The primary raw data are aerial counts of beaver (Castor canadensis) colonies on streams across the Chequamegon-Nicolet National Forest (CNNF). These aerial counts were performed in the fall of each year from 1987 (Nicolet side of CNNF) or 1997 (Chequamegon side of CNNF). Based on the colony counts, we also provide derived beaver colony density values. The surveyed streams were classified into four categories: managed trout, non-managed trout, managed non-trout, and non-managed non-trout. Trout versus non-trout status was assigned by the CNNF using Wisconsin Department of Natural Resources information. Managed streams were those on which targeted removal of beavers was conducted in the spring of each year under a contract with USDA-Wildlife Services. Data also include proportion of stream-side aspen, temperature, snowfall, and soil moisture as measured by the Palmer Drought Severity Index.

This data publication contains the results of chemical and mineral analyses on masticated particles from mixed-conifer forests in 15 study locations. These data were collected from 2012 through 2016 as part of the MASTIDON project. The MASTIDON project was a four-year research project to study how masticated material differs when treated with different cutting machines and how the masticated particles decompose when left on the ground for multiple years. It investigated masticated materials in four states of the western United States. The project was funded by the Joint Fire Sciences Program (JFSP) and RMRS between 2013 and 2016. The masticated particles within this project had been decomposing in situ in wet and dry areas of Idaho, Colorado, New Mexico, and South Dakota since their initial treatment. Particles were tested from four shapes (circular, three-sided, four-sided, and small wood chips) and three size classes. Each shape and size class was ground, dried, and analyzed for percent carbon and nitrogen, cellulose and lignin, heat content, and mineral content (from the duff component) using three pieces of equipment. This data publication includes the results of each of these tests and files describing the MASTIDON project and its goals.

This data publication contains the results from 45 experimental burns and 48 smoldering tests on masticated materials from mixed-conifer forests. These data were collected from 15 study locations from 2012 through 2016 as part of the MASTIDON project. The MASTIDON project was a four-year study to describe the phyical characteristics of masticated materials that were treated with four different cutting heads in xeric and mesic environments. The main focus of the project was to evaluate how leaving the particles on the ground for varying lengths of time affected the burnability of the particles. The project was funded by the Joint Fire Sciences Program (JFSP) and RMRS between 2013 and 2016. The masticated particles were created by four different machines, including a vertical rotating head, horizontal drum, chipper, and mower. They had been decomposing in situ in wet and dry areas of Idaho, and dry areas of Colorado, New Mexico, and South Dakota since their initial treatment and were between 0 and 10 years old. The materials were burned at the RMRS Missoula Fire Sciences lab, Missoula, MT. The experimental burns were conducted in a combustion facility on a large fuel bed 0.68 square meters in size. The smoldering tests were conducted on beds 497 square centimeters in size under a fume hood in the soils laboratory. This download includes (1) data on fire behavior within the experimental burns, including rate of spread, flame height, flame duration, consumption, heat flux, moisture content, and more; (2) temperature data, burn durations, duff moistures and thicknesses from the smoldering tests; (3) photos of the experimental burn beds and smoldering beds; and (4) files describing the MASTIDON project and its goals.

This data publication contains the results of field work in masticated materials of mixed-conifer forests in 14 study locations. Mixed-conifer masticated materials were investigated in four states of the western U.S., including Idaho, Colorado, New Mexico, and South Dakota. The data were collected from 2012 through 2016 as part of the MASTIDON project, which was a four-year research project to characterize how burning properties of masticated material are affected when different cutting machines are used to treat the forests and when masticated particles are left on the ground for multiple years to decompose. The project was funded by the Joint Fire Sciences Program (JFSP) and RMRS between 2013 and 2016. The masticated particles within this project were created by four different machines, including a vertical rotating head, horizontal drum, chipper, and mower. They had been decomposing in situ in wet and dry areas of the mixed-conifer forests since their initial treatment. This publication gives GPS locations and laser elevation data for each field site and the GPS locations where depth measurements were taken within each macroplot. It gives depths for each of the five fuel layers distinguished within the masticated materials at two scales. The first scale is at three-meter intervals along each of six transect lines. The second scale is within each microplot and quarter plot where samples were taken from the quarter plots for further lab work. The data also contain estimates of vegetation cover and height at each of the depth-measurement locations.

