High resolution hyperspectral and discrete analysis of Adamello glacier ice-core sections from the ADA270 drilling project

Paleoclimatic information recorded in polar and mountain glaciers are very important to study climate and environmental changes, telling a story about the ancient past and the ongoing cryosphere response to the anthropogenic impact. In this contest, a new high spatial resolution ice-core analysis has been done using a hyperspectral system, which allows to detect ice characteristics related to paleoclimatic information in a new non destructive way. This improves the accuracy of measurements and the preservation of samples. The hyperspectral analysis has been applied on sections from the ADA270 ice core project, drilled in 2021 at the Adamello-Mandrone glacial group (Pian di Neve, Eastern Alps, Italy). The optical method acquires reflectance images collecting spectral radiance in 840 bands in visible and NIR wavelengths (400-1000 nm). Different optical descriptors have been computed from all reflectance spectra to measure variations due to ice features and impurities encloses in ice layers. In particular, we considered the surface albedo and two spectral indices: Snow Darkening Index (SDI) and Impurity Index (II). From such descriptors continuous records along the core, in particular Albedo and SDI, samples are accurately selected and melted to make a discrete analysis with a Coulter Counter Beckman Multisizer 4e to make a calibration. Analyzed features are ice lenses, visible air bubbles, medium supposed dust concentration and high content of visible materials. The discrete data have been matched with SDI values for all analyzed sections to apply a regression model. A calibrated curve for impurities concentration has been created, providing a very important tool to not systematically apply destructive measurements in all ice-core sections. For a depth ranging from 14 m to around 35 m, many high concentration peaks (to a maximum of 30-35 ppm) have been identified and associated to visible impurities layers. The rest of the curve, highlights concentrations lower than 5-6 ppm. We have great confidence that the analysis of ice trough hyperspectral imaging has a great potential for improving the analysis of ice cores, guaranteeing the preservation of pristine cryosphere records over time.