Estimating seasonal values of thermal diffusivity in thawed and frozen soils using temperature time series

Various methods have been developed to estimate the thermal diffusivity in soils from temperature time series, but all have Limitations. The method described here is designed to obtain a bulk estimate of thermal diffusivity representative of the soil column during a season, or period of around three months. Hourly precision temperature measurements were recorded between August 1993 and August 1994 at eight probes in the active layer and near-surface permafrost at two sites in northern Alaska. Hourly temperature changes were calculated during the period when the soil was thawed, and used to compute the average temperature change rate (degrees C h(-1)) at each probe level. The diffusivity was estimated by relating the rates at each level using the ratio amplitude method. Bulk values for the thawed soils above permafrost were 1.5 and 2.1 x 10(-7) m(2) s(-1) at Happy Valley and Barrow, respectively. Several strategies were employed to verify the methodology. The results indicate that this method is most effective in the upper part of the soil column in summer where hourly temperature variations are fairly large. Thermal diffusivity has strong time and depth dependence. By reducing this inherent variability to a single value representing the thawed soil column over the entire season, information is lost. However, it does provide an estimate as input for modeling active layer thaw at high latitudes given similar soils, vegetation cover and topography.