*Manuscript Click here to view linked References 1 Supportive comment on: “Morphology and population of binary asteroid 2 impact craters” , by K. Miljković , G. S. Collins, S. Mannick and P. A. 3 Bland [Earth Planet. Sci. Lett. 363 (2013) 121 –132] – An updated 4 assessment 5 6 Martin Schmieder 1,2 , Mario Trieloff 3, Winfried H. Schwarz 3, Elmar Buchner 4 and Fred Jourdan 2 7 1School of Earth and Environment, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia 8 2Western Australian Argon Isotope Facility, Department of Applied Geology and JdL Centre, Curtin University, GPO Box 9 U1987, Perth, WA 6845, Australia 10 3Institut für Geowissenschaften, Universität Heidelberg, Im Neuenheimer Feld 234-236, D-69120 Heidelberg, Germany 11 4HNU-Neu-Ulm University, Edisonallee 5, D-89231 Neu-Ulm, Germany 12 13 In their recent paper, Miljković et al. (2013) assess the appar ent contradiction that the near-Earth asteroid population 14 consists of 15% binaries, while the terrestrial (and Martian) impact crater populations have only 2-4% of observable 15 doublet craters. The authors suggest that only a small fraction of sufficiently well separated binary asteroids yield 16 recognizable doublets. We generally agree with the conclusions by Miljković et al. (2013) and acknowledge the high 17 quality and relevance of the study. However, we would like to bring into focus additional geochronologic constraints 18 that are critical when evaluating terrestrial impact crater doublets. Miljković et al. (2013) appraised five potential 19 terrestrial doublets using the Earth Impact Database (EID; as of 2010). We hereby warn against the indiscriminate 20 usage of impact ages compiled in this database without an assessment based on solid isotopic and stratigraphic 21 constraints and comment on the geological, geochronological, and geochemical evidence for doublet impact craters 22 on Earth. 23 Geologic evidence 24 Firstly, the confirmation of macroscopic and microscopic shock effects in rocks and minerals recovered from 25 candidate terrestrial impact sites is a prerequisite to establish an impact origin of such structures (French and 26 Koeberl, 2010). Miljković et al. (2013) included the Crawford and Flaxman structures in South Australia in their 27 selection of ‘possible’ terrestrial crater doublets. These structures had been in troduced as impact structures in a 1999 28 conference abstract. Due to the lack of compelling evidence for impact, both structures were later classified as 29 ‘possible impact structures’ of uncertain age (Haines, 2005). Only proven impact structures should be part of 30 statistical considerations on the terrestrial impact rate. 31 32 Geochronological criteria 33 Secondly , we point out that some of the impact ages listed by Miljković et al. are outdated and incorrect. Precise and 34 accurate impact ages are crucial when it com es to ‘double impact’. All potential crater pairs of Miljković et al. 35 (2013) have apparent crater ages that comply with a theoretical double impact scenario; however, comparatively 36 large errors on the ages and, more importantly, a lack of unambiguous accuracy of some of the ages leave evidence 37 for double impact doubtful. The dating of terrestrial impact structures has undergone a dynamic evolution in recent 38 years, and more and more new and refined isotopic data are now available. Nevertheless, only a few of the ~188 39 known terrestrial impact structures have been dated with a satisfying uncertainty of ≤±2% by means of isotopic 40 and/or biostratigraphic methods (Jourdan et al., 2012). Isotopic approaches can yield results as precise as ~±0.5% in 41 their relative error on the age. Only synchronicity of two neighboring impact events within such narrow 42 uncertainties hardens the evidence for double impact. 43 44 Comments on the potential crater doublets listed in Miljković et al. (2013) 45 The ≥36 km Clearwater West and ~26 km Clearwater East impact structures in Canada are listed as a ‘very likely’ 46 doublet in Miljković et al. (2013). The maximum age of both structures is stratigraphically constrained by the 47 shocked Ordovician limestones that overlie the Archean basement (e.g., Grieve, 2006). The 290 ± 20 Ma age for 48 both impact structures widely cited in the literature and the EID seems to be adopted from early, poorly robust, K-Ar 49 ages for Clearwater West published in a 1960s Geologic Survey of Canada report. Our group obtained two 40 Ar/ 39 Ar 50 plateau ages with a mean of 286 ± 2 Ma for Clearwater West, very similar to the age reported in Bottomley et al. 51 (1990). Age data for Clearwater East are still somewhat ambiguous. A mineral isochron Rb-Sr age for melt rocks 52 from Clearwater East yielded an age of 287 ± 26 Ma (Reimold et al., 1981), coeval within error with the ages for 53 Clearwater West. However, the Rb-Sr technique has frequently failed to provide accurate impact ages (e.g., Mark et 54 al., 2013). 