:09 - 00:17:10:06 And also on the flip side, 00:17:10:06 - 00:17:14:08 an increasing trend in herb cover overall 00:17:15:19 - 00:17:19:17 as the grasses expand in burned areas. 00:17:19:18 - 00:17:22:18 A lot of that increase in is related 00:17:22:23 - 00:17:26:11 to increases in the annual portion of that. 00:17:26:11 - 00:17:28:15 So not so great news. 00:17:28:15 - 00:17:31:01 And another side effect of all of this 00:17:31:01 - 00:17:34:04 is a declining pattern in bare ground. 00:17:34:12 - 00:17:38:11 So the sagebrush biome in particular typically 00:17:38:11 - 00:17:44:08 had a lot of bare ground between the inner spaces of shrubs. 00:17:44:09 - 00:17:47:09 After invasion and following fire. 00:17:47:09 - 00:17:51:12 Those spaces get filled in with invasive grasses, often 00:17:51:23 - 00:17:55:09 leading to a declining pattern in bare ground. 00:17:56:09 - 00:17:59:09 Another interesting pattern to note here 00:17:59:19 - 00:18:02:19 in the Great Plains, this is really well documented. 00:18:03:16 - 00:18:06:21 There's an increasing pattern in tree cover, 00:18:07:23 - 00:18:12:15 especially in the southern portion of the Great Plains, as trees, 00:18:12:16 - 00:18:17:17 especially eastern red cedar, are expanding due to fire suppression 00:18:17:17 - 00:18:20:17 and other management impacts. 00:18:21:13 - 00:18:24:13 How is Arcee map data been used in the past? 00:18:24:16 - 00:18:27:02 These are examples from the literature. 00:18:27:02 - 00:18:32:13 Lots of diverse applications of the data in wildlife management, 00:18:32:17 - 00:18:37:07 climate analysis, work on energy development, all sorts of things. 00:18:37:07 - 00:18:40:07 I think my favorite is probably a study on lampreys 00:18:40:07 - 00:18:43:12 in the Goose Lake Basin in Oregon. 00:18:43:16 - 00:18:47:21 There's a broad array of applications for these sort of data. 00:18:48:08 - 00:18:50:21 I'm going to focus on three today. 00:18:50:21 - 00:18:55:09 So mentioned the tree cover expansion in the Great Plains. 00:18:55:11 - 00:18:58:00 This is an example of exactly that pattern. 00:18:58:00 - 00:19:01:18 So in southwest Nebraska here 00:19:03:09 - 00:19:06:09 I've got the tree cover slope 00:19:06:13 - 00:19:09:13 between 85 and present 00:19:10:16 - 00:19:11:06 here. 00:19:11:06 - 00:19:15:00 So increases in red and decreases in blue 00:19:15:00 - 00:19:18:07 and or basis cover same pattern 00:19:18:21 - 00:19:21:21 blue showing decreased red showing increase. 00:19:22:19 - 00:19:24:19 So in this area 00:19:24:19 - 00:19:28:11 tree cover has been expanding since 85. 00:19:28:20 - 00:19:32:03 We can see that's the general pattern here in red. 00:19:32:07 - 00:19:37:03 And as one component goes up something else has to go down. 00:19:37:11 - 00:19:40:14 As I mentioned before everything it adds up to 100. 00:19:40:16 - 00:19:42:15 So it's a zero some game. 00:19:42:15 - 00:19:47:04 And in this case what's going down is herb cover grass cover. 00:19:47:11 - 00:19:52:03 The exception to that being this area that burned in 2002, 00:19:52:10 - 00:19:57:13 we can clearly see that impact on the tree cover slope, where that is 00:19:57:13 - 00:20:02:10 an exception to the general rule in this area of increase, 00:20:03:03 - 00:20:07:07 and similarly an increasing trend in our basis. 00:20:07:07 - 00:20:11:04 They're looking at a wildlife management application 00:20:11:05 - 00:20:17:03 in the northern portion of Montana stretching into southern Canada. 00:20:17:10 - 00:20:20:10 We did expand our study area a couple years ago 00:20:20:15 - 00:20:26:12 to capture the sagebrush biome extent into Alberta and Saskatchewan as well. 00:20:26:13 - 00:20:29:07 So Julie Heinrichs et al. 00:20:29:07 - 00:20:32:20 In this case, used our data to understand 00:20:32:20 - 00:20:36:21 long term changes in sagebrush connectivity. 00:20:37:00 - 00:20:42:00 Put those data into a sage grouse resource use analysis model. 00:20:42:00 - 00:20:47:13 Considering that as as one of several factors in that model, ultimately leading 00:20:48:00 - 00:20:50:20 to an understanding of sage grouse 00:20:50:20 - 00:20:53:20 mortality risk in this landscape. 00:20:53:22 - 00:20:59:00 Another wildlife focused case study here, the U.S. 00:20:59:00 - 00:21:04:09 Fish and Wildlife Service leads an effort called the Sagebrush Conservation Design. 00:21:04:15 - 00:21:07:22 This effort aims to 00:21:09:01 - 00:21:11:15 categorize the sagebrush biome 00:21:11:15 - 00:21:14:22 into three priority classes core 00:21:15:16 - 00:21:19:18 sort of the best of the best remaining sagebrush habitat, 00:21:19:21 - 00:21:23:14 where the sagebrush cover is intact, 00:21:23:20 - 00:21:26:20 not very badly fragmented, 00:21:27:01 - 00:21:30:01 and little cover of invasive grasses. 00:21:30:18 - 00:21:35:09 Growth Opportunity area is sort of a moderate degree of intactness. 00:21:35:09 - 00:21:39:04 Maybe some invasives presence in that landscape 00:21:39:10 - 00:21:43:11 and other rangeland areas where the sagebrush habitat 00:21:43:12 - 00:21:47:04 has been badly degraded by invasives, 00:21:47:04 - 00:21:50:08 fire, fragmentation, etc.. 00:21:50:09 - 00:21:55:16 So the key map data both the sagebrush cover or basics 00:21:55:20 - 00:22:00:12 cover a bare ground cover all feed into this classification, 00:22:00:13 - 00:22:06:09 and this helps guide landscape scale management actions. 00:22:06:21 - 00:22:09:14 Recognizing that we have limited 00:22:09:14 - 00:22:14:17 conservation dollars available, we want to apply those dollars 00:22:14:17 - 00:22:17:20 to the places that are most likely 00:22:17:20 - 00:22:21:07 to see the greatest net benefit from them. 00:22:22:21 - 00:22:25:21 Another important thing to note here is 00:22:26:19 - 00:22:30:07 in the most recent version of the sagebrush conservation design, 00:22:30:07 - 00:22:34:13 which is currently in development, the authors have decided 00:22:34:13 - 00:22:39:06 to pull in data from the rangeland analysis platform as well. 00:22:39:08 - 00:22:45:01 The wrap products are similar in many respects to what RCMAP does, 00:22:45:02 - 00:22:50:22 and in this case, the authors of the Sagebrush Conservation Design 00:22:51:02 - 00:22:54:22 found a benefit from doing 00:22:54:22 - 00:22:59:01 a consensus of both our products and the wrap products. 00:22:59:01 - 00:23:02:10 So that's one thing I really want to highlight here, is it's 00:23:02:10 - 00:23:06:15 not a choice of either or necessarily which data set you use. 00:23:06:15 - 00:23:10:22 There is a benefit of, in this case, a geometric 00:23:10:23 - 00:23:14:21 average of both the Wrap and RCMAP products. 00:23:14:21 - 00:23:17:12 I found the same thing in this before. 00:23:17:12 - 00:23:22:10 Definitely less air in a consensus product than a single product by itself. 00:23:22:10 - 00:23:23:15 Speaking of error, 00:23:24:21 - 00:23:26:04 that's one thing 00:23:26:04 - 00:23:30:19 we focus a lot on in RCMAP is accuracy assessment. 00:23:30:23 - 00:23:34:05 So I mentioned the BLM Aim data previously. 00:23:34:19 - 00:23:39:09 We withhold a 10% set 00:23:39:10 - 00:23:43:20 of all of that Aim data for the purposes of validation, 00:23:44:01 - 00:23:48:13 which is about 6600 points in the current iteration of RCMAP, 00:23:49:09 - 00:23:52:04 the rangeland analysis platform folks withhold 00:23:52:04 - 00:23:56:10 that same set of data for cross product comparisons. 00:23:56:15 - 00:24:00:16 Any other project who wants to participate could do so as well. 00:24:00:18 - 00:24:03:20 So in this graphic I am showing 00:24:03:23 - 00:24:07:07 the accuracy stats the arms 00:24:07:09 - 00:24:10:20 which is the root mean squared error on the left and the correlation, 00:24:10:22 - 00:24:15:06 the R value on the right of our component 00:24:15:06 - 00:24:18:22 cover relative to those withheld points. 00:24:19:03 - 00:24:24:12 So the model did not see these points whatsoever in development or version eight 00:24:24:19 - 00:24:29:07 is the most recently released product in May this year. 00:24:30:05 - 00:24:34:19 Version seven was the prior year, six before that, and so on. So 00:24:35:18 - 00:24:36:05 looking 00:24:36:05 - 00:24:41:12 at our relative degree of enhancements through time over the last five years, 00:24:42:08 - 00:24:45:08 so we can see in both our M.Sc. 00:24:45:16 - 00:24:48:02 Where lower values would be better 00:24:48:02 - 00:24:52:09 and in correlation were higher values would be better. 00:24:52:20 - 00:24:55:09 We are definitely trending the right direction. 00:24:55:09 - 00:24:59:19 There are some components that are just easier to produce than others. 00:24:59:22 - 00:25:00:12 Their ground. 00:25:00:12 - 00:25:05:23 As an example, trees an example tend to be more accurate overall, 00:25:06:04 - 00:25:10:09 just a more clear spectral signal with those components 00:25:10:09 - 00:25:14:18 as compared to something like shrub or litter as an example. 00:25:14:20 - 00:25:21:00 So we spent a lot of time improving our products between each iteration, 00:25:21:09 - 00:25:25:08 pulling in new data sets, new modeling techniques, 00:25:25:14 - 00:25:28:09 new training data, etc. 00:25:28:09 - 00:25:32:12 each new iteration of RCMAP is a completely new time series. 00:25:32:12 - 00:25:36:06 So for example, when we do the V8 predictions, 00:25:36:09 - 00:25:39:14 we would redo not just 2025. 00:25:39:20 - 00:25:42:20 We would go back all the way to 1985 because 00:25:43:09 - 00:25:47:17 new training data in 2025 from a new site 00:25:48:00 - 00:25:51:20 would have impacts going all the way back through the time series. 00:25:51:22 - 00:25:56:08 Probably the most common question I get from users is 00:25:56:09 - 00:26:01:13 what is the appropriate spatial scale to apply your product? 00:26:01:15 - 00:26:05:13 And we've always known that bigger is better. 00:26:06:04 - 00:26:08:02 What do I mean by that? 00:26:08:02 - 00:26:12:23 Is the mean of a 1000 acre pasture likely 00:26:12:23 - 00:26:18:03 to be closer to truth than a single random pixel within that pasture? 00:26:18:04 - 00:26:20:16 Yes, but that's not very clear guidance. 00:26:20:16 - 00:26:25:06 So we undertook an analysis to get a better quantitative 00:26:25:06 - 00:26:29:06 idea of what the appropriate scale would be. 00:26:29:14 - 00:26:32:14 We did this by scaling our 00:26:32:20 - 00:26:35:11 high resolution training sites 00:26:35:11 - 00:26:38:10 to a variety of resolutions starting at 30m, 00:26:38:10 - 00:26:42:15 which is the native resolution of our products all the way up to 1500 00:26:42:15 - 00:26:47:19 meter resolution, and compared the accuracy of our products 00:26:47:19 - 00:26:52:17 relative to that training data, which is approximately truth. 00:26:53:07 - 00:26:58:04 And not surprisingly, we found that aggregating 00:26:58:05 - 00:27:02:05 the map predictions to a 90 meter resolution 00:27:02:06 - 00:27:07:17 improves the correlation and RMSE relative to the native 30 meter resolution. 00:27:07:17 - 00:27:13:22 And that affects sort of plateaus as you go beyond about 210 meter resolution. 00:27:13:22 - 00:27:18:09 Just to show that if you can get by with data 00:27:18:10 - 00:27:23:21 at a coarser resolution than 30m, you can produce a more accurate product. 00:27:23:21 - 00:27:26:21 All of this varies by landscape complexity, 00:27:26:22 - 00:27:32:09 how heterogeneous landscape is, what component you're looking at, etc. 00:27:32:10 - 00:27:34:20 but even at a 90 meter scale, 00:27:34:20 - 00:27:38:17 the accuracy is quite a bit better than a 30 meter scale. 