4.1 Description of the MN tektite fragments
The MN tektite fragments found at the HO06 section are black in color, vesicular, and show submillimeter-scale layering (Fig. 5). Under an optical microscope, thin sections of the fragments show layering composed of the dark-brownish–colored and light greenish-colored glass stripes, which are parallel to each other. Each stripe has 10-μm scale subparallel lamination characterized by alternation of dark-brownish and pale greenish lenses. Elongated vesicles of submillimeter size, the long axes of which are roughly parallel to the layering, are occasionally observed (Fig. 6). Mineral inclusions which are several micrometers in diameter are sparsely contained, but they are too small to identify mineral species under an optical microscope. These characteristics are typical for MN Australasian tektites found in Indochina (e.g., Wasson 1991; Glass et al. 2020).
Three-hundred and thirty-one MN tektite fragments were found in the small (~ 40 cm × 30 cm) area with 10 cm in thickness in the upper part of Unit 2 (Fig. 4) (One-hundred and seventy-seven fragments were recovered in the field after taking the block sample, and 154 fragments were recovered in the laboratory from the dismantled block sample.) The total weight of the 331 fragments is 713 g.
The length of a, b, and c axes of the fragments ranges from 2.00 to 52.80 mm, 1.95 to 40.00 mm, and 0.70 to 30.80 mm, respectively. The average length and standard deviation (in parenthesis) are 16.3 (8.7), 11.3 (7.1), and 7.2 (5.5) mm, respectively. These large standard deviations of the size of the tektite fragments correspond to “very poor sorting” in the scale of sorting for sediments (Folk and Ward 1957). The fragments are very angular, and fragile edges are well preserved, as shown in Fig. 5. Forty-six large fragments recovered from inside the block sample and seven large fragments recovered in the field after taking the block sample (53 tektite fragments in total) are fitted together to form a large MN tektite mass of ~ 370 g that shows an ellipsoidal shape (Fig. 7a). The rest of the fragments are too small and so many that it is difficult to find their positions to conduct restoration of the original tektite mass. However, the preliminary restoration result strongly suggests that tektite fragments forming the cluster composed one large tektite mass of > 713 g weight before the landing. The majority of the outermost part of the restored tektite mass was peeled off showing the section of the internal layered structure (Fig. 7b), while a typical pitted and grooved surface similar to that in splash-form tektites was preserved in some part (Fig. 7c) suggesting the original surface of the tektite mass was preserved. The fracture surfaces of the MN tektite fragments from HO06 are matt compared to a fresh fracture surface due to the presence of small pits, as shown in Fig. 7d. These matt surfaces of the tektites were formed by soil etching after their burial (Rost 1969; La Marche et al. 1984), suggesting that fracturing occurred in the geological past. Nearly all fragments have iron-oxide cement and calcite cement in some cases adhering to some of the fragment surfaces, and some vesicles open to the surface are filled with the iron-oxide and calcite cement materials indicating that the laterization of Unit 2 occurred after their fragmentation.
The fact that most of large tektite fragments recovered from the block sample fitted together to form a large ellipsoidal tektite mass with the original surface partly preserved indicates the tektite fragments were formed as a result of the fragmentation of one large tektite mass. The very poorly sorted nature of the MN tektite fragments, the very angular shapes with well-preserved fragile edges, and the occurrence of tektite fragments forming a cluster suggest that these MN tektite fragments were not transported a long distance but rather fragmented in situ.
4.2 Major element composition
The major element compositions of all nine MN tektite fragments from the HO06 section, together with their average compositions and standard deviations, are given in Table 2. The major element compositions of these samples fall within the range of compositions of previous analyses of MN tektites from Ubon Ratchathani province within 1σ, for most elements (Glass and Koeberl 1989; Koeberl 1992; Fiske et al. 1996; Herzog et al. 2008) (Fig. 8a), indicating that they are not distinguishable chemically from previously reported Australasian MN tektites from Ubon Ratchathani province.
