Magnetic anomaly near the center of the Vredefort structure: Implications for impact-related magnetic signatures: Comment and Reply COMMENT

Hart et al. (1995) postulated that large meteorite impacts may generate characteristic magnetic signatures due to thermal resetting of remanent magnetism, as a consequence of the impact process. Their postulation is based on the results of a study of the ca. 2 Ga Vredefort structure which shows that the remanent magnetism of rocks from the core of the structure was reset penecontemporaneously with the formation of the structure. While we agree with Hart et al. (1995) that the Vredefort structure is the product of a large meteorite impact at ca. 2 Ga and that the reset remanent magnetism in the rocks is consistent with the 2 Ga paleopole orientation for the region, we believe that Hart et al. failed to satisfactorily discount an alternative possibility: that the postimpact thermal event manifested by the reset magnetism reflects high, preimpact, ambient rock temperatures that are related to an earlier regional metamorphic event. Hart et al. (1995, p. 279) mentioned this possibility briefly but discarded it on the grounds that (1) the magnetic anomaly and thermal metamorphism are centered on the Vredefort structure, and (2) there is no evidence of any ca. 2 Ga metamorphism predating the formation of the structure. In order to address the first point, it is necessary to establish what is meant by the term ‘‘Vredefort structure.’’ Hart et al. (1995, p. 277) stated that the Vredefort structure is ‘‘a large complex crater . . . [with] an original diameter of !300 km.’’ However, their Figure 1, described as a geologic map of the Vredefort structure, shows only the central, !70–80-km-wide Vredefort dome, which, according to Therriault et al. (1993), corresponds to the central uplift of the impact structure. We propose that the term ‘‘Vredefort structure’’ be reserved for the originally wider, now deeply eroded, impact basin, and that the central uplift be designated as the Vredefort dome. In this context, it is not clear whether Hart et al.’s (1995) statement, that the magnetic anomaly and thermal metamorphism are centered on the structure, refers to the wider impact basin or only to the dome. Because the metamorphism has most typically been described as centered on the dome (e.g., Schreyer, 1983), we assume that this is their implication. Evidence exists, however, for regional magnetic resetting at 2 Ga, as follows. (1) Layer et al. (1988) established that remanent magnetism throughout the Witwatersrand basin was reset to a ca. 2 Ga paleopole; (2) Morgan (1985) established a similar paleopole for rocks in the Limpopo Belt, some 500 km to the north of the Vredefort structure. The striking similarity of these paleopole orientations to the paleopole obtained from the 2.06 Ga (Walraven et al., 1990) Bushveld Complex gabbros led Layer et al. (1988) to propose that a regional thermal event affecting the crust of the Kaapvaal craton accompanied the Bushveld magmatic event. The P-T path inferred for the mid-amphibolite facies metasedimentary rocks in the collar of the Vredefort dome is consistent with such timing (Gibson and Wallmach, 1995), and the high-grade metamorphism observed in the core of the dome (Schreyer, 1983) could also be reconciled with the intrusion of voluminous mafic magmas into the lower crust and upper mantle during the Bushveld event. On the basis of an estimated peak geothermal gradient of !40 "C/km (Gibson and Wallmach, 1995), much of the middle to lower crust in the region is likely to have undergone temperatures above the Curie point for magnetite (!550 "C; Telford et al., 1990) during this event, leading to remagnetization. According to Hart et al. (1995), however, the remagnetization that they describe occurred after the impact event, which postdated the Bushveld event by!35 m.y. (Kamo et al., 1995; Spray et al., 1995). If during this interval the rocks cooled rapidly, the magnetic anomaly described by Hart et al. (1995) might be attributable to the impact event. We believe, however, that sufficient evidence exists to suggest that postpeak cooling was slow prior to the impact event and, thus, that large parts of the crust were still above the magnetite Curie point temperature at the time of impact. (1) Clinoeulite (magnesian clinoferrosilite) grains in granulite facies meta-ironstones from the core of the dome contain narrow Fe-augite exsolution lamellae indicative of rapid cooling from temperatures above the lower stability limit of pigeonite (!800 "C; Schreyer et al., 1978). This texture is consistent with rapid exhumation of still-hot granulites. (2) Schreyer and Abraham (1978) linked granulite facies decompression textures in garnet paragneisses from the core of the dome to pseudotachylite development. Hart et al. (1991) challenged this interpretation and suggested a 3.5 Ga age for the decompression event; however, a reappraisal of the textures in these rocks (G. Stevens, 1995, personal commun.) supports Schreyer and Abraham’s (1978) findings. (3) Metamorphic textures compatible with pseudotachylite-related decompression under lower amphibolite facies conditions also occur in the mid–amphibolite facies metapelitic collar rocks (Gibson and Wallmach, 1995). The temperatures indicated by these textures are consistent with a regional crustal geothermal gradient of #25 "C/km immediately prior to the impact-related exhumation event. Rapid cooling following exhumation of these rocks associated with the impact event could thus ‘‘freeze in’’ a 2 Ga paleopole orientation in the upturned lower crustal rocks. We are in no way implying that the Vredefort event did not cause a significant temperature rise in the rocks as a consequence of impact processes. Instead, we believe that, owing to the unusual coincidence of the Vredefort impact event with a regional metamorphic event, the magnetic signature of the rocks described by Hart et al. (1995) cannot be regarded as a characteristic feature of large impacts, as they suggest. Given the complex interaction of a variety of processes that affect the magnetic signatures in other impact sites (Pilkington and Grieve, 1992), it is imperative that detailed petrographic analysis be performed on the rocks discussed by Hart et al. (1995) to identify the magnetic minerals responsible for the observed signature and their mode of occurrence. Finally, we point out that as yet no published structural data exist to substantiate Hart et al.’s (1995, p. 277) assertion that the so-called Vredefort discontinuity is a major tectonic boundary. A similar problem exists with the inferred ‘‘southeast boundary fault’’ (Hart et al., 1995, Fig. 1) and the statement (p. 277) that ‘‘the rim strata and the basement rocks all dip steeply inward toward the center.’’ Fundamental geological, structural, and mineralogical data remain to be collected from the core and collar of the Vredefort dome—the central part of a uniquely exposed deeper level of a large multiring impact basin.