Kathryn Thomas 20020101 vector digital data https://www.wildlife.ca.gov/Data/BIOS http://ftp.wildlife.ca.gov/BDB/GIS/BIOS/Public_Datasets/100_199/ds166.zip The Department of Defense and the other desert managers are developing and organizing scientific information needed to better manage the natural resources of the Mojave Desert. One product from this endeavor is the Central Mojave Vegetation Map (developed by US Dept of Interior, USGS Western Ecological Research Center and Southwest Biological Science Center) that displays vegetation and other land cover types in the eastern Mojave of California. Map labels represent alliances and groups of alliances as described by the U.S. National Vegetation Classification. The nominal minimum mapping unit is 5 hectares. Each map unit is labeled by a primary land cover type and a secondary type where applicable. In addition, the source of data for labeling each map unit is also identified in the attribute table for each map unit. Data were developed using field visits, 1:32,000 aerial photography, SPOT satellite imagery, and predictive modeling. These data were developed as part of the Department of Defense Legacy funded Mojave Desert Ecosystem Program. Decision tree models used in the development of this dataset are described in detail in the Central Mojave Vegetation Mapping Project: Final Report. To view the final report go to: http://www.nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=13890. 19940101 20000101 Observed Complete As needed -118.076384 -114.712605 37.258493 34.074091 None arid lands vegetation National Vegetation Classification ISO 19115 Topic Category biota environment None Mojave Desert California None While the polygons boundaries have been preserved in the final vegetation map so that the source and secondary label items can be maintained for each polygon, an ArcInfo 'dissolve' could be used to eliminate polygon boundaries between adjacent identical primary map labels. Alternately, the map may be viewed in ArcView with polygon boundaries eliminated for a presentation of contiguous vegetation types without the polygon boundaries.License: This work is licensed under Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/). Using the citation standards recommended for BIOS datasets (https://www.wildlife.ca.gov/Data/BIOS/Citing-BIOS) satisfies the attribution requirements of this license.Disclaimer: The State makes no claims, promises, or guarantees about the accuracy, completeness, reliability, or adequacy of these data and expressly disclaims liability for errors and omissions in these data. No warranty of any kind, implied, expressed, or statutory, including but not limited to the warranties of non-infringement of third party rights, title, merchantability, fitness for a particular purpose, and freedom from computer virus, is given with respect to these data. USGS Forest and Rangeland Ecosystem Science Center Thomas, Kathryn mailing and physical

Colorado Plateau Field Station Northern Arizona University PO Box 5614, Bldg 24

Flagstaff Arizona 86011-5614 928-556-7466 ext 235 Kathryn_A_Thomas@usgs.gov Mojave Desert Ecosystem Program, Coordinator: Clarence Everly Version 6.2 (Build 9200) ; Esri ArcGIS 10.6.1.9270 Kathryn Thomas 20020101 Database Please refer to process steps and the Central Mojave Vegetation Map Final Report. This data set is an ArcInfo polygon coverage that has been successfully processed using the ArcInfo build command. All polygons are closed and have a single label point. We assumed, based on literature review and personal observation, that floristic variation (at the alliance level) in the Mojave ecoregion is not strongly related to multispectral reflectance as recorded in satellite imagery because vegetation cover is sparse and the substrate dominates the reflectance response. Color air photos were acquired during several missions, and using two different aircraft camera platforms, due to problems with cloud cover. Photos varied in quality, as a function of format and mission. SPOT panchromatic satellite imagery with 10 m resolution was obtained from the California Department Fish Game (copyright CNES/SPOT Image Corp. 1994) for the mapping area. This imagery had been georeferenced