Risk-Based Optimal Ranking of Air Quality Monitoring Stations in a Fuzzy Environment: A Case Study

Abstract

Air Pollution can certainly be associated with increased levels of greenhouse gases in the atmosphere, health effects and climate change. Therefore, evaluation of ambient air quality, associated with uncertainty in the form of fuzziness (imprecision), and randomness assumes significant importance in the changing environment scenario. The methods used for describing air quality, i.e. Conventional Air Quality Index (CAQI) and also newly developed fuzzy logic-based Zadeh-Deshpande (ZD) formalism consider only a few air quality criteria pollutants. However, there are other constraints which could be imprecise/fuzzy. For example there could be several Air Quality Monitoring Stations (AQMS) in a city but with varying degree of certainty expressed in numeric terms using ZD formalism. In this paper, we demonstrate the application of a combination of ZD and Bellman–Zadeh (BZ) method in final risk-based optimal ranking of AQMS which is basically a complex multi-constraint decision-making problem. Furthermore, classification of twelve cities in Maharashtra State, India into different air quality clusters to demonstrate the importance of the concept of supervised learning followed by unsupervised learning using reference level concept is presented.