HANDLING SPATIAL UNCERTAINTY OF GEOGENIC CONTAMINANTS IN WATER SYSTEMS: NEW INSIGHTS INTO FLUORIDE CONTAMINATION IN DRINKING WATER THROUGH GEOSPATIAL METHODS

For decades, researchers in many disciplines have neglected the spatial aspects of data collected on geographical phenomena of interest during the analysis stage. This has been due to lack of computational capabilities and limitation in conventional methods. Consequently, insight gained in the data is not optimal. In this study, we present a novel approach of data analysis by considering the spatial aspects of the data. The case study was on the geogenic fluoride contamination in aquifers of northern Tanzania. The large part of this region lies in the East African Rift Valley (EARV), a volcanically active area, where more than 80% of the households depend solely on the groundwater as a primary source for various purposes including drinking. Due to the consumption of groundwater with fluoride concentration below and above the World Health Organization (WHO) recommended guideline of 1.5 mg/L for a prolonged period, many people in this setting have developed dental caries and fluorosis, respectively. Therefore, we combined spatial statistical methods and Geographical Information Systems (GIS) tools to identify and map potential and significant risk zones of geogenic fluoride contamination. The high-risk zones in terms of dental caries exist around major stratovolcanoes, especially Mt. Kilimanjaro and along the EARV escarpments at high elevations. The high risk zones in terms of fluorosis exist in the lowlands at the foot of major stratovolcanoes, especially Mt. Meru and Mt. Hanang’ in Arusha and Manyara regions, respectively. In addition, another risk zone exists along the local tectonic plate boundary between the recent Pliocene volcanics and the Mozambique Belt. Through spatial models, we established that contact time between fluoride-bearing rocks and groundwater and geothermal waters contributed significantly to the spatial variability of fluoride in water systems. Other factors included topography, water table variability, irrigated agricultural practices and borehole installation practices. This study provides the new insights to the long-term problem using modern geospatial methods. These methods provide room for interacting with the data through incorporation of the neighborhood of the observed event on the environmental variable of interest.