2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia
| Format: | General Document |
|---|
| _version_ | 1860798343771324416 |
|---|---|
| building | INTELEK Repository |
| collection | Online Access |
| collectionurl | https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 |
| copyright | Copyright©PWB2025 |
| country | Malaysia |
| date | 2025-04-17 |
| format | General Document |
| id | 17380 |
| institution | UniSZA |
| originalfilename | 17380_33b5357921db088.pdf |
| person | Mohd Suzairi Mohd Shafi'i |
| recordtype | oai_dc |
| resourceurl | https://intelek.unisza.edu.my/intelek/pages/view.php?ref=17380 |
| sourcemedia | Server storage Scanned document |
| spelling | 17380 https://intelek.unisza.edu.my/intelek/pages/view.php?ref=17380 https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 General Document Malaysia Library Staff (Top Management) Library Staff (Management) Library Staff (Support) Terengganu English UniSZA East Coast Environmental Research Institute application/pdf 1.7 Malaysia Server storage Scanned document Universiti Sultan Zainal Abidin UniSZA Private Access Universiti Sultan Zainal Abidin 313 Copyright©PWB2025 Sustainable development—Malaysia Air Quality Environmental monitoring—Malaysia Multivariate analysis Environmental statistics Environmetric Techniques Artificial neural networks Machine Learning Dissertations, Academic Principal Component Analysis (PCA) Microsoft® Word LTSC 2025-04-17 Mohd Suzairi Mohd Shafi'i Continuous Air Quality Monitoring (CAQM) Artificial neural networks (ANN) Hierarchical agglomerative cluster analysis (HACA) Air Pollutant Index (API) Air—Pollution—Malaysia—Measurement Air quality—Malaysia—Analysis Data mining—Environmental sciences Risk management—Environmental aspects 2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia This research aims to assess air quality status, classify pollution levels based on spatial patterns, identify key air quality parameters, and propose mitigation strategies to enhance air quality risk management. Eleven years of secondary air quality data (2011–2021) from the Department of Environment, Malaysia, collected from 38 Continuous Air Quality Monitoring (CAQM) stations across Peninsular Malaysia, were analysed. A combination of statistical and machine learning techniques was employed, including univariate statistics, environmetric methods such as hierarchical agglomerative cluster analysis (HACA), discriminant analysis (DA), principal component analysis (PCA), artificial neural networks (ANN), and the DMAIC-Six Sigma methodology. Univariate analysis focused on five air quality parameters (PM10, O₃, CO, SO₂, and NO₂) and the Air Pollutant Index (API) to assess compliance with the Recommended Malaysian Air Quality Guidelines (RMAQG). The maximum recorded concentrations were 504.250 μg/m³ (PM10), 0.068 ppm (O₃), 5.516 ppm (CO), 0.048 ppm (SO₂), and 0.060 ppm (NO₂). Only PM10 exceeded the RMAQG limit but remained below the hazardous threshold of 600 μg/m³. The highest API value recorded was 477, indicating hazardous air quality (API ≥ 301). However, the overall mean concentrations of all parameters and mean API values were within approved limits, suggesting generally good air quality (API: 0–50). Environmetric techniques were applied to analyse eleven years of daily data, where HACA identified spatial air quality patterns by clustering CAQM stations based on API values, resulting in three distinct clusters: High Pollution Cluster (HPC), Moderate Pollution Cluster (MPC), and Low Pollution Cluster (LPC). DA determined the significance of air quality variables with an accuracy of 91.87% (p < 0.0001), while PCA identified pollution sources, revealing variations across different clusters. Further analysis using ANN predicted spatial distribution based on air quality variables, with the ANN model outperforming DA at 99.97% accuracy. A combined PCA/ANN approach using the MLP-FF-ANN receptor model determined pollution sources, where PCA identified three major contributors to air pollution: combustion processes, meteorological factors, and photochemical reactions. Excluding meteorological influences, four key pollutants—NOx, NO, NO₂, and O₃—were identified as primary pollution sources. Two MLP-FF-ANN models were tested with different input variables, with the model containing ten variables providing more accurate predictions (R² = 0.786, RMSE = 9.623) compared to the model with six variables. Results from environmetric techniques and ANN strongly validated four key mitigation strategies derived from the DMAIC-Six Sigma framework: leveraging data analytics, strengthening stakeholder engagement, restructuring air quality monitoring strategies, and enhancing enforcement measures. These strategies were proposed as evidence-based actions to improve air quality risk management. This research demonstrates that environmetric techniques and ANN analysis are highly effective for processing large, complex air quality datasets, offering critical insights into pollutant behaviour, characteristics, and distribution patterns. The findings emphasize the importance of historical data in optimizing sampling strategies and provide valuable insights for policymakers and stakeholders in developing data-driven air quality regulations and management strategies in Malaysia. uuid:19D2476A-CCCB-4F2A-95E7-B40B658239C1 17380_33b5357921db088.pdf Thesis |