00:27:38:17 - 00:27:39:23 And I should also note that 00:27:39:23 - 00:27:44:11 that Aim validation I showed before is at a 30 meter scale. 00:27:44:11 - 00:27:48:01 So we are comparing single pixel truth 00:27:48:03 - 00:27:51:05 tag to our modeled predictions. 00:27:51:06 - 00:27:54:16 Some other important things to consider in terms of accuracy. 00:27:54:17 - 00:27:59:13 First, more common values are predicted more accurately. 00:27:59:13 - 00:28:04:18 If we look at the frequency of a cover value 00:28:04:20 - 00:28:07:23 per component combination versus 00:28:08:00 - 00:28:11:17 the RMS of that cover combination. 00:28:11:17 - 00:28:15:09 In our model predictions, we see a clear pattern here 00:28:15:09 - 00:28:19:07 that the more frequent that cover been is, 00:28:19:17 - 00:28:23:05 the less error there is in that prediction. 00:28:23:05 - 00:28:26:01 So something really common, like a 00:28:27:11 - 00:28:30:09 10% stage brush cover 00:28:30:09 - 00:28:35:18 value, will all else equal be more accurate than something 00:28:35:18 - 00:28:40:10 less common, like a 99% cover sagebrush bin? 00:28:40:11 - 00:28:42:11 That would be very, very uncommon. 00:28:42:11 - 00:28:45:14 Another important thing to think about is the distinction 00:28:45:14 - 00:28:48:22 between absolute error and relative error. 00:28:49:00 - 00:28:51:09 Those can be at odds at times. 00:28:51:09 - 00:28:54:09 For example, looking at bare ground, the peak 00:28:54:10 - 00:28:58:09 RMSE is at about 80% cover, but the peak 00:28:58:09 - 00:29:02:08 relative error is at about to 10% cover. 00:29:02:08 - 00:29:05:21 So considering the ratio of error 00:29:06:09 - 00:29:09:22 of absolute error to the actual predicted value, 00:29:10:01 - 00:29:14:01 aka relative error can be important as well. 00:29:14:05 - 00:29:18:10 Also, should note that our models, like most models, 00:29:18:17 - 00:29:21:11 suffer from regression to the mean. 00:29:21:11 - 00:29:22:23 What does that look like? 00:29:22:23 - 00:29:27:09 It looks like inflation of predicted values at the low end, 00:29:27:13 - 00:29:30:18 and a deflation of values at the high end. 00:29:30:20 - 00:29:33:20 So the range of values can be truncated 00:29:33:20 - 00:29:37:04 relative to training data as an example. 00:29:37:05 - 00:29:41:07 So we have enough withheld Aim data at this point 00:29:41:08 - 00:29:45:04 to look at spatial patterns in accuracy. 00:29:45:18 - 00:29:51:02 In this case I'm looking at independent correlation per level three ecoregion. 00:29:51:03 - 00:29:57:08 Some patterns to note here are the more common that component 00:29:57:08 - 00:30:02:08 is per ecoregion, the more accurate that prediction tends to be. 00:30:02:13 - 00:30:06:10 So as an example, sagebrush cover is very common 00:30:06:10 - 00:30:11:13 in the central Great Plains, much less common in the Mojave Desert, 00:30:12:02 - 00:30:14:21 and our accuracy of sagebrush is higher 00:30:14:21 - 00:30:18:07 in the central Great Basin then in the Mojave Desert. 00:30:18:16 - 00:30:24:05 That's the general pattern that dominates the spatial patterns we see here. 00:30:24:06 - 00:30:30:03 Other things like the background, mineralogy and geology of the site 00:30:30:04 - 00:30:34:19 do complicate our predictions as well, especially in places with lots of bare 00:30:34:20 - 00:30:39:10 ground like portions of the Mojave Desert, Great Basin, etc. 00:30:39:11 - 00:30:44:04 in a lot of cases, the primary thing you see with Landsat imagery 00:30:44:04 - 00:30:50:07 is that background mineralogy, which can, like I said, complicate predictions. 00:30:50:11 - 00:30:50:20 Okay. 00:30:50:20 - 00:30:55:21 Moving over to the other major data set that we produce 00:30:56:07 - 00:30:59:07 are exotic annual grass product 00:30:59:16 - 00:31:03:03 already talked about the invasive grass 00:31:03:13 - 00:31:06:11 acceleration of the fire cycle, 00:31:06:11 - 00:31:09:12 being a key driver of habitat loss 00:31:10:01 - 00:31:14:10 and degradation across a lot of the western portion of the US. 00:31:15:17 - 00:31:18:17 So again, we're using harmonized Landsat Sentinel, 00:31:18:20 - 00:31:23:01 the BLM Aim data, our high resolution training sites 00:31:23:03 - 00:31:27:18 to model weekly cover of basis, 00:31:28:09 - 00:31:33:09 invasive grasses and some species specific maps like cheatgrass 00:31:33:09 - 00:31:38:13 and medusa head near real time classification of cover conditions 00:31:38:15 - 00:31:44:06 helps in rapid decision making related to grazing and fire. 00:31:44:06 - 00:31:46:09 So that's one of our key focuses here. 00:31:46:09 - 00:31:50:17 Again, in releasing these data as quickly as possible 00:31:51:03 - 00:31:55:10 for near real time application mentioned this already. 00:31:55:11 - 00:32:00:11 We did a major overhaul of our mapping methodology in the most recent 00:32:00:11 - 00:32:01:08 data release. 00:32:01:08 - 00:32:06:09 The key problem in the prior version of the data was associating 00:32:06:21 - 00:32:09:15 the yearly Aim observations 00:32:09:15 - 00:32:14:05 with a weekly ndVi curve, resulting in a one 00:32:14:05 - 00:32:18:18 to many relationship instead of a 1 to 1 relationship as we do now. 00:32:19:09 - 00:32:22:09 Therefore, the old product represented 00:32:22:11 - 00:32:27:08 a forecast of peak cover, a forecast that changed 00:32:27:18 - 00:32:30:12 as we got better and better information 00:32:30:12 - 00:32:33:12 towards the peak of each year. 00:32:33:13 - 00:32:38:00 In addition to how we've trained the model, we've expanded 00:32:38:00 - 00:32:41:08 the study area for these products, expanded the mapping window. 00:32:41:09 - 00:32:44:22 Previously, the maps were released March through June. 00:32:45:00 - 00:32:47:20 Now we're going March through September. 