LsOo

-1T -15 -13-20 -18 -16-16 -14 -12-16 -14 -12-10 -8 -6

1900

1800

1700

1600 1500

1400

Figure 10,13. Variations in S!S0 for the Svalbard ice cores: (I) Westfonna. (2) Austfonna, (3) Lomonosov plateau. (4) Gronfjord-Fridtjof ice divide (after Vaikmac 1990).

The LIA was interrupted in Spitsbergen in the 16"' century. However, the pronounced warming occurred mainly in the lower located glaciers (Figure 10.13). On the other hand, on the Lomonosov plateau (1000 m a.s.l.) both warm and cold spells occurred during this period. Summer melting here was above normal only in the two first dccades of the 20"' century (Figure I0.12( I)). On the other hand, the LIA on the Nordaustlandet Island began with a significant delay, in comparison with most of the above-mentioned records, i.e. about

-1T -15 -13-20 -18 -16-16 -14 -12-16 -14 -12-10 -8 -6

1900

1800

1700

1600 1500

1400

Figure 10,13. Variations in S!S0 for the Svalbard ice cores: (I) Westfonna. (2) Austfonna, (3) Lomonosov plateau. (4) Gronfjord-Fridtjof ice divide (after Vaikmac 1990).

1600 A.D. (sec Figure 10.13). The greatest temporal asynchronicity of glacioclimatic conditions can be clearly noted between Spitsbergen and Nordaustlandet in the period 1200-1500 A.D. However, there is fill! agreement that the culmination of the LI A occurred between the 17"' and 19!h centuries, although during this period some warming phases have also been observed. For example, surprisingly high values of summer melting were noted in the first two decades of the 19lh century on the Lomonosov plateau, while in the Canadian Arctic (as has been mentioned) and also in Greenland very severe conditions prevailed. However, such an opposite tendency of air temperatures in the parts of ihe Arctic mentioned is also occurring at present (see Przybylak 1997b). The isotopic record also shows slightly lower than normal temperatures, but they are significantly higher than in the mid-191'1 century. An excellent agreement between summer melt and isotopic data (see Figures 10.12 and 10.13) exists for the mid-19"' century. The area affected by melting was almost two times lower than normal. These exceptionally cold conditions in the Svalbard caused the significant advances of glaciers. As a result, the LIA moraines are typically the most extensive and best preserved (e.g. Szupryczynski 1968; Liest0] 1969; Niewiarowski 1982; P^kala and Repclewska-Pekalowa 1990; Werner 1990, 1993; Elverh0i et al. 1995; Svendsen and Mangerud 1997). At the Austfonna (Figure 10.13(2), Nordaustlandet) this mid-19th century cooling was shifted to the turn of the 19lh and 20"1 century.

Werner (1993) also found a second period when moraines were deposited (ca. 650 years BP). It was probably connected with the evident climate deterioration in the northern part of Spitsbergen between A.D. 1250 and 1350 (see Figure 10.13(3), Lomonosov plateau). On the other hand, the southern part of Spitsbergen and also Nordaustlandet had a slightly warmer than normal climate during this period.

Summarising the results presented, one can conclude that most proxy data from the Arctic indicates that the LIA period occurred between 13001400 and 1900 A.D. In accordancc with recent findings, this cold period was interrupted by shorter or longer warm periods, which were observed in different periods, but mainly before 1800 A.D. The majority of the proxy data presented show that the greatest cooling in the Arctic occurred in the first half (the Canadian Arctic) and around the mid- or in the second half (Svalbard and probably other Eurasian Arctic islands) of the 19th century.

10.3 Period 0.1 ka - Present

It is clear from Section 4.1. that a reliable estimate of areally averaged Arctic temperature can only be offered from circa 1950. Nonetheless, in the literature many works can be found which also provide an areally averaged

