成矿预测中ArcGIS ModelBuilder的建模评价——以武夷山成矿带大深幅为试验区

Space Analysis Modelling of Mineralization Forecast Appraisal Using ArcGIS ModelBuilder: Taking Wuyishan Mineralization Belt Dashen Sheet as the Pilot Area

为实现数字化、智能化、自动化成矿远景预测评价,提出以证据权法、信息量法等数据驱动结合知识驱动,利用ArcGIS空间分析建模实现地质变量筛选、构置,并得到图层表示的找矿地质模型与相应的量化模型。通过对变量特征、量化方法及GIS单元划分与赋值等的分析,以规则网格、等价的中心点分别为基本单元,以及重要性法、中心点法、面积法识别变量,经ArcGIS空间分析建模实现"0,1"法赋值。同时用多维空间线性映射的定量模型,以克服地质变量不等价、不独立对评价结果的影响,以武夷山成矿带1∶5万大深幅为实验区,应用ArcGIS ModelBuilder方法,对该区铁、铜、铅锌成矿远景进行预测评价,圈定远景区12处,其中,A类3处、B类3处、C类6处,与信息量法、证据权法的预测结果基本一致。

In order to realize digitized,intellectualized and automated mineralization prospect forecast appraisal,first we carried out mineralize prospect quota forecast appraisal question decomposition under the GIS environment,based on data-driven and knowledge-driven approaches such as the evidence weighted method and information mensuration,to realize geological variable screening and Layer through the ArcGIS space analysis model builder,as well as get the comprehensive information prospecting geologic model and information mensuration or evidence weighted quantitative model by the layout description.Next,to the geological variable characteristic,different mathematical quantification method,unit division and the evaluation method and so on,through analysis and comparison under the GIS environment,proposed simultaneously the rule grid（surface shape） or the equal grid center（point shape） as the basic unit,important method,central point method,area method and corresponding＂0 or 1＂two-state evaluations as recognition variable existence or not through ArcGIS space analysis modelling realization.Thirdly,based on the variable standardization transformation and the multi-dimensional space linearity mapping quantitative evaluation model,we overcomed the influence of geological variables,which,are not equal and not independent,to the appraisal.On that basis,taking the map of Dashen sheet in Wuyishan metallogenic belt of scale 1∶50000 as the experimental plot,applying ArcGIS Model Builder method,we finally established the ArcGIS space analysis model on iron,copper,lead-zinc mineralization prospect quota forecast appraisal,selected 12 prospect areas,which included 3 areas of class A,3 of class B and 6 of class C.The forecast result is basiclly consistent with that of the information mensuration and evidence weighted method.