Long-Term Retrospective Predicted Concentration of PM<sub>2.5</sub> in Upper Northern Thailand Using Machine Learning Models

Long-Term Retrospective Predicted Concentration of PM<sub>2.5</sub> in Upper Northern Thailand Using Machine Learning Models

This study aims to build, for the first time, a model that uses a machine learning (ML) approach to predict long-term retrospective PM<sub>2.5</sub> concentrations in upper northern Thailand, a region impacted by biomass burning and transboundary pollution. The dataset includes PM<sub...

Full description

Saved in:
Bibliographic Details
Main Authors: Sawaeng Kawichai, Patumrat Sripan, Amaraporn Rerkasem, Kittipan Rerkasem, Worawut Srisukkham
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Toxics
Subjects:
PM<sub>2.5</sub> prediction
retrospective prediction
long-term prediction
machine learning
fire hotspots
Online Access:https://www.mdpi.com/2305-6304/13/3/170
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This study aims to build, for the first time, a model that uses a machine learning (ML) approach to predict long-term retrospective PM<sub>2.5</sub> concentrations in upper northern Thailand, a region impacted by biomass burning and transboundary pollution. The dataset includes PM<sub>10</sub> levels, fire hotspots, and critical meteorological data from 1 January 2011 to 31 December 2020. ML techniques, namely multi-layer perceptron neural network (MLP), support vector machine (SVM), multiple linear regression (MLR), decision tree (DT), and random forests (RF), were used to construct the prediction models. The best ML prediction model was selected considering root mean square error (RMSE), mean prediction error (MPE), relative prediction error (RPE) (the lower, the better), and coefficient of determination (R<sup>2</sup>) (the bigger, the better). Our study found that the ML model-based RF technique using PM<sub>10</sub>, CO<sub>2</sub>, O<sub>3</sub>, fire hotspots, air pressure, rainfall, relative humidity, temperature, wind direction, and wind speed performs the best when predicting the concentration of PM<sub>2.5</sub> with an RMSE of 6.82 µg/m<sup>3</sup>, MPE of 4.33 µg/m<sup>3</sup>, RPE of 22.50%, and R<sup>2</sup> of 0.93. The RF prediction model of PM<sub>2.5</sub> used in this research could support further studies of the long-term effects of PM<sub>2.5</sub> concentration on human health and related issues.
ISSN:2305-6304

Similar Items