Is it possible to identify temporal differences among combustion features in Middle Palaeolithic palimpsests? The archaeomagnetic evidence: A case study from level O at the Abric Romaní rock-shelter (Capellades, Spain)

Archaeomagnetic dating is probably one of the most known applications of magnetic methods to archaeology but there are others still underutilized and of particular interest to Palaeolithic archaeology. Here, we report a novel application of archaeomagnetism as a technique to determine temporal diachronies among combustion features from the same surface within palaeolithic palimpsests. The approach is based on the subtle directional changes of the Earth's magnetic field through time (secular variation, SV) and on the ability of burned materials to record such variations under certain conditions. Three Middle Palaeolithic hearths from level O (ca. 55 ka BP) at the Abric Romaní rock-shelter (NE Spain), were archaeomagnetically investigated. The studied surface (black homogeneous carbonaceous facies), recorded the magnetic enhancement produced by fire with a tenfold increase in concentration-dependent magnetic parameters in the uppermost centimetre with respect to its unburned or deeper counterparts. Pseudo-single domain (PSD) Ti-low titanomagnetite was identified as the main remanence carrier. The irreversibility of thermomagnetic curves suggests that these samples did not undergo enough high temperatures as to record a full thermoremanence (TRM). Additionally, the occasional occurrence of maghaemitized magnetite is interpreted as an indication of a thermochemical remanent magnetization (TCRM), making these samples unsuitable for absolute palaeointensity determinations. Two well-defined (α95 < 5°) and statistically indistinguishable archaeomagnetic directions were obtained with their mean directions within their respective confidences circles at the 95% level. The lack of directional changes and the similarity in the magnetic properties suggest that these hearths recorded simultaneously or closely confined in time the Earth's magnetic field direction at the time of cooling. These results agree well with archaeological evidence which indicates a synchronic occupation of this activity area. The possibility of determining temporal differences among combustion features in prehistoric sites arises as a promising tool in palimpsest dissection studies and may help to reconstruct occupation patterns of prehistoric groups. The practical limits of the method are discussed as well as its potential to identify post-depositional mechanical alteration processes.