PALEOCLIMATE AND STRUCTURE - THE MOST IMPORTANT FACTORS CONTROLLING SUBSURFACE TEMPERATURES IN CRYSTALLINE ROCKS - A CASE-HISTORY FROM OUTOKUMPU, EASTERN FINLAND

The present work contributes to the study of heat-transfer mechanisms in crystalline bedrock. We present evidence from a thoroughly investig ated case history in Outokumpu, eastern Finland, in the Fennoscandian (or Baltic) Shield, which shows that the subsurface temperature held i s controlled by the thermal conductivity structure and downward diffus ion of palaeoclimatic ground-temperature variations. The subsurface te mperature profiles from three continuously cored boreholes (790-1100 m deep) were used for a detailed 2-D modelling of structural, hydrogeol ogical and palaeoclimatic effects on the subsurface temperature held T he boreholes are situated in a subdued topography and they intersect E arly Precambrian folded lithologies, such as mica gneisses, black schi sts, skarn rocks, quartzites and serpentinite-talc rocks. Finite-diffe rence techniques were used in the numerical solution of the heat- and mass-transport equations. The 2-D models of thermal conductivity and h ydraulic permeability that were compiled were based on an extensive se t of data relating to geological structure, in situ hydraulic permeabi lity, groundwater composition and thermal conductivity of drill-core s amples down to about 1 km depth. It was found that the thermal effect of topographically driven groundwater convection is very small, with P eclet numbers typically of the order of 10(-4)-10(-5). The effect of a nisotropy of thermal conductivity was found to be one order of magnitu de smaller than the effects of heatflow refraction in the inclined roc k layers which produce horizontal components of heat-flow density rang ing from -15 to 5 mW m(-2) (basal heat flow 35 mW m(-2)). Since the ad vective heat transfer could be neglected, we tried to find a palaeocli matic ground-temperature history that would explain the measured data. First, an inversion algorithm based on heat conduction in laterally h omogeneous media was used for the reconstruction of palaeoclimatic gro und-temperature history, but it yielded spurious results because of th e strong 2-D conductivity effects. The only means to reconstruct the p ast ground-temperature changes was forward modelling using a conductiv e transient 2-D model with time-dependent surface-temperature variatio ns. The surface temperatures at different times (between 100 000 yr BP and the present) and the basal heat-flow density were varied in order to reach a reasonable fit between the modelled and measured borehole temperatures. To reach this goal, the ground-temperature history must include the main climatic events during the last 100 000 yr, such as t he last glacial epoch and the Little Ice Age (from about 1300 AD to 17 00 AD), followed by warmer temperatures in more recent times. Our resu lts suggests that subsurface temperatures in conditions similar to tho se of the Outokumpu case can yield a wealth of palaeoclimatic informat ion when an appropriate approach for modelling heat transfer in the be drock is chosen.