Figure 8b shows SiO2 vs Al2O3, FeO, K2O, and MgO values for the nine MN tektite fragments from the HO06 section plotted with values reported for other Australasian MN tektites from Ubon Ratchatani province ( Glass and Koeberl 1989; Koeberl 1992; Fiske et al. 1996; Herzog et al. 2008). The major element composition of the nine fragments is closely plotted in a narrow area and indistinguishable from each other within 1σ. This narrow range of the major element composition of the nine fragments is consistent with the idea that the MN tektite fragments collected at HO06 section originally formed one large MN tektite mass.
4.3 Size distribution of the MN tektite fragments
Ds for the fragments generated by fragmentation of rocks or glass is proportional to the intensity of fracturing in which fragments generated by higher magnitude and rate of stress loading tend to have higher value of Ds, although Ds is affected also by the inherent strength properties of the rock or glass (e.g., Takagi et al. 1984; Jébrak 1997; Roy et al. 2012; Xu 2018). For instance, Ds for rock fragments generated by weathering of andesite in the field and hammering of a limestone cube by hand is 2.5–2.7 (calculated from the data of Domokos et al. 2015), and that for rock fragments generated by rockfalls and rock avalanches with various lithology ranges from 1.6 to 4.7 (Crosta et al. 2007; Ruiz-Carulla and Corominas, 2020). Rock fragments generated by complete fragmentation of target rocks in hypervelocity impact experiments show bi- or tri-fractal distributions in which Ds for the finer fraction is 1.7 to 4.3 and Ds for the coarser fraction is 4.8 to 11.9 (calculated from the data of Takagi et al. 1984; Michikami et al. 2016). The Ds for fragments of a block on the Moon fragmented by a small meteorite impact is around or higher than 4 (Ruesch et al. 2020).
Figure 9 shows the cumulative size distribution of the MN tektite fragments at HO06 section. The size of the fragment is represented by the length of its equivalent spherical diameter (r = \( \sqrt[3]{abc} \)). The cumulative size distribution increases rapidly from 37 to 26 mm and then increases slowly in a range smaller than 26 mm. It shows a bi-fractal distribution described by two power laws in the range from 10 to 26 mm and from 26 to 37 mm, with Ds of 2.2 and 7.5, respectively. The position of the point dividing the two power laws is calculated by adjusting the dividing point every 0.5 cm to maximize the average of the R2 values. The bi-fractal distribution of fragment size implies that the size distribution of the fragments was affected by two different fragmentation mechanisms (e.g., Schultz and Gault 1990). Although the effect of the difference of physical properties between the tektite glass and rocks need to be considered, the high Ds value (7.5) for the coarse fraction of the tektite fragments is larger than the range of Ds previously reported for rock fragments generated by rockfalls and rock avalanches (1.6–4.7; Crosta et al. 2007; Ruiz-Carulla and Corominas 2020) and similar to the Ds for the coarser fraction fragments generated by hyperspeed impact experiments (4.8–11.9; calculated from the data of Takagi et al. 1984; Michikami et al. 2016) and impact fragmentation on the Moon (3.3–6; Ruesch et al. 2020), suggesting that the tektite fragments were formed through intense fragmentation by a relatively high energetic process.
Based on the close spatial association of the MN tektite fragments at Huai Sai, Fiske et al. (1996) proposed that the fragmentation of an original MN tektite mass was a low-energy process such as weather or climate-related temperature-induced fracturing after the deposition of the MN tektite mass. However, as is mentioned above, the size distribution of the coarse fraction of the MN tektite fragments from HO06 section measured in this study suggests a relatively high-energy process rather than a low-energy process such as weather or climate-related temperature-induced fracturing. The possibility that the MN tektite mass was broken by tectonic movement after deposition is also unlikely because we did not see any evidence of tectonic deformation at the outcrop. The other possibility is that the fragmentation occurred during the flight by high-speed collision of ejecta materials before the deposition. However, this possibility is also unlikely because if the fragmentation occurred during the flight, the fragments would have separated before they landed on the ground and would not have formed a cluster. Consequently, the timing of the fragmentation of the MN tektite mass is constrained to the time of the landing on the ground. We propose that an original tektite mass was fragmented by collision of the MN tektite mass with the ground. Either the MN tektite fragments were considered to have been co-deposited with other ejecta materials immediately after (or almost the same time) the fragmentation or the MN tektite mass was fragmented during penetration into the unconsolidated ejecta of Unit 2 so that the fragments remained as a cluster.
