Ž . Global and Planetary Change 29 2001 311–325 www.elsevier.comrlocatergloplacha A decade of air–ground temperature exchange from Fargo, North Dakota William L. Schmidt a, ) , William D. Gosnold a,1 , John W. Enz b,2 a Department of Geology and Geological Engineering, Box 8358, UniÕersity of North Dakota, Grand Forks, ND 58202, USA b Department of Soil Science, North Dakota State UniÕersity, Fargo, ND 58105, USA Received 10 January 2000; received in revised form 3 March 2000; accepted 21 March 2000 Abstract In borehole paleoclimatology, it is commonly assumed that a direct coupling exists between air and ground temperatures. This assumption is valid only if variables affecting ground-surface temperature exchange have remained constant through time. In an analysis of a 9-year record of air and ground temperature data, we found that several critical variables changed in ways that cause decoupling between air and ground temperatures. Mean-annual ground temperatures in the upper 12 m increased by 0.93 "0.098C during the study. Air temperatures used as model-forcing signals generated ground temperatures that exhibit no significant increase. The decoupling of winter air and ground temperatures is due to snow cover and latent energy effects. Maximum residual temperatures for freezing, summer and thawing modeling periods averaged "0.188C, "0.308C, and "0.758C, respectively. Duration of winter snow cover increased during the time of record and correlates with Ž 2 . winter air–ground temperature differences r s0.71 . Annual values of modeled latent energy of ground freezing show a dependence upon total precipitation 60 days prior to ground freezing. q 2001 Published by Elsevier Science B.V. Keywords: air–ground temperature exchange; heat conduction model; ground-surface temperature 1. Introduction A common assumption in borehole paleoclimatol- ogy is that the upper few hundred meters of the geothermal gradient retain a faithful record of the surface air temperature for the past several centuries. Ž In several cases, this has been tested Beltrami et al., ) Corresponding author. Tel.: q 1-701-777-2821; fax: q 1-701- 777-4449. E-mail addresses: william schmidt@und.nodak.edu – Ž . W.L. Schmidt , william gosnold@und.nodak.edu – Ž . Ž . W.D. Gosnold , john enz@ndsu.nodak.edu J.W. Enz . – 1 Tel.: q 1-701-777-2631; fax: q 1-701-777-4449. 2 Tel.: q 1-701-231-8576; fax: q 1-701-777-4449. 1992; Chisholm and Chapman, 1992; Gosnold et al., 1997; Harris and Chapman, 1997; Majorowicz and . Skinner, 1997 , but most reports have accepted the validity of this assumption. Recent studies have questioned this assumption and have found that in certain localities the ground-surface temperature may not provide an accurate record of the surface air Ž temperature history Gosnold et al., 1997; Majorow- . icz and Skinner, 1997 . Under certain conditions, ground temperatures can contain non-climatic noise due to changes in ground-surface variables, including forest and vegetative cover, ground-surface albedo, duration and thickness of snow cover, precipitation, air temperature, wind speed, net radiation, and soil moisture content. 0921-8181r01r$ - see front matter q 2001 Published by Elsevier Science B.V. Ž . PII: S0921-8181 01 00097-2