Main

Rochechouart possesses a thin, sporadically developed impact melt layer (4 m thick at most). The palaeomagnetic data of Pohl and Soffel2 were obtained from glass-bearing impact breccias and lithic breccias. The high-clast/low-melt content would have resulted in these rocks cooling below their Curie points (for example, approximately 580 °C for pure magnetite) within a short geological period, probably less than 100 years. This would be due to the cooling effect of the entrained cold clasts and the relatively rapid conductive and convective cooling of such a thin melt layer.

In contrast, Manicouagan, as a much larger impact structure, possesses an extensive and significantly thicker melt sheet (>230 m and probably 500 m in total original thickness). Glass-rich breccias were not sampled as part of palaeomagnetic studies at Manicouagan4,5, but crystalline melt sheet was sampled and many of the samples were medium-grained. It would have taken thousands, perhaps tens of thousands, of years for such a body of superheated melt to cool below the relevant Curie points. For example, Onorato et al.3 calculate that it would have required about 1,600 years for the centre of the melt sheet to reach its solidus (915 °C) and up to 10,000 years to cool to about 600 °C. This is for a body estimated at 200 m thick. We know now that 500 m is a more realistic thickness, so the estimates3 are on the low side.

Consequently, although both Manicouagan and Rochechouart could have been formed within hours of each other, the resulting impact-generated rocks would have reached their Curie points at times sufficiently different to allow for a natural geomagnetic reversal to have taken place, accounting for the different magnetic polarities of these impact structures.