Remote-sensing Geological Information Mining And Metallogenic Prediction In The Grassland Covered Area In Inner Mongolia

Content:The Erlian-Dongwuqi Metallogenic Belt is a typical grassland landscape area with a large coverage, endowed with a variety of mineral resources due to advantageous metallogenic geological conditions. However, the geological research and prospecting work face a lot of difficulties due to the regional vegetation and Quaternary cover. The most popular and difficult study for the region is the application of remote sensing technique to geological survey in the covered area. The traditional method of remote sensing interpretation and information extraction has been proven to be unsuitable for the landscape in the region, new ideas and creative methods are expected.The paper studies both less covered grassland area in the northern Erlian and densely covered grassland area in Dongwuqi, previous remote sensing geological interpretation methods and information extraction techniques were carefully studied, the ASTER data was selected as the main data source, a systematic study of digging for information of remote-sensing geological interpretation in the grassland cover area was carried out. The main work carried out and the achieved results and innovations are as follows:1. The research on application of vegetation information weakening and lithologic information enhancement was carried out. The "forced invariance" technique was used to weaken vegetation information, a set of ASTER data optimized band combination and a lithological enhancement technique was proposed, and have been proven to be suitable for grassland covered area.2. The remote sensing geological interpretation technique in the grassland covered area was summarized and innovated. A new idea was proposed that stratigraphic information mining on the basis of petrophysics and diagenetic environments and information mining of intrusive rocks on the basis of morphological characteristics and geological processes, and was proven to be applicable in the covered area. At the same time, the multi-source information fusion and comprehensive comparative research by the application of principal component analysis and wavelet transform technique etc. was carried out, the degree of interpretation in the covered area is improved.3. The study on method of alteration information extraction and map compilation was carried out. The spectral characteristics of common alteration minerals in the ASTER data format were summarized and analyzed, and their classification was also made. The modelling information extraction technique based on the decision tree model and accurate information extraction technique according to different lithologic settings are proposed. And they have resulted in good results. At the same time, the concept of the information field is introduced into the alteration information presentation method, by which the remote sensing mineralized alteration information isodensity was compiled, effectively improved the application of alteration information in the geological survey.4. The metallogenic prediction by the application of remote sensing information was carried out, several mineralization targets were delineated, with a good result of prospecting found.