Development of the NCEP Unified Land Data Assimilation System (NULDAS)

As part of an initiative to develop a Unified Forecast System for the the U.S. modeling enterprise, NCEP’s regional (NLDAS) and global (GLDAS) land-surface data assimilation systems are being consolidated and improved to develop the NCEP Unified Land Data Assimilation System (NULDAS). This unified system will be a global 0.04-degree Land Data Assimilation System which runs multiple land surface models (e.g., Noah, Noah-MP, CLSM-F2.5, VIC, and CLM) with soil moisture and snowpack DA procedure in both uncoupled and coupled mode. The system consists primarily of three parts: (1) generation and validation of global 0.04-degree hourly surface meteorological forcing, (2) production of multiple land surface models run with DA processes (e.g., water fluxes, energy fluxes, carbon fluxes, and state variables), and (3) evaluation of NULDAS products and their application to NCEP coupled prediction and DA systems, drought monitoring and prediction system, and water resource management. The forcing will be optimally generated by blending and bias-correction of community-matured research-based and NCEP-operational products (e.g., gauge-based, satellite retrievals, and reanalysis data). The models and DA process will be run via the NASA Land Information System (LIS) infrastructure. The latest upgraded Noah and Noah-MP model, and realtime vegetation datasets (e.g., LAI, Greenness Vegetation Fraction) and recently developed soil datasets will be used in this system. The evaluation and validation of forcing and model products will be carefully performed by comparing operational NLDAS, GLDAS, in-situ observations, satellite retrievals, and reanalysis data from various operational centers at spatial scales ranging from watershed to globe and temporal scales from hourly (e.g., diurnal cycle) to interannual. After the comprehensive evaluation is completed, we will work with our LDAS collaborators to pursue their potential applications in many aspects. The NULDAS is leveraging National Water Model development (e.g., 4 km forcing) for the continental United States, and will directly support the NCEP Next Generation Global Prediction System (NGGPS) such as FV3GFS by providing optimal initiation conditions.

This presentation summarizes NCEP LDAS development background, its R2O transition and operational implementation, the NULDAS concept, design, and infrastructure, and future plans. This initiative will recall community efforts to support the NCEP NULDAS development via R2O collaboration and mature data and methodology sharing.