'Arctic' temperature for earlier years (e.g. Kelly and Jones 1981 a-d, 1982; Kelly etal. 1982; Jones 1985; Alckseev and Svyashchennikov 1991; Dmitriev 1994). 'Arctic' is used here in inverted commas because, in reality, these series represent the temperature changes in selected northern latitude bands which, as follows from Figure 1.1, significantly differ from the Arctic as defined in this book. Until 1911 the only Arctic stations for which it has been possible to compute these scries were located in Greenland. Other data used in these analyses were taken from stations located in the Subarctic and even in the mid-latitudes. Moreover, the station coverage of these regions was very low, especially in the 19lh century and was biased towards the lower latitudes. For example, Jones (1985) states that the 'Arctic' temperature was computed from grid points (5°xl0°) covering only 6%, 10%, and 20% of the latitude band 65°N-85°N in the years 1851, 1874, and at the end of the I9,h century, respectively. This author opposes the definition of such series as 'Arctic' because such definitions lead inevitably to the identification of misleading estimates of Arctic air temperature tendencies. This is very well illustrated in Figure 10.14, from which it can be seen that a warming in the 1930s was most pronounced in the real Arctic (see the top curve which represents the real Arctic in the greatest degree). This warming is reduced when more areas from the Subarctic and from the mid-latitudes arc included in the Arctic. The sccond phase of contemporary warming (after 1975) in the real Arctic series is not seen, while in the other series it is distinct. For the whole Northern Hemisphere (bottom curve) the warming in last decades is even greater than in the 1930s.

10.3.1 Temperature Variations Prior to 1950

Przybylak (2000a) chose six stations to illustrate the variation of Arctic air temperature prior to 1950 (Figure 10.15). All of them represent the analysed climatic regions and offer long series. Figure 10.15 shows slightly rising temperatures in Greenland prior to 1920. After this time, the rate of wanning significantly increases. This trend was noticed very early in Greenland and in the Atlantic Arctic region and has been described by various authors (e.g. Knipovich 1921; Scherhag 1931, 1937, 1939; Hesse!berg and Birkeland 1940; Vize 1940; Weickmann 1942; Lysgaard 1949). The maximum temperature occurred in the 1930s and was higher by about 2-5°C than those occurring prior to the 1920s. The most pronounced rise in temperature oc-cuned in the Atlantic region and throughout the Arctic in winter. During (his season, the mean temperature rose locally by up to 9°C (Przybylak !996a, 2002a). Since the 1930s a statistically significant decrease in temperature has been noted.

Atmosphere Circulation Climate

Figure 10.14. Year-to-year course of annua! ( I. solid line) and 5-year running (2, heavy solid tine) mean anomalies of air temperature for the zones (after Przybylak 1996a): (a) 70°N-85°N (after Dmitriev 1994); (b) 65°N-85°N (after Alcksecv and Svyashchennikov 1991); (c) 60°N-90°N (after Jones 1995, personal communication); (d) 0°N-90°N (after Jones 1994).

Figure 10.14. Year-to-year course of annua! ( I. solid line) and 5-year running (2, heavy solid tine) mean anomalies of air temperature for the zones (after Przybylak 1996a): (a) 70°N-85°N (after Dmitriev 1994); (b) 65°N-85°N (after Alcksecv and Svyashchennikov 1991); (c) 60°N-90°N (after Jones 1995, personal communication); (d) 0°N-90°N (after Jones 1994).

1*70 1890 1910 1930 1950 1970 199« 1870 189« 1910 19.10 I9W 1970 1990

1950 1970 1990

1870 IS90 191«

1950 1970 1990

1870 1890 1910 1930 1950 1970 1990

IB70 1890 1910 1930 1950 1970 1990

Figure 10.15. Year-io-year course of the annual (solid tine) and 5-year running (heavy solid line) mean anomalies of air temperature in the Arctic stations having the longest observational series (after Przybylak 2000a).

All stations (except Barrow and Coppermine, representing only a small part of the Arctic) show the greatest warming in the 1930s. The reason most often given for this warming wave is a change in atmospheric circulation (see e.g. Scherhag 1931; Weickmann 1942; Pettersscn 1949; Lamb and Johnsson 1959; Girs 1971; Lamb 1977; Lamb and Morth 1978; Kononova 1982) which is now thought to have been at least partly related to the North Atlantic Oscillation (NAO). In this decade, as has recently been reported by Slonosky and Yiou (2001) the values of the NAO index were comparable to those occurring in the late 20,h century. However, patterns of temperature changes in the two periods differ, particularly in the area of western Greenland, where warming also occured in the 1930s (see Figure 10.15). The NAO and its influence on the Arctic climatc is described in sub-chapter 10.3.3. A secondary air temperature maximum can be seen in the central part of the Atlantic region (Svalbard Lufthavn) in the 1950s and in Greenland in the 1960s. This is not present in the Siberian region. Spatial coherency in Arctic temperature changes was significantly greater before the 1950s than it was afterwards (Figure 10.15).

Was this article helpful?

0 0

Post a comment