Climate Distance Mapper is an interactive web mapping application designed to facilitate informed seed sourcing decisions and to aid in directing regional seed collections. Implemented as a shiny web application (Chang et al. 2017), Climate Distance Mapper is hosted on the web at: https://usgs-werc-shinytools.shinyapps.io/Climate_Distance_Mapper/. The application is designed to guide restoration seed sourcing in the desert southwest by allowing users to interactively match seed sources with restoration sites climatic differences – in the form of multivariate climate distance values – between restoration sites and the surrounding landscape. Climatic distances are based on a combination of variables likely to influence patterns of local adaptation among plant populations, including: mean annual temperature, summer maximum temperature, winter minimum temperature, temperature seasonality, annual temperature range, mean annual precipitation, winter precipitation, summer precipitation, precipitation seasonality, long-term winter precipitation variability, and long-term summer precipitation variability. The climate variables are first transformed into principal components (PCA analysis), which standardizes the variables and accounts for collinearity while emphasizing the most important climate gradients. All climate data is obtained through ClimateNA (Wang et al. 2016), an application for dynamically downscaling PRISM climate data (Daly et al. 2008). The fundamental unit of measure in Climate Distance Mapper is the multivariate climate distance, which is defined as the multivariate Euclidean distance between climate-based principal components at input points and those at other grid cells throughout the chosen spatial extent. All distance calculations incorporate 5 principal components derived from an original set of 12 climate variables. The conversion to principal components standardizes and accounts for collinearity in climate variables, while ensuring that the most important climate gradients are given the most weight (i.e., because principal components are ordered in terms of the variability they express). Conceptually, multivariate Euclidean distance with principal components may be viewed as an approximation of the multivariate Mahalanobis distance calculated on the original climate variables. Mahalanobis distance is the distance between groups weighted by the within-group dispersion. Our procedure for calculating climate distance is thereby meant to emphasize natural gradients that distinguish climatic regimes across landscapes without giving any variable undue weight due to differences in units or scale. All multivariate climate distance values are relativized to the 95th percentile of the maximum possible climate distance in a given region (regions may be set dynamically by the user), such that values roughly correspond to a percentage of the total climate variability (using the 95th percentile of the maximum climate distance reduces the influence of outlier grid cells). This means that a climate distance of 0.2 is roughly analogous to 20% of the total climate variability in the selected region. Within Climate Distance Mapper, users can also constrain results to a specific level of climate similarity (e.g., areas with 90% similar climates). Climate Distance Mapper also supports projections into future climate – either by comparing the current climate at input points with the future climate across the landscape (forward projection, from current climate forward to future climate), or by comparing the future climate at input points with the current climate across the landscape (backward projection, from future climate back to current climate). Future climate is defined as the predicted 30-year average for the 2040-2070 period using an ensemble average of three models from the Coupled Model Intercomparison Project phase 5 (CMIP5) database corresponding to the 5th IPCC Assessment Report for future projections (IPCC 2014). We selected the RCP 4.5 (moderate emission) and RCP8.5 (high emission) scenarios for projections. The future climate models include CCSM4 (Community Climate System Model, version 4.0), GFDL-CM3 (Geophysical Fluid Dynamics Laboratory Climate Model, version 3), and HadGEM2-ES (Hadley Centre Global Environmental Model, version 2 (Earth System). All future climate data were generated using ClimateNA (Wang et al. 2016). References: Chang, W., Cheng, J., Allaire, JJ., Xie, Y., McPherson, J. 2017. shiny: Web Application Framework for R. R package version 1.0.0. https://CRAN.R-project.org/package=shiny. Daly, C., M. Halbleib, J. J. Smith, W. P. Gibson, M. K. Doggett, G. H. Taylor, J. Curtis, and P. A. Pasteris. 2008. Physiographically-sensitive mapping of temperature and precipitation across the conterminous United States. International Journal of Climatology 28:2031–2064. IPCC. 2014. Climate Change 2014: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press. Wang, T., A. Hamann, D. Spittlehouse, and C. Carroll. 2016. Locally downscaled and spatially customizable climate data for historical and future periods for North America. PLoS ONE 11: e0156720. https://doi.org/10.1371/journal.pone.0156720