Remote Sensing Information Mining of Soil Moisture and Surface Temperature Based on Temporal Sequence Association

This paper fully applies very high time resolution of satellite data and time sequence-based association mining of association rule mining method to concentrate on studying time sequence image data. According to characteristics of remote sensing image data, various surface parameters are inversed at first including soil moisture, land surface temperature LST and dust weather situation. Then, the paper describes association rule mining on practical remote data set in detail and proposes a design prototype of remote data mining. It transforms image data into discrete data sequence in the form of discrete data as well as similarity detection and extracts time sequence association rule through data sequence. Finally, it analyzes and evaluates functions of soil moisture and temperature on land surface in comprehensive wind erosion model. It also provides quantitative basis for occurrence and development of Asian sand storm.