doi: 10.1111/j.1365-3121.2008.00791.x Comment article Evidence that Lake Cheko is not an

G. S. Collins,1 N. Artemieva,2 K. Wu¨ nnemann,3 P. A. Bland,1 W. U. Reimold,3 and C. Koeberl4 1Impacts and Astromaterials Research Centre, Department of Earth Science and Engineering, Imperial College London, SW7 2AZ London, UK; 2Institute for the Dynamics of Geospheres, Russian Academy of Sciences, Moscow, Russia; 3Museum for Natural History, Humboldt University, Invalidenstrasse 43, 10115 Berlin, Germany; 4Center of Earth Sciences, University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria

ABSTRACT In a provocative paper Gasperini et al. (2007) suggest that Lake is required for an asteroid fragment to traverse the EarthÕs Cheko, a 300-m-wide lake situated a few kilometres down- atmosphere and reach the surface intact and with sufficient range from the assumed epicentre of the 1908 Tunguska event, velocity to excavate a crater the size of Lake Cheko. Inferred is an impact crater. In this response, we present several lines of tensile strengths of large stony during atmospheric observational evidence that contradicts the impact hypothesis disruption are 10–100 times lower. We therefore conclude that for the lakeÕs origin: un-crater-like aspects of the lake mor- Lake Cheko is highly unlikely to be an impact crater. phology, the lack of impactor material in and around the lake, and the presence of apparently unaffected mature trees close Terra Nova, 20, 165–168, 2008 to the lake. We also show that a tensile strength of 10–40 MPa

over 170 confirmed impact structures The first piece of evidence does give Introduction on the Earth (Earth Impact Database, pause for thought, but could easily be An impact origin for Lake Cheko and 2007). Every new, ÔconfirmedÕ impact coincidence. The second argument is connection with the Tunguska event crater provides important information key: if the lake pre-dates 1908, it was was dismissed by a Russian expedition to further our understanding of obviously not formed during the in the 1960s when a tentative age of impacts and the hazard they pose. Tunguska event. At this stage, there 5–10 ka for the lake was made based However, every false impact crater is no convincing evidence that the lake on the thickness (7 m) of mud depos- that is proposed clouds our under- is only 100 years old; anecdotal evi- its (Florenskij, 1963). On the basis of standing and confuses the public (Rei- dence cannot be relied on, and recent new geophysical data and shallow lake mold, 2007), and for this reason the collapse features do not imply that the sediment cores, Gasperini et al. (2007) burden of proof in identifying a new lake formed recently. The third piece argue against such an old age for the must lie with the of evidence is the weakest; geophysical lake. They hypothesize that a 10-m- proponents. Disappointingly, in the anomalies can be interpreted in many diameter fragment of the Tunguska case of Lake Cheko, very little evi- ways and never provide conclusive asteroid or comet survived the main dence has been supplied in support of evidence of impact. For the bright atmospheric disturbance, continued on the impact hypothesis for its origin, reflector to be caused by the impacting its course to collide with the ground at and none of it is compelling. Gaspe- body implies an unrealistically large a velocity of many kilometres per rini et al. (2007) provide four argu- and robust impactor, to survive second, and formed Lake Cheko. ments in support of an impact origin impact intact and be resolvable in Impact cratering is an important for the lake: the seismic data. It is far more likely geological process that has cata- that the bright reflector is sedimentary 1 The location of the lake is consis- strophically affected the global envi- in origin. Finally, the morphology of tent with the continuation of the ronment. Much of our current the lake is, in many ways, very differ- assumed trajectory of the Tunguska knowledge of the impact process and ent from small impact craters on impactor beyond the epicentre of hazard is derived from the study of Earth, as we discuss below, and there- the explosion. fore does not provide compelling evi- 2 The age of the lake is unknown. dence for an impact origin. Correspondence: Gareth S. Collins, Depart- 3 A bright seismic reflection is appar- ment of Earth Science and Engineering, ent in the seismic data beneath the Imperial College London, South Kensing- lake. The authors claim that this The Tunguska Event ton Campus, SW7 2AZ London, UK. might be evidence for impactor Years of theoretical, analogue and Tel.: +44 (20) 75941518; fax: +44 (20) material or impact-compaction of numerical modelling work have been 75947444; e-mail: [email protected] the sediments. devoted to explaining the major 4 The lake has a funnel-like morphol- consequences of the Tunguska event A response to: A possible impact crater for ogy. The authors claim this is on 30th June 1908. The most recent the 1908 Tunguska Event L. Gasperini, F. unusual for the area and similar to numerical model, which agrees with Alvisi, G. Biasini, E. Bonatti, G. Longo, M. other small impact craters on Earth. Pipan, M. Ravaioli and R. Serra. the consensus of earlier models

