Researchers at James Cook University (JCU) in Australia have developed an artificial intelligence (AI) algorithm that merges two types of satellite imagery to better detect and assess oil spills at sea. The method combines synthetic aperture radar (SAR) and hyperspectral imaging (HSI) to identify spills more accurately and estimate their thickness, offering a new tool for environmental monitoring and disaster response.
SAR sensors detect surface roughness changes on the ocean, making them useful for spotting oil slicks, while hyperspectral imagery captures reflected light across many wavelengths to distinguish materials. By fusing both datasets, the AI model benefits from the strengths of each and produces more reliable results.
The model first uses SAR data to scan large ocean areas for possible spills, then applies HSI data to classify the oil type and thickness. This information supports faster and more targeted cleanup efforts, helping responders allocate resources more efficiently.
For the UN-SPIDER community, this development supports the mission of using space-based technologies for disaster management. More accurate spill detection can reduce environmental harm and strengthen marine protection. The researchers also see potential for applying the method to other environmental challenges, including water quality and coastal monitoring.
Further testing is planned to validate the algorithm under varying ocean conditions and integrate it into real-time systems. As hyperspectral satellites become more available, this fusion approach could enhance global capabilities for environmental hazard detection.
Read the full article here: AI algorithm combines different satellite imagery for precise oil spill detection
Reference:
Liu, Q., Huang, K., et al. (2025). Enhancing oil spill detection with controlled random sampling: A multimodal fusion approach using SAR and HSI imagery. Remote Sensing Applications: Society and Environment.
