Shock barometry of the Siljan impact structure, Sweden


Corresponding author. E-mail:


Abstract– The Siljan impact structure in Sweden is the largest confirmed impact structure in Western Europe. Despite this, the structure has been poorly studied in the past, and detailed studies of shock metamorphic features in the target lithologies are missing. Here, we present the results of a detailed systematic search for shock metamorphic features in quartz grains from 73 sampled localities at Siljan. At 21 localities from an area approximately 20 km in diameter located centrally in the structure, the orientations of 2851 planar deformation feature sets in 1179 quartz grains were measured. Observations of shatter cones outside of the zone with shocked quartz extend the total shocked area to approximately 30 km in diameter. The most strongly shocked samples, recording pressures of up to 20 GPa, occur at the very central part of the structure, and locally in these samples, higher pressures causing melting conditions in the affected rocks were reached. Pressures recorded in the studied samples decrease outwards from the center of the structure, forming roughly circular envelopes around the proposed shock center. Based on the distribution pattern of shocked quartz at Siljan, the original transient cavity can be estimated at approximately 32–38 km in diameter. After correcting for erosion, we conclude that the original rim to rim diameter of the Siljan crater was somewhere in the size range 50–90 km.


The Siljan structure in south-central Sweden (centered at N61°02.196′; E014°54.467′; Fig. 1) is the largest verified impact structure in Western Europe with a commonly quoted size of 52 km (Grieve 1982, 1988). An impact origin for Siljan was suggested by Wickman et al. (1963) and Fredriksson and Wickman (1963) based on morphological features of the structure. Later, planar deformation features (PDFs) in quartz grains and observations of shatter cones gave unambiguous evidence that the structure is in fact impact-derived (Svensson 1971, 1973). Siljan is poorly studied despite its size, and basic facts regarding, for example, the original and present size of the structure remain elusive. Investigations about the distribution of shocked quartz and shock barometry at Siljan are also lacking. Some information about shocked quartz can be retrieved from publications based on few data, and limited data presentation (Svensson 1971, 1973; Tamminen and Wickman 1980; Åberg and Bollmark 1985; see also Collini 1988; Juhlin et al. 1991).

Figure 1.

 Geologic map of the Siljan impact structure (after Kresten and Aaro 1987; Kresten et al. 1991). Localities where samples were taken are indicated with dots. The annular depression is defined by the occurrence of Paleozoic sediments and lakes. The inset shows the location of Siljan in Scandinavia. TIB = Trans-Scandinavian Igneous Belt.

Previous studies of the shock level distribution at impact structures based on field studies and laboratory shock experiments have shown that PDFs oriented along different crystallographic planes develop at different shock pressures (e.g., Hörz 1968; Müller and Défourneaux 1968; Robertson et al. 1968; Grieve and Robertson 1976; Huffman and Reimold 1996). Rocks from impact structures can therefore be divided into different stages of shock metamorphism and be assigned specific shock pressures, defined by the PDF orientations observed in different samples (Robertson et al. 1968; Robertson 1975; Grieve and Robertson 1976). A method based on these early studies (an overview of this method is provided below) has been used for shock barometry studies of, for example, the Charlevoix, Slate Islands, and Manicouagan impact structures (e.g., Robertson 1975; Robertson and Grieve 1977; Dressler et al. 1998).

The present study is the first detailed characterization of the shock level distribution across the Siljan impact structure by studying shock deformation features in quartz grains. Shocked quartz is essential in investigations of shock distribution at impact structures because of the well-documented effects of shock metamorphism on quartz, the simple optical features, the stability of the mineral, and the frequent occurrences in continental rocks. Quartz is particularly useful in shock barometry studies because it develops shock effects over a wide pressure range. Here, we investigate the PDF distribution pattern to determine where the shock center of Siljan is located, and what shape the pattern of lateral shock attenuation has, to discuss whether these features are comparable with other terrestrial complex impact structures. As Siljan is deeply eroded, the distribution of shocked minerals might reveal new information about the features of complex impact structures at depth. With these data, we hope also to better constrain the original dimensions and form of Siljan.

Constraining the original size of Siljan has become even more relevant as the age of the structure was revised from 368±1.1 Ma (Bottomley et al. 1990) to 377±2 Ma by Reimold et al. (2005). This age corresponds within error to the newly revised age of the Frasnian/Famennian boundary (374.5±2.6 Ma; House and Gradstein 2004), and the associated extinction event, as discussed in Reimold et al. (2005). To date, there is no evidence for a connection between the formation of the Siljan structure and the Late Devonian extinction event, or even a connection between any other large impact event and a major mass extinction other than at the Cretaceous/Paleogene boundary (Alvarez 2003). However, the combination of several large impact events (see Earth Impact Database for details on structures with similar ages as Siljan) might be capable of disturbing the ecosystem (Rampino 2002 and references therein; Schmitz et al. 2008).

Geological Setting and Previous Studies

The Siljan impact structure is located in the boundary zone between felsic intrusives that belong to the Trans-Scandinavian Igneous Belt (intrusion ages 1.85–1.65 Ga) in the west, and Svecofennian rocks of older ages to the east (Fig. 1) (2.1–1.87 Ga; Högdahl et al. 2004 and references therein). For details on the regional geology of the area, the reader is referred to Kresten and Aaro (1987), Kresten et al. (1991), and Hjelmqvist (1966). The structure itself consists of a central area with a diameter of 28–30 km. It is dominated by Dala granites of Järna and Siljan types, but sporadically, mafic intrusive, extrusive magmatic, and sedimentary rocks also occur. Ordovician and Silurian sediments, although completely eroded in the surrounding area, have been preserved in a 10 km wide partly lake-filled annular depression that surrounds the central area. The sedimentary sequence that is preserved in the annular moat has a thickness of up to 350 m and largely constitutes a megabreccia of meter- to kilometer-sized blocks (Collini 1988). Core drillings in the Paleozoic ring revealed that the sedimentary sequence is reversed in some places, with Ordovician units resting on top of upper Silurian sandstone (Collini 1988). At the time of impact, the sedimentary sequence that overlaid the crystalline basement is commonly quoted to have been 400–500 m thick (e.g., Rondot 1975; Lindström et al. 1991). However, a case for a significantly thicker sedimentary sequence has been made based on studies of thermal indicators, such as δ18O/δ13C, conodont alteration indices and fission track data, indicating raised temperatures of as much as 125 °C (Tullborg et al. 1995). This temperature increase is attributed to burial by a much thicker sedimentary cover of approximately 2–4 km at the approximate time of impact, composed of erosion products from the Caledonides (Tullborg et al. 1995; Larson et al. 1999; Cederbom et al. 2000; Cederbom 2001).

The present day Siljan impact structure is deeply eroded. Exposed at the present surface are rocks that were originally located beneath the crater floor (Grieve 1988). Grieve (1988) concludes that a minimum estimate of the amount of erosion at the center of Siljan is 1 km, as the structure lacks both a topographic feature, such as a peak and/or ring(s) and a melt sheet. The rim area is estimated to have been eroded at least 1.5 km (Grieve 1988).

