*, %, # - trends statistically significant at the levels of 0.05, 0.01, and 0.001, respectively
*, %, # - trends statistically significant at the levels of 0.05, 0.01, and 0.001, respectively
Figure 5.28. Courses of the 5-year running means of cloudiness (1) and their linear trends over the period 1961-1990 (2) in selected stations representing the climatic regions of the Arctic.
It follows both from the above example and from the calculations of the correlation between the magnitudes of mean seasonal cloud cover and the DTR that the relationship between them is not as clear and simple in the Arctic as it is in lower latitudes. The computations showed that statistically significant negative correlations occur mostly in the summer. In the spring and autumn, they occur only in certain parts of the Arctic. In the winter, on the other hand, in most of the area examined, a positive correlation was observed between cloud cover and the DTR because at this time cloud cover is heavily influenced by atmospheric circulation. Intensive cyclonic activity, especially in the Atlantic sector of the Arctic, causes the inflow over this area of warm and humid air masses from the southern sector, i.e., from the moderate zone. It is worth adding that Przybylak (1992a), using the data for daily extreme T and cloud cover for Hornsund (Spitsbergen) between 1978 and 1983, obtained very similar results. On this basis it may be concluded that, apart from cloud cover, an equally important factor lowering the DTR in the past decade are its day-to-day aperiodic changes, caused by the variability of atmospheric circulation.
18.104.22.168.1 Fluctuations ofT and T .
Comparing the courses of fluctuations of and it may be ascer-
tained that they are similar in all stations (Figure 5.27). The trends discussed in the last sub-chapter describe the general nature of changes of these temperatures. The present chapter analyses the year-to-year variability of T extremes in the period examined (1951-1990). Figure 5.27 shows that irregular oscillations lasting a varying number of years dominate in their course. Long-term oscillations are revealed after their elimination (by means of a 5-year moving average). They are more evident in the stations located in ATLR (ca. 16-20 years) and in SIBR (ca. 13-14 years). In PACR and CANR, on the other hand, they are indistinct and, in all likelihood, manifest a cyclicity of ca. 10 and 20 years respectively. More advanced mathematical methods, however, need to be used to evaluate credibly the periodicity of these temperatures (cf. sub-chapter 5.1.4).
Similar to the case of the courses of extreme T values in particular regions ofthe Arctic differ from each other. In the western and central part of ATLR, extreme T values were evidently highest in 1951-1960 and in the next decade they were lowest (Figure 5.27). In the eastern part of ATLR (Ostrov Dikson) the general course is similar to the part of ATLR discussed above, except that here the increase of T in the 1980s is far more evident and matches the warming of the 1950s. In SIBR (Ostrov Kotelny), extreme T values continue at more or less the same level (except for a significant cooling at the beginning of the 1960s), manifesting a weak maximum at the beginning of the 1970s. Five-year moving averages in PACR and CANR (Figure 5.27) are characterised by the lowest year-to-year variability. For this reason, their maximums and minimums are poorly formed. Despite the fact that the variability of extreme T values in BAFR (similar to PACR and CANR) is low, it is possible, thanks to their significant negative trend, to identify here the warmest period (the second half of the 1960s) and the coldest (the second half of the 1980s). It is worth noting that the decrease in Tmjn in BAFR in the 1980s was much stronger than that of (Clyde A, Figure 5.27).
22.214.171.124 Trends and fluctuations of some characteristics of the Arctic climatic system
The present sub-chapter traces the changes in the past decades of sea-water temperature, sea-ice extent and thickness, and snow cover. These are features of the Arctic climatic system which are characterised by strong feedback with processes taking place in the atmosphere. Thus, every major climate change should result in reciprocal changes in the environmental elements listed above. Of special importance is tracing the behaviour of the cryosphere. Since the beginning of the 1970s satellite images, updated weekly, have been used to provide information on the cryosphere. Currently available are weekly maps depicting the concentration and extent of sea ice starting from January 1972. The maps have been produced on the basis of satellite images from the U.S. Navy / NOAA Joint Ice Center. Similar maps showing the weekly extent of snow cover have been drawn up using satellites of the NOAA (National Oceanic and Atmospheric Administration). The data have been available since 1973 (Matson & Wiesnet 1981). These data, covering the whole Arctic and Northern Hemisphere, are exceptionally valuable as indicators of climatic changes, especially in the face of the still insufficient coverage of the Arctic by meteorological stations, particularly its central part.
The year-to-year course of sea-water temperature behaves similarly to T. As follows from the data in Tables 5.17-5.19, there is a very close correlation between the two temperatures. This is why the trends computed for the temperature of the 200 m surface layer of water along a profile through the Barents Sea are predominantly similar, not only to the course of T in ATLR, but also throughout the whole Arctic (Tables 5.10 and 5.11).
