Set pair analysis-Markov chain model for groundwater quality assessment and prediction: A case study of Xi’an city, China

Abstract

This study reported a set pair analysis (SPA)-Markov chain model for groundwater quality assessment and prediction. This model combines the SPA concept and the Markov chain theory. Xi’an City was taken as a case study area, and groundwater quality monitoring data from 1996 to 2015 were used as an example to justify the feasibility of the SPA-Markov model. SPA was used to determine the groundwater quality grades, and entropy weighted Markov chain was used in groundwater quality prediction based on the weighted average connection number calculated by SPA. The assessment results indicate that groundwater quality in the study area is very poor in the year from 1996 to 1998, and it becomes improved after 1998. The prediction results show that the groundwater quality in 2014 and 2015 is both classified into the good quality grade, which is consistent with the actual groundwater quality grade. The proposed model is able to treat uncertainties in groundwater quality assessment and is effective in short term groundwater quality prediction.