Spectral properties catalogue of Earth-Venus analogues: Etna example

Volcanic regions provide natural laboratories for studying interactions between surface materials, atmospheric aerosols, and radiative processes on Earth and other planetary bodies. This study combines ground-based gas monitoring, spaceborne hyperspectral imaging, and unsupervised machine learning to characterize the eruptive activity and plume properties of Mount Etna (Italy) and to develop analogs for volcanic processes in the atmosphere of Venus. Etna, Europe’s most active volcano, exhibited persistent activity from 2023 to 2025, dominated by Strombolian explosions, lava fountains, ash plumes, and small lava flows centered on the Southeast Crater Complex.

Key volcanic gas parameters are compiled from Istituto Nazionale di Geofisica e Vulcanologia (INGV) reports into a harmonized, machine-readable dataset. The time series include daily sulfur dioxide flux (SO₂), carbon dioxide flux (CO₂), mean partial pressure of CO₂ (pCO₂), and helium isotope development (He), all of which are fundamental indicators of the state of the magmatic system.

Concurrently, all available Earth Surface Mineral Dust Source Investigation (EMIT) hyperspectral images (350–2500 nm) acquired from the International Space Station over Etna during periods of unrest and eruption are used to characterize the detailed spectral behavior of volcanic surface materials. Radiance and reflectance data cubes are converted to spectral absorption wavelength images to isolate diagnostic absorption features directly related to specific minerals or material types. These products emphasize Fe-bearing silicates, oxides, alteration phases, and other mineral constituents of fresh/weathered lava flows, pyroclastic deposits and volcanic plumes. Unsupervised machine learning classification is then applied to the processed hyperspectral data to derive material and mineral maps without prior training data. For each class, representative spectra (average, minimum, maximum) are computed over the full spectral range to capture characteristic signatures and internal variability, allowing comparison with available spectral libraries (in-house, USGS, ECOSTRESS).

To derive land surface temperature concurrent ECOSTRESS data are selected and analyzed. The ECOSTRESS instrument is a multispectral thermal imaging radiometer that provides high-resolution measurements of surface thermal emission.

The integration of hyperspectral, gas, and thermal datasets provides a promising framework for characterizing volcanic plumes and lava flows. EMIT-based spectral information is combined with concurrent ECOSTRESS thermal observations to derive plume temperatures and discriminate plume types based spectral–compositional signatures. Unsupervised techniques successfully distinguish plumes from the background and identify different plume regimes. Preliminary results indicate that mineral particulates within plumes, including ferric iron (Fe³⁺) phases, can be detected, implying that both gaseous and mineralogical components of volcanic plumes are resolvable in space and time. This is particularly relevant for comparative planetology. The inferred mineralogical composition of terrestrial volcanic plumes may constrain plausible mineral particulates and aerosol types on Venus, yielding testable predictions for the composition and spectral behavior of Venusian volcanic aerosols and mineral dust.

Acknowledgement: The manufacturing of the VenSpec electronics and the preparation of spectral libraries for the EnVision mission in the Czech Republic are funded by ESA PRODEX under contract PEA4000147310.