Assimilation of SWOT Observations for the Creation of Spatially and Temporally Consistent Discharge Records

Abstract:

The Surface Water and Ocean Topography (SWOT) mission is designed to provide global estimates of water surface elevation, slope and discharge from space. This mission will provide increased spatial and temporal coverage compared to current altimeters. However, the temporal sampling is less frequent than current in-situ discharge observations. Thus, there is a need for methods that can utilize spatially and temporally inconsistent observations of discharge to reconstruct fields that are consistent in time and space.

Using the Inverse Streamflow Routing (ISR) model of Pan and Wood [2013], discharge records are derived for the Ohio River basin using data assimilation with a fixed interval Kalman smoother. ISR utilizes observed (or SWOT retrieved) discharge values at discrete (gauge) locations to generate spatially and temporally distributed fields of runoff by inverting a linear routing model. These runoff fields are then routed to produce river discharge estimates throughout the basin.

Three experiments have been carried out to evaluate assimilating SWOT observations. The experiments are: (1) assimilating 75 in-situ gauges only, (2) using 50 in-situ gauges and 25 SWOT-retrieved “gauges”, and (3) using 75 SWOT-retrieved “gauges” only. The estimated discharges are compared to in-situ USGS gauge data from 2006 to 2009. Results show that the ISR assimilation method can be used to effectively reproduce the spatial and temporal dynamics of discharge in each of the experiments. In particular, the results of the SWOT-only data experiment indicate that despite the coarse temporal SWOT overpasses (0 to 3 over a 22 day period) significant discharge information throughout the entire basin can be retrieved. The ISR-SWOT assimilation approach will provide extremely useful discharge estimates, especially in sparsely gauged regions where spatially and temporally consistent discharge records are most valuable.

Pan, M; Wood, E F 2013 Inverse streamflow routing, HESS 17(11):4577-4588