The influence of long-term changes in snow cover on the ground thermal regime: Implications for borehole climatology

In the past decades it has been widely recognized that geothermal data can provide complimentary information to the meteorological records of past long-term ground surface temperature (GST) trends. The inversion methods used to retrieve the climatic signal from subsurface temperatures assume a direct coupling between the surface air temperature (SAT) and GST. Because snow cover insulates the ground in the winter, systematic and persistent variations of snow cover could alter the air-ground temperature coupling and complicate the reconstruction of GST histories. We used a one-dimensional, finite differences heat transfer model with phase changes to carry out a series of synthetic simulations to examine the effect of long-term variations of the snow-cover thickness and of the snow-cover duration. Results indicate that long-term changes in snow cover can have a significant effect on ground temperatures. For seasonal frost cases, we find that GST can change by 1.5 K if the snow-cover thickness is increased by 1cm/year over a period of 200 years. A change in the snow-cover duration of 1day/decade for a period of 200 years can potentially alter the GST by 0.5 K. In areas of permafrost with similar variations in snow-cover thickness and duration, the ground surface temperatures can change much more. We find that critical depths of snow thickness are about 70 cm and 130 cm for the seasonal frost and permafrost cases respectively. Increasing the snow-cover thickness beyond these values has no significant effect on the GST. Our model suggests that for a constant annual SAT the Canadian regional snow trends of the last 50 years (www.socc.ca) would induce GST decreases of 0.5 K and 0.7 K in eastern and western Canada respectively for cases with seasonal frost and permafrost. This cooling is mainly due to a decrease in snow cover. In parts of Atlantic Canada and Nunavut, where snow trends have increased in the last 50 years, GST are warmer by 0.1K and 0.4 K respectively. If we include western and eastern Canada SAT trends in the above experiment, results show that the GST decreases by 0.5 K even though the SAT increased by 0.5 K during the past 50 years. In parts of Atlantic Canada, although the air temperature decreased by 0.4 K in the past 50 years, our model shows a GST increase of 0.2 K under the influence of an increasing snow cover. Our results suggest that changes in snow cover alone can alter the temperature of the ground. Lower or higher GST are likely a coupled response to the variation in snow cover and to changes in SAT. Early snow melt may be a significant contributor to the warming of the permafrost area.

Although long-term snow cover data are few in Canada, century-long trends do not appear significant. However, the trends of the last 50 years could potentially create subsurface transients and should be considered in the reconstruction of past climatic changes from geothermal data.