The evaluation of trends in the fluctuation of P in recent decades is no less important than that of T. As has been mentioned before, P is one of the factors shaping the mass balance of glaciers. The balance is more positive if P is higher and occurs in the form of snow. Moreover snow-cover is thicker in non-glaciated areas and in those covered with sea ice, which causes it to lay longer. This is one of the ways in which changes in the value of P also influence the diversity of the heat balance in the Arctic.
Climatic models predict that along with the doubling of C02, P in the Arctic should increase by between 5% and 20% (IPCC 1990, 1992). In the past 100 years, the globe has warmed by 0.5°C on average, and C02 concentration has risen by about 25% at the same time. In the present sub-chapter we would like to examine whether empirical data confirm the results obtained from climatic models. We have already concluded, on the basis of 10-year anomalies, that the dependence between P and T in the Arctic is contrary to that suggested by climatic models. Presently, with the seasonal and annual series of P for many of the Arctic stations at our disposal, we can examine their behaviour in the century.
Calculations of regression, confidence intervals of slopes, standard errors in dependent variables, and other statistical characteristics, including the share of the linear trends in general variability of seasonal and annual P totals, were made analogously to those carried out for T (see sub-chapter 5.1.3).
184.108.40.206 The variability of P from the start of instrumental observations in the Arctic until 1950
As has already been mentioned in sub-chapter 5.1.3, there were few stations in the Arctic until the 1920s. All of them, apart from Green Harbour (Spitsbergen), were located along the coast of Greenland. Out of the five operative stations at the end of the 19th century, only two - Godthab and Angmagssalik - have continued working up to the present day. Unfortunately, P series for these stations contain many gaps until as late as 1921 and 1911, respectively. Thus, the oldest data have been excluded from the present work.
In practical terms, not before the 1920s did more information about P become available, with the establishing of several other stations for a relatively large area of the Arctic (i.e. for the regions of ATLR, PACR, and BAFR). Stations in the remaining area were established even later: SIBR in the 1930s, and CANR after the Second World War.
The year-to-year course of annual P at the stations with the longest series of P is presented in Figure 6.9. In almost all years of the period 19211950, P totals were lower than the totals of P from 1951 to 1990 at the majority of stations. In Greenland and at Jan Mayen they were even lower on average by about 200 mm. At that time, a higher P occurred only at Ostrov Chetyrekhstolbovoy and Coppermine stations. It is probable that, in the three decades analysed, its course was also similar at Ostrov Kotelny (Figure 6.10e). Out of these three decades, the lowest P was observed in the 1920s, except for Barrow and Godthab (Table 6.1, Figure 6.9). It was in the 1930s and 1940s, which so far had been the warmest decades in the Arctic, that P was clearly lower in most regions than in the later, cooler period. At some stations (Jan Mayen, Ostrov Dikson) over the period of increased T (from 1920 to 1940), an increase in P was also observed. However, magnitudes of P totals were still lower than the mean long-term P from 1951 to 1990. The relation between these two elements is not stable and is reversed in subsequent years. The decrease in T over the period 1940-1965 did not cause a change in the tendency of P which had lasted since the 1920s. In ATLR, where the analysed stations are located, the value of P is determined to a greater degree by atmospheric circulation, and thus its changes are more important than the changes in T. At the remaining Arctic stations P remained at a similar level during the three decades analysed (Figure 6.9).
220.127.116.11 P tendencies in the Arctic from 1951 to 1990
The most accurate picture of trends in P changes can be obtained only if we calculate the trends for periods of different lengths. (This issue has been discussed in sub-chapter 5.1.3 and in Przybylak & Usowicz (1994)). It is for this reason that trends in P, similar to those in T, were calculated for five different periods: 1922-1990, 1936-1990, 1951-1990, 1961-1990, and 1971-1990.
In all the stations of the longest observation series, positive trends in annual P were observed from 1922 to 1990 (Table 6.5). They are most pronounced at Jan Mayen and at the stations in Greenland, where the calculations revealed their statistical significance at the level of 0.001. Over the same period, negative trends in P occurred only in the case of particular seasons at Ostrov Dikson (summer and autumn) and Barrow (autumn and winter) stations.
Was this article helpful?