{65A8193B-0652-4748-AE8E-FD328CCC448D} 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 -- Portland, OR -- Meter-Scale Urban Land Cover (MULC) Data (2012) 2014-10-28 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) originator The EnviroAtlas Portland, OR Meter-Scale Urban Land Cover (MULC) dataset includes data for the Portland metropolitan area plus the city of Vancouver, Washington and various smaller towns and rural areas in Oregon and Washington. The total area classified was approximately 2160 square kilometers. The land cover data were generated from 1-m, four-band (red, green, blue, and near-infrared) aerial photography acquired from the United States Department of Agriculture's National Agriculture Imagery Program. Imagery for Oregon was collected in 2012, and imagery for Washington was collected in 2011. In addition, ancillary datasets were derived for the classification from two LiDAR datasets collected in 2007 and one LiDAR dataset collected in 2010. Eight land cover classes were mapped: water, impervious surfaces, soil and barren land, trees and forest, grass and herbaceous non-woody vegetation, agriculture, and wetlands (both woody and emergent). An accuracy assessment using 600 completely random and 54 stratified random land cover reference points yielded an overall accuracy of 78.6%. Using a liberal interpretation with similar classes (e.g. soil / grass, soil / agriculture) the overall fuzzy accuracy is 91.4%. For more information on fuzzy accuracy assessment see the overview section. This dataset was produced by 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 land cover map is to provide a detailed snapshot of the urban environment and ecosystem for analysis of multiple human and ecosystem health metrics and services. This analysis is performed typically 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 Oregon Washington Portland, OR 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 -123.258539 -122.274581 45.271965 45.820044 2014-02-20 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). For more information about EnviroAtlas data, go to https://enviroatlas.epa.gov/EnviroAtlas/DataFactSheets. 1) Study Area: The Portland, OR land cover dataset includes data for the Portland Urban Statistical Area, as defined by the US Census Bureau, plus a 1-km buffer. This includes the entire Portland metropolitan area plus the city of Vancouver, Washington and various smaller towns and rural areas in Oregon and Washington. The total area classified was approximately 2160 square kilometers. 2) Input Datasets: The land cover classification was developed using 94 Digital Orthophoto Quarter Quadrangles ("quarter quads") acquired from the United States Department of Agriculture's National Agriculture Imagery Program (NAIP). The 94 quarter quads included 20 from the 2011 imagery dataset for Washington State and 74 from the 2012 imagery dataset for Oregon. Imagery data for the 20 Washington quarter quads were collected on August 17, 2011. Imagery data for the 74 Oregon quarter quads were collected on three dates during the summer of 2012: June 16th (33 quarter quads), July 6th (23 quarter quads), and July 7th (18 quarter quads). Each quarter quad included four bands (visible RGB plus near-infrared) with one-meter spatial resolution. In addition to the NAIP imagery, LiDAR point cloud data were incorporated into the classification for approximately 75% of the 2160 square kilometer area. These LiDAR point cloud data were derived from three LiDAR datasets, two from 2007 and one from 2010. The two LiDAR datasets from 2007 were collected on March 15th and April 25th. The former was collected by the Oregon Department of Geology and Mineral Industries. The latter was collected by the Oregon Department of Forestry. For both, the average point densities were approximately eight per square meter. The 2010 data were collected between November 11, 2009 and July 2, 2010 by the US Army Corps of Engineers. The Average point density was approximately 10 per square meter. For all three datasets, up to four returns were collected per pulse. The 2007 data were distributed with points already classified as "ground" and "un-classified"; in the 2010 dataset, the points were already as "ground," "water," and "un-classified." 3) Method: 3)a) NAIP Preparation: The first step in the classification process was to prepare the NAIP imagery for classification. This involved first removing the overlap between contiguous NAIP tiles. To do this, a python script was develop in ArcGIS that read in the original boundaries of the NAIP tiles and then created a new boundary for each NAIP tile such that 1) the new boundary was a subset of the original boundary, and 2) all the new tile boundaries could be arranged in a rectangular grid pattern without any stair-step pattern or overlap among contiguous tiles. For each of the 94 tiles, the new tile boundaries were then intersected with the study area boundary (the area defined by the Portland Urban Statistical Area plus a 1-km buffer). The original NAIP tiles were then clipped to the portion of the new boundary that intersected the study area. This was done to ensure that only pixels falling within the study area boundary were classified. Finally, for each of the 94 clipped tiles, the normalized difference vegetation index (NDVI) was calculated using the red and near-infrared bands. This served as a fifth band in the classification in addition to the NAIP red, green, blue, and near-infrared bands. 