U.S. Geological Survey Matthew Rigge Brett Bunde Kory Postma 20260508 Raster Digital Data Set Earth Resources Observation and Science Center, Sioux Falls, SD U.S. Geological Survey RCMAP Data Releases ---------- Rigge, M.B., Bunde, B., and Postma, K., 2025, Rangeland Condition Monitoring Assessment and Projection (RCMAP) Fractional Component Time-Series Across Western North America from 1985-2024: U.S. Geological Survey data release, https://doi.org/10.5066/P13QF8HT. ----- Rigge, M., Bunde, B., Postma, K., and Shi, H., 2024. Rangeland Condition Monitoring Assessment and Projection (RCMAP) Fractional Component Time-Series Across the Western U.S. 1985-2023: U.S. Geological Survey data release, https://doi.org/10.5066/P9SJXUI1. ----- Rigge, M.B., Bunde, B., Postma, K., and Shi, H., 2022, Rangeland Condition Monitoring Assessment and Projection (RCMAP) fractional component time-series across the Western U.S. 1985–2021: U.S. Geological Survey data release. https://doi.org/10.5066/P9ODAZHC. ----- Rigge, M., B. Bunde, D. Meyer, H. Shi, and K. Postma. 2021. Rangeland Condition Monitoring Assessment and Projection (RCMAP) Fractional Component Time-Series Across the Western U.S. 1985-2020. U.S. Geological Survey data release, https://doi.org/10.5066/P95IQ4BT ----- Homer, C., Rigge, M., Shi, H., Meyer, D., Bunde, B., Granneman, B., Postma, K., Danielson, P., Case, A., and Xian, G., 2020, Remote Sensing Shrub/Grass National Land Cover Database (NLCD) Back-in-Time (BIT) Products for the Western U.S., 1985 - 2018: U.S. Geological Survey data release, https://doi.org/10.5066/P9C9O66W. ----- RCMAP Validation ---------- Rigge, M.B., Bunde, B., and Postma, K., 2024, Rangeland Condition Monitoring Assessment and Projection (RCMAP) Independent Validation Data: U.S. Geological Survey data release, https://doi.org/10.5066/P13Y9PSG ----- Rigge, M., C. Homer, H. Shi, and D. Meyer. 2019. Validating a Landsat time-series of fractional component cover across Western U.S. rangelands. Remote Sensing 11, 3009. ----- RCMAP Foundational Papers - Cross References and Larger Work Citation are listed here ---------- Abbott, B., Perry, J. and Wallace, J. 2008. Land Condition Monitoring in the Rangelands of the Burdekin Dry Tropics Region. Report for the Burdekin Dry Tropics Board, 4. ----- Arkle, R., Pilliod, D., Germino, M., Jeffries, M., and Welty, J., 2022. Reestablishing a foundational species: limitations on post-wildfire sagebrush seedling establishment. Ecosphere 13: e4195. https://doi.org/10.1002/ecs2.4195 ----- Board, David I.; Urza, Alexandra K.; Bradford, John B.; Brown, Jessi L.; Chambers, Jeanne C.; Schlaepfer, Daniel R.; Short, Karen C. 2024. Maps of abiotic susceptibility versus fire-induced conversion to cheatgrass dominance in the sagebrush biome and associated data. Fort Collins, CO: Forest Service Research Data Archive. https://doi.org/10.2737/RDS-2024-0041 ----- Filippelli, S.K., Schleeweis, K., Nelson, M.D., Fekety, P.A. and Vogeler, J.C., 2024. Testing temporal transferability of remote sensing models for large area monitoring. Science of Remote Sensing, 9, p.100119. https://doi.org/10.1016/j.srs.2024.100119 ----- Keeley, J.E. and Keeley, S.C. (1981), POST-FIRE REGENERATION OF SOUTHERN CALIFORNIA CHAPARRAL. American Journal of Botany, 68: 524-530. https://doi.org/10.1002/j.1537-2197.1981.tb07796.x ----- Kearney, S.P., Porensky, L.M., Augustine, D.J., Derner, J.D. and Gao, F., 2022. Predicting spatial‐temporal patterns of diet quality and large herbivore performance using satellite time series. Ecological Applications, 32(2), p.e2503. https://doi.org/10.1002/eap.2503 ----- Macander, M.J., Nelson, P.R., Nawrocki, T.W., Frost, G.V., Orndahl, K.M., Palm, E.C., Wells, A.F. and Goetz, S.J., 2022. Time-series maps reveal widespread change in plant functional type cover across Arctic and boreal Alaska and Yukon. Environmental Research Letters, 17(5), p.054042. https://doi.org/10.1088/1748-9326/ac6965 ----- Maestas, J.D., Campbell, S.B., Chambers, J.C., Pellant, M. and Miller, R.F., 2016. Tapping soil survey information for rapid assessment of sagebrush ecosystem resilience and resistance. Rangelands, 38(3), pp.120-128. https://doi.org/10.1016/j.rala.2016.02.002 ----- Murphy, S. and Lodge, G. 2002. Ground cover in temperate native perennial grass pastures. I. A comparison of four estimation methods. The Rangeland Journal, 24(2), pp.288-300. http://dx.doi.org/10.1071/RJ02016 ----- Rigge, M., Bunde, B., McCord, S.E., Harrison, G., Assal, T.J. and Smith, J .L., 2025a. Spatial scale dependence of error in fractional component cover maps. Rangeland Ecology & Management, 99, pp.77-87. https://doi.org/10.1016/j.rama.2025.01.004 ----- Rigge, M., Case, M.F., Shaff, S.E., Ellsworth, L., Bunde, B. and Postma, K., 2025b. Correspondence Between Satellite-Derived and Long-Term Field Observations of Vegetation Cover at Great Basin Experimental Treatments. Rangeland Ecology & Management, 103, pp.341-355. https://doi.org/10.1016/j.rama.2025.09.007 ----- Rigge, M., C. Homer, L. Cleeves, D. Meyer, B. Bunde, H. Shi, G. Xian, and M. Bobo. 2020. Quantifying western U.S. rangelands as fractional components with multi-resolution remote sensing and in situ data. Remote Sensing. 12: 412. ----- Rigge, M., H. Shi, C. Homer, P. Danielson, and B. Granneman. 2019. Long-term trajectories of fractional component change in the Northern Great Basin, USA. Ecosphere: e02762. ----- Savitzky, A., Golay, M.J., Smoothing and Differentiation of Data by Simplified Least Squares Procedures. Analytical Chemistry 1964. 36(8): 1627–1639. https://doi.org/10.1021/ac60214a047 ----- Shi, H., Rigge, M., Postma, K., and Bunde, B. 2022. Trends Analysis of Rangeland Condition Monitoring Assessment and Projection Fractional Component Time-Series (1985-2020). GIScience and Remote Sensing, 59 1: 1243-1265. https://doi.org/10.1080/15481603.2022.2104786 ----- Urza, Alexandra K.; Board, David I.; Bradford, John B.; Brown, Jessi L.; Chambers, Jeanne C.; Schlaepfer, Daniel R.; Short, Karen C. 2024. Disentangling drivers of annual grass invasion: Abiotic susceptibility vs. fire-induced conversion to cheatgrass dominance in the sagebrush biome. Biological Conservation. 