and terrain corrected and was used as the base mapping layer for the project. This imagery conforms to the 1:24,000 national map accuracy standard. Thomas, Kathryn 20020101 field data None 19970101 unknown 20000101 unknown ground condition releve' plots Data from these plots were used to create a reference data set which was used to develop the decision tree model and validate map label assignments derived from photo interpretation. Mojave Desert Ecosystem Program 20010101 vector digital data http://www.mojavedata.gov/ mdep/geomorphic/glabout.html 24000 None 19970801 unknown present Observed glscgis Landforms, which were used to help delineate vegetation type boundaries. NASA High Altitude Airborne Sciences Program 19970101 aerial photography 32000 None 19970414 unknown 19970724 unknown Observed Air Photos The aerial photography was used to delineate map polygons based on vegetation and land-surface characteristics. Following field reconnaissance, we determined that polygons could be delineated based on tone, texture, and terrain features related to landform, soil or surface color, and sometimes plant size, tone or density. Thomas, Kathryn unpublished material vector digital data Details of this model are presented in the Mojave Desert Ecosystem Program Central Mojave Vegetation Mapping Project. 24000 None unknown publication date model Decision tree models (also know as classification or regression trees) were developed to predict the probability of occurrence of each alliance or group of alliances (alliance complex). Maps of predicted alliance/alliance complex occurrence were developed using the decision tree models and maps of environmental features. California Department Fish and Game 19940101 satellite imagery None 19940101 unknown Observed spot SPOT panchromatic satellite imagery with 10 m resolution was obtained from the California Department Fish and Game (copyright CNES/SPOT Image Corp. 1994) for the mapping area. This imagery had been georeferenced and terrain corrected and was used as the base map layer for the project. SPOT imagery was received in Albers projection and reprojected to UTM projection WGS84 datum. spot Photo Interpretation - We developed a classification of photo-interpreted preliminary labels with three attributes: vegetation type, landform, and vegetation cover. These preliminary labels served two purposes: a) they indicated what criteria were used by the photo-interpreter to delineate the polygon, and b) they were used to stratify the mapped area so that test models could be applied separately to non-overlapping areas of the landscape containing subsets of the vegetation types. We included a photo-interpreted landform label because although there was a landform map available, it was developed at a coarser scale (10 ha MMU). Rolls of photo transparencies were cut into frames, and placed in protective polypropylene sleeves. Polygons were drawn onto the sleeve and then visually transferred and digitized on-screen using the georectified SPOT image as a base map. Interpreters first delineated landform boundaries, and then added any additional boundaries related to vegetation physiognomy (large shrubs, trees) and cover (as indicated by tone and texture). We initially mapped a landform class that we called "Wash Systems" - a high density of small washes with the intervening areas of upland included in the polygons. During editing, we have tried to separate washes and upland while maintaining the specified MMU. Air Photos Second iteration predicted vegetation map - The reference dataset was updated with the additional observation points so that 3819 point observations of vegetation types and their georeferenced location (UTM northing and easting) was available for the second modeling effort. Twelve environmental variables were selected from all the variables examined for predictive modeling. These variables were obtained in digital format (with integer values and 30 m. resolution) to cover the study area. The value for each environmental variable at each of the field point observations was extracted (ArcInfo 'sample'). This resulted in a database containing the observed vegetation type, UTM coordinates at that site, the observed