The diameter of 52 km for Siljan (Grieve 1982, 1988) essentially corresponds to the diameter of the outer limit of Paleozoic sediments. A larger diameter estimate of 65 km was suggested by Kenkmann and von Dalwigk (2000) based on structural investigations, especially on detailed investigation of the degree of fracturing of the basement surrounding the central area. Henkel and Aaro (2005) suggest a present diameter of 75 km based on topographic features surrounding the structure, and argue that the original diameter of Siljan could have been as large as 85 km.

The evidence for an impact origin for Siljan includes both PDFs in quartz and shatter cones. Shatter cones were first described from only a few localities (Svensson 1973), but now they are known from a large portion of the central uplift, although their occurrence is hard to map in detail due to extremely poor exposure. Based on the orientation of axes of shatter cones from four localities, Wickman (1980) determined the location of the impact center at Siljan to approximately 5 km north of the geographical center. Grieve (1988) later commented on this by saying that the data set used by Wickman (1980) was limited and that significant movement among blocks during central uplift formation disturbs the original shatter cone pattern. Aside from this (mega)brecciation of the Dala granites constituting the central part of Siljan, in situ melt breccias or pseudotachylitic breccias are rare (e.g., Rondot 1975; Bottomley et al. 1978; Reimold et al. 2005). In situ true melt rock has not yet been found, probably due to the deep erosion level of the structure.

Siljan was deeply and extensively drilled when the Deep Gas Drilling project was initiated in 1982 (Bodén and Eriksson 1988; Juhlin et al. 1991), following the proposal that fractures below the structure could contain large amounts of trapped hydrocarbons derived from the mantle (Gold and Soter 1980). However, hydrocarbons of economic interest were not detected.

Material and Methods

Sample Collection

For this study, exposed bedrock from the Siljan impact structure was sampled. The sampled lithologies represent granites and granodiorites of Järna and Siljan types, occasional mafic rock types (diorite and diabase), and sedimentary rocks. The granitic samples are medium to coarse-grained, of red (occasionally gray) color, and often display porphyritic texture. The mineral assemblage is dominated by potassium feldspar, plagioclase feldspar, quartz, hornblende, and biotite. Samples were taken only at outcrops where the in situ nature could be confirmed, across predominantly the inner part of the structure. Occurrences of shatter cones were noted, but there was no systematic mapping to characterize the detailed distribution of them across the structure.

Petrographic Analysis and U-Stage Measurements

In total, 300 thin sections from 73 localities were prepared. The thin sections were first characterized based on mineral assemblage and searched for shock metamorphic features under an optical microscope. A more detailed study of thin sections showing shock metamorphic features, such as planar fractures (PFs) and/or PDFs in quartz, was performed with the use of a Leitz five-axes universal stage (U-stage; Emmons 1943) following the techniques described in Engelhardt and Bertsch (1969), Stöffler and Langenhorst (1994), and Ferrière et al. (2009a). The orientations of optic axes (c-axes) and poles perpendicular to PDF planes in individual quartz grains were determined and then indexed using a stereographic projection template. In this study, the updated version of the stereographic projection template that was presented by Ferrière et al. (2009a), which includes five additional orientations of PDF planes, was used. One of these orientations, inline image, with a polar angle to the c-axis of 17.62°, is overlapping the inline image plane (polar angle to c-axis 22.95°) within the 5° measurement error envelope. All PDFs with orientations that fall in the overlapping zone between these two were treated as inline image in this study, as recommended by Ferrière et al. (2009a). Ferrière et al. (2009a) concluded based on a detailed study of reproducibility and accuracy of U-stage measurements of PDFs, that to reach an acceptable level of precision, at least 50 PDF sets per sample should be measured. Therefore, in this study, we aimed at measuring PDFs in 60 grains per sample, to guarantee that samples only averaging one set per grain also reach an acceptable level of precision. This was achieved for all samples with shocked quartz grains, with the exception of samples 20, 35, and 71 due to scarcity of shocked grains and/or large grain size. The amount of quartz grains displaying PFs and/or PDFs in relation to the unshocked quartz grains in each sample was estimated by point counting.

The indexing of PDFs was done by hand and polar angles between PDFs, and the c-axis of grains were also checked with the help of a computer program (Stereo32, available online at to guarantee precise angular relations.

All percentage calculations represent absolute frequencies, as defined by Engelhardt and Bertsch (1969), meaning that displayed percentages are calculated as the number of symmetrically equivalent planes measured in n quartz grains, divided by the total number of measured PDF sets in n quartz grains.

Shock Barometry

Previous Shock Barometry Methods

When dividing samples into groups based on PDF observations, we have decided not to use the division described by Robertson et al. (1968), Robertson (1975), and Grieve and Robertson (1976). Robertson et al. (1968) recognized five stages of shock metamorphism in quartz grains from 12 Canadian impact structures, termed type A–E. Type A quartz grains were characterized by development of PDFs exclusively oriented parallel to the basal plane, whereas in type B, PDFs oriented parallel to inline image occur together with basal PDFs. Basal PDFs are rare in type C, where sets oriented parallel to inline image occur together with those parallel to inline image. In type D and E quartz grains, PDFs oriented parallel to inline image occur together with other sets of which inline image-oriented ones are the most common. Type E quartz grains are partly isotropic. Grieve and Robertson (1976) assigned shock pressures for the onset of development of PDFs characterizing types A–D to: 7.5 GPa for type A, 10 GPa for type B, 14 GPa for type C, and 16 GPa for type D. Robertson (1975) included a ±3 GPa error margin for each of these pressure estimates. Individual average pressure values for each shocked quartz grain, and then in turn each sample, based on the proportions of different PDFs were then assigned. This was done with a precision of ±0.1 GPa, although the experimental data that this builds on had an accuracy that was far less precise (Hörz 1968; Müller and Défourneaux 1968). Furthermore, the method does not take into account less common high-angle planes, such as inline image, inline image, and inline image. It remains unclear, for example, how to group a grain in which two sets of PDFs were observed, and where these sets are oriented parallel to inline image and inline image, respectively (see also discussion in Ferrière et al. 2008). In addition, the number of PDF sets/grain is only to a limited degree considered, even though the number of PDF sets per grain is strongly indicative of pressure (e.g., Dence 1968; Robertson et al. 1968; Grieve et al. 1996; Huffman and Reimold 1996; Ferrière et al. 2008).

Assigning Shock Pressures to the Siljan Samples

Before assigning shock pressures to the samples that display shock metamorphic features, they were divided into groups. When grouping the samples with observations of PDFs, we primarily considered the number of sets/grain, and to a limited degree the general orientation pattern for the PDF sets. Samples lacking PFs or PDFs were grouped if shatter cones had been observed at these localities in the field. Localities where previous authors (Kresten et al. 1991) had observed shatter cones were also included.

