MISSIONS NEED MAPS: TOWARDS A GLOBAL CRATER DATABASE FOR MARS AND THE MOON AT ULTIMATE RESOLUTION.

Abstract:

We can derive geological units age for planetary bodies we’ve never visited by counting craters on their surfaces. The more craters, the older the surface. But although the resolution of planetary imaging has increased by around 5 orders-of-magnitude since then, the approach is the same: craters are counted manually. The current best dataset for Mars is a complete count of 385,000 craters >1km in diameter [1]. We can use this data to derive model ages of the surface of Mars, but it is at low resolution. We developed and trained an automated Crater Detection Algorithm (CDA) using a convolutional neural network to identify them on higher resolution imagery [2]. Our CDA reproduced, within error, the manually compiled database [2]. We have recently extended it to CTX imagery which has global coverage at 5m/pix [3] and in specific areas, HiRISE (0.25m/px) imagery.

Applied on the global CTX mosaic [3] our CDA detected more than 94 million of craters larger than 25m in diameter, constituting therefore the biggest crater database ever built. Numerous small secondary craters locally dominate the crater population (see figure where the blue rays correspond to 800,000 small craters formed by the ejection of material following the Corinto impact, D=13km). We will present at the conference the ASCI tool, allowing to isolate clusters of secondary craters and then using the remaining crater populate for an age derivation [4]. Applied on high resolution imagery dataset such as NAC images (0.5m/px) covering the surface of the Moon, this method would allow us to date, with a high temporal and spatial resolution, geological events having shaped its surface. This could also aid with finding suitable landing sites for future Artemis missions.

We have produced a CDA that is an extraordinary tool with the potential to help to the identification of particular region of interest for future automated or human missions on the Moon. At the conference, we will present the huge potential of our technique by focusing on recent results acquired on Mars and recent advances of our group to create the biggest lunar crater database ever built.

References: [1] Robbins, S. and Hynek, B. (2012) JGR, 117, E05004. [2] Benedix, G.K., et al. (2020) Earth and Space Science, e2019EA001005. [3] Dickson et al. (2018) 49^th LPSC. [4] Lagain, A. et al., submitted to PNAS (July 2020).