{9780F143-C78D-4FB6-91C1-DDCDCAE30EBE} eng; USA utf8 dataset EnviroAtlas Coordinator U.S. Environmental Protection Agency, Office of Research and Development-Sustainable and Healthy Communities Research Program, EnviroAtlas Geospatial Data Owner (919) 541-3832 109 T.W. Alexander Drive Research Triangle Park NC 27709 EnviroAtlas@epa.gov https://www.epa.gov/enviroatlas pointOfContact 2025-02-13 ISO 19115 Geographic Information - Metadata ISO 19115:2003/Cor.1:2006(E) EnviroAtlas - Milwaukee, WI - Meter-Scale Urban Land Cover Data (MULC) Data (2010) 2017-03-14 publication U.S. EPA Office of Research & Development (ORD) - National Exposure Research Laboratory (NERL) Research Triangle Park NC publisher US Environmental Protection Agency (US EPA) (additional partners are listed in Supplemental) originator The EnviroAtlas Milwaukee, WI Meter Urban Land Cover (MULC) data and map were generated from USDA NAIP (National Agricultural Imagery Program) four band (red, green, blue and near infrared) aerial photography from Late Summer 2010 at 1 meter spatial scale. Eight land cover classes were mapped: water, impervious surfaces (dark and light), soil and barren land, trees and forest, grass and herbaceous non-woody vegetation, agriculture, and wetlands (woody and emergent). An accuracy assessment yielded an overall user's accuracy (MAX) of 76.2 percent and a fuzzy user's accuracy (RIGHT) of 85.5 (see Overview section). The area mapped is defined by the US Census Bureau's 2010 Urban Statistical Area for Milwaukee. This dataset was produced by Oneida Total Integrated Enterprises (OTIE), its subcontractor, the University of Arkansas Center for Advanced Spatial Technologies (CAST), and the US EPA to support research and online mapping activities related to EnviroAtlas. EnviroAtlas (https://www.epa.gov/enviroatlas) allows the user to interact with a web-based, easy-to-use, mapping application to view and analyze multiple ecosystem services for the contiguous United States. The dataset is available as downloadable data (https://edg.epa.gov/data/Public/ORD/EnviroAtlas) or as an EnviroAtlas map service. Additional descriptive information about each attribute in this dataset can be found in its associated EnviroAtlas Fact Sheet (https://www.epa.gov/enviroatlas/enviroatlas-fact-sheets). The purpose of the Meter-scale Urban Land Cover (MULC) map is to provide a detailed snapshot of the urban environment and ecosystem for EnviroAtlas' analysis of multiple human and ecosystem health metrics and services. This analysis is typically performed by U.S. Census block group. Greenspace and impervious surface land cover classes figure prominently. Land cover maps are integral to sound, scientific decision making and effective stewardship. A high-resolution urban land cover map quantitatively and qualitatively represents the physical environment and ecosystem matrix of a city. Land cover varies at a fine scale in urban environments, and the existing NLCD (National Land Cover Database, www.mrlc.gov) land cover maps at 30 m spatial resolution are too coarse to measure many urban phenomena, and may not be sufficiently current in rapidly changing urban environments. There are many possible applications, for example: - Knowing the exact location and size of green and gray spaces; - Identifying highly impervious areas that would most benefit from additional trees and other vegetation to mitigate surface runoff and urban heat island effects; - Measuring the width of urban riparian buffers; - Locating potential wildlife corridors, gaps and linkages; - Planning park development for maximum benefits. The overall goal of EnviroAtlas is to employ and develop the best available science to map indicators of ecosystem services production, demand, and drivers for the nation. completed EnviroAtlas Coordinator U.S. Environmental Protection Agency, Office of Research and Development-Sustainable and Healthy Communities Research Program, EnviroAtlas Geospatial Data Owner (919) 541-3832 109 T.W. Alexander Drive Research Triangle Park NC 27709 EnviroAtlas@epa.gov