Estimating the approximate firing temperature of burnt archaeological sediments through an unmixing algorithm applied to hysteresis data

Iron minerals commonly found within archaeological sediments and soils are highly sensitive to thermal alteration from a comparatively low temperature (~200 deg C). Such thermal alterations may often be identified through magnetic measurements that may reveal the presence, or absence, of diagnostic magnetic minerals associated with exposure to a particular temperature regime. Controlled laboratory heating experiments, conducted on three soil types as part of this study, demonstrate a consistent magnetic enhancement over a temperature range from ambient to 500 deg C. This ongoing thermal enhancement is reflected within both Isothermal Remanent Magnetisation (IRM) acquisition curves and hysteresis loop data collected from the laboratory heated samples. The altering shape of the IRM acquisition curves and hysteresis loops, represents a gradual change in the complex mixture of magnetic minerals within the progressively heated samples. It is proposed that the magnetic data from the laboratory heated samples, may be used as a reference to compare against burnt archaeological samples, developed over the same soil types, to reveal their previous thermal history. This is achieved by constructing a mathematical unmixing algorithm, that describes the IRM curve or hysteresis loop data from the burnt archaeological sample in terms of a constrained mixture of end-members, based on the laboratory samples heated to a known temperature. The ratio of the final end-members is then used to estimate the approximate exposure temperature of the archaeological sample. Whilst the final magnetic mixture produced within a burnt sample will be dependent upon a wide range of variables, comparison with field experiments suggests this method provides an accurate determination of thermal history. Additional applications of this technique for the interpretation of semi-industrial archaeological sites, or determining the fidelity of samples recovered for archaeomagnetic dating through their apparent firing temperatures, are also considered.