A reconnaissance method for delineation of tracts for regional-scale mineral-resource assessment based on geologic-map data

• Published in: Natural resources research (New York, N.Y.) N.Y.), 2002-12, Vol.11 (4), p.241-248
• Main Authors: RAINES, Gary L, MIHALASKY, Mark J
• Format: Article
• Language: English
• Subjects: Areal geology. Maps; Deposits; Earth sciences; Earth, ocean, space; Exact sciences and technology; Geochemical exploration, methodology, general; Geologic mapping; Geologic maps, cartography; Geological mapping; Geological surveys; Metallic and non-metallic deposits; Reconnaissance; Regional analysis; Regression; Regression analysis; Statistical analysis
• ISSN: 1520-7439; 1573-8981
• DOI: 10.1023/A:1021138910662

The U.S. Geological Survey (USGS) is proposing to conduct a global mineral-resource assessment using geologic maps, significant deposits, and exploration history as minimal data requirements. Using a geologic map and locations of significant pluton-related deposits, the pluton-related-deposit tract maps from the USGS national mineral-resource assessment have been reproduced with GIS-based analysis and modeling techniques. Agreement, kappa, and Jaccard's C correlation statistics between the expert USGS and calculated tract maps of 87%, 40%, and 28%, respectively, have been achieved using a combination of weights-of-evidence and weighted logistic regression methods. Between the experts' and calculated maps, the ranking of states measured by total permissive area correlates at 84%. The disagreement between the experts and calculated results can be explained primarily by tracts defined by geophysical evidence not considered in the calculations, generalization of tracts by the experts, differences in map scales, and the experts' inclusion of large tracts that are arguably not permissive. This analysis shows that tracts for regional mineral-resource assessment approximating those delineated by USGS experts can be calculated using weights of evidence and weighted logistic regression, a geologic map, and the location of significant deposits. Weights of evidence and weighted logistic regression applied to a global geologic map could provide quickly a useful reconnaissance definition of tracts for mineral assessment that is tied to the data and is reproducible.