Prediction of groundwater quality in the east of Thailand with discriminant analysis method

Abstract

The 316 samples of groundwater wells in the east of Thailand were analyzed for consumption purpose. Concentrations of 21 various chemical parameters were considered to predict and then evaluate the quality of groundwater with discriminant analysis method and correct classification rate (CCR), respectively. The study results found stepwise discriminant analysis determined 5 significant parameters, Manganese (Mn), Copper (Cu), Iron (Fe), Fluorine (F) and Carbonate (CO₃), which discriminated the groundwater quality into 2 groups; nonstandard and standard quality groups. Derived discriminant model was capable correctly classified the group of groundwater quality with a rather high value of average CCR (78.80%) for the validation data set.