Time-series satellite remote sensing reveals gradually increasing war damage in the Gaza Strip

Area of Science:

• Remote Sensing and Geospatial Analysis
• Artificial Intelligence in Conflict Monitoring
• Urban Planning and Disaster Assessment

Background:

• War-related urban destruction poses global challenges to security, stability, and development.
• Understanding the impact of conflict on urban geomorphology and geological contexts requires advanced analytical methods.
• Previous assessments of conflict damage lacked near real-time, detailed damage level analysis.

Purpose of the Study:

• To integrate deep learning with satellite data for near real-time detection of explosions and building damage assessment.
• To quantify the extent and types of structural damage in the Israel-Palestine conflict.
• To provide data-driven insights for conflict monitoring, demining, and humanitarian aid.

Main Methods:

• Utilized data from the LuoJia3-01 satellite.
• Applied deep learning algorithms for object detection (missile craters) and damage classification.
• Analyzed changes in urban structures and agricultural land from October 2023 to March 2024.

Main Results:

• Detected 3747 missile craters and assessed escalating damage from October 2023 to March 2024.
• Identified 58.4% of damaged structures as residential or educational (15.4% destroyed, 18.7% severely damaged).
• Documented a 34.1% decline in cultivated agricultural land, impacting food security.

Conclusions:

• Integrated deep learning and satellite imagery offer a scalable, cost-effective method for conflict damage assessment.
• The findings highlight severe humanitarian consequences, including housing crises and food insecurity.
• Emphasizes the need for an immediate ceasefire, hostage release, and reconstruction efforts.