Scholarlink

<p>Air pollution remains a critical environmental and public health challenge in rapidly urbanizing Indian cities. This study presents a comprehensive assessment of Hyderabad’s air quality using daily observations from the Central Pollution Control Board (CPCB) spanning 2015–2020. Key pollutants analyzed include PM₂.₅, PM₁₀, NO, NO₂, NOx, NH₃, CO, SO₂, O₃, benzene, toluene, and xylene. Statistical analyses revealed pronounced seasonal variability, with PM₂.₅ and PM₁₀ concentrations peaking in winter due to temperature inversions, while monsoon rainfall contributed to natural cleansing. Correlation analysis showed a strong positive relationship between NO₂ and NOx (<i>r</i> = 0.85), confirming vehicular emissions as a dominant source, and a negative correlation between O₃ and NOx (<i>r</i> = − 0.60), reflecting photochemical titration processes. Extreme event analysis identified 22 days with PM₂.₅ levels exceeding 300&#xa0;µg/m³, including a peak of 389.1&#xa0;µg/m³ in December 2017, underscoring acute exposure risks. Principal Component Analysis (PCA) highlighted traffic and industrial emissions as primary contributors, explaining 65% of the variance. Machine learning evaluation using a Random Forest Regressor demonstrated strong predictive performance for PM₁₀ (R² = 0.92), NO₂ (R² = 0.95), and benzene (R² = 0.93), though PM₂.₅ predictions were less accurate (R² = 0.68), emphasizing the need to incorporate meteorological drivers and lagged features. Overall, the study demonstrates the utility of combining conventional statistical approaches with machine learning for diagnostic and forecasting purposes. Findings highlight the importance of continuous monitoring, stricter emission regulations, adaptive seasonal policies, and complementary strategies such as urban agriculture to enhance resilience and sustainability of air quality management in Hyderabad and comparable Indian megacities.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi Scale Analysis of Particulate and Gaseous Pollutants in Hyderabad: Trends, Correlations, and Machine Learning Enhanced Predictive Insights

  • 期刊论文

  • 发表日期：

    2026年3月26日

查看全文
  • Guhan V.,
  • Dharma Raju A.,
  • Naga Ratna K.,
  • Sheshakumar Goroshi,
  • Ravindra Kumar M.

摘要

Air pollution remains a critical environmental and public health challenge in rapidly urbanizing Indian cities. This study presents a comprehensive assessment of Hyderabad’s air quality using daily observations from the Central Pollution Control Board (CPCB) spanning 2015–2020. Key pollutants analyzed include PM₂.₅, PM₁₀, NO, NO₂, NOx, NH₃, CO, SO₂, O₃, benzene, toluene, and xylene. Statistical analyses revealed pronounced seasonal variability, with PM₂.₅ and PM₁₀ concentrations peaking in winter due to temperature inversions, while monsoon rainfall contributed to natural cleansing. Correlation analysis showed a strong positive relationship between NO₂ and NOx (r = 0.85), confirming vehicular emissions as a dominant source, and a negative correlation between O₃ and NOx (r = − 0.60), reflecting photochemical titration processes. Extreme event analysis identified 22 days with PM₂.₅ levels exceeding 300 µg/m³, including a peak of 389.1 µg/m³ in December 2017, underscoring acute exposure risks. Principal Component Analysis (PCA) highlighted traffic and industrial emissions as primary contributors, explaining 65% of the variance. Machine learning evaluation using a Random Forest Regressor demonstrated strong predictive performance for PM₁₀ (R² = 0.92), NO₂ (R² = 0.95), and benzene (R² = 0.93), though PM₂.₅ predictions were less accurate (R² = 0.68), emphasizing the need to incorporate meteorological drivers and lagged features. Overall, the study demonstrates the utility of combining conventional statistical approaches with machine learning for diagnostic and forecasting purposes. Findings highlight the importance of continuous monitoring, stricter emission regulations, adaptive seasonal policies, and complementary strategies such as urban agriculture to enhance resilience and sustainability of air quality management in Hyderabad and comparable Indian megacities.