Newly identified “Tunnunik” impact structure, Prince Albert Peninsula, northwestern Victoria Island, Arctic Canada


Corresponding author. E-mail:


Regional geological mapping of the glaciated surface of northwestern Victoria Island in the western Canadian Arctic revealed an anomalous structure in otherwise flat-lying Neoproterozoic and lower Paleozoic carbonate rocks, located south of Richard Collinson Inlet. The feature is roughly circular in plan view, approximately 25 km in diameter, and characterized by quaquaversal dips of approximately 45°, decreasing laterally. The core of the feature also exhibits local vertical dips, low-angle reverse faults, and drag folds. Although brecciation was not observed, shatter cones are pervasive in all lithologies in the central area, including 723 Ma old dikes that penetrate Neoproterozoic limestones. Their abundance decreases distally, and none was observed in surrounding, horizontally bedded strata. This circular structure is interpreted as a deeply eroded meteorite impact crater of the complex type, and the dipping strata as the remnants of the central uplift. The variation in orientation and shape of shatter cones point to variably oriented stresses with the passage of the shock wave, possibly related to the presence of pore water in the target strata as well as rock type and lithological heterogeneities, especially bed thickness. Timing of impact is poorly constrained. The youngest rocks affected are Late Ordovician (approximately 450 Ma) and the impact structure is mantled by undisturbed postglacial sediments. Regional, hydrothermal dolomitization of the Ordovician limestones, possibly in the Late Devonian (approximately 360 Ma), took place before the impact, and widespread WSW–ENE-trending normal faults of probable Early Cretaceous age (approximately 130 Ma) apparently cross-cut the impact structure.


Meteorite impact craters record collisional events on rocky planetary and lunar surfaces. Examples on Earth range in age from to late Archean to Recent. Impact structures are preserved in a variety of ways on land, some of the youngest ones as pristine craters, whereas others are heavily eroded; some have been detected buried in the subsurface, entombed by younger strata. Examples on Earth are vital analogs of extraterrestrial impact craters, reveal the bombardment history of the planet, can be related to abrupt changes in the biosphere, and approximately 25% of impact craters are associated with mineral or hydrocarbon deposits (Grieve 1991).

Recognizing craters in the rock record is not always straightforward. Deeply eroded impact structures in remote areas may only be noticed by ground surveys. Circular topographic and geological patterns, and shock metamorphic features are distinctive indications of an impact origin (French and Koeberl 2010). During the course of fieldwork on northwestern Victoria Island, Arctic Canada, in the summers of 2009–2011 as part of a Geological Survey of Canada bedrock mapping program, an area of anomalously dipping strata roughly 25 km in diameter was observed on Prince Albert Peninsula just south of Richard Collinson Inlet. The traditional Inuvialuit name for this inlet is “Tunnunik.” This feature is identified as a meteorite impact structure on the basis of abundant shatter cones, circular topographic expression, concentric geological map pattern, and steeply dipping strata in its core. The purpose of this article is to present a preliminary description of this impact structure, the thirtieth impact structure definitively identified in Canada (see Grieve 2006;

General Geology


Victoria Island is located in the western Canadian Arctic Islands (Fig. 1). It is the ninth largest island in the world, slightly smaller than Great Britain. It exhibits two physiographic regions. The rugged upland terrain of the Shaler Mountains runs diagonally across the island and is composed of broadly folded Neoproterozoic sedimentary and igneous rocks of the Shaler Supergroup. To the northwest and southeast are lowlands underlain by nearly flat-lying Cambrian to Devonian sedimentary rocks and covered by an extensive blanket of Quaternary postglacial sediments (Fig. 2).

Figure 1.

Location (star) of newly identified impact structure, Prince Alberta Peninsula, northwestern Victoria Island, Northwest Territories, Arctic Canada. Areas underlain by strata younger than Ordovician are left white. Outlined is the area shown in Fig 3.

Figure 2.

Stratigraphy, northwestern Victoria Island. Stratigraphic names remain informal in the lower part of the section.

