Extreme Value Analysis of PM10 Concentration in Thailand

Authors

  • Kuntalee Chaisee Data Science Research Center, Department of Statistics, Faculty of Science, Chiang Mai University, Chiang Mai, Thailand
  • Kamonrat Suphawan Data Science Research Center, Department of Statistics, Faculty of Science, Chiang Mai University, Chiang Mai, Thailand

Keywords:

Extreme values, air pollution, PM10, GPD, POT, return levels

Abstract

This research aims to analyze the extreme values of the air pollutants, in particular, PM10 concentration in Thailand. Due to the limitation of data, we restrict our attention to 23 air quality monitoring stations in Thailand. The daily PM10 concentration data from 2008 to 2019 are used to analyze and are divided into two types; 24-hour averages and daily maxima. The Peak Over Threshold (POT) approach is used to assess the risk of air pollutants; hence the Generalized Pareto Distribution (GPD) is used to fit the data. One of the challenging issues in POT is the choice of threshold. In this work, we combine the mean residual life plot and the goodness of fit test methods to determine the threshold. The maximum likelihood estimation and the bootstrap method are used to deal with parameter estimation in GPD and uncertainty quantification. We then estimate the return levels, which present extreme predictive events in terms of the values expected to exceed average once every return period. The results show that daily PM10 concentration at station 24t in Saraburi, 73t in Chiang Rai, and 36t in Chiang Mai have very high predictive extreme values. Many stations located in the north of Thailand also have relatively high levels. Consequently, the northern region is most likely to encounter high exposures to PM10.

Downloads

Download data is not yet available.

Downloads

Published

2021-06-29

Issue

Section

Articles