A new assessment method for the vulnerability of confined water: W-F&PNN method

At present, groundwater vulnerability assessment is mainly aimed at phreatic aquifer, but in most areas, people draw water from confined aquifer, so it is necessary to assess the vulnerability of confined water. As the existing groundwater vulnerability assessment methods are greatly influenced by subjectivity. The purpose of this study is to put forward W-F&PNN method to avoid subjectivity. Weber-Fechner (W-F) law in psychophysics and Probabilistic Neural Network (PNN) are combined in this paper. W-F law is a theory for describing the people's response to a stimulus, which is used to calculate the cluster center and determine the assessment standard; Probabilistic Neural Network is a widely used algorithm for classification, which is used to classify the vulnerability of confined water. According to the influence factors of confined water, the assessment parameters D(3)A(2)C(2) were determined, and the vulnerability of confined water in plain area of Xi'an city in 2005 was assessed by W-F&PNN method and were divided into four categories. The vulnerability of confined water gradually reduces from the flood plain, the first terrace, the second terrace, the third terrace, the alluvial fan, to the loess tableland. Results of vulnerability assessment were validated by nitrate concentration in 2016 over the area. The results show that the distribution of vulnerability region is mostly consistent with that of nitrate concentration region. The W-F&PNN method does not assign rate, but calculate the cluster center; and not need to assign weight for parameters, avoiding subjectivity, is a reliable method for groundwater vulnerability assessment and suitable for the area with insufficient data.