MarsSI: Martian surface data processing service

Abstract

MarsSI (“Mars Système d’Information”, French for Mars Information System) is a service to explore, request, process and retrieve planetary surface data https://marssi.univ-lyon1.fr/MarsSI.

1 Introduction

Geological investigations of planetary surfaces require the combination of orbital datasets. Missions being often multiple-instruments platforms from multiple space agencies, the quantity of data available increased quickly. It is now large enough that a dedicated system has to be used to explore, calibrate, process and retrieve the relevant information.

MarsSI [1] is a platform developed in the context of the e-Mars project (2012-2017) to facilitate the use of combined orbital data. It was certified in 2017 as french national Research Infrastructure by the Centre National de la Recherche Scientifique (CNRS) as part of the Planetary Surface Portal (PSUP) [2]. The permanent staff currently consists of one scientist lead and one engineer.

2 Architecture

Hosted by ENS-Lyon, MarsSI hardware consists of a frontend server, a storage bay a storage bay and a 80 core cluster. For the software stack, we decided to rely mainly on software provided by the FOSS communities and standard and documented protocols such as those proposed by the Open Geospatial Consortium (http: //www.opengeospatial.org/), managed in a PostGIS (https://postgis.net/) database.

3 Catalog

As of 2024, MarsSI indexes and give access to optical data (visible, multi and hyper-spectral) and derived products from three missions: Mars Odyssey, Mars Express and Mars Reconnaissance Orbiter. Figure 2 shows all the referenced products and the processing that can be applied to them. Our emphasis was to provide ”ready-to-use” products in regards of calibration, corrections and georeferencing. The user will be able to visualize and interpret the data in GIS or remote sensing software.

3.1 Imagery

MarsSI provides access to various optical datasets for visible, multi- and hyper-spectral data from the various martian orbital missions over the years. For example, for the MRO mission, CTX and HiRISE images are calibrated and map projected using the ISIS software suite (https://isis.astrogeology.usgs.gov/). CRISM hyperspectral data is processed using the CRISM Analysis Toolkit (CAT) pipeline developed by the CRISM team [3][4]. Spectral cubes are calibrated to I/F and corrected from the atmospheric contribution with the volcano-scan method [5] and spectral parameter maps are produced. This procedure was also applied to the OMEGA dataset.

3.2 DEMs

We offer multiple Digital Elevation Model (DEM) datasets in MarsSI. Some of them are provided from external sources (such as those provided by the HiRISE and HSRC teams). But users can also requests med- and high- resolution DEMs generated using the Ames Stereo Pipeline software [6]. Possible DEMs are computed by analyzing the CTX and HiRISE coverage footprints to find image pairs with overlapping by at least 60% and a minimum deviation of 10° in emission angle. While these products are the results of automated procedures and cannot benefit from human quality control and fine tuning, they do provide much larger coverage of Mars surface: HiRISE team provided DEMs cover 0.04% of Mars surface, while MarsSI HiRISE-based DEMs cover 3.84% and CTX-based DEMs 16.93%. The DEM generation workflow was updated in 2020 with a completely new version[7].

4 User interface

MarsSI client interface is a web application that aims to be as straightforward as possible and received many updates in the previous years, starting with a complete rewrite. As shown on figure 3, the user can explore the datasets over a map interface. Data can be selected and sent to a workspace. The workspace view, shown on figure 4, allow to review products in detail. This is also where users will be able to request dataset processing. User can create more workspaces to organize their data selections.

When datasets processing are finished, the user will request a copy operation. The selected data will be copied to a directory that is aviable through SFTP using the user credentials. Files in that directory are kept for 30 days.

5 Documentation

A wiki has been set to offer tutorials, with data specific pages including brief data descriptions and relevant links. A introduction video tutorial is available. In 2023, a ticketing system was also launched to report issues.

6 User community

MarsSI is open to the scientific communities around the world. As of april 2024, we count more than 600 registered users. Naturally for french service, 25% of the users are from France, but we also offer data to scientists from the USA, UK, India and China.

7 Perspectives

After many improvements to the platform, we now aim to complement our datasets. We are adding the SHARAD catalog and aim to provide a service to create radargram simulation based on our CTX DEM dataset to help interpret the real data. We also expect to expand the service to the Moon next year and provide LROC based datasets.

8 Conclusion

Built upon opensource frameworks and using standardized protocols, MarsSI offers the scientific communities a way to explore space agencies catalogs and automatically process them to high value products, notably DEMs derived from CTX and HiRISE datasets and spectral parameter maps from CRISM and OMEGA datasets.

Acknowledgments

MarsSI is part of national Research Infrastructure PSUP, recognized as such by the French Ministry of Higher Education and Research under the ANO5 label. It was supported by the Programme National de Planétologie (PNP) of CNRS/INSU, co-funded by CNES.

9 References

[1] C. Quantin-Nataf et al. In: Planetary and Space Science 150 (2018). [2] F. Poulet et al. In: Planetary and Space Science 150 (2018). [3] S. Pelkey et al. In: Journal of Geophysical Research: Planets 112.E8 (2007). [4] S. Murchie et al. In: Journal of Geophysical Research: Planets 112.E5 (2007). [5] P. C. McGuire et al. In: Planetary and Space Science 57.7 (2009). [6] Z. M. Moratto et al. In: Lunar and Planetary Science Conference. Vol. 41. 2010. [7] M. Volat, C. Quantin-Nataf, and A. Dehecq. In: Planetary and Space Science 222 (2022).