Integrated analysis of Landsat TM,airborne magnetic,and radiometric data,as an exploration tool for granite-related mineralization,Salamanca province,Western Spain |
| |
| Authors: | Marc A Goossens |
| |
| Institution: | 1. Geoconnaissance Consulting, Iepstraat 3, 3581 LK, Utrecht, The Netherlands
|
| |
| Abstract: | For successful application of remote sensing and data integration in regional mineral exploration, it is crucial that, prior to regional application, relations between geological setting, mineralizing processes, related anomalies, and available remote sensing data are thoroughly studied in a calibration area with known mineralization that is representative for the region of interest. Recognizing features that are diagnostic for mineralization and its setting using Landsat Thematic Mapper (TM) and airborne geophysical data is complicated for the following reasons: (1) The basic relation between geology and the original data is often obscured by complex data manipulation, obstructing the understanding of the final output. (2) The risk of removing potentially important information during the data interpretation is considerable. (3) The decision criteria with respect to the correctness of classifications are often arbitrary. (4) Error and quality assessments are often subjective and not reproducible under different circumstances. A data integration study is presented for a 20-km×20-km calibration area in Spain with known granite-related mineralization. The contact metamorphic setting is diagnostic for most mineral deposits. Therefore, TM and airborne data are used to map features predictive for contact metamorphic rocks. A concept of spatial reasoning is presented that reduces the damage caused by these problems: (1) The data manipulation is very straightforward and aimed at recognizing specific features known to be diagnostic for the setting in which mineralization is likely to occur. (2) Every step in the data manipulation is described in a quantitative way. (3) The nature of the pixels surrounding a classified pixel is considered when deciding whether the classification is correct. (4) Weights are assigned to the degree of uncertainty within an interpretation. (5) Integrating the interpreted and weighed data into a probability map highlights zones that are confirmed by all data sets, and thus are most reliably classified. This concept enables users of a careful and systematic analysis of the multiple spatial data sets to detect diagnostic features that are predictive for granite-related mineralization in this region. |
| |
| Keywords: | |
| 本文献已被 SpringerLink 等数据库收录！ |
|
