Polar föhn winds and warming over the Larsen C Ice Shelf, Antarctica

Recent hypotheses that the foehn effect is partly responsible for warming to the east of the Antarctic Peninsula (AP) and enhanced melt rates on the Larsen C Ice Shelf are supported in a study combining the analysis of observational and high resolution model data. Leeside warming and drying during foehn events is observed in new aircraft, radiosonde and automatic weather station data and simulated by the UK Met Office Unified Model at ~1.5 km grid spacing (MetUM 1.5 km). Three contrasting cases are investigated. In Case A relatively weak southwesterly flow induces a nonlinear foehn event. Strongly accelerated flow above and a hydraulic jump immediately downwind of the lee slopes lead to high amplitude warming in the immediate lee of the AP, downwind of which the warming effect diminishes rapidly due to the upward ‘rebound’ of the foehn flow. Case C defines a relatively linear case associated with strong northwesterly winds. The lack of a hydraulic jump enables foehn flow to flood across the entire ice shelf at low levels. Melt rates are high due to a combination of large radiative heat flux, due to dry, clear leeside conditions, and sensible heat flux downward from the warm, well-mixed foehn flow. Climatological work suggests that such strong northwesterly cases are often responsible for high Larsen C melt rates. Case B describes a weak, relatively non-linear foehn event associated with insignificant daytime melt rates. Previously unknown jets – named polar foehn jets – emanating from the mouths of leeside inlets are identified as a type of gap flow. They are cool and moist relative to adjacent calmer regions, due to lower-altitude upwind source regions, and are characterised by larger turbulent heat fluxes both within the air column and at the surface. The relative importance of the three mechanisms deemed to induce leeside foehn warming (isentropic drawdown, latent heating and sensible heating) are quantified using a novel method analysing back trajectories and MetUM 1.5 km model output. It is shown that, depending on the linearity of the flow regime and the humidity of the air mass, each mechanism can dominate. This implies that there is no dominant foehn warming mechanism, contrary to the conclusions of previous work