sign in sign up
<<
Home , Dike (geology), Sediment, Stormberg Group

dikes in dolerite sills: Evidence for high-pressure gradients and mobilization during solidifi cation of magmatic sheet intrusions in sedimentary basins

Henrik Svensen Ingrid Aarnes Yuri Y. Podladchikov Physics of Geological Processes, University of Oslo, PO Box 1048 Blindern, 0316 Oslo, Norway Espen Jettestuen Physics of Geological Processes, University of Oslo, PO Box 1048 Blindern, 0316 Oslo, Norway, and International Research Institute of Stavanger (IRIS), Prof. Olav Hanssensvei 15, 4068 Stavanger, Norway Camilla H. Harstad AGR Group, Karenslyst allè 4, 0278 Oslo, Norway Sverre Planke Physics of Geological Processes, University of Oslo, PO Box 1048 Blindern, 0316 Oslo, Norway, and Volcanic Basin Research (VBPR), Oslo, Norway

ABSTRACT aureoles on a short time scale, representing That study was based on fi eld examples from an intermediate situation between fl uid loss a emplaced in during formation Sediment dikes are common within doler- during formation of microfractures and fl uid of the Central Atlantic Magmatic Province. The ite sill intrusions in the Karoo Basin in South loss during violent vent formation. role of pore fl uid boiling in causing high aureole Africa. The dikes are subvertical and as pressures and subsequent fl uid movement was much as 2 m wide, sometimes with abundant INTRODUCTION explored in more detail by Delaney (1982; and fragments of sedimentary rocks and dolerite. more recently, e.g., Jamtveit et al., 2004). The matrix consists of contact-metamorphic Subsurface sediment mobilization and fl u- Understanding sediment mobilization from sandstone. There is no petrographic evidence idization have been recognized from many contact aureoles may put important constraints for melting within the sediment dikes. The geological settings, ranging from overpressured on pressure evolution of aureoles. The past maximum temperature during heating is clastic reservoirs (Jolly and Lonergan, 2002; decade has seen an increasing interest in degas- restricted to the plagioclase and biotite stabil- Mazzini et al., 2003; Nichols et al., 1994) to sing of volatiles from sedimentary basins with ity fi eld, or above ~350 °C. Thermal model- contact around magmatic sill magmatic intrusions, where high pore fl uid ing of a sandstone in a dolerite sill shows intrusions (Jamtveit et al., 2004; Svensen et al., pressure plays a key role (Ganino and Arndt, that a temperature of 350–450 °C is reached 2006). In sedimentary basins affected by mag- 2009; McElwain et al., 2005; Retallack and in the dike after a few hundred years of sill matic sill intrusions (i.e., volcanic basins), like Jahren, 2008; Svensen et al., 2004, 2007, 2009). cooling. The calculated pressure history of a the Karoo Basin in , sediment dikes Gas venting triggered by overpressure in con- cooling sill and its contact aureole shows that are reported from within doleritic sills (Van Bil- tact aureoles within has been proposed substantial fl uid pressure anomalies develop jon and Smitter, 1956). It is interesting that these to have caused global changes in the on a short time scale (1–15 yr) and are main- dikes comprise metamorphic sandstone, demon- end-, Early (Toarcian), and at tained for more than 100 yr. Calculated pres- strating that the intruded the dolerite while the Paleocene-Eocene boundary (Svensen et al., sure anomalies in the sill (−7 to −22 MPa) the sills were still hot. The importance of these 2004, 2007, 2009). and the aureole (4–22 MPa) are signifi cant observations is that they form direct evidence The aim of this study is to understand the and may explain sill fracturing and sediment for high pore fl uid pressure during sill emplace- formation of sandstone intrusions in dolerite mobilization from the aureole into the sill. ment and subsequent contact metamorphism. sills. We present several case studies of sedi- We conclude that sediment dikes represent In a classic study by Walton and O’Sullivan ment dikes and sediment within sills common features of sedimentary basins with (1950), it was suggested that pressure drop in the Karoo Basin. However, the results can be sill intrusions in which fl uid pressure gra- during sill cooling and fracturing (i.e., thermal applied to other sedimentary basins where sedi- dients have been high. Sediment dikes thus contraction) led to boiling of aureole pore fl u- ments have been injected into magmatic sheet signify that pore fl uids may escape from the ids that ultimately led to sediment fl uidization. intrusions, including the Vøring Basin offshore

Geosphere; June 2010; v. 6; no. 3; p. 211–224; doi: 10.1130/GES00506.1; 10 fi gures; 1 table.

For permission to copy, contact [email protected] 211 © 2010 Geological Society of America

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Svensen et al.

Norway, the Tunguska Basin of east Siberia, 2005), although in sents a two-dimensional (2D) slice through the and the Amazonas Basin in Brazil. The process continued for several million years (Jourdan et dike. We then used image analysis techniques of sediment injections is addressed by adopting al., 2005). Sills and dikes are present throughout and a MATLAB (http://www.mathworks a new theoretical model for sill pressure evolu- the sedimentary in the Karoo Basin .com/) code to quantify the clast content (i.e., tion during cooling and crystallization (Aarnes (Fig. 1) (Chevallier and Woodford, 1999; Pol- area). Probability densities were calculated et al., 2008). teau et al., 2008b), where they locally compose using a smoothing procedure, where data were as much as 70% of the (Rowsell binned in either 10 consecutive areas (for sedi- GEOLOGICAL SETTING and De Swardt, 1976). ment clasts) or 5 consecutive areas (for dolerite clasts). The aspect ratio between the long and The Karoo Basin (Fig. 1) covers more than METHODS short axes of the fragments was also calculated. half of South Africa. The basin is bounded by Since our mapping analyses are done in 2D, the Cape Belt along its southern mar- Sampling and Petrography and we only have one slice through the dike, gin and comprises as much as 6 km of clastic the results should be regarded as approximate. sedimentary strata capped by at least 1.4 km of Sediment dikes are common within thick Thin sections of collected samples were stud- basaltic (e.g., Johnson et al., 1997; Smith, (70–120 m) dolerite sills within the Beau- ied by optical and electron microscopes (scan- 1990). The sediments were deposited from the fort Group sediments. The depth of ning electron microscope, SEM) at the Depart- late to the Middle Jurassic, in an emplacement is estimated as 600–1000 m ment of Geology, University of Oslo. The SEM environment ranging from dominantly marine below the paleosurface, based on present-day is a JEOL JSM 840, and was also utilized for (the Dwyka and Ecca Groups) to fl uvial (the stratigraphic levels. We have done detailed cathodoluminescence (CL) imaging. and parts of the Stormberg studies of three localities with sediment dikes Group) and eolian (upper part of the Stormberg in dolerites: (1) the Waterdown area, Phase Stability Calculations Group) (Catuneanu et al., 1998; Veevers et al., (2) the Elandsberg roadcut (Nico Malan Pass), 1994). The Beaufort Group is a thick sequence and (3) the Golden (Fig. 1). Many more We used Perple_X (Connolly, 2005) to com- of dominantly . The overlying Storm- localities with sediment dikes have been dis- pute phase diagrams for rocks with a pelite berg Group includes the covered during our fi eld work in the Karoo composition to predict the temperature stabil- (coarse sandstone, shale, and ), the Elliot Basin during the past decade (e.g., south of ity of the assemblages identifi ed in the Formation (sandstone, shale; red beds), and the Cathcart), but the chosen localities are repre- sandstone dikes. The calculated phase diagram (sandstone with occasional sentative. One of the sediment dikes from the is projected from an average pelite composi- horizons). Waterdown Dam locality contains numerous tion (Caddick and Thompson, 2008), with