| spellingShingle | 2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia |
| state | Terengganu |
| subject | Sustainable development—Malaysia Environmental monitoring—Malaysia Multivariate analysis Environmental statistics Artificial neural networks Dissertations, Academic Air—Pollution—Malaysia—Measurement Air quality—Malaysia—Analysis Data mining—Environmental sciences Risk management—Environmental aspects |
| summary | This research aims to assess air quality status, classify pollution levels based on spatial patterns, identify key air quality parameters, and propose mitigation strategies to enhance air quality risk management. Eleven years of secondary air quality data (2011–2021) from the Department of Environment, Malaysia, collected from 38 Continuous Air Quality Monitoring (CAQM) stations across Peninsular Malaysia, were analysed. A combination of statistical and machine learning techniques was employed, including univariate statistics, environmetric methods such as hierarchical agglomerative cluster analysis (HACA), discriminant analysis (DA), principal component analysis (PCA), artificial neural networks (ANN), and the DMAIC-Six Sigma methodology. Univariate analysis focused on five air quality parameters (PM10, O₃, CO, SO₂, and NO₂) and the Air Pollutant Index (API) to assess compliance with the Recommended Malaysian Air Quality Guidelines (RMAQG). The maximum recorded concentrations were 504.250 μg/m³ (PM10), 0.068 ppm (O₃), 5.516 ppm (CO), 0.048 ppm (SO₂), and 0.060 ppm (NO₂). Only PM10 exceeded the RMAQG limit but remained below the hazardous threshold of 600 μg/m³. The highest API value recorded was 477, indicating hazardous air quality (API ≥ 301). However, the overall mean concentrations of all parameters and mean API values were within approved limits, suggesting generally good air quality (API: 0–50). Environmetric techniques were applied to analyse eleven years of daily data, where HACA identified spatial air quality patterns by clustering CAQM stations based on API values, resulting in three distinct clusters: High Pollution Cluster (HPC), Moderate Pollution Cluster (MPC), and Low Pollution Cluster (LPC). DA determined the significance of air quality variables with an accuracy of 91.87% (p < 0.0001), while PCA identified pollution sources, revealing variations across different clusters. Further analysis using ANN predicted spatial distribution based on air quality variables, with the ANN model outperforming DA at 99.97% accuracy. A combined PCA/ANN approach using the MLP-FF-ANN receptor model determined pollution sources, where PCA identified three major contributors to air pollution: combustion processes, meteorological factors, and photochemical reactions. Excluding meteorological influences, four key pollutants—NOx, NO, NO₂, and O₃—were identified as primary pollution sources. Two MLP-FF-ANN models were tested with different input variables, with the model containing ten variables providing more accurate predictions (R² = 0.786, RMSE = 9.623) compared to the model with six variables. Results from environmetric techniques and ANN strongly validated four key mitigation strategies derived from the DMAIC-Six Sigma framework: leveraging data analytics, strengthening stakeholder engagement, restructuring air quality monitoring strategies, and enhancing enforcement measures. These strategies were proposed as evidence-based actions to improve air quality risk management. This research demonstrates that environmetric techniques and ANN analysis are highly effective for processing large, complex air quality datasets, offering critical insights into pollutant behaviour, characteristics, and distribution patterns. The findings emphasize the importance of historical data in optimizing sampling strategies and provide valuable insights for policymakers and stakeholders in developing data-driven air quality regulations and management strategies in Malaysia. |
| title | 2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia |
| title_full | 2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia |
| title_fullStr | 2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia |
| title_full_unstemmed | 2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia |
| title_short | 2025_Spatial air quality analysis using environmetric techniques in Peninsular Malaysia |
| title_sort | 2025_spatial air quality analysis using environmetric techniques in peninsular malaysia |