Analysis and prediction of spatial data using GIS statistical analysis software 'Geostatistical Analyst': A case study of MTBE contamination in groundwater at the city of Temecula, Riverside County, California

MTBE has been classified by U.S. Environmental Protection Agency (USEPA) as a possible human carcinogen. It is highly soluble in water and widely spread into surface water and groundwater resources.; This thesis utilizes the Geostatistical Analyst software to perform statistical analysis of MTBE concentration data and map its distribution in groundwater in the City of Temecula. It first explores the spatial data distribution and autocorrelationship of the dataset, and then it uses a kriging method to conduct the structural data analysis. The validation function is used to assess the accuracy of the prediction. Finally a prediction map of MTBE concentrations is generated for the study area. A probability map is created to evaluate the probability that groundwater MTBE concentration exceeds a predefined threshold at any location in the subject area.; The thesis concludes that the Geostatistical Analyst is well developed and has great potential for analyzing, predicting and mapping spatial data.