Both southern Africa and under- fragments of sediments and dolerite. It was SiO2 = 59.8, Al2O3 = 16.6, FeO = 5.8, MgO =

went extensive volcanic activity in early Juras- mapped in detail by covering it with transpar- 2.6, CaO = 1.1, Na2O = 1.7, K2O = 3.53,

sic times, starting ca. 182.5 Ma. Dolerites and ent A4 plastic sheets and tracing individual TiO2 = 0.75, H2O = 5.0 (all in wt%). We calcu- of the Karoo-Ferrar clasts by hand. This method was preferred over lated the reactions using quartz saturation, which were emplaced within a relatively short time photo analysis due to better accuracy and the means that the phase assemblages obtained are span. The main phase of fl ood volcanism lasted benefi t of doing on-site interpretations on clast not dependent upon the bulk content of quartz. <1 m.y. (Duncan et al., 1997; Jourdan et al., type and clast outline. The resulting map repre- Hence the phase diagram is valid for sandstones

Drakensberg Gr. Ecca Gr. Waterdown Dam Stormberg Gr. Dwyka Gr. A B Beaufort Gr. Sill intrusion The Whitehill Fm. 30 Waterdown Dam 3 2 Farm

Golden Valley 1 32 Cathcart Waterdown Dam (R 6 Nico Malan Pass 7)

Cape Fold belt Indian Ocean Cape Town Dolerite dike 20 150 km 2km 24 28 Sediment dike

Figure 1. (A) Simplifi ed geological map of the Karoo Basin in South Africa; black dots are sill intrusions and hydrothermal vent complexes. The three study localities are shown. One more locality with sediment dikes in sills is shown (Cathcart), but not included in this study. (B) Simplifi ed geological map of the Waterdown Dam area based on the 1:250 000 geological map of Council for Geo- science, South Africa. Note that there is one locality with dikes that have not been included in this study.

212 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Sediment dikes in dolerite sills

as well as pelites, as long as the ratios of the occurring at subsolidus conditions. The ther- pressure is not achieved in our model, which other oxides do not change signifi cantly. mal diffusivities are equal for the sill and the suggests that we are using conservative values. sedimentary host , as differences in thermal The main equations used for the modeling are Numerical Modeling properties are negligible (see Table 1). However, shown in the Appendix. the hydraulic diffusion coeffi cients of melts and We have developed a numerical model using pore fl uids differ by approximately one order RESULTS the fi nite element method (FEM) in MATLAB. of magnitude in our model. We assume no heat We couple standard heat conduction to pressure advection by fl uids in either the sill or the con- Sediment Dikes in Dolerite Sills (or hydraulic) diffusion using the equation for tact aureole. This is justifi ed from studies show- thermal similar to that of Aarnes et al. ing that heat advection by fl uids is a second- Waterdown Dam (2008). We calculate the pressure anomalies order effect (Connolly, 1997; Podladchikov Several sediment dikes within dolerite sills arising from pore fl uid expansion of pure water and Wickham, 1994). Apart from the sandstone are located in roadcuts along the Waterdown in the contact aureole, and the pressure changes dikes, there is little evidence of high fl uid circu- Dam north of the Elandsberg area in South related to phase transitions (melt to crystal) in lation in the intruded sediments, which makes Africa (Fig. 1). The main sites are numbered the sill. The pressure anomalies diffuse over heat advection within the intrusion negligible 1–3 in Figure 1B, where thick sediment dikes time according to Darcy’s law. The equations (cf. Norton et al., 1984). are exposed close to the lower contact of a trans- are solved on a 2D square grid with a resolution The major assumption concerning the equa- gressive dolerite sill. The intruded sediments of 25 × 200 elements. Initial conditions for the tion of thermal stresses is that expansion of pore are mainly sandstones, all from the Permian

thermal solver is a host-rock temperature Thr of fl uids and contraction of melt due to crystal- and Beaufort Group. An overview of

