Contents lists available at ScienceDirect Journal of Archaeological Science: Reports journal homepage: www.elsevier.com/locate/jasrep Last 3000 years of geomagnetic field intensity from India: New reference palaeointensity data from two east coast archaeological sites and archaeomagnetic dating insights N. Basavaiah a, ⁎ , K. Deenadayalan a , J.L.V. Mahesh Babu a , M.O. Kanu b , K.Ch.V. Naga Kumar c , G. Demudu c , K. Nageswara Rao c , N. Mallikarjunarao d a Indian Institute of Geomagnetism, Kalamboli, New Panvel, Navi Mumbai 410 218, India b Department of Physics, Taraba State University, P.M.B. 1167, Jalingo, Taraba State, Nigeria c Department of Geo-Engineering, Andhra University, Visakhapatnam 530 003, India d Department of Archaeology and Museums, Govt. of Andhra Pradesh, Vijayawada 521225, India ARTICLE INFO Keywords: Archaeomagnetism Southern India Buddhist monument Medieval fort Secular variation Archaeomagnetic dating Rock magnetic properties ABSTRACT A comprehensive rock magnetic and absolute archaeointensity (AI) investigation was carried out on baked bricks from the east coast of India with an aim to obtain a set of well-defined palaeointensity (PI) data. The new PI data from India comprise a significant first step to improve the regional intensity reference curve, which can be used as reliable dating tool mainly for archaeomagnetic material. Standard Thellier-Coe experiments were applied on 12 independent brick samples collected from a Buddhist monument at Adurru in the Godavari delta and a Medieval Fort at Kolleti Kota site situated on an island in the Kolleru Lake. The samples thus collected represent two different archaeological periods, one around the 3rd century BCE and the other between the 4th and 7th centuries CE respectively. The data obtained from thermal demagnetization of Saturation Isothermal Remanent Magnetization (SIRM), temperature dependent magnetic susceptibility (χ–T), Isothermal Remanent Magnetization (IRM) acquisition curves, and hysteresis measurements showed that the primary carriers of Natural Remanent Magnetization (NRM) are fine-grained Pseudo-Single Domain (PSD) (Stable Single Domain (SSD) + Superparamagnetic (SP)) titanomagnetite with various titanium contents. This implies that the samples had significant stable Thermoremanent Magnetization (TRM), and, therefore, good candidates for PI analysis. Strict selection and quality-control criteria for detecting magnetomineralogical alteration and intensity data corrected for anisotropy and the cooling rate effects on TRM acquisition, offer high quality PI determinations with respect to the previous available data that missed such corrections. A total 33 specimen level PI determinations that correspond to linear NRM–TRM plots, were made using the Thellier GUI software. The corrected data yielded mean PI values of 44.6 ± 1.4 μT for the Buddhist Monument, and 36.6 ± 2.9 μT for the Fort site. By comparison and evaluation of new PI data with Global Geomagnetic Field Model predictions, the sampled bricks were archaeomagnetically dated using the Matlab tool developed by Pavón-Carrasco et al. (2011). Both ARCH3K.1 and SHA.DIF.14 K global models were used to evaluate the age estimates. The SHA.DIF.14 K model provided the most accepted age estimates that varied between 305 and 182 BCE for the early historic Buddhist Structure at Adurru, and 516–889 CE for the Medieval Fort at Kolleti Kota. These new reference archaeointensity data highlight the potential of archaeomagnetic dating to improve the archaeologically determined time periods of bricks, and also show the need for new high quality reference data to improve the geomagnetic field models from India. 1. Introduction Archaeomagnetism is the field of study concerned with the appli- cation of time variant geomagnetic phenomena to archaeology and the use of archaeological material for elucidating the past behavior of the geomagnetic field. Many studies have demonstrated that archaeological artifacts can reliably record the direction and intensity of the geo- magnetic field (e.g., Aitken and Weaver, 1962). For example, baked https://doi.org/10.1016/j.jasrep.2019.101943 Received 31 December 2018; Received in revised form 2 July 2019; Accepted 12 July 2019 ⁎ Corresponding author. E-mail address: bas@iigs.iigm.res.in (N. Basavaiah). Journal of Archaeological Science: Reports 27 (2019) 101943 2352-409X/ © 2019 Published by Elsevier Ltd. T