The datasets on this page, representing velocity of climate change for North America, are based on Parameter Regression of Independent Slopes Model (PRISM) climate data from the historical normal period (see this page), and climate change projections of the Coupled Model Intercomparison Project phase 5 and 6 (CMIP5/CMIP6) database corresponding to the 5th and 6th IPCC Assessment Reports for future projections. The data has been prepared for AdaptWest, a project supported by the Wilburforce Foundation.

The velocity of climate change is an analytical concept that can be used to evaluate the exposure of organisms to climate change. As originally proposed, velocity is calculated by dividing the rate of climate change by the rate of spatial climate variability to obtain a speed at which species must migrate over the surface of the earth to maintain constant climate conditions. Here, we provide data from an improved algorithm that conforms to standard velocity calculations if climate equivalents are nearby. Otherwise, the algorithm extends the search for climate refugia globally. Below we provide velocity surfaces for a multivariate PCA implementation that searches for climate matches based on the first 2 axes derived from a PCA of 11 biologically relevant climate variables¹.

The improved algorithm also distinguishes forward and backward velocities, allowing useful inferences about conservation of species (present-to-future velocities) and management of sites (future-to-present velocities). For the forward calculation we ask: what is the rate at which an organism in the current landscape has to migrate to maintain constant climate conditions? Conversely, in the backward calculation we ask: given the projected future climate habitat of a grid cell, what is the minimum rate of migration for an organism from equivalent climate conditions to colonize this climate habitat? 

Citations for the CMIP6-based velocity data are:

Carroll, C. (2023). Velocity of climate change for North America based on CMIP6 GCMs [Data set]. Zenodo. https://doi.org/10.5281/zenodo.10631707

Carroll C, & Mahony CR. 2025. Sources of uncertainty in estimation of climate velocity and their implications for ecological and conservation applications. Conservation Science and Practice e13296. http://doi.org/10.1111/csp2.13296

For background information on the methodology and previous analyses see the two papers below. The R script used to produce the data on this page, along with other details concerning the data, can be found in the readme file provided here.

The gridded climate layers downloadable below are in Lambert Azimuthal Equal Area projection, at 1km resolution, and covering North America.

Velocity rasters based on CMIP6 GCM

CMIP6 ensemble data are based on mean projections from 8 CMIP6 models that were chosen to represent the very likely range of Earth's equilibrium climate sensitivity per the most recent IPCC assessment. Velocity rasters generated from these 8 individual GCMs are also provided below. Additionally, velocity rasters developed from the UKESM1-0-LL model (not used in the ensemble) are provided as a representative of high-sensitivity models characteristic of the long tail of uncertainty in the upper limit of climate sensitivity

Values represent climate velocity in km/yr from the histroical normal period (here 1961-1990) and the relevant future multi-decadal period (2041-2070 or 2071-2100). Areas of no-analog (backward velocity) and disappearing (forward velocity) climates are shown as NODATA in the velocity rasters. Values showing the proportion of runs in which a cell was characterized as no-analog or disappearing are given in additional rasters. Typically, the two related rasters (velocity and disappearing climates) would be viewed and analyzed together. Further information is contained in the readme file provided here.

The data is provided as GeoTiff formats that can be imported into ArcGIS or other GIS applications.

Please cite the CMIP6 data below as:
AdaptWest Project. 2023. Gridded CMIP6-based climatic velocity data for North America at 1km resolution. Available at adaptwest.databasin.org.

Velocity rasters based on CMIP5 GCM

CMIP5 ensemble data are velocities based on mean projections from 15 CMIP5 models (CanESM2, ACCESS1.0, IPSL-CM5A-MR, MIROC5, MPI-ESM-LR, CCSM4, HadGEM2-ES, CNRM-CM5, CSIRO Mk 3.6, GFDL-CM3, INM-CM4, MRI-CGCM3, MIROC-ESM, CESM1-CAM5, GISS-E2R) that were chosen to represent all major clusters of similar AOGCMs (Knutti et al 2013, Geophys Res Let 40: 1�6, doi:10.1002/grl.50256). Projections from 8 of these 15 GCMs were also used to generate velocities based on individual GCM projections.

Values represent climate velocity in km/yr from the histroical normal period (here 1981-2010) and the relevant future multi-decadal period (2041-2070 or 2071-2100). Areas of no-analog (backward velocity) and disappearing (forward velocity) climates are shown as NODATA in the velocity rasters. Values showing the proportion of runs in which a cell was characterized as no-analog or disappearing are given in additional rasters. Typically, the two related rasters (velocity and disappearing climates) would be viewed and analyzed together. Further information is contained in the readme file provided here.

The data is provided in several formats: 1) for ensemble-based velocites only, via a link to the DataBasin page from which an ArcGis layer file can be downloaded, and 2) for all velocities, via a link to a zipfile containing either ASCII (.asc) or TIFF (.tif) format files that can be imported into ArcGIS or other GIS applications. The DataBasin link also allows one to open the associated dataset in a map window on DataBasin.

CMIP5 velocity archives have been compressed with the 7-zip utilty. To extract the data, use the 7-zip software, which is freely available for Windows (link) as well as Mac and Linux systems (link). Use of 7-zip in place of the standard zip format allows better compression, lower storage costs, and shorter download times.

Please cite the CMIP5 data below as:
AdaptWest Project. 2015. Gridded climatic velocity data for North America at 1km resolution. Available at adaptwest.databasin.org.

For gridded velocity data for western North America based on CMIP3 projections, see this page .