摘要:Temperature changes at the Earthʼs surface propagate and are recorded underground as perturbations to the equilibrium thermal regime associated with the heat flow from the Earthʼs interior. Borehole climatology is concerned with the analysis and interpretation of these downward propagating subsurface temperature anomalies in terms of surface climate. Proper determination of the steady-state geothermal regime is therefore crucial because it is the reference against which climate-induced subsurface temperature anomalies are estimated. Here, we examine the effects of data noise on the determination of the steady-state geothermal regime of the subsurface and the subsequent impact on estimates of ground surface temperature (GST) history and heat gain. We carry out a series of Monte Carlo experiments using 1000 Gaussian noise realizations and depth sections of 100 and 200 m as for steady-state estimates depth intervals, as well as a range of data sampling intervals from 10 m to 0.02 m. Results indicate that typical uncertainties for 50 year averages are on the order of ±0.02 K for the most recent 100 year period. These uncertainties grow with decreasing sampling intervals, reaching about ±0.1 K for a 10 m sampling interval under identical conditions and target period. Uncertainties increase for progressively older periods, reaching ±0.3 K at 500 years before present for a 10 m sampling interval. The uncertainties in reconstructed GST histories for the Northern Hemisphere for the most recent 50 year period can reach a maximum of in some areas. We suggest that continuous logging should be the preferred approach when measuring geothermal data for climate reconstructions, and that for those using the International Heat Flow Commission database for borehole climatology, the steady-state thermal conditions should be estimated from boreholes as deep as possible and using a large fitting depth range (~100 m).