2008 Blackwell Publishing Ltd 165 Evidence that Lake Cheko is not an impact crater • G. S. Collins et al. Terra Nova, Vol 20, No. 2, 165–168 ......

(Korobeinikov et al., 1976; Chyba but this would occur close to, and up the form of iron fragments et al., 1993; Boslough and Crawford, range of, the epicentre. or Fe-Ni metal grains and spherules 1997), suggests that the Tunguska For a high-velocity impact crater to has been found (Table 1). No trace of event was caused by a cosmic body form as part of the atmospheric dis- impactor material has yet been found 50–80 m in diameter entering the turbance over Tunguska in 1908, in or around Lake Cheko, and if it EarthÕs atmosphere at 20 km s)1 and therefore, is contrary to our current exists it is extremely unlikely to be at an inclination of 30–45 to the understanding of the event. Neverthe- iron. If Lake Cheko is a 300-m- horizontal (Artemieva and Shuvalov, less, Gasperini et al. (2007) estimate diameter impact crater formed by a 2007). At an altitude of 20 km, the that a 10 m diameter1 Asteroid frag- stony meteorite, it would be anoma- impactor starts to deform, disrupt and ment impacting at a velocity of 1– lous in terms of its impactor compo- evaporate strongly; the resulting jet of 10 km s)1 is required to form an sition. vaporized impactor material is totally impact crater the same size as Lake Another important observation is decelerated at an altitude of 8–10 km Cheko. For this putative impactor to that all young (<10 ka), 0.1- to 0.3- and releases all its energy into the survive passage through the atmo- km-diameter meteorite craters are not atmosphere (5–15 Mt, high explosive sphere from an altitude of 8km isolated craters, but are part of crater equivalent, consistent with estimates (the assumed height of the air blast Ôstrewn fieldsÕ, formed by the near- from seismic records Ben-Menahem, centre) to the ground, the fragment simultaneous collision of a number of 1975). The mixture of hot air and must have a tensile strength of at least dispersed fragments of the same ori- vaporized impactor material is buoy- 10 MPa if it strikes the ground at ginal meteoroid (Table 1). Given the ant and accelerates back along the 5kms)1, and of at least 40 MPa if it obvious disruption that occurred to wake, while an atmospheric shock strikes the ground at 10 km s)1. the bulk of the Tunguska impactor, it wave reaches the surface. The interac- Although these values are within the is very hard to explain how Lake tion of the shock wave with the range of laboratory measurements of Cheko could have formed by impact surface causes the observed famous tensile strengths of cm-sized meteorite in isolation. If Lake Cheko is an forest devastation and fallen tree samples (Grady and Lipkin, 1980; impact crater, and depending on the pattern (Florenskij, 1963); the model Medvedev et al., 1985; Petrovic, composition of the impactor and exact results are consistent with observa- 2001), the dynamic strength of larger impact velocity, a number of solid tions of total area and ÔbutterflyÕ plan- objects (1–20 cm), observed disrupt- fragments of the impactor should be form of the damage area, and the ing as they enter the EarthÕs atmo- preserved in and around the lake. telegraph poles and trees that sphere, are substantially lower, rarely remained vertical near the epicentre. exceeding 1 MPa (Ceplecha et al., Other observational evidence All the extra terrestrial impactor 1993). The low strength of stony material (in the form of tiny droplets) objects is further indirectly confirmed In addition to hard evidence of impact is carried ÔuprangeÕ away from the by an absence of large stony meteor- that is lacking for Lake Cheko, there epicentre, reaches high altitudes and ites on the Earth. Hence, the impact are several observable differences may then disperse worldwide. It is hypothesis for Lake Cheko requires between the lake and small meteorite most likely that this dispersal of hot that the impactor must have been an impact craters on the Earth. Gasperini condensed impactor material caused exceptionally strong and large frag- et al. (2007) compare Lake ChekoÕs the strange optical effects (white ment of a much weaker body that cross-section with Odessa crater, nights) observed in Northern Eurasia disrupted catastrophically to cause the Texas (actually the largest crater in (Whipple, 1930). Tunguska event. the Odessa crater strewnfield; 170 m diameter), and use this alleged simi- larity as evidence in support of the Surviving atmospheric entry Hard evidence Lake Cheko impact hypothesis. How- Importantly, the model of Artemieva To confirm unequivocally that Lake ever, in many other ways, Odessa and Shuvalov (2007) requires that the Cheko was the result of the collision crater and Lake Cheko are very comet or asteroid had little strength of an extraterrestrial body with Earth different. (<5 MPa). This is consistent with would require discovery of meteoritic All fresh impact craters have raised stony meteorite strength estimates of impactor material and ⁄or impact melt rims, surrounded by a continuous 0.1–1 MPa, based on observed atmo- in or around the lake. In ten out of blanket of ejected material. At Odessa spheric break-up events (Ceplecha twelve confirmed, 0.1- to 1-km-diam- crater, for example, the rim of the et al., 1993), and implies that an iron eter meteorite impact craters on crater is uplifted by 4 m and sur- impactor is unlikely. The model does Earth, bona fide impactor material in rounded by a proximal layer of ejected not predict that solid fragments great- material up to 1 m thick (Holliday er than a few centimetres can reach et al., 2005). This is despite erosion of the ground, consistent with the 1In fact, the 10-m diameter estimate is the crater rim and in the absence of extraterrestrial material appropriate only for an impact velocity of 63 000 years since formation (Holli- near the epicentre (Florenskij, 1963). 10 km s-1. At lower impact velocities, the day et al., 2005). At Lake Cheko, Some fine material not resolved by the asteroid fragment must be larger than 10-m there is no evidence of uplift of the model may fall out from the plume, diameter; for example, a 20-m-diameter rim, although it is possible that some travelling at terminal velocity (tens of asteroid is required at an impact velocity of of the uplifted rim collapsed into the metres per second for cm-size objects), 5kms-1 (Holsapple, 1993). crater if the target material was weak.

