Abstract:
Soil moisture plays a major role in the global hydrologic cycle, principally through its effect on the partitioning of energy and on precipitation at the land surface. As a result, soil moisture is a key variable in weather and climate prediction and in flood forecasting. It is well known that soil moisture is difficult to measure in situ over the scales needed for these applications. Remotely sensed radiobrightness measurements (for example 1.4 GHz L-band passive microwaves) provide good spatial coverage but are only sensitive to soil moisture in the top 5 cm of the surface layer.

We assess the performance of the Ensemble Kalman filter (EnKF) for soil moisture estimation by assimilating L-band microwave radiobrightness observations into a land surface model. A dynamic variational method (an optimal smoother) is used as a benchmark for evaluating the filter's performance. In a series of synthetic experiments we investigate the effect of ensemble size and non-Gaussian forecast errors on the estimation accuracy of the EnKF. With a state vector dimension of 4608 and a relatively small ensemble size of 30 (100), the actual errors in surface soil moisture at the final update time are reduced by 55 % (70 %) from the value obtained without assimilation (as compared to 84 % for the optimal smoother). For robust error variance estimates, an ensemble of at least 500 members is needed.

The dynamic evolution of the estimation error variances is dominated by wetting and drying events with high variances during drydown and low variances when the soil is either very wet or very dry. Furthermore, the ensemble distribution of soil moisture is typically symmetric except under very dry or wet conditions when the effects of the nonlinearities in the model become significant. As a result, the actual errors are consistently larger than ensemble-derived forecast and analysis error variances. This suggests that the update is suboptimal. However, the degree of suboptimality is relatively small and our results indicate that the EnKF is a flexible and robust data assimilation option which gives satisfactory estimates even for moderate ensemble sizes.