Research trends in spatial modeling of PM2.5 concentration using machine learning: a bibliometric review

Aug 3, 2026

Spatial modeling supports mapping research variables, including particulate matter 2.5 (PM2.5) concentrations across geographic areas. This study analyzes trends in applying machine learning to spatial PM2.5 modeling through bibliometric review methods. Publications indexed in Scopus from 2014–2023 form a corpus of 335 articles. Co-authorship and co-occurrence analyses are performed using VOSviewer, showing steadily increasing research, strongest productivity in China, a multidisciplinary environmental-science focus, and a very high collaboration rate (0.98).

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Introduction

PM2.5 background and health impact

Spatial modeling needs and limitations of monitoring networks

Prior spatial modeling approaches

Move toward machine learning and ensemble methods

Indonesian Journal of Electrical Engineering and Computer Science Vol. 37, No. 2, February 2025, pp. 1317~ 1327 ISSN: 2502-4752, DOI: 10. 11591/ijeecs.v37 . i2 .pp1317-1327 􀂈 1317 | Research trends in spatial modeling of PM2.5 concentration using machine learning: a bibliometric review
Retno Tri Wahyuni1,2, Dirman Hanafi1, M. Razali Tomari1, Dadang Syarif Sihabudin Sahid3
1Faculty of Electrical and Electronic Engineering, Universiti Tun Hussein Onn Malaysia, Batu Pahat, Malaysia 2Department of Industrial Technology, Politenik Caltex Riau, Pekanbaru, Indonesia
3Department of Information Technology, Politenik Caltex Riau, Pekanbaru, Indonesia | | | --- | --- | | Article Info
Article history:
Received Jun 5, 2024 Revised Sep 24, 2024 Accepted Sep 30, 2024
Keywords:
Bibliometric Machine learning PM2.5
Spatial modeling VOSViewer
Corresponding Author: | ABSTRACT
Spatial modeling is commonly used to map research variables, including particulate matter 2.5 (PM2.5) concentrations, in specific areas. The article that surveys publications on the application of machine learning in spatial modeling of PM2.5 using bibliometric methods has not been identified yet. This paper aims to analyze trends in applying machine learning in the spatial modeling of PM2.5 using bibliometric methods. The review was conducted on publications indexed in the Scopus database over the decade (2014–2023) comprising 335 articles. The analysis included co-authorship and cooccurrence using VOSviewer. From the two stages of analysis, it can be concluded that research on this topic has constantly increased over the past 10 years, with the highest productivity coming from researchers in China. This research topic is multidisciplinary, with most publications appearing in environmental science. The research also shows a very high collaboration rate of 0.98. A deeper examination of the keywords reveals the most commonly used machine learning techniques by researchers. The random forest method is the most frequently found in the analyzed documents, followed by deep learning, long short-term memory (LSTM), extreme gradient boosting (XGBoost), and ensemble model.
This is an open access article under the CC BY-SA license.
| | Retno Tri Wahyuni
Department of Industrial Technology, Politeknik Caltex Riau Pekanbaru, Indonesia
Email: [retnotri@pcr.ac.id](retnotri@pcr.ac.id) | | 1. INTRODUCTION Particulate matter 2.5 (PM2.5) concentration describes the amount of fine particulate aerosol particles with diameters up to 2.5 microns produced by various sources, which can result in different chemical compositions and physical characteristics. Among the prevalent components found within PM2.5 are sulfates, nitrates, black carbon, and ammonium, which collectively form a significant portion of these particles [1] . PM2.5 can be generated from anthropogenic activities such as traffic emissions, industrial processes, agricultural practices, and natural sources such as dust storms, sandstorms, and wildfires. PM2.5 is one of the dangerous pollutants whose concentration is monitored and regulated as an air quality standard by the World Health Organization (WHO), along with five other parameters such as PM10, ozone (O3), nitrogen dioxide (NO2), dioxide sulfur (SO2), and carbon monoxide (CO) . Several studies discuss the impact of PM2.5 on health and causes of mortality. As a tiny particulate matter, PM2.5 hurts human health. PM2.5 can penetrate the lungs and bloodstream without being filtered, causing respiratory diseases [2]􀀐[4] . Several epidemiological studies have proven that PM2.5 increases mortality and morbidity rates [5]􀀐[7] . Spatial modeling is widely applied in various fields, one of which can be found in the modeling concentration distribution of pollutants such as PM2.5 . One background for developing spatial modeling for the distribution of pollutant concentrations in an area is the limited number of air regulatory monito

FAQs

What is the main objective of this bibliometric review?

To analyze trends in using machine learning for spatial modeling of PM2.5 concentration through bibliometric methods over 2014–2023.

Which database and time range were used for the publication dataset?

The review used publications indexed in the Scopus database spanning a decade from 2014 to 2023.

How was research collaboration and keyword structure analyzed?

Co-authorship and co-occurrence were analyzed with VOSviewer, and the keywords were examined to identify commonly used machine learning techniques.

Aug 3, 2026

Upload by Clementine