Revealing information and equipment redundancies in air pollution monitoring networks in Turkey

Abstract

Air pollution monitoring networks are the primary tools for measuring, managing and assessing air quality. However, these networks need considerable financial resources due to expensive devices and analyses, as well as such issues as the likely redundancy in the number of samples. The primary objective of this study was to identify possible information and equipment redundancies in Turkish monitoring networks. Thus, it is expected that the results of this study may help reduce air pollution monitoring expenses and increase monitoring efficiency. For this purpose, the Fuzzy C-Means clustering algorithm and time series analyses were used. This study has two novelties. First, this is the first study to be conducted for this purpose in Turkey. Further, Dickey-Fuller test statistics and model parameters have not been used as clustering variables before. Thus, it is expected that both stochastic behavior and concentration levels of PM10 time series will be reflected simultaneously, and similarities among monitoring stations will be better identified.