Improving seasonal streamflow forecasts by assimilating high-resolution radar precipitation and snow data

Abstract:

Reliable and skillful seasonal streamflow forecasts is crucial for efficient management of water resources. Specifically, seasonal streamflow forecasts can provide advanced water availability information that can help water managers for a range of applications- water supply, hydropower generation and drought risk management. In snowmelt dominated river basins, streamflow is sensitive to seasonal changes in precipitation as well as temperature and incoming radiative fluxes because runoff in these basins largely depends on the amount of precipitation and snow accumulation, which mainly occurs during the winter, spring and late part of the fall season. Snowmelt also depends on temperature and solar radiation fluxes, however, seasonal anomalies in temperature are much more consistent and predictable than precipitation. Thus, high-resolution and skillful precipitation products can play a more important role towards improving the accuracy of seasonal forecasts. In this study, we will evaluate the impact of a new radar product utilized within an experimental version of Multi Radar Multi Sensor (MRMS) precipitation product from NOAA’s National Severe Storm Laboratory (NSSL). MRMS precipitation will be used as an input into the WRF-hydro hydrological modeling system to obtain seasonal streamflow forecasts in the selected watersheds and reservoirs of Colorado River Basin. To explore the improvements, MRMS driven hydrological model outputs (i.e. streamflow and snowpack) will be compared with North American Land Data Assimilation System-2 (NLDAS-2) driven outputs. NLDAS-2 driven WRF-hydro seasonal streamflow forecasts have already shown competitive skill when compared to other operational forecast products. In addition to assessing the impact of new radar information on seasonal water supply skill, we will also evaluate a bias correction-snow data assimilation methodology to the seasonal streamflow forecasts using in situ snow and remotely sensed snow observations.