297: 110737. https://doi.org/10.1016/j.biocon.2024.110737 ----- Zhou, B., Okin, G.S. and Zhang, J., 2020. Leveraging Google Earth Engine (GEE) and machine learning algorithms to incorporate in situ measurement from different times for rangelands monitoring. Remote Sensing of Environment, 236, p.111521. https://doi.org/10.1016/j.rse.2019.111521 ----- Zhu, Z., Zhang, J., Yang, Z., Aljaddani, A.H., Cohen, W.B., Qiu, S. and Zhou, C., 2020. Continuous monitoring of land disturbance based on Landsat time series. Remote Sensing of Environment, 238, p.111116. https://doi.org/10.1016/j.rse.2019.03.009 ----- RCMAP Applications ---------- Rigge, M., Bunde, B., McCord, S.E., Harrison, G., Assal, T.J. and Smith, J.L., 2025. Spatial scale dependence of error in fractional component cover maps. Rangeland Ecology & Management, 99, pp.77-87. ----- Rigge, M., Bunde, B., Postma, K., Oliver, S., and Mueller, N. 2024. Application of Normalized Radar Backscatter and Hyperspectral Data to Augment Rangeland Vegetation Fractional Classification. Remote Sensing. 16, 2315. https://doi.org/10.3390/rs16132315. ----- Story on RCMAP. 2024. Maximizing Accuracy of Rangeland Data | U.S. Geological Survey (usgs.gov) ----- Allred, B., Cruetzberg, M., Carlson, J., Christopher, C., Dovichin, C., Duniway, M., Jones, M., Maestas, J., Naugle, D., Nauman, T., Okin, G., Reeves, M., Rigge, M., Savage, S., Twidell, D., Uden, D., Zhou, B. 2022. Guiding principles for using satellite-derived maps in rangeland management. Rangelands. 44: 78-86. https://doi.org/10.1016/j.rala.2021.09.004 ----- Shi, H., C. Homer, M. Rigge, K. Postma, and G. Xian. 2020. Analyzing change in a shrubland ecosystem with both long-term field observations and Landsat time-series in Wyoming, USA. Ecosphere: e03311. ----- Videos ---------- RCMAP Background and Data Access - https://www.youtube.com/watch?v=kPxhA_am2k0 https://doi.org/10.5066/P138CYEL https://www.mrlc.gov/data The RCMAP (Rangeland Condition Monitoring Assessment and Projection) dataset quantifies the percent cover of rangeland components across western North America using Landsat imagery from 1985-2025. The RCMAP product suite consists of ten fractional components: annual herbaceous, bare ground, herbaceous, litter, non-sagebrush shrub, perennial herbaceous, sagebrush, shrub, tree, and shrub height; in addition to the temporal trends of each component. Several enhancements were made to the RCMAP process relative to prior generations. 1) The training database was expanded by incorporating new observations from the BLM Analysis Inventory and Monitoring dataset, the Global Biodiversity Information Facility (GBIF) data, Hopping data (Sorrentino et al. In Prep), Johnston data (Johnston et al. 2025), and additional Landsat scale observations collected by RCMAP. 2) 10 new high-resolution training sites were added in the Great Plains. 3) Enhanced noise filtering post-processing using the Savitsky-Golay filter, and 4) integration of Exotic Annual Grass products into the RCMAP hierarchy. RCMAP data can be used to answer critical questions regarding the influence of climate change and the suitability of management practices. Component products can be downloaded at https://www.mrlc.gov/data and are available in an interactive viewer: MRLC Rangeland Viewer. Additionally, data are available on Google Earth Engine, and as WMS/WCS services Data Services Page | Multi-Resolution Land Characteristics (MRLC) Consortium. The goal of RCMAP is to provide a Landsat imagery-based time series of rangeland fractional components across Western North America from 1985 to 2025. This dataset contributes to improved monitoring of rangeland change at broad temporal and spatial extents. Components are defined as: Bare Ground; exposed soil, sand, and rocks. Annual Herbaceous; including grasses and forbs whose life history is complete in one growing season. This component is primarily dominated by annual invasive species including Cheatgrass (Bromus tectorum), Medusahead (Taeniatherum caput-medusae), Red Brome (Bromus rebens), or annual mustards such as Tumble Mustard (Sisymbrium altissimum) and Tansy Mustard (Descurainia pinnata), but it may contain substantial native annual herbaceous vegetation at higher elevations and in California. This component is nested within Herbaceous as a secondary component. Herbaceous; grasses, forbs and cacti which were photosynthetically active at any point in the year of mapping. Non-sagebrush shrub; includes all shrub species not of the sagebrush (Artemisia spp.) genus. Shrub; vegetation with woody stems and less than 6-m in height. Perennial herbaceous; grasses, forbs and cacti which were photosynthetically active at any point in the year of mapping and whose lifecycle includes more than one growing season (includes biennials). Litter; dead standing woody vegetation, detached plant organic matter and biological soil crusts. Sagebrush (Artemisia spp.) including Big Sagebrush (A. tridentata spp.), Low Sagebrush (A. arbuscula), Black Sagebrush (A. nova), Three-tip Sagebrush (A. triparta), Silver Sagebrush (A. cana), Sand Sagebrush (A. filifolia), Bigelow sagebrush (A. bigelovii), California sagebrush (A. californica), Channel Island sagebrush (A. nesiotica), Scabland sagebrush (A. rigida), Sticky sagebrush (A. rothrocki), Snowfield sagebrush (A. spiciformis), and Bud sage (A. spinescens). Sagebrush is nested within Shrub as a secondary component. Excludes the low stature prairie sage (A. frigida) and white sagebrush (A. ludoviciana). Tree; vegetation with persistent woody stems greater than 6m in height. Mature stands of pinyon (Pinus spp. and juniper (Juniperus spp.) are included regardless of height. Shrub height; the average height of all shrub in centimeters. This component only occurs where the shrub cover component is greater than 0%. Height is given for the portion of pixel with shrubs present. For example, in a pixel where shrub cover is 10% and the average height of those shrubs is 100cm, height will be given as 100cm, not 100cm/10% cover as 10cm. Although this Federal Geographic Data Committee-compliant metadata file is intended to document the data set in nonproprietary form, as well as in Esri format, this metadata file may include some Esri-specific terminology. 