SSDC's SciComHub and MATISSE: A Web-Based, Reproducible Framework for Multi-Mission Space Science

Background and Motivation: Modern planetary exploration, and in particular the selection of safe and scientifically valuable landing sites, depends on the joint exploitation of spectral, morphological, thermal, and radar datasets within a consistent geodetic reference. The data reduction of the main planetary missions traditionally relies on open-source software such as ISIS and the Ames Stereo Pipeline (ASP) [1, 2], which are powerful but notoriously difficult to install and to reproduce consistently across different machines, while the relevant datasets remain scattered across heterogeneous archives. To address these issues, the Space Science Data Center (SSDC) of the Italian Space Agency (ASI) maintains two complementary, side-by-side frameworks: the Science Computing Hub (SciComHub) [3, 4], which provides ready-to-use, containerized processing environments, and MATISSE v2 (Multi-purpose Advanced Tool for the Instruments of the Solar System Exploration) [5, 6], a research and visualization framework for multi-mission planetary data. Used together, they form a web-based, reproducible pipeline for landing site selection and other planetary investigations. 

Processing Framework — the Science Computing Hub: SciComHub is a centralized, browser-accessible workspace that ships ready-to-use Docker builds of the main planetary software stacks, avoiding the local installation effort and ensuring reproducible analyses. 

Isolation and Scheduling: Docker and Podman provide sandboxed user sessions; upcoming releases will add SLURM for compute-intensive batch jobs and Kubernetes for concurrent workloads. 

Pre-Built Software Stacks: SciComHub exposes Docker images that users can launch on demand without local setup: a Planetary stack with USGS ISIS, NASA ASP, GDAL and GRASS GIS for DTM generation and mosaic reprojection, and an Astrophysical stack with dedicated libraries for stellar and galactic data reduction, including SEDBuilder integration. 

Heritage Code: Dedicated Fortran, C, and C++ environments keep legacy scientific codes operational and reusable. 

Accelerated Computing: Planned GPU support for TensorFlow and PyTorch will enable automated feature recognition tasks, such as crater counting, on HPC resources. 

Research and Visualization Framework — MATISSE v2: MATISSE v2 is a GIS-oriented framework for interactive 2D and 3D query, exploration, and visualization of multi-mission planetary datasets. 

Stratigraphy-Aware Queries: MATISSE allows users to filter datasets by mapped geological units. With global geological maps of Mercury, Ceres, the Moon, and Mars, searches can be restricted to specific stratigraphic domains, so that compositional and physical observations are framed within their geological context.  

Impact and Subsurface Studies: New crater-centered queries use catalogs of terraced craters to investigate the impact stratigraphy of Mars and probe near-surface structures. 

 

Core Application — Landing Site Selection: Landing site selection is the main use case driving the joint use of SciComHub and MATISSE v2, implemented as a FAIR-compliant (Findable, Accessible, Interoperable, Reusable) pipeline in which SciComHub performs the data processing and MATISSE v2 supports the scientific query and visualization of the results. Engineering Safety Layer: DEM-derived products are computed to evaluate slope distributions, surface roughness, and illumination conditions against mission-specific safety thresholds, so that candidate sites comply with spacecraft landing constraints 

Scientific Value Layer: In parallel, the stratigraphy-aware query engine is used to isolate terrains of high scientific return, selecting regions by geological unit, mineralogy, or thermophysical signature.  

Weighted Suitability Mapping: Engineering and scientific layers are then combined within a unified GIS environment to produce Weighted Suitability Maps, in which each candidate area is scored against tunable criteria. 

This layered workflow supports iterative, collaborative trade-offs between engineering and science teams during mission planning. The pipeline has been prototyped and validated on lunar test cases, integrating LRO/LROC high-resolution imagery and topography, Chandrayaan-1 M3 mineralogical data, and legacy global datasets such as Clementine and MOLA; the same workflow is being extended to Mars and Mercury candidate sites. 

 

Roadmap: The next development phase will concentrate on:  

New Data Products: The MATISSE archive will be extended with 3D Digital Terrain Models (DTMs and VTP meshes) generated from CaSSIS observations on board the ExoMars Trace Gas Orbiter, public released in in early 2026, together with a dedicated continuum-removed M3 product from Chandrayaan-1 for lunar surface investigations. 

Virtual Observatory Interoperability: Adoption of IVOA protocols so that SSDC-derived products can be exposed and discovered through the global Virtual Observatory. 

Machine Learning Services: Integration of AI-based workflows for predictive terrain modeling and automated detection of surface features. 

Single Sign-On: Deployment of eduGAIN-based federated authentication to grant researchers worldwide institutional access. 

 

Summary: By combining SciComHub for reproducible, scalable data processing with MATISSE v2 for scientific query and visualization, the SSDC provides an integrated yet modular pipeline that reduces the human error and the latency associated with heterogeneous archives. This side-by-side architecture supports landing site selection and the broader exploration of the Solar System. A new release, MATISSE v3, is currently in development and will introduce tighter integration with SciComHub, extended 2D and 3D visualization capabilities, and native processing functionalities aimed at reducing the need for external tools in standard analysis workflows. 

 

Funding: This work is supported by the ASI-INAF agreement n. 2025-33-HH.0. 

 

References: [1] Sucharski T. et al. (2020) USGS-Astrogeology/ISIS3, ISIS 4.2.0 public release. [2] Beyer R. A. et al. (2018) Earth and Space Science, 5(9), 537–548. [3] Brandt C. H. et al. (2024) Europlanet-Gmap/Docker-JupyterHub, Zenodo. [4] Nodjoumi G. et al. (2025) Earth and Space Science, 12, e2025EA004251, https://doi.org/10.1029/2025EA004251.   [5] A. Zinzi, M.T. Capria, E. Palomba, P. Giommi, L.A. Antonelli, (2016) Astronomy and Computing, Volume 15, 2016, Pages 16-28, ISSN 2213-1337. [6] V. Camplone, A. Zinzi, M. Massironi, A.P. Rossi, F. Zucca, (2024) Astronomy and Computing, Volume 48, 2024, 100852, ISSN 2213-1337.