Integrated metrics and benchmarking for the North American Land Data Assimilation System (NLDAS) (Invited Presentation)

The North American Land Data Assimilation System (NLDAS) produces hourly land-surface meteorology and surface states, including precipitation, soil moisture and temperature, snow cover/amount, evapotranspiration, runoff, and streamflow. The NLDAS domain extends over North America from 25-53° North on a 1/8th-degree grid. NLDAS is a collaborative project between NCEP/EMC, NASA/GSFC, Princeton University, the University of Washington, NWS/OHD, and NCEP/CPC. The current Phase 2 of NLDAS employs four unique land-surface models (LSMs), each driven separately by the surface forcing. The four LSMs are: EMC's Noah version 2.8, NASA/GSFC's Mosaic, Princeton's VIC version 4.0.3, and OHD's SAC/SNOW-17. For the next phase of NLDAS, GSFC's Catchment LSM (the land model within NASA's GEOS-5 GCM) has replaced the Mosaic LSM – and the Noah, VIC, and SAC-HTET/SNOW-17 LSMs were brought to their latest model versions. These models are run using the Land Information System (LIS) software framework developed at GSFC. The LIS software allows a common driver for all LSMs as well as the assimilation of remotely-sensed soil moisture and snow states and of terrestrial water storage anomalies to improve model states and fluxes.

For this presentation, the new/upgraded LSMs will be compared against the Phase 2 versions of the LSMs under the NLDAS configuration. The Land Verification Toolkit (LVT), developed at GSFC, will be used to evaluate the LSMs against available observations including streamflow, in situ soil moisture, snow products from in situ and remote sensing (e.g., depth, SWE, snow cover fraction), and surface fluxes from towers and from gridded flux products. This suite of tests will examine the effects of upgraded model physics and of data assimilation on the evaluation metrics (i.e., anomaly correlation, RMSE, bias, Nash-Sutcliffe efficiency, Taylor Skill Scores, etc.). Integrated metrics will be presented that will normalize these evaluations across separate components of the water cycle to investigate if the upgraded model physics and/or included data assimilation updates can enhance the simulation skill and improve the overall accuracy of the model products. Additionally, results will be presented from a regression model to estimate the information provided/lost from the model physics and model parameters. The goal of this work is to quantify the changes in model performance from the upgraded physics and data assimilation, and to identify the land model output that may be best for each end user's particular application.