19850101 20251231 publication date Complete As needed -128.0026 -90.1430 51.5761 25.8371 ISO 19115 Topic Category biota environment geoscientificInformation imageryBaseMapsEarthCover USGS Thesaurus shrubland ecosystems terrestrial ecosystems Alexandria Digital Library Feature Type Thesaurus shrublands time series back-in-time trends grassland change shrubland change vegetation change climate change rangeland management None shrub sagebrush big sagebrush herbaceous annual herbaceous litter grass vegetation bare ground rangeland shrubland shrub height vegetation height tree Common geographic areas United States Washington Oregon Montana North Dakota South Dakota Wyoming Idaho Nebraska Nevada Utah California Colorado Arizona Texas New Mexico Kansas Oklahoma Minnesota Iowa Missouri Louisiana Desert North Plains Plains The Rockies Canada Alberta Saskatchewan None NM TX AZ CO CA UT NV NE ID WY SD ND MT OR WA KS OK MN IA MO LA Great Basin Arizona Plateau Black Hills Blue Mountains Chihuahuan Desert Colorado Plateau Columbia Plateau Grand Canyon Middle Rockies Rocky Mountains Gunnison Sonoran Desert Southwest Tablelands Wasatch Western US Yellowstone Mediterranean California Northern Great Plains Northern Rocky Mountains Wyoming Basin Mojave Sonoran Sierra Nevada Chihuahuan Southern Rocky Mountains Great Plains Southern Great Plains Northern Great Plains Sandhills Gulf Coast AB SK Prairie Provinces Any downloading and use of these data signifies a user's agreement to comprehension and compliance of the USGS Standard Disclaimer. Ensure all portions of metadata are read and clearly understood before using these data to protect both user and USGS interests. There is no guarantee of warranty concerning the accuracy of the data. Users should be aware that these data were developed from models which can contain some local error. Users should not use these data for critical applications without a full awareness of their limitations. Users should also be aware of the impact of spatial scale on application error (Allred et al. 2022). Rigge et al. (2025a) quantified the trade-offs between data granularity and error related to scale for fractional vegetation cover, specifically finding substantial reduction in error when data are aggregated into a 90 m (3 by 3 pixel) mean compared to single pixels. This finding of accuracy benefit in aggregation was underscored in Rigge et al. (2025b), who further noted differences in accuracy by disturbance/treatment type. Users should also note that regional variation in accuracy relative to the values reported below can occur. Artifacts in Landsat imagery composites and other input data can result in corresponding artifacts in predicted cover. Our validation data represents the time period of 2008-2025 and primarily covers the spatial extent of BLM-managed lands. Our training data covers the full temporal extent of our mapping period and the full spatial extent but primarily focuses on the period of 2008-2025 and BLM-managed lands. Therefore, our validation results may degrade outside of the core temporal period of training (i.e. pre-2008, Filippelli et al. 2024) and outside the core training extent (i.e. in the Great Plains and Pacific Northwest). Acknowledgement of the originating agencies would be appreciated in products derived from these data. Any user who modifies these data is obligated to describe the types of modifications they perform. User specifically agrees not to misrepresent the data, nor to imply that changes made were approved or endorsed by the U.S. Geological Survey. Please refer to https://www.usgs.gov/privacy.html for the USGS disclaimer. U.S. Geological Survey Customer Services Representative mailing

47914 252nd Street

Sioux Falls SD 57198-0001 U.S. 605-594-6151 605-594-6589 custserv@usgs.gov U.S. Geological Survey (USGS) and Bureau of Land Management (BLM) Microsoft Windows 11 and Esri ArcGIS Pro Maps are rigorously validated using field data not included as training (i.e., independent) with data from long-term monitoring sites, and by assessing model fit to training data. The independent data consists of 1) 2,014 points, each specifically designed to represent a single Landsat pixel, collected from 2013-2023, 2) long-term monitoring data in southwest Wyoming at 126 plots observed 13 times between 2008 and 2025, and 3) a 10% withholding of BLM AIM and LMF data not used for training (n = 6,415) collected between 2011 and 2024. Correlations between RCMAP predictions and RCMAP independent validation sites (n = 2,014) were robust across all components, with an R2 of 0.71 for bare ground (RMSE = 17.6%) and 0.46 for shrub cover (RMSE = 10.3%) and average R2 of 0.52 and RMSE of 11.6% across components. Bare Ground - R2 0.71, RMSE 17.6%, Herbaceous - R2 0.66, RMSE 15.5%, Litter - R2 0.30, RMSE 9.2%, Shrub - R2 0.46, RMSE 10.3%, Sagebrush - R2 0.50%, RMSE 6.9, Annual Herbaceous - R2 0.48, RMSE 10.9%, Shrub Height – R2 0.32, RMSE 32.9 cm. 2). At Wyoming plots (n = 126 sites), the spatial-temporal correlation (n = 1,162) was robust for all components, with an R2 of 0.66 for bare ground (RMSE = 17.7%) and 0.39 for shrub cover (RMSE = 9.4%) and average R2 of 0.42 and RMSE of 10.2% across components. Next, at the 10% withholding of BLM AIM and LMF data not used for training (n = 6,598), correlations between RCMAP and AIM/LMF data were again robust across all components, with an R2 of 0.73 for bare ground, R2 of 0.77 for tree, and 0.56 for shrub cover and average R2 of 0.60 and RMSE of 9.5% across cover components. Bare Ground - R2 0.73, RMSE 12.1%, Herbaceous - R2 0.71, RMSE 13.2%, Litter - R2 0.24, RMSE 9.7%, Shrub - R2 0.56, RMSE 8.8%, Sagebrush - R2 0.62, RMSE 6.6%, Annual Herbaceous - R2 0.60, RMSE 10.8%, Tree - R2 0.77, RMSE 5.4%, Shrub Height - R2 0.29, RMSE 23.0 cm. In addition to independent validation, data were assessed using model cross-validation which evaluates model performance relative to a 10% withholding of training data. Cross-validation R2 averaged 0.91; Bare Ground - R2 0.95, RMSE 5.97%, Herbaceous - R2 0.92, RMSE 6.52%, Litter - R2 0.90, RMSE 3.17%, Shrub - R2 0.86, RMSE 6.79%, Sagebrush - R2 0.90, RMSE 2.41%, Annual Herbaceous - R2 0.90, RMSE 4.59%, Tree - R2 0.92, RMSE 5.82%, and shrub height – R2 0.93, RMSE 8.71 cm. Changes observed in both the field and RCMAP data were typically gradual, within-state, changes which are most difficult to resolve, which were often successfully captured. It is important to consider that all accuracy assessments described above are designed to evaluate single-pixel level correspondence. Due to fine-scale landscape heterogeneity this is the most rigorous approach, and most applications looking at broader spatial scales would tend to lower error relative to this validation analysis (Rigge et al. 2025a). Additionally, the independent data used to evaluate RCMAP accuracy also contains error, which is included in the reported accuracy metrics (i.e. our error metrics include error in both