Mars Crater Database: A participative project for the classification of the morphological characteristics of large Martian craters

The most recent comprehensive database of Martian impact craters was the result of the work of impact crater scientists (S.J. Robbins and B.M. Hynek) who carefully examined the available high-resolution imagery of Mars. Building on this previous work, we present the result of an alternative approach involving 56 planetary scientists and trained students. A web platform was designed for this purpose. All impact craters larger than 1 km in diameter were classified according to a simplified classification scheme, recording the primary or secondary nature of the crater, and the morphology of the ejecta (single, double, or multiple layered ejecta rampart sinuous [LERS], or low-aspect-ratio layer ejecta [LARLE]). In total, 8445 LERS craters, 24,530 partially buried craters, 55,309 secondary craters, and 288,155 craters in the category “standard” were identified. Our assessment differs for 8145 entries in the original database compiled by Robbins and Hynek, which are not considered to be impact structures. In this work, ~39,000 secondary craters have been associated with 108 primary craters. Coupled to the existing database, the database we propose here offers a complementary way to investigate the geological history of Mars. More specifically, the completion of layered ejecta crater morphologies down to 1 km and the connection established between secondary and primary impact crater sources will allow the implementation of statistical studies to reveal the spatial and temporal evolution of the impacted material characteristics. Thanks to the simplified classification we performed here, this version of the database can be easily used as a training data set for crater identification algorithms based on machine-learning techniques with the aim to identify smaller impact craters and to automatically define their morphological characteristics. Since it is not possible to confirm an impact structure from remote-sensing data alone, any Martian impact database at this stage remain