Monday, 15 January 2007: 2:00 PM
Influence of surface water on US regional climate forecasts
214B (Henry B. Gonzalez Convention Center)
Three long term North American (NA) climate simulations with the Experimental Climate Prediction Center (ECPC) Global to Regional Spectral Model (G-RSM) have now been developed for the period 1979-2004. These NA simulations include: a) control (CTL), b) climatological surface forcing (CLM), and c) precipitation replacement (PR). Snow and soil moisture fields from these three basic climate simulations were then used as initial conditions for longer-range predictions. Every month, 1982-2004, the G-RSM was started three times globally from the National Centers for Environmental Prediction-Department of Energy Reanalysis II (R-2) initial conditions, except over NA, where the three different surface conditions derived from the long term simulations were used instead of the R-2 initial conditions. The RSM simulation fields affected both the global and regional fields over this area. All forecasts are for duration of one month. Additional ensemble members and forecast lead-time for each of the basic experiments are being added as computer resources become available. For the first few weeks, forecasts of surface air temperature (SAT) started from the PR compared best with PR simulation, forecasts started from the CTL simulation compared best with the CTL simulation, etc. Also, in comparison to the North American Regional Reanalysis (NARR), the PR initialization provides the best temperature and precipitation forecasts in the central part of the US, especially during the spring and summer. This appears to be due to the influence of soil moisture and snow feedbacks.
Supplementary URL: