UNIVERSITY OF GOTHENBURG Department of Earth Sciences Geovetarcentrum/Earth Science Centre

Mafic Impact Ejecta from

the Lockne Crater, Sweden

Axel S.L. Sjöqvist

ISSN 1400-3821 B644 Bachelor of Science thesis Göteborg 2011

Mailing address Address Telephone Telefax Geovetarcentrum Geovetarcentrum Geovetarcentrum 031-786 19 56 031-786 19 86 Göteborg University S 405 30 Göteborg Guldhedsgatan 5A S-405 30 Göteborg SWEDEN Abstract The Lockne crater formed ca. 455 million years ago in a target composed of a crystalline Proterozoic basement, overlain by early Palaeozoic sedimentary rocks, and seawater. The diameter of the inner crater is ca. 7 km, with an up to 2.5 km wide brim of ejected, crushed crystalline basement spread around the inner crater (Lindström et al., 2005). Large ejecta blocks of crushed crystalline basement with sizes ranging from a few meters to ca. 100 m in diameter are found as far as 15 km away from the edge of the crater. The largest ejecta blocks have created smaller, secondary craters around Lockne with diameters of 100–200 m (i.e. Målingen, Karsåtjärnen, and Kloxåsen) (Sturkell, 1998a). The field distribution of the Kloxåsen ejecta blocks is unknown and their relation to the rest of the is important for a fuller understanding of the formation of the Lockne and other similarly sized craters in mixed (crystalline and sedimentary) targets. Here, the distribution of ejecta around Kloxåsen is mapped. Further, the mineralogical composition of the Kloxåsen ejecta blocks is unknown; therefore a mineralogical analysis of this material was done to establish the source rock of the mafic ejecta blocks, to characterise the hydrothermal alteration, and to determine the levels of shock in the ejecta. Field mapping of the Kloxåsen area showed that the distribution of the mafic ejecta is much more widespread than previously thought, although the granitic breccia dominates. Optical petrography, SEM-EDS, and Raman spectroscopy revealed curious hydrothermal alteration in the mafic clasts. Olivine is completely replaced by pseudomorphs of serpentine that in turn are wholly or partly replaced by calcite (and translucent sphalerite) and chlorite. Titanaugite is replaced by quartz, chlorite, calcite, and anatase, although unaffected titanaugite does also occur. Some biotites appear “split open” and are replaced by quartz and anatase, but no kinkbanding has been observed. Titanaugite affected by mechanical twinning suggests shock pressure of at least 5 GPa (Melosh, 1989). Fluid-inclusion analysis in calcite revealed that the hydrothermal system in Kloxåsen is the same as in the rest of the crater (M. Ivarsson, personal communication, October 4, 2010). Sector-zoned titanaugite oikocrysts, which are only known to occur in CSDG rocks (Claeson et al., 2007), strongly suggests the Åsby dolerite as a protolith for the observed mafic impact ejecta at Kloxåsen.

Keywords: Lockne, impact, ejecta, breccia, Åsby dolerite, CSDG, hydrothermal, calcite Abstrakt Locknekratern bildades för ca. 455 miljoner år sedan i en målberggrund som bestod av Proterozoiska kristallina bergarter som täcktes av tidig-Paleozoiska sediment och havsvatten. Den inre kraterdiametern är ca. 7 km och omges av en upp till 2,5 km bred rand av krossat, kristallint berg (Lindström et al., 2005). Stora ejektablock som består av krossat berg, som varierar i storlek från några meter upp till ca. 100 m i diameter, kan finnas upp till 15 km utanför kratern. De största ejekta har skapat mindre, sekundära kratrar runt Locknekratern med diametrar mellan 100 och 200 m (Målingen, Karsåtjärnen och Kloxåsen) (Sturkell, 1998a). Fältdistributionen av Kloxåsens ejektablock är okänd och deras relation till resten av astroblemet är viktig för en bättre förståelse av Locknekratern bildning och andra likstora kratrar i blandad (kristallin och sedimentär) målberggrund. Här redovisas en detaljkartering av Kloxåsen. Samt är den mineralogiska sammansättningen av Kloxåsens ejektablock okänd, därför har en genomgående mineralogisk analys utförts för att etablera ursprungsbergarten för den mafiska ejektan, att karakterisera den hydrotermala omvandlingen och att determinera graden av shock i denna bergart. Fältkartering av Kloxåsenområdet visade att förekomsten av den mafiska ejektan var större än vad som tidigare var känt, även om den granitiska ejektan dominerar. Optisk petrografi, SEM-EDS och Raman spektroskopi visade underlig hydrotermal omvandling i den mafiska ejektabreccian. Olivin är helt omvandlat till pseudomorfer av serpentin som i sin tur är helt eller delvis ersatta av kalcit (och genomskinlig zinkblände) och klorit. Titanaugit är i vissa fall omvandlat till ett aggregat av kvarts, klorit, kalcit och anatas, men helt opåverkad titanaugit förkommer också. En del biotitkristaller ser ut som de har blivit “klyvna”, men ingen kinkbanding har observerats. En titanaugitkristall som visar effekten av mekanisk tvillingbildning avslöjar ett shocktryck på minst 5 GPa (Melosh, 1989). Analyser av vätskeinneslutningar i kalcit visar att hydrotermala systemet i Kloxåsen har samma signatur som i resten av Locknekratern (M. Ivarsson, personlig kommunikation, 4 oktober, 2010). Sektorzonerade titanaugit oikokrister, vilka bara förekommer i CSDG:ns bergarter (Claeson et al., 2007), argumenterar starkt för Åsbydiabasen som protolit för den mafiska ejektan i Kloxåsen.

Keywords: Lockne, impakt, ejekta, breccia, Åsbydiabas, CSDG, hydrotermal, kalcit

Table of Contents

1 INTRODUCTION 1

AIMS AND OBJECTIVES 1 HISTORY OF IMPACT CRATERING 1 IMPACT CRATERING 2 CONTACT/COMPRESSION STAGE 2 EXCAVATION STAGE 3 MODIFICATION STAGE 6 SIMPLE AND COMPLEX CRATERS 6 7 HISTORY OF THE LOCKNE CRATER 7 THE LOCKNE CRATER 8 PRE-IMPACT ROCKS 8 9 POST-IMPACT ROCKS 10

2 METHODS 10 MAPPING 10 OPTICAL PETROGRAPHY 11 RAMAN SPECTROSCOPY 11 SCANNING ELECTRON MICROSCOPY 11

3 RESULTS 11

DISTRIBUTION OF THE EJECTA 11 PRE-IMPACT ROCKS 11 IMPACTITES 11 MINERALOGY 15 ÅSBY DOLERITE 15 DOLERITIC TANDSBYN BRECCIA 15

4 DISCUSSION 28

5 CONCLUSION 30

6 ACKNOWLEDGEMENTS 30

7 REFERENCES 31 12

1 Introduction 1.2 History of impact cratering What was once regarded as a relatively minor 1.1 Aims and objectives astronomical process that only resulted in exotic The aim of this study is to obtain a fuller yet insignificant geological structures has over understanding of the origin, distribution, and the last 40 years gradually grown into one of the petrologic characteristics of the mafic impact mainstream themes in geology. Impact cratering ejecta blocks in the Kloxåsen area of the Lockne is now accepted as more abundant, larger, older, impact crater in central Sweden. and more economically and biologically Ejecta blocks are masses of rock that were part significant than anyone would have thought of the target rock material previous to the before the 1960s. impact that have been ejected from the crater Meteorite impacts, geologists then agreed, did interior. During the formation of the Lockne occur on Earth, but merely resulted in small impact crater, ejecta blocks of crushed, crater structures that were short-lived over crystalline basement, up to about 100 m in geologic time and were mostly preserved in diameter, were thrown up to 15 km out of the desert areas. Cratering events were regarded as crater (Sturkell, 1998a). Distant ejecta of the a process that only produced a local hazard; very Lockne have been found 40 km few people believed the striking of extra- away from the crater (Sturkell, Ormö, Nõlvak, & terrestrial objects could result in major Wallin, 2000). geological effects. This study will particularly focus on the mafic An event that fuelled this skepticism toward ejecta in the Kloxåsen area. Impacts in mafic terrestrial impact cratering is what is known as target rocks are not yet well understood, since the Tunguska event (60°53′09″N 101°53′40″E): only two other impact craters on Earth have an enormous explosion in Siberia on June 30, mafic targets (Hagerty & Newsom, 2003): Lonar 1908. In 1921 Russian mineralogist Leonid Kulik Lake impact structure, India (19°58’N, 76°31’E), began to show an interest in the area and and Logancha crater, Russia (65°31’N, 95°56’E). participated in several expeditions in the Mafic targets are common on for instance the following decades. From local accounts of the Moon and Mars. Thus the mafic ejecta in the event, Kulik deduced that the explosion was Kloxåsen area are of interest for studies of caused by a meteorite impact. During the first impact processes in mafic targets. expedition in 1927 Kulik and his fellow The field distribution and source rock of the researchers, while standing on a ridge Kloxåsen mafic ejecta have never been formally overlooking the devastated area, were established. The mineralogical characteristics of confronted with the problem that an impact the ejecta blocks are also unknown. Therefore crater was absent. Today it is believed that the the main objectives of this study are: Tunguska event was caused by a body of asteroidal or cometary origin that exploded in 1) Mapping of the Kloxåsen area bedrock to the atmosphere (Krinov, 1966; Trayner, 1997; determine the field distribution of the mafic Vasilyev, 1998; Bronshten, 2000c: in Farinella et ejecta blocks. al., 2001). 2) Thorough mineralogical analyses of the As a result of planetary exploration, the mafic ejecta material to establish its textural scientific community and the public nowadays characteristics, indications of hydrothermal both—rightfully—regard impact cratering as alteration, and to identify possible signs of shock one of the geologically most significant metamorphism. processes. Events of extra-terrestrial objects 3) Comparison of the mineralogy in the mafic striking the Earth’s surface have created vast ejecta with mafic rocks from the vicinity of the geological disturbances, generated incredible Lockne crater, in order to identify the protolith volumes of igneous rocks (e.g. Vredefort, South of the mafic ejecta and to characterise any Africa), formed massive ore deposits (e.g. alteration in them. Sudbury, Canada), and caused—at least—one A detailed geochemical characterisation and biological mass extinction (e.g. Chicxulub, analysis of the mafic ejecta breccia at Kloxåsen is Mexico (Schulte et al., 2010)). beyond the scope of this study. The At present, there are 175 confirmed meteorite geochemistry of Kloxåsen’s ejecta will be the craters known on Earth (Rajmon, 2009), but subject of future research. there are many more geological structures that are suspected to be impact craters and require

1 43

closer examination; presently there are 574 other as one single, continuous process. probable and possible craters that are still 1.3.1 Contact/compression stage unconfirmed. The , by the This is the first stage in the formation of an Planetary and Space Science Centre, reports 178 impact crater and it begins as soon as a cosmic confirmed craters (accessed May 17th, 2011). object strikes the surface of another object at New possible impact craters are identified on a hypervelocity. Hypervelocity generally refers to regular basis, especially since in this age of a velocity of approximately 11 km/s or more. technology and information everyone can have Cosmic objects hit the Earth at 11.2—72 km/s access to high-quality satellite images of the (French, 1998). The impact velocity of an entire Earth for free on their home computers. astronomical object depends not only on its own The first geological structure on Earth to be velocity but also on the Earth’s relative velocity, identified as an impact crater is Barringer in i.e. whether the projectile is “catching up” with Arizona, U.S.A. (35°02'N, 111°01'W), and was the Earth from the rear or whether it is a “head- documented by D.M. Barringer in 1909. It is on collision”. Therefore impact velocities follow more commonly referred to as Meteor Crater a normal distribution curve. and is perhaps the most widely recognised The lowest impact velocity (“catching up”) is impact crater. The crater is 1.186 km in by definition equal to the escape velocity for an diameter and has an age of 0.049 ± 0.003 Ma object to be launched into space from the (Phillips et al., 1991: in Rajmon, 2009): a surface of the target astronomical object: 11.2 geologically very young age, to which the crater km/s for the Earth (French, 1998). largely thanks its exceptional preservation. The maximum impact velocity is the sum of The largest identified impact crater on Earth is two velocities; the velocity of the projectile in its the in South Africa (27°00’S, orbit around the Sun (heliocentric velocity), 27°30’E) with an estimated diameter of 300 km which can be thought of as the escape velocity and it is also the second oldest crater known on from the solar system, and the Earth’s orbital Earth with an age of 2023 ± 4 Ma (Kamo, velocity around the Sun. The heliocentric Reimold, Krogh, & Colliston, 1996: in Rajmon, velocity at the orbit of the Earth is roughly 42 2009). The oldest, identified impact structure on km/s and the Earth’s orbital velocity is about 30 Earth is the Suavjärvi crater, Russia (63°7’N, km/s (French, 1998). The sum of these two 33°23’E), with an age of approximately 2.4 Ga constrains the maximum impact velocity on (Mashchak & Naumov, 1996). Larger and older Earth: 42 + 30 = 72 km/s, for a dead, head-on possible craters have been identified, but have collision. However, in the majority of impacts not yet been confirmed, for instance the 2.975 the orbits of the colliding objects are inclined to Ga Maniitsoq structure, West Greenland. each other and will thus be geometrically added The largest crater in the Solar System is to each other, producing the variation of Earth- suspected to be the Borealis Basin (67°N, 208°E) encounter velocities (geocentric velocities). on Mars (Andrews-Hanna, Zuber, & Banerdt, When hit by an object travelling at hyper- 2008): an elliptical basin with a long axis of velocity, the yield strength of a material is tiny 10,600 km and a short axis of 8,500 km. compared to inertial stresses imposed upon it by the impact (Melosh, 1989), thus even solid rock 1.3 Impact cratering will momentarily behave like a fluid during a Recognising impact craters and structures may hypervelocity impact. be difficult on the surface of the Earth. Many The place where the cosmic projectile strikes is craters have been erased permanently from the called the target. If the target is solid rock, the Earth’s geological record through erosion and projectile is stopped in a split second and plate tectonics. On other astronomical objects, penetrates the surface no more than 1—2 times such as the Moon, where crustal reworking is a its own diameter (Kieffer & Simonds, 1980; less significant process, impact craters appear O’Keefe & Ahrens, 1982, 1993; Melosh, 1989, more obviously at the surface than on Earth. Chapter 4: in French, 1998). For example, a Impact cratering is a complex process, but, in cosmic object travelling at 30 km/s with a an attempt to simplify the formation of impact diameter of 500 m would be stopped within craters, it can arbitrarily be divided into three 0.033 s. stages: contact/compression stage, excavation The object’s kinetic energy, the energy an stage, and modification stage (see French (1998) object possesses due to its mass and velocity, for a more detailed description; Fig. 1). In reality measured in joules (J), is equal to half its mass these described stages grade fluently into each multiplied by its velocity squared (1/2!"²),

2 65

with its mass m in kg and its velocity v in m/s. peak shock-wave pressure reaches 1—2 GPa Because the object travels at cosmic hyper- and at this point the shock waves become velocity—which is squared in the equation—its seismic waves and propagate through the kinetic energy becomes massive. bedrock at the speed of sound in the rocks The kinetic energy is transferred to the target (Kieffer & Simonds, 1980: in French, 1998), rocks upon impact by compressional shock similarly to seismic waves from an earthquake waves that travel faster than the speed of sound or volcanic eruption. in the bedrock—supersonic shock waves— The duration of the contact/compression stage which are generated at the interface between is dependent on the reflected shock wave that the projectile and target. Importantly, a shock travels through the projectile itself (Melosh, wave is also reflected back into the cosmic 1989, p. 57—59: in French, 1998). When this object itself (French, 1998). shock wave reaches the back end of the cosmic The amount of energy released by the impact object, it is reflected forward into the object as a of a cosmic object is very difficult to compre- tensional wave, or release wave. As the release hend and this necessitates it to be compared to wave propagates from the back to the front of other geologic processes, to place it into a the projectile, it gradually unloads the high context that a geologist might understand. shock pressure. This is the opposite of compres- For example, a roughly spherical projectile sion and, in physical terms, similar to the unloa- with a density of approximately 3,500 kg/m³ ding of a coil spring. As a result of the high that is 6.0 m in diameter and travels at a velocity pressure and temperature that were present in of 30 km/s carries a kinetic energy of 1.4 × 1015 the object, it is virtually completely vaporised J, which is roughly equivalent to 17 Hiroshima and melted during the release. atomic bomb explosions (French, 1998). (The When the projectile itself is unloaded, the above-described cosmic object is not at all release wave continues into the target rocks and unrealistic or uncommon in our solar system, decompresses those as well. This is the end of although, due to atmospheric effects on Earth, it the contact/compression stage. would probably burn up in the sky before ever For most impact events it lasts less than a hitting the surface.) second. A cosmic projectile merely a few kilometres across, which is still a moderately-sized object 1.3.2 Excavation stage compared to known and comets, During the excavation stage the crater is opened would upon impact release more energy in a up due to complex interactions between the matter of seconds than the entire Earth releases shock waves and the ground surface (Melosh, in thousands of years through volcanism, 1989, Chapter 5; Grieve, 1991: in French, 1998). earthquakes, tectonic processes, and heat flow At the end of the contact/compression stage, (French, 1998). (“If that doesn't make the hair the projectile is surrounded by shock waves that stand up on the back of your neck, read it again are rapidly radiating outwards. The shock waves until it does, because it is important." — Steve that hit the ground surface are reflected Grand.) downward into the target bedrock as release The shock waves that are produced upon waves. In the near surface bedrock, the stress in impact lose energy rapidly as they travel away the tensional release exceeds the mechanical from the impact site due to the shock-wave front strength of the rocks and the rocks are covering a larger hemispherical area with fractured. The reflection process also converts increasing radial distance, but the shock waves part of the shock-wave energy into kinetic also lose energy due to heating, deformation, energy, which accelerates the affected rocks and acceleration of the target rocks. outward: a large portion as individual ejecta The peak shock-wave pressure during a typical blocks. cosmic impact event may easily exceed 100 GPa, The forces that drive the target rock outward which produces total melting and vaporisation produce a symmetric excavation flow around of the meteorite and large volumes of surroun- the centre of the developing crater structure. ding bedrock. Pressures of 10—50 GPa may be The bowl-shaped depression that forms is called present over many kilometres from the point of the transient crater or transient cavity. The impact, producing distinctive shock metamor- diameter of the final crater is usually 20—30 phic effects in a large volume of target bedrock times the diameter of the projectile (French, that was not melted by the impact. 1998). Even further out from the point of impact the The transient crater is divided into the

3 Figure 1. Presentation of impact cratering in its different stages in a simple crater (French, 1998). 4 Figure 2. Presentation of the different stages of impact cratering in a (French, 1998).

5 78

excavated zone and displaced zone. In the excavated zone the velocities given to the fragmented rock can be as high as several kilometres per second and the material is ejected beyond the rim of the final crater. In the deeper displaced zone the excavation flow velocity is not high enough to eject material and the direction of the flow is not upward; instead the material is driven downward and outward. The excavation stage, although relatively much longer than the contact/compression stage, happens in the blink of an eye in a geological context. The transient crater of Barringer took about 6 s to form, while a crater with a diameter Figure 3. Backjet in a liquid after a drop impact of a of 200 km takes about 90 s to form (French, coloured liquid (source: http://upload.wikimedia.org/ 1998). The excavation stage ends when the wikipedia/commons/d/de/Blue_Droplet.jpg). transient crater has reached its maximum size. in diameter. During the modification stage, a simple crater is quickly filled with ejecta that 1.3.3 Modification stage were shot up in the air during the impact, called The modification stage takes over instantly after fallback ejecta, and debris from the walls and the excavation stage has ended. The expanding rim. This infill is called the breccia lens or crater- shock waves that were generated at the impact fill breccia and it consists of both shocked and now exist as low-pressure elastic waves beyond unshocked rocks and lenses or fragments of the rim of the crater and play no role in the shock-melted rock (impact melt). further development of the final crater; the Complex craters (Fig. 2) are larger than simple transient crater is instead modified by gravity craters and are characterised by a centrally and rock mechanics. uplifted region, a largely flat crater floor, and The immediate process of modification lasts inward collapse around the rim (Dence, 1968; only slightly longer than the excavation stage: Grieve, Dence, & Robertson, 1977, 1981; Grieve, less than a minute for small craters, a few 1991: in French, 1998). On Earth the transition minutes for large ones (Melosh, 1989, Chapter 8, from simple to complex crater occurs at crater p. 141—142). diameters around 4 km in massive crystalline The modification stage has no clearly marked target rock and 2 km in sedimentary rock. This end. The processes of uplift and collapse grade transition is dependent on gravitational accele- into slower processes of geological modification ration and thus varies depending on the size of like mass movement, isostatic uplift, erosion, the target astronomical object (French, 1998). and sedimentation. This process of complex-crater formation with a A simple way of defining what marks the end central uplift is much similar, although more of the modification stage is “when things stop complex, to the backjet after a single drop of falling”. liquid hits a liquid surface (Fig. 3) (e.g. Melosh, 1989, p. 148; Taylor, 1992, p. 168: in French, 1.3.4 Simple and complex craters 1998). However, these processes take place on a Depending on the size of the crater and on the far larger scale—and in solid rock—in impact characteristics of the target rock, the transient events. crater is modified differently. A small crater is An approximation of the degree of strati- mainly modified by the gravitational collapse of graphic uplift (SU) that occurs in complex the upper wall. In large craters the modification impact craters is roughly one-tenth of the final stage may introduce major structural changes, crater diameter (D), !" = 0.1 ! (French, 1998). such as uplift of the central part of the floor and Detailed, statistical analyses of the relation collapse around the rim of the crater. between the stratigraphic uplift of the central Impact structures can be divided into three uplift and the crater diameter have yielded groups, depending on the degree of modifi- !" = 0.06 !!.! (Grieve, Robertson, & Dence, cation: simple craters, complex craters, and 1981, p. 44: in French, 1998) and !" = multi-ringed basins. 0.086 !!.!" (Grieve & Pilkington, 1996, p. 404: in Simple craters (Fig. 1) appear as bowl-shaped French, 1998), but both equations are practically depressions that are less than a few kilometres identical and !" = 0.1 ! is a reasonable and

6 109

Figure 4. Comparison between the temperature-pressure conditions for endogenous metamorphic reactions and shock metamorphism. Note that the horizontal axis is logarithmic. (Adapted from Stöffler, 1971, Fig. 1; Grieve, 1990, p. 72; Grieve & Pesonen, 1992, Fig. 9: in French, 1998)

useful approximation. This, then, implies that for quartz ( and ) and diaplectic craters with large diameters (D = 100—200 km) glasses. More detailed descriptions of various the vertical uplift of the crustal rocks beneath forms of shock metamorphism are given by the crater is in the range of 10—20 km and an French (1998) and Melosh (1989). uplift of approximately this magnitude has been The rocks that are exposed to the highest estimated for Vredefort, South Africa, based on pressures are vapourised and melted. Diaplectic geological characteristics of the structure’s glasses are formed by amorphisation—without central uplifted granitoids. melting—when the rocks are relieved of the shock pressure and the crystal structure 1.4 Shock metamorphism ”relaxes”. Planar deformation features can occur During a hypervelocity impact an incredible in many minerals, such as feldspars and zircon, amount of energy is released locally and in an but are often especially well-developed in quartz extremely short period of time. The tempera- crystals. PDFs occur as thin, penetrative lamellae tures in a cosmic impact event due to shock (50—500 nm) filled with diaplectic glass and heating can be over 6000 K and pressures may occur in 15 different crystallographic reaching up to several hundred GPa (French, orientations in quartz. Shatter cones can occur 1998). These conditions of temperature and in all types of rock as macroscopic, conical, pressure cannot be matched near the surface by striated features that can be up to a few metres any other geological process, e.g. earthquakes long, commonly radiating out from the impact and volcanic eruptions. centre. These extreme conditions of temperature and pressure give rise to impact-specific metamor- 1.5 History of the Lockne crater phic effects and lithologies. Fig. 4 compares the The Lockne area has for a long time fascinated temperature and pressure conditions for endo- geologists. It has been known since 1900 for its genous metamorphic reactions with those invol- unique, coarse-clastic facies of Baltoscandian ved in shock metamorphism. Ordovician rocks, which were interpreted by The most important shock metamorphic both Wiman (1900) and Thorslund (1940) as features are shatter cones, planar deformation facies of sedimentary rocks that were deposited features (PDFs), high-pressure polymorphs of in a high-energy, coastal environment.

7 1112

Lindström (1971) argued that a high-energy coastal environment was unlikely in the Baltoscandian Ordovician and Lindström, Simon, Paul, and Kessler (1983) interpreted the coarse- clastic facies as debris flows, which were named Lockne Breccia, and turbidites. Simon (1987a) elaborated a gravity-flow model of Lockne and also discussed two further curious facies of the Lockne area. Firstly, a regular occurrence of apparently volcanic ash in arenitic beds of a facies of rock that has turbidite-like characteristics, locally known as Loftarsten (anglicised: Loftarstone). Secondly, Simon discussed the occurrence of a mono- mictic, crystalline breccia that is composed mainly of granite, named Tandsbyn Breccia by Lindström and Sturkell (1992). Previously it had been interpreted by Thorslund (1940) as a facies that had resulted from Early Palaeozoic, terrestrial weathering. Simon admitted his disability to explain both facies from what he could reconstruct of the geology in the Lockne area, but stated that he did not see signs of weathering in the crystalline breccia; rather, he Figure 5. Location of the marine Lockne impact structure saw signs of mechanical crushing. in central Sweden. Wickman (1988) was the first who published a mentary rocks, impactites, post-impact sedimen- paper that discussed the existence of a meteorite tary rocks and, on top of this sequence, Caledo- impact structure in the Lockne area, on evidence nian overthrusts (Sturkell, 1998a). from Simon (1987a), however, Wickman placed Kloxåsen is located on a Caledonian overthrust, the centre of the hypothesised crater structure approximately 7 km from the centre of the roughly 20 km northwest of Lockne. Looking at Lockne crater (see Fig. 6); however, the distance the distribution of rock facies that are possibly to the crater was not much greater previous to related to an impact event, this claim cannot be the tectonic movement (Lindström, Sturkell, supported. Törnberg, & Ormö, 1996: in Sturkell, 1998a). The meteoritic origin of the geology of the Lockne area was confirmed by Lindström and 1.6.1 Pre-impact rocks Sturkell (1992) and the shape of the inner crater The Proterozoic basement in the Lockne area is could be identified with a diameter of just over 7 dominated by the post-orogenic Revsund suite km. that is made up of granites that intruded at 1.8— 1.77 Ga (Sturkell, 1998a). The oldest rocks in the 1.6 The Lockne crater vicinity of the Lockne crater are part of the Early The area that has been studied in this thesis— Svecofennian Börön volcanic suite (Mansfeld, Kloxåsen—is part of the Lockne impact Sturkell, & Broman, 1998: in Sturkell, 1998a). structure. The Lockne crater structure is Other rocks from the Early Svecofennian are situated in Jämtland in central Sweden, about 20 foliated granitoids from the Stor-Handsjön area km south of Östersund (63°00’20”N, 14°49’30”E; that are associated with metasedimentary see Fig. 5). The crater formed in the Middle gneiss, veined paragneiss, and migmatite Ordovician at approximately 458 Ma (recal- (Högdahl, Gromet, & Claesson, 1996: in Sturkell, culated from 455 Ma in Sturkell, 1998a). It is 1998a). The youngest Proterozoic rocks in the located at the erosional front of the Caledonides Lockne area are sills of Åsby dolerite with an age and has been protected from erosion by of 1.2 Ga that rarely exceed a thickness of 100 m Caledonian overthrusts that have now been in the Lockne area (Patchett, 1978: in Sturkell, eroded away (Lindström, Ormö, Sturkell, & von 1998a). Dalwigk, 2005). A peneplain had formed on the Proterozoic The geology of the Lockne area contains basement before the Palaeozoic marine sedi- Proterozoic crystalline rocks, pre-impact sedi- mentation had begun in the area (Sturkell &

8 Figure 6. Geologic map of the Lockne area. Kloxåsen is indicated with an arrow. Modified from Lindström et al. (2005).

9 1314

Lindström, 2004). It is in Sweden commonly Marine impacts generate facies of rock that are referred to as the Subcambrian Peneplain and not known to form in continental impacts, extends roughly 1—2 km out westward from the collectively referred to as resurge deposits. Palaeozoic deposits in the Lockne area. Pre- Resurge deposits are formed when the transient impact sedimentary rocks consisted of 30 m crater in the water mass collapses and large unconsolidated Cambrian clays—which are now volumes of material are suspended and moved shale—and 50 m lithified Ordovician limestone by the water that is surging back into the crater (Sturkell, 1998a). cavity. The suspended material is deposited as sediment when the energy of the water declines. 1.6.2 Impactites In the Lockne crater the resurge deposits can Most of the identified impact craters on Earth be split into the Lockne Breccia and the arenitic have been formed in a continental environment. Loftarstone. The transition between these two is What distinguishes the Lockne crater from most either sharp or gradual (Sturkell, 1998a). A other craters is that it is a marine impact crater. sharp transition between the two facies occurs This has various implications for the preser- where there is an erosional surface in the vation of the crater structure, the facies of rock Lockne Breccia. that are formed, and the cratering process. The Lockne Breccia is a polymictic, sedimen- A crater structure that is formed in a conti- tary breccia that was deposited by the resurging nental environment is most likely subject to water. Clasts in the Lockne Breccia consist of erosion and modification immediately after its limestone, fragments of intrusive and extrusive formation. In contrast, directly after a marine igneous rocks, and even crystalline Tandsbyn impact crater is formed, sedimentation can Breccia (Sturkell, 1998a). continue as it did before the impact. This means The Loftarstone, which can be observed as far that the newly formed crater structure has good as 45 km from the Lockne crater, is an arenitic potential to be covered with a protective layer of unit composed of limestone clasts, fossil frag- sediment, which results in better preservation of ments, clasts of Proterozoic rocks, more or less the crater over geologic time. altered melt rock fragments, monomineralic The target of the Lockne impact crater is grains of quartz and feldspar, and calcite cement mixed; this means it consists of sedimentary and in a fine-grained matrix (Sturkell, 1998a). crystalline rocks (and, indeed, water). The target Quartz with planar deformation features is consisted of a Proterozoic crystalline basement abundant in the Loftarstone. An iridium anomaly that, at the time, was overlain by approximately ranging from 0.8 to 4.5 ppb has been observed 30 m of unconsolidated Cambrian clay and by Sturkell (1998b). approximately 50 m of lithified limestone (Sturkell, 1998c). The water depth at the time of 1.6.3 Post-impact rocks the impact is constrained by 500 and 700 m The oldest facies of post-impact rock that through computer modelling (Ormö, Shuvalov, & covered the Lockne impact crater immediately Lindström, 2002). after it was formed is the Dalby Limestone Immediately surrounding the impact structure (Sturkell, 1998a). is an impact-generated, crystalline, authigenic The whole area has at one point been covered breccia, named Tandsbyn Breccia by Lindström by Caledonian overthrusts, which were up to 5 and Sturkell (1992). The Tandsbyn Breccia is km thick (Sturkell, Broman, Forsberg, & generally monomictic on a small scale (Sturkell, Torssander, 1998). 1998a), but its composition depends on the local basement geology. 2 Methods The that normally forms around an impact crater has in the Lockne impact 2.1 Mapping largely been wiped out by sedimentary rewor- Detailed geological mapping of the Kloxåsen king in the modification stage, when the tran- area was carried out in 8 days during July 2010. sient crater in the Ordovician ocean collapsed GPS coordinates of outcrops were taken in the (Sturkell, 1998a). Only the largest ejecta have field and imported to a digital map of Kloxåsen not been moved by post-impact reworking; for using Generic Mapping Tool and Adobe Illus- instance the ejecta at Kloxåsen. The crystalline trator. In total 257 outcrops were observed and ejecta breccia is similar in character to the used to produce a map over Kloxåsen, covering Tandsbyn Breccia. The ejecta breccia will in this approximately 1.5 km2. In the dense forest of study be referred to as Tandsbyn Breccia. Kloxåsen the accuracy of the coordinates was

10 1516

generally better than 10 m, whereas in the open Tandsbyn Breccia are marked with a white fields accuracies down to 5 m were achievable. check mark (✓) and the underlying portion of the ejecta breccia that is coloured grey is 2.2 Optical petrography mapped as doleritic. The grey check mark shows The mineralogy of the mafic Tandsbyn Breccia at the location of a polymictic Tandsbyn Breccia Kloxåsen was analysed in 17 polished thin with both granitic and doleritic clasts. The black sections and compared to 4 thin sections of Åsby check mark shows the location of the “original dolerite and 2 thin sections of Börön basalt that mafic ejecta”, where samples from the KM series were not affected by the impact cratering are taken as well. Sampling sites for AS10-7 and processes at Lockne, as well as 3 thin sections AS10-8 are also marked. Sample AS10-7 is taken from the granitic Tandsbyn Breccia that is most from a chunk of doleritic Tandsbyn Breccia that common in the Lockne crater structure. was found in the Lockne Breccia. Samples Å1, KM1, KM3, KM4, KM5, KM6, KM8, A geological map of the Kloxåsen area is B1, and B2 were collected and polished thin presented in Fig. 8. The mapped area covers sections were produced before the start of this approximately 1.5 km2. study. Units that were mapped in the field are the Petrographic microscopy was carried out at Cambrian alumshale, Ordovician Latorp lime- the Department of Geographical and Earth stone, Ordovician Töyen shale, Ordovician Sciences, University of Glasgow and at the “orthoceratite limestone”, Lockne Breccia, Ynn- Department of Earth Sciences, University of tjärnen Breccia, and Tandsbyn Breccia that con- Gothenburg. sists of granitic, doleritic, or volcanic material and a polymictic Tandsbyn Breccia of doleritic 2.3 Raman spectroscopy and granitic clasts. Thin sections of sample KM2 and KM3 were Where the Töyen shale has not been observed analysed with a Raman spectroscope. Raman in outcrop it has been drawn on the map simply spectra were obtained by operating a Renishaw as a line. InVia Raman Microscope, using a 514 nm edge laser that was used at 100% laser power. The 3.1.1 Pre-impact rocks spectral acquisition was centred on 830 Raman Underlying the Kloxåsen overthrust is the shift × cm-1 with 20 accumulations at an expo- crystalline Subcambrian Peneplain. The pene- sure time of 1.0 s. plain is superposed by approximately 30—50 m The Raman spectroscopy was performed at the of tectonised and thickened Cambrian alum- Department of Geographical and Earth Sciences, shale. University of Glasgow. On top of the Cambrian shale rest approxima- tely 2 m of Latorp limestone, about 5 m of Töyen 2.4 Scanning electron microscopy shale, and the “orthoceratite limestone”. X-ray microanalysis was carried out on thin The sequence of Ordovician rocks is not 50 m sections of samples KM2 and KM3 with a Carl thick, as it is outside of the Lockne crater. The Zeiss Sigma VP field-emission FE analytical limestone underlying the impact-related rocks is scanning electron microscope (SEM). Secondary approximately 30 m thick. electron (SE) and back-scatter electron (BSE) imaging modes were used for quantitative point 3.1.2 Impactites identification of mineral chemistry and chemical Ten ejecta blocks in total were observed in the element mapping. The SEM was operated with a Kloxåsen area consisting of Tandsbyn Breccia high vacuum at 20.00 kV. with varying composition, the largest of which SEM analyses were done at the Department of are approximately 200 m across. The ejecta Geographical and Earth Sciences, University of masses consist mainly of granitic Tandsbyn Glasgow. Breccia, although a considerable amount of doleritic Tandsbyn Breccia also occurs. A rare 3 Results polymictic Tandsbyn Breccia of both granitic and doleritic clasts occurs in one outcrop. 3.1 Distribution of the ejecta Another mafic Tandsbyn Breccia exists in the The Kloxåsen area is characterised by dense Kloxåsen area that probably is composed of forest and glacial sediment, nonetheless it was— dark, volcanic rock, however protolith deter- surprisingly—possible to observe 257 outcrops mination of this unit is beyond the scope of this in the area (see Fig. 7). Outcrops with doleritic study.

11 Figure 7. Outcrops in the Kloxåsen area observed and used to produce a geologic map. Outcrops with doleritic Tandsbyn Breccia are marked with a white check mark (✓) and the underlying portion of Tandsbyn Breccia that is coloured grey is mapped as doleritic. The grey check mark shows the location of a polymictic Tandsbyn Breccia with granitic and doleritic clasts. The black check mark shows the location of the “original mafic ejecta”, where samples marked KM are taken. Sampling sites for AS10-7 and AS10-8 are also marked.

12 Figure 8. Geologic map of the Kloxåsen area, based on outcrop coverage as presented in Fig. 7.

13 Figure 9. Poikilitic olivine in the unaltered Åsby dolerite. (XPL)

Figure 10. Sector zonation in a titanaugite oikocryst in the unaltered Åsby dolerite. (XPL) 14 1718

Lockne Breccia occurs mainly in the western altered by serecitisation. SEM-EDS analysis of part of the mapped area, flanking a mass of the plagioclase in the mafic Tandsbyn Breccia Tandsbyn Breccia. It was also found as breccia revealed signs of albitisation. sills in Ordovician limestone in the southwest. Titanaugite is found as unaltered oikocrysts in In-situ Ynntjärnen Breccia was found in some samples. The oikocrysts commonly show association with Tandsbyn Breccia and Lockne clear sector-zonation (see Fig. 11). In other Breccia. The limestone underlying the ejecta is samples the titanaugite is completely or almost commonly fractured into Ynntjärnen Breccia. completely replaced by a pseudomorph. Altera- Similarly the limestone that is injected with tion of this kind progresses inward from Lockne Breccia is often brecciated. fractures in the crystals (see Fig. 12), leaving The arenitic Loftarstone is not present in the isolated patches of titanaugite surrounded by Kloxåsen area. aggregates of alteration minerals on all sides No post-impact rocks are present in the (sieve structure), until the entire crystal has Kloxåsen area, thus the Lockne Breccia is been replaced (see Fig. 13, 14, 15, and 16). stratigraphically the youngest rock facies. Not a single crystal of olivine has been found in It was clear from observations in the field that any of the thin sections of doleritic Tandsbyn the sampled outcrop of mafic Tandsbyn Breccia Breccia that were analysed in this study. has a doleritic protolith, due to its textural Calcite is abundant in all clasts. It often occurs characteristics and mineralogy. in clusters of small, differently orientated Both the volcanic and polymictic Tandsbyn crystals with a poikilitic texture (see Fig. 13 and Breccia may be subject of future research. 14) or in veins. Calcite is normally found in association with serpentine and/or chlorite: 3.2 Mineralogy commonly a green rim of serpentine and Because a volcanic protolith for the mafic Tands- chlorite can be observed around calcite crystals byn Breccia that is subject of this study was that are not fully developed (see Fig. 17 and 18). ruled out during field work, the Börön basalt Calcite can also be found within titanaugite that is present in the Lockne area was excluded oikocrysts (see Fig. 12). from in-depth mineralogical studies. In some samples the calcite crystals with green rims of chlorite and serpentine are less common. 3.2.1 Åsby dolerite Instead pseudomorphs of serpentine (and minor The fine-grained Åsby dolerite from the chlorite) can be observed, not uncommonly with Jämtland suite outside the Lockne crater a small crystal of calcite at the centre (see Fig. consists mainly of euhedral plagioclase and 19). Other aggregates of calcite with associated titanaugite oikocrysts with an ophitic texture, serpentine give the impression that the serpen- and small olivine crystals with a poikilitic tine was gradually replaced by calcite (see Fig. texture (see Fig. 9). Many titanaugite oikocrysts 20; compared to Fig. 19). display sector-zonation (see Fig. 10). The pseudomorphs of calcite, chlorite, and The dolerite also contains lesser amounts of serpentine in the altered clasts have very similar apatite and ilmenite, often with titanium-rich textural characteristics to olivine in the unal- biotite rims. tered Åsby dolerite. Opaque minerals in the clasts are predomi- 3.2.2 Doleritic Tandsbyn Breccia nantly ilmenite crystals. Not uncommonly a Clasts in the doleritic breccia vary in size, shape, titanium-rich biotite rim can be found around and mineralogical composition. The clasts are ilmenite crystals. The titanium-rich biotite is generally fairly rounded, as opposed to the interpreted to be part of the primary mineralogy granitic Tandsbyn Breccia that is mainly and no titanium-poor biotite or suspected composed of angular clasts. Clasts in the secondary biotite has been observed. Opaque doleritic breccia vary more in size than clasts in minerals occur about as commonly in the clasts the granitic breccia and range from roughly 0.5 as they do in the matrix. cm to over 10 cm in diameter. Also in contrast Accessory minerals are mostly apatite, but also with the granitic Tandsbyn Breccia, in which minor occurrences of goethite, sphalerite, and clasts are generally single crystals, clasts in the pyrite have been observed. doleritic breccia consist of rock. The matrix is relatively uniform in minera- The clasts have an ophitic texture and consist logical composition. It consists of finely-crushed largely of fine-grained, euhedral plagioclase. material that generally is partly or wholly Plagioclase crystals are commonly partially replaced by chlorite (see Fig. 21). Smaller clast

15 Figure 11. Sector zonation in a titanaugite oikocryst in the doleritic Tandsbyn Breccia. (XPL)

Figure 12. Poikilitic calcite in a partially altered titanaugite oikocryst in the doleritic Tandsbyn Breccia. (XPL) 16 Figure 13. Hydrothermally replaced titanaugite oikocryst with calcite. (PPL)

Figure 14. Hydrothermally replaced titanaugite oikocryst with calcite. (XPL) 17 Figure 15. Hydrothermally replaced titanaugite oikocryst with sericitised plagioclase. (PPL)

Figure 16. Hydrothermally replaced titanaugite oikocryst with sericitised plagioclase. (XPL) 18 Figure 17. Pseudomorphs consisting of calcite surrounded by chlorite and serpentine. (PPL)

Figure 18. Pseudomorphs consisting of calcite surrounded by chlorite and serpentine. (XPL) 19 Figure 19. Pseudomorphs consisting mainly of serpentine with some calcite cores. (XPL)

Figure 20. Pseudomorphs consisting mainly of calcite with some serpentine rims. (XPL) 20 Figure 21. Matrix of the doleritic Tandsbyn Breccia. (XPL)

Figure 22. Calcite-rich matrix in AS10-8e. Note the calcite rim on the clast. (XPL) 21 Figure 23. Translucent sphalerite in calcite. (PPL)

Figure 24. Sphalerite in calcite. (PPL; reflected light) 22 Figure 25. Evidence of shock pressure: mechanical twinning in a titanaugite crystal. (XPL)

Figure 26. Vein with quartz edges and a calcite core. The matrix is almost completely altered to chlorite. (XPL) 23 Figure 27. View of a clast in KM3, consisting mainly of augite, plagioclase, and calcite. (SEM-EDS)

Figure 28. Pseudomorph after titanaugite in KM2, consisting of quartz, chlorite, anatase, and calcite. (SEM-EDS) 24 Figure 29. Raman spectra of anatase, apatite, calcite, and three spectra of the finely-crushed dolerite matrix. 25 Figure 30. Clast-matrix boundary in KM2 with signs of comminution. (SEM-EDS)

Figure 31. Microclast with an ilmenite core. Biotite has largely been replaced by quartz and anatase. (SEM-EDS) 26 Figure 32. Clast-matrix boundary featuring a fractured ilmenite crystal and veins. (SEM-EDS)

Figure 33. Detailed view of a vein from Fig. 32. (SEM-EDS)

27 2019

fragments have been observed in the matrix, ilmenite spectrum in the RRUFF database, which commonly containing a fractured titanaugite is a database under construction, did not match oikocryst or ilmenite crystal. Samples from the perfectly, however SEM-EDS analysis deter- AS10-8 series are more highly altered and also mined that the opaque minerals are ilmenite. contain small crystals of calcite in the matrix Interestingly and surprisingly, Raman analyses (see Fig. 22). of the finely-crushed dolerite matrix resulted in Many of the calcite crystals found in the clasts Raman spectra that are largely diagonal and of AS10-8 series abundantly contain small, have no distinct peaks (see Fig. 29 D), even with translucent sphalerite crystals (see Fig. 23 and variations in laser power (see Fig. 29 E and F). 24). With the SEM it was clear that the clast-matrix The degree of alteration varies in both the boundary is characterised by fractures and small clasts and the matrix. For instance, the matrix fragments of clast that have been chipped off shown in Fig. 21 contains fragments of titan- during the process of brecciation (see Fig. 30). augite, whereas the matrix shown in Fig. 22 A number of affected biotites has been contains none. Similarly, some samples are observed with a petrographic microscope, of characterised by pristine titanaugite oikocrysts which one was analysed with the SEM (see Fig. in the clasts (see Fig. 11) whereas other samples 31). SEM-EDS analysis revealed that quartz and do not contain any titanaugite (merely a titanium-rich mineral gradually had replaced pseudomorphs). There exists a direct corre- the biotite rim, although it remains unclear what lation between the degree of alteration in the “triggered” this type of alteration. Note also that clasts and matrix: samples that still inhibit next to the altered biotite is pristine biotite titanaugite oikocrysts also contain fragments of (clearest in Fig. 31 E). Raman spectroscopy those in the adjacent matrix. determined that the titanium mineral is anatase. There are not many signs of shock meta- SEM-EDS analysis of veins in the doleritic morphism in the thin sections that were Tandsbyn Breccia offered a more detailed view analysed. One titanaugite crystal has been of the bimodal mineralogy observed in the observed with mechanical twinning (see Fig. petrographic microscope (see Fig. 32 and 33; 25). No kinkbanding was found in the biotite, compare with Fig. 26). Fig. 33 shows a more however biotite crystals that are found in the detailed picture of a vein in Fig. 32. It is matrix are commonly “split open”, but it is important to note the thin layer of quartz on the unclear whether this is a result of shock sides of the vein (see Fig. 33 E) and the good fit pressure or brecciation and subsequent hydro- between quartz on opposite sides of the vein thermal processes. (see Fig. 33 F). Veins are common throughout the mafic Tandsbyn Breccia, mainly occurring in the 4 Discussion matrix but also cutting through individual clasts. Protolith determination for the mafic Tandsbyn Veins are generally “lined” with quartz on the Breccia could almost be done in the field, where sides, with a calcite “core” that may sometimes a volcanic source rock could be excluded due to contain fragments of quartz (see Fig. 26). a doleritic texture and mineralogy. The only SEM-EDS analysis of thin section KM3 confir- dolerite known in the Lockne area is the Åsby med that the clasts that are not highly altered olivine dolerite. Also, a few outcrops of are mainly composed of plagioclase, titanaugite, Tandsbyn Breccia were observed with a and calcite (see Fig. 27), with lesser occurrences probably-volcanic protolith, which appeared of apatite, ilmenite, and biotite. distinctly differently than the mafic Tandsbyn Analysis of section KM2 revealed that the Breccia that is the subject of this thesis. pseudomorphs that have replaced the titan- In addition, optical petrography revealed that augite oikocrysts in the clasts are a mixture of the titanaugite oikocrysts in the mafic Tandsbyn quartz, chlorite, calcite, and a titanium oxide Breccia display clear sector zonation (see Fig. (see Fig. 28). Raman spectroscopy of the same 11). This feature is only known to occur in sample indicates that this titanium oxide is clinopyroxene oikocrysts in rocks of the Central anatase (see Fig. 29 A). Scandinavian Dolerite Group (CSDG) (Claeson, Raman spectroscopy was utilised with Meurer, Hogmalm, & Larson, 2007), to which the satisfaction on various minerals, such as apatite Åsby dolerite belongs (Gorbatschev, Solyom, & and calcite (see Fig. 29 B and C), however it was Johansson, 1979). not possible to confirm, for instance, ilmenite. Previously it has been hypothesised that the Obtaining a spectrum was possible, but the explanation for the rounded clasts in the

