Scholarlink

<p>Pyrocumulonimbus (PyroCB) clouds, which result from intense wildfires, pose significant challenges due to their potential to inject aerosols into the upper atmosphere and influence climate dynamics. This study proposes a comprehensive deep learning-based framework for the detection, forecasting, and spatial analysis of Pyrocumulonimbus (PyroCB) events, specifically designed for wildfires over the Indian region. The proposed framework utilizes high-resolution Meteosat Second Generation (MSG) SEVIRI satellite and ERA5 reanalysis data to develop novel models for PyroCB detection, and forecasting. This work presents a new architecture, DIMPy (Detection and Integrated Masking for PyroCB), which combines innovative detection and mask generation models, extending prior frameworks like Pyrocast over the Indian region. With the availability of advanced imager data over&#xa0;the Indian Region (MSG-SEVIRI-IODC), this paper presents a unique model for detecting the spread of wildfires with the identification of PyroCB clouds. This study presents the first in-depth analysis of PyroCB events in the western Himalayas. With the backdrop of climate change, the&#xa0;severity of the fire and its associated meteorological linkages are anticipated to rise. Accordingly, DIMPy enables real-time analysis using fire-data independence, lossless approximations, spatial context, and minimal data, integrating AI with satellite data to improve PyroCB prediction accuracy, generalizability, and disaster response strategies.</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

Artificial Intelligence for Pyrocumulonimbus Cloud Detection: Enhancing Indian Wildfire Management

  • 期刊论文

  • 发表日期：

    2025年7月4日

查看全文
  • Darshit Amit Pandya,
  • Bipasha Paul Shukla

摘要

Pyrocumulonimbus (PyroCB) clouds, which result from intense wildfires, pose significant challenges due to their potential to inject aerosols into the upper atmosphere and influence climate dynamics. This study proposes a comprehensive deep learning-based framework for the detection, forecasting, and spatial analysis of Pyrocumulonimbus (PyroCB) events, specifically designed for wildfires over the Indian region. The proposed framework utilizes high-resolution Meteosat Second Generation (MSG) SEVIRI satellite and ERA5 reanalysis data to develop novel models for PyroCB detection, and forecasting. This work presents a new architecture, DIMPy (Detection and Integrated Masking for PyroCB), which combines innovative detection and mask generation models, extending prior frameworks like Pyrocast over the Indian region. With the availability of advanced imager data over the Indian Region (MSG-SEVIRI-IODC), this paper presents a unique model for detecting the spread of wildfires with the identification of PyroCB clouds. This study presents the first in-depth analysis of PyroCB events in the western Himalayas. With the backdrop of climate change, the severity of the fire and its associated meteorological linkages are anticipated to rise. Accordingly, DIMPy enables real-time analysis using fire-data independence, lossless approximations, spatial context, and minimal data, integrating AI with satellite data to improve PyroCB prediction accuracy, generalizability, and disaster response strategies.