3)b) LiDAR Preparation: LiDAR tiles that intersected the study area were downloaded from NOAA's Digital Coast website (http://csc.noaa.gov/digitalcoast/). The individual tiles were unzipped and then re-projected into the UTM coordinate system (NAD_1983_UTM_Zone_16N) using "LAStools" stand-alone programs. The LiDAR point data were then imported into ArcGIS and converted from vector to raster data. For the 2007 datasets, a total of four bands were created from the raw LiDAR point cloud data. The first band was height-above-ground. A canopy surface model (CSM) was created by taking the maximum z-value in each 1m pixel. A bare-earth digital elevation model (DEM) was calculated by averaging the z-values of all the LiDAR points classified as "ground" in the dataset. NoData values in the CSM and DEM rasters were estimated using an inverse-distance weighted interpolation algorithm. The height-above-ground band was calculated by subtracting the canopy surface model from the bare-earth elevation model. The second LiDAR-derived band was the percentage of points with non-last returns (return value less than total number of returns for the pulse). Areas with high percentages of LiDAR points with non-last returns are highly correlated with trees and forest. The third LiDAR-derived band was percentage of points classified as "Ground." The fourth LiDAR-derived band was percentage of points tagged as "Un-Classified." For the 2010 datasets, an additional 3 bands were created from the raw LiDAR point cloud. The first was the average intensity value of all the points falling within a given pixel. The second was the percentage of points classified as "Water." The third was distance to a pixel with at least one point classified as "Water." For the first four LiDAR-derived bands, the 2010 and 2007 LiDAR datasets were mosaicked together with the 2010 data taking precedence over the 2007 data where they overlapped. There was no overlap between the 2007 datasets. The four bands derived from the 2007 and 2010 datasets, plus the three additional bands derived from the 2010 dataset, were used as ancillary data for the classification. 3)c) Classification: The classification was done using the randomForest classification and regression tree library in the stats program R. The classification was done on a tile by tile basis. First, for each clipped tile boundary, the LiDAR-derived ancillary data were extracted from each of the datasets that intersected the respective tile boundary. Next, for all the input bands (four NAIP bands, NAIP NDVI band, and any LiDAR-derived bands), the raw pixel values were extracted and copied to binary data files. Storing the pixel values in this format allowed the data to be imported into R more quickly. Using the software program GeniePro 2.4, training data was created for four land cover classes: water, impervious surfaces, tree/forest, and grass/herbaceous. The training data was exported from GeniePro in raster format. In ArcGIS, a table was created that contained for each training pixel the corresponding pixel values of all the input datasets. In ArcGIS, a python script was developed that created a randomForest model from the training data. This was done using the "pyper.py" module that interfaces with the R statistical program. Then, for each pixel in the clipped NAIP tile, the script read in the values of the input datasets from the binary data files, and then used the randomForest model to predict the corresponding land cover type. The land cover predictions were then written to a binary output file. Once the process was complete for all the pixels in the clipped NAIP tile, the script then created a new output raster containing all of the classified pixel values. Once the initial classification process was complete, a filtering algorithm was run on the initial classification raster to remove speckle. To do this, a script was created that grouped the initial classification raster into regions based on similar classification values. Any group of a size less than 5 was merged into the largest contiguous group and re-classified. Then, for any group between the size of 5 and 150, if the group was initially classified as "grass/herbaceous" and completely surrounded by tree/forest, it was reclassified as tree/forest. The third step in the classification process was manual on-screen editing the classified and filtered image. The hand-editing was done using Genie Pro software and it involved correcting any major errors in the classification. It also involved hand-digitizing areas for a new class: "soil/barren." In Portland, the amount of land that falls within this class was minimal (less than 1%). Common Land Unit (CLU) polygons created by the US Department of Agriculture were 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 manually selected and converted to an agriculture raster layer. At the completion of the classification process, all the individual classifications were mosaicked together. The final step in the classification process involved adding in the wetland classification. Wetland data for the study area was acquired from the National Wetland Inventory. Polygons corresponding to woody or emergent wetlands were extracted. Classified pixels falling within the boundaries of these polygons were extracted from the 6-class classification. Pixels initially classified as "grass and herbaceous" were reclassified as "emergent wetlands"; pixels initially classified as "trees and forest" were reclassified as "woody wetlands." 