40 Ar/ 39 Ar dating of melt rocks from Clearwater East produced a set of significantly older Ordovician 55 apparent ages around ~460-470 Ma (Bottomley et al., 1990). Similar ages were obtained by our group, suggesting 56 that Clearwater West and East are probably not a doublet. Currently, the idea of synchronicity of the two impact 57 events entirely relies on a single Rb-Sr age for Clearwater East and, probably, a general ‘agreement’ since the 1960s 58 based on the expectation that the two impact structures, closely spaced, must represent a doublet. 59 The ~25 km Kamensk and ~3 km Gusev impact structures, also a ‘very likely’ crater doublet in Miljković et al. 60 (2013), have been considered a doublet since the 1970s. Both craters are filled with sediments of the Globukaya 61 Formation of postulated Cretaceous-Paleogene boundary age, an impact-derived marine resurge breccia that overlies 62 the uppermost Cretaceous and is topped by Paleogene sands (Movshovich et al., 1991). Notwithstanding with the 63 previous biostratigraphic age estimates, 40 Ar/ 39 Ar dating of fresh impact glass from Kamensk by Izett et al. (1994) 64 yielded a fairly robust Eocene age of 50.36 ± 0.33 Ma (recalculated; Jourdan et al., 2012), contradicting a ~65 Ma 65 impact age (as given in Miljković et al. 2013). This also constrains a minimum age for the Globukaya Formation and 66 the smaller Gusev crater, and suggests sedimentary reworking processes in the Eocene. Despite some uncertainties 67 regarding the exact timing of the Kamensk-Gusev impact, stratigraphic correlation and geographic evidence support 68 a likely double impact scenario. 69 The ~24 km Ries and ~3.8 km Steinheim impact craters in Southern Germany, associated with the Central European 70 tektite strewn field, have been generally regarded as a typical crater doublet since the 1960s (Stöffler et al. 2002). 71 Miljković et al. (2013) conservatively labelled them a ‘likely’ crater doublet. Multiple dating campaigns have 72 established a robust Miocene 40 Ar/ 39 Ar age of 14.83 ± 0.15 Ma (Di Vincenzo and Skála, 2009; recalculated by 73 Jourdan et al., 2012) for the Ries crater. In contrast to a well-established isotopic age data set for the Ries, the nearby 74 Steinheim Basin has failed to yield any reliable isotopic dating results so far ( 40 Ar/ 39 Ar and (U-Th)/He dating by our 75 group; Buchner et al., 2011). No coherent Steinheim impact ejecta are preserved outside the crater that could be 76 correlated with the Ries ejecta blanket. As a result, the assumed synchronicity of the Ries and Steinheim impacts 77 completely relies on the biostratigraphy of the Miocene post-impact crater lake sediments at both sites. However, 78 assuming that both crater lakes formed shortly after impact, the oldest known freshwater deposits of the Ries contain 79 fossil mammals of the Neogene mammal zone MN6 (Langhian), whereas no evidence for such fossils could be 80 observed at Steinheim. Instead, the lowermost Steinheim lake deposits are representative of the younger mammal 81 zone MN7 (Serravallian; Heizmann and Hesse, 1995; Heizmann and Reiff, 2002). Therefore, one must consider that 82 the Ries crater might be slightly older than the Steinheim Basin. Until such issues are resolved, it is unsafe to treat 83 these craters as a ‘proven ’ doublet based on their proximity alone. 84 Miljković et al. (2013) also listed the ~12 km Serra da Cangalha and ~4.5 km Riachão impact structures as a 85 ‘possible’ doublet. For both impact structures precise and accurate ages are currently lacking, and only a 86 stratigraphic maximum age of ≤250 Ma can be assigned for the Serra da Cangalha impact that affected the ~250- 87 260 Ma Permian sandstones of the Pedra de Fogo Formation in the Parnaíba Basin of Brazil (Kenkmann et al., 88 2011). This formation also forms the youngest impact-deformed target rocks at Riachão (Maziviero et al., 2012). 89 Due to deep erosion of both impact structures, no post-impact deposits are preserved that could constrain a 90 minimum impact age; nor are melt lithologies known from either of these structures to provide material for isotopic 91 dating. Based on the poor stratigraphic age constraints, evidence for a crater doublet is far from proven. 92 93 Additional candidate doublets? 94 A recent addition to the list of potential impact doublets on Earth are the Paleozoic marine Lockne and Målingen 95 impact structures in central Sweden. The ~1 km Målingen structure, ~16 km southwest of the ≥7.5 km Lockne 96 impact structure, was confirmed as of impact origin after the release of the paper by Miljković et al. (2013). The 97 Målingen impact affected Ordovician orthoceratid limestones and produced a sequence of impactites and a resurge 98 breccia overlain by the post-impact Dalby Limestone – a pre- to post-impact sequence essentially identical with that 99 at Lockne (Alwmark et al., 2013).