Multiple Data Layer Modeling and Analysis in Assessments

Abstract

Geoscientific information is essential for assessing not only mineral resource favorability, but also geoenvironmental hazard potential. The health of any ecosystem is directly related to its underlying geology. Within every watershed, concentrations of metals and other elements in ground and surface waters, sediments, soils, plants, many animals, and humans are intimately connected to the geologic “landscape” of the area. Comprehensive, unbiased geoscientific information is therefore fundamentally and critically important in effectively assessing current, as well as past and future impacts on the environment caused by natural and/or anthropogenic changes.

A multidisciplinary team of U.S. Geological Survey (USGS) scientists was assembled to study the geoenvironmental assessment process and to develop applications for the State of Montana, north-central USA. Areas of expertise represented in the team include geology, geochemistry, remote sensing, and geophysics. A variety of “layers” of digital information related to these various disciplines was compiled for the project.

Several questions typically arise in assessment projects that involve numerous information components: (1) what categories of which data are relevant to the study objectives; (2) what relative importance (weights) should appropriately be ascribed to the data layers; and (3) how should the layers be combined to effectively achieve the objectives of the study. The Montana project has relied heavily on geographic information systems (GIS) applications to attempt to address these questions in pursuit of the objective of mapping geoenvironmental potential for acidic, metal-rich drainage. However, methodologies utilized in this project could effectively be applied to other geoenvironmental questions or for assessing potential for mineral resources.

Mining districts within the State were characterized according to their acid generating and acid buffering potentials. These properties were assigned according to considerations of abundance of sulfides and of calcium carbonate in the mining areas. Those areas that were characterized as high in acid generating potential and low to moderate in acid buffering capacity were chosen as prototypical of areas to be identified in the geoenvironmental assessment. Various classes of each information layer were considered as “candidates” for inclusion in the overall assessment model and were tested for significance of spatial association with the reference, or prototype, areas. This was accomplished using the GIS to calculate the “probability ratio” of each class, representing the relative likelihood of finding the class within the reference areas versus elsewhere in the study area. These computations not only guided determination of which data layers were relevant to the assessment, but also what weights should appropriately be assigned to them. The resulting weighted “submodels” were combined, using various methods, to produce composite models and derivative maps showing relative potential for acidic, metal-rich drainage.