Scholarlink

<p>The presence of aerosol concentrations in mining regions substantially affects air quality and human health. Aerosol Optical Depth (AOD) is directly related to the amount and type of aerosols present in the atmosphere. Higher aerosol concentrations generally lead to higher AOD values. The main aim of this study is to utilize the Support Vector Machine (SVM) algorithm to forecast AOD near coal mines located in Assam. The SVM was developed for application on specific datasets between 2003 and 2019. The AOD data is acquired from Moderate Resolution Imaging Spectroradiometer (MODIS-Terra) in the proximity of coal mines over four distinct coalfields at Assam. Using a comprehensive dataset, the SVM model is trained and validated with great attention to detail. The hyperparameters are optimized, and measures are taken to mitigate overfitting. The optimized model is developed to predict monthly AOD<sub>550</sub> values based on historical AOD<sub>550</sub> observations. The analysis of seasonal patterns reveals that Dilli-Jeypore, Mikir, and Sheelveta exhibit the most elevated values of AOD during the Winter season, with respective values of 0.58 ± 0.16, 0.63 ± 0.14, and 0.49 ± 0.13. The Singrimari sample site reaches its maximum AOD during the Pre-Monsoon season. In contrast, the Dilli-Jeypore, Mikir, and Sheelveta regions show the lowest values during the Post-Monsoon season respectively. However, Singrimari is an exception, with higher AOD during the Monsoon season (0.50 ± 0.20). It is observed that the Sheelveta coalmine fields exhibit the lowest Root Mean Square Error (RMSE) values in both the training and testing phases, with values of 0.0469 and 0.0744, respectively. The results show that this method establishes better ecosystems and sustainable mining methods.</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

Prediction of Aerosol Optical Depth Using a Support Vector Machine Model in the Coal Mining Regions of Assam

  • 期刊论文

  • 发表日期：

    2026年5月29日

查看全文
  • Nishant Kumar,
  • Kirti Soni,
  • Kulwinder Singh Parmar,
  • Pranalee Thorat,
  • Dipankar Saha,
  • Arvind Kumar Jha,
  • Anikender Kumar,
  • Vijay Kumar Soni

摘要

The presence of aerosol concentrations in mining regions substantially affects air quality and human health. Aerosol Optical Depth (AOD) is directly related to the amount and type of aerosols present in the atmosphere. Higher aerosol concentrations generally lead to higher AOD values. The main aim of this study is to utilize the Support Vector Machine (SVM) algorithm to forecast AOD near coal mines located in Assam. The SVM was developed for application on specific datasets between 2003 and 2019. The AOD data is acquired from Moderate Resolution Imaging Spectroradiometer (MODIS-Terra) in the proximity of coal mines over four distinct coalfields at Assam. Using a comprehensive dataset, the SVM model is trained and validated with great attention to detail. The hyperparameters are optimized, and measures are taken to mitigate overfitting. The optimized model is developed to predict monthly AOD550 values based on historical AOD550 observations. The analysis of seasonal patterns reveals that Dilli-Jeypore, Mikir, and Sheelveta exhibit the most elevated values of AOD during the Winter season, with respective values of 0.58 ± 0.16, 0.63 ± 0.14, and 0.49 ± 0.13. The Singrimari sample site reaches its maximum AOD during the Pre-Monsoon season. In contrast, the Dilli-Jeypore, Mikir, and Sheelveta regions show the lowest values during the Post-Monsoon season respectively. However, Singrimari is an exception, with higher AOD during the Monsoon season (0.50 ± 0.20). It is observed that the Sheelveta coalmine fields exhibit the lowest Root Mean Square Error (RMSE) values in both the training and testing phases, with values of 0.0469 and 0.0744, respectively. The results show that this method establishes better ecosystems and sustainable mining methods.