Figure 5.12. Year-to-year course of the annual (1) and 5-year running (2) mean anomalies of T for the zones:
(b) 65-85°N (after Alekseev and Svyashchennikov 1991)
a cooling was observed, which ended in ca. 1917 (Figures 5.12b and 5.12c). From that year on there was an intense warming of the Arctic, which reached its maximum in the second half of the 1930s (in 1938 the annual anomaly of T was 1.21°C). According to the data analysed by the Arctic and Antarctic Research Institute at St. Petersburg, in the latitude band 70-85°N the maximum of the warming occurred also in 1938 (Figure 5.12a). The annual anomaly of T calculated was 2.3°C (Dmitriev 1994). The difference of T between the 1880s and the 1930s amounted to ca. 2.5°C in the winter, 1.7°C in the spring, 1.65°C in the autumn, and 1.30°C in the summer (Kelly et al. 1982). In some areas of the Arctic the changes were significantly greater. A good example is Spitsbergen, where in the period 1917-1922 there occurred a sudden increase in the mean T by 7°C for the winter, by 2°C for the summer, and by 4°C for the annual mean T (Hesselberg & Johannessen 1958). This warming occurred universally in the whole area of the Arctic (Figures 5.13-5.17). This is also confirmed by numerous analyses of meteorological data from various parts of the Arctic (Lysggard 1949; Hesselberg & Johannessen 1958; Pietrov 1959; Thomas 1961; Steffensen 1969, 1982; Bradley 1973b; Burns 1973; Pietrov & Subbotin 1981; Brazdil 1988; Hanssen-Bauer et al. 1990; Nordli 1990; Przybylak & Usowicz 1993, 1994; and others). The consequences of such a catastrophic warming were, among others, a significant decrease of the area and thickness of sea-ice, changes in atmospheric circulation, the retreat of glaciers, and a northward migration of flora and fauna. According to Aleksandrov et al. (1986), warming in lower geographical latitudes occurred later than in the Arctic and was several times less.
The first signs of cooling were observed in the 1940s most conspicuously in the Atlantic sector of the Arctic (Pietrov 1959; Zakharov 1976; Lamb 1977; Wigley et al. 1981). The analyses of the climatic variability in the Arctic up to the middle of the 20th century, carried out by various researchers, are significantly convergent.
126.96.36.199 The variability of T in the Arctic over the period 1951-1990
Unfortunately, according to Aleksandrov et al. (1986), the evaluation of the tendencies of T in the Arctic in the second half of the century, and particularly in the last 30 years, is equivocal (unlike the evaluation of the period discussed above). According to the author, there are at least two reasons for this divergence in evaluation. The first is the different lengths of the series of T chosen for analyses. As demonstrated in Figures 5.13-5.19 and in earlier publications by Przybylak and Usowicz (1993, 1994), the sign (±) and magnitude of a given trend is highly dependent on the starting and finishing points of the data series for which the trend is calculated. The finishing point, characterised by small changes of T in the Arctic and in most contemporary analyses located in the 1980s or 1990s, is of less significance. It is the starting point that is of primary importance. If it is set at the beginning or in the middle of the 1960s, when a significant cooling occurred in the Arctic, the resultant 30- and 20-year-long trends of T are most often positive (Figures 5.13-5.19; Jones 1988a; Chapman & Walsh 1993). If the calculation of a trend begins with the data from the 1950s or earlier, when there was the greatest 20th-century warming in the Arctic, the resultant trends of T are predominantly negative (Figures 5.13-5.19; Kukla et al. 1992; Kahl et al. 1993a, b). The second reason for the discrepancy in evaluation may be the fact that so far the southern border of the Arctic has not been delimited unequivocally. The review of the relevant literature carried out by Aleksandrov et al. (1986) shows that the authors most often consider as the Arctic the latitudinal zone of 60-90°N, less frequently the zones 65-90°N, 72.5-87.5°N, 70-85°N, or the area delimited on the basis of climatic criteria. Thus, regions reacting in drastically different ways to global warming are included in the analysis of mean T of the Arctic (cf. Chapman & Walsh 1993). A particularly large distortion of the actual climatic variability results from counting as the Arctic the Subarctic continental regions, which were characterised by significant warming in past years, while T in the Arctic (delimited on the basis of climatic criteria) underwent minor changes.
A major research problem is determining whether the currently observed warming of the earth can also be noted in the Arctic and whether it is, in fact, significantly greater than in other parts of the globe, as has been suggested by climatic models (IPCC 1990, 1992). In order to carry out the project thus formulated, it has been decided to examine in detail the trends and fluctuations of T (T ., T., and T ) in the Arctic, the boundaries of which were v mm i max' '
delimited on the basis of climatic criteria (Atlas Arktiki 1985). All the documentary material is shown in Tables 5.10-5.15 and in Figures 5.12-5.28. The analysis will begin with an examination of the nature of the trends of T in the Arctic, starting with the longest periods and ending with the trends for the period 1971-1990. The calculations ofthe trends for the longest periods (19221990 and 1936-1990) can only be made for a small number of stations with the longest observation series (Table 5.10, Figures 5.13-5.17). For comparison, at the bottom of the table the results of the trends and are also shown, according to Jones (1994, and personal communication ) and other parameters determining the Arctic and the global climatic system.
The analysis of the data from this table demonstrates that the annual T. trends over the periods 1922-1990 and 1936-1990 are negative for all stations except Barrow in Alaska. What is more, the majority of them are statistically significant. The results of the analysis of seasonal trends, however, are more complicated. In the period between 1922 and 1990 negative trends, similar
1940 1950 1960 1970 1980 1990 1940 1950 1960 1970 1ÏK0 19«!
1940 1950 1960 1970 1980 1990 1940 1950 1960 1970 1ÏK0 19«!
Was this article helpful?