The general outcrop pattern of northwestern Victoria Island is portrayed in Geological Survey of Canada map 1135A by Thorsteinsson and Tozer (1962). The regional Arctic map of Okulitch (1991) incorporated later airphoto interpretation. Oil industry assessment reports by Ehman and Wise (1971) and PetroCanada Exploration Inc. (1978) have incomplete geological maps of parts of the area. New aeromagnetic surveys (Kiss and Oneschuk 2010) and bedrock mapping (Bédard et al. 2012) over the western part indicate an extensive array of WSW–ENE-trending normal faults (Fig. 3). These faults are most obvious near the contact between Paleozoic and Neoproterozoic strata and they have displacements from hundreds of meters, close to the Shaler–Paleozoic contact, to tens of meters within the Paleozoic succession. The age of faulting is unknown, but the faults clearly cut Devonian strata. Similar normal faults on Banks Island have a parallel trend and cut Cretaceous strata and are inferred by their age and orientation to be associated with the opening of the Canada Basin of the Arctic Ocean (Miall 1979). At the small scale, none of the units exhibits fracture sets due to regional tectonism.

Figure 3.

Bedrock geological map showing distribution of Cambrian to Devonian strata, northwestern Victoria Island. Strata generally dip to the NW, but are cut by an extensive set of ENE–WSW-trending normal faults. The impact structure appears as a circular feature with a central core of Neoproterozoic strata of the Shaler Supergroup, just south of Richard Collinson Inlet.

Both the Ehman and Wise (1971) and the PetroCanada Exploration Inc. (1978) reports noted a small area with a roughly circular map pattern, formed by a dome of Neoproterozoic strata, just south of Richard Collinson Inlet. Thorsteinsson and Tozer (1962) had earlier visited the tilted strata at the northern part of the feature, but did not offer an interpretation. These strata are exposed mainly in an un-named river gorge. The Alminex Ltd (1969) report informally called it “Glaukos River.” Our mapping in 2010 shows the distribution of Neoproterozoic and lower Paleozoic units in more detail, as well as normal faults which cross-cut them.

Gravity data were collected in 1975 on land and included the area on both sides of Collinson Inlet. These data have a relative equal station spacing of 15 km. In 1977, gravity data were added in the intervening marine area of the Inlet with a station spacing of 6–7 km. Both surveys were conducted by the Government of Canada and are of reconnaissance nature. No pattern indicating an underlying impact is apparent on Bouguer anomaly or any other processed version of these gravity data, which are publically available at the Geoscience Data Repository of Natural Resources Canada. There are no aeromagnetic data in the immediate vicinity of the impact structure.


Stratigraphic terminology for northwestern Victoria Island has evolved with continued study and correlation across the Canadian Arctic Islands. The Precambrian Shaler Group of Thorsteinsson and Tozer (1962) now comprises the Neoproterozoic-aged Shaler Supergroup, which consists, in ascending stratigraphic order, of the Rae Group, the Reynolds Point Group, and the Minto Inlet, Wynniatt, Kilian, Kuujjua, and Natkusiak formations (Rainbird 1992; Rainbird et al. 1994). These strata range in age from roughly 1000 to 720 Ma (Heaman et al. 1992; Rainbird et al. 1997). The informally named “Upper member” of the Wynniatt Formation is exposed in the core of the circular feature.

Thorsteinsson and Tozer (1962) recognized four Paleozoic units in the region: (1) Cambrian sandstone, siltstone, shale, and dolostone at the base of the succession; (2) Ordovician and Silurian dolostone; (3) Silurian carbonate and shale; and (4) the Devonian Blue Fiord Formation. Plauchut and Jutard (1967) divided the Cambrian section into a lower sandy unit, a middle dolostone unit, and an upper argillaceous dolostone unit. Fossils indicate an early Cambrian age for the lower unit and possibly the middle unit (Fritz 1967, 1971). Chernoff (1974) recognized Cambrian sandstone overlain by dolostone, the latter broken into an upper Cambrian? lower “cyclic and rhythmic unit,” and a Lower Ordovician “cherty unit.” Plauchut and Jutard (1967) assigned Ordovician and Silurian dolostone to an informal “Franklin Straits” formation. Ehman and Wise (1971) assigned the Upper Ordovician limestones to the Allen Bay Formation, whereas these were referred to as “Mount Kindle equivalent” in Chernoff (1974). Upper Silurian and Lower Devonian strata were assigned to the Read Bay Formation by Plauchut and Jutard (1967).

One hydrocarbon exploration well has been drilled on northwestern Victoria Island. It started in Devonian Blue Fiord Formation and intersected fossiliferous dolostone assigned to the Ordovician Thumb Mountain Formation, and carbonates and evaporites assigned to the Bay Fiord Formation (Batten 1975; Dewing and Embry 2007).