35 °C, and a sill temperature, Tm, of 1200 °C. lization are prevented either by the sediment the locality is given in Figure 2A. At all sites, For temperature boundary conditions we fi xed matrix or the crystal network. This assumption the fi eld evidence suggests upward movement both the upper and lower boundaries at initial is valid until the expanding fl uids break the sedi- of sediment, based on the presence of dolerite host-rock temperature, as the geothermal gradi- ment matrix and reduce the overpressure, either bridges. The maximum upward penetration is ent is negligible on the scale of a few hundred by fl uidization or by pervasive fl ow along the not known, but is estimated to 10%–15% of the meters. We assume a hydrostatic pressure gradi- overpressure gradient. We account for fractur- sill thickness based on the exposed dike heights ent with a fl uid density of 1000 kg/m3 as initial ing of the host rock by resetting pressures that and sill thickness. conditions for pressure. The upper and lower exceed the tensile strength of the host rock to At site 1 (S32°18.2′, E26°52.6′), a sediment boundaries are fi xed according to initial hydro- hydrostatic pressure. We assume the tensile dike can be traced for ~150 m westward static pressure. The boundaries do not infl uence strength of our model sandstone host rock to from the main road, cutting vertically through at the calculations. be on average 35 MPa (Ai and Ahrens, 2004). least 15 vertical meters of dolerite. The strike is We expect a drop in overpressure gradients 80° east, and it pinches out in both directions. Model Assumptions with time, depending on how freely the mobi- The maximum thickness is 0.5 m and it splits We have developed a numerical model to lized sediments can move and reequilibrate the in two branches toward the west. Sediment and quantify the fi rst-order effects associated with overpressure anomalies. For the underpressure, dolerite fragments as much as 40 cm long are sill cooling and pressure evolution. The model we expect the assumption of prevented vol- common, and bridge-like portions of dolerite is conceptual and does not attempt to describe ume change to be valid for the intrusion until are present locally (Fig. 2E). The latter suggests the full system. We assume an instant emplace- the thermal contraction produces fracturing of an eastward direction of emplacement. ment model of the sill because sediment dikes the sill. Tensile strength of gabbroic rocks is Several thin sandstone dikes crop out at site are related to postemplacement processes >125 MPa (Ai and Ahrens, 2004). Such under- 2 (Fig. 2B). The maximum width is 0.5 m, the strike is 84° east, and their vertical extension can be traced for 10–15 m in the roadcut. A few pieces of fresh dolerite are located within the TABLE 1. SYMBOLS AND VALUES USED IN THE NUMERICAL MODEL dikes. These represent fragments of wall-rock Symbol Description Initial value Unit References dolerite broken off during dike emplacement, as z Vertical system size 500 m 1 also seen at site 1. d Sill thickness 100 m 1 At site 3, a vertical dike as thick as 2.2 m Tm Initial temperature of melt 1200 + 273 K 1 T Initial temperature of host rock 35 + 273 K 1 crops out along the road (Fig. 2C), striking hr 78° east. This is, to our knowledge, the thick- TL Liquidus temperature of melt 1200 + 273 K 1 est sediment dike ever found in a sill intrusion. TS Solidus temperature of melt 900 + 273 K 1 ρ) –6 2 –1 KT Thermal diffusivity melt (K/CP/ 10 m s 2 The contact with the dolerite is sharp, although χ / β) –3 2 –1 KHm Hydraulic diffusivity melt ( m µm/ 2.3 × 10 m s 2, 3 weathered, and it comprises a breccia with sedi- K Hydraulic diffusivity melt (χ /µ /β ) 3.7 × 10–2 2 –1 Hhr hr f f m s 2, 4 mentary fragments as long as 40 cm. Some of β –10 –1 Isothermal compressibility (fl uid/water) 4.3 × 10 Pa 2 the fragments show sedimentary layering. The α 4.1 × 10–4 K–1 2 Thermal expansion coeffi cient (water) lateral extension of the dike is unknown due to L Latent heat of crystallization 320 000 Jkg–1 4 poor exposures, but the dike is located ~10 m Ste Stefan number L/C /(T – T ) 0.27 1, 4 pm L S above the dolerite-sediment contact cropping P Pressure 1000 × g × z Pa 1 g Standard gravity 9.81 ms–2 out to the west. In addition to the sedimentary t Time0 s 1 clasts, the dike contains numerous fragments of Note : References: 1—this study; 2—Delaney (1982); 3—Hersum et al. (2005); dolerite (Fig. 2D). The dike has been mapped 4—Turcotte and Schubert (2002). in detail, and the results are presented in Figure

Geosphere, June 2010 213

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Svensen et al.

3A. The area percentage occupied by clasts pose 3.0% (70 clasts). There is a four order-of- clasts length and width is calculated and shown and their size distribution have been quantifi ed magnitude variation in clast size for the sedi- in Figure 3C. It is interesting that the aspect ratio (Fig. 3B). The results show that the sandstone mentary fragments, but a lesser variation for the is independent of the clast size. The sediment matrix (including clasts <0.5 cm) composes dolerite fragments. Note that the probability clasts are more elongated compared to the doler- 86.4% of the area, sediment fragments occupy versus size relationship is similar for both sedi- ite clasts (aspect ratios of 2.51 and 1.95, respec- 10.6% (320 clasts), and dolerite fragments com- ment and dolerite clasts. The aspect ratio of the tively), which is also evident from Figure 3A.

A SeSedimentdiment di dikeskes

SeSedimentd ments DoleriteDDoleeritet s sill R67

B C

SeSSedimentdid mentt DoDoleriteleer tee s silll ddikeke

DoDoleriteerir tee sis silll DoleriteDoD lerir tee s sill

SedimentSeS d mentt did dikeskes D E MeMeta-sandstoneta-sandsstot ne

DoDDoleriteeritee SeSedimentdiment didikeke DoDoleriteleerir tet

DoDoleriteler tee

Figure 2. The Waterdown Dam locality. (A) Overview of the locality, showing the transgressive dolerite sill and the roadcut along R67 with sediment dike localities. (B) Site 2, with sediment dikes that can be traced 10–15 vertical meters. (C) Site 3, with the >2-m-thick breccia dike within the dolerite. (D) Close-up of the dike at site 3, showing a dolerite fragment within the baked sandstone. Note the irregular fragment in the lower right, possibly representing altered magmatic material. Coin for scale. (E) The sediment dike at site 1. Note the sediment fragments and the dolerite bridge extending from the walls and into the dike. Hammer for scale.

214 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Sediment dikes in dolerite sills

A B Cover

Dolerite

Sediment

Cover C

Cover

Metamorphic sandstone Sediment Dolerite

0.5 meter

Figure 3. (A) Graphical representation of the rock fragments in the sediment breccia dike at site 3. The dike content was drawn on transparent plastic sheets (1:1 scale), scanned, and redrawn. Note the abundant dolerite fragments and sedi- mentary fragments with preserved sedimentary layering. (B) Image analysis shows that the rock fragments constitute 13.6% of the dike area. (C) The fi gure shows a scatter plot of the minor and major half-axis for ellipses fi tted to the frag- ments, with dolerite in red and sediment in gray. The plot suggests a linear trend between the minor and major half-axis, hence the aspect ratio is independent of the fragment size. The mean aspect ratios for dolerite and sediment fragments are 1.95 and 2.51, respectively. Very small fragments (<0.1 cm2 [i.e., 10 pixels]) are disregarded.