https://www.epa.gov/enviroatlas pointOfContact asNeeded Ecology Ecosystem Environment Health Impact Land theme EPA GIS Keyword Thesaurus Agriculture Biophysical Communities Ecosystem Services EnviroAtlas Environment Environmental Atlas Green Space Land Land Cover Surface Water Water Wetlands theme EnviroAtlas Wisconsin Milwaukee, WI place None https://edg.epa.gov/EPA_Data_License.html otherRestrictions Use Constraints: None. Please check sources, scale, accuracy, currency and other available information. Please confirm that you are using the most recent copy of both data and metadata. Acknowledgement of the EPA would be appreciated. otherRestrictions Access Constraints: None. PUBLIC https://project-open-data.github.io/schema/#accessLevel Standard Technical Controls eng; USA environment geoscientificInformation imageryBaseMapsEarthCover -88.54936 -87.79614 42.84283 43.44913 2010-01-01 Oneida Total Integrated Enterprises (OTIE) and the University of Arkansas Center for Advanced Spatial Technologies (CAST) contributed to this dataset. true Entity and Attribute Information Entity and Attribute Overview: For easier readability, this XML metadata can be opened in a text editor (e.g., Notepad). The original Milwaukee, WI land cover dataset was developed in two parts: one by the U.S. Environmental Protection Agency (EPA) and the other by Oneida Total Integrated Enterprises (OTIE) and its subcontractor, the University of Arkansas Center for Advanced Spatial Technologies (CAST). An update was completed by EPA in 2017 to correct errors in the original dataset. The Milwaukee, WI land cover dataset was developed using 85 United States Department of Agriculture (USDA) National Agriculture Imagery Program (NAIP) Digital Orthophoto Quarter Quadrangle (DOQQ) ("quarter quads"). EPA developed 19 quarter quads and OTIE/CAST developed 66 quarter quads. Each quarter quad contains 4 bands including visible RGB and a Near Infrared (NIR) band with 1 meter pixel scale. Section 1. OTIE/CAST Methods and Accuracy Assessment Oneida Total Integrated Enterprises (OTIE) and its subcontractor, the University of Arkansas Center for Advanced Spatial Technologies (CAST), developed a land cover classification using eCognition, an Object Based Image Analysis (OBIA) software, to classify 66 of the NAIP quarter quads in the study area. Ancillary data was incorporated into the classification process and includes the following: 1. 2006 National Land Cover Database (NLCD) from the U.S. Geological Survey. This dataset was used to initially outline large contiguous features for further image segmentation. 2. High Resolution National Hydrography Dataset (NHD), used to initially outline and identify large to medium-sized water bodies. 3. Road Vectors from the County of Milwaukee and OpenStreetMaps.org, that were rasterized and buffered at various widths: (Residential Roads buffered 4 meters in each direction; Secondary and Primary Roads buffered 5 meters in each direction; Major Roads buffered 6 meters in each direction.) 4. For 19 of the 66 quarter quads in the study area a Height-Above-Ground map derived from Light Detection and Ranging (LiDAR) was used to augment the image classification. This 2010 dataset came from Milwaukee County. 5. In addition to the acquired datasets, eCognition was used to create a number of other out-of-the-box and custom "features" including homogeneity layers for the NAIP bands, contrast layers for the NAIP bands, a segment area border ratio layer, segment size layer, etc. In total 7 custom features were generated. The imagery was classified into the following land cover classes using the methodology below to process all data: (these were reclassified to match classes used by the EPA): 1 - Dark impervious surfaces: Areas classified as Impervious Surface that have a value in the NAIP visible green band equal to or greater than 200. 2 - Light impervious surfaces: Areas classified as Impervious Surface that have a value in the NAIP visible green band less than 200. 