JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 100,NO. Bll, PAGES 22,017-22,032, NOVEMBER 10, 1995 Evidence for climaticwarming in northcentralOklahoma from analysis of boreholetemperatures David Demirig and Robb A. Borel 1 School of Geology and Geophysics, University of Oklahoma, Norman Abstract. Temperatures were measured in six boreholes (-380 m deep) in northcentral Oklahoma (36.36 øN, 96.70 øW) in May 1993. Temperatures in the upper 150 m of the boreholes appeared tobeanomalously warm when the average thermal gradient was extrapolated from below 150 m depth. The average thermal gradient below 150 m depth was 37.5øC/km; the average thermal gradient from 20to 150 m depth was 27.4øC/km. Several hundred laboratory measurements of thermal conductivity were made oncores collected at the.study site; 65 specific heat capacity measurements were also made. In situ porosity was estimated fromdensity logs and measurements of rock matrix density. Average thermal diffusivity of the upper 110 m of the stratigraphic section was estimated to be 20.3 +2.0 m2/yr (6.43 x 10 '7 m2/s). Heat flow was estimated to be 52 +6 mW/m 2 from 20to110 mdepth and 69+7mW/m 2from 277 to305 mdepth. The observed energy imbalance and anomalously warm temperatures in the upper 150 m of the boreholes could not beexplained solely by hypotheses related to topographic gradients, vegetation, heat refraction, groundwater flow, orland use changes. The only hypothesis which satisfactorily explained all of the observations was an apparent increase (1.25-1.50 _+0.5øC) in ground surface temperature (GST) related to a climatic warming starting in the middle to early 19th century or before (1835, +50, -150 years). When constraints from surface airtemperatures (SATs) were used to interpret borehole temperatures, a better match toobservations was obtained, suggesting that changes in SATs at thestudy site weretracked by changes in GSTs. Introduction Earth's climate changes continuously on all timescales [Savin, 1984; Foiland et al., 1990; Molnar and England, 1990]. For the last 20 yearsor so, increased attention has been given to the possibilityof a climatic warming related to the accumulation of greenhouse gasses in Earth'satmosphere from anthropogenic activities such as deforestation and industrialization. Increasesin the concentration of greenhouse gasses have the theoretical potential to increase average global air temperatures 3-6øC by the end of the 21st century [Houghton et al., 1990; Barron, 1995]. The resulting warming couldchange climatic conditions at ratesthat are unprecedented geologically, with perhaps large social, economic, and environmental consequences. The primary database that has been used to assess climatic warmingover the last 100-150 yearsis the history of surface air temperatures (SATs) as recorded on a daily basisfor the purpose of weather forecasting [Karl et al., 1989]. Hansen and Lebedeff [1987] found a meanglobal increase of-0.5-0.7 øCin SAT over the period 1880-1985. Similar results have been obtained by Ellsaesser et al. [1986], Jones et al. [1986], and Jones [1994]. However, the SAT record has limitations when usedfor the purpose of assessing climate change. The global SAT record prior to about 1870 is sparse to nonexistent, and the SAT history of North America can be reliably reconstructed •Now at Exxon Exploration Company, Houston, Texas. Copyright 1995by the American Geophysical Union. Papernumber 95JB02625. 0148-0227/95/95JB-02625505.00 only backto 1880 [Hansen and Lebedeff, 1987]. The SAT record also suffers from bias and noise due to the existence of urban heat islands and changes in station locations, instrumentation, and times of measurement. SAT records provide a precise measurement of temperature changes on a daily or yearly basis, but theirinterpretation canbe ambiguous because the instrumental record is not long enough to determine thelong-term mean andthus assess if current trends represent significant departures fromit. The limitation of the SAT recordto reveal climatic trends prior to the late 1800s is particularly grievous, because it is unclear if the global warming trendof the last 100 years is a significant warming above a long-ter. m meanthat may be related to increasing concentrations of greenhouse gasses, or simply a return to normal temperatures following a cold spell overthelastpartof the 19th century. Fortunately, the solidEarthcontains the information that is lacking in the SAT record. Changes in groundsurface temperature (GST) propagate into the subsurface, exponentially decreasing in amplitude with increasing time anddepth. The solid Earth acts like a low-pass filter, efficiently and continuously filtering out daily andseasonal changes in GSTs while maintaining a running record of the long-term mean and departures from it. If the average GST increases over aperiod of several years, thenormal upward flux of heat in thesolid Earth is lessened or even reversed, leadingto anomalously high temperatures andanenergy imbalance in theupper 100m or so of the Earth's crust [Cerrnak, 1971' Beck, 1982; Lachenbruch and Marshall, 1986; Sass, 1991' Beltrarni et 'al., 1992; Chisholm and Chapman, 1992;Shenand Beck,1992; Wang and Lewis, 1992; Wang et al., 1992; Cermaltet al., 1993; Pollack and Chapntan, 1993;Harris and Chapntan, 1995; Denting, 1995]. 22,017