TM imagery uses 0-255 grey levels to represent different spectrum of the ground truth. In the ease of using TM data to enhance and map mineral bearing alterations

Amos （1989） and Zhao Yuanhong （1991） applied TM ratio+ PCA to enhance alterations with eliminating the influencial factor of vegetation; Crosta （1991） and loughlin （1991） used PCA+ negative+ PCA to successfully identify alterations. However

those methods have unsatisfactory results in the areas where dominanted with forestry、grass、glacial ice、loess、water bodies、and clouds. The new methodology to eliminate those environment factors which have strong spectral reflectance comparing with relative weak spectral reflectance of alterations

and which have impact on spectral signatures of the TM data in terms of differentiating alterations from surrounding rocks is developed mainly by TM data masking-PCA-supervised and unsupervised classification in the testing areas selected both in Canada and China. The results are proved vary effective.