Discrimination of Local and Faint Changes From Satellite-Borne Microwave-Radiometer Data

In order to extract a local and faint variation generated by a specific source from a physical value obtained by a remote-sensing satellite, we should focus on the difference between physical values at two points. We have developed a new data-restructuring algorithm based on this concept. This algorithm enables us to obtain a distribution of physical values at any interval and to calculate the difference between any two points regardless of the sensor-sampling interval. This paper presents this algorithm and how to implement it on a computer. Based on this data-restructuring algorithm, we have established a new data-processing method and analyzed the brightness-temperature data obtained by a microwave radiometer. As a result, we have successfully extracted the feature associated with a volcanic eruption. This feature cannot be extracted from brightness temperatures directly. This paper demonstrates this data-processing method by analyzing a volcanic eruption.