3 - Water: Areas classified as water that are at least 4 square meters in size. 4 - Grass (including all herbaceous non-woody vegetation): Areas classified as grass that are at least 4 square meters in size. 5 - Trees and forest: Areas classified as Tree that are at least 2 square meters in size. 6 - Bare earth (sand, beaches, mining operations, excavated construction sites, etc.): Areas classified as Bare Earth that are at least 4 square meters in size. 7 - Agricultural fallow/bare: Areas classified as Agriculture that are at least 4 square meters in size and little or no vegetation. 8 - Agricultural vegetated: Areas classified as Agriculture that are at least 4 square meters in size and have significant vegetative cover. 9 - Emergent wetlands: Areas identified as grass in the raster classification that are labeled as "wetland" in the National Wetland Inventory (NWI) database. 10 - Woody wetlands: Areas identified as trees and forest in the raster classification that are labeled as "wetland" in the National Wetland Inventory (NWI) database. A. Data Preprocessing: The datasets acquired were processed, as needed, in order to ensure that they were in the same geographic projection and datum. The NLCD raster dataset was resampled to fit in the 1 meter grid of the NAIP data. The remaining vector datasets were rasterized and resampled to fit in the 1 meter grid of the NAIP data. The LiDAR points were imported into eCognition and the point cloud was converted to a 1 meter grid height-above-ground-layer. Finally, all datasets were combined into an eCognition "Workspace" for automated processing. B. Rule Set Development (OBIA) and Image Classification: Image processing rulesets were developed in eCognition. The basic steps in ruleset development are: - Class hierarchy generation based on the category definitions described above, - Multiple image segmentation layering for optimal delineation of variously sized objects across various scales and reshaping procedures. - Exploration and band/layer threshold determination and classification - Object attribution based on identified thresholds. - Application of minimum mapping unit - Testing of ruleset across multiple heterogeneous images. C. Automated Workflow: Once the ruleset was determined to be operationally stable and to produce meaningful results across the study area, an eCognition workspace was created for automated processing. The following steps were used to automate the batch workflow. - Loading all Quarter-Quad rasters into the workspace, - Creating 1950 x 1950 pixel tiles for all Quarter-Quads (12 for each QQ) - Configuring the processing grid structure (primary node, data server node, and worker nodes) for use in eCognition GRId Server. - Running ruleset on each tile. - Mosaicking tiles back into full quad. - Exporting completed projects to ERDAS Image (.IMG) format. D. OTIE/CAST Accuracy Assessment of Rasters: The accuracy assessment was completed only on the raster data classifications resulting from the eCognition processing. (See Categories and Definitions, above). The following processing steps comprised the accuracy assessment process. Initial reference point selection and assignment: Before the classification was complete (independent from the classification process) 480 points were assigned to coordinates at random across the entire study area (six points per quad). Once generated each point was assigned to one of the land-cover categories. Additionally, a confidence value was attached to each point in order to give some indication of how confident the technician was that the point was assigned to the appropriate category. It was also noted in the point (vector) file if the point was within a 4 meter radius of a different land-cover category, and if the point was located in shadow. Upon completion of the image classification and generation of the output raster images, a histogram was used to determine the categories for which extra accuracy assessment points were needed. Additional points were also added in order to make each category have at least 50 accuracy assessment points. The final