Evaluation of High-Resolution Simulated Runoff in the Midwest using the National Water Model

Abstract:

High-resolution hydrological models, such as the National Oceanic and Atmospheric Administration’s National Water Model (NWM), are being developed primarily for improving river forecasts. However, these models also offer an opportunity for improving the understanding and prediction of other important water balance variables, such as surface runoff at a fine scale. Water quality issues across the Great Lakes and upper Midwest are attributed to runoff of nutrients every year. Runoff risk tools help provide guidance to agricultural decision makers and represent the importance of skilled high-resolution runoff forecasts. This study aims to assess the skill of the NWM simulated runoff by using WRF-Hydro, with the NWM 2.1 configuration, to run experiments across the Midwest region. Runoff is evaluated by analyzing meteorological and hydrological variables that contribute to the water balance, such as ground water storage and precipitation. This regional assessment employs observational data such as USGS station data, edge-of-field runoff measurements, and the Snow Data Assimilation System (SNODAS) in order to establish model skill and recommend potential improvements for regional water prediction. The study takes place within a larger project aimed at advancing the NOAA modeling framework that informs state runoff risk tools.