Different rocks in the Xiongcun ore district in Xietongmen county, Xizang (Tibet) have their respective characteristic spectra, which are very helpful to the explanation of remote sensing data and the search for copper—polymetallic deposits by remote sensing technology. The paper deals with basic principles and technique method of mineralization information discrimination based on apparent reflectivity image, and presents a case study carried out by authors. Based on the study of the apparent reflectivity image and the ground spectrum from copper—polymetallic deposits in the Xiongcun ore district, the ETM apparent reflectivity image in the region has been transformed into ground—space correlation image with the nine ground—space models which were established by means of a correlation study of the ground—space spectrum. Through the analyses of the ground—space correlation image, some mineralization image anomalies have been distinguished,and some basic problems concerning the practical application of the discrimination—classification method for remote sensing have been discussed in the paper. The authors hold that the analyses of ETM ground—space spectrum correlation is valuable in the prospecting and prognosis of copper —poymetallic deposits in Gangdese, Xizang (Tibet).