Onset and end of the summer melt season over sea ice: Observations of thermal structure, surface energy fluxes, and synoptic influence

In the Arctic, the date of onset and the length of the summer melt season has been proposed to impact local and large-scale environmental features such as the growing season, treeline, vegetation, and animal behaviour; subsequent winter ice thickness anomalies; the establishment of fall and winter atmospheric general circulation patterns in the northern hemisphere; mass loss from the Greenland ice sheet; the amount of solar energy absorbed by Arctic sea ice during the subsequent summer; and the sea-ice volume variance in subsequent seasons. The onset and end of the summer melt season determines the length, and the length of the melt season has been observed to be increasing. The precise physical mechanisms for the melt transitions are unknown, as are the mechanisms by which the melt season is lengthening. Observations from a variety of on-ice field programs over the Arctic sea ice are studied to better understand the mechanisms producing the summer melt transitions, and to suggest the mechanism(s) by which the melt season lengthening may occur.

The examination of the SHEBA data set shows that the melt-season transitions are evident in various atmospheric and ice parameters and that melt transitions are associated with changes to all terms of the surface energy budget. In particular, changes occurring in downwelling longwave radiation and surface albedo are key. Furthermore, it is evident that melt transitions are driven by atmospheric processes over at least a several week period, with specific transition events apparently triggered by atmospheric synoptic events and associated clouds. As for SHEBA, observations from the Arctic Ocean Expedition (AOE-2001) and Arctic Summer Cloud-Ocean Study (ASCOS) also show the importance of synoptic events impacting the downwelling longwave radiation and the surface albedo for producing end-of-summer melt events. The impact of these synoptic events is even evident in the ocean temperatures below the sea ice. Analysis of data from the Soviet drifting stations allows a multi-decade generalization of the detailed results from the SHEBA, AOE, and ASCOS field programs. This analysis showed that half of the melt onset cases occurred in conjunction with above freezing air advected over the sea ice at 850 mb in association with synoptic events. Similarly, half of the end-of-melt events occurred in conjunction with an above-freezing air event aloft, similar to that occurring at SHEBA. The Soviet data also show that the melt onset occurred earlier in later years at a rate of 3.1 days per decade (significant at the 97% level), while the end-of-melt dates occurred later in later years at a rate of 1.5 years per decade. However, the end-of-melt trend was not significant. Of special significance, the change towards earlier melt onsets and later end-of-melt dates was more rapid and more significant when only considering the events with linkages to above freezing air aloft and synoptically forced events. Hence, one mechanism by which the onset of melt can come earlier is through the northward advection of warm air from lower latitudes above the inversion. This mechanism is less dependent on the annual cycle of solar radiation than the other melt-season transition scenarios, and hence makes sense as a process leading to the lengthening of the summer melt season over sea ice.