RCMAP predictions and independent data). The methods employed to map fractional vegetation components in rangeland ecosystems in the Western U.S. include modelling rangelands as a series of independent continuous field components from 1985 – 2025 using field observations, neural network classifiers, and multiple Landsat composites. This fractional estimation of ten rangeland habitat variables ranging from 1985 - 2025 in Western North America is the version dated May 2026. Data set is considered complete for the information presented, as described in the abstract. Users are advised to read the rest of the metadata record carefully for additional details. A formal accuracy assessment of the horizontal positional information in the data set has not been conducted. A formal accuracy assessment of the vertical positional information in the data set has either not been conducted or is not applicable. The RCMAP (Rangeland Condition Monitoring Assessment and Projection) dataset quantifies the percent cover of rangeland components across western North America using Landsat imagery from 1985-2025. The RCMAP product suite consists of ten fractional components: annual herbaceous, bare ground, herbaceous, litter, non-sagebrush shrub, perennial herbaceous, sagebrush, shrub, tree, and shrub height and the temporal trends of each. The five primary components, bare ground, shrub, litter, herbaceous, and tree are designed to sum to 100% in each pixel. The secondary components of annual herbaceous and perennial herbaceous are subsets of the primary component herbaceous. While the secondary components non-sagebrush shrub and sagebrush are subsets of the primary component shrub. Processing occurred in ten regions which were subsequently mosaicked across the study area. The ten regions: Mediterranean California, Great Basin and Columbia Plateau, Northern Rocky Mountains, Northern Great Plains, Southern Great Plains, Wyoming Basin, Warm Deserts (Mojave, Sonoran, and Chihuahuan), Southern Rocky Mountains and Sierra Nevada, Colorado Plateau, and Pacific Northwest. 2026 High-resolution training data - We used neural network models to predict component cover at high resolution sites, (n =367), pooled across similar ecological conditions (same level 3 ecoregion) into regional models. The imagery was split into 15 Worldview (WV) groups (average of 20.1 images per group), 4 Pleiades groups (10.5 images per group), and 1 QuickBird group (5 images). These image groups had an average of 2,106 training polygons (22,451 pixels), maximum of 4,907 (54,073 pixels), and minimum of 375 (1,440 pixels), compared to an average of 95 plots per image. Pooling data into regional models offers the advantages of increasing the size and range of training data and spectral conditions and improving regional consistency. While pooling data into ecoregion-based models can regress results toward the mean we found on average 15% improvement in high-resolution model accuracy in the pooled models as compared to independent runs, a pattern consistent with Zhou et al. (2020) and Kearney et al. (2022). Processing regionally also allows the addition of 51 WV images where no field data were collected due to access or other logistical challenges. These collects with a viable image, but no field data, had component cover predictions generated from the data available from sites within the same ecological grouping. In total, the high-resolution training footprint was ~62,090 km2. To improve consistency among regions some image sites and corresponding field data near ecotones were used to train multiple regions, and predictions for these sites were averaged across model runs. Independent variables for the high-resolution models included imagery, 3-by-3 focal coefficient of variation for all imagery bands, indices (NDVI, red/green, NIR/yellow), topographic elevation downscaled to 2-m, latitude, and longitude. Our final high-resolution model was a DNN (Deep Neural Network) with a neuron width of 1,024, 4-layers deep, learning rate of 10^-3, 20% dropout between layers, batch normalization, 10,000 epochs with early stopping given no improvement over 100 epochs, and clamping of 0-100% (0-500 cm for shrub height). A weighted mean squared logarithmic error (WMSLE) loss function was used, where zeros have half the weight of nonzero values to reduce the abundance of zeros in predictions. The point of the WMSLE approach is to penalize higher relative errors, for example a prediction of 5% cover with a training value of 0% is given a higher penalty than a prediction of 50% cover with a training value of 55%. Whereas commonly used loss functions such as MAE would consider the above example to be equivalent. A model was developed with all cover components and used the subsequent classifications as independent variables in the shrub height model. It was found that adding cover components as variables significantly improved shrub height classification accuracy as they, especially shrub cover, are strongly related to height. To test high-resolution model performance, a 5% sample was utilized. The 5% sample was selected randomly from a pool of plots greater than 90 m away from other sites (representing 41% of the total pool) to minimize spatial autocorrelation bias (e.g., Macander et al. 2022). High-resolution prediction cross validation accuracy (R2) was 0.73 for annual herbaceous, 0.83 for herbaceous, 0.66 for litter, 0.71 for sagebrush, 0.69 for shrub, 0.91 for bare ground, 0.89 for tree, 0.61 for shrub height, with an overall average of 0.75. Independent accuracy results were on average 26% weaker than cross validation. Spatial patterns in the high-resolution predictions closely matched expected patterns observed in the field. Following initial high-resolution classification, a series of post-processing steps was utilized to improve accuracy. 1) Since field plots in high-resolution sites had minimal tree cover, tree cover training values were extracted from a Unet-based tree cover classification with significant hand editing for each site. 100 tree cover training points (all at 100% cover) were added for each collect in the processing region. 