The Enduring Mystery of Australasian Tektites

DOI: 10.2138/gselements.14.3.212

Figure 1. Australasian tektite strewn field. After Cavosie et al. (2018).

Molten glass rained down from the sky over parts of Southeast Asia, Australia, Antarctica, and into the neighbouring ocean basins during the Pleistocene, about 790,000 years ago. These glass occurrences, long recognized to be remnants of melt formed during meteorite impact, are known as the Australasian tektites. Their distribution defines the largest of at least four known strewn fields across the globe, strewn fields being regions over which tektite glass are scattered from what are thought to be single-impact events. The three other big tektite strewn fields are associated with known source craters, including the Bosumtwi (1.07 Ma, Ghana), Ries (15 Ma, Germany), and Chesapeake Bay (35.5 Ma, USA) impact structures (Glass and Simonson 2013). At only 790,000 years old, the Australasian tektite strewn field (Fig. 1) is both the youngest and the largest known. Despite much effort, the source crater has yet to be discovered. The search to locate it represents something akin to a “holy grail” in impact cratering studies.

Tektite glass is widely accepted to represent impact melt that was ejected during crater formation. Such glass is generally dark to black but also occurs in other colours, as well as in a variety of shapes and sizes. Differences in tektite morphology have led to their being classified into three groups. First, the “splash form” tektites, which include the classic spherical and dumbbell-shaped tektites (Fig. 2A). Second, the “ablated form” tektites (Fig. 2B), which preserve evidence of the glass having partially re-melted during atmospheric re-entry. Third, the “Muong-Nuong type” (MN-type) tektites, which preserve conspicuous layering and have other distinguishing features such as relict minerals from the target rocks (e.g. quartz, zircon, rutile, chromite, monazite and others). Tektites are generally centimetre-size, but kilogram-size Australasian tektites have been reported. A given strewn field will contain all tektite types, but most MN-type are known from the Australasian field. In contrast to tektites, microtektites are sub-millimetre in size and are found almost exclusively in deep-sea cores.

Figure 2. Tektites and shocked minerals. (A) Example of a splash-form tektite from the Nullarbor desert in Western Australia. Photo credit: Morgan A. Cox (B) Example of an ablated-form tektite. Photo by H. Raab licensed under CC BY-SA 3.0 (C) Backscattered electron image of a granular zircon grain in an MN-type tektite from Thailand that is comprised of zircon neoblasts with inclusions of zirconia (ZrO2). (D) Electron backscatter diffraction orientation map of the same grain shown in Figure 2C. Systematic orientation relations among adjacent neoblasts, represented by color variations, provide evidence for the former existence of reidite, the high-pressure ZrSiO4 polymorph, in MN-type tektites. Figures 2C and 2D after Cavosie et al. (2018).

Australasian tektites vary in composition but are isotopically distinct from other tektites (Koeberl 1990). Generally, they have high-Si glass, with SiO2 averaging ~73.5 wt%. Other major element oxides include Al2O3 (~11.5 wt%), FeO (~4.7 wt%), CaO, MgO, and K2O (all between 2.0 to 3.5 wt%), NaO (1.3% wt%) and TiO2 (0.7% wt%). Tektites differ from volcanic glass in several important aspects, including very low water content (~50 ppm OH). The presence of 10Be can be used as a unique tracer of tektite provenance (e.g. Ma et al. 2004), which is generally taken as near-surface material rather than excavated bedrock. A near-surface source, such as siliciclastic sediments (e.g. Koeberl 1990), is also consistent with the relict mineral assemblage represented by included detrital grains.

The formation mechanism of the Australasian tektites remains poorly understood (e.g. Koeberl et al. 1994) and this is further compounded by a lack of direct knowledge on the whereabouts of the source crater. Outstanding questions include, “How much melt was produced and subsequently ejected?” and “What were the target rocks?” Perhaps the most critical questions are “Where is the location of the source crater” and “What is its size?” Many locations have been proposed, including sites in China, Antarctica, and Siberia, although most studies appear to favour a location in Southeast Asia. Modelling the crater size, based on tektite distribution, has resulted in estimated crater diameters ranging from ~40 km to >100 km. Wherever its actual location, it is widely agreed that the crater is young and large, and, thus, should be a conspicuous feature on Earth’s surface.

