Figure 5.30. The spatial distribution of periods of cyclic oscillations occurring in seasonal and annual T and P in the Arctic. A - T.; B - T. - upper map, P - lower map; C - P.

Figure 5.30. The spatial distribution of periods of cyclic oscillations occurring in seasonal and annual T and P in the Arctic. A - T.; B - T. - upper map, P - lower map; C - P.

The series of mean regional T. were characterised by a cyclicity which was highly consistent with that presented above: ATLR - 32.0 and 2.6; SIBR -7.1 ; PACR - 9.1 and 5.3; CANR - 3.8 and 2.1 ; and BAFR 5.5, 16.0, and 2.8 years (Figures 5.31a and b). The series of 7] from the Arctic (delimited according to the boundaries assumed in the present work) has an oscillation period of 32.0 years, whereas the series of T. from the "Arctic" (additionally encompassing large Subarctic areas) are characterised by a cyclicity of 64.0 years (the series computed from 17 stations and the series according to Alekseev & Svyashchennikov 1991) (Figure 5.31c). The length ofthis cycle approximates the oscillation cycle of global (65-70 years) calculated using the SSA method by Schlesinger and Ramankutty (1994 and 1995). It is worth adding that these researchers detected, for 11 selected regions of the globe, oscillation periods changing from 9 years (the equatorial part of the western Pacific) to 88 years (North America).

As the most plausible reason for the 65-70-year oscillation, they assume the cyclicity occurring within the ocean-atmosphere system. To support this assumption, we may quote the calculations which were made of the periodicities of the occurrence frequency of circulation macrotypes W and C (according to the Vangengeim-Girs classification) amounting to 64.1 and 31.9 years respectively; for macrotype E, the oscillation periods were 31.9 and

16.0 years (Figure 5.32). The latter two cycles are the higher harmonics of the

64.1 years. Moreover, also detected is a 64.1-year oscillation period of the zonal circulation index. The dominance of atmospheric circulation in the shaping of the Arctic climate has often been underscored in the present work. Thus, it seems plausible that the 32- and 64-year cycles determined in the series of means result from the same cycles in the circulation characteristics presented above. The assumption is further confirmed by the spatial distribution of the oscillation periods of T in the Arctic, also mentioned above. Their magnitude is the closest to the circulation cyclicity in the areas where the influence of circulation is the strongest (ATLR and BAFR). It is worth adding that the mean water temperature in the Barents Sea from a depth of 0-200 m along a profile from Nordkapp to Bear Island is characterised, as Table 5.16 and Figure 5.32 show, by an identical dominant oscillation period as T in Bjornoya (18.3 years). Also of a similar order (21.3 years) is the cyclicity of the ice cover of the Barents Sea. It follows that, apart from atmospheric circulation, the two characteristics of the Arctic climatic system discussed above are also conducive to the periodicity of T detected in the Arctic. A reverse relationship also obtains between these elements, i.e., the cyclicity of the changes of T initiated by their influence induces, in turn, modifications of atmospheric circulation and temperature and sea-ice cover. This provides support for Burroughs's (1992) opinion that a possible explanation of most of the quasi-periodic characteristics observed in climatic series may be the reciprocal influence of the atmospheric variability and various feedback mechanisms "working" in the climatic system.

Figures 5.31a-c. Singular Spectrum Analysis (SSA) results for the regional T means (RI-RV) and for 7" for the whole of the Arctic (Al, A2, A3, and A4). Key as in Figure 5.29. Al - means have been calculated using data from 27 stations; A2 - means have been calculated using data from 17 stations; A3 - after Alekseev and Svyashchennikov (1991); A4 - after Jones (1994).

0.1 0,2 0,3 0,4 Frequency (rycles/year)

Was this article helpful?

0 0

Post a comment