00:32:47:20 - 00:32:51:16 We've added additional components, made the data more consistent, 00:32:51:18 - 00:32:54:02 improved input variables to the model. 00:32:54:02 - 00:32:58:10 We are now participating in that 10% withholding aim validation 00:32:58:10 - 00:33:03:09 I talked about before, and offered the data in a lot of new places 00:33:03:09 - 00:33:08:12 in terms of user access like Google Earth Engine, MRC, etc. 00:33:09:10 - 00:33:11:18 all of these changes have 00:33:11:18 - 00:33:15:06 dramatically improved our validation results. 00:33:15:08 - 00:33:21:04 So the top panel here is showing the accuracy of our e g products 00:33:21:04 - 00:33:25:09 relative to that same set of 10% aim withheld points 00:33:25:09 - 00:33:29:23 I talked about before, compared to the accuracy of our prior set of data 00:33:30:23 - 00:33:32:04 total basis, 00:33:32:04 - 00:33:35:16 we did not have that product before, so it can't make that comparison. 00:33:36:02 - 00:33:41:07 But for all the other components, we are significantly better than previously. 00:33:41:16 - 00:33:45:23 Also, the prior version of the data, in this case, it's 00:33:45:23 - 00:33:49:09 not a fully fair comparison because we did 00:33:49:09 - 00:33:53:06 not withhold these points from training the old version. 00:33:53:06 - 00:33:57:09 So it's a cross-validation on the bottom compared to independent validation. 00:33:57:09 - 00:33:58:21 And yet we're still doing better. 00:33:58:21 - 00:34:03:01 So really good news there in terms of improved product accuracy 00:34:03:03 - 00:34:04:23 based on how the model is trained. 00:34:04:23 - 00:34:09:18 Our model will initially predict green cover of all of our components, 00:34:09:18 - 00:34:14:16 and we derive a total and thinnest portion of 00:34:15:16 - 00:34:19:09 cover through a post-processing approach that calculates 00:34:19:15 - 00:34:23:19 maximum year to date green cover as total, 00:34:23:19 - 00:34:30:01 and then cover would be the difference between weekly total and green cover. 00:34:30:02 - 00:34:33:02 Here's an example of what that phenology looks like. 00:34:33:18 - 00:34:36:10 We're total and green are synonymous. 00:34:36:10 - 00:34:41:19 Until we see a decrease in green cover total would stay flat. 00:34:41:19 - 00:34:44:18 In that scenario, senescent would increase. 00:34:44:18 - 00:34:48:12 And then eventually, once vegetation is growth 00:34:48:12 - 00:34:51:21 is done that year, all the cover would go to senescence. 00:34:52:23 - 00:34:54:18 I think this crossing point here, 00:34:54:18 - 00:34:59:09 where green and nest meat and eventually nest is higher 00:34:59:09 - 00:35:04:00 than green, would have important fire implications. 00:35:06:17 - 00:35:08:23 So looking at the weekly time 00:35:08:23 - 00:35:11:23 series here from 2025 00:35:12:00 - 00:35:15:16 for cheatgrass on the left and total e.g. 00:35:15:20 - 00:35:18:21 on the right, we can see a rapid increase 00:35:18:21 - 00:35:22:09 early in the spring and then a plot tailing effect. 00:35:22:18 - 00:35:27:03 So we're looking here at total cover, which would not decrease 00:35:28:02 - 00:35:31:02 throughout the course of the year. 00:35:32:12 - 00:35:32:21 Moving 00:35:32:21 - 00:35:35:21 to a different set here again from 25, 00:35:36:03 - 00:35:39:03 we've got green total and senescent. 00:35:39:16 - 00:35:42:16 Green is going to increase 00:35:43:02 - 00:35:46:02 starting in the south and at low elevation. 00:35:46:07 - 00:35:49:06 And the green will work its way further north 00:35:49:06 - 00:35:52:15 and to higher elevation as the season progresses. 00:35:53:13 - 00:35:58:12 Total or will accumulate and not decline. 00:35:59:01 - 00:36:03:12 And the difference between total and green will go to senescent. 00:36:04:09 - 00:36:06:13 The brown or dead 00:36:06:13 - 00:36:10:03 plant material that was previously alive that year. 00:36:11:15 - 00:36:13:20 Looking at the 00:36:13:20 - 00:36:18:05 patterns across some select ecoregions of our study area. 00:36:18:06 - 00:36:20:07 This is the green component. 00:36:20:07 - 00:36:23:19 So again we would expect an increase and then decrease. 00:36:24:16 - 00:36:28:01 So these patterns follow the expected 00:36:28:01 - 00:36:33:12 phenology here in terms of when the peak green cover is 00:36:33:13 - 00:36:37:19 and the timing of the start and end of season. 00:36:38:00 - 00:36:42:10 Interesting also to compare across years. 00:36:42:20 - 00:36:45:10 So in this case I've plotted the study 00:36:45:10 - 00:36:48:14 area wide means by year 00:36:48:19 - 00:36:52:09 for six of our EEG components showing. 00:36:52:13 - 00:36:54:22 And then the lines are color coded by year. 00:36:54:22 - 00:36:57:22 So focusing on cheatgrass as an example here 00:36:59:05 - 00:37:01:10 we can see there's a lot of variability 00:37:01:10 - 00:37:05:19 between years in terms of when green up occurs. 00:37:06:01 - 00:37:12:01 And the relative amount of cheatgrass is pretty variable, especially compared to 00:37:12:01 - 00:37:17:03 something like total or basis, where the the degree of variability is much lower. 00:37:17:22 - 00:37:22:05 All of this underscores what we see in the literature that cheatgrass 00:37:22:12 - 00:37:27:03 and invasive grass cover in general, tends to be much more dynamic temporarily 00:37:27:09 - 00:37:30:09 than perennials are that e.g., data 00:37:30:12 - 00:37:33:15 reflect fire nearly immediately. 00:37:33:21 - 00:37:35:06 So this is an example. 00:37:35:06 - 00:37:38:06 In California, the cornucopia fire. 00:37:38:14 - 00:37:40:00 And that occurred 00:37:41:02 - 00:37:42:18 about week 28. 