Michael Ek, NOAA/NCEP

Abstract: The NCEP/EMC Land Data Assimilation System (LDAS) team has collaborated with its partners in developing two systems: the North American LDAS (NLDAS, www.emc.ncep.noaa.gov/mmb/nldas) and the Global LDAS (GLDAS). NLDAS provides initial land states that may be used for regional weather and climate forecasts, drought monitoring, and seasonal hydrological prediction via an uncoupled land modeling system over the continental U.S. (CONUS). NLDAS uses the NCEP North American Regional Reanalysis and observed gauge precipitation as surface forcing to drive four land surface models, including the NCEP Noah land surface model (LSM), at 0.125-deg resolution to produce a 29-year (1979-2007) retrospective and more than three-year (2008-present) near real-time hydrometeorological/climatological set of products. GLDAS provides initial land states to the NCEP Climate Forecast System version 2 (CFSv2) for global seasonal climate prediction and generates hydrometeorological/climatological reanalysis products. GLDAS is a semi-coupled land modeling system which uses hybrid precipitation (combination of gauge, satellite and model) and observed snow, along with CFSv2 model analysis for the other surface forcing to drive the NCEP Noah LSM, with land states provided daily to the operational CFSv2; GLDAS was part of the CFSv2 reanalysis (1979-present, cfs.ncep.noaa.gov/cfsr).

GLDAS operates under the NASA Land Information System (LIS) in executing the Noah LSM within CFSv2, and will be upgraded to use of a new version of LIS that has an Ensemble Kalman Filter (EnKF) data assimilation capability that will allow assimilation of satellite sources of land surface states, such as land surface skin temperature (e.g. NESDIS GOES products) and soil moisture (e.g., NESDIS SMOPS and NASA SMAP products). As a first step in assessing the impact of soil moisture data assimilation on weather forecasts, an EnKF soil moisture data assimilation algorithm has been implemented in the NCEP Global Forecast System (GFS) in an EnKF-GFS coupled approach. In another LIS effort, Noah and SAC-HT/SNOW17 LSMs are run at 4-km over CONUS as part of a high-resolution NLDAS; in this setting, to address early snowmelt biases due to immature snow evolution physics, high-resolution MODIS snow cover is assimilated to improve snow water equivalent values, and ultimately other land-states and hydrological outputs such as streamflow. Finally, model physics improvements and assimilation of surface data sets are closely related, e.g. to address GFS daytime low-level temperature biases due to model physics limitations, modifications are made to surface-layer formulations and surface characterization in microwave emissivity calculations, resulting in improved brightness temperatures at surface-sensitive channels and increased satellite data utilization.