Improving hydrological simulations via the integration of remotely sensed data assimilation and coupling of the WRF-Hydro model with NASA's Land Information System (LIS)

The use of remotely sensed observations is less frequent in the hydrological model environments whereas land surface models can benefit from the recent advancement of hydrologic models through incorporating critical hydrologic processes including surface-subsurface water interactions, multiscale spatial variability in hydrologic connectivity, and surface and subsurface flow routing that finally impact the estimating and forecasting stream flows. The NASA's Land Information System (LIS) is a software framework for high-performance hydrological modeling and data assimilation, and it was developed to integrate land surface models and hydrological models, satellite, and ground-based observational data products using advanced modeling techniques to obtain optimum land surface state and fluxes.

We coupled NASA Land Information System (LIS) and NCAR WRFHydro model components using Earth System Modeling Framework (ESMF) infrastructure and further refined the component interoperability and data sharing using the NUOPC (National Unified Operational Prediction Capability) compliant software layer. We validated the coupled LISHydro model by comparing it with the baseline outputs of the LIS stand-alone model and WRFHydro stand-alone model.

Finally, we utilize the coupled LISHydro model environment to demonstrate the impact of snow data assimilation on the stream flows and their forecasts during the NASA Airborne Snow Observatory (ASO) campaigns in the western United States. The ASO's combined spectrometer and lidar instruments provide snow albedo, snow depth, and snow water equivalent (SWE) during this campaign. We will test our implementation on the Tuolumne Basin, CA because primarily due to the availability of a substantial record of observational data in the region and validate with the USGS streamflow measurements in many first-order watersheds distributed across the study area. The results from assimilating ASO data in LIS will be shown in this presentation.