On the basis of nine stratigraphic sections and an additional 213 field locations, we recognize ten map units in the region (Fig. 2).

Neoproterozoic/Upper Member, Wynniatt Formation

Rhythmically bedded and normally graded, quartz-sandy grainstone exhibits meter-scale alternations of stromatolitic dolostone and cross-laminated intraclastic grainstone (intraformational conglomerate). Local herringbone cross-bedded quartz arenite, microbially laminated lime mudstone, and chert occur, but are not exposed in the impact structure. This member is approximately 300 m thick.

Neoproterozoic/Franklin Igneous Rocks

Dark-gray to black weathering diabase dykes are 3 m wide in the impact structure. These may contain metasedimentary xenoliths; they exhibit finely crystalline chilled margin, which grades to medium-crystalline diabase in the center, with sector-zoned clinopyroxene phenocrysts. The age of examples elsewhere in the region is 723 Ma (Heaman et al. 1992).

Cambrian/Clastic Unit

Red-brown to orange-weathering fine- to coarse-grained quartz arenite and mudstone comprise the base of the Paleozoic succession. Sedimentary structures are planar lamination, wave and current ripples, and cross-stratified beds 10 cm to 2 m thick. Reactivation surfaces and foresets with rounded tops indicate an influence by tidal currents. Mudstones contain early Cambrian (series 2) trilobites and trace fossils. Distribution and thickness are variable, with thickness ranging from 0 to 90 m.

Cambrian/Tan Dolostone Unit

Light-brown dolomudstone to dolograinstone contains locally well-developed thrombolite mounds and meter-scale cross-stratification. Although no fossils were recovered from this unit, the lower contact is gradational with the mudstones that contain early Cambrian (series 2) trilobites. Thickness is 30–45 m.

Cambrian/Stripy Unit

Thin- to medium-bedded red mudstone and shale interbedded with green, gray, and reddish dolomudstone give outcrops a distinctive stripy appearance. Sedimentary structures include mudcracks, wave ripples, small stromatolites, intraclasts, burrows on bed soles, microkarsted exposure surfaces, and planar microbial lamination. Mudstones contain middle Cambrian (series 3) trilobites. The lower contact is covered in most places; thickness ranges between 15 and 95 m.

Cambro-Ordovican/Victoria Island Formation (Informal)

Light-gray to almost white-weathering, fine- to coarsely crystalline, fabric-destructive dolostone locally preserves primary structures including horizontal bedding, cross-laminated oolitic grainstone, thrombolite and stromatolite bioherms, microbial lamination, and intraformational conglomerate. Fossils are rare, but include silicified gastropods. Silicification is widespread in the upper two thirds of the unit. Chert occurs as prominent white-weathering beds and nodules, 5–60 cm thick, and is composed of microcrystalline quartz or as silicified stromatolites. Some vugs contain coarsely crystalline quartz. The lower contact is gradational with the (informally named) Stripy unit, with the contact pinned at the highest shale interbed. The upper contact is sharp and inferred to be disconformable. Early Ordovician conodonts occur directly below this contact. The thickness is 550 m on the south shore of Minto Inlet.

Ordovician–Silurian/Undivided Thumb Mountain – Allen Bay Formations

Fossiliferous dolostone consisting of finely crystalline, burrow-mottled, nodular to thinly bedded dolomudstone, dolowackestone, and dolopackstone. Fossils include crinoids, solitary rugose and colonial tabulate corals, aulacerid calcisponges, gastropods, and nautiloids. These fossils and conodonts indicate a Late Ordovician age. The thickness is estimated at 250 m. Overlain by poorly exposed carbonates of probable Silurian age and tentatively assigned to the Allen Bay Formation, consisting of thick-bedded, burrow-mottled, fine- to medium-crystalline dolomudstone. Outcrop is sparse and discontinuous, but well logs indicate that this unit is 650 m thick on northwestern Victoria Island. Sparse outcrop makes mapping the contact between the two formations difficult in most areas.

Silurian/Unnamed Shale

This unit is not exposed and is inferred from well logs and topographic expression. It consists of dark-brown calcareous and bituminous shale and interbedded light-brown lime mudstone to wackestone.

