Assimilation of passive microwave observations into atmospheric models
Final Report Abstract
In total, this study shows no significant improvements of modeled soil moisture and surface heat fluxes by assimilating passive microwave observations. Reasons for the unsatisfying results discussed in the previous sections can be summarized as follows: • The effective thermal emission soil depth of passive microwave data at X-band (as from TMI) is too low to be representative of the soil moisture conditions needed for modelling purposes (at least for the site analyszed here). • Despite of the quality checks and the data corrections applied, the observations at the NOAA/ATDD site may still be non-optimal due to technical shortcomings in the soil moisture and turbulent heat flux measurement procedure which, however, is hardly avoidable for these variables on long-term observation sites. • The soil model TESSEL is not able to consistently simulate both soil moisture as well as sensible and latent heat fluxes due to shortcomings in the corresponding model parameterization schemes. Potential future operational soil moisture analysis systems in atmospheric models should rely on L-band passive microwave observations which are more appropriate for soil moisture sensing due to their higher soil penetration depth. They will soon be available through the ESA Soil Moisture and Ocean Salinity Mission. Furthermore, the land surface schemes of current atmospheric models, which at present are primarily tuned to get the surface heat fluxes right, need to be upgraded to be able to simulate hydrologic processes at the soil surface realistically (i.e. improvement or replacement of current parameterizations). And thirdly, calibration and validation of the new assimilation systems require thoroughly measured and quality-checked soil moisture and surface heat flux observations over an area equivalent to the satellite footprint resolution which, however, requires a significant effort of technical and human resources.
