Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #347466

Research Project: Integrating Remote Sensing, Measurements and Modeling for Multi-Scale Assessment of Water Availability, Use, and Quality in Agroecosystems

Location: Hydrology and Remote Sensing Laboratory

Title: A 3-step framework for understanding the efficiency of surface soil moisture data assimilation for improving large-scale runoff prediction

Author
item MAYO, Y. - UNIVERSITY OF WASHINGTON
item Crow, Wade
item NIJSSEEN, B. - UNIVERSITY OF WASHINGTON

Submitted to: Journal of Hydrometeorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/1/2018
Publication Date: 1/18/2019
Citation: Mayo, Y., Crow, W.T., Nijsseen, B. 2019. A 3-step framework for understanding the efficiency of surface soil moisture data assimilation for improving large-scale runoff prediction. Journal of Hydrometeorology. 20:79–97. https://doi.org/10.1175/JHM-D-18-0115.1.
DOI: https://doi.org/10.1175/JHM-D-18-0115.1

Interpretive Summary: Data assimilation is a powerful tool for constraining dynamic models using geophysical observations. It can potentially be applied to integrate satellite-based soil moisture observations into hydrologic models to produce the best-possible stream flow forecast for an agricultural basin. However, efforts to realize this potential have been hampered by large variations in the success of data assimilation when applied to different hydrologic models. This paper reviews this problem and describes a new technique to isolate specific features of a hydrologic model which determine whether the model can be improved via the assimilation of soil moisture observations. Understanding these features is critical for maximizing the benefit of satellite-based soil moisture observations for hydrologic forecasting. Therefore, the approach outlined in this paper will eventually be used by operational hydrologic modelers to improve the accuracy of their stream flow forecasts.

Technical Abstract: Data assimilation (DA) techniques have been widely used to assimilate soil moisture (SM) measurements into hydrologic models to improve runoff predictions. Past studies typically treated the assimilation procedure as a black box, which makes it difficult to understand why a particular DA procedure improved or failed to improve runoff predictions. Here we propose a framework that decomposes the surface SM DA procedure into three steps: 1) improvement of surface SM state via surface SM assimilation; 2) improvement of deeper-layer SM states via surface SM assimilation; 3) improvement of runoff via improved multi-layer SM states. The first two steps are evaluated by a synthetic twin experiment that examines various antecedent SM state correction via DA. The third step is evaluated by a set of synthetic perfect-state runs where error-free antecedent states are given to the hydrologic model to evaluate the maximum achievable runoff improvement. The framework allows the diagnosis of bottlenecks in a DA system that may inhibit runoff improvement. As a case study, the 3-step framework is successfully applied to the Arkansas-Red River basin using the Variable Infiltration Capacity (VIC) hydrologic model. We conclude that in this application both surface and deeper-layer SM states are effectively corrected via surface SM DA. The runoff prediction is moderately improved, which is mainly attributed to the strong control of deeper-layer SM states on slow-response runoff while the contribution of surface SM update on runoff improvement is limited.