The discovery of shocked minerals in Australasian tektites has provided tantalizing clues to their origin. Shock-damaged quartz in Australasian tektites was recognized from X-ray data back in the 1970s, and coesite, the high-pressure SiO2 polymorph, was later found in microtektites within a circular distribution around Southeast Asia (e.g. Folco et al. 2010). Most recently, evidence for the former presence of reidite, a high-pressure ZrSiO4 polymorph stable at 30 GPa, has been discovered in granular-textured zircon (Figs. 2C, 2D) in MN-type tektites from Thailand, again supporting a crater location in Southeast Asia (Cavosie et al. 2018).

Solving the mystery of where Australasian tektites originated is of broad interest, far beyond the interest of those only concerned with glass and shocked minerals. The event that created the Australasian tektites is the only potential environmentally catastrophic extra-terrestrial impact event possibly witnessed by anyone on the specifically human ancestral tree. While the northern latitudes were experiencing glacial conditions, many details of human evolution during the mid-Pleistocene are still debated. It is tantalizing to think that members of the Homo erectus lineage living in Asia, best known from the discovery of the Peking Man skull (Fig. 3), could have witnessed the Australasian impact.

Figure 3. Image of the Peking Man skull, being representative of Homo erectus, some of whom might have witnessed the Pleistocene fireball that initiated the Australasian tektites. Photo by Kevin Walsh licensed under CC-BY 2.0.

The burial age of the sedimentary layer that Peking Man came to rest in (i.e. 770,000 ± 80,000 years old) (Shen et al. 2009), fully overlaps with the age of Australasian tektite formation at 785,000 ± 7,000 years (Schwarz et al. 2016). Did our distant ancestors see this event, perhaps as a second sunrise, or a flash in the night sky? Or did they wake up to find bits of shiny black glass scattered about, objects that weren’t there the day before? While unknowable, it is interesting to consider the idea that Homo erectus may well have been the first of us to see such a huge fireball, and then to have puzzled over the enigmatic shiny glass that fell from the sky afterward.


Support was provided by the NASA Astrobiology program (NNAI3AA94A), the Australian Research Council, and the John de Laeter Centre, the Space Science and Technology Centre, and The Institute for Geoscience Research at Curtin University.


Cavosie AJ, Timms NE, Erickson TM, Koeberl C (2018) New clues from Earth’s most elusive impact crater: evidence of reidite in Australasian tektites from Thailand. Geology 46: 203-206

Folco L and 5 coauthors (2010) Shocked quartz and other mineral inclusions in Australasian microtektites. Geology 38: 211-214

Glass BP, Simonson BM (2013) Distal Impact Ejecta Layers: A Record of Large Impacts in Sedimentary Deposits. Springer-Verlag, Berlin Heidelberg, 716 pp

Koeberl C (1990) The geochemistry of tektites: an overview. Tectonophysics 171: 405-422

Koeberl C (1994) Tektite origin by hypervelocity asteroidal or cometary impact: target rocks, source craters, and mechanisms. In: Dressler BO, Grieve RAF, Sharpton VL (eds) Large Meteorite Impacts and Planetary Evolution. Geological Society of America Special Paper 293, pp 133-152

Ma P and 11 coauthors (2004) Beryllium-10 in Australasian tektites: constraints on the location of the source crater. Geochimica et Cosmochimica Acta 68: 3883-3896

Schwarz WH and 9 coauthors (2016) Coeval ages of Australasian, Central American and Western Canadian tektites reveal multiple impacts 790 ka ago. Geochimica et Cosmochimica Acta 178: 307-319

Shen G, Gao X, Gao B, Granger DE (2009) Age of Zhoukoudian Homo erectus determined with 26Al/10Be burial dating. Nature 458: 198-200



  1. John Zawiskie, Cranbrook Institute of Science on March, 2020 at 6:52 am

    Unlikely that it was Pithecanthropus (too young – 300,000 to 600,000 ybp) more likely – Sanigran H.erectus – ~800,000 ybp – more in line with age of the tektite fall.

    • Aaron Cavosie on March, 2020 at 7:12 pm

      Thanks for your comment. I agree, Pithecanthropus (e.g., Java man) is too young. However, the burial age of H. erectus pekinensis at 770,000 ± 80,000 years old, as described in Shen et al., 2009, does make it plausible that this species may have been present for the tektite forming event. One has to acknowledge that the uncertainty in the burial age (±80k years) is large, but it at least allows the proposal that early humans of some form likely saw, and were perhaps influenced, by the Australasian tektite-forming event. Thanks again for your detailed comment.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.