Brick Kiln Dataset for Pakistan's IGP Region Using AI.

Area of Science:

• Environmental Science
• Remote Sensing
• Artificial Intelligence

Background:

• Brick kilns are significant air polluters in Pakistan and the Indo-Gangetic Plain.
• Limited air quality monitoring and data transparency hinder pollution source identification.
• Unregulated brick kilns contribute to severe air pollution challenges.

Purpose of the Study:

• To develop and validate an AI-driven approach for mapping brick kiln locations using satellite imagery.
• To create a dataset distinguishing between Fixed Chimney and Zigzag kilns for accurate pollution estimation.
• To support regulatory interventions and air quality management strategies in Pakistan.

Main Methods:

• A two-fold AI approach combining low-resolution Sentinel-2 and high-resolution satellite imagery.
• Initial identification of potential brick kilns using low-resolution data.
• Post-processing steps to reduce false positives and validate findings with high-resolution imagery.

Main Results:

• Successfully mapped approximately 11,000 brick kilns.
• Distinguished between Fixed Chimney and Zigzag kiln types.
• Demonstrated the effectiveness of AI and multi-resolution imagery for source detection.

Conclusions:

• AI-powered satellite imagery analysis is effective for identifying specific pollution sources like brick kilns.
• The developed dataset provides crucial insights for regulating unregistered kilns and managing air pollution episodes.
• This methodology can be applied to similar environmental monitoring challenges globally.