point count was 605 points. The following confusion matrix summarizes the accuracy assessment data, as performed by CAST. For easier readability, this XML metadata can be opened in a text editor (e.g., Notepad) and the confusion matrix may be copied from text editor to an Excel spreadsheet. Confusion matrices can also be viewed in original formatting by opening in ArcCatalog: OTIE/CAST Confusion Matrix for 66 Quarter Quadrangles RASTER INTERP Ag D_IMP GH L_IMP SB TF W N(n) PA EoO Ag 119 1 12 0 4 2 0 138 86.23% 13.77% D_IMP 0 35 6 1 0 9 0 51 68.63% 31.37% GH 9 5 70 0 1 21 0 106 66.04% 33.96% L_IMP 0 5 1 20 1 2 0 29 68.97% 31.03% SB 0 1 0 3 45 1 0 50 90.00% 10.00% TF 0 0 7 0 0 171 0 178 96.07% 3.93% W 1 1 1 0 3 0 47 53 88.69% 11.32% N(n) 129 48 97 24 54 206 47 605 UA 92.25% 72.92% 72.16% 83.33% 83.33% 83.01% 100% EoC 7.75% 27.08% 27.84% 16.67% 16.67% 16.99% 0% OVERALL ACCURACY: 83.80% UA= User's accuracy PA= Producer's accuracy EoO= Errors of Omission EOC= Errors of Commission Section 2. EPA Methods and Accuracy Assessment The 19 quarter quads remaining after the OTIE/ CAST classification were classified by the U.S. Environmental Protection Agency using NAIP imagery and ancillary data. These remaining quarter quads were located on two separate sides of the study area: western Waukesha county and central Ozaukee county. The ancillary data sets consisted of the National Wetlands Inventory (NWI) data downloaded in sections from the Unites States Fish and Wildlife Service (USFWS) Wetland Mapper web mapping service (http://www.fws.gov/wetlands/Wetlands-Mapper.html) covering the study area and completed in 2005; LiDAR data for Waukesha, Jefferson and Ozaukee counties provided by the individual County Land Information Officers and used to derive Height Above Ground (HAG) and Intensity rasters; and Normalized Difference Vegetation Index (NDVI) and Variance rasters (standard deviation for each cell using a 3x3 moving window) created from the NAIP tiles. The Metadata for the NWI layer can be found at http://www.fws.gov/wetlands/Data/Metadata.html and the LiDAR metadata can be provided upon request. A supervised classification was performed using Genie Pro 2.4 machine-learning feature extraction software (www.observera.com) to map the following land cover classes: Water, Impervious Surface, Soil and Barren, Tree and Forest, Grass and Herbaceous. GeniePro software creates ancillary bands such as texture, band ratios, and edges to increase the dimensionality of the data for better classification results. Land cover classes were delineated in a five step procedure whereby each class was removed from further processing after delineation. Step 1: Training pixels were collected for WATER on one representative quarter quad for each of the two sides of the study area using band stacked imagery consisting of NAIP plus NDVI. The solution file was applied to the remaining imagery and obvious errors were hand edited on the result files. The tiles were mosaicked together to create a water raster, then the areas of NoData were used as a mask for the next step. Step 2: Training pixels were collected for VEGETATED areas on one representative quarter quad for each of the two sides using band stacked imagery consisting of NAIP, NDVI, and Variance with WATER masked out. The solution file was applied to the remaining imagery and obvious errors were hand edited on the result files. The tiles were mosaicked together to create a Vegetation/Non-vegetation raster. Two new rasters were created, Vegetation only and Non-vegetation only. These served as mask layers for Tree/Grass delineation as well as Soil/Impervious delineation. Step 3: Training pixels were collected for TREE and GRASS features on one representative quarter quad for each of the two sides using band stacked imagery consisting of NAIP, NDVI, HAG, and Intensity with WATER and NON-VEGETATED areas masked out. The solution file was applied to the remaining imagery and obvious errors were hand edited on the result files. The tiles were mosaicked together to create a TREE/GRASS