2) Field plots contained minimal samples on hay, pasture, or otherwise irrigated land covers, resulting in overestimation of tree, shrub, and sagebrush cover over these areas. To remedy this issue, areas in which no woody cover is likely to occur (Annual NLCD hay/pasture class and classified as non-woody types (e.g. alfalfa, soybeans) in the cropland Data Layer [CDL, Boryan et al. 2011a]) were identified. In these areas, in the initial prediction the values for shrub, sage, and tree cover, were set to zero and we proportionally added their cover to herbaceous, annual herbaceous, litter, and bare ground, additionally, in these areas we set shrub height to zero. Next, 100 training points for each collect in the processing region were added from the hay-pasture extent, with the proportioned training data as input. 3) To fully leverage our Unet-based classification of trees, a convergence of evidence approach was used to label shrub and tree. Specifically, in pixels identified as tree in the Unet product, the shrub cover prediction was set to zero and added the initial prediction to tree cover. Conversely, in pixels identified as non-tree, the tree cover prediction was set to zero and added the initial prediction to shrub cover. Finally, in pixels sharing any boundary with a tree classified pixel, the initial shrub and tree predictions were kept. 4) The relationship between high-resolution derived training and the spatio-temporally corresponding Landsat imagery is the foundation of the RCMAP approach. Yet, in some cases, the Landsat imagery overlapping high-resolution sites contains various artifacts. We applied additional masking to remove portions of high-resolution sites with topographic shadows, clouds, cloud shadows, and areas of extreme variance. This was accomplished using a focal coefficient of variance and thresholds in various indices of the 10th, 50th and 90th percentile Landsat composites of the year of high-resolution data collection. Collectively, this approach removed ~4% of the high-resolution training footprint. The 2 m predictions of component cover were resampled to 30 m using bilinear interpolation. Next, at the 30 m scale we ensured the sum of primary components was 100% and the rectification was intact (shrub greater than sage, herbaceous greater than annual herbaceous, and shrub height was greater than 0 only if shrub cover was greater than 0. Since the collect data are based on ocular estimates, which have known bias relative to objective measurements (e.g., line point intercept) (Murphy and Lodge 2002, Abbott 2008), a correction function was applied to our data: visual cover = 0.925 X objective cover ^2). Specifically, the bias addresses a deflation in values in the mid-ranges (~20-80%) of visual data relative to objective data. Finally, to ensure proper training of burned areas, the 30 m scale high resolution collect data were conformed to the limits of shrub, sage, and tree cover based on time-since fire (Rigge et al. 2019) stratified by ecosystem resistance and resilience classes (Maestas et al. 2016). 2026 Image Processing – Collection 2 Landsat Analysis Ready Data (ARD) data provide improved spatial and radiometric qualities and more image availability, which enhance training accuracy and time-series consistency relative to collection 1 (Crawford et al. 2023). Landsat 5, 7, 8, and 9 data including 6 surface reflectance bands from the Collection 2 ARD and corresponding Quality Assessment (QA) bands were used in the current generation of RCMAP. A custom snow/cloud/shadow masking procedure was developed to augment pixel QA masking. First, a reference image was developed using all “leaf-on” images from 1985-2025 where the mean minus 2 standard deviations was calculated for each pixel and band. Next, for each new candidate collection 2 pixel, the Normalized Burn Ratio (NBR, Garcia et al. 1991), snow index (SWIR 2 x 100) / Red, and the Cloud Shadow Index (CSI) (Zhai et al. 2018) were calculated. Pixels were flagged with a mask value if 1) candidate collection 2 values for the NIR or SWIR 1 bands were lower than the reference image for that region, 2) CSI values indicate high confidence for clouds or shadows, or 3) snow index values (less than 81) indicate the presence of snow. To remove masking commission in dark lava, bright playas, and recently burned areas, the preliminary masking in pixels with an NBR less than 0.04 were removed. Finally, cloud/snow/shadow pixels were buffered by 1 pixel. QA (pixels not marked clear or water but may be shadow or snow) and additional custom snow/shadow/cloud masking were applied to all ARD imagery. Additionally, for each pixel, the Z-score of each band relative to the ARD observation mean and standard deviation were calculated, ignoring the areas masked as described above. Pixels where the range between the maximum and minimum Z-score among bands exceeded 6 were also masked. This additional Z-score masking procedure removed image artifacts, especially near scene edges, and snow cover. From the remaining clear pixels, the 10th, 50th, and 90th percentile values were calculated on a band-by-band basis for all images available within each evaluated year (1985-2025). Calculation of percentile values varied based on the number of clear observations available. 1) If pixel clear observation count was 0, the median of observations that passed the QA masking, but not custom masking was calculated. 2) If pixel clear observation count was 1 or 2, pixels greater than 2 standard deviations above or below the mean were removed, then calculated the percentile values from the remaining observations. 3) If pixel clear observation count was 3-11, pixels in the lowest and highest observations that year were removed, then calculated the percentile values from remaining observations. 4) If pixel clear observation count was greater or equal to 12, pixels in the 2 lowest and 2 highest observations that year were removed and calculated the percentile values from remaining observations. This staggered approach based on clear observation count produced the highest quality composites by removing noise (snow, shadows, cloud cover, image artifacts) often found in the highest and lowest observations within a year. The paradigm allows one to be more selective on pixel quality as clear count increases. For 2025, images were obtained through December 15. Landsat imagery was limited in 2012, so composites were generated as the average of 2011 and 2013, except for areas identified as burned in 2012 by MTBS, which were filled with 2013 only. 2026 Ancillary Data -- In addition to Landsat composites, several ancillary data layers were included as independent variables in the Landsat scale modelling of fractional components. 