As for the fragmentation mechanism for the finer fraction of the MN tektite fragments, one possibility is that some of the tektite fragments were further fractured by weather or climate-related thermal fracturing after the deposition. Another possibility is that some of the tektite fragments generated by the collision of the original tektite mass with the ground surface were secondarily fractured by collision with the surrounding other ejecta materials immediately after the first fragmentation, resulting in the bi-fractal size distribution of the tektite fragments. In any case, the whole size distribution of the tektite fragments, especially the high Ds value for the coarse fraction, is different from the size distributions of rock fragments generated by low-energy processes.
4.4 Preliminary observation of CT scan 3D image of the block sample
In CT cross-section images, tektite fragments appear as angular-shaped areas with moderate X-ray transmittance, exhibiting elongated vesicles. The preliminary CT scan 3D image observation of the block sample revealed that the MN tektite fragments are distributed in 20 × 10 × 10 cm space as if they were expanded from the original tektite mass by fragmentation. This distribution of the tektite fragments suggests either that the tektite fragments were buried immediately after fragmentation on landing by co-deposited other ejecta materials of Unit 2 or that fragmentation occurred during the tektite mass penetrated into the ground cover of unconsolidated ejecta (granule to pebble-bearing sand) of Unit 2 so that tektite fragments have not been spread significantly. This distribution of the tektite fragments is inconsistent with the previous interpretation in which these tektites were considered to have been reworked on the paleo surface because if reworked on the paleo surface, the fragments would be distributed on a plane.
4.5 Stratigraphic position of MN Australasian tektite in Indochina
As is proposed in the previous sections, the fragmentation of the MN tektite mass occurred at the time of the landing on the ground, and either the fragments were buried immediately with other ejecta materials or the tektite mass penetrated into the ground covered by unconsolidated ejecta deposit and fragmented within the ejecta. In either case, the fragments were preserved as a cluster. Thus, the MN tektite fragments in the upper part of Unit 2 (“laterite” layer) at the HO06 section are considered as in primary position.
Fiske et al. (1996) considered that layered tektites found at the top of Unit 2 at the Huai Om and Huai Sai sections to be reworked because, in their opinion, the top of Unit 2 represents a paleo-erosional surface. However, the fact that MN tektite fragments are found not at the top of the “laterite” layer but within the “laterite” layer at the HO06 section does not support this interpretation. Also, there was no evidence of erosion at the top of Unit 2 at the nearby Huai Om section (Tada et al. 2019) as well as several other sections in the region. Furthermore, our field observation of the “laterite” layer in other localities (for example, an active sand pit at Noen Sa-nga in Chaiyaphum province) suggests that laterization (precipitation of iron hydroxides) of this layer can occur within a few years after road cuts and pit walls were formed, indicating that the hard-cemented appearance of the “laterite” layer was not formed as an erosional pavement before the impact, but formed on the outcrop surface by recent ferricretization (Fig. S2). Koeberl and Glass (2000) and Keates (2000) pointed out that the Australasian tektites in Australia generally are found in deposits younger than the age of the impact, and interpreted by analogy that the tektites in Indochina may be of similar reworked origin. However, this analogy is not based on observational evidence. It should also be considered that dating of Quaternary sediments in Indochina is replete with methodological difficulties that may result in young ages being reported for strata that are considerably older (Carling et al. 2020).
The in situ occurrence of the MN tektite fragments in Unit 2 at the HO06 section supports the idea that the Quaternary depositional sequence of Units 1–3 in this region is an ejecta deposit of the Australasian tektite event based on the finding of shocked quartz from Units 1–3 at Huai Om section (Tada et al. 2019). The fact that both the MN and splash-form tektites generally were found in the upper part of or at the top of the “laterite” layer at sites in the wide area of eastern Indochina (Table 1, Fig. 1) further supports the idea that the Australasian tektites found from the “laterite” layer were deposited in situ as an ejecta.