A Slope Correction Method for Ground Photon Extraction Over Mountainous Forest AREAs

The ICESat-2 (Ice, Cloud, and Land Elevation Satellite-2) plays an important role in the scientific task of global terrain mapping. ICESat-2 records high-precision object information in the form of photon clouds. Ground photon extraction is a key step in retrieving high-precision terrain. However, it is a difficult task to extract ground photons in mountainous forest areas due to the influence of terrain slopes. We proposed a slope correction method to mitigate the effect of terrain slope on ground photon extraction. First, the background noise photon is removed by a photon cloud filtering method. Second, we used the Douglas–Peucker algorithm to find the special terrain points (STPs), the photon cloud then was divided into a series of segments by STPs. We calculated the slope in each segment to form a rotation matrix, and corrected the slope of the photon cloud in each segment. The results show that the obtained STPs have high accuracy, and these STPs divided the photon cloud into a series of segments accurately. Compared with the reference slopes in each segment, the obtained slopes have a small RMSE of 2.3° and a high R
²
of 0.96. Additionally, the results indicate that the discrimination between ground photons and canopy photons in mountainous forest areas is significantly obvious after a slope correction.