An improved fuzzy comprehensive evaluation of water quality in karst aquifer based on level characteristic values

The aim of this study is to improve traditional fuzzy logic approach to water quality evaluation, due to some drawbacks such as unrepresentative evaluation indexes for water pollution assessment, indeterminate evaluation results and obscure grade boundaries for fuzzy evaluation. The first layer of karst groundwater in Liuzhi special district Weining-Langdai folding area was selected as the modelling object. Based on the karst water quality analysis, factors for the fuzzy logic process were selected from those analysed constituents which were in excess of relevant water quality regulations, from which establishing the evaluation index system which includes physical, chemical, and microbiological factors, and modifying the method of traditional fuzzy comprehensive evaluation. Consequently, results of the new model were assessed and compared with those from the methods of both comprehensive pollution index and traditional fuzzy comprehensive evaluation. Results of the fuzzy comprehensive evaluation show that water quality reaches class Ⅲ in light of national drinking water standard in Liuzhi special district. However, there are still 71% sampling points where the concentrations of ammonia nitrogen (NH4+), fluoride (F-), permanganate index (CODMn) and total dissolved solids (TDS) exceed the water class; and these points mostly occur in the strata of the middle and the late Triassic. On the other hand, the groundwater of 57.1% sampling points falls in class Ⅰ, resulted from previous method of fuzzy comprehensive evaluation, which is not in accord with the reality. In addition, through the fuzzy comprehensive evaluation calibrated with level characteristic values, the water quality of 57.1% and 28.6% points fall within class Ⅱ and class Ⅲ, respectively, of which the standard deviation is 0.011 lower than that from the comprehensive index method which stresses much on weight factor for the element with highest concentration. Therefore, current method for comprehensive evaluation of karst water quality can effectively reflect overall water quality level, water chemical anomalies and quantitative quality classification. It is trusted that the evaluation results are reliable.