vegetation type, and the value for each of 12 environmental variables at that site. Decision tree models were developed predicting presence/absence each of the 20 alliances and were pruned and snipped. The resulting decision trees ranged in size from 12 to 41 terminal nodes. A C++ program was used to convert the decision tree rules into an AML (Arc Macro Language) script for use in ArcInfo. The AML that was generated consists of a series of "if-then" statements based on the threshold values of environmental variables, and the terminal node associated with those environmental criteria. Each AML was interactively edited so that the terminal node number was replaced with the proportion of observations in which the predicted alliance/alliance complex was present at that node. When combined with grids of the environmental data, the AML generates an output grid where each grid cell meets the environmental criteria of one of the decision tree terminal nodes and has an associated probability of presence based on the observed data used to construct the model. The AML's were used to generate 20 probability maps, one for each vegetation type modeled. The predicted probabilities for each vegetation type ranged from 0-100 initially and were reclassified (ArcInfo Grid 'reclass') to 5 categories: 1 = <10% probability, 2= 11-49%, 3=50-64%, 4=65-84%, and 5=85-100% probability. The 20 probability grids were combined into combined probability grids (ArcInfo Grid 'combine') that contained the 1-5 probability category for each alliance/alliance complex in each grid cell. The database tables for the combined probability grids were extracted in an Access database for further processing. model Final labeling of probability maps - Using these tables, one alliance or alliance complex was assigned to each grid cell by inspecting the probability categories for all 20 alliances/alliance complexes for each cell and selecting the best possible prediction(s) for that grid cell. This recoding was done by visually inspecting the probabilities in each cell using a set of recoding rules. The recoding scheme was chosen to select for any grid cell the alliance/alliance complex with the highest prediction probability for that location. If a single vegetation type was predicted with a probability category of 5 or 4 (greater than 65%), then the recoded assignment became that vegetation type. If more than one vegetation type had a probability category of 4 or 5 then the grid cell was assigned a combination of all the types with probability 4 or 5. If the highest predicted probability was 3, 2 or 1; the grid cell was assigned the vegetation type with the highest probability. The final probability maps were used as one of the data sources for final labeling of the map polygons. model First iteration predicted vegetation map - A reference dataset of 2008 point observations derived from the releve data and analysis of existing data was used to create an initial map of predicted alliance and alliance complex distributions. This map was field validated in the fall of 1999 using check maps of vegetation polygons overlain on satellite imagery. Based on the field validation and a review panel of Mojave vegetation experts, additional updating and editing of the map was planned based on collection of additional field observation points and reworking of the decision tree models. In the spring of 2000 over 3500 additional observations points were collected. model 19990101 Final Map Label Assignment - The map polygons draw during the photo-interpretation step were used in the final vegetation map. They were labeled by assigning a map label using the best available information for each polygon. Map labels were assigned in the following order of priority: 1. Field observed map label assignments, where the field observation was for an area at least the size of the minimum map unit (5 hectares). These field observations were from San Diego State University observations to support photointerpretation, photos of the relevé plots acquired as part of the project (1997, 1998 and 1999), and the reference dataset. 