Neglect of potential seasonal streamflow forecasting skill in the United States national water model

Authors: Crow, Wade; KOSTER, R - National Aeronautics And Space Administration (NASA); REICHLE, R - National Aeronautics And Space Administration (NASA); CHEN, F - Oak Ridge Institute For Science And Education (ORISE); LIU, Q - National Aeronautics And Space Administration (NASA)

Submitted to: Geophysical Research Letters
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/29/2024
Publication Date: 2/14/2024
Citation: Crow, W.T., Koster, R., Reichle, R., Chen, F., Liu, Q. 2024. Neglect of potential seasonal streamflow forecasting skill in the United States national water model. Geophysical Research Letters. 51. https://doi.org/10.1029/2023GL105649.
DOI: https://doi.org/10.1029/2023GL105649

Interpretive Summary: Hydrologic forecasting systems attempt to predict the magnitude of future streamflow based on current conditions. Recent work has underscored the importance of antecedent soil moisture for such forecasts. That is, knowledge of current soil moisture conditions can guide our expectation regarding future streamflow amounts. However, it is currently unclear whether existing hydrologic forecasting models accurately capture the usefulness of soil moisture for such a purpose. Here we use new data provided by the NASA Soil Moisture Active/Passive satellite mission to assess the degree to which the National Water Model (currently being developed for operational streamflow forecasting in the United States) accurately captures the degree of correlation between current soil moisture conditions and future streamflow. Results suggest that systematic changes to the National Water Model could improve its ability to forecast agricultural water resource availability.

Technical Abstract: Recent research underscores the key role that soil moisture plays in forecasting streamflow. For example, using data from the NASA Soil Moisture Active/Passive mission, Koster et al. (2023) conclude that for medium-scale basins in the contiguous United States, a quarter of interannual variability in springtime streamflow is explained by interannual anomalies in late-fall root-zone soil moisture (RZSM). The strength of this lagged relationship can be leveraged for seasonal hydrologic forecasting, but only if effectively captured by existing hydrologic forecasting models. Here, we extend the analysis of Koster et al. (2023) to diagnose systematic error present in the National Water Model (NWM). Results demonstrate that the NWM tends to underestimate both the trans-winter temporal memory of RZSM as well as the correlation between concurrently averaged RZSM and streamflow – thereby reducing the NWM’s ability to leverage an important source of streamflow predictability.