Devonian/Blue Fiord Formation

This consists of light-gray to greenish-weathering limestone with abundant fossils including brachiopods, corals, and crinoids.

Devonian/Kitson Formation

This unit is poorly exposed black argillaceous limestone and shale.

Total thickness of strata involved within the preserved diameter of the impact structure is about 700 m. The thickness of stratigraphic units preserved on Victoria Island (including Devonian units) not known to be affected by the impact is about 2000 m. As much as 2900 m of Middle and Upper Devonian strata is preserved on Banks Island to the west, and estimates based on thermal maturity indicators indicate that there were likely several kilometers of strata once present above the youngest preserved strata.

Impact structure


Paleozoic strata on northwestern Victoria Island generally dip a few degrees to the northwest. Thus, the circular pattern on the geological map (Fig. 3) is a strikingly anomalous feature roughly 25 km in diameter, centered on 72° 28′ N, 113° 58′ W. The digitally rendered elevation model indicates a circular topographic feature of low relief, rising some 40 m above the surrounding flat-lying tundra (Fig. 4). (The vertical and horizontal accuracy and elevation determination method of each of the source data tiles that comprise the DEM are available at All or part of the following sixteen 50 K tiles: 88-A-01 to -16 were used in this DEM work.) The circular pattern is accentuated by a series of concentric ridges that curve around it. These ridges are formed of resistant dolostone beds of the Victoria Island formation.

Figure 4.

Digital elevation model (DEM) illustrating the circular nature of the preserved terrain expression of the impact structure. The DEM was sourced from 50 K scale Canadian digital elevation data (CDED) available at, resampled to a 50 m cell size, and rendered with sun-illumination azimuth of 0° and a vertical exaggeration factor of 3. Map projection used is UTM zone 11, NAD 83. Inset shows location of Fig. 5.

The core of the feature consists of Neoproterozoic rocks belonging to the Shaler Supergroup and is flanked by Cambrian and Ordovician strata (Fig. 5). The Neoproterozoic strata consist of thin- to medium-bedded, plane-laminated lime mudstone and fine-grained grainstone (Fig. 6A) and locally cross-laminated oolitic grainstone, assigned to the Upper member of the Wynniatt Formation (location 90). These strata are cut by a ferro-diabase dyke trending 075°/335°, with a finely crystalline (chilled) margin and medium-crystalline center containing sector-zoned clinopyroxene phenocrysts and randomly oriented plagioclase (location 91; Fig. 7F). The geochemical signature of the dyke (Fig. 8) is identical to other igneous rocks belonging to the Franklin igneous event.

Figure 5.

Detailed geology of the impact structure based on airphotograph A17345-76. Sites of field observations referred to in the text are denoted by circles and numbers.

Figure 6.

Neoproterozoic and Cambro-Ordovician lithologies. A) Medium-bedded lime mudstone of the Upper member of the Wynniatt Formation. Location 90. B) Thin-bedded dolomudstone and shale of the Stripy unit. Location 78. C) Buff-weathering dolomudstone with interbedded red and green mudstone of the Stripy unit. Dolostone thin beds are broken along shatter cone surfaces (appearing as oblique feathery striae). Location 78. D) Weakly mottled dolomudstone of the Victoria Island formation. Location 82. Pocket knife is 9 cm long.

Figure 7.

Shatter cones. A) Upward- and subordinate downward-pointing (lower right) cones. Victoria Island formation, location 80. B) Upward-pointing cones. Victoria Island formation, location 80. C) Downward-splaying meandering surfaces. Tan dolostone unit, location 78. (B). Upward-pointing cones in the Victoria Island formation. Location 80. D) Crudely developed interfering cones. Thumb Mountain Formation, location 87. E) Upward-pointing interfering cones. Stripy unit, location 78. F) Downward-splaying meandering surface. Upper member of the Wynniatt Formation, location 90. G) Upward-pointing interfering cones. Upper member of the Wynniatt Formation, location 90. H) Segment of poorly developed cone surface in diabase sill. Franklin igneous province, location 91. Pocket knife (A, C–G) is 9 cm long; hammer (B) is 33 cm long; finger (H) is 1.8 cm wide.

Figure 8.

Comparison of the geochemistry of the dyke found in the center of the impact structure (solid line; location 91; Fig. 5) to other Franklin igneous rocks on Victoria Island and on the mainland (dashed lines). Data from Coronation sills from Shelnutt et al. (2004). All data normalized to N-MORB of Sun and McDonough (1989).

