A fast generic millimetre-wave emissivity model

In recent years an increasingly diverse range of passive microwave satellite data has become available for applications in numerical weather forecasting, climate studies and environmental monitoring. Top of atmosphere radiance is measured, which has originated from both the surface and the atmosphere. When retrieving atmospheric quantities such as temperature, humidity or cloud liquid water content near the surface it is necessary to account correctly for the contribution to the measured radiance from the surface. This depends on the surface emissivity, which varies widely with surface type, roughness and temperature. For real time applications such as numerical weather forecasting, it is necessary to be able to model the surface emissivity very quickly. There is therefore a need for a fast surface emissivity model for any surface type, instrument geometry and observation wavelength. A fast generic emissivity model has therefore been developed. This is a semi-empirical model. Some aspects are physically based, for example many surfaces can be assumed to be dielectric media. Other aspects such as geometric roughness have been parameterised for speed. For some complex or poorly understood aspects of the electromagnetic interaction empirical adjustments are made to fit observed values. The model has been compared with emissivities derived from aircraft radiometer measurements at 24, 50, 89 and 157 GHz. It is also intended to compare with data from the Special Sensor Microwave Imager and Advanced Microwave Sounding Unit instruments. For the ocean surface the fast model fits estimated emissivities from airborne radiometers to within 1-2%, and it fits the unparameterised model to within 0.1-0.3%.