Samples from the different groups were then collectively assigned a specific range of shock pressures that represents that group (see details below). The pressure estimates are based primarily on the specific orientation of the population of PDFs in the different samples, and on other shock metamorphic features (shatter cones).


All 73 sampled localities are displayed in Fig. 1 and are briefly described in Appendix 1. Of these, samples collected at 21 localities display microdeformation features attributed to shock metamorphism (PFs and PDFs; Table 1). Another five localities are macroscopically shocked, i.e., they display shatter cones, but no PFs or PDFs were observed in thin sections prepared from these samples. Outside of the annular depression no shock metamorphic features were observed. The 26 localities where shock effects were detected are restricted to the central part of Siljan (Fig. 2). We have divided them into four groups (A–D; see Table 2).

Table 1.   U-stage data for planar deformation feature (PDF) set abundances and crystallographic orientations of these sets, as measured in quartz grains from the Siljan structure.
  1. Notes: n.d. = none detected. In total, 2851 sets in 1179 grains were measured. Note that also the percentage of shocked quartz grains in each sample is displayed.

  2. aExcluding unindexed sets.

  3. bSee text and Engelhardt and Bertsch (1969) for description of method.

  4. cPDFs with measured orientations that plot in the overlapping zone between inline image and inline image were treated as inline image. See text and Ferrière et al. (2009a) for details.

No. investigated grains6017606046060606060606060606060606060606018
Shocked quartz grains (%)100101009010100100100100649010010090819110010010010010030
No. measured sets17617916941991361071836060782231669711316622922221821918
No. measured setsa17117856741911301071776058782171599511216322121420921718
No. PDF sets/grain (N)
No. PDF sets/grain (N′)a2.
Number of PDF sets/grain (%)
 1 set16.710063.388.31006.726.760.015.0100.010081.73.316.755.
 2 sets30.0n.d.23.38.3n.d.21.733.316.718.3n.d.n.d.1015.033.328.330.
 3 sets20.0n.d.11.73.3n.d.38.328.313.335.0n.d.n.d.526.720.016.718.325.028.318.331.728.3n.d.
 4 sets13.3n.d.1.7n.d.n.d.16.710.05.018.3n.d.n.d.3.326.716.7n.d.
 5 sets18.3n.d.n.d.n.d.n.d.
 6 setsn.d.n.d.n.d.n.d.n.d.8.3n.d.n.d.5.0n.d.n.d.n.d.6.7n.d.n.d.n.d.5.06.711.710.06.7n.d.
 7 sets1.7n.d.n.d.n.d.n.d.1.7n.d.n.d.1.7n.d.n.d.n.d.1.7n.d.n.d.n.d.n.d.
 8 setsn.d.n.d.n.d.n.d.n.d.1.7n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.1.7n.d.n.d.n.d.n.d.
Indexed PDF crystallographic orientations (absolute frequency percentb)
 c (0001)13.110058.287.010021.628.752.327.310096.775.613.918.744.351.331.314.019.88.718.7100
 inline image5.1n.d.2.2n.d.n.d.
 inline image68.2n.d.27.58.7n.d.58.855.141.159.0n.d.n.d.19.262.360.843.344.
 inline image1.1n.d.n.d.n.d.n.d.
 inline image1.7n.d.n.d.n.d.n.d.1.5n.d.n.d.n.d.n.d.n.d.n.d.2.21.2n.d.n.d.n.d.
 inline imagen.d.n.d.n.d.n.d.n.d.0.5n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.dn.d.n.d.n.d.
 inline image1.7n.d.3.31.4n.d.1.02.2n.d.n.d.n.d.n.d.n.d.1.31.8n.d.n.d.n.d.n.d.0.92.3n.d.n.d.
 inline imagen.d.n.d.n.d.n.d.n.d.0.5n.d.n.d.n.d.n.d.n.d.1.3n.d.n.d.n.d.n.d.n.d.n.d.n.d0.5n.d.n.d.
 inline image1.7n.d.2.2n.d.n.d.1.0n.d.n.d.0.5n.d.n.d.n.d.2.70.6n.d.n.d.n.d.
 inline imagen.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d0.5n.d.n.d.
 inline imagen.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.0.9n.d.n.d.n.d.n.d.n.d.n.dn.d.n.d.n.d.
 inline image4.0n.d.n.d.n.d.n.d.1.5n.d.n.d.n.d.n.d.n.d.n.d.2.71.8n.d.n.d.n.d.
 inline image0.6n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.1.32.4n.d.n.d.n.d.0.4n.d2.3n.d.n.d.
 inline imagen.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d0.9n.d.n.d.
 inline imagen.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.dn.d.n.d.n.d.
Figure 2.

 Zones of shock at the Siljan structure based on planar deformation features in quartz and occurrence of shatter cones. Plusses mark shatter cone observations made by the Swedish Geological Survey (Kresten et al. 1991). Note also the location of the geographical center, locality 68 and the previous estimated shock center according to Wickman (1980). The distribution of the pressure contours was determined by using the method behind drawing a basic contour map. The contours were drawn by dividing the distance between two localities, belonging to different groups, with the number of groups that ideally would pass through that distance even if one sample is “missing” between them. So, in other words, the distance between a point belonging to group A and C was divided by three and then the two pressure contours were drawn at equal distance from each other between these points.

Table 2.   Grouping of, and pressure estimates for shocked samples from the Siljan structure based on planar deformation feature (PDF) sets in quartz and occurrence of shatter cones.
GroupSamplesDiscriminative featuresAverage No. (maximum) of sets per grainEstimated pressure range (GPa)
  1. Note: n.a. = not applicable.

A16, 63, 72, 73, 74Shatter cones, no PFs or PDFsn.a.2–5
B20, 35, 57, 58, 71PDFs: (0001)1.0 (1.0)5–10
C131, 59PDFs: >75% (0001), inline image1.0–1.5 (4.0)10–15
C221, 54, 64A, 64BPDFs: >50% (0001), inline image1.5–2.0 (5.0)10–15
C311, 51, 52, 55, 62, 65PDFs: >50%inline image, (0001) + others2.5–3.3 (8.0)10–15
D61, 66, 67, 68, 69PDFs: inline image, <20% (0001) + 10% others>3.5 (8.0)15–20

Shock Attenuation at the Siljan Impact Structure

Grouping of the samples at Siljan results in a pattern of shock zones with roughly circular shapes surrounding the central area of the central plateau (Fig. 2). When evaluating the shock pressures that affected the different groups of samples, it becomes evident that recorded pressures decrease smoothly outward from the center of the structure (Fig. 3). The radius of the area where PDFs in quartz were observed is 8–13 km, and observations of shatter cones extend the shocked zone to a radius of 15–16 km.

Figure 3.

