Advancing probabilistic drought monitoring with the North American Land Data Assimilation System (NLDAS) through multisensor ensemble data assimilation

Background and Objectives
The North American Land Data Assimilation System (NLDAS) has a long successful history of producing surface meteorology and precipitation datasets used as forcing for land-surface models (LSMs) to produce soil moisture, snow cover, and runoff/streamflow products. These products have been used by both MAPP and non-MAPP researchers for applications including drought and streamflow monitoring, as well as initial conditions for drought forecast systems and for short-term weather forecasts. To maintain the state-of-the-art nature of this land data assimilation system, several NLDAS upgrades are proposed, including the addition of three community LSMs that include a prognostic water table to help characterize hydrological drought. The assimilation of recent terrestrial water storage and surface soil moisture products, as well as incorporating the effects of irrigation, should also better represent evolving drought conditions. Probabilistic historic drought analysis and probabilistic real-time drought monitoring capabilities enabled by the Land Information System (LIS) architecture used for NLDAS can provide a “probability of drought” rather than a single deterministic drought index.

Brief Summary of Work to be Completed
The proposed work will include the following elements: 1) Adding improvements and capabilities to the NLDAS data production and drought monitoring system. This effort will include adding the latest versions of the Noah-MP and the CLM LSMs, in addition to the NASA Catchment LSM added under previous MAPP funding, to take advantage of their groundwater sub-modules to obtain a more complete picture of drought conditions; 2) Assimilating GRACE terrestrial water storage and SMAP surface soil moisture products into the NLDAS system for better diagnosis of drought and improvement of initial land conditions. This task will also include the effects of irrigation using maps of irrigated area derived from MODIS; and 3) Performing probabilistic historical drought analysis as well as real-time monitoring to both advance our understanding of drought and provide a measure of drought uncertainty.

Relevance to Program Announcement Research Area
The proposed work is in response to MAPP Competition: Research to Advance Understanding, Monitoring, and Prediction of Drought and will be primarily relevant to focus area B: Advancing the Development of a National Drought Monitoring and Prediction System. The improvements to the NLDAS drought monitoring and data products will be made by assimilating multiple data sources into community models to objectively evaluate the impact of model and data upgrades on modeled soil moisture, snow, and groundwater products with the goal to improve drought analysis and monitoring. The improved land-surface states will also provide initial conditions to a complimentary drought prediction system also being proposed under this call. The probabilistic drought analysis supported by our ensemble data assimilation system is also relevant to the goals of the MAPP program, and will represent a significant advance in drought monitoring and assessment.