Elandsberg Roadcut (Nico Malan Pass) Golden Valley patterns (Fig. 5B). These nodules were The locality is located in the great escarp- The Golden Valley sill complex (Galerne originally composed of carbonate, but were ment defi ned by thick sill intrusions in the et al., 2008; Polteau et al., 2008a, 2008b) is modifi ed during metamorphism. Beaufort Group sediments. A sediment dike characterized by a fl at inner sill that is partly was found intruding into the lower contact of exposed along a small in the southern end. Sediment Petrography and Petrology the upper sill encountered when driving north Here (S31°58.4′, E26°16.4′) a well-exposed toward the Nico Malan Pass along the R67 part of the sill-roof hosts several small (<30 cm We studied thin sections of sediment dikes (S32°30.2′, E26°50.2′). The dike has pen- wide) sandstone intrusions (Fig. 5A). Note that from the Waterdown Dam (site 3), the Nico etrated 2.3 m into the inclined dolerite sill, the sediment source is located above the sill Malan Pass, and Golden Valley. The main aims and has a slightly curved and irregular shape contact, demonstrating downward sediment were to identify metamorphic , char- (Fig. 4A). The maximum width is ~20 cm, movement. Note that in general, downward acterize the texture, and characterize the meta- and the dike pinches out upward. No doler- sediment movement is not unique for this loca- morphic conditions. The diageneses of non- ite fragments were found in the dike, and the tion (e.g., Harms, 1965; Peterson, 1968; Vita- metamorphic sandstones located far from sill sandstone texture was markedly different at nage, 1954). The dikes are irregularly shaped, intrusions in the Karoo Basin are characterized the tip of the dike compared to the surrounding and are characterized by a network-like pattern. by authigenic minerals stable at relatively shal- contact-aureole sandstone, becoming increas- Brownish alteration haloes are common around low burial ( minerals, K-feldspar, , ingly recrystallized. No fl ow structures were the dikes. The intruded sediments are Beaufort albite, and quartz) (e.g., Rowsell and De Swardt, observed in the sediment beds below the dike Group sandstones and , where the sand- 1976; Svensen et al., 2008; Turner, 1972). Typi- or within the dike. stones contain abundant nodules with radial cally, detrital grains (like quartz and K-feldspar)

Geosphere, June 2010 215

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Svensen et al.

A 200200 µmµ m B

AlbiteAlbite AlbiteAlbite SandstoneSandstone ApatiteApatite dikedike QuartzQuartz

DoleriteDolerite BiotiteBiotite K-FeldsparK-Feldspar

QuartzQuartz Meta-sandstoneMeta-sandstone

Figure 4. The Elandsberg roadcut (Nico Malan Pass). (A) Sandstone dike extending ~2.5 vertical meters from the base of an ~100-m-thick dolerite sill. (B) Scanning electron microscope photograph showing metamorphic biotite, chlorite, and feldspar in the meta-sandstone from 2.3 m into the dike.

A B

SedimentSediment dikedike

DoleriteDolerite

QuartzQuartz 50 µm C 100100 µmµ m D QuartzQuartz QuartzQuartz GarnetGarnet PlagPlag PlagPlag ChlChl ApAp ZeoliteZeolite

IlmIlm AlAlb

QuartzQuartz ChlChl PlagPlag

Figure 5. The Golden Valley locality. (A) Network of sandstone dikes within the upper 1 m of a sill in the fl oor of the Golden Valley saucer. (B) Reaction in sandstone from 1 to 2 m above the contact with the sill. (C) Scanning electron microscope (SEM) photograph showing metamorphic plagioclase (plag), chlorite (Chl), and apatite (Ap) in a sample from the thickest dike shown in A. Ilm—ilmenite. (D) SEM photograph showing authigenic garnet, zeolite, quartz, and chlorite in the nodule in B. Alb—albite.

216 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Sediment dikes in dolerite sills

get coated and overgrown by authigenic miner- als during burial without affecting the composi- 200200 µmµ m tion or texture of the grain interiors. At 1 km of ChloriteChlorite A burial in the Karoo Basin, the original sandstone could have been ~10%–25%, presum- FspFsp ably fi lled with low-salinity pore fl uids. After contact metamorphism of sandstone within QuartzQuartz the sediment dikes, detrital components of the quartz grains are still easily recognized, whereas feldspar grains (plagioclase and K-feldspar) were recrystallized in mosaic patterns. More- over, the rock porosity is negligible, and chlo- QuartzQuartz FspFsp rite and biotite are commonly present. Further details of the effects of contact metamorphism of sandstone injections from the examined localities are given in the following. IlmIlm The sediment dike at Waterdown Dam con- tains metamorphic sandstone. Former grains 200200 µmµ m and grain boundaries, representing the original ChloriteChlorite B sedimentary components, are easily recognized FspFsp (Fig. 6A). This is confi rmed by CL imaging of quartz (Fig. 6B). In addition to quartz, the dom- inant minerals that recrystallized in the dike are QuartzQuartz K-feldspar, plagioclase, and chlorite. Detrital feldspar grains are recrystallized and contain QuartzQuartz a mosaic of K-feldspar and plagioclase. Based on the gray-scale variations on SEM backscat- FspFsp ter images, the plagioclase is characterized by several different compositions, apparently in textural equilibrium (Fig. 6C), and thus recrys- tallized during high-temperature metamorphic conditions. Biotite was not identifi ed in the studied sample, but abundant chlorite could possibly be a product of biotite retrogression. 100100 µmµm C The altered dolerite fragments in the dike are dominated by chlorite. QuartzQuartz The Elandsberg sandstone dike contains PlagPlag PlagPlag identifi able detrital sand grains with quartz overgrowths. A sample from 2.3 m into the dike was studied using SEM, where CL imaging ChloriteChlorite revealed detrital quartz cores. The presence of metamorphic epidote and biotite is important. The biotite is partly altered to chlorite, although K-fsKK-fsp-fsp the dominating mode of chlorite occurrence is in fresh patches unrelated to alteration. Feldspar grains are recrystallized and comprise mixtures of K-feldspar, albite, and plagioclase. Generally, QuartzQuartz the textures within the dike sample are tight and typical -like. The sandstone dikes from Golden Valley have Figure 6. Scanning electron microscope (SEM) petrogra- the same mineral content as the one at Water- phy from the sediment dike at Waterdown Dam, site 3. down Dam. Quartz, plagioclase, and chlorite are (A) The sandstone matrix shows well-defi ned quartz grains that the main minerals. One difference, however, is have survived metamorphism. Fsp—feldspar; Ilm—ilmenite. the plagioclase textures. In Golden Valley, the (B) The cathodoluminescence imaging demonstrates that melt- detrital plagioclase is apparently completely ing never took place, as the quartz grains have retained their recrystallized and zoned, present as tabular detrital core. (C) The feldspar grains are, however, pervasively crystals (Fig. 5C). Ilmenite and apatite are recrystallized, present as mosaics of K-feldspar (K-fsp) and minor minerals. The chlorite is locally present plagioclase (Plag) with various compositions. as tabular crystals, possibly suggesting biotite replacement. We have compared this mineral