1) Topographic data; slope, Cartesian aspect, elevation, and position index. 2) For each Landsat composite, seven spectral indices were calculated, the Soil Adjusted Vegetation Index (SAVI), Normalized Difference Built-up Index (NDBI), Normalized Difference Water Index (NDWI), and tasseled cap wetness, dryness, brightness, and greenness. 3) Geographic location information (i.e., latitude and longitude). 4) Time since most recent fire (TSF) in years which was found to improve classification results in burned areas. 5) 3- by-3 focal coefficient of variation of Landsat percentile composite bands, which we found further improved results, especially for shrub and tree cover. 2026 Land Cover Masking -- A land cover mask based on pixels identified as non-rangeland was applied to the final products. Unmasked version is available upon request. The non-rangeland mask includes urban areas, major roads, snow, and ice identified in Annual NLCD (USGS, 2024) land cover classes in addition to cultivated crops and water. Cultivated croplands were masked using a combination of the 2013 Cropland Data Layer (Boryan et al. 2011) and NLCD classes. Open water areas were identified using Normalized Difference Water Index (NDWI) thresholds. NLCD hay-pasture and forest areas are not masked in the current generation of RCMAP. 2026 Training – The majority of our training pool is derived from a 10% random sample of high-resolution collect data (n= 68,882,813). To counteract model tendency to revert to the mean, we pull an additional 5% random sample of collect data with attention to the extremes of data ranges by component. Specifically, we calculate the z-score for each component cover value and the sample density from each percent cover bin increases with distance to mean. This additional data sample is completed on a component-by-component basis and pooled together across components. We converted AIM and Landscape Data Commons sites to a 3-by-3 grid of 30 m pixels to represent the sampled footprint following the RCMAP approach (Shi et al. 2022). [29,47,48]. We filter each of the 8 edge pixels in the grid for similarity to the center pixel, retaining only those pixels with <3% relative difference in NDVI. On average, this approach yields 8.3 training pixels per AIM plot. We sample all non-collect data (non-withheld AIM (n = 539,077)), LANDFIRE public reference database (n = 169,804), Landscape Data Commons (n = 5,808), Landsat scale training pixels (n = 303,850), Aldridge (2005) data from southeast Alberta from 1998-2003 (n = 1,590), UAV derived data from Johnston et al. (2025), (n = 2,520), GBIF observations from 1985-2025 within the RCMAP study area for the following species ; field brome (Bromus arvensis), rattlesnake brome (Bromus briziformis), rescuegrass (Bromus catharticus) Bald brome (Bromus commutatus and Bromus racemosus), ripgut brome (Bromus diandrus), soft brome (Bromus hordeaceus and Bromus hordeaceus spp. hordeaceus), Japanese brome (Bromus japonicus), compact brome (Bromus madritensis and Bromus madritensis ssp. Rubens), red brome (Bromus rubens), rye brome (Bromus secalinus), cheatgrass (Bromus tectorum), medusahead (Taeniatherum caput-medusae), Ventenata (Ventenata dubia), and all species of sagebrush listed previously ( n = 9,297), field samples from central Idaho (Sorrentino et al. In Prep) (n = 128). Similar to the high-resolution scale, agricultural area training at the Landsat scale serves to reduce the excess prediction of shrub and tree cover in agriculture and hay-pasture land covers. 10,000 training points (20,000 in the Northern and Southern Great Plains) were added per region in 2001 and 2019 (to correspond with NLCD epochs in areas with non-woody crop types as defined by CDL. As with the high-resolution scale, the initial prediction was used with zeroed out shrub, sage, shrub height, and tree cover, and proportionally added their cover to herbaceous, annual herbaceous, litter, and bare ground. AI models are often “data hungry” and one solution is incorporation of crowdsourced data (Derwin et al. 2025), in our case from GBIF. Derwin et al. (2025) note that crowdsourced data can differ from traditional professional samples but also note differences between individual collectors of traditional data. Our set of GBIF data constitutes a small portion of our overall set and thus had minimal impact on our overall accuracy numbers but may impact local results. The 10% random sample, z-score binned data, and non-collect training are appended and serve as the final training for the DNN models. Landsat scale neural net model architecture was like that described above for the high-resolution scale, though the list of independent variables differed. Variables used; topographic data; slope, Cartesian aspect, elevation, and position index, elevation, latitude, longitude, from each Landsat composite; SAVI, NDBI, NDWI, and tasseled cap wetness, brightness, dryness, and greenness, a TSF layer (time in years since most recent fire), and 3-by-3 focal coefficient of variance of all predictor variables. Models were developed for each region independently. 2026 Component Prediction – The spatio-temporal training pool was split by region and a DNN was developed for each and used to predict all components, using the component values as the dependent variable, and the full set of independent variables. Neural network models were developed for the spatio-temporal training dataset but applied the model to data from each target year to produce component cover predictions. The DNN models had a neuron width of 1,024, 4-layers deep, learning rate of 10^-3, 20% dropout between layers, batch normalization, 10,000 epochs with early stopping given no improvement over 100 epochs, and clamping of 0-100% (0-500 cm for shrub height). We used a zero-inflated log normal loss function. 