00:37:42:18 - 00:37:45:18 Here we can see as soon as that fire occurred 00:37:45:18 - 00:37:50:03 we well, actually even prior to that fire occurring, green was going down 00:37:50:09 - 00:37:53:15 making that eventual fire more likely. 00:37:54:02 - 00:37:57:07 And then after fire, we lose all our green cover. 00:37:57:13 - 00:37:59:14 And that goes to senescence. 00:37:59:14 - 00:38:05:06 That's one caveat in these products, really, the fire would consume that 00:38:05:17 - 00:38:09:09 to cover instead of it going up, it should go down as well. 00:38:09:09 - 00:38:12:09 That's something we're working on in future generations 00:38:12:11 - 00:38:15:21 in terms of improving that post-fire representation. 00:38:15:22 - 00:38:17:05 Change accuracy. 00:38:17:05 - 00:38:22:22 So all the accuracy statistics I've shown so far are comparing the accuracy 00:38:22:22 - 00:38:26:16 of our map, a single date of our map 00:38:26:16 - 00:38:30:14 to a single date of validation data. 00:38:30:15 - 00:38:35:18 The BLM aim database does contain lots of examples of plots 00:38:35:18 - 00:38:41:22 that have been observed multiple times, both across years and even within a year. 00:38:41:22 - 00:38:45:19 So here's an example of an aim point in Nevada 00:38:45:20 - 00:38:50:00 that's been observed six times starting in 2013. 00:38:50:16 - 00:38:53:10 We've got plot photos here showing the conditions 00:38:53:10 - 00:38:56:10 of what that plot looked like through time. 00:38:56:12 - 00:38:59:23 And I've got the Aim data plotted in these orange dots, 00:39:00:03 - 00:39:04:02 the RCMAP components at that same location 00:39:05:01 - 00:39:06:20 with the blue line. 00:39:06:20 - 00:39:11:06 So we can see in both annular basis and in bare ground. 00:39:11:21 - 00:39:15:16 The sea map components do a nice job of capturing 00:39:15:18 - 00:39:19:19 that temporal pattern observed in AME data. 00:39:20:09 - 00:39:24:23 This is an example of within year change from AME data. 00:39:25:10 - 00:39:28:05 This is in the Mojave Desert in Nevada. 00:39:28:05 - 00:39:31:05 Looking at the year 2020, 00:39:31:09 - 00:39:35:09 this site has been observed five times within 2022, 00:39:35:10 - 00:39:39:13 rather in June and September of 22. 00:39:39:18 - 00:39:43:21 Quite a bit of difference even in that shorter time period. 00:39:45:05 - 00:39:49:04 A little hard to see, but there's a patch of cover here 00:39:49:04 - 00:39:53:03 that really is less dense in June, so they must 00:39:53:03 - 00:39:58:14 have had a pretty good monsoon season this year to get more cover. 00:39:59:00 - 00:40:01:10 So in this case we can see the aim. 00:40:01:10 - 00:40:06:13 Observations of urban cover have increased from here 00:40:06:13 - 00:40:10:15 to here, and our predictions have done the same. 00:40:10:15 - 00:40:14:18 They have increased nearly synonymously with AME data. 00:40:14:21 - 00:40:17:23 In this example, if we plot these out through time, 00:40:18:00 - 00:40:23:20 all the course bonding AME observations versus the course week of our 00:40:24:21 - 00:40:25:20 cover, we 00:40:25:20 - 00:40:28:20 see a pretty nice correlation there. 00:40:29:01 - 00:40:31:05 Moving over to Oregon, 00:40:31:05 - 00:40:33:18 focusing on phenology here. 00:40:33:18 - 00:40:38:08 This is an Aim site that was observed in May and June of 2025. 00:40:38:23 - 00:40:43:03 You can see the phenology is much more progressed on the right. 00:40:43:14 - 00:40:48:09 We've lost most of that green cover in this month span, 00:40:48:22 - 00:40:51:12 and our data correspond 00:40:51:12 - 00:40:54:19 pretty well with that in terms of less green cover 00:40:54:22 - 00:40:57:22 in June as compared to May. 00:40:58:14 - 00:41:00:21 One thing to note here is that 00:41:00:21 - 00:41:04:04 satellite products tend to lag reality 00:41:04:04 - 00:41:10:03 by about two weeks, so probably more fair to move this green curve a little bit 00:41:10:03 - 00:41:13:22 to the left in this comparison, which would better align the products. 00:41:13:22 - 00:41:18:13 And if we plot all of these out, there's like 8000 examples of interannual 00:41:18:14 - 00:41:22:23 change in AME data and about a couple 00:41:23:00 - 00:41:25:23 thousand examples within your change. 00:41:25:23 - 00:41:29:19 So focusing on the cover component from RCMAP, 00:41:30:04 - 00:41:34:09 we can see that aim change on the 00:41:34:09 - 00:41:37:21 x axis versus map change on the y axis. 00:41:38:13 - 00:41:41:16 We do a pretty good job of capturing 00:41:41:16 - 00:41:45:19 that Interannual change in the mapping product 00:41:46:20 - 00:41:49:20 also should highlight, as I mentioned before, 00:41:49:22 - 00:41:54:13 the map change slope is less than ame, 00:41:55:01 - 00:41:59:01 meaning that the map change tends to be less 00:41:59:01 - 00:42:02:01 than the real world observed change. 00:42:02:22 - 00:42:06:13 A pretty nice correlation with the interannual 00:42:06:13 - 00:42:10:14 within season change as well with e.g. 00:42:10:16 - 00:42:11:09 cover. 00:42:11:09 - 00:42:17:11 We've done some work with the weekly EEG products to understand 00:42:17:16 - 00:42:22:22 how the landscape condition currently differs from the long term mean. 00:42:22:22 - 00:42:24:16 So what we've done here. 00:42:24:16 - 00:42:27:05 This is an example from 2026. 00:42:27:05 - 00:42:29:11 Week 23. 00:42:29:11 - 00:42:34:08 We took the mean of all prior week 23 in our products. 00:42:34:08 - 00:42:40:06 So the mean of the basis cover from 2016 through 2025. 00:42:40:06 - 00:42:46:01 And then compare that from our 2026 week 00:42:46:02 - 00:42:49:17 23 map and mapped out the spatial patterns here. 00:42:49:17 - 00:42:53:02 So blue meaning we in week 23 of 00:42:53:02 - 00:42:58:02 26 had more cover than the long term mean. 00:42:58:20 - 00:43:03:07 And red being we have less verbatim cover than the long term mean. 00:43:03:23 - 00:43:07:04 So this would be about June of this year. 00:43:07:21 - 00:43:10:22 I did some field work in this part of the world around Gillette, 00:43:11:00 - 00:43:15:20 Wyoming about that time, and indeed it checks out in the field. 