Remote sensing data fusion by co-location of footprints from different sensors

A new tool for remote sensing data fusion has been developed. The footprint which corresponds to a single pixel from a coarse-resolution sensor is defined using the geographic information in the data as well as the radiometric characteristics of the sensor. It is then overlaid on a background image acquired from a fine-resolution sensor. The single radiometric parameter (or any derived physical parameter) associated with the pixel in the coarse-resolution data can then be linked to the statistics of the radiometric observations from all co-located pixels of the fine-resolution image (i.e. pixels within the footprint of the coarse-resolution pixel). Examples from sea ice applications are presented using coincident data from SSM/I, AVHRR, and RADARSAT. Contours of footprints from a coarse-resolution sensor can be presented in color to designate either the observed radiation (e.g. brightness temperature from an SSM/I channel) or a derived parameter such as ice thickness or surface temperature. The technique is particularly useful in linking observations from coarse-resolution sensors to the sub-pixel surface information available from the fine-resolution sensor