Individual boulder detections of Global Lunar Boulder Map from LRO NAC optical images using Deep Learning: Implications for Regolith and Protolith
收藏资源简介:
This archive contains the individual lunar boulder detections between 60°S and 60°N derived with the boulder detection algorithm developed by Aussel et al. (2025). The data is split into 72 parquet files named as "<lon1>_<lon2>_<lat1>_<lat2>_dets.parquet", containing the boulder locations, sizes, areas and additional metadata. In addition, it contains maps of the selected LRO NAC images ("pos_metadata.tif"), with layer 1 indicating the respective grid cell (0.02°x0.02°) coverage in percent, layer 2 indicating the resolution and layer 3 the incidence angle of the selected LRO NAC image. Moreover, a 128 ppd map of the areal fraction of the detected boulders is provided in "boulder_areal_fraction.tif", containing gaps (NaN) values at locations where no LRO NAC image was selected or where the region was covered in shadow. This information is not available in the individual detection files.



