Published onlineDownloaded September from gsabulletin.gsapubs.org 25, 2009; doi:10.1130/B26517.1 on 8 October 2009 The Sudbury impact layer in the Paleoproterozoic iron ranges of northern Michigan, USA W.F. Cannon1,†, K.J. Schulz1, J. Wright Horton Jr.1, and David A. Kring2 1U.S. Geological Survey, Reston, Virginia 20192, USA 2Lunar and Planetary Institute, 3600 Bay Area Boulevard, Houston, Texas 77058, USA ABSTRACT the character of sediments across the region. crater margin. The fi rst documentation of the The impact layer marks the end of a major existence of such ejecta was by Addison et al. A layer of breccia that contains fragments period of banded iron formation deposition (2005), who described occurrences of ejecta- of impact ejecta has been found at 10 sites in that was succeeded by deposition of fi ne clas- bearing breccias in the Gunfl int iron range in the Paleoproterozoic iron ranges of north- tic rocks, commonly black shales. The impact Ontario and Mesabi iron range in Minnesota. In ern Michigan, in the Lake Superior region may have produced regional, if not global, the past two years, following the descriptions of of the United States. Radiometric age con- changes in the environment that resulted in Addison et al., we have located the breccia layer straints from events predating and postdat- this widespread synchronous change in sedi- at 10 sites in the Upper Peninsula of Michigan; ing depo si tion of the breccia are ca. 1875 Ma mentation style. the layer contains ejecta particles, including and 1830 Ma. The major bolide impact that quartz grains with relict shock-induced planar occurred at 1850 Ma at Sudbury, Ontario, INTRODUCTION deformation features (PDFs), together with 500–700 km east of these sites, is the likely spheres and shards of altered devitrifi ed glass causative event. The Michigan sites de- The major impact event at Sudbury, Ontario, and accretionary lapilli. We term this the “Sud- scribed here, along with previously described has been studied in great detail for the past 40 bury impact layer” and interpret the breccias to sites in Minnesota and Ontario, defi ne an years since its existence was fi rst proposed by have formed rapidly, in part within hours, after extensive ejecta-bearing deposit through- Dietz (1964). The time of impact is precisely the impact (Cannon et al., 2006; Kring et al., out the Paleoproterozoic iron ranges of the dated at 1850 ± 1 Ma (Krogh et al., 1984; Davis , 2006; Cannon and Addison, 2007). The impact Lake Superior region that we refer to as the 2008), the age of impact-generated melts. Brec- layer was also described by Pufahl et al. (2007) Sudbury impact layer. The layer at the sites cias and related rocks of the Onaping Forma- from two drill holes very near some of the lo- in Michigan exhibits a range of thicknesses, tion near Sudbury were fi rst interpreted as calities described here. Many of the impact layer lithologic characters, and sedimentary set- impact-related rocks soon after Dietz’s pro- localities had been observed, mapped, analyzed, tings. The diversity of rock types and internal posal (French, 1967, 1970; Peredery, 1972) and and described previously by numerous geolo- stratigraphic details of the layer imply that are now widely accepted to be, at least in part, gists, in some cases as much as 60 years ago. several different processes of transport and crater-fi lling material resulting from direct fall- The previous interpretations ascribed volcanic deposition are represented, but the detailed back of ejecta, slumping of initial crater walls, or submarine slumping processes as the cause of investigations needed to document them are and resurge of ocean water into the new crater the breccias, failing to recognize the widespread incomplete. Many of the sites had been de- (for recent summaries of various aspects of this but sparse shock-induced PDFs in quartz grains. scribed and interpreted previously as prod- large body of work, see Deutsch et al., 1995; The ejecta-bearing rocks described here, along ucts of common terrestrial processes, but Riller, 2005; Spray et al., 2004; Grieve and with previously documented sites (Addison et al., the presence of relict shock-induced planar Therriault, 2000; Naldrett, 2003; Grieve, 2006). 2005; Pufahl et al., 2007; Jirsa et al., 2008), de- deformation features in quartz indicates that There is growing evidence that igneous rocks, fi ne a regionally extensive fi eld of ejecta from the breccia layer is in fact the product of an including the Sudbury igneous complex, are the Sudbury impact event that can reasonably extra terrestrial impact. At most localities, this largely the impact-generated melt sheet (Faggart be inferred to have originally covered roughly layer also contains relict fragments of altered et al., 1985; Grieve, 1994; Keays and Lightfoot, 100,000 km2. The ejecta fi eld lies between 500 devitrifi ed glass and/or accretionary lapilli. 