Synchronizing Geomagnetic Field Intensity Records in the Levant Between the 23rd and 15th Centuries BCE: Chronological and Methodological Implications

International audienceArcheomagnetic records are an important source of information on the past behavior of the geomagnetic field. Frequently, however, coeval archeomagnetic intensity (archeointensity) datasets from nearby locations display significant discrepancies, hampering precise reconstruction of high-resolution secular variation curve. This is the case for the time interval between the later phase of the Early Bronze and the early phase of the Late Bronze Ages (23rd-15th centuries BCE) in the Levant and Mesopotamia. We address the problem by cross-correlating archeointensity datasets from four major multilayered archeological sites in the southern Levant (Hazor and Megiddo), northern Levant (Ebla), and western Upper Mesopotamia (Mari). We report new archeointensity data, obtained using the Thellier-IZZI-MagIC and the Triaxe methods, from six strata at Hazor and four radiocarbon-dated strata at Megiddo. From 39 pottery fragments, 199 specimens passed our selection criteria, from which we calculated the mean archeointensity for each stratum. To strengthen the comparison of these data with previously published data from Mari and Ebla, obtained using the Triaxe method, we conducted a blind test of the methods that resulted in indistinguishable results or a difference of less than 1 μT. The synchronized compilation, constrained by radiocarbon data from Megiddo, displays a V-shaped pattern with a prominent minimum of at least 200 years centered around the 18th century BCE. The study highlights the importance of stacking archeomagnetic data obtained by different archeointensity methods only after cross-testing the methods and ensuring that archeological samples were dated in a consistent manner