This data publication contains the results of sorting masticated particles from mixed-conifer forests in 15 study locations. These data were collected from 2012 through 2016 as part of the MASTIDON project. The MASTIDON project was a four-year research project to study how masticated material differs when treated with different cutting machines and how the masticated particles decompose when left on the ground for multiple years. It investigated masticated materials in four states of the western United States. The project was funded by the Joint Fire Sciences Program (JFSP) and RMRS between 2013 and 2016. The masticated particles within this project were created by four different machines, including a vertical rotating head, horizontal drum, chipper, and mower. They had been decomposing in situ in wet and dry areas of Idaho, Colorado, New Mexico, and South Dakota since their initial treatment. Particles were broken down into 15 shape and three size classes. Each shape and size class was counted for total particles and weighed (in grams) for total fuel load by class. The total weights by shape and size class were then aggregated for a total fuel load for the 0.5 x 0.5 sample area at each location and converted to fuel loads for a 1 x 1 meter area. Subsamples of each shape and size class were taken to obtain specific information on the characteristics of particles in each class, such as average length, width, weight, particle density, volume, and surface area. This data publication includes field data on fuel loads, depth measurements, and bulk densities of five fuel layers; lab data from the sorting, characterization, and bulk density measurements of the fuel particles; and files describing the MASTIDON project and its goals.

This feature class represents the historical (1970-1999) scenario for bull trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields BT_0BRK - BT_100BRK indicate the probabilities of bull trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the mid-century (2030-2059) scenario for bull trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields BT_0BRK - BT_100BRK indicate the probabilities of bull trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the end-of-century (2070-2099) scenario for bull trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields BT_0BRK - BT_100BRK indicate the probabilities of bull trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the historical (1970-1999) scenario for bull trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields BT_0BRK - BT_100BRK indicate the probabilities of bull trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the mid-century (2030-2059) scenario for bull trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields BT_0BRK - BT_100BRK indicate the probabilities of bull trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the end-of-century (2070-2099) scenario for bull trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields BT_0BRK - BT_100BRK indicate the probabilities of bull trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the historical (1970-1999 scenario for cutthroat trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields CT_0BRK - CT_100BRK indicate the probabilities of cutthroat trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the mid-century (2030-2059) scenario for cutthroat trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields CT_0BRK - CT_100BRK indicate the probabilities of cutthroat trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the end-of-century (2070-2099) scenario for cutthroat trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields CT_0BRK - CT_100BRK indicate the probabilities of cutthroat trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the historical (1970-1999) scenario for cutthroat trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields CT_0BRK - CT_100BRK indicate the probabilities of cutthroat trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the mid-century (2030-2059) scenario for cutthroat trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields CT_0BRK - CT_100BRK indicate the probabilities of cutthroat trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

This feature class represents the end-of-century (2070-2099) scenario for cutthroat trout, derived from the Climate Shield fish distribution models. These models provide stream-specific probabilistic predictions about the occurrence of juvenile bull trout and cutthroat trout in association with three different scenarios for climate change and brook trout invasions. These datasets indicate all potential cold-water habitats less than 11 degrees Celsius. The attribute fields CT_0BRK - CT_100BRK indicate the probabilities of cutthroat trout occurrence within a cold-water habitat based on the prevalence of brook trout at 0%, 25%, 50%, 75%, or 100% of the sites within a habitat. The probabilities were predicted using the Climate Shield native trout models developed from known species occurrence in greater than 500 cold-water streams. The stream centerlines were based on the National Hydrography Dataset (NHD) but were modified for purposes of modeling and cross-walking to other datasets.

A stationary camera overlooking the Kendall sub-watershed in the Walnut Gulch Experimental Watershed used to track vegetation phenology (RGB and IR imagery). Images are taken every 30 minutes between 4:00am and 10:30pm local standard time. A link to the Phenocam's network FAQ: https://phenocam.sr.unh.edu/webcam/faq

A stationary camera overlooking the Lucky Hills sub-watershed in the Walnut Gulch Experimental Watershed used to track vegetation phenology (RGB and IR imagery). Images are taken every 30 minutes between 4:00am and 10:30pm local standard time. A link to the Phenocam's network FAQ: https://phenocam.sr.unh.edu/webcam/faq

