Short-Term Retrospective Land Data Assimilation Schemes

Subsurface moisture and temperature and snow/ice stores exhibit persistence on various time scales which has important implications for the extended prediction of climatic and hydrologic extremes. Hence, to improve their specification of the land surface, many numerical weather prediction (NWP) centers have incorporated complex land surface schemes in their forecast models. However, because land storages are integrated states, errors in NWP forcing accumulates in these stores, which leads to incorrect surface water and energy partitioning. This has motivated the development of Land Data Assimilation Schemes (LDAS), which are uncoupled land surface schemes that are forced primarily by observations, and are therefore less affected by NWP forcing biases. The implementation of an LDAS also provides the opportunity to correct the model's trajectory using remotely-sensed observations of soil temperature, soil moisture, and snow in a 4-Dimensional Data Assimilation (4DDA) framework. The inclusion of 4DDA in an LDAS will greatly increase its predictive capacity, as well as provide high-quality land surface assimilated data.

The Land Data Assimilation Scheme (LDAS) project is a multi-institutional research effort centered on the development of a data assimilation scheme suitable for near-real time and retrospective modeling. Through the use of land surface models (LSMs) as well as terrestrial and space-based observations, this data assimilation scheme will reduce errors in surface fluxes and storage quantities that are often present in LSM simulations. Most numerical weather prediction models include, and depend upon, an LSM for realistic forecasts, so application of this scheme will directly improve forecast accuracy. The LDAS currently operates at a 1/8th-degree resolution over the continental United States and makes use of NCEP-EDAS forcing data, NESDIS/GOES radiation data and Stage IV precipitation data. An LDAS web site--which features a real-time image generator and project information--has been created and is located at http://ldas.gsfc.nasa.gov.

To address a variety of land-surface research questions, LDAS is being implemented in real-time, short-term retrospective (1996-current),and long-term retrospective (50 year) modes. The short-term retrospective LDAS simulations use the same 1/8th degree LDAS grid, parameters, and atmospheric forcing approach as is used in the real-time LDAS simulations. This consistency between the short-term and real-time LDAS simulations facilitates: 1) land surface storage spin-up and drift analysis; 2) soil moisture, temperature, and snow initialization studies; 3) evaluation of new land surface theory and forcing; and 4) development and validation of land surface data assimilation techniques. An overview of the LDAS short-term retrospective forcing data, simulations, and results will be presented.