The behaviour of sea ice, which separates the atmosphere from the ocean, is a result of the processes taking place in the atmosphere and the ocean, hence its special role in the Arctic climatic system. In recent years, particularly intensive research into both the extent and thickness of sea ice has been conducted (Manak & Mysak 1989; Mysak & Manak 1989; Parkinson & Cavalieri 1989;
Gloersen & Campbell 1991; McLaren et al. 1992; Barry et al. 1993; Chapman & Walsh 1993; Wadhams 1994, and others).
As far as the changes in the extent of sea ice are concerned, the most frequent conclusion is that the data covering the last 20 or 30 years of the observation period do not manifest a significant trend indicating either an increase or a decrease of its surface area (Mysak & Manak 1989; Parkinson & Cavalieri 1989; Barry et al. 1993). Chapman and Walsh (1993), on the other hand, having analysed the same data as Barry et al. (1993), ascertained the existence of a significant downward trend in the extent of sea ice in the summer. In the remaining seasons (except for the winter) the changes in the extent of sea ice are negative but statistically insignificant. The winter extent of sea ice, on the other hand, is characterised by an insignificant upward trend. From Figure 10, published by Chapman and Walsh (1993), it can be seen that the significance of the summer trend of the extent of sea ice was caused by a marked decrease in the extent of sea ice in 1988, 1989, and 1990. Through excluding these three years from the analysis, we obtain (as indicated by Parkinson & Cavalieri 1989) a weak upward trend for 1973-1987 and a weak downward trend for 1961-1987. Similar results were obtained from the analysis of the Barents Sea ice-cover (Tables 5.10 and 5.11). This clearly shows that in all the periods examined (including those far longer than the ones analysed in the literature cited) weak positive, though statistically insignificant, trends prevailed, except for the period 1961-1990. In the present work, similar to Manak and Mysak (1989) and Chapman and Walsh (1993), a strong dependence was ascertained of the extent of the Barents Sea ice on T in its area - and even in the whole Arctic - and on the temperature of sea water (Tables 5.10, 5.11, 5.19).
Recently, an examination of the variability of sea-ice thickness has been initiated on a larger scale. Unfortunately, data of this sort are sporadic and pertain to a short period. Most of them come from measurements conducted by means of sonars installed in submarines and only a very small portion come from exploratory drilling. Sonar measurements, taken using submarines at more or less at the same time of the year, are available only for some years starting from the end of the 1970s. The interpretation of the data available poses difficulties. Wadhams (1994) ascertained a statistically significant decrease in the thickness of sea ice in the Eurasian basin between 1976 and 1987. This issue was also investigated by American researchers McLaren et al. (1993), who stated, on the basis of ice thickness measurements in the North Pole, that no trend is visible between 1977 and 1990. Wadhams (1994), in turn, analysing their data, ascertained the existence of a statistically significant decrease in sea-ice thickness in the region between the 1970s and the 1980s. His final conclusion is, however, that so far measurements have not unequivocally yielded a trend connected with climatic change, though research results pertaining to the North Pole and the Eurasian basin suggest a decrease in ice thickness in the later years of our period of observations. However, even if the latter part of the conclusion is true, this does not mean that the trend is caused by the greenhouse effect or that it will persist in subsequent years.
More unambiguous results are available for the examination of the extent of snow cover. The observations of its extent in the Northern Hemisphere over the periods 1972 to 1989 (Robinson & Dewey 1990) and 1973 to 1992 (Groisman et al. 1994a, b) showed that in the latter, its extent decreased by ca. 10%, a process which occurred both in North America and in Eurasia. Robinson and Dewey (1990), on the other hand, calculated that mean seasonal extents of snow cover for most of the 1980s were lower by 3.7-8.4% than the means computed for the years between 1972 and 1980. The results of research pertaining to the area of Canada (Brown & Goodison 1993) are consistent with the findings stated above. The decrease in the snow-covered area, according to Groisman et al. (1994b), is most evident in the spring. For this reason, the greatest increase in T is observed over the continents in this season. It is worth adding that the changes in the extent of snow cover should be perceived as correlated with the climate of the continental areas in the moderate zone as the decrease in the area covered with snow occurs mostly in this area. The snow-covered area in the Arctic, on the other hand, probably does not undergo significant changes.
Summing up the above discussion, it should be said that the changes in the characteristics of the cryosphere discussed here, similar to the changes in the temperature of the air and the sea water, do not manifest any clear significant tendencies in the final decades of the period of observations.
Was this article helpful?