Sea ice and the Arctic Atmosphere: Trends, Relations and Teleconnections.

This study statistically analyses Arctic sea ice, sea level pressure (SLP) and large scale circulation systems throughout the period 1983-2013. The Arctic is divided into nine regions, and the year is split into four seasons, so that each season in each region can be separately analysed for trends a...

Full description

Bibliographic Details
Main Author: Lees, Kirsten
Format: Thesis
Language:English
Published: Apollo - University of Cambridge Repository 2015
Subjects:
Online Access:https://dx.doi.org/10.17863/cam.8764
https://www.repository.cam.ac.uk/handle/1810/263422
Description
Summary:This study statistically analyses Arctic sea ice, sea level pressure (SLP) and large scale circulation systems throughout the period 1983-2013. The Arctic is divided into nine regions, and the year is split into four seasons, so that each season in each region can be separately analysed for trends and correlations. Across the whole Arctic a negative trend in sea ice of 3.9% per decade is found, but a positive trend in SLP. The focus on seasonal trends shows that the September ice decrease is the greatest across the whole Arctic, although this is not the case in every region, and the focus on regional patterns shows that the Bering Sea in fact has an increasing ice extent across the study period. The SLP appears to have a greater impact on sea ice than vice versa, especially in summer, although the autumn sea ice does show some important feedbacks onto the winter SLP pattern. There appear to be negative trends in the North Atlantic Oscillation (NAO), the Arctic Oscillation (AO) and the El Nino Southern Oscillation (ENSO), but a positive trend in the Arctic Dipole (AD). The NAO/AO pressure pattern is confirmed to show low pressures within the Arctic Circle and high pressures at lower latitudes when in a positive phase. In general the NAO and AO are considered to be similar and strongly related in their effects on the Arctic SLP and sea ice, however the AO seems to link more strongly with SLP whilst the NAO has more of an effect on sea ice. Principal Component Analysis (PCA) confirms this, showing that the first principal component (PC) of SLP has a clear AO-type pattern. The effect of a positive NAO/AO pattern seems to be to increase sea ice across the Arctic, and the NAO also creates a pattern known as the Greenland Sea-ice Dipole (GSD). The AD is defined as the second PC of SLP, and is strongly connected to sea ice throughout the seasons but particularly in summer. In general a strongly positive AD causes ice decreases across the Arctic, which then creates a positive feedback. The ENSO is less well-connected to the Arctic than the NAO/AO and the AD, but nevertheless does show some links. Whilst the other patterns are most influential in summer, the ENSO has greatest impact in winter and spring. A positive ENSO index appears to be linked with lower pressures and ice extents in the Pacific sector, but higher pressures and ice extents on the North American/Atlantic side of the Arctic. The ENSO also has a negative correlation with the AD in several seasons.