basalts). The groundwater bias cannot be responsible for the strong reconstructed warming in permafrost regions of northern Alaska (Lachenbruch and Marshall 1986) and in the semiarid U.S. and Canadian plains (Gosnold et al. 1997), although it may be partly responsible for the stronger reconstructed warming in eastern versus western North America (Pollack and Huang 2000). A comparison of borehole-based and instrumental 20th century warming trends for specific regions (Pollack and Smerdon 2004) shows no consistent offset related to precipitation: borehole-inferred warming exceeds instrumental warming in the wet regions of North America but is less than instrumental warming in the wet regions of Europe and Southeast Asia. This is further evidence that the groundwater bias is quantitatively small in borehole-based temperature reconstructions on larger scales. The borehole temperature database has been screened to eliminate other sorts of groundwater influences that are more readily apparent (Huang and Pollack 1998).
A separate issue from the uncertainty of reconstructions is that air temperature itself can change for many local reasons, including deforestation and urban expansion (Skinner and Majorowicz 1999, Majorowicz et al. 2006), as discussed in Chapter 2. Borehole temperatures do record such changes, which are real changes of local climate. The borehole temperature database has been screened to eliminate sites with urban influence. Effects of rural land use change are not eliminated and represent part of the human influence on climate in rural regions.
A small number of boreholes in the ice sheets have also been analyzed in conjunction with ice core studies (Cuffey et al. 1994, 1995; Cuffey and Clow 1997; Dahl-Jensen et al. 1998, 1999). Ice sheet boreholes permit longer timescale temperature reconstructions because of the purity of ice, the great depth of the boreholes, and the opportunity for combination with isotopic information from the ice core itself. These analyses can only be conducted in dry cold ice, and so are limited to the polar ice sheets and some high-altitude sites. As with the continental boreholes, the time resolution of reconstructions is strongly limited by the heat flow process; ice boreholes can only be used to reconstruct long-term averaged temperatures. Ice borehole reconstructions have used several different methods, yielding similar results (Cuffey et al. 1995, Dahl-Jensen et al. 1998, Clow and Waddington 1999). The main assumption of these analyses is that the physical process of heat transfer in the ice is well understood.
As with continental boreholes, the ice boreholes give a temperature history that is the local surface temperature. For central Greenland (Cuffey et al. 1995, Cuffey and Clow 1997, Dahl-Jensen et al. 1998), results show a warming over the last 150 years of approximately 1°C ± 0.2°C preceded by a few centuries of cool conditions. Preceding this was a warm period centered around A.D. 1000, which was warmer than the late 20th century by approximately 1°C. An analysis for south-central Greenland (Dahl-Jensen et al. 1998) shows the same pattern of warming and cooling but with larger magnitude changes. Uncertainties on these earlier numbers are a few tenths of a degree Celsius for averages over a few centuries.
A borehole from Law Dome (Dahl-Jensen et al. 1999), in coastal East Antarctica, reveals a warming of approximately 0.7°C from the middle 19th century to present (uncertainty of approximately 0.2°C). This was preceded by a period of comparable