Firn densification in East Antarctica – a detailed model-data comparison at Dome C and Little Dome CFirn densification in East Antarctica – a detailed model-data comparison at Dome C and Little Dome C

Deep ice cores from the East Antarctic plateau provide unique continuous records for paleoclimate. While some proxies are recorded in the ice phase, others are recorded in the gas phase. Air enclosure occurs at several dozens of meters below the snow surface which leads to gas being always younger than the surrounding ice. Presenting the records measured on the gas and ice phases on the same chronology relies on the determination of the lock-in depth (LID) where air is isolated from the atmosphere. Two different methods can be used to determine past LID in ice core. On the one hand, firn densification models have been developed over the past 45 years to model progressive evolution densification from surface snow to ice over the top 60 -120 m and empirical determinations permit to link firn density to the lock in process. On the other hand, measurements of d¹⁵N of N₂ in trapped air in ice cores provide information on the past evolution of the LID through the gravitationnal fractionation which leads to a linear relationship between d¹⁵N of N₂ and firn diffusive height, itself directly linked to LID in the absence of any surface convective zone.    

Here, we present independent estimate of the LID at two neighboring central sites in East Antarctica, Dome C (DC) and Little Dome C (LDC) where the EDC and BELDC deep ice cores have been drilled. We present results from different firn densification models and measurements of d¹⁵N of N₂ both in the open porosity in the upper snow and in bubbles trapped in ice over the penultimate deglaciation and last interglacial period. For both studies, the measurementsshow a coherent 10% shallower LID at LDC than at DC which is relatively large given the similar climatic conditions on these neighboring sites. The firn densification models used in this study are not able to reproduce both the LID difference of about 10% between the two sites and the LID increase over the glacial -interglacial transitions. Missing processes in the firn densification model might be related with variations in the physical properties of the surface snow and surface snow metamorphism. To explore this hypothesis, our study hence also includes high resolution profile of density and specific surface area measurements at both sites.