raster. The TREE and GRASS features were then reclassified into separate raster layers for later processing. Step 4: Training pixels were collected for SOIL and IMPERVIOUS features on one representative quarter quad for each of the two sides using band stacked imagery consisting of NAIP and NDVI with WATER and VEGETATED areas masked out. The solution file was applied to the remaining imagery and obvious errors were hand edited on the result files. The tiles were mosaicked together to create a SOIL/IMPERVIOUS raster. The SOIL and IMPERVIOUS features were then reclassified into separate raster layers for later processing. Step 5: Common Land Unit (CLU) polygons created by the US Department of Agriculture were acquired through a publicly available GIS source (www.geocomm.com) and used as an overlay to determine agricultural fields. The CLU polygons serve as the boundaries for locating agricultural fields. Those polygons demarcating agricultural fields were selected and converted to an agriculture raster layer. Step 6: Each individual cover class raster was combined into a single raster image using the MOSAIC TO NEW RASTER TOOL, whereby the order of individual layers is as follows: WATER, IMPERVIOUS, TREE, AGRICULTURE, SOIL, GRASS. Both woody and emergent wetlands were included in the final classification, again using the NWI dataset. Trees and Forest within NWI wetland polygons were reclassified to woody wetlands. Grass and herbaceous within NWI wetland polygons were reclassified to emergent wetlands. Wetland boundaries were accepted as accurate and not photo interpreted as wetlands, but were photo interpreted to assess the accuracy of the underlying Grass/Herbaceous or Tree/Forest classes. The land cover mosaic was reclassified using a modified Anderson Level 1 classification similar to that used for the North American Land Cover Data 2001 (NLCD) with the following classes: 0 - Unclassified, 10 - Water, 20 - Impervious Surface, 30 - Soil/Barren, 40 - Trees/Forest, 70 - Grass/Herbaceous, 80 - Agriculture, 91 - Woody Wetlands, 92 - Emergent Wetlands. Accuracy Assessment - A completely random sample of 600 points (100 points for each class) was created for the two sides of the secondary (EPA) study area. A shapefile of sampling points was provided to two analysts who had not been involved with generating the land cover map. They performed photo interpretation of the six hundred validation points on the original NAIP imagery used to develop the land cover map, assigning a land cover class to each point. Ancillary data such as Google Satellite and Street Views, and Bing Aerial and Birdseye views, were used as needed to substantiate the interpretation. A fuzzy classification scheme was used as defined by Gopal and Woodcock (1994), whereby each point is given a value between 1 and 5 for a given land cover class: 1: Absolutely Wrong: classification value is unacceptable (Very Wrong); 2: Understandable but Wrong: classification value is not good. There is something about the site that makes the answer understandable, but there is clearly a better answer. Classification would pose a problem for users of the map. (Not Right); 3: Reasonable or Acceptable: Maybe not the best possible classification but it is acceptable; the classification does not pose a problem to users of the map. (Right); 4: Good Answer: Would be happy to find this classification given on the map (Very Right); 5: Absolutely Right: No doubt about the match. (Perfect)) In situations where a particular class does not result in at least 50 points falling within that class, additional stratified random points were added to reach the minimum 50 points. The following confusion matrix summarizes the MAX accuracy assessment data for the EPA study area: Confusion Matrix for 19 EPA-Produced Quarter Quadrangles Agricult GH Imperv Soil Tree Water N(n) UA EoC Agricult 169 11 0 1 0 0 181 93.37% 6.63% Grass 2 94 5 4 16 1 122 77.05% 22.95% Imperv 0 3 47 2 0 1 53 88.68% 11.32% Soil 18 3 2 29 0 1 53 54.72% 45.28% Tree 1 21 3 0 123 0 148 83.11% 16.89% Water 1 0 0 0 