Using the simple anomaly method (modifying a historical baseline with differences or ratios projected by General Circulation Models), scientists from the California Academy of Sciences downscaled monthly total precipitation from 16 different global circulation models (GCMs). The GCMs were described in the latest Intergovernmental Panel for Climate Change (IPCC 2007) and archived at the WCRP PCMDI ( http://www-pcmdi.llnl.gov/ipcc/about_ipcc.php). Monthly total precipitation was downscaled using the ratio between each 20 year-period in the future (e.g. 2040-2059) and the simulated historical average (1950-2000) from each GCM. Each layer was then interpolated to a spatial grain of 10x10 km2 using the regularized spline function of Spatial Analyst tool in ArcGIS 9.3, and then multiplied by the baseline historical climate (Wordclim 1.4) to create future projected precipitation: Pdw=((Pf/Pc )*Pobs )
Monthly precipitation was averaged seasonally over 20 year periods of GCM simulations from 2000 through 2099. Seasonal climate means were generated by averaging values for three months (ex. winter conditions correspond to the average climate for December, January, February). Two sets of future climate projections are available, corresponding to either the regional economic A2 or the global environmental and equitable B1 greenhouse gas emissions scenarios. Every data layer has a corresponding layer representing the standard deviation across GCMs, which gives an estimate of the variability of the climate across the various GCMs (16 for precipitation) used to calculate the average.
Related datasets and documentation can be found in the Databasin gallery at: http://app.databasin.org/app/pages/galleryPage.jsp?id=2a47360040364876b37a52657793faa6