Spatial Representativeness of Climatic Data from Baffin Island, N.W.T., with Implications for Muskoxen and Caribou Distribution

Abstract

Climatic records from often widely scattered arctic stations are commonly used to draw conclusions about such matters as wildlife habitat and distribution, yet little is known about the validity of extrapolating from such limited data. Statistical tests of meteorological network representativeness can be applied where station numbers and record length permit. Such an analysis was carried out on 29 years of temperature and precipitation data from five stations in central Baffin Island and Foxe Basin. It was found that seasonal temperatures and degree-days correlate highly across the region, indicating that interpolation and extrapolation can be carried out with confidence. Such was not the case for rainfall, snowfall, and depth of snow cover, all of which showed large extrapolation errors over modest (mesoscale) distances. This is attributed to the intrinsic variability of precipitation in the region and, in the case of snowfall and snowdepth, the inadequacy of the measurement method. The result have been applied in an evaluation of suggestions concerning climatic constraints on the distribution and numbers of muskoxen and caribou, with the conclusion that the data do not support a causal relationship based on climate.

Key words: climate, spatial representativeness, Baffin Island, muskoxen, caribou