2026 Post Processing – To improve the accuracy of post-fire component response, we implement a series of post-processing procedures. The primary goal of the procedures is to remedy the observation of shrub, tree, and sagebrush cover being predicted in excess (and vice versa for bare ground and herbaceous) in the years following fire. Comparison of RCMAP data to SageSTEP observation post-fire highlight that post-fire recovery signal in Landsat data outpaces that of field data (Rigge et al. 2025a). We developed fire recovery equations, stratified by ecosystem resistance and resilience (R and R) classes (Maestas et al. 2016), to limit shrub, sagebrush, and tree cover based on time since most recent fire. Ecosystem R and R maps are only available for the sagebrush biome. R and R classes were intersected with 1985-2020 average water year precipitation to identify precipitation thresholds corresponding to R and R classes. Outside of the sagebrush biome, precipitation was used to produce R and R equivalent (low, medium, high). Due to the fast recovery following fire in chapparal ecosystems (e.g., Keeley and Keeley 1981, Storey et al. 2016), EPA level 3 ecoregions were used to define a 4th chapparal R and R zone. Recovery rates are based on Arkle et al. (2022) who evaluated the recovery of plant functional groups in post-fire rehabilitation plots by time since disturbance stratified by ecosystem resistance and resilience. This analysis was expanded by evaluating the postfire-recovery in all AIM and LMF data across the West to establish second order polynomial equations defining maximum sage, shrub, and tree cover by time-since fire by R and R class. Recovery limits in California follow Keeley and Keeley 1981 and Storey et al. 2016. Fire correction percent cover limits (y) for sagebrush by each R and R zone were; zone 1) (high R and R) y = 0.0332x^2 – 0.1866x + 3.9794, zone 2) (medium R and R) y = 0.0234x^2 – 0.0988x + 1.9323, zone 3) (low R and R) y = 0.00074x^2 – 0.1535x + 0.676, zone 4) (California chapparal) y = 0.0332x2 – 0.1866x + 3.9794, where x is the number of years since most recent fire. Fire correction percent cover limits (y) for shrub and tree by each R and R zone were; zone 1) (high R and R) y = 0.0019x^2 + 1.611x + 3.564, zone 2) (medium R and R) y = 0.0141x^2 + 0.5558x + 2.540, zone 3) (low R and R) y = -0.0011x^2 + 0.5891x + 1.1267, zone 4) (California chapparal) y = 0.6.7857x^2 – 7.6429x + 20.571, where x is the number of years since most recent fire. It should be noted that a) less than 10% of the post-fire landscape is impacted by the recovery limits, b) the majority of the limitation occurs in the first 4 years following fire, and c) by limiting shrub, sage, and tree, the cover of the non-woody components (i.e. bare ground, herbaceous, litter) are proportionally increased in order to maintain a sum of 100% in these areas. Next, we apply a model to reduce noise in the time-series for each component. Most time-series noise detection procedures are not appropriate in rangelands due to their interpretation of rapid changes related to interannual weather variation noise (e.g. Zhu et al. 2020). We employ the Savitzky and Golay (1964) filter to detect and filter noise in the time-series, using a 5-year window. These noise removal procedures were not implemented post-fire for 2 years (annual herbaceous, bare ground, herb, litter) and for 10 years (sagebrush, shrub, tree, and shrub height). Additionally, the noise filtering cannot include any post-burn or year of burn. The intention of the noise filtering approach is to conservatively remove obvious examples of noise. For example, shrub cover predictions of 10%, 50%, and 12% in 2001, 2002, and 2003, respectively, where the 2002 observation would be identified and replaced with a value of 11% (mean of 2001 and 2003). 2026 Finalizing Steps – The data were mosaiced from the ten regions, which were designed with a 1.5 km overlap. The weighting function was calculated based on distance to the edge of the overlap to feather regional predictions together. The secondary components were reconciled to the primary components for sagebrush to shrub, annual herbaceous to herbaceous, and shrub height to shrub cover. Next, perennial herbaceous and non-sagebrush shrub components were calculated by subtracting annual herbaceous from herbaceous cover and sagebrush from shrub cover, respectively. The land cover mask based on pixels identified as non-rangeland in circa 2024 was applied. Finally, extent masking was applied to sagebrush and annual herbaceous components following Rigge et al. (2020). Sagebrush and annual herbaceous extent masking was based on thresholds combining elevation, aspect, and latitude. Specifically, these maps restricted the distribution of annual herbaceous at elevations greater than ~2400 m and sagebrush at greater than ~2800 m. Both the sagebrush and annual herbaceous extent masks are generally higher elevation than in prior versions of RCMAP. The annual herbaceous extent mask was modified to include pixels modelled to have a > 45% chance of annual herbaceous presence (in locations < 3000 m elevation) in the mapping extent (i.e., allow non-zero annual herbaceous predictions in these areas) (Board et al. 2024, Urza et al. 2024). In the Great Plains, we relied on ecological mapping systems data to restrict the distribution of sagebrush cover predictions; from Diamond et al. 2021 for Kansas and Nebraska, from the Colorado Natural Heritage program in Colorado, from Elliot et al. (2014) for Texas and Oklahoma. Shrub height values in locations with 0-10% shrub cover tend to be more unreliable, and in some cases will be predicted as 0 cm height. While we zeroed shrub height predictions in areas with no predicted shrub cover. 2026 Trends -- The temporal patterns were assessed in each RCMAP component with two approaches, 1) linear trends and 2) a breaks and stable states method with an 8-year temporal moving window based on structural change at the pixel level. Linear slope is calculated using the Thiel-Sen method. The slope represents the average percent cover change per year over the times-series and the p-value reflects the confidence of change in each pixel. The structural change method partitions the time-series into segments of similar slope values, with statistically significant breakpoints indicating perturbations to the prior trajectory. The break point trends analysis suite relies on structural break methods, resulting in the identification of the number and timing of breaks in the time-series, and the significance of each segment. The following statistics were produced: 1) for each component, each year, the presence/absence of breaks, 2) the slope, p-value, and standard error of the segment occurring in each year, 3) the overall model R2 (quality of model fit to the temporal profile), and 4) an index, Total Change Intensity. This index reflects the total amount of change occurring across components in that pixel. The linear and structural change methods generally agreed on patterns of change, but the latter found breaks more often, with at least one break point in most pixels. The structural change model provides more robust statistics on the significant minority of pixels with non-monotonic trends, while detrending some interannual signal potentially superfluous from a long-term perspective. 