资源评价方法32-ARC-SDM (arcview3.2).ppt

ARC-SDM Key Terms - Weights of evidence Logistic regression - Fuzzy Logic - Neural Network(RBFLN) Neural Network(Fussy) Quantitative Mineral Potential Mapping in GIS Predict the probability of occurrence of point objects. The GIS-based mineral potential mapping process can be broken down into four main steps: (1) building a spatial digital database, (2) extracting predictive evidence for a particular deposit type, based on an exploration model, (3) exploring each evidential theme individually in order to generalize or classify it, (4) combining the evidential themes to predict mineral potential. Key Terms Exploration data set Digital data sets, such as digitized geological maps, geophysical images, geochemical survey data and remotely sensed images, frequently employed by exploration geologists in the mineral exploration process. There is always a process of extracting evidence from the raw data set to be used in prediction. This process depends strongly on the exploration model being used. Exploration model There are a large number of deposit models that have been defined for the various mineral deposit types,based on deposit characteristics. Deposit models help to classify and identify new occurrences, lead to an improved understanding of how the deposits formed, and act as an aid to exploration. However, they are usually based on characteristics of the deposit and its immediate surroundings, and many of the characteristic diagnostics of a mineral deposit type cannot be used with regional exploration data sets. The exploration model refers to the characteristics of a deposit type identifiable in regional data sets, as used in the exploration process. It is important to note that the absence of evidence can be equally important as the presence of evidence and data on negative associations between deposit and particular types of data are often poorly documented. Evidential Theme This is a map or area layer (in either vector or raster format) used for predic