• permafrost modelling;
  • climate change;
  • BTS;
  • mountain permafrost


The basal temperature of snow (BTS) method was used to predict the distribution of permafrost within a mountainous basin located in the southern Yukon Territory. A modelled BTS surface, based on several hundred measured values, was created within a Geographic Information System (GIS) environment using elevation and potential incoming solar radiation as independent variables. The distribution of frozen ground at 200 test sites was compared to the modelled BTS values using logistic regression. The resultant map of permafrost probability shows that all four conventional permafrost distribution classes (isolated patches, scattered and widespread discontinuous permafrost, and continuous permafrost) are present within the basin. Supplementary logistic regression analyses reveal that at certain elevations and aspects, the probability of permafrost occurrence varies markedly over short distances in response to snowpack depth. They also show that widespread alterations in snow cover would be expected to substantially affect permafrost distribution even if air temperatures were to remain unchanged. Copyright © 2004 John Wiley & Sons, Ltd.