0 99 100 99.0% 1.0% N(n) 191 132 57 36 139 102 657 PA 88.48% 71.21% 82.46% 80.56% 88.49% 97.06% EoO 11.52% 28.79% 17.54% 19.44% 11.51% 2.94% OVERALL ACCURACY 85.39% Section 3. January 2017 Update Methods and Accuracy Assessment Two major classification errors were detected upon re-inspection of the Milwaukee land cover data: 1) road layers overlaying tree canopy and 2) houses misclassified as trees. Before error correction, a new accuracy assessment was conducted for the Milwaukee land cover data using the EPA methods described above, but with different points taken from the whole study area. This was done to assess the accuracy of the Milwaukee data as a whole rather than as two separate areas (EPA and OTIE/ CAST). This new accuracy assessment produced two confusion (error) matrices. The "MAX" view is more conservative and the "RIGHT" view is more liberal. The MAX interpretation is correct if the classified land cover matches the interpreter's highest maximum score, illustrating the interpreter finds this class to be the most appropriate for that location. The RIGHT interpretation is correct if the classified land cover matches any class the interpreter has given a value of 3 or greater, illustrating that the interpreter finds the classification to be acceptable but another class may be more appropriate. The following confusion matrices summarize the MAX and RIGHT accuracy assessment results, respectively. These matrices can be copied and pasted into Excel. Confusion Matrices for Entire Milwaukee Study Area (85 Quarter Quadrangles), Before EPA Edits MAX Agricult Grass Imperv Soil Tree Water Row Total User's Accuracy Agricult 43 5 0 0 2 0 50 0.860 Grass 27 101 6 1 22 2 159 0.635 Impervious 1 5 90 3 12 1 112 0.804 SoilBarren 1 5 3 34 2 5 50 0.680 TreeForest 1 37 10 0 153 2 203 0.754 Water 0 0 0 0 0 52 52 1 Row Total 73 153 109 38 191 62 626 Producer's Accuracy 0.589 0.660 0.826 0.895 0.801 0.839 Overall Accuracy 0.756 K_Hat 0.689 K Variance 4.91e-4 RIGHT Agricult Grass Imperv Soil Tree Water Row Total User's Accuracy Agricult 43 5 0 0 2 0 50 0.860 Grass 1 136 4 0 17 1 159 0.856 Impervious 1 4 95 3 8 1 112 0.848 SoilBarren 0 4 0 41 2 3 50 0.820 TreeForest 1 26 10 0 164 2 203 0.808 Water 0 0 0 0 0 52 52 1 Row Total 46 175 109 44 193 59 626 Producer's Accuracy 0.935 0.778 0.872 0.932 0.850 0.881 Overall Accuracy 0.848 K_Hat 0.805 K Variance 3.48e-4 Following the preliminary accuracy assessment, the land cover data were modified to correct misclassifications caused by road layers incorrectly overlaying vegetation. To address these misclassifications, all impervious pixels from the OTIE study area were extracted from the dataset. These pixels were separated into impervious and vegetation classes using a soil adjusted vegetation index (SAVI). The SAVI was calculated as follows: SAVI = (1.5 * (NIR - Red)) / (NIR + Red + 0.5). Next, the vegetation class was separated into tree and grass classes using a height above ground (HAG) threshold of 2 m. Then, the newly-classified tree pixels were overlaid on the original dataset. Next, the land cover data were modified to correct houses misclassified as trees. To do so, a buildings layer was created by applying HAG and SAVI data to the OTIE study area. Any pixels that had a HAG value greater than 10 m and a SAVI value less than zero were classified as impervious. All pixels that did not meet these criteria were classified as "no data." To address speckle issues in the buildings layer, shrink and expand operations were performed. To correct errors involving water and wetlands in the buildings layer, the original water and wetlands values were overlaid on the buildings layer. Then, this adjusted buildings layer was overlaid on the entire Milwaukee land cover dataset. Following the dataset update, a second accuracy assessment was performed using the methods described at the beginning of this section (Section 3). The following confusion matrices summarize the MAX and RIGHT accuracy results for the updated dataset. Confusion Matrices for Entire Milwaukee Study Area (85 