Variability of Arctic sea ice and the relationship with the large-scale climate patterns

Variability of Arctic sea ice and the relationship with large-scale climate patterns

Abstract:

Arctic sea ice (ASI) has been declining significantly in recent decades, especially in the summer, which is likely attributed to the Arctic warming at twice the global rate since the 1980s, a phenomenon known as the Arctic Amplification. Besides global warming, large-scale climate patterns also affect the ASI, but the physical mechanism is complicated and not well understood. An analysis of the ASI by quantile regression shows that a decreasing trend of the ASI extent exists at all quantile levels, and the ASI extent in warm seasons has been decreasing faster than the mean annual rate. To analyze effects of climate patterns on ASI, dominant oscillations of sea ice concentration (SIC) were extracted by applying Principal Component Analysis (PCA) on ASI transformed by a Morlet wavelet. Results shows that SIC is more strongly correlated with climate patterns such as the Arctic Oscillation (AO), North Atlantic Oscillation (NAO), Pacific/North American Pattern (PNA) and ENSO (Nino3.4) at 1-3 years rather than at 3-8 years (interannual) or 8-32 years (interdecadal). A composite analysis shows that in cold seasons, under the positive phase of AO and ENSO (El Niño), the sea level pressure SLP in mid-latitudes of Asia and the north Atlantic is lower than normal, and is positively correlated to the PC1 of SIC. It seems that the negative relationship between AO, ENSO and SIC is due to anomalies of SST and SLP over the mid-latitudes of Asia and the Atlantic Ocean. In warm seasons, under the positive phase of NAO and PNA, SST in mid-latitudes of the Pacific Ocean and off eastern North America are lower than normal while SLP in the North Pacific Ocean is higher than normal. It seems that the PC1 of SIC is teleconnected to NAO and PNA through anomalous SST and SLP in the mid-latitudes of the Pacific Ocean.