Geostatistical analysis for spatially characterizing hydrochemical features of springs in Taiwan

Abstract

The Taiwan Island is abundant in spring resources. Many industries of spring spas highlight unique hydrochemical characteristics of springs because consumers frequently consider bathing in all kinds of natural springs with therapeutic and relaxing functions. This work adopted geostatistical approaches, ordinary kriging (OK) and sequential Gaussian simulation (sGs), to spatially explore hydrochemical features of springs in Taiwan and establish classifications of springwater. First, OK and sGs were used to determine chloride, sulfate and bicarbonate distributions of springs emerging from metamorphic and sedimentary rocks. Then, ratios of chloride, sulfate and bicarbonate equivalents per liter were calculated based on the determined distributions. Finally, various combinations of chloride, sulfate and bicarbonate were adopted to spatially characterize classifications of springwater. This research result indicates that OK and sGs are capable of spatially characterizing the hydrochemical features and classification uncertainty of springs. However, the smooth effect of OK underestimates high values and overestimates low values, reducing the difference among Cl−, SO42− and HCO3− in springwater. Meanwhile, sGs realizations not only present high spatial heterogeneity, but enhance the influence of dominant hydrochemical features on classifications of springwater. Furthermore, springwater emerging from metamorphic and sedimentary rock areas in Taiwan can be divided into eight categories based on combinations of determined chloride, sulfate and bicarbonate. The principal HCO3− water is the major springwater category. The subordinate springwater types include the principal HCO3− water with Cl− or HCO3− and the principal Cl− water with HCO3−.