SMOS Validation and Soil Moisture Product Enhancement with data assimilation methods

The terrestrial part of ESA's Soil Moisture and Ocean Salinity (SMOS) mission is focused on the spatial and temporal dynamics of soil moisture, which will have a large impact on the understanding of climate-related processes and will help to improve the forecasts of climate change, weather and extreme-events. MIRAS, the radiometer system onboard SMOS, records brightness temperatures which cannot be implemented into the relevant climate or weather forecast models directly. ESA provides a SMOS Level-2 product processed by an operational routine that includes a radiative transfer model, which in turn needs further information about the vegetation cover and surface conditions in order to generate a high accuracy soil moisture product. So far, this additional information is just basically parameterized and does not adequately consider the seasonal variability of the vegetation.

SMOS Validation and Soil Moisture Product Enhancement with data assimilation methods

Result of a simulation experiment: Uncertainty evolution of vegetation opacity [-] (top), roughness parameter [-] (middle) and bias [%] (bottom) for parameter estimation by the Particle Filter. Shaded areas correspond to 95, 90, 68, and 10 percentile confidence intervals; red line shows the average and the blue line the simulated true value.

The accuracy of this soil moisture product can be enhanced by data assimilation techniques. A coupled model system containing a hydrometeorology model as well as a radiative transfer model is integrated into a data assimilation framework using a sequential Monte Carlo algorithm, which is able to update both model states (e.g. soil moisture or brightness temperatures) and model parameters (e.g. surface roughness, vegetation opacity). This study will focus on the estimation of parameters for the radiative transfer model and their spatiotemporal dynamics by assimilating SMOS brightness temperature and in situ soil moisture observations. Enhanced SMOS data products taking into account the uncertainty of the data will have a high impact on the scientific outcome produced by SMOS-data users. Strategies for an operational application of the proposed approach will be formulated. The approach of using additional world-wide available in situ observations during the processing may enhance also the accuracy of Level-2+ products from other ESA missions.

Dr. Carsten Montzka
Miltin Cho Mbo
Kathrina Rötzer
Long Zhao

Last Modified: 14.02.2023