 Shock pressure variations with distance from the geographical center of the Siljan structure. Only localities within the annular depression are included. Localities 60, 70, and 56 are excluded (see Appendix 1). For the purpose of the figure, samples have been given the mean pressure for the range defined for the group (see Table 2). Subgroups C1, C2, and C3 (assigned pressure range 10–15 GPa) are given different pressures to illustrate the increasing shock metamorphism within the group. Note also that the rows of dots (samples) are labeled with the group that the samples belong to.

The outer limit of the shocked zone is defined by samples that belong to group A. These come from localities where shatter cones were observed in the field, but that lack PFs or PDFs. The shock pressure range that affected these samples therefore has a lower limit of 2 GPa, the lowest pressure where shatter cones are generated (French 1998), and a higher limit defined by the lower pressure level where PDFs start to form, which is set to 5–8 GPa (Grieve et al. 1996; French 1998 and references therein). Therefore, group A samples are assigned a pressure of 2–5 GPa. Group B were subjected to shock pressures of 5–10 GPa based on the occurrences of PDFs with orientations exclusively parallel to (0001) (Hörz 1968; Stöffler and Langenhorst 1994). The first PDFs parallel to inline image start to develop around (8)10–12(15) GPa (Hörz 1968; Stöffler and Langenhorst 1994; Huffman and Reimold 1996; French 1998). Less common sets parallel to, for example, inline image, inline image, and inline image, are expected to require pressures of at least 13–15 GPa to form (Hörz 1968). Above (16)20 GPa, experiments indicate that also planes parallel to inline image form. These planes increase in frequency with increasing pressure above 20 GPa (Hörz 1968; Müller and Défourneaux 1968; Langenhorst and Deutsch 1994; Huffman and Reimold 1996). Group C is assigned a shock pressure range of 10–15 GPa based on these experimental data. In the local setting of Siljan, it is possible to divide group C into three subgroups—C1, C2, and C3. This division is based on decreasing frequency of basal PDFs, and increasing frequencies of planes parallel to inline image and less common orientations, for example, inline image, and the number of sets/grain. The combination of a low percentage of basal PDFs, high frequencies of less common sets, such as inline image and inline image, and more than 3.5 PDF sets per grain on average, defines group D, which based on these parameters has been assigned a shock pressure range of 15–20 GPa. The maximum of 20 GPa is set due to lack of frequent sets oriented parallel to inline image (samples typically contain two to five grains with one inline image-oriented plane each). For example, Hörz (1968), Engelhardt and Bertsch (1969), and Grieve et al. (1996) state that above 20 GPa, probably at 22–25 (or even 28) GPa, PDFs oriented parallel to inline image are as common as, or even more frequent than those oriented parallel to inline image. Due to heterogeneity in the shock wave distribution in a sample, the pressure range is larger than that defined for occasional grains in the samples, i.e., a sample belonging to group C may have occasional grains that are more indicative of group B.

The area of highest shock at Siljan is located in the center of the structure (Fig. 2). Even though the area defined by the most strongly shocked samples is slightly west of the geographical center, it is locality 68 that we determine as the most strongly shocked one, based on the lowest percentage of basal plane PDFs (8.7%), the highest percentage of planes parallel to less frequent high-angle planes (17.6%), and a high average number of sets/grain (3.6). This locality is situated 400 m south of the geographical center. The decrease in pressure from this locality essentially at the geographical center out to approximately 15 km radial distance (i.e., incorporating also the area with shatter cones) is equivalent to a horizontal shock attenuation gradient across the present erosion level of the central topographic high region of Siljan of 1.3 GPa km−1.

Although several other factors than just shock pressure, such as petrographic features of the rock (e.g., grain size and porosity), the direction of the shock wave, and the preimpact temperature of the shocked rocks influence the development of PDFs (e.g., Hörz 1968; Short 1970; Grieve and Robertson 1976; Walzebuck and Engelhardt 1979; Langenhorst and Deutsch 1994; Huffman and Reimold 1996), the consistent annular pattern leads us to conclude that the shock pressure ranges given above (and in Table 2) are representative of the level of shock that the Siljan samples have been subjected to, considering their mineralogical and petrographic homogeneity.

Petrographic Description of Shock Metamorphic Features

Planar microstructures (PFs and PDFs) were observed mainly in quartz grains as one to eight individual PDF sets per grain, with the average number of PDF sets in quartz grains per sample ranging from 1.0 to 3.8 (Table 1; Figs. 4a and 5; Appendix 2). The PDFs were also rarely observed in feldspar, but extensive alteration of the feldspar minerals did not allow any detailed studies of shock effects in these minerals. Extensive chloritization of biotite also limited the possibility of studying kink bands in biotite. With increasing distance from the center of the structure, the samples become less altered, especially with regard to feldspar minerals and biotite. This observation is reflecting the degree of impact-induced fracturing in the bedrock, causing the samples that are located closer to the impact center, where the effects of the impact are greater, to be more prone to alteration.

Figure 4.

 Thin section photomicrographs of shocked bedrock from the Siljan structure. a) Quartz grain from sample 11 with two decorated planar deformation feature (PDF) sets parallel to planes of the positive inline image and the negative inline image rhombohedron (uncrossed polars). b) Closeup of quartz grain from sample 11 with one decorated PDF set with inline image-equivalent orientation (uncrossed polars). c) Quartz grain from sample 64A with two nearly undecorated PDF sets with inline image-equivalent orientations. Note also the grain to the top right with one decorated PDF set with (0001) orientation (uncrossed polars). d) Quartz grain from sample 64A with one decorated PDF set with basal plane (0001) orientation (uncrossed polars).

Figure 5.

 Histograms of the absolute frequency percent of indexed planar deformation features (PDFs) in quartz from 21 localities across the Siljan structure. PDFs that plot in the overlapping zone between the inline image and inline image crystallographic orientations are plotted in gray on top of the uniquely indexed inline image planes (see text).

In general, the shocked samples have a high percentage of quartz grains displaying planar microstructures, and it is only samples located the furthest away from the center that record <50% shocked quartz grains (Table 1). The number of individual PDF sets per grain, and the abundance of less frequent high-angle sets such as inline image also decreases with increasing distance from the center of the structure (see below).