This dataset is part of the common observation in the centralized repository for public access, also known as the Common Observatory Repository (CORe), of the USDA ARS Long-Term Agro-ecosystem Research (LTAR) network. This is part of the National Program 216 (NP#216): Agricultural System and Competitiveness and Sustainability. Also The National Wind Erosion Research Network was established in 2014 as a collaborative effort led by the US Department of Agriculture (USDA) Long Term Agro-Ecosystem Research (LTAR) network and the Bureau of Land Management (BLM). The research domain incorporates the diverse soils and vegetation communities in the rangelands and croplands of the western United States, with sites located in New Mexico, Texas, Oklahoma, Arizona, California, Colorado, North Dakota, Utah, Idaho and Washington. We have a tower that collects data for the North Dakota Agricultural Weather Network. Our site is part of the NEON project with a tower that is designed to collect and provide open data that characterize and quantify complex, rapidly changing ecological processes across the US. We have a National Center for Environmental Information (NOAA) tower that collects daily summaries of weather data. A Natural Resource Conservation (NRCS) National Water and Climate tower that collects snow and water data.

Note: To download this raster dataset, go to ArcGIS Open Data Set and click the download button, and under additional resources select any of the download options. Data can also be downloaded from the FSGeodata Clearinghouse.More information about rangeland productivity and the effects of drought are available in this StoryMap; additional drought and rangeland products from the Office of Sustainability and Climate are available in our Climate Gallery. Time enabled image service showing estimates of annual production of rangeland vegetation.Production data were generated using the Normalized Difference Vegetation Index (NDVI) from the Thematic Mapper Suite from 1984 to 2021 at 250 m resolution. The NDVI is converted to production estimates using two regression formulas depending on the level of the NDVI; there is one equation for lower values (and thus lower production values) and one for higher values. This raster dataset yields estimates of annual production of rangeland vegetation and should be useful for understanding trends and variability in forage resources. This raw lbs/acre data that the Z-scores were derived from as well as the Z-scores dataset can be downloaded from: https://data.fs.usda.gov/geodata/rastergateway/rangelands/index.phpMore information about rangeland productivity and the effects of drought are available in this story map.

Production data were generated using the Normalized Difference Vegetation Index (NDVI) from the Thematic Mapper Suite from 1984 to 2021 at 250 m resolution. The NDVI is converted to production estimates using two regression formulas depending on the level of the NDVI; there is one equation for lower values (and thus lower production values) and one for higher values.This raster dataset yields estimates of annual production of rangeland vegetation and should be useful for understanding trends and variability in forage resources.The Rangeland Productivity data can be downloaded here:https://data.fs.usda.gov/geodata/rastergateway/rangelands/index.php

This study used data from field plots in urban areas to describe forest structure (e.g., tree numbers, density, basal area, species composition) for six land use categories in six California climate zones: Southern California Coast, Inland Empire, Inland Valley, Southwest Desert, Northern, and Interior West. Two types of field plot data were utilized. The first set of data include 702 randomly sampled 0.04 hectare (ha) plots obtained from i-Tree Eco plot data for Los Angeles (in 2007-2008), Santa Barbara (2012) and the Sacramento area (2007). The second set of data (687 plots, in 2011) consisted of 0.067 ha (four 0.017 ha subplots) plots based on the Forest Service Forest Inventory and Analysis (FIA) plot design. The number of plots collected varied by climate zone and a total of 3,796 trees were sampled. Data collection included percentage of tree canopy cover over the plot, tree species, stem diameter at breast height (1.37 meters above ground, dbh), tree height, crown width, distance and azimuth to buildings that fit the requirements as specified in the i-Tree Eco and Urban FIA manuals.

This data publication contains urban tree inventory data for 929,823 street trees that were collected from 2006 to 2013 in 49 California cities. Fifty six urban tree inventories were obtained from various sources for California cities across five climate zones. The five climate zones were based largely on aggregation of Sunset National Garden Book's 45 climate zones. Forty-nine of the inventories fit the required criteria of (1) included all publicly managed trees, (2) contained data for each tree on species and diameter at breast height (dbh) and (3) was conducted after 2005. Tree data were prepared for entry into i-Tree Streets by deleting unnecessary data, matching species to those in the i-Tree database, and establishing dbh size classes. Data included in this publication include tree location (city, street name and number), diameter at breast height, species name and/or species code, and tree type.

