Central ring structure identified in one of the world’s best-preserved impact craters

A.C. Gebhardt F. Niessen Alfred Wegener Institute for Polar and Marine Research, Columbusstraße, 27568 Bremerhaven, Germany C. Kopsch Alfred Wegener Institute for Polar and Marine Research, Telegrafenberg A43, 14473 Potsdam, Germany

ABSTRACT Seismic refraction and re¯ection data were acquired in 2000 and 2003 to study the morphology and sedimentary ®ll of the remote El'gygytgyn crater (Chukotka, northeast- ern ; diameter 18 km). These data allow a ®rst insight into the deeper structure of this unique . Wide-angle data from sonobuoys reveal a ®ve-layer model: a water layer, two lacustrine sedimentary units that ®ll a bowl-shaped apparent crater mor- phology consisting of an upper layer of fallback with P-wave velocities of ϳ3000 m/s, and a lower layer of brecciated bedrock (velocities Ͼ3600 m/s). The lowermost layer shows a distinct anticline structure that, by analogy with other terrestrial and lunar cra- ters of similar size, can be interpreted as a central ring structure. The El'gygytgyn crater exhibits a well-expressed morphology that is typical of craters formed in crystalline target rocks.

Keywords: impact crater, El'gygytgyn, , seismic refraction, seismic re¯ection.

INTRODUCTION (Belyi, 1998; Gurov et al., 1979a, 1979b). The Grette, 2006; Nolan et al., 2002) formed in- The El'gygytgyn crater, located in the Rus- Anadyr Mountains are part of the Okhotsk- side the crater (the crater and crater are sian Arctic, is one of the world's best- Chukotka Volcanic Belt, composed of Late not concentric; see Fig. 1), and has become a preserved impact craters of its size (Dietz and Cretaceous calc-alkaline volcanic rocks (Be- major focus of multidisciplinary international McHone, 1976), and is the only known crater lyi, 1994; Layer, 2000; Layer et al., 2001). research (e.g., Nolan et al., 2002; Nowaczyk on Earth that was formed by an impact into The target rocks are mainly rhyodacite tuffs et al., 2002) because it could provide a 3.6 siliceous volcanic rocks (e.g., Gurov et al., and ignimbrites, with some rhyolites, andesite m.y. paleoclimatic record unique to the Arctic. 1979a, 1979b; Gurov and Koeberl, 2004); its tuffs, and basalts. A lake of ϳ12 km diameter When the impact formed the crater, the Arctic lake is one of only a few inside an impact and 170 m water depth (Nolan and Brigham- was forested, and ϳ1 m.y. passed before the crater (Lerman et al., 1995). The crater's age, 3.6 Ma (Layer, 2000), and its excellent pres- ervation provide unique information about im- pact mechanisms and in siliceous volcanic rock. The lacustrine sedi- ments have never been glaciated (Glushkova and Smirnov, 2005; Minyuk et al., 2003), and so may contain a high-resolution paleoclimate record of the past 3.6 m.y. In a ®rst step to- ward continental deep drilling, a seismic re- fraction and re¯ection survey was carried out to obtain information on the structure of the crater and the geometry of its sedimentary ®ll (Niessen et al., 2006). Seismic refraction data form the basis of a ®ve-layer velocity-depth model that reveals the structural geometry of the impact crater, and multichannel re¯ection data provide information on its lacustrine sed- imentary ®ll.

GEOLOGICAL BACKGROUND AND PREVIOUS WORK The 18-km-diameter El'gygytgyn crater is Figure 1. A: Geological overview of crater and its surroundings (modi®ed after Nowaczyk located in the remote Anadyr Mountains (Fig. et al., 2002). B: General overview of crater location. Note that crater and crater lake are not 1) in central Chukotka, northeastern Siberia concentric.

᭧ 2006 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected]. Geology; March 2006; v. 34; no. 3; p. 145±148; doi: 10.1130/G22278.1; 4 ®gures. 145 Figure 2. Sonobuoy refraction data. Left: W-E pro®le. Right: N-S pro®le. Two seismic record sections are given in upper part (A, B); picked phases are marked as Ps (sedimentary units) and Pg (breccia and bedrock). Observed and calculated traveltimes are shown in C and D. Picks as error bars indicate pick uncertainties, and modeled synthetic rays are lines. Five-layer seismic velocity model is shown in E and F. For locations of pro®les and sonobuoys, see Figure 1.

®rst major glaciation of the Northern Hemi- velocities for the lower layers (breccia and ture are ϳ50 m in the central part and ϳ80 m sphere (Brigham-Grette and Carter, 1992). To- bedrock). in the more distal part of the refraction data. day's lake level is ϳ500 m above sea level In 2000, a single 5 in3 (82 cm3) Bolt 600B (asl), and the highest elevations of the crater airgun was used as seismic source for the re- RESULTS AND INTERPRETATION rim reach 850±950 m asl; the crater rim forms fraction pro®les. The shot interval was 6 s, The ®nal ®ve-layer model (Fig. 2) includes the outer margin of the catchment area of the equating to an average of 8 m shot distance. a water layer and two upper sedimentary units lake. Two sonobuoys were used on two perpendic- of low seismic velocities (1550 m/s and 1650 ular pro®les across the center of the crater lake m/s); the upper unit (Ia) is 170 m thick, and SEISMIC DATA ACQUISITION AND (Fig. 1) and recorded the airgun-borne acous- the lower unit (Ib) is 190±250 m thick. These PROCESSING tic pulses by a single hydrophone under the units are underlain by a unit with a distinctly During two expeditions in 2000 and 2003, buoy. Signals were ampli®ed and transmitted higher seismic velocity of ϳ3000 m/s and a seismic refraction and multichannel re¯ection via radio to the platform (Niessen et al., thickness of 100±400 m (unit II). The lower- data were obtained using 2 sonobuoys and a 2006). Both sonobuoy pro®les were reversed. most unit (III) has a seismic velocity of 260-m-long streamer (Niessen et al., 2005). The seismic refraction data recorded by the Ͼ3600 m/s. Due to the remote location (Fig. 1), all geo- sonobuoys covered an offset range to 4.4 km The most striking feature revealed by the physical equipment, including the vessel (a 3 (E-W pro®le) and 3.3 km (N-S pro®le, Fig. 2) sonobuoy data is the clearly visible uplift ϫ 4 m aluminum platform), had to be trans- and were improved by standard processing structure in the lowermost layer of our model. ported to the lake by helicopter. For the mul- (bandpass ®ltering 3-5-50-60 Hz, AGC) to Its position in the two perpendicular pro®les tichannel survey, a 26 in3 (426 cm3) mini-GI identify different re¯ection and refraction shows that it is located in the southwestern gun (triggered in G-gun mode) at a pressure phases. Picking and modeling of the identi®ed part of the lake (Fig. 1), ϳ2 km away from of 110 bar was used. The shot interval was 10 phases (Pg and Ps, Fig. 2) were done using the crater center. Complex impact craters are s, resulting in ϳ12 m shot distance. A 14 the ``zplot'' and ``rayinvr'' program package characterized by a central uplift area consist- channel streamer with an initial offset of 130 (Zelt and Smith, 1992). A forward ray-tracing ing of a central topographic peak and/or m and a hydrophone spacing of 10 m was modeling technique was used to develop a ring(s) (Grieve and Pesonen, 1992). The crater used as a receiving array. Re¯ection data were model of layer thicknesses and velocities. The morphology is strongly related to its size and processed in a standard sequence including initial ®ve-layer model was used to calculate target rock (Grieve and Pesonen, 1992; Spu- bandpass ®ltering (70-90-240-300 Hz), veloc- synthetic seismic traveltimes. It was iterative- dis, 1993). Grieve and Pesonen (1992) report- ity analysis, CMP (common midpoint) stack- ly modi®ed until the calculated and observed ed a central uplift morphology succession ing, and predictive deconvolution. The ®nal traveltimes were concordant (Fig. 2B). Mod- from craters with a simple central peak (e.g., stacking was done using the refraction data eling uncertainties for the central ring struc- Steinheim crater, Flynn Creek crater, both with

146 GEOLOGY, March 2006 the central uplift structure and is 100 m thick on top of the central ring, increasing to 400 m thick in the surrounding annular basin. Com- parable layers have been found, for example, in the Bosumtwi crater (velocities of ϳ3200 m/s; Karp et al., 2002; Scholz et al., 2002) and in the NoÈrdlinger Ries crater (e.g., Engelhardt et al., 1994; WuÈnnemann and Morgan, 2004). Seismic velocities are much lower within the lacustrine layers, averaging 1550 m/s in the upper and 1650 m/s in the lower sedimen- tary unit. The lower sedimentary unit shows a slight drape on top of the central uplift, Figure 3. Re¯ection pro®le along W-E sonobuoy pro®le. Locations of pro®le and sonobuoys whereas the contact between the two sedimen- are given in Figure 1. Colors at sonobuoy positions correspond to Figure 2. MÐmultiples. tary units is almost ¯at. Re¯ection seismic data (Fig. 3) show a well-strati®ed 170-m- thick upper unit that is locally intercalated diameters of 3.8 km), to craters whose central ring; there is no evidence for a central peak with debris ¯ows, mainly at the lake margins. within this ring. Figure 1 shows the position peak basins consist of both a central peak and The lower unit is more massive and is 190± a surrounding ring (e.g., Mistastin crater, 28 of the proposed central ring. The layer form- 250 m thick. Considering an age of 3.6 Ma km), to large craters with one or several rings, ing the ring is characterized by seismic veloc- and a lacustrine ®ll thickness of 360±420 m, but without a central peak (e.g., West Clear- ities of Ͼ3600 m/s and is interpreted to con- an average sedimentation rate of 0.1±0.12 water crater, 32 km; Puchezh Katurki crater, sist of parautochthonous bedrock forming the mm/yr can be estimated. 80 km; , 100 km; Sudbury cra- crater ¯oor. P-wave velocities in hand speci- ter, 250 km). The transition between the dif- mens of bedrock outcrops around the lake are ferent types depends on the target rock (Spu- generally much higher (ϳ4950 m/s, Niessen DISCUSSION dis, 1993, and references therein). According et al., 2006). This suggests that the bedrock is The El'gygytgyn created an to our work, the El'gygytgyn crater (18 km) brecciated at least in its upper part, as is the with a diameter of 18 km and is likely to have a central uplift structure sev- case in other craters (e.g., Bosumtwi crater, a central uplift morphology that consists of a eral kilometers in diameter, rather than a sin- Ghana, Africa; Karp et al., 2002; Scholz et al., concentric inner ring (Fig. 4). Because a thick gle central peak. The uplift structure observed 2002). The central ring structure is overlain by blanket of gravel limits acoustic penetration in in the sonobuoy data is clearly not in the cen- a layer interpreted as allochthonous fallback the western part of the lake, the western part ter of the crater, and we think that it corre- breccia characterized by seismic velocities of of the central ring cannot be observed in the sponds to the northeastern ¯ank of a central ϳ3000 m/s. This breccia is almost level over seismic data (Figs. 2 and 4C). The existence of a central ring is in good agreement with gravity data that showed a generally negative anomaly with a local positive anomaly ϳ3km northwest of the lake center (Alyunin and Da- bizha, 1980). The Ries crater in southern Germany has a diameter of 25 km and shows an outer ring forming the crater rim, an inner concentric ring, and a central peak. WuÈnnemann and Morgan (2004) showed that the results of nu- merical models of impact craters are depen- dent on the target rock. Impacts into crystal- line rock result in more pronounced structural elements, whereas impacts into sedimentary rock exhibit a relatively ¯at morphology. The Ries crater target rock consists mainly of crys- talline bedrock, and its sedimentary cover sup- ported the formation of a large megablock zone (WuÈnnemann and Morgan, 2004); this zone is less pronounced in the El'gygytgyn crater (Gurov and Koeberl, 2004). Even though the El'gygytgyn crater is smaller than the Ries crater, its morphology is more com- plex, according to Grieve and Pesonen (1992). Figure 4. Model of El'gygytgyn crater formation (modi®ed after Heiken et al., 1991) during This probably means that the crater morphol- impact (A), immediately after impact (B), and at present (C). Erosion seems to be much ogy is even more pronounced in siliceous vol- higher in northwestern part of crater, causing large areas to be ®lled with coarse sediments, and consequently lake is not located in center of crater. Permafrost soil and soli¯uction canic target rock than in crystalline target rock were observed in northwestern part of lake. Coarse sediments (e.g., gravel) are also ob- due to different petrological and rheological served in northwestern part of lake, limiting acoustic penetration in this area. behavior of the target rock.

GEOLOGY, March 2006 147 On an analysis of ®ve complex terrestrial work at the lake. Financial support by the German and chemistry of lakes: Berlin, Heidelberg, craters, Grieve and Pesonen (1992) concluded Ministry of Education and Research (BMBF grant Springer, 334 p. 03G0586B) is gratefully acknowledged. Minyuk, P., Nowaczyk, N., Glushkova, O.Y., Smir- that, within structures formed nov, V., Brigham-Grette, J., Melles, M., Cher- in crystalline bedrock, the width and height of REFERENCES CITED epanova, M., Lozhkin, A., Anderson, P., Ma- the crater structures roughly follow the rela- Alyunin, A.V., and Dabizha, A.I., 1980, Geophysi- trosova, T., Hubberten, H., Belaya, B.V., tionship cal characteristics of meteoritic crater Borkhodoyev, B.Y., Forman, S., Asikainen, El'gygytgyn, USSR, in Proceedings, Lunar C.A., Layer, P., Nolan, M., Prokein, P., Liston, and Planetary Science Conference XI: Hous- G., Huntzinger, R., Sharpton, V., and Niessen, d ϭ 0.15 D0.43, (1) a ton, Texas, Lunar and Planetary Institute, F., 2003, Paleoclimatic characteristics of Lake El'gygytgyn, NE RussiaÐMultidisciplinary p. 21±23. investigation (in Russian), in Malakhov, M.I., where d is the apparent crater depth (i.e., Belyi, V.F., 1994, Geology of Okhotsk-Chukchi vol- a ed., The process of post-depositional magne- canogenic belt: Magadan, Northeastern Inte- from the crater rim to the top of the breccia) tization and characteristic changes of the grated Scienti®c Institute, 76 p. and D is the diameter of the crater. Consid- Earth's magnetic ®eld and climate in the past: Belyi, V.F., 1998, Impactogenesis and volcanism of ϳ Magadan, Northeastern Integrated Scienti®c ering an apparent crater depth of 1 km, and the El'gygytgyn depression: Journal of Petrol- a crater diameter of 18 km, the El'gygytgyn Institute FEB RAS, p. 91±135. ogy, v. 6, p. 96±110. Niessen, F., Gebhardt, A.C., and Kopsch, C., 2005, impact formed a topographic relief that is Brigham-Grette, J., and Carter, L.D., 1992, Pliocene Geophysical survey: Reports on Polar and about twice as pronounced as expected from marine transgression of northern Alaska: Cir- Marine Research, v. 509, p. 121±129. equation 1. This could be due to differences cumarctic correlations and paleoclimate: Arc- Niessen, F., Gebhardt, A.C., Kopsch, C., and Wag- tic, v. 43, p. 74±89. between crystalline bedrock and siliceous vol- ner, B., 2006, Seismic investigation of the Dietz, R.S., and McHone, J.F., 1976, El'gygytgyn: El'gygytgyn (Central Chu- canic bedrock, or the fact that ®ve craters are Probably world's largest crater: Ge- kotka, NE Siberia): Preliminary results: Jour- insuf®cient to calculate reliable relationship ology, v. 4, p. 391±392. nal of Paleolimnology, Special Issue (in equations. Engelhardt, W.V., Arndt, J., Fecker, B., and Pankau, press). H.G., 1994, Ries impact crater, Germany: The uplift structure observed in the sono- Nolan, M., and Brigham-Grette, J., 2006, Basic hy- ϳ Thermal history of the breccia: Mete- drology, limnology, and meteorology of mod- buoy pro®les is 3 km wide (Fig. 3). Its oritics, v. 29, p. 463±464. ern Lake El'gygytgyn, Siberia: Journal of Pa- asymmetric position within the crater is ab- Glushkova, O. Y., and Smirnov, V., 2005, General leolimnology, Special Issue (in press). normal in comparison with other crater stud- geology and geography: Reports on Polar and Nolan, M., Liston, G., Prokein, P., Brigham-Grette, ies, and suggests that only parts of a wider Marine Research, v. 509, p. 14±18. J., Sharpton, B., and Huntzinger, R., 2002, Grieve, R.A.F., and Pesonen, L.J., 1992, The terres- Analysis of lake ice dynamics and morpholo- and more complex central uplift structure are trial impact cratering record: Tectonophysics, gy on Lake El'gygytgyn, NE Siberia, using visible in the seismic records. Interpolating a v. 216, p. 1±30. SAR and Landsat: Journal of Geophysical concentric ring structure from the visible Gurov, E.P., and Koeberl, C., 2004, Shocked rocks Research, v. 107, p. doi: 10. 1029/2001JD northeastern ¯ank of the uplift results in a ring and impact glasses from the El'gygytgyn im- 000934. Nowaczyk, N.R., Minyuk, P., Melles, M., Brigham- structure with a diameter of ϳ7±7.5 km (Fig. pact structure (Russia): Meteoritics and Plan- etary Science, v. 39, p. 1495±1508. Grette, J., Glushkova, O., Nolan, M., Lozhkin, 1). Data from lunar craters show that complex Gurov, E.P., Gurova, E.P., and Rakitskaya, R.B., A.V., Stetsenko, T.V., Andersen, P.M., and craters with diameters of 18 km, like the 1979a, and in shock meta- Forman, S.L., 2002, Magnetostratigraphic re- El'gygytgyn, commonly have central peak di- morphosed rocks of the El'gygytgyn crater in sults from impact crater Lake El'gygytgyn, ϳ Chukotka (in Russian): Akademiya Nauk northeastern Siberia: A 300 kyr long high- ameters of 2.5±4 km (Spudis, 1993, and ref- resolution terrestrial palaeoclimatic record SSSR Doklady, v. 248, p. 213±216. erences therein). However, lunar and terrestrial from the Arctic: Geophysical Journal Inter- Gurov, E.P., Valter, A.A., Gurova, E.P., and Kotlov- national, v. 150, p. 109±126. craters cannot be directly compared due to the skaya, F.I., 1979b, El'gygytgyn impact crater, different lunar and terrestrial gravity ®elds. Scholz, C.A., Karp, T., Brooks, K.M., Milkereit, B., Chukotka: Shock metamorphism of volcanic Amoako, P.Y.O., and Arko, J.A., 2002, Pro- rocks [abs.], in Proceedings, Lunar and nounced central uplift identi®ed in the Bos- SUMMARY Planetary Science Conference X: Houston, umtwi impact structure, Ghana, using multi- The El'gygytgyn crater, the only known ter- Texas, Lunar and Planetary Science Institute, channel seismic re¯ection data: Geology, p. 479±481. restrial crater with a siliceous target rock, v. 30, p. 939±942. Heiken, G., Vaniman, D., and French, B., 1991, The Spudis, P.D., 1993, The geology of multi-ring im- shows a distinct morphology with a central lunar sourcebook: Cambridge, Cambridge pact basins: The moon and other planets: New ring structure of ϳ7±7.5 km in diameter and University Press, 736 p. York, Cambridge University Press, 263 p. a minimum height of ϳ350 m. This ring struc- Karp, T., Milkereit, B., Janle, P., Danuor, S.K., Pohl, WuÈnnemann, K., and Morgan, J.V., 2004, Investi- ture is formed in the bedrock, which is brec- J., Berckhemer, H., and Scholz, C.A., 2002, gation of the Ries impact crater based upon Seismic investigation of the old and new geophysical data and numerical ciated to depths of at least a few hundreds of impact crater: preliminary results: Planetary modeling, in Proceedings, Lunar and Plane- meters. The ring structure topography is al- and Space Science, v. 50, p. 735±743. tary Science Conference XXXV: Houston, most buried by an overlaying fallback breccia Layer, P., 2000, 40Ar/39Ar age of the El'gygytgyn Texas, Lunar and Planetary Institute, p. 1229. unit. The crater lake sediments consist of an crater event, Chukotka, Russia: Meteoritics & Zelt, C.A., and Smith, R.B., 1992, Seismic travel- time inversion for 2-D crustal velocity struc- upper and a lower sedimentary unit character- Planetary Science, v. 35, p. 591±599. Layer, P., Newberry, R., Fujita, K., Parfenov, L., ture: Geophysical Journal International, ized by the low velocities of unconsolidated Trunilina, V., and Bakharev, A., 2001, Tecton- v. 108, p. 16±34. mud. ic setting of the plutonic belts of Yakutia, Manuscript received 26 September 2005 40 39 northeast Russia, based on Ar/ Ar geochro- Revised manuscript received 1 November 2005 ACKNOWLEDGMENTS nology and trace element geochemistry: Ge- Manuscript accepted 7 November 2005 We thank the 2000 and 2003 expedition members ology, v. 29, p. 167±170. for their excellent cooperation and support during Lerman, A., Imboden, D., and Gat, J., 1995, Physics Printed in USA

148 GEOLOGY, March 2006