Background

Intense meso-scale cyclones known as polar lows are frequently observed in the Arctic sector of the North Atlantic Ocean. During winter, cold-air outbreaks may be triggered by the large scale atmospheric flow exposing dry and very cold air to the relatively warm ocean surface. One of the ocean areas mostly favoured for polar low development is between the Norwegian mainland and the Svalbard island. Due to the presence of the North Atlantic Current (NAC) the waters in this area may be warmer than 6 deg C in January. The large air-sea temperature differences experienced during cold-air outbreaks result in the formations of atmospheric fronts, large heat fluxes and sometimes the development of polar lows. As polar lows are almost exclusively observed over open waters and are characterised by deep convective clouds, it is usually assumed that atmosphere-ocean heat and moist exchange are important contributions to the energetics of polar lows. The striking similarities to Tropical hurricanes displayed by the satellite images, such as clear eye and spiral bands of deep-convective clouds, has lead several authors to suggest the release of latent heat as one major energy source for at least a class of polar lows. studied a polar low south of the Bear island and concluded that in the final phase of its development some polar lows could act as Carnot engines, working between a warm (the ocean) and a cold (the tropopause) reservoir, and referred to these as Arctic hurricanes.

Still, it is important to keep in mind the baroclinic nature of cold-air outbreaks. Often, polar lows develop from disturbances on Arctic fronts that are formed during cold-air outbreaks when dry Arctic air is exposed to oceanic heat fluxes. The Arctic fronts are characterised by strong horizontal temperature gradients in the lower troposphere and an intense low level jet. In such cases the presence of the front suggests that baroclinic instability plays a role in the polar low formation. Nonetheless, the deep convective clouds revealed at the same time by satellite images suggests that large amounts of latent heat must be released in the troposphere. Possibly, both baroclinic and diabatic processes are important during the early stages of polar low development. Pure Arctic hurricanes as described by Emanuel and Rotunno (1989) may represent the final stage for vertically aligned symmetric cyclones with no baroclinic potential left.

Despite the fact that polar lows are almost exclusively marine phenomena, surprisingly few investigations of the interaction process between polar lows and the ocean. A noteworthy exception is PhD thesis by Linders (2009) who investigated the role of ocean surface temperature and air-sea fluxes on the development and intensity of polar lows. Saetra et al. (2008) demonstrated that during winter, the NAC sub-ducts under colder and less saline waters, leading to a warm subsurface core under cold upper-level waters. During strong wind events such warm sub-surface waters may entrain the surface by intense turbulent mixing. In the paper by Saetra et al. (2008), microwave satellite images revealing rapid surface warming during a polar low event outside the coast of Norway is presented. These finding raises some interesting questions. How may a rapid surface warming affect the life-cycle and intensity of polar lows? Also, a surface warming by entrainment of warm waters to the surface represents a cooling of the ocean by increased surface fluxes of latent and sensible heat. Is this a significant contribution to the cooling of the NAC and subsequent ocean overturning?

The oceanic response to hurricanes has long been recognised (Price, 1983; Sanford et al., 1987; Brink, 1989). Strong turbulent mixing entrainment of cold waters from deep layers leads to a cooling of the sea-surface. This rapid surface cooling reduces the surface fluxes and inhibits further hurricane intensification. When hurricanes moves over deep cores of warm waters, such as the Loop Current in the Gulf of Mexico, or warm core rings this surface cooling is strongly reduced. The warm water will then act to insulate the entrainment of cold waters form even deeper layers (Hong et al., 2000; Shay et al., 2000). In such cases, strong hurricane intensification has been observed. In 2005, Katrina intensified into a category 5 hurricane as it entered the warm Gulf of Mexico (Kafatos et al., 2006).

The ocean surface warming reported by Saetra et al. (2008) has only been observed by microwave satellite data. During could air outbreaks the ubiquitous cumulus convection prevents the sea-surface to be observed by infrared sensors (IR) such as AATSR, AVHRR and MODIS. However, verification of such ocean response to polar lows is urgent. Here, we propose to use altimeter combined with SST products from both microwave and infrared sensors to investigate possible surface warming in connection with polar lows. As the altimeter measures the surface anomaly (SLA) this can be related to the ocean heat content.