2026 Source input list ----- Aldridge, C. L. 2005. Identifying habitats for persistence of greater sage-grouse (Centrocercus urophasianus) in Alberta, Canada. Dissertation, University of Alberta, Edmonton, Canada ----- Boryan, C., Yang, Z., Mueller, R., Craig, M., 2011. Monitoring US agriculture: the US Department of Agriculture, National Agricultural Statistics Service, Cropland Data Layer Program. Geocarto International, 26(5), 341–358. https://doi.org/10.1080/10106049.2011.562309 ----- Crawford, C.J., Roy, D.P., Arab, S., Barnes, C., Vermote, E., Hulley, G., Gerace, A., Choate, M., Engebretson, C., Micijevic, E. and Schmidt, G., 2023. The 50-year Landsat collection 2 archive. Science of Remote Sensing, 8, p.100103. http://dx.doi.org/10.1016/j.srs.2023.100103 ----- Derwin, J.M., Thomas, V.A., Wynne, R.H., Schleeweis, K.G., Coulston, J.W., Peery, S.S., Luther, K., Liknes, G.C., Bender, S. and Sen, S., 2025. Factors influencing the consistency in crowdsourced interpretations of aerial photographs to measure tree canopy cover. Ecological Informatics, p.103300. https://doi.org/10.1016/j.ecoinf.2025.103300 ----- Diamond, D.D., L.F. Elliott, G. Steinauer, K. Kindscher, P. Hanberry, D. True, and M. Sunde. 2021. Ecological Systems of Kansas and Nebraska: Final Report. Submitted to Kansas Department of Wildlife, Parks and Tourism, and Nebraska Game and Parks Commission. http://dx.doi.org/10.7944/P9VLM7ZF ----- Elliott, Lee F., Amie Treuer-Kuehn, Clayton F. Blodgett, C. Diane True, Duane German, and David D. Diamond. 2009-2014. Ecological Systems of Texas: 391 Mapped Types. Phase 1 – 6, 10-meter resolution Geodatabase, Interpretive Guides, and Technical Type Descriptions. Texas Parks & Wildlife Department and Texas Water Development Board, Austin, Texas. ----- Garcia, M.L. and Caselles, V., 1991. Mapping burns and natural reforestation using Thematic Mapper data. Geocarto International, 6(1), pp.31-37. ----- GBIF.org (8 August 2025). GBIF Occurrence Download. https://doi.org/10.15468/dl.ydjdh4 ----- Johnston, A.N., Preston, T.M., Wilson, S., and Haynam, R., 2025, Sagebrush Maps for the Northern Great Plains, 2023: U.S. Geological Survey data release, https://doi.org/10.5066/P149BWDK. ----- Sorrentino, M., de Graaff, M.A., Hulet, A., Kehler, R., Walter, K., Simler-Williamson, A., Swette, B., Arispe, S., and Hopping, K.A., Ecosystem effects of targeted grazing with sheep to manage cheatgrass. For submission to Rangeland Ecology and Management. ----- Storey, E. A., Stow, D. A., and O'Leary, J. F., (2016), Assessing postfire recovery of chamise chaparral using multi-temporal spectral vegetation index trajectories derived from Landsat imagery. Remote Sensing of Environment, 183(15): pp.53-64. https://doi.org/10.1016/j.rse.2016.05.018 ----- U.S. Geological Survey (USGS), 2024, Annual NLCD Collection 1 Science Products: U.S. Geological Survey data release, https://doi.org/10.5066/P94UXNTS ----- Zhai, H., Zhang, H., Zhang, L. and Li, P., 2018. Cloud/shadow detection based on spectral indices for multi/hyperspectral optical remote sensing imagery. ISPRS journal of photogrammetry and remote sensing, 144, pp.235-253. 2026 Raster Grid Cell 99344 115597 1 NAD 1983 Contiguous USA Albers 29.5 45.5 -96.0 23.0 0.0 0.0 row and column 30.0 30.0 meters North American Datum of 1983 (NAD 83) NAD_1983 6378137.000000 298.257222 Attribute Table Table containing attribute information associated with the data set. Producer defined Value Column for value indicating per-pixel percent for bare ground, shrub, herbaceous, litter, sagebrush, annual herbaceous, non-sagebrush shrub and perennial herbaceous components range from 0 to 100. Producer defined NoData NoData value Producer defined 0 100 Percent Value (Shrub Height) In shrub height, column for value indicating per-pixel average height of all shrub in centimeters. Shrub height values greater than 0 only occur where the shrub cover component is greater than 0% Height is given for the portion of pixel with shrubs present. The shrub height component ranges from 0 to 500 cm. Producer Defined 0 500 Centimeters Count All raster attribute tables include a column for count describing a nominal integer value that designates the number of pixels that have each value in the file; histogram column in ERDAS Imagine raster attributes table. Producer defined 0 125551462 Integer Red Red color code for RGB. The value is arbitrarily assigned by the display software package, unless defined by user. Producer defined 0 100 Percent Green Green color code for RGB. The value is arbitrarily assigned by the display software package, unless defined by user. Producer defined 0 100 Percent Blue Blue color code for RGB. The value is arbitrarily assigned by the display software package, unless defined by user. Producer defined 0 100 Percent Opacity A measure of how opaque, or solid, a color is displayed in a layer. Producer defined 0 100 Percent The entity and attribute information provided here describes the tabular data associated with the data set. Please review the detailed descriptions that are provided (the individual attribute descriptions) for information on the values that appear as fields/table entries of the data set. The entity and attribute information were generated by the individual and/or agency identified as the originator of the data set. Please review the rest of the metadata record for additional details and information. U.S. Geological Survey GS ScienceBase mailing address

Denver Federal Center, Building 810, Mail Stop 302

Denver CO 80225 United States 1-888-275-8747 sciencebase@usgs.gov Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although these data and associated metadata have been reviewed for accuracy and completeness and approved for release by the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data for other purposes, nor on all computer systems, nor shall the act of distribution constitute any such warranty. Raster Digital Data Set https://doi.org/10.5066/P138CYEL None 20260423 U.S. Geological Survey Customer Services Representative mailing and physical

47914 252nd Street

Sioux Falls SD 57198-0001 U.S. 605-594-6151 605-594-6589 custserv@usgs.gov FGDC Content Standard for Digital Geospatial Metadata FGDC-STD-001-1998