Advancing the Timeliness of the Real-time N. American Land Data Assimilation System 2: Development and Evaluation

The North American Land Data Assimilation System (NLDAS) has a long successful history of producing A) surface meteorological forcing for and B) land surface datasets from land-surface models (LSMs) to provide soil moisture, snow cover, surface evaporation, runoff/streamflow products, among others. Since Phase 2 of NLDAS (NLDAS-2) was implemented in the NCEP product suite in August 2014, these products have been widely used for drought monitoring and water resource management purposes. The NLDAS website shows that in 2015, about 5400 distinct users downloaded over 71 million files with a total volume of 144 Tb.

However, this operational system has a 3.5-4 day time lag, which hampers its application for realtime operational tasks, such as the weekly US drought monitor and initializing land states in the NCEP coupled mesoscale North American Modeling (NAM) system. To close this time lag, the NCEP/EMC NLDAS team worked with its collaborators to integrate 1) NESDIS GOES downward shortwave radiation, 2) CPC daily gauge precipitation, 3) EMC hourly stage II radar-based national precipitation analyses, and 4) fields from both the NAM Data Assimilation System (NDAS) and the NAM mesoscale model forecast product suite to upgrade the current operational NLDAS system to an actual real-time NLDAS system with zero temporal lag. The upgraded NLSAS-2 suite will be implemented following the next NCEP operational NDAS and NAM mesoscale model upgrade event (anticipated for January 2017) to make the final modifications that achieve the next operational version of the NLDAS-2 suite. The resulting zero-lag real-time status of NLDAS-2 products will positively benefit the NLDAS Drought Monitor, which is updated daily in support of national operational drought monitoring and prediction tasks.  Real-time NLDAS-2 drought products will result in better timeliness for depictions of drought extent and severity over CONUS.  At the same time, the NLDAS-2 real-time state variables may provide more reasonable land-state initial conditions for NCEP regional weather/climate models (e.g., NAM) to enhance prediction skills for different timescales. 

This presentation summarizes the development, testing and evaluation of the upgraded fully real-time NLDAS-2 system including its design and infrastructure changes, product distribution, and assessment of water fluxes, energy fluxes and state variables. In particular, using the current operational NLDAS drought-monitor indicators as a benchmark, we evaluate the impact of real-time surface meteorological forcing data, including NDAS and NAM forecast products, on drought indicators.