Arctic snow and sea ice response to melt season atmospheric forcing across the land-ocean boundary

Concern over the rapid changes in the Arctic cryosphere in recent years has spurred much research into the response of sea ice and snow cover to warming temperatures and the resulting climate feedbacks. However, the vast majority of Arctic climate studies do not assess the response of both continental snow cover and sea ice in concert through the data record. This study is focused on three study regions located throughout the Arctic in North America, western Russia, and eastern Russia (Siberia) over the passive microwave record (1979-2007). Each study region is divided by a land-ocean boundary that is roughly parallel to latitude and extends northward into an Arctic marginal sea that is subject to considerable inter-annual variability in the extent and retreat of sea ice during the warm season. The two Russian study regions contain mainly thin first-year ice, while the North American Region has a large portion of thick multi-year ice. The regions were also selected to include a minimal extent of mountainous areas within the continental portion where persistent snowfields throughout the year could misrepresent the seasonal progression of continental snow cover lost, relative to other land domains in the study.

Aerial extent anomalies within the domain boundaries were calculated from the monthly visible snow cover (available from the Rutgers University Global Snow Lab) and monthly Bootstrap algorithm sea ice concentrations (available from the National Snow and Ice Data Center) using a 50% concentration threshold to allow for consistent comparisons across the two data sets. Reanalysis-2 mean sea level pressures, 500 hPa geopotential heights, surface U and V vector wind components, and 925 hPa air temperatures used as a surface temperature proxy (data available from NOAA-ESRL Physical Sciences Division) were averaged and anomalies were calculated over the full extent of each study region.

The analysis of monthly snow cover and sea ice extent anomalies indicates that, on average, sea ice extent is lost earlier in the year (in May) than continental snow cover extent (in June) in all three study regions over the study period. This suggests that over the land-ocean boundary, the assumption that loss of the snow and sea ice cover progresses northward through the melt season is false. Instead, it appears that the snow cover and sea ice respond differently to melt forcing.

The biggest atmospheric contributor to anomalous snow cover extent loss appears to be warmer 925 hPa air temperatures in months with below freezing mean monthly 925 hPa air temperatures. The greatest contributors to sea ice loss appear to be strong winds and warmer 925 hPa air temperatures during mid-melt season months when mean monthly 925 hPa air temperatures are still below 0°C. Although one would expect a southerly wind component to advect warmer 925 hPa air temperatures into the study region leading to more melt and sea ice loss, a cooler 925 hPa air temperature coincident with a northeasterly wind direction due to the presence of lower pressure located over the southern portion of the region is also very effective at forcing anomalous sea ice extent loss.

The results of this study set the framework for continued analyses of the influence of snow cover loss on sea ice extents, since the snow cover is completely reduced before the sea ice extents in the months following the initial melt. Large inter-annual variations in snow cover and sea ice extent occur in all three study regions, however, as a result of this study the annual loss of snow and sea ice are each dependant on numerous factors in addition to direct atmospheric forcing. The loss of snow cover extent varies within the region as well as between regions due to local geography and the regional climatological mean temperature. While sea ice extent loss varies with the sea ice motion, type, and thickness. These characteristic differences result in the complete loss of snow cover prior to sea ice extent minima, which may influence the remaining sea ice loss nearing the end of the melt season. Additionally, further resolving the atmospheric conditions controlling the loss of snow cover and sea ice extents during the melt season both spatially and temporally could lead to determining the predictability of Arctic cryospheric feedbacks to a more complex degree in response to changing climatic conditions.