Geosphere, June 2010 217

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Svensen et al.

assemblage with the assemblage within a for- the hydrostatic pressure gradient (~7–8 MPa) dispersed, and the main gradient is now from mer carbonate nodule with sandstone from the (Figs. 8A, 8B). A fracture pressure of 35 MPa is the host rock into the sill (Fig. 8D). The dif- same locality. The sample (Fig. 5B) is located indicated by a red dashed line in Figure 8B. The ference between the maximum overpressure ~2–3 m above the contact with the dolerite sill, sill is in a state of underpressure due to cooling (~4 MPa) in the host rock and underpressure and is characterized by radial fractures extend- and crystallization of interstitial melt in a solid in the sill (~–7 MPa) is ~10 MPa, relative to ing out from a zoned nodule. In thin section, the crystal network. The major mechanism of under- hydrostatic pressure. main minerals are quartz (with detrital cores pressure within the sill (−22 MPa relative to the and overgrowths), feldspar, chlorite, and zeo- hydrostatic pressure gradient) is due to a den- Fluid Pressure Evolution During lite (Fig. 5D). The boundaries between detrital sity change when interstitial melt (2600 kg/m3) Sediment Heating cores and metamorphic quartz are marked by is crystallizing (2900 kg/m3) within a confi ning rims of metamorphic garnet. The plagioclase is crystal network during cooling. Note that the The fl uid pressure in the aureole is increasing partly dissolved, and the pores fi lled by zeolite. tensile strength of a gabbroic rock is >125 MPa after sill emplacement due to the density change

Chlorite is common within the zeolite. (Ai and Ahrens, 2004). associated with heating of the H2O pore fl uid. The studied textures from the examined local- Assuming a pore fl uid pressure of 25 MPa at

ities show that the sandstone dikes underwent 100 yr After Emplacement ~1 km depth, H2O undergoes a density reduc- medium-temperature metamorphism following After 100 yr, the sill has solidifi ed and the tion from 1004 kg/m3 to 162 kg/m3 when heated injection. Original quartz grain boundaries and temperature gradients become less steep (Fig. from 35 °C to 400 °C (Wagner and Pruss, 2002). grain cores are still preserved and document 8C). Correspondingly, the pressure gradient The thermal expansion of the sediment matrix that the sediment dikes did not undergo partial anomalies are reduced through diffusive fl uid for the same temperature interval is negligi- melting after emplacement. This is consistent fl ow. The internal gradient within the sill is ble relative to the phase transition in the fl uid with the absence of macroscopic melt patches in the dikes. Diagnostic peak metamorphic minerals are sparse in metasandstones due to the low iron and magnesium content. Recrys- 300 tallization of quartz and feldspar grains is the +H2O dominant mode of mineralogical transforma- BioBio +Qtz tion. However, the occurrence of minerals like Chl hCrdhCrd chlorite, biotite, plagioclase, and epidote is Ms MsMs typical for greenschist conditions. Based 240 Ab PlagPlag on the general presence of these phases in the Ksp sediment dikes, we can use phase petrology to Chl Ep constrain the peak metamorphic conditions. We Ms Bio have made a phase diagram projected from a ChlChl Ab Chl Bio pelite composition, and compared the calcu- 180 MsMs Ksp Ms hCrd hCrd lated phase assemblages with those identifi ed in AbAb Zeo Plag San Plag the rocks in order to determine the temperature KspKsp Ep Plag San during dike emplacement (Fig. 7). Opx

(MPa) ZeoZeo

P Ol Fluid Pressure Evolution During 120 Sill Cooling

We have developed a numerical model in Chl order to calculate the pressure gradients devel- Ms W oping between an igneous sill and the sur- 60 Ab E rounding sedimentary rocks as a function of Ksp G temperature. Here we present snapshots of the Zeo temperature and pressure state during sill cool- ing. The modeling is based on the parameters listed in Table 1. 200 400 600 800 15 yr After Emplacement T (°C) At the time of instantaneous emplacement, the 100-m-thick sill is hot (1200 °C) with a Figure 7. Phase diagram calculated using Perplex for a rock with a pelite sharp thermal boundary to the cold host rock composition. The mineral assemblages are shown between the black (35 °C) (Fig. 8A). Note that the gradient will be lines (dehydration reactions), and the positions of the sediment dike similar even if the sill is emplaced by continu- samples are indicated by stars (W—Waterdown Dam; G—Golden Val- ous infi lling and infl ation. After 15 yr the tem- ley; E— Elandsberg). Chl—chlorite; Ms—muscovite; Ab—albite; Ksp— perature increases rapidly in the host rock, caus- K- feldspar; Zeo— zeolite; Ep—epidote; Bio—biotite; Plag—plagioclase; ing thermal expansion of the pore fl uids, which San—sanidine; Opx—orthopyroxene; Ol—olivine; Qtz—quartz; hCrd—– results in overpressure of ~22 MPa relative to high-cordierite.