3)d) Accuracy Assessment: A sample of 600 completely random points was generated to assess the accuracy of the classified land cover map. The random points were provided to a second analyst who had not been involved with generating the land cover map. Analyst Two performed photointerpretation of the NAIP aerial photography used in the classification. Several hundred random points (approximately 100 per land cover class) were labeled using a fuzzy classification approach (Gopal and Woodcock, 1994). This permits the Analyst to assign a confidence value to the photointerpretation accuracy at a sample point. Confidence is expressed as an integer from 1 to 5: 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) After the initial assessment, additional samples were generated for rare classes (e.g., soil). Using stratified random sampling, sufficient sample points were generated such that all classes had a minimum of fifty reference samples (most classes had approximately 100 reference samples). These points were interpreted in a similar method and included in the final accuracy assessment. The accuracy assessment produces two confusion (error) matrices. One is more liberal ("RIGHT") and one is more conservative ("MAX") views of the interpretation. The MAX interpretation is correct if the classified land cover matches the interpreter's highest 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 the interpreter finds the classification to be acceptable but another class may be more appropriate. We typically present the MAX interpretation. The following error matrix summarizes the accuracy assessment data. 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: MAX RASTER_TXT Agricult Grass_Herb Impervious SoilBarren TreeForest Water N(n) Users_Accuracy Errors_of_Commission\n Agricult 47 9 0 2 1 0 59 79.66 20.34\n Grass_Herb 31 101 11 15 17 0 175 57.71 42.29\n Impervious 2 3 127 20 1 2 155 81.94 18.06\n SoilBarren 4 1 0 50 0 0 55 90.91 9.09\n TreeForest 3 14 2 0 140 0 159 88.05 11.95\n Water 0 0 1 1 0 49 51 96.08 3.92\n N(n) 87 128 141 88 159 51 654\n Producers_Accuracy 54.02 78.91 90.07 56.82 88.05 96.08\n Errors_of_Ommission 45.98 21.09 9.93 43.18 11.95 3.92\n OVERALL_ACCURACY 78.59\n RIGHT RASTER_TXT Agricult Grass_Herb Impervious SoilBarren TreeForest Water N(n) Users_Accuracy Errors_of_Commission\n Agricult 48 8 0 2 1 0 59 81.36 18.64\n Grass_Herb 3 149 6 7 10 0 175 85.14 14.86\n Impervious 2 0 142 8 1 2 155 91.61 8.39\n SoilBarren 0 0 0 55 0 0 55 100.0 0.0\n TreeForest 0 3 1 0 155 0 159 97.48 2.52\n Water 0 0 1 1 0 49 51 96.08 3.92\n N(n) 53 160 150 73 167 51 654\n Producers_Accuracy 90.57 93.13 94.67 75.34 92.81 96.08\n Errors_of_Ommission 9.43 6.87 5.33 24.66 7.19 3.92\n OVERALL_ACCURACY 91.44\n 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/POR_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 Acquired 2011 4-band NAIP imagery of Portland, OR (Washington portion) study area. Publisher: USDA-FSA-APFO Aerial Photography Field Office. Publication_Date: 20111214. Publication_Place: Salt Lake City, Utah. Link: http://www.fsa.usda.gov/FSA/apfoapp?area=home&subject=prog&topic=nai 2013-12-01T00:00:00 Acquired 2012 4-band NAIP imagery of Portland, OR (Oregon portion) study area. Publisher: USDA-FSA-APFO Aerial Photography Field Office. Publication_Date: 20120924. Publication_Place: Salt Lake City, Utah. Link: http://www.fsa.usda.gov/FSA/apfoapp?area=home&subject=prog&topic=nai 2013-12-01T00:00:00 Acquired 2010 US Army Corps of Engineers (USACE) Portland District Columbia River Lidar. Originator: USACE Portland District. Publication Date: 201203. Publication Place: Charleston, SC. Link: http://www.coast.noaa.gov/dataviewer 2013-12-01T00:00:00 Acquired 2007 Oregon Department of Geology and Mineral Industries (DoGAMI) LiDAR: Northwest Oregon and Portland Metro Area. Originator: DoGAMI. Publication Date: 2007. Publication Place: Charleston, SC. Link: http://www.coast.noaa.gov/dataviewer 2013-12-01T00:00:00 Acquired 2007 Oregon Department of Forestry (ODF) Northwest Oregon Lidar. Originator: ODF. Publication Date: 2007. Publication Place: Charleston, SC. Link: http://www.coast.noaa.gov/dataviewer 2013-12-01T00:00:00 Acquired National Wetlands Inventory - Wetlands. Publisher: U.S. Fish and Wildlife Service. Publication Place: Washington, D.C.. Link: http://www.fws.gov/wetlands/ 2013-12-01T00:00:00 Acquired Common Land Unit for Oregon, Washington. Publication Date: 20080225 Originator: USDA-FSA Aerial Photography Field Office. Publication Place: Salt Lake City, Utah. 2013-12-01T00:00:00 See overview section for detailed processing workflow. 2013-12-01T00:00:00 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