00:43:15:20 - 00:43:20:19 It was very, very dry in this part of the world at that time. 00:43:22:10 - 00:43:23:21 Another interesting thing to 00:43:23:21 - 00:43:28:23 point out here is all these blue dots in Oregon, Washington, etc. 00:43:29:00 - 00:43:31:16 are burned areas. 00:43:31:16 - 00:43:35:08 So places where there was tree cover at some point 00:43:35:08 - 00:43:38:20 in that reference period between 16 and 25. 00:43:39:01 - 00:43:41:05 And now we're a post burn. 00:43:41:05 - 00:43:44:05 So or cover has increased 00:43:44:09 - 00:43:47:05 as tree cover was reduced with fire 00:43:47:05 - 00:43:51:11 leading to that positive deviation in herbaceous cover. 00:43:52:00 - 00:43:55:03 So on the left we can see what that looks like temporarily. 00:43:56:02 - 00:43:59:21 This is from 2026 focusing on total or basis. 00:44:00:07 - 00:44:03:15 I've got the long term mean per week plotted in blue 00:44:03:16 - 00:44:07:04 with standard deviation in that shade. 00:44:07:20 - 00:44:12:19 We started the year above normal, probably as a result of a warm 00:44:12:19 - 00:44:16:09 winter across a lot of the West, and as the season progressed, 00:44:16:22 - 00:44:21:15 we eventually went below the long term mean by June 00:44:21:16 - 00:44:25:15 our data both the RCMAP and RCMAP EAG. 00:44:25:18 - 00:44:30:14 The weekly products are available on the MRC Rangeland Viewer, 00:44:30:15 - 00:44:35:16 where you can download custom extents of the products, do time series 00:44:35:16 - 00:44:40:05 plotting by a series of different polygon types, so plot long 00:44:40:06 - 00:44:44:03 term trends by watershed by a lot meant, etc. 00:44:44:06 - 00:44:47:09 the data are also available for static download on MRL. 00:44:47:09 - 00:44:50:16 See we've got w w CS services 00:44:50:18 - 00:44:54:01 on MRC, also on ArcGIS online. 00:44:54:03 - 00:44:57:01 The data are also in the Google Earth 00:44:57:01 - 00:45:00:20 Community Catalog, specifically the awesome Community catalog 00:45:00:21 - 00:45:06:04 there on the Desert Research Institute Climate engine there on Science Base. 00:45:06:04 - 00:45:10:02 We try to make these as broadly accessible as possible 00:45:10:03 - 00:45:11:21 and lots of different venues. 00:45:11:21 - 00:45:17:03 So we are in the midst of releasing 2026 data for the E.g. 00:45:17:08 - 00:45:19:02 side of the project. 00:45:19:02 - 00:45:20:09 We'll continue that through. 00:45:20:09 - 00:45:23:05 Next week is actually our last release in the background. 00:45:23:05 - 00:45:26:21 We've been releasing the 2016 through 2025, E.g. 00:45:26:22 - 00:45:27:09 data. 00:45:27:09 - 00:45:30:18 We've got a paper summarizing the updated e.g. 00:45:30:20 - 00:45:36:07 methods in works, another paper summarizing the entire RCMAP project, 00:45:36:08 - 00:45:39:09 and a third one looking at regional accuracy 00:45:39:09 - 00:45:44:03 patterns of RCMAP will produce a new generation of e.g. 00:45:44:04 - 00:45:46:07 products starting early next year. 00:45:46:07 - 00:45:50:17 Same with a new generation of RCMAP early next year. 00:45:50:18 - 00:45:53:23 We are looking into expanding the number 00:45:53:23 - 00:45:57:00 of unique species we're mapping on the E.g. 00:45:57:01 - 00:46:01:22 side, with some more emphasis on some of the species in the southwest 00:46:02:00 - 00:46:05:18 like red brome, buffel grass, etc. 00:46:05:20 - 00:46:08:22 we're also exploring additional modes 00:46:08:22 - 00:46:14:01 of data input, so specifically hyperspectral data, 00:46:14:11 - 00:46:18:08 SAR data, etc., further improving our products. 00:46:18:08 - 00:46:21:21 And then lastly, this is still in its infancy, 00:46:22:00 - 00:46:25:20 but we are looking into better incorporation 00:46:25:20 - 00:46:30:20 of RCMAP data into both land fire and NLCD. 00:46:30:23 - 00:46:34:12 And with that I am happy to take any questions. 00:46:34:12 - 00:46:38:07 We've got contact information noted here, website link, 00:46:38:08 - 00:46:42:12 viewer link and please reach out with any questions. 00:46:42:13 - 00:46:43:22 Perfect. Thanks, Matthew. 00:46:43:22 - 00:46:44:09 And Matthew 00:46:44:09 - 00:46:45:14 is going to leave this slide up 00:46:45:14 - 00:46:47:11 for the rest of the webinar, just so you can write this down 00:46:47:11 - 00:46:49:05 if you need any of this contact information. 00:46:49:05 - 00:46:53:05 But for those listening in, I have added a few final links to the chat 00:46:53:08 - 00:46:55:12 which include more information about RCMAP 00:46:55:12 - 00:46:58:02 as well as links to sign up for a few of our listeners, 00:46:58:02 - 00:47:02:05 a link to watch the recorded webinars, and a few social media channels 00:47:02:05 - 00:47:03:12 to follow up. You're interested. 00:47:03:12 - 00:47:06:23 I do want to thank Matthew for a really great presentation at this time. 00:47:06:23 - 00:47:11:15 Please add any final questions you have about the data or tools covered today, or 00:47:11:15 - 00:47:15:05 perhaps any future data or tutorial ideas you'd like to see with RCMAP? 00:47:15:05 - 00:47:18:23 Please feel free to add those to the Q&A feature and teams or the chat now 00:47:19:01 - 00:47:22:08 you can also email them into serve at USGS gov. 00:47:22:10 - 00:47:25:00 If the chat and QA feature does not work for you, 00:47:25:00 - 00:47:28:12 but with that, we will transition over to the Q&A portion of the webinar. 00:47:28:14 - 00:47:32:10 The first question we have is from Scott. 