In shocked quartz, PDFs occur either as thin lamellae composed of amorphous silica or as multiple mechanical Brazil twin lamellae (e.g., Carter 1965; Alexopoulos et al. 1988; Goltrant et al. 1992; Grieve et al. 1996; Trepmann and Spray 2006). The Brazil twin lamellae are oriented parallel to the basal plane of the crystal, and are thus herein referred to as basal PDFs. Postshock annealing and aqueous alteration of a sample results in the transformation of the amorphous lamellae to decorated PDFs (e.g., Goltrant et al. 1992; Trepmann and Spray 2006). The decorated PDFs consist of aligned bubbles that preserve the original orientation of the amorphous lamellae. The majority of observed PDFs in the Siljan samples are decorated (Figs. 4a and 4b), although several localities (especially sample 64A; Fig. 4c) also display nondecorated PDFs, at least on the resolution level available with the petrographic microscope. The decorated nature of PDFs was expected, considering the age of the structure. Postimpact alteration, likely involving hydrothermal alteration, causing the decorations, is also supported by the altered nature of feldspars and biotite. The basal PDFs are more strongly decorated than those oriented along other directions and can therefore easily be distinguished from other PDFs just from their appearance (Figs. 4c and 4d). Strong decoration of PDFs is occasionally associated with toasting of quartz grains in samples 53, 55, 61, and 65–69. This phenomenon is regarded as the result of either high densities of fluid inclusions predominantly located along decorated PDFs (Whitehead et al. 2002), and/or of vesiculation caused by pressure release at high postshock temperatures (Ferrière et al. 2009b, 2010).

In samples where all quartz grains are shocked, the PDFs generally cross entire crystals, or the majority of the affected crystal (Fig. 4a). In contrast, with decreasing percentage of shocked grains, the PDFs become less and less penetrative, appearing only near grain boundaries and/or fractures. Basal PDFs generally penetrate entire crystals in samples with a low number of sets/grain, but with a high percentage of shocked grains, e.g., samples 57 and 58. In samples with a high number of sets per grain, on the other hand, for example, 66–69, the Brazil twins are nearly indistinguishable, barely visible with ×200 magnification, and only in a limited area of the crystal, commonly near the grain boundary or fractures. Samples that display both frequent rhombohedral PDFs (especially inline image and inline image orientations), and higher angle PDFs (e.g., inline image), reveal that the high-angle sets are almost exclusively only visible in parts of grains, whereas the lower angle plane orientations are always penetrating entire grains.

The majority of the PDF sets (Table 2; Figs. 5 and 6; Appendix 2) are oriented parallel to the inline image orientation (53.2%; Fig. 4a). The majority of the poles of these PDF planes are oriented at an angle to the c-axis of between approximately 18° and 23°, meaning that they cannot be uniquely indexed as inline image, as they are overlapping the inline image orientation (see template in Ferrière et al. 2009a for details). These planes are considered as inline image orientations, as recommended by Ferrière et al. (2009a), but are reported as gray columns in Fig. 5.

Figure 6.

 Histogram showing the distribution of planar deformation features (PDFs) in different crystallographic orientations both in relation to the total number of measured PDFs (dark gray columns) and the abundance of them among the samples (light gray columns). Inset is a closeup of dark gray columns showing the PDFs that occur in minor amounts in the samples, and also the frequency of unindexed planes.

The second most abundant orientation is (0001), i.e., basal PDFs, constituting 30.3% of the total population of PDFs. In addition to these two dominant sets, planes parallel to inline image occur in minor amounts (each between 0.04% and 5.9% of the total measured PDFs). Of these, planes parallel to inline image, together constituting 8.9% of the total planes, are the most common. Only 2.8% of the planes could not be indexed (Fig. 6). This either reflects errors of the initial measurement, or the fact that the planes do not correspond to any rational crystallographic orientation.

PFs in quartz occur sporadically as open, filled with secondary minerals, or closed fractures in several samples (Appendix 1). PFs that are oriented parallel to the c-axis are most common in the samples. These always occur together with a set of PDFs that are also oriented in the (0001) direction. In more strongly shocked samples, the PFs are associated also with sets parallel to higher angle planes. PFs oriented parallel to the inline image orientation were also observed, for example, in samples 68 and 52. Grains with multiple sets of PFs were not observed.

Other shock-related features that were observed are melt-breccia veinlets in samples 51, 61, 62, 66, 68, and 69 (Fig. 7a), and cataclastic zones in samples 62 and 68 (Fig. 7b). The melt-breccia veinlets are generally submillimeter-sized (0.2–0.8 mm thick), with a maximum width of approximately 2.5 mm, and can be followed through entire thin sections. They consist predominantly of microcrystalline material that we interpret as recrystallized melt. Flow structures can be observed, especially associated with the material that is most fine-grained. Clasts found in the veins are almost exclusively rounded and the majority of them represent feldspars and quartz. Other clasts observed were composed of titanite and secondary minerals (e.g., chlorite). In the boundary to the veins, the surrounding minerals are rounded. In several places, especially where the veins are at their thinnest, the boundary consists of material that has suffered cataclasis, or crushing, with strong reduction of particle size as a consequence. We believe that sample 51B represents the same material dated by, and discussed in, Reimold et al. (2005), referred therein to Si-1. Reimold et al. (2005) interpret this material to be impact melt breccia (see Reimold et al. 2005 for a more detailed description). As no clasts were found in any of the thin sections produced from this sample, no PFs or PDFs were observed.

Figure 7.

 Thin section photomicrographs of shocked bedrock from the Siljan structure. a) Melt-breccia veinlet from sample 61 (uncrossed polars). b) Cataclastic zone in sample 68 showing decrease in grain size relative to the average grain size of the thin section as a whole (crossed polars). fsp = feldspar; amph = amphibole; qz = quartz; cl = clast; vn = veinlet.


With the confirmation of the shock center location at Siljan, based on shocked quartz, to almost exactly the geographical center, the previously suggested center of the structure, located approximately 5 km north of the geographical center (Fig. 2) (Wickman 1980) that was based on measurements of shatter cone orientations can be discarded. Instead, the shock center is now equivalent in its position to other complex impact structures (cf. e.g., Dressler 1990). Circular patterns of shock attenuation across the central sections of complex impact structures have also been observed, for example, at Charlevoix, Manicouagan, and Slate Islands, all located in Canada (e.g., Robertson 1975; Dressler 1990; Dressler et al. 1998). We can conclude that also for Siljan, it is true that, although the central uplifted area of complex impact structures is deformed during formation (Grieve et al. 1981; Melosh 1989), the movement does not disrupt the general circular pattern of shock attenuation at a deep erosion level.

The major difference in the measured PDF pattern between the three Canadian structures and Siljan is the scarcity of planes parallel to inline image at Siljan. We conclude, based on the current literature, that the reason for the lack of these planes indicative of high pressures (e.g., Hörz 1968) reflects the fact that Siljan is more deeply eroded than the three mentioned Canadian impact structures. This statement is corroborated by the fact that Siljan lacks both a topographic high, representing either a central peak or ring/s, and preserved in situ crater-fill deposits. Although the presently similar-sized Charlevoix structure (estimated diameter of central plateau approximately 30 km; Grieve 2006) and Siljan are not eroded to the same level, indications suggest that Siljan was larger than Charlevoix. At Siljan, the shocked zone extends out to 11.3 km from the center at the present erosional surface. At Charlevoix, PDFs in quartz have been documented out to a radial distance of approximately 10 km (Robertson 1975). If erosion of complex impact structures decreases the lateral extent of the shocked area (e.g., Robertson 1975; Therriault et al. 1997), then Siljan, which is more deeply eroded than Charlevoix, was originally significantly larger than Charlevoix (see also discussion below).

