L-band brightness temperature assimilation into a distributed hydrological model : application to the Community Land Model over Australia

An increase in global population and climate change will lead to added stresses on water resources. This requires hydrological monitoring systems which can provide critical information on the hydrological state to decision-makers. One key hydrological variable is soil moisture, which codetermines several hydrological processes such as infiltration, runoff production and evaporation. Moreover, soil moisture plays an important role in the formation of droughts and floods. Land surface models are able to provide soil moisture information across large spatial scales but can lack accuracy. Remote sensing offers great potential in providing independent information. Through data assimilation techniques, the model can be steered towards better predictions by assimilating remotely sensed soil moisture information. In this thesis, observations from the Soil Moisture and Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) missions are assimilated into the Community Land Model (CLM) over Australia. Both SMOS and SMAP carry radiometers and are the first dedicated missions to observe soil moisture from space. They operate within the L-band frequency, which is considered optimal for soil moisture. CLM is a state-of-the-art land surface model that provides all necessary information to integrate these observations. The focus on Australia allows the use of high-quality observations that are not contaminated by Radio Frequency Interference, as is the case in many other parts of the world. The thesis explores in detail (i) the impacts of assimilating long time series on soil moisture probability levels, (ii) the use of SMOS and SMAP within a common assimilation scheme, and (iii) the effects of an improved land scape representation within CLM on the assimilation performance. Briefly explored is also the potential of downscaling the observations by using model information at higher resolution. The assimilation improves the soil moisture simulations and the findings support the integration of L-band observations into operational monitoring systems