Signal Photon Extraction Method for ICESat-2 Data Using Slope and Elevation Information Provided by Stereo Images

Area of Science:

• Earth Observation
• Geospatial Science
• Remote Sensing

Background:

• Laser altimetry data, such as from ICESat-2, often contain significant noise.
• Effective signal photon extraction is crucial for accurate surface elevation measurements.
• Existing algorithms may struggle with noisy data, particularly in complex terrain.

Purpose of the Study:

• To develop and validate a novel density clustering method for extracting reliable surface signal points from noisy ICESat-2 laser altimetry data.
• To improve the accuracy and reliability of laser altimetry data processing, especially in challenging environments.

Main Methods:

• A density clustering approach was developed, integrating optical stereo image data for slope and elevation.
• The method adaptively adjusts neighborhood search directions and calculates local classification density thresholds.
• Performance was evaluated using both strong and weak beam ICESat-2 data against the ATL08 algorithm.

Main Results:

• The proposed method demonstrated superior signal extraction quality compared to the ATL08 algorithm in steep slope and low signal-to-noise ratio (SNR) regions.
• It achieved a better balance between recall and precision, resulting in a higher F1-score.
• Accurate extraction of continuous and reliable surface signals was confirmed across diverse terrains and land cover types for both strong and weak beams.

Conclusions:

• The novel density clustering method effectively extracts reliable surface signals from noisy ICESat-2 laser altimetry data.
• This approach offers improved performance over existing methods, particularly for challenging datasets.
• The method provides a robust solution for enhancing the quality and utility of laser altimetry data for various applications.