Warm and dry föhn winds on the lee side of the Antarctic Peninsula (AP) mountain range cause surface melt that can destabilize vulnerable ice shelves. Topographic funneling of these winds through mountain passes and canyons leads to localized wind-induced melt which is difficult to identify without direct measurements. Our Föhn Detection Algorithm (FonDA) identifies the surface Föhn signature using data from twelve Automatic Weather Stations on the AP and uses machine learning to detect föhn in 5km Regional Atmospheric Climate Model 2 (RACMO2.3p2) output and ERA5 reanalysis data. We estimate and compare the climatology and impact of föhns on the AP surface energy budget, surface melt pattern, and melt quantity from 1979-2018. We show that föhn-induced melt is strongest at the eastern base of the AP and the northern portion of the Larsen C ice shelf. We identify previously unknown wind-induced melt possibly katabatic in nature on the Wilkins and George VI ice shelves. Neither RACMO2 nor ERA5 datasets show a significant increase in föhn melt thus far despite a more positive Southern Annular Mode and increasing surface temperatures. The warming climate and associated southward shift of westerly winds on the AP suggest a likely increase in the wind-induced melt that can densify firn, form melt ponds, and weaken ice shelf stability, however that trend remains insignificant for the past 40 years.

Larsen A and B ice shelves were affected by surface melt which preconditioned them for rapid disintegration due to hydrofracture and densification. Recently, warm and dry foehn winds have been discovered to melt the vulnerable Larsen C Ice Shelf (LCIS) surface via sensible heat transfer during polar night. The climatological extent and intensity of polar night surface melt and their effects on the ice surface energy budget are unknown. Here we quantify the spatial pattern and temporal variability of foehn winds and associated melt events during polar night to understand the ice shelf surface mass balance and indirect implications for ice shelf vulnerability. Our Foehn Detection Algorithm (FonDA) uses events identified from in situ Automated Weather Stations (AWS) to calibrate foehn detection from reanalysis data covering all of Antarctica and Greenland. We present a climatology of foehn-driven surface melt days, melt water equivalent, fraction of melt that occurs during polar night, and the surface energy budget. Preliminary results show that foehns perturb sensible heat fluxes by up to 300 Wm-2 and surface air temperatures by up to 13 °C in the absence of shortwave radiation.