Although the pattern for shock attenuation at Siljan is a smooth decrease from the center, there are occasional “jumps” where a shock zone is either very thin or maybe even missing. As samples are generally located quite far from each other outside of zones D and C3, it is not surprising that in a linear array drawn from the center toward the annular graben, samples from all groups are not represented. There is simply too much distance between the samples to discuss apparent “jumps.” However, between localities 11 and 59, the distance is <1 km although they represent (sub)groups C3 and C1. This is interpreted as the result of small differential movement among large blocks at this location during central uplift formation, and in fact, a fault separates these two localities (see Kresten et al. 1991). A sample taken between these two localities could possibly clarify this.

Original Morphology and Size of the Siljan Structure

At the present level of erosion, the original morphology of Siljan cannot be precisely determined, although it is clear that it is a complex impact structure, that originally had the morphology of either a central peak or peak ring crater (see also discussion in Grieve 1988). The transition from a central peak structure to a peak ring structure occurs at about 25 km on Earth, based on observations of terrestrial impact structures (Grieve et al. 1981; French 1998), indicating that Siljan, as it is significantly larger than 25 km, can have been a peak ring structure before erosion.

Grieve and Head (1983) and Grieve (2006) interpret the annular moat at Manicouagan as the result of differential erosion due to glacial action at a competency boundary at the present moat area, and not an original feature of the crater. Siljan also has been strongly eroded during glaciations, and we consider that the present annular depression is a result of significant erosion of the less competent Paleozoic sediments that were down-faulted during central uplift formation compared to the surrounding crystalline basement. This means that the original dimensions of Siljan cannot be determined based on present topography. However, the study of the distribution of shocked quartz in combination with observations of shatter cones across the structure allows interpretations of the original size of Siljan.

Shock effects in quartz are developed only in the area of a complex impact structure that was originally located well inside of the transient cavity (Grieve et al. 1996). The transient cavity is formed when the target material is driven away from the impact point upon impact before it later collapses to form the modified crater (Melosh 1989; French 1998). The outermost limit where PDFs develop corresponds to approximately 0.6–0.7 transient crater radii (Dence 1972; Robertson and Grieve 1977). With this study, we have showed that the maximum distance from the geographical center where PDFs in target rocks occur is at 11.3 km radial distance (i.e., the distance between the geographical center and locality 20). The diameter of a circular area defined by this outer limit is thus 22.6 km. We can therefore define a minimum estimate of the diameter of the transient cavity of Siljan at the present level of erosion, to 32–38 km. Further, we can also estimate the transient cavity diameter based on the outer limit of shatter coning, as shatter cones are also formed within the original transient cavity (e.g., Stöffler et al. 1988). The sample that is furthest away from the center where shatter cones were observed in the field is located at a distance of 14.8 km. The diameter of the transient cavity defined by this observation is thus about 30 km. However, the sampling done for this study did not focus on constraining the distribution of shatter cones at Siljan, and therefore this number represents a minimum value. Kresten et al. (1991) present shatter cone observations up to a distance of 17.2 km from the center, corresponding to a transient cavity diameter of maximum 34 km, i.e., within the range of possible transient cavity diameters defined by the previous calculations based on PDF distribution. The transient cavity of Siljan can therefore be estimated to be between 32 and 38 km in diameter (Table 3). Grieve (1988) has previously discussed the dimensions of the transient cavity at Siljan, and proposed a possible maximum radius for the transient cavity at Siljan based on the innermost occurrences of sediments. The Ordovician and Silurian sediments that are preserved at the Siljan structure represent material that was down-faulted during modification of the crater, resulting in a mega-breccia, with repetition of strata (e.g., Collini 1988; Kenkmann and von Dalwigk 2000). Therefore, the innermost observation of isolated Paleozoic sediments, at approximately 13 km (Grieve 1988) from the center of the structure, does not correspond to the maximum radius for the transient cavity. Other previously published estimations of the diameter of the transient cavity at Siljan were based on geophysical data produced in relation to the drillings (25 km; Juhlin et al. 1991) and on calculations based on the 65 km diameter estimate for the final structure (42 km; Kenkmann and von Dalwigk 2000). Both of these estimates relate to the uneroded target surface.

Table 3.   Diameter (D) estimates for the transient cavity of the Siljan structure based on planar deformation features in quartz grains (I) and shatter cones (II).
Estimated D (km)Estimated D (km; corrected for erosiona)RelationshipReferences
  1. Note: tc = transient cavity.

  2. aErosion estimate between 1 and 4 km; see text for details on the method of calculation.

(I) 32–3834–4611.3 = 0.6–0.7 tc radiiDence (1972); Robertson and Grieve (1977)
(II) 3436–4217.2 ≤ tc radiusStöffler et al. (1988) 

Erosion has likely reduced the spatial distribution of shock metamorphic features, but to what extent is unknown, and therefore the estimate of 32–38 km is a minimum value. If we assume that hypervelocity impact produces a parabolic transient cavity (e.g., Dence 1973; Dence et al. 1977) that can be described with the formula y2 = 2px, where p is the crater depth below the original ground surface, and the radius of the transient cavity (r) can be given by r2 = 2p2, we can estimate the effects of erosion (Table 3). This has been done also, for example, for the Vredefort and Charlevoix impact structures (Robertson 1975; Therriault et al. 1997).

Several authors have presented relationships between transient cavity diameter and final rim diameter (e.g., Grieve et al. 1981; Croft 1985; Lakomy 1990), and therefore we can estimate the original size of Siljan based on the data of transient cavity diameters. Estimated diameters of the final crater rim of Siljan are on the order of 47–75 km (with erosion correction 50–91 km), and are summarized in Table 4. Therriault et al. (1997) presented another relationship which defines the final rim diameter of an impact structure based on the central uplift area. If we assume that the area of Siljan that is located inside of the annular graben is the central uplift area (e.g., Grieve 1988), then we can estimate the final rim diameter of Siljan to approximately 89 km. The diameter estimate of 50–91 km means that the range in possible diameters for the Siljan crater is 40 km. The reason for this is the spans in the scaling relationships and insecurities in the amount of erosion (calculations were made on 1 and 4 km of erosion). A possible way to better constrain the amount of erosion would be to calculate, based on the present shock level of the rocks, how much of the central uplift is missing.

Table 4.   Diameter (D) estimates for the present and the uneroded Siljan structure.
Estimated present D (km)Estimated uneroded Da (km)RelationshipReferences
  1. Note: tc = transient cavity; n.a. = not applicable; cp = central uplift area.

  2. aDtc corrected for erosion (see text).

  3. bEstimated simple to complex transition at approximately 4 km.