2. Expert assigned map label where a) the expert was known to be an area authority, b) the alliance type is known to be spatially constrained, and/or c) the assignment was based on the expert's use of ancillary data sources. 3. Photointerpreted labels for Pinyon, all land use categories (Development, Mining, Agriculture and Urban), Mesquite, and Limber Pine/Bristlecone Pine. 4. Wash System groupings based on identification of wash characteristic alliances that occur within particular elevation zones. This was based on elevation modeling of the field collected relevé data and on the descriptions of wash characteristic alliances presented in Appendix B of final report. 5. Landform based habitat labels based on the presence of a GLSCGIS delineated landform polygon (Dunes, Lava Beds and Cinder Cones, Playa). 6. Predictive model (probability maps) derived map label assignment. The map label indicates the predicted type, and the source label indicates the percentage of the polygon, which was predicted to be that vegetation type. releve' plots glscgis Air Photos model 20000101 Edited original metadata by adding ISO keywords. 20070723 California Department of Fish and Game Kristina White bios@dfg.ca.gov Vector GT-polygon composed of chains 15579 NAD 1983 California Teale Albers 34.0 40.5 -120.0 0.0 0.0 -4000000.0 coordinate pair 0.0001 0.0001 meter D North American 1983 GRS 1980 6378137.0 298.257222101 ds166 ArcInfo Polygon Attribute Table ESRI Shape_Length Length of feature in internal units. Esri Positive real numbers that are automatically generated. Shape Feature geometry. ESRI Coordinates defining the features. SHAPE_area Area of feature in internal units squared. ESRI Positive real numbers that are automatically generated. AREA_ Area of polygon in square meters. PERIMETER Perimeter of the polygon in meters. 658.8316 1053083 meters LABEL_1 The finest level of mapping representing a single alliance or groupings of similar alliances (alliance complex) or a land use type. See Table 2 in Final Report for complete explanation. Agricultural Agriculture generally irrigated Alkali Meadow/Sink Distichlis spicata Intermittently Flooded Herbaceous Alliance Big Sagebrush Artemisia tridentata Shrubland Alliance, Ephedra viridis-Artemisia tridentata Shrubland Alliance Blackbrush Coleogyne ramosisssima Shrubland Alliance Creosote Larrea tridentata Shrubland Alliance Larrea tridentata-Ambrosia dumosa Shrubland Alliance, Larrea tridentata-Encelia farinosa Shrubland Alliance (occasionaly), Ambrosia dumosa Dwarf-Shrubland Alliance, Encelia farinosa Shrubland Alliance (occasionally) Creosote-Brittlebush Larrea tridentata-Encelia farinosa Shrubland Alliance, Encelia farinosa Shrubland Alliance Desert Holly Artiplex hymenelytra Shrubland Alliance Dunes Barren; Herbaceous Dunes Sparse Vegetation Alliance; Panicum urvilleanum Sparsely Vegetated Herbaceous Alliance; Achnatherum speciosum Herbaceous Alliance: Pleuraphis rigida Herbaceous Alliance; Ambrosia dumosa Dwarf-Shrubland Alliance; Artiplex canescens, A. polycarpa or A. confertifolia Shrubland Alliance; Larrea tridentata Shrubland Alliance; Larrea tridentata-Ambrosia dumosa Shrubland Alliance; Prosopis glandulosa Shrubland Alliance; Abronia villosa Sparsely Vegetated Alliance Galleta Pleuraphis jamesii or Pleuraphis rigida Herbaceous Alliance High Elevation Wash System Sparsely Vegetated Wash, Unvegetated, Prunus fasciculatum Shrubland Alliance, Salazaria mexicana Shrubland Alliance, Hymenoclea salsola Shrubland Alliance, Salvia dorrii Dwarf-Shrubland Alliance, Viguiera reticulata Intermittently Flooded Shrubland Alliance (occasionally), Baccharis sergiloides Intermittently Flooded Shrubland Alliance, occasionally more typically upland types such as Artemisia tridentata Shrubland ALliance Hopsage Grayia spinosa Shrubland Alliance Iodine Bush-Bush Seepweed Allenrolfea occidentalis Shrubland Alliance, Suaeda moquinii Intermittently Flooded Shrubland Alliance Joshua Tree Yucca brevifolia Wooded Shrubland Alliance Juniper Juniperus californica or Juniperus osteosperma Wooded Shrubland Alliance Lava Beds and Cinder Cones Barren, Artiplex hymenelytra Shrubland Alliance, Encelia farinosa Shrubland Alliance, Larrea tridentata-Encelia farinosa Shrubland Alliance Limber Pine/Bristlecone Pine Pinus flexilis Woodland Alliance, Pinus longaeva Woodland