Lower Cambrian siliciclastic strata were not observed at the localities visited (including on the traverse on the north side; Fig. 5), but are inferred from both the recessive interval between the Wynniatt Formation and the base of the Stripy unit, and from mention of conglomeratic sandstone in this recessive interval by Alminex Ltd (1969) and Ehman and Wise (1971). We suspect that slumped glacial sediment and tundra vegetation covered their outcrop in 2010.

Exposures at the upstream end of the gorge (locations 78 and 79; Figs. 5 and 9F) consist of thinly interbedded; red, gray, and green shale; and buff-weathering, thin- to medium-bedded, finely crystalline dolostone (Figs. 6B and 6C), locally with burrows and ripple cross-lamination, belonging to the Stripy unit. This unit is overlain by a thick unit of medium- to thick-bedded, medium-crystalline dolomudstone and dolograinstone (locations 80–83; Figs. 5 and 9C–E). The dolostone is massive, but locally, burrows, plane- and cross-lamination, and oolite can be distinguished (Fig. 6D). This unit is assigned to the Victoria Island formation. It is overlain by thin- to medium-bedded dolowackestone with crinoids, gastropods, stromatoporoids, and silicified solitary rugose corals and tabulate corals belonging to Catenipora and Favosites. These strata are assigned to the Thumb Mountain Formation.

Figure 9.

Field photographs of central uplift strata. A) Oblique aerial view (looking northwestward) of gorge exposing steeply dipping Neoproterozoic and Cambro-Ordovician strata. Strata to left are highly faulted. Richard Collinson Inlet is in the distance. Locations 78–80 are at the right end. B) Oblique aerial view (looking southeastward) showing steeply dipping Cambrian and Neoproterozoic strata. Location 80 is in this segment. C) View (looking westward) of nearly vertically dipping Cambrian strata faulted against more gently dipping Neoproterozoic strata. Locations 81–83. D) Thrust fault placing northwest-dipping beds over steeply north-dipping beds. Stripy unit, location 78. E) South-dipping thrust truncating folded beds. Victoria Island formation, location 80. F) North-dipping strata. Stripy unit, location 78.

Bedding dips quaquaversally from the core and becomes shallower with distance. Dips in the central portion range from 57 to 66° (Figs. 9A and 9B) and locally, bedding can be nearly vertical (Fig. 9C) due to radial normal faults that separate “rafts” of strata with variable attitudes. On the flanks there are low-angle reverse faults (Fig. 9D) and drag folds (Fig. 9E). There are no distinct fracture arrays or breccias.

Shatter Cones

Shatter cones are pervasive in all units in the central area, whereas toward the margin, their abundance decreases and they become more sporadic and isolated. In the thicker bedded dolomites of the Victoria Island and Thumb Mountain formations, they reach approximately 50 cm in vertical extent (Figs. 7A, 7B, and 7D), while those in thinner bedded strata are several to approximately 20 cm depending on bedding thickness (Figs. 6C, 7C, 7D, and 7F). We were unable to determine if they are present in interbedded shales because this lithology is recessively weathered.

Most of the shatter cones have stratigraphically upward-pointing apices, but downward-facing apices are common (Fig. 7G). This means that their long axes are oblique due to subsequent tilting of the host strata. In thicker beds, shatter cones tend to be nested and interfering (Fig. 7D) and their apices typically begin at different levels in the same bed (Figs. 7C, 7E and 7F). In some thinner beds, the shatter cones are strongly oblique to bedding and locally the interference angles between intersecting surfaces reach approximately 60°. In many thin beds, a conical shape is not well developed, and meandering surfaces have formed instead (Figs. 6C and 7E). Shatter cones in the diabase dyke are poorly developed, and consist of rare vertically striated segments about 5 cm wide and a few centimeters “high,” lacking distinct conical morphology (Fig. 7H).

In fine-grained host strata, shatter cone surfaces have a feathery appearance imparted by closely spaced grooves and ridges (“striae” or “striations” of some studies) that range from slightly smooth to sharp-edged and cuspate. They are mainly straight, but some are slightly curved, and exhibit a relief of about 0.1 mm up to some 3 mm. In coarser dolostones of the Victoria Island and Thumb Mountain formations, they lack this morphological detail. Shatter cone surfaces are arranged in predominantly splaying, bundle-like arrays (“horsetail structures” of some studies) up to 10–15 mm wide, with subordinate ones that are more or less vertical (Figs. 10A–D). The angle where the ridges and grooves converge ranges from about 15 to 25°, and while upward-pointing orientations are most common, downward-pointing ones are also present (Figs. 10A and 10B).

