Spatial distribution of extreme ground-level ozone (O3) in Peninsular Malaysia using stationary and nonstationary generalized extreme value (GeV) models

• Main Authors: Zakaria, Siti Aisyah Binti, Mohd Amin, Nor Azrita Binti, Radi, Noor Fadhilah Ahmad, Noor, Aishah Binti Mohd
• Format: Conference Proceeding
• Language: English
• Subjects: Air monitoring; Air pollution; Air quality; Extreme value theory; Extreme values; Goodness of fit; Outdoor air quality; Ozone; Parameters; Spatial distribution; Statistical tests
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0224370

Ground-level ozone (O3) is a major air pollutant that can have significant impacts on human health, ecosystems well-being, and agricultural productivity. This study aims to map the spatial distribution of extreme O3 in Peninsular Malaysia using stationary and nonstationary Generalized Extreme Value (GEV) models. The models are applied to air quality data collected from 24 air monitoring stations across the region between 2000 and 2016. The stationary GEV model assumes that the distribution of extreme O3 values is constant for all parameters while the nonstationary GEV model allows for cyclic effect on location parameters to capture trends or changes in the underlying distribution while other parameters remain constant. The results show that both stationary and nonstationary GEV models perform well in terms of goodness-of-fit statistics using probability plotting method. Maps generated by the stationary and nonstationary GEV model reveal significant spatial variation in extreme O3 concentrations across the region, with hotspots in urban areas and near major industrial facilities. The findings provide important information for policymakers and other stakeholders working to mitigate the impacts of air pollution in Peninsular Malaysia and demonstrate the use of extreme value theory techniques in modelling spatial distribution of extreme environmental events.