The Impact of Earthquakes on Air Pollution: Machine Learning-Based Air Quality Analysis Following the Kahramanmaraş Earthquakes


Eryilmaz F., Güzel M., Şimşek M. U., KÖK İ., ÖZDEMİR S.

1st Mediterranean Smart Cities Conference, MSCC 2024, Martil - Tetuan, Fas, 2 - 04 Mayıs 2024 identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/mscc62288.2024.10697006
  • Basıldığı Şehir: Martil - Tetuan
  • Basıldığı Ülke: Fas
  • Anahtar Kelimeler: Air pollution, air quality, earthquake effects, time series prediction
  • Ankara Üniversitesi Adresli: Evet

Özet

Air pollution is a significant environmental and societal problem that affects human health, wildlife, and ecological balance. While human activities such as industry, agriculture, traffic, and energy production are major contributors to air pollution, natural causes such as volcanic activities, forest fires, desert dust, and vegetation disruption also play a role. In this paper, we explore whether earthquakes can be considered natural factors contributing to air pollution. Specifically, we examine the effects of the magnitude 7.7 and 7.6 earthquakes centered in Kahramanmaraş that occurred on February 6, 2023, in Turkey. We collected data for the earthquake-affected province of Malatya for the five years before and 84 days after seismic events. We developed pollution prediction models by training artificial intelligence algorithms such as LSTM(Long Short-Term Memory), Facebook Prophet, and SVR(Support Vector Regression) on the collected data. These models aim to predict the expected normal air pollution levels in the absence of an earthquake. Then, post-earthquake air pollution measurements are taken and the predictions obtained from the developed models are compared with the actual measurements. The experimental results reveal a dramatic increase in air pollution, especially in the first month after the earthquake. This research contributes to our understanding of the complex interactions between seismic activity and air quality and emphasizes the importance of considering natural factors in the assessment of air pollution.