Satellite Remote Sensing SST Quality Evaluation Based on Consensus Measurement

In areas lack of fixed buoys monitoring Sea Surface Temperature (SST), but rich of multi-satellite measured SST, consensus evaluation among multi-source of data can supply a large number of highly credible data as benchmarks to substitute for buoys. This paper proposes a method named Weighted Joint Probability Distribution (WJPD) to quantify the degree of consensus amongst multi-source SST data. In addition to the virtue that flexible regulation can be realized to distinguish different types of conflict, WJPD can detect false alarm and serious conflict that can not be found by using variance method. WJPD can also give a clearer idea about the exact confliction extent than the traditional variance method. WJPD is used in processing SST data from eleven sensors in area of East China Sea for 2006. As a result, this paper confirms a large number of highly consistent points, identifies failure data, and finds controversial areas which deserve further research.