Oil Spill Detection Performance in a Multitype Polarimetric-Feature Space Using a Polarimetric Synthetic Aperture...

Area of Science:

• Remote Sensing
• Oceanography
• Environmental Monitoring

Background:

• Marine oil spills pose significant threats to ecosystems and economies.
• Fully polarimetric synthetic aperture radar (FP SAR) is crucial for all-weather oil spill monitoring.
• Variability in oil types affects FP SAR data interpretation.

Purpose of the Study:

• To compare polarimetric feature distribution based on oil slick type and incident angle using RADARSAT-2 data.
• To assess the influence of incident angle on oil spill detection performance.
• To identify robust polarimetric features for oil spill monitoring.

Main Methods:

• Analysis of RADARSAT-2 images from a North Sea oil-on-water experiment.
• Comparison of polarimetric features across different oil slick types and incident angles.
• Evaluation of feature robustness and effectiveness for oil spill detection.

Main Results:

• Incident angle influences polarimetric features and detection, especially for low contrast spills at large angles.
• Polarimetric features related to scattering mechanisms show robustness across incident angles.
• A combination of scattering entropy (H) and modified anisotropy (A12) demonstrated strong performance.

Conclusions:

• Appropriate polarimetric feature selection, considering incident angle, enhances oil spill identification.
• The study facilitates better oil spill detection and monitoring by understanding feature behavior.
• This approach aids in differentiating oil slick types and improving overall monitoring accuracy.