Abstract:
At the NASA Global Modeling and Assimilation Office (GMAO), near real-time seasonal forecasts are routinely produced with the GMAO CGCMv1, a fully coupled atmosphere-ocean-land general circulation model. In the tropics, skill in seasonal forecasting derives primarily from initialization and prediction of sea surface temperature (SST). During summer over mid-latitude land, SST forcing has limited impact on predictions of precipitation and surface temperature, but some predictability at sub-seasonal to seasonal time scales may result from slowly varying land boundary conditions. At the GMAO, we now use realistic soil moisture and ground temperature initial conditions for our seasonal forecasts. These are derived off-line using observed surface meteorological forcing data from the Global Land Data Assimilation System (GLDAS). In this paper, we describe our new initialization scheme and demonstrate its impact on the seasonal forecasts generated for May through September 2004.