166 2008 Blackwell Publishing Ltd Terra Nova, Vol 20, No. 2, 165–168 G. S. Collins et al. • Evidence that Lake Cheko is not an impact crater ......

Table 1 Confirmed terrestrial impact craters with diameter between 0.1 and 1 km and inferred impactor types. Name Country Diameter (km) Age (ka) Location Impactor type References

Morasko Poland 0.1* <10 N 5229¢ E1654¢ IIICD Iron Korpikiewicz (1978) Kaalijarvi Estonia 0.11* 4 ± 1 N 5824¢ E2240¢ IAB Buchwald (1975) Wabar Saudi Arabia 0.12* 0.14 N 2130¢ E5028¢ IIIAB Iron Morgan et al. (1975) Henbury Australia 0.16* 4.2 ± 2 S 2434¢ E 1338¢ IIIAB Taylor (1967) Odessa USA 0.17* 63 N 3145¢ W 10229¢ IIIAB Buchwald (1975) Boxhole Australia 0.17 54 ± 1 S 2237¢ E 13512¢ IIIAB Buchwald (1975) Macha Russia 0.3* <7 N 606¢ E 11735¢ Iron Gurov (1996) Aouelloul Mauritania 0.39 3000 ± 300 N 2015¢ W1241¢ Iron Morgan et al. (1975); Koeberl et al. (1998) Amguid Algeria 0.45 <100 N 265¢ E423¢ Unknown Monturaqui Chile 0.46 <1000 S 2356¢ W6817¢ IAB Bunch and Cassidy (1972); Buchwald (1975) Kalkkopà South Africa 0.64 <1800 S 3243¢ E2434¢ Chondrite? Koeberl et al. (1994); Reimold et al. (1998) Wolfe Creek Australia 0.88 <300 S 1910¢ E 12748¢ IIIAB Attrep et al. (1991)

Adapted from Koeberl (1998) and Earth Impact Database (2007). *Crater strewnfield; largest dimension of largest structure is given. No data are available for Amguid. àKalkkop has only tentatively been connected with a chondritic impactor type (Koeberl et al., 1994; Reimold et al., 1998).