218 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Sediment dikes in dolerite sills Temperature after 15 years Pressure after 15 years 600 600 AB

650 650

700 700 hydrostatic pressure gradient tensile strength solid melt 750 sill 750 Depth [m] 800 800

850 850

900 900 0 500 1000 -2 0 2 4 [°C] [Pa] 7 x 10 Temperature after 100 years Pressure after 100 years 600 600 C D

650 650

hydrostatic 700 700 pressure gradient tensile strength solid melt 750 sill 750 Depth [m] 800 800

850 850

900 900 0 500 1000 -2 0 2 4 7 [°C] [Pa] x 10 Figure 8. (A) Pressure gradients developed after 15 yr of sill cooling. The sill is still close to 100% molten (see vertical dashed line). (B) There is a strong pressure gradient between the sill margins and the aureole, where the aureole pressure is generated by thermal expansion of pore water. The arrows indicate pressure gradients along which melt and fl uids are expected fl ow. The tensile strength of dolerite is indicated. (C) The sill is solidi- fi ed after 100 yr of cooling. (D) Note that there is still a strong gradient from the underpressure at the margins to the overpressure in the aureole, but most of the overpressure that was generated previously has diffused away.

Geosphere, June 2010 219

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Svensen et al.

(i.e., boiling). The overpressure is released berg), injection after 300 yr of sill cooling is are heated around sills, pore fl uid expansion through diffusive fl ow, with rates depending on indicated. After 600 yr the sill has cooled to and boiling occur on a time scale of years, the permeability of the host rock. Similarly, the such an extent that the temperature in the sedi- dominating the fl uid production compared underpressure within the sill is created during ment dike never exceeds 350 °C (cf. Golden to devolatilization reactions (e.g., Delaney, cooling and crystallization. The thermal contrac- Valley and Waterdown Dam). 1982; Hanson, 1995; Jamtveit et al., 2004). tion associated with the melt to crystal transition Overpressure related to boiling and pore fl uid is several magnitudes larger than the thermal con- DISCUSSION expansion may ultimately lead to hydrofrac- traction of the surrounding network for the same turing and the formation of hydrothermal vent temperature interval. Hence, an underpressure Contact Metamorphism in complexes in the upper 1 km in the basin (e.g., will develop as a response to the density change Sedimentary Basins Jamtveit et al., 2004). In the Karoo Basin, the of interstitial melt (>55% crystals; Marsh, 1996; hydrothermal vent complexes commonly crop Philpotts and Carroll, 1996). This underpressure In contrast to the 30–70 m.y. time scale of out in the Stormberg Group sediments. In can be relaxed through internal melt fl ow from fl uid production and pressure buildup during addition, numerous breccia pipes are rooted in the molten to the crystallizing regions of the sill. regional metamorphism and orogenesis (e.g., contact aureoles of black shale, demonstrating When the sill is 100% crystallized, the thermal Connolly and Thompson, 1989; Walther and that high pore fl uid pressures developed dur- stresses will continue to develop as long as the Orville, 1982), contact metamorphism around ing rapid cracking of organic matter to meth- thermal contraction is larger than what can be igneous sill intrusions in sedimentary basins ane (Svensen et al., 2007). Thus contact meta- accommodated by volume change. The stresses have dramatic and short-term effects on fl uid morphism around sill intrusions is a process can be released through brittle fracturing of the fl ow. This is particularly important in basins that causes rapid pressure buildup and drives rocks, which in turn can be fi lled in by, for exam- with rapidly cooling sill intrusions compared fl uid fl ow on a very short time scale. In this ple, fl uids, interior melt, or fl uidized sediments to settings with >100 k.y. of contact meta- setting, sediment dikes represent direct evi- (e.g., Norton et al., 1984). morphism around plutons (e.g., Hanson, dence for the rapid release of aureole pressure When estimating the thermal expansion of 1992, 1995). When sedimentary host rocks and fl uids. pore fl uids in the aureole, we use a conservative coeffi cient value of 4 × 10–4 (Delaney, 1982), resulting in pressure anomalies to ~25 MPa. Using the defi nition of thermal expansion coef- α 1000 fi cient , 100 meter sill

1200 °C 1100 °C 1200 °C 1 ⎛⎞∂v 900 1100 °C 50 meter sill α = ⎜⎟, (1) ∂ vT⎝⎠P 800 where v is specifi c volume (per unit mass; 1/ρ; ρ is density), the expansion coeffi cient for pore 700 fl uid is 2.3 × 10–2 K–1, where boiling occurs, and for melt-to-crystal transition it is 3.5 × 10–4 K–1. 600 This study (W+G) The maximum pressure anomaly by boiling and expansion of pore fl uids may thus be as much as 500 two magnitudes larger than our estimates. 400 Thermal Modeling of Sediment Dikes This study (E)

Years after sill emplacement Years 300 We have made a thermal model with a real- istic sediment dike geometry to estimate the 200 maximum temperature attained within the Walton and O’Sullivan (1950) dike at a given sill temperature. We emplace a 100% solid 20-m-tall and 2-m-thick sandstone dike with an 100 melt initial temperature of 35 °C into a 100-m-thick sill with sill temperatures between 1100 and 0 100 200 300 400 500 600 700 800 900 1000 1200 °C (Fig. 9). As expected, the dike rapidly Max temperature in sandstone dike [°C] reaches peak temperature (i.e., within 1 yr). Hence, the initial temperature of the sandstone Figure 9. Calculated maximum (max) temperature of a 2 × 20 m sediment dike injected dike is not important for the fi nal maximum into a 100 m sill (solid lines) and a 1 × 10 m sediment dike injected into a 50 m sill temperature recorded in the sill. If the dike is (dashed lines) for intrusion temperatures of 1100–1200 °C as a function of injection time injected 15 yr after sill emplacement, the sedi- after sill emplacement. The gray area indicates when the sill is still molten. By knowing ment dike reaches a temperature of ~850 °C. the maximum temperature of the sandstone dike, we can infer that the injection time Injection at the time of sill solidifi cation (i.e., was 200–300 yr after sill emplacement for the Elandsberg dike (E) and ~600 yr for at ~100 yr), the peak temperature in the dike the Golden Valley (G) and Waterdown Dam (W) dikes. For comparison, the calculated is ~650–675 °C. For the sandstone dike to be injection time of the dikes from Walton and O’Sullivan (1950a) occurred closely after heated to a maximum of ~450 °C (cf. Elands- sill solidifi cation (~150 yr).

