Indoor sulfur dioxide prediction through air quality modeling and assessment of sulfur dioxide and nitrogen dioxide levels in industrial and non-industrial areas.
Jamal Kamal Mohammedamin1, Yahya Ahmed Shekha2
1Environmental Science and Health Department, College of Science, Salahaddin University, Erbil, Iraq. jamal.mohammedamin@su.edu.krd.
Related Experiment Videos
Automated, High-resolution Mobile Collection System for the Nitrogen Isotopic Analysis of NOx
Published on: December 20, 2016
08:23Chemiluminescence-based Assays for Detection of Nitric Oxide and its Derivatives from Autoxidation and Nitrosated Compounds
Published on: February 16, 2022
08:23Analyzing the Photo-oxidation of 2-propanol at Indoor Air Level Concentrations Using Field Asymmetric Ion Mobility Spectrometry
Published on: June 14, 2018
View abstract on PubMed
This study measured indoor and outdoor sulfur dioxide (SO2) and nitrogen dioxide (NO2) in industrial and non-industrial areas. Machine learning models, particularly Random Forest, accurately predicted indoor SO2 levels, offering insights for exposure reduction.
Area of Science:
- Environmental Science
- Air Quality Monitoring
- Computational Chemistry
Background:
- Indoor and outdoor air quality significantly impacts human health.
- Sulfur dioxide (SO2) and nitrogen dioxide (NO2) are key air pollutants with varying emission sources.
- Understanding pollutant distribution and developing predictive models is crucial for public health interventions.
Purpose of the Study:
- To quantify indoor and outdoor SO2 and NO2 concentrations in industrial and non-industrial settings across seasons.
- To compare the predictive performance of different machine learning (ML) models for indoor SO2 concentrations.
- To identify factors influencing indoor SO2 levels and inform strategies for reducing exposure.
Main Methods:
- Passive samplers were used to measure SO2 and NO2 levels in homes and outdoor environments.
- Data collection was conducted during summer and winter seasons in Tymar village and Haji Wsu.
- Machine learning models (MLR, ANN, RF) were trained using factor analysis outputs for predicting indoor SO2.
Main Results:
- Tymar village exhibited significantly higher indoor and outdoor SO2 and NO2 concentrations than Haji Wsu in both seasons.
- Peak outdoor SO2 was observed in summer, while peak indoor NO2 occurred in winter.
- The Random Forest model demonstrated superior accuracy in predicting indoor SO2 concentrations.
Conclusions:
- Industrial areas show higher SO2 and NO2 pollution levels compared to non-industrial areas.
- Indoor NO2 levels can exceed outdoor levels, particularly during winter.
- The Random Forest model effectively captures complex relationships, aiding in the development of targeted air quality management strategies.