Neither is there evidence of any for meteoritic material, or any other Bottke, W.F., Love, S.G., Tytell, D. et al., ejected material; in fact, aerial photos hard evidence of impact, has yet been 2000. Interpreting the elliptical crater of the lake from 1938 and 1999 show found in or near Lake Cheko. The populations on Mars, Venus, and the mature trees that pre-date 1908 lining lake is in many ways substantially Moon. Icarus, 145, 108–121. the rim of the lake (Longo and Di different from confirmed, young, ter- Buchwald, V.F., 1975. Handbook of Iron Meteorites: their History, Distribution, Martino, 2003). It is hard to imagine restrial impact craters of the same size. Composition and Structure. University how a violent could It is also surrounded by trees that of California Press, Berkeley. excavate a 300-m-wide hole without would have been flattened as a conse- Bunch, T.E. and Cassidy, W.A., 1972. affecting trees so close. The ground quence of the impact. Taken individ- Petrographic and electron-microprobe movement and ejecta deposition alone ually, none of these arguments can study of Monturaqui . Contrib. should have been enough to flatten all conclusively rule out the impact Mineral. Petrol., 36, 95–112. proximal vegetation. hypothesis for the origin of the lake. Ceplecha, Z., Spurny, P., Borovicka, J. Very few impact craters are elliptical The easiest way to do so would be to et al., 1993. Atmospheric fragmentation in plan view like Lake Cheko, which establish an accurate age for the lake. of meteoroids. Astron. Astrophys., 279, 615–626. has an ellipticity of 4 3. Impact However, together this evidence  ⁄ Chyba, C.F., Thomas, P.J. and Zahnle, experiments and statistical analysis of makes it extremely unlikely that Lake K.J., 1993. The 1908 Tunguska explo- crater shapes on the terrestrial planets Cheko is of impact origin. sion ) atmospheric disruption of a stony show that crater ellipticity is controlled asteroid. Nature, 361, 40–44. by impact angle (to the target plane) Earth Impact Database, 2007. http:// References and that departure from circularity www.unb.ca/passc/ImpactDatabase/ occurs only for very oblique (<10) Artemieva, N. and Shuvalov, V., 2007. 3D (accessed 6 July 2007). impacts (Gault and Wedekind, 1978; effects of Tunguska Event on the ground Florenskij, K.K.P., 1963. Predvaritelnyje Bottke et al., 2000). If Lake Cheko was and in the atmosphere. Lunar and rezultaty Tungusskoj meteoritnoj Planetary Science Conference XXXVIII, formed at the same time as the 1908 kompleksnoj ekspeditsii 1961. Meteori- Lunar and Planetary Institute, Houston, tika, 23,3. Tunguska event, then its location TX, abstract no. 1537. Gasperini, L., Alvisi, F., Biasini, G., relative to the blast epicentre (8 km Attrep, M., Orth, C.J., Quintana, L.R., Bonatti, E., Longo, G., Pipan, M., downrange) and the estimated altitude Shoemaker, C.S., Shoemaker, E.M. and Ravaioli, M. and Serra, R., 2007. A of the main explosion (5–10 km) imply Taylor, S.R., 1991. Chemical fraction- possible impact crater for the 1908 an impact angle of 30–50. A high- ation of siderophile elements in impac- Tunguska Event. Terra Nova, 19, velocity impact at this angle produces tites from Australian meteorite craters. 245–251. an almost circular crater. Lunar and Planetary Science XXII, Gault, D.E. and Wedekind, J., 1978. Lunar and Planetary Institute, Houston, Experimental studies of oblique impact. TX, 39–40. Proc. Lunar and Planet. Sci. Conf. 9th, Summary Ben-Menahem, A.A., 1975. Source Pergamon Press, New York, 3843–3875. parameters of the Siberian explosion of Grady, D.E. and Lipkin, J., 1980. Criteria The Tunguska event is well explained June 30, 1908, from analysis and syn- for impulsive rock fracture. Geophys. by models that employ a weak impac- thesis of seismic signals at four stations. Res. Lett., 7, 255–258. tor strength, consistent with observed Physics Earth Planet. Int., 11,1. Gurov, E.P., 1996. The group of Macha stony meteorite break-up events; these Boslough, M.B.E. and Crawford, D.A., craters in Western Yakutia. Lunar and models do not predict that large, high- 1997. Shoemaker-Levy 9 and plume- Planetary Conference XXVII, Lunar velocity fragments of the impactor forming collisions on Earth. Ann. N.Y. and Planetary Institute, Houston, TX, could reach the surface. No evidence Acad. Sci., 822, 236–282. 473–474.

2008 Blackwell Publishing Ltd 167 Evidence that Lake Cheko is not an impact crater • G. S. Collins et al. Terra Nova, Vol 20, No. 2, 165–168 ......

Holliday, V.T., Kring, D.A., Mayer, J.H. Korobeinikov, V.P., Chushkin, P.I. and Petrovic, J.J., 2001. Mechanical properties et al., 2005. Age and effects of the Shurshalov, L.V., 1976. Mathematical of meteorites and their constituents. Odessa meteorite impact, western Texas, model and computation of the Tunguska J. Mater. Sci., 36, 1579–1583. USA. Geology, 33, 945–948. meteorite explosion. Acta Astronaut., 3, Reimold, W.U., 2007. The impact crater Holsapple, K.A., 1993. The scaling of 615–622. bandwagon (some problems with the impact processes in planetary sciences. Korpikiewicz, H., 1978. Meteoritic terrestrial cratering record). Meteorit. Annu. Rev. Earth Planet. Sci., 21, shower Morasko. Meteoritics, 13, 311– Planet. Sci., 42, 1467–1472. 333–373. 326. Reimold, W.U., Koeberl, C. and Redder- Koeberl, C., 1998. Identification of mete- Longo, G. and Di Martino, M., 2003. ing, J.S.V., 1998. The 1992 drill core oritic components in . In: Remote sensing investigation of the from the Kalkkop impact crater, Eastern Meteorites: Flux with Time and Impact Tunguska explosion area. International Cape Province, South Africa: stratigra- Effects (M.M. Grady, R. Hutchison, Conference Remote Sensing for phy, petrography, geochemistry and age. G.J.H. McCall and D.A. Rothery, eds), Agriculture, Ecosystems, and Hydrology J. Afr. Earth Sci., 26, 573–592. pp. 133–153. Geological Society, IV, SPIE, USA. Taylor, S.R., 1967. Composition of mete- London. Medvedev, R.V., Gorbatsevich, F.I. and orite impact glass across the Henbury Koeberl, C., Reimold, W.U., Shirey, S.B. Zotkin, I.T., 1985. Determination of the strewn field. Geochim. Et Cosmochim. et al., 1994. Kalkkop Crater, Cape physical properties of stony meteorites Acta, 31, 961–968. Province, South-Africa – Confirmation with application to the study of pro- Whipple, F.J.W., 1930. The great Siberian of impact origin using osmium isotope cesses of their destruction. Meteoritika meteor and the waves, seismic and aerial, systematics. Geochim. Cosmochim. Acta, (in Russian), 44, 105–110. which it produced. Quart. J. Royal 58, 1229–1234. Morgan, J.W., Higuchi, H., Ganapathy, R. Meteorol. Soc., 56, 287–304. Koeberl, C., Reimold, W.U. and Shirey, and Anders, E., 1975. Meteoritic S.B., 1998. The , Mau- material in four terrestrial meteorite Received 11 July 2007; revised version ritania: on the problem of confirming the craters. 6th Lunar Science Conference, accepted 6 December 2007 impact origin of a small crater. Meteorit. Lunar and Planetary Institute, Houston, Planet. Sci., 33, 513–517. 1609–1623.

168 2008 Blackwell Publishing Ltd