Has Northern Hemisphere Heat Flow Been Underestimated?

We present three lines of evidence to suggest the hypothesis that heat flow in the northern hemisphere may have been underestimated by 15 to 60 percent in shallow wells due to a large post-glacial warming signal. First, temperature vs. depth (T-z) measurements in parts of Europe and North America show a systematic increase in heat flow with depth. This phenomenon is best recognized in analyses of deep (greater than 2km) boreholes in non-tectonic regions with normal to low background heat flow. In Europe, the increase in heat flow with depth has been observed by analysis of more than 1500 deep boreholes located throughout the Fennoscandian Shield, East European Platform, Danish Basin, Germany, Czech Republic, and Poland. There are significantly fewer deep boreholes in North America, but the increase in heat flow with depth appears in a suite of 759 sites in the IHFC Global Heat Flow Database for the region east of the Rocky Mountains and north of latitude 40 N. Second, surface heat flow values in southern hemisphere shields average approximately 50 mWm-2, but surface heat flow values in northern hemisphere shields average 33 mWm-2. Unless crustal radioactivity or mantle heat flow or both factors are greater in southern hemisphere continents, there is no reason for the northern and southern shield areas having similar ages to have different heat flow values. Third, two recently published surface heat flow maps show anomalously low heat flow in the Canadian Shield in a pattern that is coincident with the Wisconsinan ice sheet. The coincidence of low heat flow and ice accumulation has no geophysical basis, thus the coincidence may suggest the existence of a transient signal caused by a warming event. Recent studies of heat flow in North America indicate that in several sites, the ice base temperature was close to the pressure melting point. We hypothesize that there may have been cold ice-free periods during the Pleistocene that would account for the apparent colder surface temperatures. If our first hypothesis is correct, a majority of northern hemisphere heat flow values require revision by as much as 60 percent because they were determined from boreholes too shallow for recognition of the gradient disturbance caused by a large post-glacial warming signal. Consequently, estimates of the total global heat flux may need revision by as much as 20 to 30 percent due to the underestimate of heat flow in the northern hemisphere. A critical aspect of such a revision is that in northern continental regions where a linear relation between heat flow and radioactive heat generation has been observed, revision of heat flow values would require a significant change in the estimate of mantle heat flow.