An optimized denoising method for ICESat-2 photon-counting data considering heterogeneous density and weak...

Area of Science:

• Geospatial science
• Remote sensing
• Oceanography

Background:

• The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) provides valuable underwater elevation data due to its photon penetration capabilities.
• Existing density-based denoising methods struggle with heterogeneous photon distribution, weak connectivity, and sparse signals in deep water or areas with drastic topographic changes.

Purpose of the Study:

• To develop and evaluate a novel fused denoising method for ICESat-2 photon data.
• To improve the accuracy and reliability of underwater elevation extraction from ICESat-2 data, particularly in challenging environments.

Main Methods:

• A fused denoising approach combining the local outlier factor and inverse distance metric was developed.
• The method utilizes K-nearest neighbors (KNNs) to consider data density differences and directional uniformity.
• Performance was validated using six ICESat-2 trajectories across diverse seabed topographies and compared against state-of-the-art algorithms and official ATL03 data.

Main Results:

• The proposed method achieved an overall accuracy exceeding 96%.
• It demonstrated higher recall rates and lower false positive rates compared to existing methods.
• The technique effectively processed areas with abrupt topographic changes and deep water, yielding more continuous signal extraction.

Conclusions:

• The novel fused denoising method significantly enhances ICESat-2 photon data processing for underwater elevation.
• This approach offers a superior alternative to density-based algorithms, particularly for complex terrains and deep-water applications.
• The study opens new avenues for ICESat-2 photon-counting data denoising research.