Archaeomagnetic Dating Laboratory

Lab Director: Shelby Jones

The Archaeomagnetic Dating Laboratory was established in 1988 by Dr. Daniel Wolfman, with Jeffrey Royce Cox as the laboratory technician. Since Dan's death in 1994, the lab has continued under the direction of Jeffrey Cox. The lab is one of three dedicated laboratories in the Americas—the others are at the Illinois State Museum (Dr. Stacey Lengyel) and Yale University (Dr. Roderick McIntosh). Other archaeomagnetic researchers active in the United States include Dr. Rob Sternberg, Dr. Jeff Eighmy, and William Deaver.

Cox (left) and WolfmanArchaeomagnetic dating was advanced in the United States in the 1960s by Dr. Robert DuBois, professor emeritus, University of Oklahoma. The method is based on comparing the magnetic properties of burned archaeological deposits with calibration curves for the region of the earth's surface that includes the site. When magnetic minerals in soils and sediments are heated, their magnetic fields become oriented to the prevailing magnetic field. When the sediments cool, the orientations are fixed and remain as a "thermoremanent magnetic moment" (TRM). TRM alignments generally persist until the material is again heated to the original or a higher temperature, so that the archaeomagnetic orientation retains evidence of the last heating or burning event. Since the earth's magnetic field is constantly changing, the past apparent or "virtual geomagnetic pole" (VGP) position will be different at different times in the past. VGP positions from archaeological samples can be compared with the regional calibration curves, and the position of the sample VGP then can be interpreted as a date range. Successful archaeomagnetic dating requires appropriate earthen materials, fires sufficiently hot to create a strong and coherent alignment, recovery of a carefully aligned set of specimens from the burned archaeological feature, careful laboratory measurement of the specimens to determine a mean VGP and its error term, and interpretation of a date range from the relationship between the error ellipse and calibration curve.