1999; Therriault et al., 2002; Naldrett, 2003; and 900 km from Sudbury within an arc of 30° One immediate use of the impact layer is as Mungall et al., 2004; Zieg and Marsh, 2005; radially outward from Sudbury. The currently an ultraprecise time line that ties together Grieve, 2006). The crater produced by this im- known ejecta fi eld is likely to be the only repre- the well-known stratigraphic sequences of pact event, now largely destroyed by erosion and sentative of proximal to distal ejecta that is pre- the various geographically separated iron strongly modifi ed by younger tectonic events, served, there being no other rocks of suitable age ranges, the correlation of which has re- has been variously estimated to have a diameter nearer than ~1200 km from Sudbury (Labrador mained controversial for many decades. The between 150 km and 260 km (see summaries in Trough). This ejecta fi eld is a signifi cant addi- Sudbury impact layer most commonly lies at Abramov and Kring, 2004; Grieve et al., 2008). tion to the still small inventory of known ejecta a horizon that records a signifi cant change in An important missing component of the Sud- blankets from giant impacts on Earth. The study bury story has been information on the character of these newly discovered sites of the Sudbury †E-mail: [email protected] and distribution of ejecta deposited beyond the impact layer is in its infancy, and each site merits GSA Bulletin; January/February 2010; v. 122; no. 1/2; p. 50–75; doi: 10.1130/B26517.1; 17 fi gures; 2 tables. 50 For permission to copy, contact [email protected] © 2009 Geological Society of America Downloaded from gsabulletin.gsapubs.org on 8 October 2009 The Sudbury impact layer in northern Michigan, USA considerable additional detailed description and Crystal Falls), the Sudbury impact layer is thick of planar microstructures (PMs), which are in- interpretation to decipher the processes repre- and extensive enough to constitute a mappable terpreted to be highly annealed relicts of planar sented in transport and deposition of the ejecta- unit and, although not recognized as impact- deformation features (PDFs) characteristic of a bearing units and the implications for impact related rocks at the time of mapping, appears shock-metamorphic origin. Such PDFs indicate processes. Our studies to date have emphasized as breccia beds on U.S. Geological Survey that the quartz grains experienced the extreme the mapping of the layer through the iron ranges (USGS) 1:24,000 scale geologic maps (Puffett, pressures and strain rates diagnostic of a hyper- of Michigan to determine its geographic distribu- 1974; Clark et al., 1975; James et al., 1968). velocity impact (Grieve et al., 1996; French, tion and stratigraphic position. In this paper, we 1998). We have identifi ed single sets or multiple present the fi rst descriptions of 10 sites in Michi- GENERAL CHARACTER OF THE intersecting sets of these PMs interpreted to be gan at which we have identifi ed the Sudbury im- SUDBURY IMPACT LAYER annealed PDFs in quartz grains from fi ve of the pact layer by examination of outcrops and drill sites described here. These features originally cores. We also provide general descriptions of the The Sudbury impact layer in Michigan is a bed formed as thin planar lamellae of shock-induced physical and geochemical character of the layer of breccia and related rocks. Because sites vary glass within the quartz grains. With time, the based on petrographic microscope examination from a single drill hole or outcrop to clusters of glass devitrifi ed and recrystallized to quartz. and chemical analyses of impact layer materials. drill holes and outcrops, the amount of informa- During that process, the annealed PDFs became A diversity of ejecta-bearing rock types suggests tion is quite variable between sites. Table 1 pro- “decorated” with small fl uid inclusions, so that that multiple processes of transport and deposi- vides some general information about each site, they now appear as thin, regularly spaced planar tion are probably represented across the region, and Figure 2 shows the stratigraphic setting of the zones of inclusions (decorated PDFs) within the but our studies and understanding of these rocks impact layer at all known localities in the Lake host quartz grains (French, 1998). Examples of are too preliminary to attempt anything but the Superior region. The impact layer in Michigan these features from the Sudbury impact layer broadest genetic interpretations at this time. appears to have been deposited in a submarine to at fi ve localities in Michigan are shown in Fig- peritidal setting. At localities in the Baraga Basin ure 3. Some of the shocked quartz grains have GEOGRAPHIC DISTRIBUTION OF and Dead River Basin (Fig. 1), the Sudbury layer subrounded shapes (Figs.