Alliance Low Elevation Wash System Barren, Psorothamnus spinosus Intermittently Flooded Shrubland Alliance, Hymenoclea salsola Shrubland Alliance, Ephedra californica Intermittently Flooded Shrubland Alliance, Acacia greggii Shrubland Alliance, Chilopsis linearis Intermittently Flooded Shrubland Alliance, Encelia virginensis Shrubland Alliance, Ericameria nauseosa Shrubland Alliance, Eriogonum fasciculatum Shrubland Alliance, Hyptis emoryi Intermittently Flooded Shrubland Alliance, Lepidospartum squamatum Intermittently Flooded Shrubland Alliance, Prospois glandulosa Shrubland Alliance, Tamarix spp. Teporarily Flooded Shrubland Alliance, Viguiera parishii Shrubland Alliance and occasionally more typically upland types such as Artiplex hymenelytra Shrubland Alliance, Artiplex canescens/confertifolia/or polycarpa Shrubland Alliances Menodora Menodora spinescens Dwarf-Shrubland Alliance Mesquite Prosopis glandulosa Shrubland Alliance Mid Elevation Wash System Barren, Acacia greggii Shrubland Alliance, Prosopis glandulosa Shrubland Alliance, Chilopsis linearis Intermittently Flooded Shrubland Alliance, Ericameria paniculata Intermittently Flooded Shrubland Alliance, Viguiera parishii Shrubland Alliance, Baccharis sergiloides Intermittently Flooded Shrubland Alliance, Viguiera reticulata Intermittently Flooded Shrubland Alliance, Ephedra californica Intermittently Flooded Shrubland Alliance, Hymenoclea salsola Shrubland Alliance, Salazaria mexicana Shrubland Alliance, Encelia virginensis Shrubland Alliance, Ericameria nauseosa Shrubland Alliance, Eriogonum fasciculatum Shrubland Alliance, Lepidospartum squamatum Intermittently Flooded Shrubland Alliance, occasionally more typically upland types such as Artiplex hymenelytra Shrubland Alliance; Artiplex canescens, A. confertifolia or A. polycarpa Shrubland Alliances Mining Land use type Mojave Yucca Yucca schidigera Shrubland Alliance Nevada Joint-Fir Ephedra nevadensis Shrubland Alliance Pinyon Pinus monophylla Wooded Shrubland Alliance, Pinus monophylla-(Juniperus osteosperma) Woodland Alliance Playa Barren around edges may find Artiplex polycarpa, Artiplex confertifolia, or Artiplex canescens Shrubland Alliances; Allenrolfea occidentalis Shrubland Alliance, Suaeda moquinii Intermittently Flooded Shrubland Alliance, Pluchea sericea Seasonally Flooded Shrubland Alliance, Prosopis glandulosa Shrubland Alliance, Sporobolus airoides Intermittently Flooded Herbaceous Alliance Rural Development Buildings, structures or surface development other than urban Saltbush Artiplex canescens Shrubland Alliance, Artiplex polycarpa Shrubland Alliance, Artiplex confertifolia Shrubland Alliance (when around playas) Shadscale Artiplex confertifolia Shrubland Alliance Sparse Vegetation Less than 2% perennial vegetation Urban Land use type White Burrobush Ambrosia dumosa Dwarf-Shrubland Alliance LABEL_2 The secondary label for any alliance or alliance complex that may have been assigned to the map unit. See Table 2 in Final Report for complete explanation. Big Sagebrush Artemisia tridentata Shrubland Alliance, Ephedra viridis-Artemisia tridentata Shrubland Alliance Blackbrush Coleogyne ramosisssima Shrubland Alliance Creosote Larrea tridentata Shrubland Alliance Larrea tridentata-Ambrosia dumosa Shrubland Alliance, Larrea tridentata-Encelia farinosa Shrubland Alliance (occasionaly), Ambrosia dumosa Dwarf-Shrubland Alliance, Encelia farinosa Shrubland Alliance (occasionally) Creosote-Brittlebush Larrea tridentata-Encelia farinosa Shrubland Alliance, Encelia farinosa Shrubland Alliance Desert Holly Artiplex hymenelytra Shrubland Alliance Disturbed Galleta Pleuraphis jamesii or Pleuraphis rigida Herbaceous Alliance Hopsage Grayia spinosa Shrubland Alliance Iodine Bush-Bush Seepweed Allenrolfea occidentalis Shrubland Alliance, Suaeda moquinii Intermittently Flooded Shrubland Alliance Joshua Tree Yucca brevifolia Wooded Shrubland Alliance Juniper Juniperus californica or Juniperus osteosperma Wooded Shrubland Alliance Menodora Menodora spinescens Dwarf-Shrubland Alliance Mesquite Prosopis glandulosa Shrubland Alliance Mining Land use type SYSTEM The coarser level of vegetation mapping representing groupings of alliances. See Table 2 in Final