Figure 10.

Fresh shatter cone surfaces. A) Obliquely vertically oriented surface developed in lime mudstone (up is toward top of photograph). Surface consists of upward-splaying grooves and ridges with slightly smoothed edges and about 1 mm of relief. Linear feature at right is later (cross-cutting) calcite-filled vein. Upper member of Wynniatt Formation, location 90. B) Detail of A (but to left of its field of view). C) Obliquely oriented surface developed in dolomitized lime mudstone (up and down orientation unknown). Surface consists of splaying sharp-edged grooves and ridges with up to 3 mm of relief, and subordinate small curved splays oriented almost at right angles to the main array. Stripy unit, location 78. D) Detail of lower right of C, showing concavo-convex splays (and molds of splays [right side]).

An unusual feature present in some shatter cones penetrating dolomudstones of the Stripy unit are strongly concavo-convex surfaces ranging in size from about 1 to 5 mm (Figs. 10C and 10D). These are revealed when the host rock is split open along the shatter cone surface. The grooves and ridges are oriented more or less at right angles to the main arrays. They splay from a short straight to gently curved margin, not from a point. On the main shatter cone surface, they are seen to splay in both directions, and both outward and inward curvature is present.

We have not studied the shatter cones petrographically. However, the limestone and dolostone composition of the target rocks and lack of quartz suggest that typical microscopic planar deformation features in silicate minerals may not be recorded. None was observed in a single thin section of the diabase dyke. The lower Cambrian clastic unit was found to be covered in 2010.


The feature is interpreted as the lower level of a terrestrial meteorite impact crater on the basis of the anomalous, localized, concentric orientation of dipping and faulted strata, from older in the middle to younger outward, which are consistent with it being the uplifted central area of a complex crater. The pervasiveness of well-developed shatter cones supports this interpretation, despite the lack of melting and small-scale brecciation or fracturing, which are also commonly generated by impact (e.g., Grieve and Pilkington 1996; Crósta et al. 2010; French and Koeberl 2010).

Undeformed glacial deposits and glacial striae overlie and cross-cut deformed strata implying an age older than the last glacial retreat at about 8000 yr. Deformation displaces rocks as young as the Late Ordovician Thumb Mountain Formation, so the impact must have occurred after approximately 450 Ma. The WSW–ENE-trending normal faults that can be followed across the feature are probably related to the opening of the Canada Basin prior to the Early Cretaceous (Miall 1979). This indicates that the impact can be bracketed in age to between 450 and approximately 130 Ma. The cruder sculpture of shatter cone surfaces in the Thumb Mountain Formation suggests that it took place after regional hydrothermal dolomitization, because recrystallization would probably have annealed the surfaces. The timing of this diagenetic phase in the western Arctic Islands is unknown, but is probably Late Devonian (approximately 360 Ma; Wendte 2012). Because of the substantial erosion that has taken place after impact—well below the original level of impact melt and brecciation—it is reasonable to surmise that the impact formed during a time of subaerial exposure in the Mesozoic or early Cenozoic. Correlation with more complete stratigraphies elsewhere in the western Arctic suggests that the amount of erosion was at least 700 m and could have reached 5000 m. Only one other impact feature is known from the Canadian Arctic Islands, the Haughton impact structure, which is Eocene in age and it is not so deeply eroded (Osinski et al. 2005). The ages of other Phanerozoic impact structures in North America are variably constrained, but there is no obvious pattern to suggest episodes of heavier bombardment (Earth Impact Database:

Estimating the size of the original crater and the impactor is hampered by the incomplete exposure of the impact structure as well as the uncertainty about when the impact occurred and consequently how much of the overlying stratigraphy has since been removed by erosion. On the basis of the apparent relationship between the stratigraphic thickness of rocks in the structural uplift (i.e., those deformed in the central uplift), Grieve and Therriault (2000; see also Grieve and Pilkington 1996; Therriault et al. 1997) suggested that the size of the original crater can be calculated from the diameter of the structural uplift:

display math(1)

