A Matlab tool for archaeomagnetic dating Fco. Javier Pavón-Carrasco a, * , Juan Rodríguez-González a , Maria Luisa Osete a , J. Miquel Torta b a Dpto. de Física de la Tierra I: Geofísica y Meteorología, Universidad Complutense de Madrid, 28040 Madrid, Spain b Observatori de l’Ebre, CSIC e Universitat Ramon Llull, Horta Alta 38, 43520 Roquetes, Spain article info Article history: Received 21 June 2010 Received in revised form 15 September 2010 Accepted 17 September 2010 Keywords: Archaeomagnetic dating Geomagnetism Archaeomagnetism Dating tool Matlab abstract A Matlab tool for archaeomagnetic dating has been developed in this work. Well-dated palaeosecular variation curves (PSVCs) can be used to date archaeological artefacts with unknown ages. The archae- omagnetic direction (declination and/or inclination) and the archaeointensity obtained from the archaeological artefact are compared with a master PSVC. In addition, historical lava flows with controversial ages can be dated using this methodology. The dating process follows the descriptions given by Lanos (2004), which is based on the combination of temporal probability density functions of the three geomagnetic field elements. Here, we develop an interactive tool in Matlab code to carry out archaeomagnetic dating by comparing the undated archaeomagnetic (or lava flow) data with a master PSVC. The master PSVCs included with the Matlab tool are the different European Bayesian curves and those generated using both regional and global geomagnetic field models. A case study using all the PSVCs available in Europe and some undated archaeomagnetic data has been carried out to analyze how the different PSVCs affect the dating process. In addition, the dating uncertainty and the relocation error have been analyzed in the European region. Results show that some regional Bayesian PSVCs and the regional SCHA.DIF.3K archaeomagnetic model are the best choices to obtain an accurate date in Europe. Moreover, when it is available, the full geomagnetic field vector must be used for archaeomagnetic dating. Ó 2010 Elsevier Ltd. All rights reserved. 1. Introduction The geomagnetic field varies in both space and time over the Earth’s surface, and its internal fields show a long-term variation called secular (or palaeosecular) variation. This physical phenom- enon, along with the ability of certain archaeological artefacts (e.g. ceramics, bricks, burnt horizons, etc), lava flows and sediments, to acquire a stable remanent magnetisation at the time of its last use (heated archaeological materials) or its formation (rocks), are the basic principles for palaeomagnetism. When studies are carried out on archaeological materials they are referred to as archaeomagnetic investigations. A compilation of these types of data, i.e., archaeomagnetic data, allows us to construct a Palaeosecular Variation Curve (PSVC) for any particular region. If the curve is sufficiently well-defined it can be used as dating tool for archaeological features from the same region (see e.g. Aitken, 1990; Eighmy and Sternberg, 1990; or McIntosh and Catanzariti, 2006 for more details). To determine the most probable age of the last use of an archaeological artefact its remanent magnetisation is compared with a reference PSVC. This classical approach to archaeomagnetic dating can be completed by the use of PSVCs generated by global (Lodge and Holme, 2009) or regional models (Pavón-Carrasco et al., 2009). The classical technique requires a high density of archae- omagnetic data concentrated in a region of about 2  10 5 km 2 . The use of synthetic PSVCs generated by geomagnetic models allows extending the archaeomagnetic dating technique to regions where there is not enough density of data to construct a classical PSVC, or where there are important gaps in its temporal distribution. In these cases the classical dating archaeomagnetic tool could not be applied or would provide high uncertainty. However if the behav- iour of the geomagnetic field is well defined at a continental scale where the region is located (or at a global scale) a synthetic PSVC can be estimated for the locality of the archaeological site. In addition, this new approach avoids the relocation error included when data are relocated to a reference point to construct the PSVC and when the site to date is also relocated for comparison with the PSVC. Recently new, and more refined, global (Korte et al., 2009) and regional models for the European region (Pavón-Carrasco et al., 2009, 2010) have been published; therefore this new technique can be implemented as a routine dating tool for archaeological sites. * Corresponding author. Tel.: þ34 913945190; fax: þ34 913944398. E-mail addresses: fjpavon@fis.ucm.es (F.J. Pavón-Carrasco), juan.rodriguez@fis. ucm.es (J. Rodríguez-González), mlosete@fis.ucm.es (M.L. Osete), jmtorta@ obsebre.es (J.M. Torta). Contents lists available at ScienceDirect Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas 0305-4403/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jas.2010.09.021 Journal of Archaeological Science 38 (2011) 408e419