The United States Department of Agriculture-Agricultural Research Service (USDA-ARS) North Central Soil Conservation Research Laboratory - Soil Management Unit established a weather data collection system at the Swan Lake Research Farm in 1997. Weather data collected include wind speed and direction, barometric pressure, relative humidity, air temperature, soil temperatures, soil heat flux, solar radiation, photosynthetic active radiation, and precipitation. In 2015 the site became part of the Long Term Agroecosystem Research (LTAR) project. The Swan Lake Research Farm is located in Stevens County Minnesota, in the Upper Mississippi River Basin (UMRB) watershed.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.

TreeMap 2016 provides a tree-level model of the forests of the conterminous United States.Metadata and DownloadsWe matched forest plot data from Forest Inventory and Analysis (FIA) to a 30x30 meter (m) grid. TreeMap 2016 is being used in both the private and public sectors for projects including fuel treatment planning, snag hazard mapping, and estimation of terrestrial carbon resources. We used a random forests machine-learning algorithm to impute the forest plot data to a set of target rasters provided by Landscape Fire and Resource Management Planning Tools (LANDFIRE: https://landfire.gov). Predictor variables consisted of percent forest cover, height, and vegetation type, as well as topography (slope, elevation, and aspect), location (latitude and longitude), biophysical variables (photosynthetically active radiation, precipitation, maximum temperature, minimum temperature, relative humidity, and vapour pressure deficit), and disturbance history (time since disturbance and disturbance type) for the landscape circa 2016. The main output of this project (the GeoTIFF included in this data publication) is a raster map of imputed plot identifiers at 30X30 m spatial resolution for the conterminous U.S. for landscape conditions circa 2016. In the attribute table of this raster, we also present a set of attributes drawn from the FIA databases, including forest type and live basal area. The raster map of plot identifiers can be linked to the FIA databases available through the FIA DataMart (https://doi.org/10.2737/RDS-2001-FIADB). The dataset has been validated for applications including percent live tree cover, height of the dominant trees, forest type, species of trees with most basal area, aboveground biomass, fuel treatment planning, and snag hazard. Application of the dataset to research questions other than those for which it has been validated should be investigated by the researcher before proceeding. The dataset may be suitable for other applications and for use across various scales (stand, landscape, and region), however, the researcher should test the dataset's applicability to a particular research question before proceeding. This raster dataset represents model output generated by a random forests method that assigns Forest Inventory Analysis plot identifiers to a 30x30m grid (Riley et al. 2016 and Riley et al. 2021). Some attributes provided have been validated as detailed below, and we have high confidence they would be suitable for stand, county, and national scale analyses. Other attributes have not been validated as of this writing on 2/25/2022. Accuracy may vary regionally. This dataset is for the landscape circa 2016 and does not capture disturbances such as fire and land management after that date. Based on a set of FIA validation plots, these data have moderate to high accuracy at point locations for forest cover, height, vegetation group, and recent disturbance by fire and insects and disease (Riley et al. 2021). Summary statistics at Baileys section and subsection levels indicate high accuracy in most sections and subsections when compared to FIA statistics for live basal area, number of live trees greater than or equal to 1 diameter, live cubic-foot volume, and live-tree biomass. Estimates of number of dead trees greater than or equal to 5 diameter and dead tree above-ground biomass have lower correlations with FIA estimates, which are driven largely by the fact that TreeMap does not include areas where live tree cover is less than 10% while FIA does, meaning that severely disturbed areas are not included in mapping. In general, the TreeMap data are appropriately used for planning and policy-level analyses and decisions. Local map accuracy is suitable for many local-scale decisions regarding questions around forest cover, height, vegetation group, and recent disturbances. For other attributes provided here, formal validation has not been completed, and assessment at local scales is advised and must be driven by project-specific needs. References: Riley, Karin L., Isaac C. Grenfell, and Mark A. Finney. 2016. Mapping Forest Vegetation for the Western United States Using Modified Random Forests Imputation of FIA Forest Plots. Ecosphere 7 (10): e01472. https://doi.org/10.1002/ecs2.1472. Riley, Karin L., Isaac C. Grenfell, Mark A. Finney, and John D. Shaw. 2021. TreeMap 2016: A Tree-Level Model of the Forests of the Conterminous United States circa 2016. https://doi.org/10.2737/RDS-2021-0074.