An integrated remote sensing investigation of the Alto Paranaiba kimberlite province, Minas Gerais, Brazil

Arvind Chaturvedi, Purdue University

Abstract

This thesis presents an integrated exploration model, using Landsat Thematic Mapper (TM) and aeromagnetic data, for better detection and delineation of kimberlite-lamproite diatremes in vegetated areas. The model has been developed and tested for the Alto Paranaiba province which is noted for its occurrences of kimberlite, lamproite, and alluvial diamonds. The exploration model assumes that kimberlite-lamproite diatremes have certain characteristics. First, these diatremes occur in highly fractured areas. Second, in areas with deep weathering profiles, red-colored soils frequently develop over them. And, third, these rocks have significant magnetite content. The objective of the exploration model then is to identify these characteristics using different data sets and methodologies. In the initial stage, delineation of areas with higher incidence of fractures was performed through spatial analysis of a Landsat TM image. Areas with lower incidence of fractures were thus eliminated from further consideration. For the accurate detection and mapping of red-colored, iron oxide-rich soils that may be indicative of kimberlite-lamproite diatremes, spectral analysis of Landsat TM digital data was accomplished using a new procedure, the redness transformation, developed in this work. The procedure utilizes the base-line approach and principal components analysis to produce transformed data that are more amenable to clustering techniques than the original data. Red-colored soil anomalies thus detected are then investigated further using aeromagnetic data. Aeromagnetic data are used to detect sources with high magnetite content that may be related to kimberlite-lamproite diatremes. In order to define approximate source outlines, reduction-to-the-pole of low-latitude aeromagnetic anomalies that exhibit remanent magnetization is accomplished using a procedure, developed in this work, that incorporates the damped least-squares equivalent point source (EPS) inversion technique. Constraints on the geometry and physical properties of the source can then also be placed using a three-dimensional modeling algorithm. The results from the digital processing of Landsat TM and aeromagnetic data are then integrated based on the convergence of evidence approach. The exploration model proposed here has been successfully tested over the known Japecanga diatreme. Positive results obtained for this diatreme are further corroborated by field evidence provided by others.

Degree

Ph.D.

Advisors

Meyer, Purdue University.

Subject Area

Geology|Geophysics|Remote sensing

COinS