50–7553–91Dtc = 0.5–0.65DGrieve et al. (1981)
47–5650–70Dtc = 1.23D0.85bCroft (1985)
57–6660–80Dtc = 0.57DLakomy (1990)
89n.a.Dcp = 0.31D1.02Therriault et al. (1997)

For a structure with the size of Siljan, the expected peak shock pressures in the parautochtonous bedrock of the upper part of the central uplift of the uneroded structure should have been on the order of 30 GPa or slightly higher (e.g., Stöffler et al. 1988; French 1998). The rocks presently exposed at the central uplift record maximum pressures of approximately 20 GPa. Stöffler et al. (1988) and French (1998) suggest that the pressure decreases from 25 to 30 GPa at the surface, to 0 GPa over no longer vertical distance than a few kilometers at the center of large structures. Only a few studies have been published that investigate variations in PDFs in quartz with depth at complex impact structures (the Puchezh-Katunki, Kara, and Bosumtwi structures) and discuss vertical shock attenuation (e.g., Sazonova et al. 1981; Pevzner et al. 1992; Fel’dman et al. 1996; Masaitis and Orlova 1999; Ferrière et al. 2008). The drilling into the central part of the central uplift of the Puchezh-Katunki structure, Russia, revealed that the pressure decreases from approximately 40 GPa to approximately 10 GPa at the bottom of the 5 km deep drill hole (Ivanov 1994). However, the diameter of the central uplift of Puchezh-Katunki is significantly smaller than Siljan (approximately 15 km compared to 28–30 km for Siljan). In addition, we do not know the lateral extent of the shock zones at Puchezh-Katunki. Therefore, to conclusively constrain the original rim to rim diameter of Siljan, studies of drill cores from the central part of the structure, providing a 3-D “shock hemisphere,” are required.

Shock Metamorphic Features

The lack of Brazil Twin lamellae in samples from the central part of the structure can be explained by the fact that Brazil Twins form when the deviatoric component of the stress associated with the shock wave is large compared to the hydrostatic component, which is related to the distance from the center of impact structures (e.g., Leroux et al. 1994; Trepmann and Spray 2006). The reason for low amounts of observed basal PDFs at central localities is most likely not the lack of decorations making the lamellae hard to detect (Grieve et al. 1996), as the basal PDFs in the Siljan samples are always more decorated than other sets, but related to the hydrostatic component of the stress associated with the shock wave, favoring development of planes parallel to other orientations.

The melt-breccia veinlets are found in the group C3 and D shock zones (pressures below 20 GPa), and therefore they had to have been formed by localized high-pressure processes. In fact, we consider the veinlets to be small scale pseudotachylite breccia veins (e.g., Mohr-Westheide and Reimold 2011). Mohr-Westheide and Reimold (2011) explain the formation of such veins as a response to high pressure localized in limited areas, and related high postshock temperatures, and in addition, a temperature component controlled by frictional movement. We believe that the cataclastic zones observed in several samples represent zones in which strong reduction of particle size took place as a response to microfracturing and cataclasis, but these zones never reached the next step—melting (e.g., Mohr-Westheide and Reimold 2011). The conditions facilitating such action were attained only in limited areas of the samples collected for this study, controlled by local heterogeneities such as non–shock-related fractures. The fact that these veinlets are found only in samples that belong to groups C3 and D indicate that a pressure of >10 GPa was required for the conditions of localized pressure increase.


  • 1 Sampled bedrock from Siljan contains abundant shock metamorphic features, including PFs and PDFs in quartz, that record pressures of up to 20 GPa. The PDF distribution pattern among the samples shows that the most strongly shocked locality is essentially located at the geographical center of the structure.
  • 2 The pre–central-uplift-formation spatial relationships of the parautochtonous rocks of the present central plateau are well preserved considering the roughly circular shape of the shock attenuation envelopes surrounding the shock center.
  • 3 The distribution of shocked quartz suggests that the transient cavity at Siljan was 32–38 km in diameter, and that the original diameter of Siljan can have been as large as 90 km.
  • 4 For an even better constrained original rim to rim diameter of the Siljan structure, studies of drill cores from the central part of the structure, providing a three-dimensional image of the shock distribution, are needed.

Acknowledgments–– S. H. thanks professor R. Wieler, ETH Zürich, for the support and use of facilities during her stay at ETH Zürich. We greatly appreciate the unregistered use of computer program Stereo32. We also thank Lunds Geologiska Fältklubb for the grant that financed part of the field work. We are grateful to M. Poelchau and L. Ferrière for reviews that improved the manuscript. This study was supported by grants to B. S. by the Swedish Research Council.

Editorial Handling–– Dr. Christian Koeberl


Appendix 1

Description of sampled localities from the Siljan structure. Displayed is also the distance from the geographical center of the structure for each sample and short notes on optical and U-stage microscopic observations of shock metamorphic features.

Sample No.Distance from center (km)aCoordinatesDescriptionNotesb
  1. Note: PDF = planar deformation feature; PF = planar fracture.

  2. aCoordinates for center are N61°02.196′; E014°54.467′.

  3. bPlanar microstructure observations only noted for quartz (see text).

  4. cPFs are rare in all samples, if observed they occur only in a few grains.