220 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Sediment dikes in dolerite sills

Pressure Evolution of a Cooling Sill been diffused by fl uid fl ow. Thus, the main pres- and result in substantial sediment displacement sure gradient is now from the host rock toward (e.g., Harms, 1965; Kokelaar, 1982; Ross and It has been shown that sill cooling and crys- the sill, both above and below the intrusion. At White, 2005; Vitanage, 1954). The sandstones tallization result in an underpressure within the this stage, heated pore fl uids will fl ow into the of the Beaufort Group in the Karoo Basin were sill (Aarnes et al., 2008). Underpressure gener- sill if permeability allows the fl uids to enter, i.e., still in the early to intermediate stages of dia- ation is caused by the following. At the earlier if fractures develop. genesis (i.e., reached quartz ) at stages of the sill cooling, a solid crystal net- the time of sill emplacement. Thus the condi- work (>55% crystals) with interstitial melt will Aureole Overpressure and Sediment tions were right for fl uidization to occur, at least form (Marsh, 1988, 1996; Philpotts and Car- Injections Into Sills where clay minerals limit relaxation of pres- roll, 1996). With further cooling the interstitial sures through fl uid fl ow (Jolly and Lonergan, melt undergoes a signifi cant density change Fluidization due to heating of water-rich sed- 2002), or as mentioned, if heating was rapid due to the melt-to-crystal transition. However, imentary rocks is most likely to occur at depths compared to pressure drop by fl uid fl ow (Jamt- a strong crystal network prevents a volume where pressure is less than the pressure corre- veit et al., 2004). change and causes a large underpressure to sponding to the critical point of water (Jamtveit Heat-induced overpressure and subsequent develop. Experiments have shown that a crys- et al., 2004; Kokelaar, 1982). The paleodepth of fl uidization of sediments in the contact aure- tal network have considerable strength already the study areas with sediment dikes in dolerites ole is here suggested to be the main formation at 35% crystals, and effectively behaves as a is ~600–900 m, thus shallower than the critical mechanism of sandstone dikes in magmatic solid even with large amounts of interstitial depth. In some geological settings overpressure intrusions. We show that there is an additional melts (Philpotts and Carroll, 1996). Such an can cause horizontal fracturing through fl uids strong gradient from the aureole into the intru- underpressure may induce melt fl ow, have con- seeping away from the overpressurized source sion, and that this gradient makes sediment sequences for the chemistry of the magmatic (e.g., Cobbold and Rodrigues, 2007; Mourgues mobilization more likely to happen compared to system, and induce sediment injections into the and Cobbold, 2003), while in the case of boil- injections driven by pore fl uid boiling and frac- sill (Aarnes et al., 2008). ing and very high overpressures, modeling has turing during thermal contraction. However, we During the initial stages of sill cooling, the demonstrated that the gas release may localize argue that fracturing during thermal contrac- pore water in the aureole sediments will expand vertically and eventually reach the atmosphere tion is of lesser importance, as sediment dikes and fl ow either away from the sill or into the (e.g., Jamtveit et al., 2004; Rozhko et al., 2007). are not present in a hexagonal network even in sill, depending on the pressure gradients. Melt The key requirements for pressure-induced areas with abundant fractures developed dur- may also fl ow within the sills along the pressure sediment mobilization in the aureole are low ing thermal contraction (e.g., the Golden Valley gradient toward the cooling margins (cf. Fig. permeabilities, high , and high ther- locality). The overpressure scenario is sche- 8). The fl uid fl ow is a result of the developed mal diffusivities (Delaney, 1982; Jamtveit et matically presented in Figure 10. Our results pressure anomalies and will act to even out the al., 2004). In the case of high permeabilities show pressure anomalies of as much as 108 Pa pressure anomalies with time. After 100 yr the in shallow sandstones, the rate of heating must after solidifi cation of the sill, in agreement with pressure gradient within the sill is reversed, exceed the rate of pressure loss by fl uid fl ow the magnitude 107 Pa overpressure commonly going from the margins to the center (Fig. 8B). in order to build up signifi cant overpressure. found for several rock types due to expansion However, the melt is now unable to fl ow as When the sedimentary host rock undergoes of pore fl uids from magmatic intrusion (Del- solidifi cation is complete. At this time, the pres- extensive pressure buildup, it may ultimately aney, 1982). It is important that the high pres- sure in the sedimentary host rock has effectively lose all cohesive strength and become fl uidized sure is suffi cient to break the tensile strength of

t = emplacement P T t = 1-10 years t = 250-600 years ABC

Sandstone

Sill intrusion Melt+crystals 100% crystals meters (100% melt)

100

Sandstone

Figure 10. Schematic evolution of a sill-aureole system with sediment injections. (A) Initial sill emplace- ment into cold sedimentary host rocks. (B) Contact metamorphism around the molten sill, and expansion of sedimentary pore fl uids (symbolized by circles). (C) Crystallization of the sill followed by fracturing due to the huge pressure (P) difference between sill and aureole. The sill is still hot enough to cause high- temperature (T) metamorphism of the injected sediments.

Geosphere, June 2010 221

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Svensen et al.

