A comparative analysis of global lunar crater catalogs using OpenCraterTool – An open source tool to determine and compare crater size-frequency measurements

Heyer T.; Iqbal W.; Oetting A.; Hiesinger H.; van der Bogert C.H.; Schmedemann N.

Abstract

Impact crater size-frequency measurements are key to date and investigate surface processes on terrestrial planetary bodies. Using high-resolution images of numerous space missions, various regional and global crater datasets of the Moon and other planetary bodies have been created in recent years. In addition, machine-learning techniques are increasingly being used to automatically detect craters and to create extensive crater catalogs. Accurate map projection-independent measurement of crater sizes is essential for an accurate dating of the investigated surface. Thus, based on the work of Kneissl et al. (2011) on CraterTools for the geographic information system ArcGIS, we developed the OpenCraterTool as an open source QGIS extension. Our tool allows a map projection-independent measurement and comparison of craters. Results can be visualized within the tool or exported for further external analysis. With our new tool, we compared five global lunar crater datasets; three datasets based on manual crater measurements and two machine-learning products. The comparison revealed substantial differences in both the number of craters and crater diameters. In particular, two catalogs have about three times as many craters as the two smaller catalogs in the analyzed crater diameter range (>5 ​km). The compared craters revealed mean relative differences of 0.4 and 2.3% between the respective catalogs. As expected, the best agreements were found between the manual and automatically generated crater catalogs, where the manual counts were used as training datasets. However, the comparison of the machine-learning products revealed the lowest agreements between themselves. Although the mean relative differences between all the catalogs are small, the number of outliers, as well as the number of under- and overcounted craters over the entire diameter range, are high between all compared catalogs. The differences in crater diameter, as well as the number of over- and undercounted craters have noticeable effects on the age determination of the surfaces.