Retrieval of surface microwave emissivity using multisensor satellite measurements

This thesis concerns the development and validation of a technique to measure the microwave emissivity of the Earth's surface from space. The dependence of microwave emissivity upon various surface geophysical parameters is discussed. The role of these geophysical parameters in the global climate system together with the importance of satellite remote sensing are also discussed. The physics of microwave emission and the models used to predict the microwave emission of the Earth's surface are described, including the assumptions made and constraints that apply. The atmospheric contribution to the microwave signal from space is investigated and found to be a major source of errors. A new technique is developed to correct the apparent microwave emissivity as seen by the radiometer from space for both atmospheric absorption and direct and reflected emission due to water vapour using radiometer data in both the microwave and infrared regions. Validation study of this new technique over the ocean surface is performed using simultaneous data from the Along Track Scanning Radiometer (ATSR) and the Microwave sounder (ATSR/M). The technique is applied to the soil moisture retrieval from space by performing a case study over the Simpson Desert in Australia using near-contemporaneous data from the Advanced Very High Resolution Radiometer (AVHRR) and TOPEX microwave sounder. Results obtained from the satellite data are compared with contemporaneous ground data. Analysis of errors and error sources is discussed. Constraints on the accuracy of emissivity measurement are set out and recommendations are suggested for future application.