 128.5N60°52.221′; E015°18.368′Coarse-grained porphyritic granite 
 228.5N60°52.333′; E015°18.368′Medium-grained porphyritic granite 
 328.3N60°52.577′; E015°17.835′Medium-grained porphyritic granite 
 427.6N60°53.243′; E015°18.896′Medium-grained porphyritic granite 
 527.4N60°53.270′; E015°18.863′Coarse-grained porphyritic granite 
 622.5N60°55.127′; E015°14.853′Medium-grained porphyritic granite and diabase 
 7A22.2N60°55.303′; E015°14.760′Medium-grained granite 
 7B22.2N60°55.303′; E015°14.760′Medium-grained granite 
 823.8N60°55.176′; E015°16.568′Coarse-grained porphyritic granite and mafic rock 
 915.4N60°58.014′; E015°09.387′Coarse-grained porphyritic granite 
1014.7N60°58.041′; E015°08.502′Medium-grained porphyritic granite 
116.4N60°58.944′; E014°53.494′Medium-grained granitePDFs: Decorated and nondecorated. Occasional toasted quartz grains
1215.4N60°54.503′; E015°00.197′Coarse-grained porphyritic granite 
1322.1N60°52.433′; E015°08.836′Medium- to coarse-grained porphyritic granite 
1414.5N60°55.067′; E015°00.610′Fine-grained mafic rock 
1514.4N60°55.657′; E015°02.813′Medium-grained porphyritic granite 
1611.1N60°57.573′; E015°02.278′Medium-grained porphyritic graniteShatter cones
1717.8N60°52.815′; E014°57.276′Medium-grained porphyritic granite 
1817.8N60°52.814′; E014°57.091′Coarse-grained porphyritic granite 
1916.1N60°53.844′; E014°51.164′Sedimentary rock 
2011.3N60°56.637′; E014°50.372′Fine-grained graniteShatter cones. PDFs: Decorated, visible only near grain boundaries
218.9N60°59.569′; E014°46.760′Coarse-grained porphyritic granitePDFs: Decorated. Undulose extinction of quartz
2215.7N61°00.336′; E014°37.786′Medium-grained granite 
2320.7N60°54.964′; E014°37.008′Medium-grained weakly porphyritic granite 
2421.1N60°53.896′; E014°38.071′Medium-grained porphyritic granite 
2521.8N60°51.087′; E014°36.016′Medium-grained granite 
2614.0N61°02.975′; E014°39.316′Porphyritic volcanic rock 
2714.1N61°03.726′; E015°10.127′Medium-grained porphyritic granite 
2812.0N61°04.109′; E015°07.529′Medium-grained granite 
2914.8N61°07.029′; E015°08.022′Fine-grained weakly porphyritic granite 
309.2N61°04.440′; E015°04.048′Medium- to coarse-grained porphyritic granite 
314.4N61°04.258′; E014°57.545′Coarse-grained porphyritic granitePDFs: Decorated. PFsc. Strong undulose extinction of quartz
328.9N61°05.236′; E015°02.643′Coarse-grained porphyritic granite 
3311.3N61°07.053′; E015°02.643′Medium-grained granite 
349.7N61°07.016′; E014°59.630′Medium- to coarse-grained porphyritic granite 
359.9N61°07.036′; E014°49.625′Medium-grained porphyritic granitePDFs: Decorated and only visible near grain margins. PFs
3610.6N61°07.133′; E014°48.314′Coarse-grained porphyritic granite 
3716.7N61°06.625′; E014°38.455′Brecciated porphyritic volcanic rock 
3817.2N61°11.152′; E014°48.948′Volcanic rock 
3916.5N61°10.982′; E014°50.784′Volcanic rock 
4012.7N61°09.108′; E014°57.071′Medium- to coarse-grained porphyritic granite 
4112.6N61°09.043′; E014°57.230′Fine-grained granite 
4215.6N61°09.435′; E015°04.050′Fine-grained granite 
4319.7N61°12.301′; E015°06.223′Medium-grained porphyritic granite 
4430.5N61°14.380′; E015°05.367′Medium-grained porphyritic granite 
4523.1N60°51.508′; E015°07.492′Coarse-grained porphyritic granite 
4624.1N60°50.646′; E015°08.144′Sandstone 
4726.5N60°48.715′; E015°03.875′Medium-grained weakly porphyritic granite 
4827.8N60°47.964′; E015°03.237′Medium-grained granite 
4926.9N60°48.062′; E015°01.530′Medium-grained granite 
5032.0N60°45.263′; E015°00.933′Medium-grained granite 
51A3.3N61°01.284′; E014°51.359′Coarse-grained porphyritic granitePDFs: Decorated and undecorated. PFs. Melt-breccia veinlet. Strong undulose extinction of quartz
51B3.3N61°01.284′; E014°51.359′Aphanitic material 
525.2N61°04.200′; E014°50.610′Fine-grained granitePDFs: Decorated and undecorated
544.1N61°00.117′; E014°54.557′Coarse-grained porphyritic graniteShatter cones. PDFs: Decorated. PFs. Undulose extinction of quartz
553.4N61°00.885′; E014°57.126′Medium-grained porphyritic graniteShatter cones. PDFs: Decorated. Strong undulose extinction of quartz. Some toasting
5611.1N60°56.583′; E014°57.960′Fine-grained dioriteNo quartz
578.6N60°57.721′; E014°53.894′Coarse-grained porphyritic graniteShatter cones. PDFs: Decorated, visible only near grain boundaries. Undulose extinction of quartz
589.2N60°57.409′; E014°54.355′Coarse-grained porphyritic graniteShatter cones. PDFs: Decorated. PFs filled with secondary minerals. Undulose extinction of quartz
597.3N60°58.524′; E014°52.866′Coarse-grained porphyritic graniteShatter cones. PDFs: Decorated. PFs filled with secondary minerals. Strong undulose extinction of quartz
607.0N60°58.565′; E014°55.323′Fine-grained quartz dioritePDFs in quartz (unmeasured). Grain size too small
612.3N61°01.268′; E014°53.411′Coarse-grained porphyritic granitePDFs: Decorated. PFs. Strong undulose extinction of quartz. Melt-breccia veinlet
625.8N60°59.524′; E014°52.030′Coarse-grained porphyritic granitePDFs: Decorated. PFs filled with secondary minerals. Strong undulose extinction and some toasting of quartz. Melt-breccia veinlet and cataclastic zone
6314.8N61°02.258′; E014°38.366′Medium-grained graniteShatter cones
64A5.3N61°04.046′; E014°50.075′Fine-grained granitePDFs: Decorated and undecorated
64B5.3N61°04.046′; E014°50.075′Medium-grained granitePDFs: Decorated. PFs. Undulose extinction of quartz
654.4N61°03.671′; E014°50.729′Medium-grained porphyritic granitePDFs: Decorated and fresh-looking. PFs filled with secondary minerals. Strong undulose extinction of quartz. Occasional toasting of quartz
662.9N61°03.190′; E014°52.150′Coarse-grained porphyritic granitePDFs: Decorated and fresh-looking. PFs filled with secondary minerals. Undulose extinction of quartz. Occasional toasting of quartz. Melt-breccia veinlet
673.4N61°04.146′; E014°53.946′Coarse-grained porphyritic granitePDFs: Decorated and fresh-looking. PFs filled with secondary minerals. Strong undulose extinction of quartz. Occasional toasting of quartz
680.4N61°02.111′; E014°54.790′Coarse-grained porphyritic granitePDF: Decorated and fresh-looking. PFs filled with secondary minerals. Strong undulose extinction of quartz. Toasting of quartz. Melt-breccia veinlet and cataclastic zone
691.7N61°03.051′; E014°53.612′Coarse-grained porphyritic granitePDFs: Decorated and fresh-looking. PFs filled with secondary minerals. Undulose extinction of quartz. Occasional toasting of quartz. Melt-breccia veinlet
703.3N61°02.149′; E014°51.117′Fine-grained quartz dioritePDFs (unmeasured) in quartz. Grain size too small
717.4N61°01.604′; E015°02.803′Coarse-grained granitePDFs: Decorated
728.45N60°59.756′; E015°02.512′Fine-grained quartz dioriteShatter cones
7313.3N60°57.705′; E015°05.988′Coarse-grained porphyritic graniteShatter cones
7412.4N60°59.394′; E015°07.184′Medium-grained porphyritic graniteShatter cones

Appendix 2

Histograms of angles between c-axis and poles to PDFs, binned by 5°, in quartz grains from sampled bedrock from the Siljan structure. Light gray columns represent PDFs that could be indexed using a stereographic projection template (see text). Black columns represent unindexed PDF sets.

inline image