The North American Land Data Assimilation System: Application to Drought over CONUS

The NCEP Environmental Modeling Center and its NOAA Climate Program Office partners developed the North American Land Data Assimilation System (NLDAS) to monitor and predict land surface states, and surface water and energy budgets. NLDAS land-surface monitoring consists of a 29-year (1979-2007) retrospective and a companion near real-time extension, with hourly water and energy fluxes and state variables (e.g. soil moisture, snowpack, runoff, evaporation) at 1/8th degree resolution over the continental US from four land surface models (NCEP/Noah, NASA/Mosaic, OHD/SAC, and Princeton/VIC). Land model forcing is from the NCEP 30-year retrospective and real-time North American Regional Reanalysis System (NARR), except precipitation which uses daily gauge-based precipitation disaggregated to hourly using radar and satellite data. The NLDAS seasonal hydrological prediction system uses three different sources for generating down-scaled ensemble seasonal hydrological forecasts of surface forcing to drive the VIC land model in an uncoupled mode to provide one to six month ensemble seasonal hydrological predictions (of e.g. streamflow, etc). NLDAS is used to support the NCEP/Climate Prediction Center and the National Integrated Drought Information System in drought and flood monitoring and seasonal hydrological forecasting.