Abstract:
Under field conditions modern digital conductivity meters give standardized, rapid and reproducible measurements. Here we investigate the accuracy of their estimates of the composition of karst waters, as total hardness (TH, as mg/L CaCO3) for limestone and dolomite. These are the fundamental measures of process in carbonate karst geomorphology. PHREEQC theoretical curves for the dissolution of pure calcite/aragonite and dolomite in water at 25 degrees C are compared with water analyses from karst studies worldwide. Other principal ions encountered are sulphates, nitrates and chlorides (the 'SNC' group). From carbonate karsts, 2309 spring, well and stream samples were divided into uncontaminated (SNC < 10%), moderately contaminated (10 < SNC < 20%), and contaminated (SNC > 20%) classes. Where specific conductivity (SpC) is less than 600 mu S/cm, a clear statistical distinction can be drawn between waters having little contamination and substantially contaminated waters with SNC > 20%. As sometimes claimed in manufacturers' literature, in 'clean' limestone waters TH is close to 1/2SpC, with a standard error of 2-3 mg/L. The slope of the best-fit line for 1949 samples covering all SNC classes where SpC < 600 mu S/cm is 1.86, very close to the 1.88 obtained for clean limestone waters; however, the value of the intercept is ten times higher. The regression line for clean limestone waters where SpC > 600 mu S/cm helps to distinguish polluted waters from clean waters with possible endogenic sources of CO2. In the range 250 < SpC < 600 mu S/cm, dolomite waters can be readily distinguished from limestone waters. Copyright (C) 2005 John Wiley & Sons, Ltd