Quarter Quadrangles), After EPA Edits MAX Agricult Grass Imper Soil Tree Water Row Total User's Accuracy Agricult 43 5 0 0 2 0 50 0.860 Grass 27 101 4 1 22 2 157 0.643 Impervious 1 5 92 3 10 2 113 0.814 Soil 1 5 3 34 2 4 49 0.694 Tree 1 37 10 0 155 2 205 0.756 Water 0 0 0 0 0 52 52 1 Row Total 73 153 109 38 191 62 626 Producer's Accuracy 0.589 0.660 0.844 0.895 0.812 0.839 Overall Accuracy 0.762 K_Hat 0.697 K Variance 4.82e-4 RIGHT Agricult Grass Imperv Soil Tree Water Row Total User's Accuracy Agricult 43 5 0 0 2 0 50 0.860 Grass 1 136 2 0 17 1 157 0.866 Impervious 1 4 97 3 6 2 113 0.858 Soil 0 4 0 41 2 2 49 0.837 Tree 1 26 10 0 166 2 205 0.810 Water 0 0 0 0 0 52 52 1 Row Total 46 175 109 44 193 59 626 Producer's Accuracy 0.935 0.777 0.890 0.932 0.860 0.881 Overall Accuracy 0.855 K_Hat 0.813 K Variance 3.36e-4 The MAX accuracy for the entire Milwaukee study area increased from 75.6% to 76.2% because of the updates. The RIGHT accuracy increased from 84.8% to 85.5%. For the Tree class, the Producer's MAX accuracy increased from 80.1% to 81.2% and the Right accuracy increased from 85.0% to 86.0%. The Tree User's MAX accuracy increased from 75.4% to 75.6% and the Right accuracy increased from 80.8% to 81.0%. For the Impervious class, the Producer's MAX accuracy increased from 82.6% to 84.4% and the Right accuracy increased from 87.2% to 89.0%. The Impervious User's MAX accuracy increased from 80.4% to 81.4% and the Right accuracy increased from 84.8% to 85.8%. Entity and Attribute Detail Citation: https://www.epa.gov/enviroatlas/enviroatlas-fact-sheets EnviroAtlas Coordinator U.S. Environmental Protection Agency, Office of Research and Development-Sustainable and Healthy Communities Research Program, EnviroAtlas Geospatial Data Owner (919) 541-3832 109 T.W. Alexander Drive Research Triangle Park NC 27709 EnviroAtlas@epa.gov https://www.epa.gov/enviroatlas distributor https://enviroatlas.epa.gov/download/MWI_MULC_tif.zip download Download The URL providing direct access to the downloadable dataset. download https://enviroatlas.epa.gov/arcgis/rest/services/Supplemental/Landcover_AllCommunities/ImageServer ESRI:ArcGIS API The endpoint of a web service to access the dataset (REST endpoint, WMS GetCapabilities URL, or a SOAP WSDL endpoint). webMapService https://www.epa.gov/enviroatlas/enviroatlas-data information Documentation URL for accessing related documents such as data dictionary, technical information about a dataset, developer documentation, etc. information https://www.epa.gov/enviroatlas information Home Page Alternative landing page used to redirect user to a contextual, Agency-hosted "homepage" for the Dataset or API information dataset Horizontal Positional Accuracy Report Data were collected using methods that are accurate to within 2-5 meters (EPA National Geospatial Data Policy [NGDP] Accuracy Tier 2). For more information, please see EPA's NGDP at https://www.epa.gov/geospatial/geospatial-policies-and-standards) Features represented have not been tested for completeness Features represented have not been tested for completeness Tests for integrity have not been performed. This lineage was generated by the FGDC CSDGM to 19139 transformation The original MULC raster was developed. See the Overview section for the complete processing workflow. 2013-11-05T00:00:00 Two major classification errors were corrected: 1) road layers overlaying tree canopy and 2) houses misclassified as trees. See the Overview section for details 2017-01-01T00:00:00 unknown 2029-02-13 This metadata was automatically generated from the FGDC Content Standard for Digital Geospatial Metadata standard (version FGDC-STD-001-1998) using the 2024-02-09T11:11:00 version of the FGDC CSDGM to ISO 19139 transform. EnviroAtlas Coordinator U.S. Environmental Protection Agency, Office of Research and Development-Sustainable and Healthy Communities Research Program, EnviroAtlas Geospatial Data Owner (919) 541-3832 109 T.W. Alexander Drive Research Triangle Park NC 27709 EnviroAtlas@epa.gov https://www.epa.gov/enviroatlas pointOfContact