Analyzing the geomagnetic axial dipole field moment over the historical period from new archeointensity results at Bukhara (Uzbekistan, Central Asia)

Abstract Since the mid-19th century, direct measurements of both intensity and directions of the Earth's magnetic field have been available, allowing an accurate determination of its spatio-temporal variations. Prior to this time, between ∼1600 and 1840, only direct directional measurements are available. Therefore, the construction of global field models over this period requires either a specific treatment of the axial dipole field component or the use of archeomagnetic intensity data. In this study, we use a regional approach based on the construction of an archeointensity variation curve in Central Asia. We analyze baked clay brick fragments sampled in Bukhara (Uzbekistan), dated between the end of the 16th century and the beginning of the 19th century. This city is of particular interest for archeomagnetism due to the well-preserved old buildings accurately dated by documentary archives. A series of archeointensity results is obtained using the Triaxe experimental protocol, which shows a decreasing trend in intensity from ∼1600 to ∼1750, with intensities during the 18th century lower than expected from global geomagnetic field models. These new data appear consistent with other Triaxe data previously obtained in western Europe and western Russia, when transferred to Bukhara using the field geometry of the gufm1 model. Together, these data are used to recalibrate the axial dipole moment evolution provided by this model. The resulting evolution appears non-linear, with a clear relative minimum in the magnitude of the axial dipole during the late 18th century. We illustrate the fact that at present this evolution can neither be satisfactorily confirmed nor refuted by other datasets available in western Eurasia (as well as at a wider spatial scale), mainly due to the significant dispersion of the data. Our interpretation relies on the accuracy of the field geometry of the gufm1 model, which appears less reliable prior to ∼1750. Nevertheless, the minimum proposed in the 18th century seems to be a true feature of axial dipole behavior.