28 2122

doleritic Tandsbyn Breccia is that the clasts are Tandsbyn Breccia has not been affected by core stones from deep-weathered Åsby dolerite regional metamorphism, which indicates to me a sills that were part of the Cambrian bottom minimal Caledonian influence. gravel (E. Sturkell, personal communication, July In addition, optical petrography of the Åsby 18, 2010). However, optical petrography of the olivine dolerite from just outside the Lockne breccia clasts shows no signs of deep weathering crater is not hydrothermally altered or meta- that could account for rounded clasts. Rather, morphosed (see Fig. 9 and 10), and still contains signs of mechanical crushing and rounding are pristine olivine, whereas the Åsby dolerite from observed (see Fig. 30). I hypothesise that the the Tandsbyn Breccia in Kloxåsen does not reason that the doleritic Tandsbyn Breccia clasts contain any olivine. are not as angular as in the granitic breccia is An interesting feature of the Raman spectra that the Åsby dolerite is more finely grained. In that were obtained from the finely-crushed the granitic Tandsbyn Breccia clasts are usually dolerite matrix is the diagonality and the monomineralic, due to the protolith granite absence of distinct peaks. Collection of Raman being fairly coarse grained. During brecciation spectra with an unfocused microscope results in mineral boundaries are relatively weak links in a diagonal spectrum that increases in intensity the rock and it will tend to break along those with higher Raman shift, with no peaks. I boundaries. Thus, clasts in the granitic breccia hypothesise that the similarity of the Raman will tend to be single, angular crystals, whereas spectra of the finely-crushed dolerite to an the Åsby dolerite is too fine grained to produce unfocused Raman spectrum is due to the high monomineralic clasts and instead clasts are degree of crushing. The dolerite matrix may rounded by comminution, which also produces have been so finely crushed during brecciation the fine-grained matrix. that the obtained Raman spectra appear The observed polymictic Tandsbyn Breccia, unfocused. composed of both granitic and doleritic clasts, is situated between a mass of granitic breccia on In conclusion, with respect to all the available one side and doleritic breccia on the other. Due data, my model for the petrogenesis of the to the relatively authigenic character of the doleritic Tandsbyn Breccia at Kloxåsen is as Tandsbyn Breccia, I have interpretated this follows. polymictic variant as a brecciated contact zone Sills of Proterozoic Åsby olivine dolerite were between granite and a dolerite sill in the original present in the crystalline basement rocks of the target bedrock. Because of the relatively rare target of the Lockne impact crater. Upon impact occurrence of this polymictic, crystalline ejecta these rocks were exposed to at least 5 GPa, and its association to the other types of Tands- following evidence from shocked titanaugite byn Breccia, this polymict does not transform (see Fig. 25). Shock pressure caused fracturing the Tandsbyn Breccia into an allogenic breccia. If of olivine as well. The rocks were authigenically it were an allogenic breccia, the polymictic brecciated and subsequently ejected during the Tandsbyn Breccia should be more common and process of crater excavation. occur homogeneously. The ejecta blocks crashed down into the The hydrothermal system that caused the Palaeozoic seabed and created a small, secon- alteration in Kloxåsen has the same signature as dary crater, based on approximately 20 m of elsewhere in the Lockne crater, as determined missing Ordovician limestone underlying the by fluid-inclusion analysis in calcite from ejecta at Kloxåsen. Kloxåsen by M. Ivarsson and C. Broman (perso- After the resurge of water into the collapsing nal communication, October 4, 2010). This transient crater, heat from the impact and strongly suggests that the observed hydrother- brecciation that was present in the ejecta mal alteration in the doleritic Tandsbyn Breccia initiated a hydrothermal system that was part of was caused by the aftermath of the Lockne the hydrothermal circulation that reigned in the impact event and not the Caledonian orogeny. entire Lockne crater after the impact, based on It could be possible that the extensive evidence from fluid inclusions at Kloxåsen and chloritisation in the finely-crushed dolerite elsewhere in the crater. Hot water circulated matrix of is an effect from the Caledonian through cracks in the ejecta breccia. orogeny, however, I have only observed one The fractured olivine was rapidly replaced by a generation of veining. If the Caledonian orogeny pseudomorph of serpentine (see Fig. 19), which triggered hydrothermal alteration, it left no is an exothermal reaction—and thus resulted in cross-cutting textures. Additionally, the mafic more heat—that produces silica gel. The hot

29 2423

silica was deposited in fractures as quartz veins, shock metamorphism has been observed in the based on observations of veins that are lined clasts. I believe that this is also partly due to the with quartz where—importantly—the opposite extensive hydrothermal alteration of the mafic sides fit (see Fig. 33). Tandsbyn Breccia, which has erased most of the The hydrothermal circulation dissolved calcite signs of shock metamorphism through subse- from the Ordovician limestone, which is quent replacement of titanaugite oikocrysts and abundantly available in the Kloxåsen area. almost-complete chloritisation of the matrix. Calcite was hydrothermally redeposited as cores in the freshly-made quartz veins (see Fig. 26, 32, 5 Conclusion and 33) and began to gradually replace the Field mapping revealed a larger abundance of serpentine pseudomorphs (see Fig. 17, 18, and mafic impact ejecta in the Kloxåsen area than 20), until they were no more than rims on the was previously known or anticipated. The newly-formed calcite or were replaced entirely granitic Tandsbyn Breccia still dominates in the (see Fig. 13, 14, and 27). The replacement of ejecta. serpentine—a silicate mineral—likely also has Optical petrography, scanning electron contributed to the formation of quartz veins. microscopy, and Raman spectroscopy revealed In parts of the ejecta blocks where the strong hydrothermal alteration. Some biotites hydrothermal circulation was more intense appear “split open”, but no kinkbanding is titanaugite was beginning to be replaced by a observed. Titanaugite affected by mechanical pseudomorph of quartz, chlorite, calcite, and twinning indicates a shock pressure of at least 5 anatase, spreading from fractures in the crystals GPa in the ejecta at Kloxåsen. (see Fig. 12), until entire titanaugite oikocrysts Textural and mineralogical comparison with were replaced by a pseudomorph (see 13, 14, mafic rocks from the vicinity of the Lockne 15, 16, and 28). impact structure strongly suggest the Åsby The matrix adjacent to clasts that contain dolerite as a protolith. Both the doleritic unaltered titanaugite also contains fragments of Tandsbyn Breccia and the Åsby dolerite host titanaugite (see Fig. 21 and 25) and vice versa sector-zoned titanaugite oikocrysts, which are for matrix adjacent to clasts with replaced only known to occur in CSDG rocks. titanaugite. Due to the movement involved in the Hydrothermal replacement of olivine resulted brecciation process, this means that the in pseudomorphs of serpentine that in turn are alteration of titanaugite also is hydrothermal wholly or partly replaced by calcite (and and occurred heterogeneously in the ejecta translucent sphalerite) and chlorite. Titanaugite depending on the local intensity of the hydro- was gradually replaced by aggregates of quartz, thermal circulation. chlorite, calcite, and anatase. Unaffected Chloritisation of the finely-crushed dolerite titanaugite does still occur in some clasts and matrix and of serpentine is accounted to the adjacent matrix, whereas all olivine has been hydrothermal system in the aftermath of the completely replaced. Lockne impact event. Samples that are the most altered contain 6 Acknowledgements calcite pseudomorphs that are rich in trans- I would like to thank Paula Lindgren, Erik Sturkell, and lucent sphalerite (see Fig. 23 and 24). In addi- Martin Lee for their supervision and making this project possible to begin with. Jens Ormö, Gabrielle Stockmann, tion, the matrix is rich in small crystals of calcite Irene Melero Asensio, Joakim Mansfeld, and Reinhard (see Fig. 22) and the clasts contain more calcite Greiling made the field work in Jämtland exceptionally than the absence of olivine can account for. pleasant. Peter Chung at Glasgow University is thanked for Plagioclase in the doleritic Tandsbyn Breccia is his technical support. John Gilleece at Glasgow University, more affected by sericitisation than in the Åsby Ali Firoozan at Gothenburg University, and Kjell Hälge stood for the production of thin sections. Johan Hogmalm dolerite that was unaffected by the Lockne is thanked for technical support and challenging impact event and subsequent alteration, which I discussions that helped the thesis forward. Similarly, Eric believe is due to hydrothermal sericitisation Fair was always willing to lend his attention when I wanted after the impact. Plagioclase in the mafic to discuss hypotheses. Magnus Ivarsson and Curt Broman are thanked for their contribution with results of fluid- Tandsbyn Breccia has also been slightly albitised inclusion analysis from Kloxåsen. Lennart Björklund is (see Fig. 28D). thanked for examining this thesis project. Elin Ekman is The grade of shock is low in the doleritic thanked for opponing me in this challenging thesis. This Tandsbyn Breccia—and in the Tandsbyn Breccia thesis project was mainly funded with a grant from the in general—and only one shock-metamorphosed Royal Astronomical Society. crystal in the matrix has been observed; no

