ALBERT

Description: Granular processes are ubiquitous within volcanic environments. For instance, they contribute to the initiation, transport, and arrest of deadly pyroclastic density currents (PDCs). However, an incomplete understanding of granular mechanics limits our ability (i) to predict the inundation area of PDCs, and (ii) to understand the formation of complex deposits, which help us decipher the past history of volcanoes around the world. In part, this incomplete understanding is due to the intrinsic complexity of granular media, which span a range of flow regimes, ranging from solids to gases and display local and non-local rheology. Natural volcanic granular flows are composed of non-spherical particles with large polydispersities and variable densities, making their rheology challenging to study. Our limited understanding of granular media, is mostly restricted to quasi-monodisperse flows that span size ratios 〈10, while natural flows have size distributions exceeding two orders of magnitude. During this presentation, we briefly discuss the complex behaviour of granular media and the challenges we have encountered in exploring their dynamics. Our study investigates the rheology of polydisperse granular media with direct application to PDC mixtures by using tools from soft-matter physics, such as the discrete element method and rheometry experiments. Specifically, we demonstrate how to encapsulate polydispersity in volcanic mixture constitutive descriptions. The findings of our study have wide-ranging implications for geosciences.

Description: Snow-atmosphere humidity fluxes change the isotopic composition of the surface snow after deposition. However, to date, it is unclear which role these post-depositional processes play in the formation of the climate signal recorded in ice cores. Quantifying the post-depositional impact on the isotope record in ice cores is challenging due to the poor humidity flux representation over ice sheets in isotope-enabled climate models. Here, we present results from the isotope-enabled SNOWISO exchange and snowpack model. We run a 30-year (8.5 m) snow core simulation for the EastGRIP drilling site on the Greenland Ice Sheet with and without humidity flux induced fractionation of the surface snow implemented. To guarantee reliable forcing, we combine isotopic input from the global climate model ECHAM-wiso with accurate surface humidity flux simulations from the polar regional climate model MAR. Implementing fractionation induced by the humidity flux increases the mean annual δ18O by +2.1 ‰ and reduces the mean annual d-excess by -6.2 ‰ with a limited effect on the year-to-year variability. We further show that capturing the diurnal cycle of the humidity flux instead of using daily or monthly means is essential to correctly simulate the post-depositional effect on the snow core simulation. Our results shed new light on the current proxy interpretation of stable water isotopes in ice cores and open new opportunities to infer temperature and vapor source region signals more accurately.