University of California Davis 20170210 remote-sensing image Thorne, J.H., R.M. Boynton, A.J. Holguin, J.A.E. Stewart, & J. Bjorkman. (2016) A climate change vulnerability assessment of California’s terrestrial vegetation. California Department of Fish and Wildlife (CDFW), Sacramento, CA. https://www.wildlife.ca.gov/Data/BIOS Climate exposure is the level of climate change expected in the areas where each vegetation macrogroup is dominating. The data depicts the results of the “vegetation climate exposure analysis”, conducted on 29 macrogroups across the state. The vegetation climate exposure analysis is calculated using the mapped extent of each macrogroup. Every grid cell of each macrogroup was ranked according to how frequently the climate in it occurs across all the grid cells occupied by that macrogroup in the state. This was done for the current time, which was then used to create a classification of common and rare climate conditions of each macrogroup. Once each type’s “climate envelope” was defined, we then assessed how the climate in every grid cell changed under various future climate projections. This allowed a measure of potential vegetation stress, or climate exposure. The area extent of each macrogroup that will be lost from the most commonly occurring climate conditions (less than or equal to 80%) and the area that will fall into current marginal, or stressed, climate conditions (greater than 95%) or outside the current climate conditions known in the state (non-analog) was calculated. This approach is particularly useful for resource managers, who often are constrained to work in specified areas, and need estimates of what areas within their jurisdiction are likely to be highly stressed, and what areas are likely to be less stressed, in effect climate refuge areas. Macrogroup 106 (Temperate Pacific Intertidal Shore) was excluded from the exposure analysis due to its limited distribution within the study areas. The small sample size made it difficult to accurately identify its climate space. Also excluded from this assessment are non-vegetated types such as snow, open water, and ice; and non-natural landcover types mapped as vineyards, tilled earth, orchards and Urban. The exposure raster data files are named with the GCM (miroc or cnrm), emission scenario (rcp 4.5 or rcp 8.5), and the last two digits of the year range of the time period (e.g. 2070-2099 = 7099) that the data represents. Further information can be found in the project report: [https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=116208] The Climate Change Vulnerability Assessment of California's Terrestrial Vegetation assessed climate exposure of vegetation macrogroups based on two global climate models (GCMs) and two emission scenarios. The GCMs, CNRM CM5 and Miroc ESM, and emission scenarios used, RCP4.5 and RCP 8.5, represent a range of warming statewide from 1.99 to 4.56°C and between a 24.8% decrease in precipitation and a 22.9% increase, respectively. A 2015 map of the state’s natural vegetation compiled from multiple sources was classified to the National Vegetation Classification Standard’s mid-level classification, called “Macrogroup”. Thirty one natural vegetation macrogroups are identified in the map, covering 99.87% of the state’s natural terrestrial vegetation, and occupying 353,271 sq km. The use of a 2015 map portraying the current extent of vegetation permitted an assessment of climate exposure based on known locations, rather than on potential locations. This data depicts the level of exposure to changing climate that each of 29 vegetation types examined is projected to experience at the locations it currently occupies. Further information can be found in the project report: https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=116208 Analysis Approach: For detailed information on the analysis, see the project report [https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=116208&inline]. The vegetation climate exposure analysis takes advantage of the 2015 vegetation map compiled for California, which is described below. The vegetation climate exposure model is implemented in the R programming language, and takes the vegetation and climate raster files as the primary input data. The climate space (climate envelope) occupied by each distinct macrogroup from the current time period was identified using a principle components analysis of 9 hydro-climatic variables and applying a kernel density estimator on a 2-d surface composed of the first two principal components of the climate conditions. The result is a smoothed continuous point density surface, showing the prevalence of each vegetation type across the range of sampled climatic conditions. This surface was partitioned by fitting contour lines so that they enclose a proportion of the original points from the current time period. Cells further away from the dense central core are considered to be more marginal in the vegetation type’s distribution. With the current climate space defined for each vegetation type, each pixel from every time period was then classified into an exposure category based on where it falls relative to the density contours of its historic, or baseline distribution. In the initial time period, 1981-2010, the pixels by definition follow a uniform distribution across the classes. However, in the future time periods climate conditions shift, resulting in the vegetation becoming exposed to different conditions. Over time, they tend to experience increasingly marginal conditions relative to their observed distribution in the recent past. The raster cells in these marginal conditions are considered to be highly exposed to stress and type conversion. The most highly exposed cells fall completely outside the range of recent historic conditions for its type. There is also the possibility that a location is exposed to a combination of climatic conditions which are outside the range of the sampled historic conditions for the whole state. In this case, the conditions are considered “non-analog”, and no explicit exposure ranking is made, although studies have shown these areas to be equivalent to highly exposed (>99%). To visualize the values on the landscape, the exposure scores are projected back into spatial raster files, which can be portrayed either in the original 5% intervals, or in fewer aggregated classes on a map. We term all grid cells within 80% of the current climate distribution (gridcell values 0-80) to be in climatically suitable areas, and all grid cells in the >95% classes, including non-analog (gridcell value = -100), to be in the climatically stressed or marginal classes. Vegetation Data Inputs: The California State Department of Forestry and Fire Protection (CalFire) Fire and Resource Assessment Program (FRAP) has produced a statewide vegetation map, which is a compilation of the most recent and most accurate maps produced for various parts of the state. This most recent edition uses maps produced between 1990 and 2014, and is portrayed as a 30m grid. The vegetation in the map was crosswalked to the macrogroup classification that is part of the national vegetation classification system ("SWAP Macrogroups with Target Ranks - 2015" [ds1933], https://www.wildlife.ca.gov/Data/BIOS). The map was resampled to a 270m grid before analysis, to reduce computing time, and to align the patterns of vegetation distribution with climate data used. The sampling selected the most prevalent vegetation type found in the 81 30m grids within each 270m pixel. Climate Data Inputs: This analysis used projections of future climate using two global climate models (GCMs) that respectively are hotter and drier, and warmer and wetter than current conditions. For each GCM we used two emission scenarios that represent lower and higher levels of greenhouse gas concentration. All analyses were conducted on projections for the end century (2070-2099), which allow the furthest assessment of future trends. To select the futures used in this analysis, we selected two that are relatively much drier or wetter than most of the models, in order to both capture a range of future conditions, and also to minimize the areas within the state that seem to be trending in opposite directions from the overall direction of a given model. The two GCMs selected are MIROC ESM and CNRM CM. The emission scenarios were selected to represent a more hopeful level of climate change (the lower emissions RCP 4.5) and emissions levels that are closer to the current trend in emissions (the RCP 8.5). We ran the downscaled CNRM CM5 and MIROC ESM climate models under the RCP 4.5 and 8.5 emission scenarios through the hydroclimatic model called the Basin Characterization Model (BCM) to obtain a series of landscape hydrology values that could be what vegetation is more directly responding to, including potential evapotranspiration (PET), actual evapotranspiration (AET), climatic water deficit (CWD), snowpack (PCK) on April 1st, runoff (RUN) and recharge. In sum, 13 climate and hydrological variables were developed for every grid cell. Unknown -124.545978 -113.127363 42.595329 32.303733 None vegetation chaparral CWHR Urban Valley Oak Hardwoods California Range Lands Land Cover Shrub climate change macrogroup forest redwood ISO 19115 Topic Categories biota None California The State of California Department of Fish and Wildlife make no representation or warranties regarding the accuracy of the data or maps. The user will not seek to hold the State or the Department liable under any circumstances for any damages with respect to any claim by the user or any third party on account of or arising from the use of data or maps. None University of California, Davis, Information Center for the Environment (ICE) Dr. James H. Thorne Research Ecologist unknown

