Modelling Cold Firn Evolution at Colle Gnifetti, Swiss/Italian Alps

The existence of cold firn and ice within the European Alps provides an invaluable source of paleoclimatic data with the capability to reveal the nature of anthropogenic forcing in Western Europe over the preceding centuries. Unfortunately, continued atmospheric warming has initiated the thermal degradation of cold firn to that of a temperate firn facie, where infiltrating meltwater compromises this vital archive. However, there is currently limited knowledge regarding the physical transition of firn between these different thermal regimes.

We present the application of a modified version of the spatially distributed Coupled Snow and Ice Model in Python (COSIPY) to the high-altitude glacierised saddle of Colle Gnifetti (4,450 m a.s.l.) of the Monte Rosa massif, Swiss/Italian Alps. Forced by an extensively quality-checked meteorological time series from the Capanna Margherita (4,560 m a.s.l.), with a distributed accumulation model to represent the prevalent on-site wind scouring patterns, the evolution of the cold firn’s thermal regime is investigated between 2003 and 2023. Our results show a continuation of previously identified trends of increasing surface melt at a rate of 0.54 cm w.e. yr ⁻², representing a doubling over the 21-year period. This influx of additional meltwater and the resulting latent heat release from refreezing drives englacial warming at a rate of 0.045 °C yr ⁻¹, comparable to in-situ measurements. Since 1991, a measured warming of 1.5 °C (0.046 °C yr ⁻¹) has been observed at 20 m depth with a marked inversion in the temperature gradient developing in the 15-30 m depth range. While this remains below the local rate of atmospheric warming (0.073 °C yr ⁻¹), in lower altitude regions (∼ 4,300 m a.s.l.) simulated warming is considerably greater suggesting a rapid transition from cold to temperate firn is occurring – potentially indicative of future conditions at Colle Gnifetti. However, uncertainty is high in this region as the simulation is particularly sensitive to changes to the model’s parameterisations – principally those controlling albedo and percolation – and crucially the length and simulated depth of the model spin-up.

Our research also greatly contributed to the development of the latest version 2.0 of the COSIPY model which includes critical bug fixes, the addition of new parameterisations and performance enhancements to benefit the wider modelling community.