Applications of trend surface analysis and geologic model building to mineralized districts in Alaska

Abstract

The Mineral Industry Research Laboratory, University of Alaska, has investigated the application of computers and statistics to mineral deposits in Alaska. Existing programs have been adapted and new ones written for the computers available at the University. The methods tested are trend surface analysis and geologic model making. An existing coeffecient of association program was converted to Fortran IV , but was not applied to an Alaskan problem. A trend surface is a mathematically describable surface that most closely approximates a surface representing observed data. In geologic model making, regression analysis is used to determine what geologic features are significant as ore controls. Coefficient of association compares samples to each other on the basis of a variable being present or absent. Trend surfaces were computed for dips and s t r i k e s of geologic features ( v e i n s , f a u l t s , bedrock) for Southeastern Alaska, the Chichagof district , and the Hyder district . Results for the f i r s t two are presented as maps. Trend surfaces and residual maps were prepared for geochemical data from the Slana district, Alaska. A mineral occurrence model was made for a portion of the Craig Quadrangle, and potential values were computed for c e l l s in the area. Appraisals of potential values by five geologists are compared with those of the model. An IBM 1620 multiple regression program is included.