30 12

Science 12A (P.H. Schultz and R.B. Merrill, 7 References eds.), pp. 37-57. Pergamon, New York. Andrews-Hanna, J.C., Zuber, M.T., & Banerdt, Hagerty, J.J., & Newsom, H.E. (2003). W.B. (2008). The Borealis basin and the Hydrothermal alteration at the Lonar Lake origin of the martian crustal dichotomy. impact structure, India: Implications for Nature, 453, 1212-1215. impact cratering on Mars. Meteoritics & Bronshten, V.A. (2000c). Tungusskiy Meteorit: Planetary Science, 38, 365-381. Istoria Issledovaniya. Selyanov AD, Moskva. Högdahl, K., Gromet, L.P., Claesson, S. (1996). Claeson, D.T., Meurer, W.P., Hogmalm, K.J., & The geochronology of a ductile shear zone in Larson, S.-Å. (2007). Using LA-ICPMS Jämtland, central Sweden. Current Problems, Mapping and Sector Zonation to Understand Ideas, and Results in Geology. 16-18 October, Growth and Trace-Element Partitioning in Stockholm, GFF 118, Jubilee Issue, pp. A14- Sector-Zoned Clinopyroxene Oikocrysts from A15. the Norra Ulvö Gabbro, Sweden. Journal of Kamo, S.L., Reimold, W.U., Krogh, T.E., & Petrology, 48, 711-728. Colliston, W.P. (1996). A 2.023 Ga age for the Dence, M.R. (1968). Shock zoning at Canadian Vredefort impact event and a first report of craters: Petrography and structural shock metamorphosed zircons in implications. In Shock Metamorphism of pseudotachylitic and Granophyre. Natural Materials (B.M. French and N.M. Earth and Planetary Science Letters, 144, 369- Short, eds.), pp. 169-184. Mono Book Corp., 387. Baltimore. Kieffer, S.W., & Simonds, C.H. (1980). The role of Elliot, B. (n.d.) Earth Impact Database. Retrieved volatiles and lithology in the impact cratering from process. Reviews of Geophysics and Space http://www.passc.net/EarthImpactData- Physics, 18, 143-181. base/index.html Krinov, E.L. (1966). Giant . Oxford, NY: Farinella, P., Foschini, L., Froeschlé, Ch., Gonczi, Pergamon Press. R., Jopek, T.J., Longo, G., & Michel, P. (2001). Lindström, M. (1971). Vom Anfang, Hochstand Probable asteroidal origin of the Tunguska und Ende eines Epikontinentalmeeres. Cosmic Body. Astronomy & Astrophysics, 377, Geologische Rundschau, 60, 419-438. 1081-1097. Lindström, M., & Sturkell, E.F.F. (1992). Geology French, B.M. (1998). : A of the Early Paleozoic Lockne impact Handbook of Shock-Metamorphic Effects in structure, central Sweden. Tectonophysics, Terrestrial Meteorite Impact Structures. 216, 169-185. Houston, TX: Lunar and Planetary Institute. Lindström, M., Simon, S., Paul, B., & Kessler, K. Gorbatschev, R., Solyom, Z., & Johansson, I. (1983). The Ordovician and its mass (1979). The Central Scandinavian Dolerite movements in the Lockne area near the Group in Jämtland, central Sweden. GFF, 101, Caledonian margin, central Sweden. 177-190. Geologica et Palaeontologica, 17, 17-27. Grieve, R.A.F. (1991). Terrestrial impact: the Lindström, M., Sturkell, E.F.F., Törnberg, R., record in the rocks. Meteoritics, 26, 175-194. Ormö, J. (1996). The marine impact crater at Grieve, R.A.F., & Pilkington, M. (1996). The Lockne, central Sweden. GFF, 118, 193-206. signature of terrestrial impacts. AGSO Journal Lindström, M., Ormö, J., Sturkell, E., & von of Australian Geology and Geophysics, 16, 399- Dalwigk, I. (2005). The Lockne Crater: 420. Revision and Reassessment of Structure and Grieve, R.A.F., Dence, M.R., & Robertson, P.B. Impact Stratigraphy. In Impact tectonics (C. (1977). Cratering process: As interpreted Koeberl & H. Henkel, eds.), pp. 357-388, from the occurrence of impact melts. In Berlin, Springer. Impact and Explosion Cratering: Planetary Mansfeld, J., Sturkell, E.F.F., Broman, C. (1998). and Terrestrial Implications (D.J. Roddy, R.O. Meteorite impact influence on a Proterozoic Pepin, and R.B. Merrill, eds.), pp. 791-814. shear zone mineralization at Gullgruvan, Pergamon, New York. Jämtland, Sweden. Abstract in the 23rd Nordic Grieve, R.A.F., Robertson, P.B., Dence, M.R. Geological Winter meeting. Aarhus, Denmark, (1981). Constraint on the formation of ring p. 195. impact structures, based on terrestrial data. Mashchak, M.S., & Naumov, M.V. (1996). The In Multi-Ring Basins: Formation and Suavjärvi structure: An early proterozoic Evolution, Proceedings of Lunar and Planetary impact site on the Fennoscandian shield.

31 43

Lunar and Planetary Science, 27, 825-826. anomaly in the Middle Ordovician Lockne Melosh, H.J. (1989). Impact cratering: A impact structure, Jämtland, Sweden. GFF, 120, geological process. New York, NY: Oxford 333-336. University Press. Sturkell, E.F.F. (1998c). Resurge morphology of O’Keefe, J.D., & Ahrens, T.J. (1982). Cometary and the marine Lockne impact crater, Jämtland, meteorite swarm impact on planetary central Sweden. Geological Magazine, 135, surfaces. Journal of Geophysical Research, 87, 121-127. 6668-6680. Sturkell, E.F.F., Broman, C., Forsberg, P., & O’Keefe, J.D., & Ahrens, T.J. (1993). Planetary Torssander, P. (1998). Impact-related cratering mechanics. Journal of Geophysical hydrothermal activity in the Lockne impact Research, 98, 17011-17028. structure, Jämtland, Sweden. European Ormö, J., Shuvalov, V.V., Lindström, M. (2002). Journal of Mineralogy, 10, 589-606. Numerical modeling for target water depth Sturkell, E., & Lindström, M. (2004). The target estimation of marine-target impact craters. peneplain of the Lockne impact. Meteoritics Journal of Geophysical Research, 107, 31-39. and Planetary Science, 39, 1721-1731. Patchett, P.J. (1978). Rb/Sr ages of Precambrian Sturkell, E., Ormö, J., Nõlvak, J., & Wallin, Å., dolerites and syenites in southern and central 2000: Distant ejecta from the Lockne marine- Sweden. Sveriges Geologiska Undersökning C target impact crater, Sweden. Meteoritics & 747, 1-63. Planetary Science, 35, 929-936. Phillips, F. M., Zreda, M.G., Smith, S.S., Elmore, D., Taylor, S.R. (1992). Solar System Evolution: A Kubik, P.W., Dorn, R.I., & Roddy, D.J. (1991). New Perspective. New York, NY: Cambridge Age and geomorphic history of Meteor Crater, University. Arizona, from cosmogenic 36Cl and 14C in Thorslund, P. (1940). On the Chasmops Series of rock varnish. Geochimica et Cosmochimica Jemtland and Södermanland (Tvären). Acta, 55, 2695-2698. Sveriges Geologiska Undersökning C 436, 1- Rajmon, D. (2009) Impact database 2010.1. 191. Online: http://impacts.rajmon.cz Trayner, C. (1997). The Tunguska event. Journal Roesler, W. (2009) Alz. In Rajmon, D. (2009) of the British Astronomical Association, 107, Impact database 2010.1. Online: 117-130. http://impacts.rajmon.cz Vasilyev, N.V. (1998). The Tunguska meteorite Simon, S. (1987a). Stratigraphie, Petrographie problem today. Planetary and Space Science, und Entstehungsbedingungen von 46, 129-150. Grobklastika in der autochthonen, Wickman, F.E. (1988). Possible impact ordovizischen Schichtenfolge Jämtlands structures in Sweden. In Deep drilling in (Schweden). Sveriges Geologiska crystalline bedrock. I. The deep gas drilling in Undersökning C 815, 1-156. the Siljan impact structure, Sweden and Schulte, P., Alegret, L., Arenillas, I., Arz, J.A., astroblemes (A. Bodén and K.G. Eriksson, Barton, P.J., Bown, P.R., Bralower, T.J., eds.), pp. 298-327, Springer Verlag, Berlin Christeson, G.L., Claeys, P., Cockell, C.S., Heidelberg. Collins, G.S., Deutsch, A., Goldin, T.J., Goto, K., Wiman, C. (1900). Eine untersilurische Grajales-Nishimura, J.M., Grieve, R.A.F., Litoralfacies bei Locknesjön in Jemtland. Gulick, S.P.S., Johnson, K.R., Kiessling, W., Bulletin of the Geological Institution of the Koeberl, C., Kring, D.A., MacLeod, K.G., Matsui, University of Uppsala, 4, 133-151. T., Melosh, J., Montanari, A., Morgan, J.V., Neal, C.R., Nichols, D.J., Norris, R.D., Pierazzo, E., Ravizza, G., Rebolledo-Vieyra, M., Reimold, W.U., Robin, E., Salge, T., Speijer, R.P., Sweet, A.R., Urrutia-Fucugauchi, J., Vajda, V., Whalen, M.T., & Willumsen, P.S. (2010). The Chicxulub Impact and Mass Extinction at the Cretaceous-Paleogene Boundary. Science, 327, 1214-1218. Sturkell, E.F.F. (1998a). The marine Lockne impact structure, Jämtland, Sweden: a review. Geologische Rundschau, 87, 253-267. Sturkell, E.F.F. (1998b). Impact-related Ir

32