Testing Soil Moisture Performance Measures in the Conceptual‐Functional Equivalent to the WRF‐Hydro National Water Model

ABSTRACT The Conceptual‐Functional Equivalent (CFE) to the National Water Model (NWM) serves as a baseline rainfall‐runoff model in the National Oceanic and Atmospheric Administration (NOAA)'s Next Generation National Water Model Framework (NextGen). The CFE model performs similarly to the earlier version of the NWM, allowing comparisons with new models introduced in future versions. In addition to streamflow, the NWM outputs other hydrologic variables including soil moisture. Soil moisture plays a key role in simulating seasonal hydrologic processes in process‐based models; therefore, it is imperative to evaluate models against observed data. However, incorporating in situ observed soil moisture data, which is highly spatially variable, into the calibration process may compromise streamflow results. We investigate how model evaluation, including in situ soil moisture observations, affects CFE's ability to reproduce streamflow and soil moisture. We evaluated the CFE model on two experimental watersheds using both traditional and signature‐based performance metrics for soil moisture. Results showed that including soil moisture data enhances the reproducibility of overall and seasonal soil moisture patterns without sacrificing the reproducibility of streamflow. Calibration against streamflow alone was insufficient to reproduce soil moisture patterns. We recommend including soil moisture metrics when available in the CFE model calibration to improve seasonal streamflow predictions.