Downscaled Climate Projections

Climate Models and De-biasing

CMIP3 Climate Models

The large-scale climate predictions are taken from the CMIP3 climate models, a critical source of data to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC AR4). There are 13 models with enough daily data for us to downscale:

Originating Group(s) Country CMIP3 I.D.
Canadian Centre for Climate Modelling & Analysis Canada CGCM3.1(T47)
Canadian Centre for Climate Modelling & Analysis Canada CGCM3.1(T63)
Meteo-France / Centre National de Recherches Meteorologiques France CNRM-CM3
CSIRO Atmospheric Research Australia CSIRO-Mk3.0
CSIRO Atmospheric Research Australia CSIRO-Mk3.5
US Dept. of Commerce / NOAA / Geophysical Fluid Dynamics Laboratory USA GFDL-CM2.0
NASA / Goddard Institute for Space Studies USA GISS-AOM
NASA / Goddard Institute for Space Studies USA GISS-ER
LASG / Institute of Atmospheric Physics China FGOALS-g1.0
Center for Climate System Research (The University of Tokyo), National Institute for Environmental Studies, and Frontier Research Center for Global Change (JAMSTEC) Japan MIROC3.2(hires)
Meteorological Institute of the University of Bonn, Meteorological Research Institute of KMA, and Model and Data group Germany / Korea ECHO-G
Max Planck Institute for Meteorology Germany ECHAM5/MPI-OM
Meteorological Research Institute Japan MRI-CGCM2.3.2

Daily data from the CMIP3 models is only available for the following time periods: 1961-2000, 2046-2065 and 2081-2100. Therefore our downscaling is unfortunately also constrained to lie within these time periods.

De-biasing the Climate Models

The CMIP3 climate models are de-biased using the "CDF remapping" of Wood et al 2002 with a very minor modification to enhance the robustness at the tails of the distribution. They are "de-biased" to look like the NCEP Reanalysis, because our statistical models were "trained" using the large-scale from the NCEP Reanalysis.