An Evaluation of Spatial Data and Analysis for Identifying Potentially Favorable Areas for Manual Well Drilling: Zinder Region of Niger
AuthorThomas, Sean Allan
AdvisorThomas, James M.
StatisticsView Usage Statistics
This thesis evaluated a variety of straightforward spatial data and analysis techniques for identifying potentially favorable areas for manual well drilling in the Zinder region of Niger. A key question was whether environmental variables derived from publically available spatial data had the capacity to augment groundwater depth data for mapping these potentially favorable areas. Some variables considered were: a new calculation of vegetation persistence derived from Moderate Resolution Imaging Spectroradiometer (MODIS) Enhanced Vegetation Index (EVI) data, MODIS night land surface temperature, and lineament properties, topographic convergence index, and landforms derived from the Shuttle Radar Topography Mission Digital Elevation Model (SRTM DEM). Regression tree analysis showed that geology and soils were the strongest variables for predicting groundwater depth in the study area. The results indicated and parsimony dictates that a geology map and adequate groundwater data are sufficient to map favorable areas for manual well drilling. However, the regression tree analysis also revealed that the combination of relatively high vegetation persistence and low night land surface temperature were related to shallow groundwater depth and can improve favorability mapping for manual well drilling. Additional research is needed to describe these relationships further. Among the output was a procedural outline for favorability mapping, which uses common hydrogeologic and terrain criteria to differentiate between topography and recharge controlled water tables, to direct the choice of variables used in future mapping efforts. Ultimately, several maps of favorable areas for manual well drilling for the Zinder region were created using geology, groundwater depth, and threshold values of environmental variables from the regression tree analysis.