00:47:32:13 - 00:47:38:08 How does RCMAP reconcile low temporal or spatial support for a given state? 00:47:38:09 - 00:47:42:14 For example, states east of Colorado or sampled substantially less frequently 00:47:42:14 - 00:47:43:16 or spatially? 00:47:43:16 - 00:47:47:02 Even with 70 million training points, there are magnitudes 00:47:47:02 - 00:47:50:21 more unmonitored pixels and potentially unobserved vegetation. 00:47:50:21 - 00:47:52:13 Community distributions. 00:47:52:13 - 00:47:54:02 Yeah, that's a good question. 00:47:54:02 - 00:47:55:15 The density of training data 00:47:55:15 - 00:47:59:10 is definitely lower in the Great Plains than it is further west. 00:47:59:10 - 00:48:03:15 So we do have a number of our high resolution 00:48:03:15 - 00:48:06:11 satellite sites in the Great Plains. 00:48:06:11 - 00:48:09:11 We've added a few more in Texas, 00:48:09:14 - 00:48:12:15 South Dakota, North Dakota, Nebraska, 00:48:12:18 - 00:48:17:15 and we are investigating always incorporating new lines of data there. 00:48:17:15 - 00:48:20:21 There are some land fire observations in the Great Plains. 00:48:20:21 - 00:48:25:00 We're looking into getting some data from the National Park Service there. 00:48:25:01 - 00:48:27:15 I am a data I believe it's called. 00:48:27:15 - 00:48:31:12 Yes, that is a bit of a weakness that the training 00:48:31:12 - 00:48:34:13 data density is lower in that part of the world. 00:48:34:13 - 00:48:38:08 Also, in general, the training data is less dense 00:48:38:09 - 00:48:42:04 earlier in the time series relative to later in the time series. 00:48:42:04 - 00:48:45:04 The BLM Aim data come online in about 2008. 00:48:45:06 - 00:48:48:16 Our high res site data come online in about 2013. 00:48:48:18 - 00:48:51:21 We do have some data going back through the whole archive 00:48:51:22 - 00:48:54:22 through 85, but it is less dense there as well. 00:48:54:22 - 00:48:57:10 So that is something to note. 00:48:57:10 - 00:48:58:18 All right. Thanks, Matthew. 00:48:58:18 - 00:49:00:21 Our next question comes from Megan. 00:49:00:21 - 00:49:02:01 She had two questions. 00:49:02:01 - 00:49:05:00 The first was if we can share where the webinar will be posted. 00:49:05:00 - 00:49:06:20 So I did add that link to the chat. 00:49:06:20 - 00:49:07:22 You know, she had several partners 00:49:07:22 - 00:49:11:03 that were not able to attend but are very interested in our sea map. 00:49:11:03 - 00:49:12:07 So excited to hear that. 00:49:12:07 - 00:49:15:04 And then Megan also had a question for you, Matthew. 00:49:15:04 - 00:49:18:17 Was there any exploration of changes in non or excuse me, 00:49:18:21 - 00:49:20:02 sorry if I'm mispronouncing that. 00:49:20:02 - 00:49:22:14 Shrub cover and height in the woody encroachment. 00:49:22:14 - 00:49:24:17 Partners in northeast have expressed interest 00:49:24:17 - 00:49:28:03 in both encroachment of trees and shrubs, and climate dependency 00:49:28:03 - 00:49:32:10 and treatment efficacy will likely vary across the functional groups. 00:49:32:12 - 00:49:37:13 Yeah, they are asking about Artemisia, which is the sagebrush genus. 00:49:37:17 - 00:49:41:04 And the question is have we explored non 00:49:41:04 - 00:49:45:05 artemisia changes in the Great Plains. 00:49:45:05 - 00:49:45:19 Was that it. 00:49:45:19 - 00:49:49:15 Was there any exploration of the changes in non artemisia shrub 00:49:49:15 - 00:49:51:23 cover and height in the woody encroachment. 00:49:51:23 - 00:49:54:23 Yes. So and Megan note in the chat in Nebraska. 00:49:55:00 - 00:49:56:01 Okay. Yes. 00:49:56:01 - 00:50:01:16 So that would be captured in our shrub cover 00:50:01:19 - 00:50:06:18 trend in places where there are no sagebrush present. 00:50:06:21 - 00:50:10:18 The shrub cover trend would be that precisely. 00:50:10:18 - 00:50:13:21 So we can see the woody expansion in the Great Plains. 00:50:14:00 - 00:50:19:12 You can kind of see this pattern where it starts off as an increase in shrubs. 00:50:19:21 - 00:50:23:07 What those shrubs really are are juvenile trees. 00:50:23:08 - 00:50:28:09 Trees less than three meters tall would be called shrubs in our product. 00:50:28:09 - 00:50:32:07 And then eventually, once the tree individuals who pass that height 00:50:32:08 - 00:50:36:12 threshold, then that becomes a tree cover trend. 00:50:36:19 - 00:50:39:20 So all of that being said, in the Great Plains, 00:50:40:03 - 00:50:43:17 the first early warning sign of this woody 00:50:43:17 - 00:50:48:11 expansion would typically be increasing trend in shrubs, 00:50:48:14 - 00:50:51:14 which would be, again, those juvenile trees. 00:50:51:14 - 00:50:52:23 All right. Thank you, Matthew. 00:50:52:23 - 00:50:55:20 The next question we have and I will note it is 1:00, 00:50:55:20 - 00:50:58:05 but we're going to extend the webinar for about five more minutes 00:50:58:05 - 00:50:59:23 while we can get through all these questions. 00:50:59:23 - 00:51:02:22 So if you need to hop off. Thank you all for joining us. 00:51:02:22 - 00:51:05:00 But our next question is from Karen. 00:51:05:00 - 00:51:07:08 Have you been able to map yet? 00:51:07:08 - 00:51:10:21 Yeah. It's not a it's another is another. 00:51:10:21 - 00:51:11:05 Yeah. 00:51:11:05 - 00:51:14:14 The botanists like to make pronunciation difficult. 00:51:14:15 - 00:51:17:17 Vince Nata is another important invasive grass. 00:51:17:17 - 00:51:20:09 We have looked into that. 00:51:20:09 - 00:51:25:04 The challenge we're facing there is it is just not very well 00:51:25:04 - 00:51:29:17 represented in the BLM Aim database at this point. 00:51:29:21 - 00:51:33:12 I think there was just a couple hundred maybe observations 00:51:33:12 - 00:51:36:04 total within that Aim database. We are trying. 