Constraints on the depth and thermal history of cratonic lithosphere from peridotite xenoliths, xenocrysts and seismology

Despite the relatively long-standing availability of numerical approaches for estimating palaeogeotherms using peridotite xenolith Pressure–Temperature (P–T) data, the practise of fitting xenolith P–T arrays to simple models of lithospheric heat generation, in a non-quantitative manner, remains widespread. The lack of quantification in both the magnitude and uncertainty of heat flow and lithosphere thickness estimates leads to difficulty in evaluating proposed models for lithosphere evolution on a local and regional scale. Here, we explore the advantages of using a numerical approach to palaeogeotherm fitting, in terms of the ability to make objective comparisons of the effect that differing thermobarometer combinations and varying states of mineral and textural equilibrium have on the shape of the palaeogeotherm, and the resulting estimates of lithospheric thickness and heat flow. We also make quantitative comparisons between lithospheric mantle properties estimated using peridotite xenoliths versus single mineral xenocrysts. Using two reference peridotite xenolith databases from Bultfontein (S. Africa) and Somerset Island (Canada) we show that the same lithospheric mantle properties are predicted using harzburgite versus lherzolite thermobarometry methods. Filtering mineral data for the effects of inter-mineral disequilibrium does not produce significantly different palaeogeotherms but does increase the quality of fit of the palaeogeotherm to the P–T data, allowing more confidence to be placed in comparisons between locations. Palaeogeotherms calculated using xenocryst data, screened for peridotitic affinities, show misfits that are 2–3 times greater than those obtained using xenoliths. Lithospheric properties calculated from the Somerset Island xenocryst-based geotherm yield results that are within error of the xenolith estimate. A mutually consistent and quantitative palaeogeotherm fitting approach is used to evaluate existing hypotheses for the evolution of the southern African lithosphere. We find very similar estimates for the heat flow and thickness of the lithosphere between SW Namibia (off-craton) and Bultfontein (on-craton). This supports suggestions of a cratonic thermal regime and equivalent lithospheric thickness across that region of southern Africa at the time of kimberlite sampling, with concurrent local thermal disturbance evident in Namibia. Complimentary, novel, seismically-obtained geotherm estimates show that the lithosphere in Namibi