Report for complete explanation. Barren Sparse Vegetation Creosote Bush Mixed Scrub White Burrobush, Creosote, Creosote-Brittlebush Desert Grassland and Shrub Steppe Galleta Desert Sink Iodine Bush-Bush Seepweed, Alkali Meadow/Sink, Playa Desert Wash System High Elevation Wash System, Mid Elevation Wash System, Low Elevation Wash System Interior Dunes Dunes Land Use Agriculture, Rural Development, Mining, Urban Lava Beds Lava Beds and Cinder Cones Limber Pine/Bristlecone Pine Woodland Limber Pine/Bristlecone Pine Mesquite Bosque Mesquite Mid Elevation Mixed Desert Scrub Blackbrush, Nevada Joint-Fir, Hopsage, Menodora, Joshua Tree, Mojave Yucca Pinyon Juniper Woodland Big Sagebrush, Juniper, Pinyon Saltbush Scrub Saltbush, Shadscale, Desert Holly SOURCE_1 The code for the source of information used to assign the primary vegetation type (i.e., LABEL_1). EXPERT1 Expert review workshop EXPERT2 MDEP mapping team editors FIELD1 San Diego State University photointerpretation team field observations FIELD2 Relevé data, retrospective data, validation and editing observations LSU Landforms from GLGSGIS MOD1 First iteration predictive modeling MOD2-100 Predictive modeling, all of polygon predicted to be primary label MOD2-25 Predictive modeling, 25-34% of polygon is predicted to be primary label MOD2-35 Predictive modeling, 35-44% of the polygon is predicted to be primary label MOD2-45 Predictive modeling, 45-54% of polygon is predicted to be primary label MOD2-55 Predictive modeling, 55-64% of polygon is predicted to be primary label MOD2-65 Predictive modeling, 65-74% of polygon is predicted to be primary label MOD2-75 Predictive modeling, 75-84% of polygon is predicted to be primary type MOD2-85 Predictive modeling, 85-99% of polygon is predicted to be primary type Photo-CPSU Photointerpretation conducted at Colorado Plateau Field Station Photo-SDSU Photointerpretation conducted at San Diego State University SOURCE_2 The code for the source of information used to assign the additional vegetation type (i.e., LABEL_2), if present. EXPERT2 MDEP mapping team editors FIELD1 San Diego State University photointerpretation team field observations FIELD2 Relevé data, retrospective data, validation and editing observations MOD1 First iteration predictive modeling MOD2 Second iteration predictive modeling Photo-SDSU Photointerpretation conducted at San Diego State University MOJVEG_NUM Unique identification number assigned to record by data developer. LAND_COVER Primary vegetation Refer to label attributes. MOJVEG_ID Unique identification number assigned to record by data developer. Esri Coordinates defining the features. SHAPE Feature geometry. ESRI Coordinates defining the features. Although these data have been processed successfully on a computer system at the USGS-FRESC-Colorado Plateau Field Station, no warranty expressed or implied is made regarding the accuracy or utility of these data on any other system or for general or scientific purposes, nor shall the act of distribution constitute any such warranty. This disclaimer applies both to individual use of these data and aggregate use with other data. It is strongly recommended that these data are directly acquired from a U.S. Geological Survey server, and not indirectly through other sources which may have changed these data in some way. It is also strongly recommended that careful attention be paid to the contents of the metadata file associated with these data. The U.S. Geological Survey and USGS-FRESC-Colorado Plateau Field Station shall not be held liable for improper or incorrect use of these data described and/or contained herein. ARCE ArcInfo interchange file format The interchange file contains all coverage information and appropriate INFO file information in a fixed-length, ASCII format. ArcInfo EXPORT compression applied to this coverage. Use ArcInfo IMPORT command to convert the interchange file. CD-ROM 20210420 20011012 U.S. Geological Survey Forest and Rangeland Ecosystem Science Center Metadata Coordinator mailing and physical

3200 SW Jefferson Way

Corvallis OR 97331 541-750-7343 541-758-7761 george_lienkaemper@usgs.gov FGDC Content Standard for Digital Geospatial Metadata FGDC-STD-001-1998 local time