In the new impact structure, the dip of the uplifted beds shallows markedly over a relatively short distance laterally, so that delineating the diameter of the central uplift is imprecise. The estimated width of the whole feature based on tilted strata is about 25 km, which would give a crater diameter of 80 km. However, if the diameter is based on the most steeply dipping strata, some 15 km gives a crater diameter of about 50 km, which seems a reasonable estimate. The impactor required to generate an impact structure of this size would be approximately 2 km in diameter, calculated using the online program Crater (

The presence of shatter cones within 10 m of the dyke was interpreted in the PetroCanada Exploration Inc. (1978) report as due to high pressure developed during injection of the igneous material. However, the distribution of shatter cones over a wide area indicates an event that affected all the lower Paleozoic strata as well, and shows that deformation was not related to igneous intrusion. Notably, shatter cones are never observed in the vicinity of Franklin intrusions elsewhere on Victoria Island. Now, it is accepted that such shatter cones are shock-induced tensile features that were created by hypervelocity impact early in crater excavation and before formation of the central uplift (e.g., French and Koeberl 2010). It is commonly considered that they record shock pressures of approximately 2–20 GPa; planar deformation features in silicate minerals record the higher pressures, up to approximately 30 GPa, before melting takes place (e.g., Reimold and Koeberl 2008). In detail, however, it has been argued that shatter cones formed right after the passage of the shock wave, rather than from shock compression itself (Baratoux and Melosh 2003; Wieland et al. 2006).

The shatter cones are morphologically identical to the many other examples occurring in sedimentary rocks (e.g., Fackelman et al. 2008; French and Koeberl 2010; Ferrière et al. 2011). Their variable orientation in the new impact feature, especially in the thinner beds, is a phenomenon that has been noted also in some other examples (e.g., Wieland et al. 2006; Osinski and Spray 2008). For this reason, the common assumption that the apices point to the source of the shock wave is not strictly valid. The variation in apices has been attributed to scattering and reflecting of the shock wave due to rock inhomogeneities (Baratoux and Melosh 2003; Wieland et al. 2006). However, while bed thickness seems to have played a role, most of the sedimentary strata in the region are individually quite homogeneous such that it would seem unlikely that rock properties alone were responsible for the morphological variation in shatter cone morphology. For example, we did not see any effect caused by silicified macrofossils in the Thumb Mountain Formation or burrowed bed soles and horizons with intraclasts in the Stripy unit. Nor have we observed obvious inhomogeneities in the dolomudstone that could have led to the small concavo-convex surfaces in the Stripy unit. On the other hand, keeping in mind that even deeply buried strata are water-wet, we suggest that pore water may have influenced stresses at the millimetric to decametric scale. This could have induced bursts in directions oblique or normal to the shock wave propagation direction, as well as possibly the nested aspect of many shatter cones. The presence of fluids under pressure is well known to weaken the tensile strength of rocks, and it may be that shatter cones, in some cases, are a form of hydraulic fracturing. The presumed near absence of pore water in the diabase dyke may explain why shatter cones are poorly developed in this rock type. It has already been suggested that pore water could play a role in the nature of cratering itself (Kenkmann et al. 2011).


Field evidence in the form of a roughly circular domain of steeply dipping beds in regionally flat-lying strata, along with pervasive development of shatter cones, indicates that this 25 km wide feature in northwestern Victoria Island is a deeply eroded meteorite impact structure. It is not known when it occurred, but it is likely late Paleozoic or Mesozoic in age. The original crater size was about 50 km wide and it has suffered deep erosion. The size of the impactor is estimated at about 2 km. Shatter cones are well developed, and their morphology suggests a role for pore water in the transmission of shock-induced stress. The Tunnunik impact structure is the 30th meteorite impact feature identified in Canada.


KD and BRP made the field observations on the impact structure. KD, TH, TB, JB, and RHR conducted the regional mapping; BP and KD were responsible for biostratigraphic determinations. TB made the DEM. We thank Polar Continental Shelf Project of Natural Resources Canada for logistical support, and the Olokhaktomiut Hunters and Trappers Committee, the Hamlet of Ulukhaktok, and Arctic Char Inn (Ulukhaktok) for various forms of assistance during the 2009, 2010, and 2011 summer field seasons. Funding was provided through the Geo-mapping for Energy and Minerals Program of the Geological Survey of Canada.

Editorial Handling

Dr. John Spray