Satellite soil moisture measurements improve streamflow estimation
New research shows that hydrological model streamflow estimates can be improved be integrating streamflow and satellite soil moisture measurements.
The team of scientists from the Netherlands, US and Australia set out to assess the difference in model-derived streamflow estimates when introducing streamflow and satellite soil moisture data. They tested this by driving the model with both coarse or high spatial resolution meteorological data in the Murrumbidgee River basin in Australia.
Results show the improvement gained by assimilating streamflow and satellite soil moisture observations can bring the closer to those achieved using a locally calibrated model.
The study, published in Hydrology and Earth System Sciences, contributes to recent efforts to improve global hydrological model estimates without modifying the model structure, such as through local calibration of model parameters. Instead, the aim is to use global earth observations products, such as the AMSR-E soil moisture data to achieve better model performance.
According to lead author, Patricia López López, “these experiments may constitute a step forward by showing the suitability of incorporating remotely sensed observations into global models for their application at a river basin scale.”
An important point raised by the researchers is that this work demonstrates the usefulness of satellite soil moisture observations in improving model estimates for river basins all over the world, due to the near global coverage of the data. This is particularly relevant in ungauged river basins or with a limited number of gauging stations, where traditional calibration based on in situ data is not possible.
The authors also noted that the improvement on model estimates of assimilating downscaled soil moisture observations is higher than when non-downscaled soil moisture observations are assimilated. This indicates that further investments and improvements in remotely sensed observations may benefit hydrological model predictions.
Future research work is needed in order to test other available near-global remotely sensed observation products in a range of river basins with diverse hydrometeorological characteristics.
This research demonstrates that further investments and improvements in remotely sensed observations, especially in soil moisture products can benefit large-scale hydrological model predictions and bring these closer to those obtained from local-scale hydrological modelling.
Article: ‘Improved large-scale hydrological modelling through the assimilation of streamflow and downscaled satellite soil moisture observations’, Patricia López López, Niko Wanders, Jaap Schellekens, Luigi Renzullo, Edwin Sutanudjaja and Marc Bierkens, Hydrology and Earth System Sciences, doi: 10.5194/hess-20-3059-2016