RESEARCH COMMUNICATIONS CURRENT SCIENCE, VOL. 115, NO. 8, 25 OCTOBER 2018 1567 *For correspondence. (e-mail: akkiraju_vyas@yahoo.com) Ground surface warming in peninsular India: evidence from geothermal records Vyasulu V. Akkiraju 1, *, Sukanta Roy 1,2 and Shalivahan 3 1 Ministry of Earth Sciences, Borehole Geophysics Research Laboratory, Karad 415 114, India 2 CSIR-National Geophysical Research Institute, Uppal Road, Hyderabad 500 007, India 3 Department of Geophysics, IIT (Indian School of Mines), Dhanbad 826 004, India Analyses of borehole temperature profiles provide useful information about regional climate change over a few centuries. Data from 146 borehole sites in the crystalline terrain of peninsular India were used to reconstruct surface ground temperature history. Depths of the boreholes ranged from 150 to 1522 m. The temperature profiles were characteristic of heat flow by conduction, being largely unaffected by per- turbations due to groundwater flow. The profiles show temperature anomalies in the top few hundred metres that is consistent with changing surface temperature over the past two–three centuries. Analysis of individ- ual profiles for a ramp change in temperature reveals predominant surface ground warming in peninsular India with a mean magnitude of 1.0 ± 0.2°C for 129 ± 18 years at 95% confidence level corresponding to onset ca 1860 AD for the change. Keywords: Borehole temperatures, ground temperature history, peninsular India, surface air temperature. METEOROLOGICAL surface air temperature (SAT) records obtained from a worldwide network of stations provide unambiguous evidence of warming over the past 150 years 1,2 . However, to constrain surface temperature change and climate variability over the past millennium for which direct observations are unavailable, proxy indi- cators of temperature change, such as tree rings, sediment cores, ice cores and corals 3–11 are considered. Reconstruc- tion of surface temperature from conventional proxy indi- cators provides high temporal resolution, but it cannot resolve low frequency components or provide a long-term mean temperature. Moreover, the proxy records only pro- vide an indirect measure of temperature and therefore re- quire additional calibration. On the other hand, borehole temperature–depth profiles, down to a few hundred metres, provide a direct measure of changing surface ground temperature averaged over a much longer period 12–15 . The geothermal signature is a result of ther- mal diffusion process; the earth efficiently and conti- nuously filters the short-period surface temperature changes on diurnal and annual timescales while retaining a record of the long-term mean and the departures from it 12 . Recent observational studies demonstrate the strong coupling between surface temperature changes and tran- sient anomalies gleaned from borehole temperature pro- files 16 . In this way, borehole temperature–depth data complement surface air temperature and multi-proxy reconstructions of climate change. Analyses of several hundred borehole temperature profiles distributed mainly in the Northern Hemisphere have revealed important information regarding past cli- mate change 11,14,15 . The salient features are 11 : (i) Majority of geothermal climate change studies are based on data from the mid-latitude band in the Northern Hemisphere. (ii) The last century experienced widespread warming compared to cooler conditions in the four prior centuries. (iii) The magnitude of warming from 1850 to 1990 is estimated to be approximately 0.7 ± 0.2°C, consistent with the meteorological surface air temperature records. (iv) Reconstructions of surface ground temperature (SGT) from geothermal records have decadal- to century- scale resolution and thus constrain the long-term tem- perature trends, but not the annual or sub-annual scale variations. In the recent past, borehole temperature profiles from different climatic provinces in India were studied to infer past climate change. Roy et al. 17 studied the database of borehole temperature profiles in India and analysed 70 temperature profiles on the basis of well-established criteria. The data covered six major climatic provinces 18 : interior peninsula (IP), east coast (EC), west coast (WC), north-central (NC), north-east (NE) and north-west (NW). Analyses of data revealed mean warming of 0.9 ± 0.1°C with onset ~1850 AD. Akkiraju and Roy 19 made a detailed study of temperature–depth data obtained from a specially drilled borehole located near Hyderabad in the Interior Peninsula and found the regional warming consistent with other parts of the IP province. They also found signature of recent cooling attributable to land-use change in the vicinity of the borehole during the past few decades. Another study 20 compiled new data from south India and analysed it together with the previous dataset from south India. Combined data analysis from 74 sites yielded a mean warming of 0.9 ± 0.3°C with onset centred around 1850 AD. In the present study, data from 146 sites selected according to criteria adopted previously 20 , were analysed to infer the past climate change in peninsular India. The dataset includes those from previous studies 17,20 and new data from Koyna in WC province and Choutuppal area near Hyderabad in IP province. Clearly, the IP and WC provinces have better coverage when compared to the other provinces. Nevertheless, the 146 temperature pro- files together provide a compelling dataset to cover the low latitude range 8°–29°N, which is under-represented in geothermal climate change studies. Importantly, the