The conventional geo-accumulation index (GI) cannot deal with the uncertainty in the geochemical background information of pollutants or conduct a synthetic evaluation of heavy metal pollution in sediments. To solve these problems, a geo-accumulation vector (GV) model is designed based on uncertainty analysis theory. GV uses the probability vector to quantify the possibility for the pollution condition of a heavy metal to belong to a certain grade and then uses the first-order moment principle to determine its category. The GV model is applied in this study to evaluate heavy metal pollution in the sediments of Western Dongting Lake. The results show that (i) compared with the conventional GI model, the GV model has better capacity in evaluating synthetic pollution and in dealing with the uncertainty in the geochemical background information of sediments; (ii) the sediments of Li River estuary, Yuan River estuary, Songzi River estuary, and the central part of Western Dongting Lake are "moderately to heavily contaminated," "moderately to heavily contaminated," "moderately contaminated," and "uncontaminated to moderately contaminated" by heavy metals, respectively. These sediments also have 0.19, 0.36, 0.48, and 0.15 possibilities to deteriorate to the next category, respectively; and (iii) all deterioration risks are introduced by cadmium.

JOURNAL OF HYDROLOGY   卷: 573   页: 40-48   出版年: 2019