DISCOVERY OF A NEW SPHERULE LAYER IN THE HAMERSLEY BASIN (WESTERN AUSTRALIA) REPRESENTING A FOURTH ASTEROID IMPACT NEAR THE ARCHEAN/PROTEROZOIC BOUNDARY

We recently found a previously unrecognized Neoarchean spherule layer enclosed in basinal dolomite in the Paraburdoo Member of the Wittenoom Formation (Hamersley Group, Western Australia). The Hamersley succession straddles the Archean/Proterozoic boundary and contains 3 known spherule layers at other stratigraphic levels (one below, two above), so the new layer represents a fourth impact event within a span of ~140 million years. Very approximately, the impact was comparable in size to the end-Cretaceous event and occurred about 2.6 Ga. Where discovered, the layer consists almost entirely of spherules, averages 2 cm in thickness, and is normally graded. Spherules average 0.8 mm in diameter at the base of the layer vs. 0.4 mm at the top; internally, they are highly crystallized. Most of their cross-sections are filled with acicular to thin, highly elongated lath-shaped pseudomorphs that consist of feldspar. These look like replaced plagioclase crystallites seen in other Hamersley spherules. However, unlike those of other layers, some of the Paraburdoo spherules contain pseudomorphs of more equant to broad, elongated crystals with pointed terminations and/or internal gaps that consist of chloritic material. The latter closely resemble skeletal olivine and/or pyroxene crystals in komatiites and chondrules. The Paraburdoo spherules contain few candidates for vesicles or large patches of replaced glass. At the discovery site, the dolomite above and below the new layer is typical of the Paraburdoo Member, i.e., thinly laminated and deposited in quiet, deep water environments. No spherules or silicate detritus of any sort were observed in the overlying beds, nor does the base of the layer erode into underlying strata. We suspect the layer formed via suspension settling of spherules to the basin floor with little or no reworking. In contrast, all of the previously reported Hamersley layers are high-energy event beds that involved considerable syndepositional reworking. Finally, the new layer may well have been formed by the same impact as the Reivilo spherule layer in the coeval Griqualand West Basin (South Africa). The Paraburdoo and Reivilo layers occur at comparable stratigraphic levels (based on sequence analysis) and their spherules have very similar textures, most notably a high degree of crystallization.