Key as in Table 6.5
Key as in Table 6.5
Figure 6.10a. Year-to-year course of the seasonal and annual mean anomalies of P at Danmarkshavn and their trends over the period 1949-1990.
a) year-to-year course; b) running 10-year mean; c) linear trend over the whole period of observation; d) linear trend over the period 1961-1990.
Figure 6.10a. Year-to-year course of the seasonal and annual mean anomalies of P at Danmarkshavn and their trends over the period 1949-1990.
a) year-to-year course; b) running 10-year mean; c) linear trend over the whole period of observation; d) linear trend over the period 1961-1990.
Similar to T, the winter, summer, and annual regression equation of P, limits of the confidence interval of regression coefficients, and other statistical characteristics (including the share of the linear trend in the general variability of P), were set for selected stations representing particular climatic regions and sub-regions of the Arctic (Table 6.7). Analysis confirms the stability of the decreasing tendency in the changes in P in the Russian Arctic. The value of the 40-year annual decrease in P at Ostrov Vize lies within the limits of -126 mm to -20 mm, with 95% probability. This interval is set by the values -230 mm and -79 mm for Mys Kamenny, and by -96 mm and -33 mm for Ostrov Kotelny. Very high values of t-statistics, exceeding 4.0, were computed for the latter two stations. In series where n amounts to 40, critical to values for the three analysed intervals of significance 0.05, 0.01, and 0.001 are 2.02, 2.70, and 3.55, respectively. At the remaining Arctic stations, limits of the confidence interval oscillate between negative and positive values, and are not statistically significant. At these stations, the durability of P trends is significantly less than at the stations analysed earlier. The last column of Table 6.7 indicates that in such cases, regression eliminates a small part of the variability in P (<10%), The share of statistically significant trends in the general variability of P is greater than 10%, and in some cases it slightly exceeds 30%.
Station |
Period |
Coefficients ol regression equations a (mm/year) ¿(mm) |
ConlIdence limits ol coefficient a mrn/year lower upper |
Standard error of dependent var mm |
fstalislic |
Share of linear trend In total variation ol P % | ||
Was this article helpful? |
Post a comment