Data mining has been an important technique that discovers implicit knowledge from large amount of data. Since the amount of spatial data grows rapidly, the concepts and methods of data mining have been involved in the geographic researches varies from GIS, remote sensing to environmental assessment. Among all spatial data mining algorithms, decision tree classification is commonly utilized to construct rules from a variety of datasets. In this paper, two spatial applications of decision tree classification would be addressed. In the first one, the idea is applied to find regularities hidden between land change spots and related spatial data such as digital terrain model, slope model and road map, etc. In the second application, the same scheme is utilized to find the rules concealed in the locations of invasive plants and other GIS layers like soil map, land use map and digital terrain model. The valuable information extracted by data mining in the first study is used to perform the prediction of land change locations while in the second study it helps to forecast the allocation of invasive plants. The knowledge mined in the both studies not only assists in environmental monitoring, but also shows the potential of the integration of GIS and data mining technique.