GEOSCAN Search Results: Fastlink


TitleApplication of a Microcomputer-based Geographic Information System to Mineral - Potential Mapping
AuthorBonham-Carter, G F; Agterberg, F P
SourceMicrocomputer Applications in Geology, II; by Hanley, J T (ed.); Merriam, D F (ed.); Computers and Geology vol. 5, 1986 p. 49-74,
Alt SeriesGeological Survey of Canada, Contribution Series 47488
Mediapaper; on-line; digital
File formatpdf
ProvinceNova Scotia
NTS11F/SW; 11E/NE; 11D/15; 11D/16
AreaNortheastern Nova Scotia
Lat/Long WENS -63.0000 -61.0000 45.5000 44.7500
Subjectsmathematical and computational geology; structural geology; computer applications; mineral potential; mineralization; gold; lineations; faults; folds; structural interpretations; structural controls; structural features; statistical methods; Goldenville Formation; Halifax Formation; Horton Group; Meguma Terrane; Paleozoic
Illustrationsformulae; sketch maps
ProgramCanada-Nova Scotia Mineral Development Agreement, 1984-1989
AbstractRecent advances in technology and methods of integrating map data are making the task of mapping mineral potential by computer more appealing to the practicing geologist. Micro-based Geographic Information Systems (GIS) make the task of building a database from diverse maps relatively straightforward, even for those who are not computer specialists. Efficient commercial software packages, hierarchical data structures such as quadtrees, low-cost disk storage, and faster micros make multiple map analysis by computer easier than manual procedures, such that the light table for overlaying maps which may soon become a thing of the past. A recent development in methodology, termed here "weights of evidence" mapping, uses an approach for combining evidence from several predictor maps to estimate probability of mineral occurrence. Not only does the GIS aid in capturing and coregistering the input maps for this task, it also facilitates the modeling and visualization of results. In the simplest situation, each input map is converted to binary form, and a pair of weights, W+ and W-, are determined corresponding to the two map classes. The posterior log odds of a mineral occurrence within a unit cell is calculated by adding the weights of evidence (W+ and W-) from the input maps to the prior log odds. The weights of evidence are determined using the distribution of known mineral occurrences with respect to the input maps. The method assumes that the input maps are conditionally independent of one another with respect to the points, and tests are made to verify this assumption. The uncertainty of the probability estimates, because of both the variances of the weights of evidence and incomplete or missing data, also can be mapped. A small worked example illustrates the calculations involved. The method also is applied to gold potential mapping in Nova Scotia. Maps showing rock type, regional geochemistry, proximity to structures, formation contacts, and fold axes are used to predict gold mineralization. The proximity maps are generated by dilating linear features, a useful function of GIS's for mineralization modeling. Hardcopy of the output maps can either be made by photographing the color monitor or by using raster plotting devices. The GIS map-query function provides a powerful tool to explore the output maps interactively with the cursor, simultaneously interrogating the "map stack" to aid in understanding the relationships between input and output.

Date modified: