An SQL-based framework to work with Galileo/NIMS data

Context

The Jupiter icy moons are the target of two major space missions in the next decade: The ESA JUICE [1] and NASA Europa Clipper [2] missions. The NASA Galileo spacecraft explored the Jupiter system between 1995 and 2003, and performed multiple encounters with Jupiter’s moons. Among the scientific instruments onboard Galileo operating during these encounters was the Near-Infrared Mapping Spectrometer instrument (NIMS), a complex imaging spectrometer working in the infrared domain, operating from 0.7 to 5.2 microns with 17 detectors [3]. To date, and despite challenges in its usage, the NIMS data remain a great resource of recent scientific research on the properties of Jupiter’s icy moons surfaces [e.g. 4-7] to prepare for the next explorers of the Jupiter system.

The Galileo NIMS data set

Galileo NIMS data are archived in the PDS as so-called “tubes” and “g-cubes” (or mosaics). We base our work on the g-cube products, where individual tubes are assembled in single mosaiced and geolocated products. G-cubes are available as products in radiance and radiance factor in the archive. Due to the varying distance and observing geometry of each targeted flyby in the Jupiter system, together with the own instrument operational settings and health, the dataset is very heterogenous in spatial, spectral, and angular resolution. Backplanes offer the possibility to easily extract the observation geometry per pixel, to be used in modeling spectral or photometric properties. Recent recalibrated NIMS data sets [e.g. 8, 9] are being studied to quantify differences with the original PDS archived NIMS data used in this work.

The NIMS database framework

We converted the Galileo/NIMS calibrated g-cube dataset into a relational SQL database. This framework allows to quickly select and extract radiance factors (I/F), radiance (I), geometry data, and metadata from the entire NIMS data set. The data that can be retrieved cover the NIMS observations of Jupiter, Io, Europa, Ganymede, and Callisto. Metadata extracted in the PDS3 labels regarding the individual observations are part of the database, together with calibration information and calibration data available from the g-cubes labels. The smallest element in the database is a spectrum (i.e. one pixel). Using SQL queries on this SQL database, and criteria based on the pixel viewing geometry (e.g. incidence, emission, phase, and azimuth) and the geographic pixel location (latitudes and longitudes on a given target), phase curves and/or collections of spectra can be easily retrieved from regions of interest. Individual g-cubes data can also be retrieved. 

Making available the NIMS database to the community

We wish to make this framework available for the community interested in working with NIMS data. We recently integrated the framework within the ESA DataLabs platform [10], which provides a JupyterLab-based environment from which users will be able to easily access the database content using notebooks. A python package to perform predefined queries will also be part of the DataLab. Queries can be executed in the DataLab to extract the NIMS spectra to be analysed, and query results can be exported.

References  

[1] Grasset et al., Plan Spac Sci 78, 1-21, 2013; [2] Howell and Pappalardo, Nat Commun 11, 1311, 2020 ; [3] Carlson et al., Space Science Reviews, 60, 457-502, 1992 ; [4] Mishra et al. The Planetary Science Journal, 2, 183, 2001 ; [5] Cruz Mermy et al., Icarus, 394, 115379, 2023 ; [6] Belgacem et al., Planetary Science Journal, 6:237, 2025 ; [7] Cruz Mermy et al., Icarus, 442, 2025 ; [8] Malaska et al., 2023a, PDART program, DOI:10.17189/4sq6-x165 ; [9] Malaska et al., 2023b, PDART program, DOI:10.17189/4sz4-5024 ; [10] Navarro et al., Space Data Management, 1-13, 2024.