Remote Sensing and Gis for Groundwater Assessment in Hard Rock Areas. Applications to Water Well Siting in Ghana and Botswana

Doktorsavhandling, 1996

Different types of remote sensing data have been evaluated in this thesis with respect to their potential use in groundwater assessment in hard rock areas. The primary objectives of the study have been (1) to evaluate the interpretation of groundwater indicators from high resolution satellite images and aerial photography and (2) to investigate how remote sensing interpretations subsequently can be integrated with other pertinent data, as a decision support in water well siting under difficult hydrogeological conditions. Two areas of low-grade metamorphic and crystalline rocks in Ghana and Botswana have been the focus of the study. The study has included a general field evaluation of various types of optical remote sensing imagery, with special emphasis on lineaments, vegetation, drainage and bedrock characteristics. The results indicate that a top-down approach using multiple sources of remote sensing data is beneficial in groundwater assessment, and that there are clear advantages in using digital image data in preference to paper prints. Remote sensing interpretations have been integrated with geological and topographical maps, borehole data, and geophysical investigations in a geographical information system (GIS). A number of spatial analyses have been performed to identify relationships that will prove useful to groundwater exploration strategies. Correlation studies of lineaments and well capacities have not been conclusive in the study areas, but have demonstrated the need for improved lineament descriptors in groundwater exploration. Lineaments classed by inferred hydrogeological significance has, between several interpreters, shown a good reproducibility of the most significant lineaments. The confidence in these features being important to groundwater availability, is thereby increased. A probabilistic approach to groundwater exploration has been evaluated in the study. The approach incorporates professional judgment in a Bayesian framework and is based on a number of different datasets indicating groundwater occurrence. The results clearly show that the approach is promising, as it restricts detailed field investigations to the most favourable targets. However, improved methods for assessing the fracture zone characteristics in hard rock areas are needed. Data integrated in a GIS has been found an ideal decision support, for a time and cost effective selection of well sites or target areas. Errors and positional accuracy of the data should, however, be carefully considered, not to yield meaningless analysis results in the GIS. The use of the global positioning system (GPS) has been very useful in limiting positional inaccuracies in borehole and image data.