Department of Environmental Science and Policy

University of California Davis CA 95616-8628 530-752-4389 jhthorne@ucdavis.edu University of California, Davis, Information Center for the Environment (ICE) Dr. James H. Thorne Research Ecologist unknown

Department of Environmental Science and Policy

University of California Davis CA 95616-8628 530-752-4389 jhthorne@ucdavis.edu ice.ucdavis.edu/ http://www.dfg.ca.gov/biogeodata/vegcamp/ http://www.fs.fed.us/r5/rsl/projects/mapping/ http://www.fs.fed.us/pnw/fia/ http://www.frap.cdf.ca.gov/ Unclassified Microsoft Windows 7 Version 6.1 (Build 7601) Service Pack 1; Esri ArcGIS 10.4.0.5524 Raster Pixel 4165 3486 NAD 1983 California Teale Albers 34.0 40.5 -120.0 0.0 0.0 -4000000.0 coordinate pair 0.000000003754707655900803 0.000000003754707655900803 meter D North American 1983 GRS 1980 6378137.0 298.257222101 value vegetation exposure value: -100 = non-analog; 0-80 = remaining within current climate envelope; >95 = highly stressed OBJECTID Internal feature number. Esri Sequential unique whole numbers that are automatically generated. Value Count 20170303 University of California, Davis, Information Center for the Environment (ICE) Dr. James H. Thorne Research Ecologist unknown

Department of Environmental Science and Policy

University of California Davis CA 95616-8628 530-752-4389 jhthorne@ucdavis.edu FGDC Content Standard for Digital Geospatial Metadata FGDC-STD-001-1998 local time