The North Atlantic and Arctic Oscillations

The circulation changes of primary interest here involve the North Atlantic Oscillation (NAO) and the Arctic Oscillation (AO), the latter also known as the Northern Hemisphere Annular Mode (NAM). The NAO represents covariance between the strength of the Icelandic Low, an area of mean low pressure at sea level centred near Iceland, and the Azores High, an area of mean high pressure at sea level centred near the Azores. In the positive mode, both are strong. Thompson & Wallace (1998) argued that NAO should be placed in the more general framework of the AO. The AO framework describes an oscillation of atmospheric mass between the Arctic and middle latitudes. In the positive mode, pressures are low over the Arctic, most strongly expressed in the vicinity of the Icelandic Low.

Although debate has ensued as to which framework is most appropriate, both on statistical and physical grounds, for many purposes, including the present discussion, we can consider the NAO and AO as different expressions of the same basic phenomenon. They can be identified at any time of the year, but are best expressed from late autumn into early spring. The strength and phase of the NAO and AO can be described by simple indices. As evaluated using data for December through to March (Fig. 23.2), the period from about 1970 onwards has been characterized by an overall trend toward the positive modes of the NAO/AO. There has certainly been high variability over this period. Note in particular that the NAO/AO was especially positive from about 1989 to 1995, but has subsequently regressed toward a more neutral state.

The characteristics of recent Arctic warming have been widely examined (see the review of Serreze & Francis, in press). Most studies have focused on changes in surface air temperature (SAT) over land areas where the longest time series are available, some of which extend into the 19th century. Computed trends are sensitive to the period examined, the data source, and the way in which the data are analysed. However, it is quite clear that land areas have experienced warming since about 1970, at rates larger than that for the Northern Hemisphere as a whole. Although warming is evident in all seasons, it has been most pronounced over the Eurasian and North American sub-Arctic in winter and spring (Serreze & Francis, in press). As outlined in a number of studies (Hurrell, 1995, 1996; Thompson & Wallace, 1998, 2000; Thompson et al., 2000; Hurrell et al., 2003), this winter and spring warming is consistent with the change to the positive NAO/AO mode. The basic reason is that the positive mode of the NAO/AO

AO Index: Winter (DJFM) 1950-2003

Figure 23.2 Time series of the winter (December through to March) index of the Arctic Oscillation from 1950 to 2003 and linear least-squares regression line.

Year

Figure 23.2 Time series of the winter (December through to March) index of the Arctic Oscillation from 1950 to 2003 and linear least-squares regression line.

promotes advection of relatively warm air across large parts of the Arctic and sub-Arctic land area.

Trends have also been observed over the Arctic Ocean. Information comes primarily from the Russian 'North Pole' measurements for 1950-1991, collected at a series of manned drifting camps (providing only one to four monthly means per year), since 1979 from arrays of drifting buoys maintained by the International Arctic Buoy Programme (IABP) and from the late 1970s or early 1980s onwards from various satellite retrievals. An analysis of the Russian North Pole records (1961-1990) by Martin et al. (1997) showed a significant increase in Arctic Ocean SAT during May and June (0.89 and 0.43°C per decade). Rigor et al. (2000) assessed trends for the period 1979-1997 north of 60°N via gridded fields that combine land-station records with data from the North Pole stations and the IABP. In basic agreement with the Martin et al. (1997) study, positive ocean trends are most pronounced and widespread during spring. The subsequent effort by Comiso (2003) made use of clear-sky surface temperature retrievals from Advanced Very High Resolution Radiometer (AVHRR) satellite data for 1981 onwards. Based on updates using improved algorithms, the Arctic Ocean has experienced warming during winter, spring and autumn, with greatest increases in spring and autumn.

All of these studies point to an earlier seasonal melt onset and lengthening of the melt season. A longer melt season would favour less total ice cover at summer's end. Once the ice cover begins to melt, the reduction in the surface albedo provides a feed-back to foster further melt. Comiso (2003) calculates an increase in the melt season of 10-17 days per decade. Further support for extended melt comes from Belchansky et al. (2004), who base their independent analysis on passive microwave satellite retrievals. Comiso's (2003) lack of summer SAT trends is not surprising over the Arctic Ocean. Presumably, his summer trend includes the effects of an earlier June transition to the melting point.

Figure 23.3 Mean annual circulation of the Arctic sea-ice cover, based on data from the International Arctic Buoy Programme, with overlay of sea-level pressure from 1979 to 2002. Contours are shown every 1 hPa.

Figure 23.3 Mean annual circulation of the Arctic sea-ice cover, based on data from the International Arctic Buoy Programme, with overlay of sea-level pressure from 1979 to 2002. Contours are shown every 1 hPa.

As discussed, the pronounced winter and spring SAT changes over land areas reflect the direct advective effects of the change in the NAO/AO. It might be tempting to attribute spring warming over the Arctic Ocean and earlier melt onset to this same process. Indeed, one finds that the largest reductions in sea-ice extent have occurred along the coasts of Siberia and Alaska, where the rises in spring SAT have been most pronounced. Temperature advec-tion probably is playing some role. A closer analysis forces us to conclude, however, that although the ice reductions are closely tied to the NAO/AO, the primary links are actually more 'indirect'.

Was this article helpful?

0 0

Post a comment