Statistical Downscaling
Real time seasonal forecast at regional scale is produced by Statistical Downscaling. Global precipitation and 2m temperature at a resolution of ~1.8 degree (FSU/COAPS Global climate model) and 2.5 degree (NCEP/CFS) is downscaled to a fine spatial scale of ~20km for the southeast US region. The NCEP/CFS global model precipitation and temperature used for downscaling is obtained from NCEP/CFS realtime seasonal forecasts.
The primary techniques involved in downscaling are Cyclostationary Emperical Orthogonal Function (CSEOF), multiple regression, and stochastic time series generation. Detailed description on this downscaling model is found in Lim et al. (2007) and Lim et al. (2009)
The views and opinions expressed in this site are strictly those of the site author and in no way represent those of COAPS, the Florida State University, the Board of Trustees, nor the State of Florida.