00:51:36:04 - 00:51:39:14 I mentioned the red brome and buffel 00:51:39:14 - 00:51:43:01 grass being ones that we're going to look into as well. 00:51:43:02 - 00:51:48:12 A lot of our invasive grass species are really Great Basin centric. 00:51:48:15 - 00:51:51:21 The cheatgrass medusa head, but also the field brome 00:51:51:21 - 00:51:55:02 component includes Japanese brome, 00:51:55:02 - 00:51:59:21 which is a pretty common invasive grass across a lot of the Great Plains. 00:51:59:23 - 00:52:03:13 So definitely looking into expanding that list, 00:52:03:16 - 00:52:06:15 I would say the less common 00:52:06:15 - 00:52:10:14 that species is in the training data, the more difficulty 00:52:10:15 - 00:52:14:08 we will have mapping it, but definitely looking into it. 00:52:14:09 - 00:52:15:03 All right. Excellent. 00:52:15:03 - 00:52:16:21 Our next question is 00:52:16:21 - 00:52:20:21 can this product or model be available to use outside of the US? 00:52:21:05 - 00:52:24:03 I want to use it for another one that I'm going 00:52:24:03 - 00:52:27:09 to have to pronounce Pentagon Harlem mapping. 00:52:27:11 - 00:52:29:12 Okay. That one I'm not familiar with. 00:52:29:12 - 00:52:33:03 And no, it's not currently available outside of the US, 00:52:33:05 - 00:52:38:02 aside from that little portion of Canada that's within the sagebrush biome, 00:52:38:05 - 00:52:43:02 we are considering expanding the study area to cover all of Conus. 00:52:43:03 - 00:52:43:20 At least. 00:52:43:20 - 00:52:47:12 I'm not sure if that would cover the users study area of interest. 00:52:47:12 - 00:52:52:18 There are some global fractional cover products available based on Motus. 00:52:52:18 - 00:52:57:11 If I recall correctly, it wouldn't be as detailed as what we do. 00:52:57:12 - 00:53:01:12 It would just be shrub or basis, maybe bare ground, 00:53:01:13 - 00:53:04:13 I think maybe three components and maybe an NPV 00:53:04:14 - 00:53:07:14 and non photosynthetic veg component as well. 00:53:07:14 - 00:53:08:12 And of course 00:53:08:12 - 00:53:13:19 or facial resolution to my knowledge that would be available at a global scale. 00:53:13:19 - 00:53:18:15 But we currently don't have any plans to expand RCMAP at a global level. 00:53:18:16 - 00:53:19:22 Okay. Thank you Matthew. 00:53:19:22 - 00:53:23:15 And our last question is broadly, how do RCMAP data 00:53:23:16 - 00:53:27:13 compare to RAP data or Rangeland Analysis Platform? 00:53:27:15 - 00:53:29:15 Apologies if you covered this and I missed it. 00:53:29:15 - 00:53:31:17 Yeah that's a really great question. 00:53:31:17 - 00:53:33:12 And it's a big topic. 00:53:33:12 - 00:53:38:03 So the rangeland analysis platform data, the RAP data produced by the ARS 00:53:38:11 - 00:53:42:17 and University of Montana and others is similar 00:53:42:17 - 00:53:46:15 in many respects to our data. 00:53:46:21 - 00:53:49:15 The key difference is our training data. 00:53:49:15 - 00:53:54:01 We've got those high resolution sites that are not used in the rangeland 00:53:54:01 - 00:53:55:07 analysis platform. 00:53:55:07 - 00:53:58:07 Our post-processing methods differ. 00:53:58:07 - 00:54:00:12 Some of our modeling methods differ. 00:54:00:12 - 00:54:04:01 All of this difference is good news for users. 00:54:04:01 - 00:54:08:23 It's good that the products are different in that you get more statistical power 00:54:09:04 - 00:54:12:06 in having a different modeling framework. 00:54:12:06 - 00:54:15:08 So generally speaking, the two data 00:54:15:09 - 00:54:20:12 sets are pretty similar, especially the bigger and temporal scale 00:54:20:13 - 00:54:23:13 you look at, the more similar they will become. 00:54:23:20 - 00:54:27:14 There are some important differences that I don't have time to go into. 00:54:27:15 - 00:54:32:21 We do have a paper in the works with rap people as coauthors 00:54:32:21 - 00:54:35:21 talking about some of these differences, so look out for that. 00:54:35:21 - 00:54:40:04 Did also mention earlier that there is benefit from looking 00:54:40:04 - 00:54:44:12 at a consensus of both Rap data and or C map data. 00:54:44:12 - 00:54:48:14 So for example, if shrub cover was of interest, a mean 00:54:48:23 - 00:54:52:10 or some other statistic, considering the data from both RAP 00:54:52:10 - 00:54:56:18 and RCMAP could be of benefit and often is of benefit. 00:54:56:19 - 00:54:57:23 Excellent. Thank you Matthew. 00:54:57:23 - 00:55:01:11 We are at 105 so we are going to conclude the webinar. 00:55:01:12 - 00:55:04:17 Clarinda, I do see that you had a question about some open spaces. 00:55:04:17 - 00:55:07:15 So I'll make sure we touch base with you offline on that one. 00:55:07:15 - 00:55:10:08 But I did want to thank everyone for attending the webinar. 00:55:10:08 - 00:55:14:11 If you have any additional questions later on about see map, the RCMAP Viewer, 00:55:14:11 - 00:55:18:10 or any data, tools or services the USGS EROS provides, 00:55:18:12 - 00:55:21:19 you can always email our team at custserv@USGS.gov. 00:55:22:01 - 00:55:23:02 I'd like to once again 00:55:23:02 - 00:55:26:08 thank our speaker, Matthew for his time and great presentation 00:55:26:10 - 00:55:28:09 and our team in the background that helped answer 00:55:28:09 - 00:55:30:20 some of those great questions you all had today. 00:55:30:20 - 00:55:34:07 If you'd like to be notified of when the next webinar will occur, please 00:55:34:07 - 00:55:36:04 sign up for our EROS User Services 00:55:36:04 - 00:55:39:21 listserv by emailing that custserv@USGS.gov email. 00:55:40:04 - 00:55:42:21 Again, thank you all for taking the time to join us today. 00:55:42:21 - 00:55:44:18 We hope you have a wonderful rest of the day. 00:55:44:18 - 00:55:46:05 We'll see you next time. Thank you.