GROUND SURFACE PALEOTEMPERATURE RECONSTRUCTION BY EMPIRICAL BAYES: APPLICATION TO THICK CLAYEY AQUITARDS

In much the same manner as the transport of solutes through aquitards via diffusion, heat conduction theory predicts that transient temperature changes at ground surface diffuse into the subsurface and perturb the subsurface thermal profile. Although the thermal diffusion coefficient is about four orders of magnitude greater that solute diffusion coefficients, thermal profiles have been shown to contain thermal memories of ground surface temperatures (GST) for the past few centuries. In fact, analyses of temperature-depth profiles in the grassland ecosystem in the southern Canadian Prairies in Saskatchewan show warming over the last century. This is in keeping with data collected in Alberta and the Yukon by others. Large scale GST reconstructions from borehole temperatures (BHTs) clearly confirms recent global warming although there is some controversy surrounding the onset of this warming period. There are also differences in the timing and magnitude of the GST signal in Canada. To date, most borehole temperature work has been conducted on boreholes in the Canadian Shield, although not exclusively. It is our hypothesis that the thick clayey aquitards that cover large portions of the Canadian prairies well suited for this purpose. The King site in Saskatchewan is an excellent site to measure thermal profiles to a depth of 160 m below ground because the geology is very simple (consisting of two well-defined clay units and about 50 piezometers (for thermal profiling) are completed throughout the 160 m. To this end, detailed temperature profiles were performed on piezometers in 1997 and again in 2002 and 2005. At each date, the temperature profiles from the piezometers exhibited the same trends with depth. Differences were however observed between sampling dates in the upper 20 m between measurement dates. To compliment these temperature data, thermal data has been collected from piezometers installed in the till and monitored for the past several years using Troll ®systems and from nested thermocouples installed in the upper 5 m. In both cases, data is collected hourly and stored on data loggers. We also carried out an analysis of the Canadian borehole dataset which consisted of 221 temperature profiles. The reconstructed GST record shows warming between 1800 and 1949 of approximately 1.0 K, with the maximum rate of warming occurring between 1900 and 1949.