University of Twente Student Theses


Modeling of prospectivity for Lode-Au mineralization in the Kibaran of Uganda

Msechu, M.E. (2011) Modeling of prospectivity for Lode-Au mineralization in the Kibaran of Uganda.

[img] PDF
Abstract:Accurate prediction of potential zones for the occurrence of new mineral deposits requires an understanding of geological processes that controlled the existing deposits. Understanding these geological controls means understanding the geological processes that were active during the formation and preservation of the deposits sought. Quartz-vein hosted gold mineralization, typical of orogenic gold deposit model occurs in the south-western part of Uganda in which the study area lies. Geologically, it is characterized by deformed and metamorphosed Karagwe Ankole Belt (KAB) lithologies consisting predominantly of biotite gneisses, amphibolites, intebedded schist, metasediments, granites and mafic volcanics. In order to understand geological processes that are conceived to control mineralization, airborne magnetics and radiometric methods were employed. In addition, digital elevation model (DEM) outlined several surficial deformation features that were in turn used to infer structural conduits related to transportation and preservations of gold bearing fluids. Interpretations of these structural features were to some extent supported by the presence of existing ground structural measurements. Otherwise, standard procedures used to interpret magnetic and radiometric datasets were used. The results of this study show that the indicative geological features/processes that control mineralization include 1) age of mineralization (Neoproterozoic) 2), proximity to NW-trending faults/shear airborne 3) proximity to NE-trending faults/shear and 4) proximity to anticlines. On the other hand, the results of the analyses of magnetic and radiometric datasets were further used to study and understand lithological variations of existing lithological units. Contrasts within the magnetic dataset are related to either variations within similar/different lithological units or major structural sutures. This understanding led to the identification of new deformation zones and lithological units and subsequently added value to the geological map of the area. Knowledge-guided data-driven wildcat modeling of mineral prospectivity was used to generate predictor maps independent of existing deposits but based on improved wildcat predictor scores through 5- percentile intervals of distances to the above mentioned evidential features (except for age of mineralization). The resultant scores maps were later integrated using principal component analysis in order to generate prospectivity map of the study area. The prospectivity map was further re-classified based on 5-percentile classes and pixels above 85 percentile, i.e. mean+1SD were considered prospective. Of the total 10 mineral deposits tested, 50% were correctly predicted. In addition, 20% of the area tested, i.e. 132265 out of 661085 pixels, is considered prospective. These prospective zones can be used as potential corridors for strategic mineral exploration programmes of the area.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page