Empirical determination of thermal noise levels in synthetic aperture radar

Thermal noise in synthetic aperture radar can be significant when working with dark targets. This paper outlines an empirical method of determining thermal noise levels for Radarsat beam modes ST 5, ST 6 and ST 7. Calm water bodies are the reference targets that are used in the empirical analysis. The backscatter from these targets is assumed to be very close to the noise floor. This empirical method is compared to a theoretical calculation based on nominal values from the Canadian Space Agency. Using the theoretical method the user is responsible to determine additional processing parameters applied to the data during processing and subtract those back out from the image. This method may not account for all the possible contributing factors to thermal noise levels.