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 #303208

Title: The impacts of assimilating satellite soil moisture into a rainfall-runoff model in a semi-arid catchment

item ALVAREZ, C. - University Of Melbourne
item RYU, D. - University Of Melbourne
item WESTERN, A. - University Of Melbourne
item Crow, Wade
item ROBERTSON, D. - Commonwealth Scientific And Industrial Research Organisation (CSIRO)

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/25/2014
Publication Date: 11/1/2014
Publication URL:
Citation: Alvarez, C., Ryu, D., Western, A., Crow, W.T., Robertson, D. 2014. The impacts of assimilating satellite soil moisture into a rainfall-runoff model in a semi-arid catchment. Journal of Hydrology. 519(D):2763-2774. DOI:10.1016/j.jhydrol.2014.07.041.

Interpretive Summary: Estimating runoff from agricultural catchments is important for efforts to: minimize the impact of agricultural management on water quality, forecast downstream flooding, and monitor the availability of water resources during periods of drought. A promising approach for improving runoff estimates is the integration of remotely-sensed soil moisture estimates into a hydrologic model. Since soil moisture conditions determine the land surface's ability to infiltrate future rainfall, the improved estimation of pre-storm soil moisture conditions should facilitate better monitoring of stream flow conditions. With this goal in mind, this paper evaluates the degree to which stream flow predictions made by a hydrologic model in a semi-arid catchment can be improved by integrating satellite-derived surface soil moisture estimates into the model. The results of this analysis can be used to improve the operational monitoring of stream-flow conditions in agricultural catchments. Better monitoring of this conditions will - in turn - allow local water resource manager to minimize the impact of agricultural management on water quality and better manage water resources in semi-arid areas.

Technical Abstract: Soil moisture plays a key role in runoff generation processes. As a result, the assimilation of soil moisture observations into rainfall-runoff models is increasingly being investigated. Given the scarcity of ground-based in situ measurements, satellite soil moisture observations offer a valuable dataset to be explored; however, there is still little evidence that assimilating these observations can actually aid flood prediction. In this study we update the soil moisture state of the probability distributed model (PDM) by assimilating an Advanced Microwave Scanning Radiometer (ASMR-E) Level 3 soil moisture retrieval product using an Ensemble Kalman Filter approach. To remove the systematic differences existing between AMSR-E and modeled soil moisture products, linear regression (LR) and anomaly-based cumulative distribution function (aCDF) rescaling strategies are applied to the observed data. For the aCDF, seasonality is removed by grouping soil moisture values by corresponding months and applying standard CDF matching to the separately grouped values. Additionally, to account for the depth-mismatch between observations and the model predictions, a soil wetness index (SWI) is estimated from the satellite soil moisture and then rescaled either by using LR or by matching the aCDF. We assimilate these four rescaled products into PDM and evaluate the resulting improvement in PDM discharge estimates. On average, prediction uncertainty - expressed as the ensemble mean of the root mean squared difference (MRMSD) in discharge - is reduced by 30% after the four rescaled soil moisture products are assimilated. However, when specific flood events are analyzed, the level of improvement varies. These findings provide novel evidence of the advantages of assimilating satellite soil moisture observations for improving flood prediction via hydrologic models.