sandstones above ~1 km depth (e.g., Kokelaar, never exceed about half the sill temperature, a stone and clasts of sediments and dolerite. 1982), thus fl uids can potentially fl ow from the doleritic sill (~1200 °C) will commonly not be Field, petrographic, and numerical evidence aureole and into the sill. able to melt the host sediments, and maximum suggests the following. We therefore argue that sandstone dikes form temperatures should be close to 600–700 °C, Both upward and downward movement of as a result of the difference in pressure between depending upon the host-rock temperature at sediments into sill intrusions is common. the sill and the aureole (~10 MPa) that develops the time of emplacement. However, this situ- The sediments intruded while the sills were during sill cooling and sediment contact meta- ation may be different in other geological sys- hot, producing mineral assemblages typical for morphism. The pressure gradient is suffi cient tems (e.g., Hersum et al., 2007). >300 °C metamorphism. for fracturing the sill (pressures beyond the The temperature estimates from the mineral- Thermal modeling, to account for the dike lithostatic) and to act as a suction force on the ogy are of importance when assessing the tim- metamorphism, shows that the sediment sediments from the moment the chilled margin ing of sediment dike emplacement. As we have dikes were injected more than 100 yr after sill of the sill fractures. Once initiated, the fracture shown, an early emplacement into a hot sill will emplacement, depending on sill thickness and will propagate as a result of the injected pore result in high-temperature metamorphism in the the initial sill temperature. fl uids and sediments; this also may lead to fur- dike. Based on our thermal modeling, injec- The presence of sediment dikes in sills is a ther tensile failure (cf. Rubin, 1993). The frac- tion after 250–600 yr of sill solidifi cation will result of the coupled pressure evolution of dol- turing process may be violent, as indicated by give 325–450 °C in the dike. Note that reaction erite sills and contact aureoles. Negative pres- the high proportion of both sedimentary and kinetics or signifi cant latent heat of vaporization sure anomalies in the sill form due to cooling, doleritic rock fragments in the dikes at Water- may contribute to discrepancies between mod- whereas high pressure develops in the aureole down Dam. Sediment fragments compose 86% eled heat from conduction and that of a natural due to thermal expansion. of the dike surface at site 3, and the size distri- system. Earlier timing of sediment injection is The pressure generated is of the correct order bution between sedimentary and dolerite clasts therefore possible. of magnitude required to explain fracturing of suggests that the same process was responsible To summarize, our data suggest that the the solidifi ed sill. The sediments were accord- for brecciation of both dolerite sill and aureole emplacement of the sediment dikes occurred ingly drawn into the sill. sediments. The four orders of magnitude varia- after the sill was 100% crystallized, which tion in clast size (Fig. 3B) demonstrate that the puts a lower boundary to the timing of injec- APPENDIX brecciation was rapid and that the bulk of the tion of ~100 yr. breccia was injected into the sill. This means that sediment injection into sills Equations has only limited potential for contaminating the The cooling of the sill and heating of the host rock Sediment Dike Metamorphism and magma, since the sill is 100% crystallized at follow the heat conduction equation: Injection Timing the time of sandstone injection. For contamina- tion to happen, the sediments would have to be ∂∂∂TTT⎧⎫22 The sediment dikes described in this study injected into a partly molten sill, for which we =+K eff ⎨⎬, (A1) ∂∂∂T 22 are all affected by contact metamorphism. have no supporting observations. txz⎩⎭ Thus they were heated while the sill intrusions Field evidence shows that sediment dikes can where T is the temperature, x is lateral direction, z is were still hot, either in situ in the contact aure- propagate tens of meters into dolerite sills from eff vertical direction, and KT is the effective thermal ole prior to injection, or within the sediment the lower contact. The vertical termination of diffusivity coeffi cient (λ / C / ρ), and λ is the thermal dike. Metamorphism of the injected sediments dikes has, however, not been found in the fi eld. P ρ conductivity, CP is heat capacity and is density. The is a common observation from all sediment However, since the metamorphic recrystalliza- effective thermal diffusivity accounts for the latent dikes in magmatic sill intrusions (Van Biljon tion led to very low permeabilities, the sediment heat of fusion: and Smitter, 1956; Walton and O’Sullivan, dikes were prevented from becoming long- K 1950). Based on the metamorphic minerals in lasting fl uid fl ow pathways. eff T KforTTTTSL=<<() the dikes and aureoles from the Karoo Basin The basin settings in which sediment dikes (1+ Ste ) (A2) eff (chlorite, biotite, plagioclase, epidote, and gar- within igneous sills are not likely to form are KKTT=> forTT() S. net), the metamorphic conditions were equiva- (1) when the overpressure difference between lent to those of the greenschist facies. Based sill and aureole is small, as when the sill intru- The nondimensional ratio quantifying the effect of the on these minerals and the phase diagram (Fig. sion is thin, or (2) the aureole has limited poten- latent heat is the Stefan number, Ste, given by: 7), a maximum temperature of ~450 °C is sug- tial for generating overpressure during heating, gested. There are no accurate thermometers e.g., when the porosity is very low or the content L that can be applied to the identifi ed mineral of organic matter is negligible. Thus the pres- Ste = , (A3) ()TTC− assemblages, so the temperature is approxi- ence of sediment injections in igneous systems LSP mate. Comparing with active hydrothermal may provide important constraints on the pres- where C is heat capacity and L is the latent heat of metamorphism of sandstone, biotite appears at sure evolution and fl uid fl ow history in sedimen- P fusion per unit mass. ~320 °C (e.g., Schiffman et al., 1985), so our tary basins with sill intrusions. Equations A1–A3 are coupled with pressure estimate is reasonable. The absence of minerals through thermal stresses, like cordierite, clinopyroxene, and muscovite CONCLUSIONS furthermore suggests temperatures <~450 °C, α although the potential for generating some Sediment dikes have been discovered dP= dT , (A4) β of these minerals depends on the bulk rock within dolerite sill intrusions at several locali- composition. As the temperature of heated ties in the Karoo Basin in South Africa. The as described by, e.g., Turcotte and Schubert (2002), sedimentary rocks around a sill intrusion will sediment dikes contain metamorphic sand- assuming isochoric conditions for crystallization.

222 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Sediment dikes in dolerite sills

Taking the partial derivative of equation A4 with during prograde regional metamorphism: Journal Marsh, B.D., 1988, Crystal capture, , and reten- respect to time, the hydraulic equation becomes: of Geophysical Research, v. 102, p. 18,149–18,173, tion in convecting magma: Geological Society doi: 10.1029/97JB00731. of America Bulletin, v. 100, p. 1720–1737, doi: ∂∂∂∂PPPT⎧⎫22α Connolly, J.A.D., 2005, Computation of phase equilibria by 10.1130/0016-7606(1988)100<1720:CCSARI> =++K linear programming: A tool for geodynamic model- 2.3.CO;2. H ⎨⎬22 , (A5) ∂∂∂∂txzt⎩⎭β ing and its application to zone decarbon- Marsh, B.D., 1996, Solidifi cation fronts and magmatic evo- ation: Earth and Planetary Science Letters, v. 236, lution: Mineralogical Magazine, v. 60, p. 5–40, doi: p. 524–541, doi: 10.1016/j.epsl.2005.04.033. 10.1180/minmag.1996.060.398.03. where P is pressure, α is the volumetric coeffi- Connolly, J.A.D., and Thompson, A.B., 1989, Fluid and Mazzini, A., Jonk, R., Duranti, D., Parnell, J., Cronin, B.T., cient of thermal expansion and β is the isothermal enthalpy production during regional metamor- and Hurst, A., 2003, Fluid escape from reservoirs: compressibility. phism: Contributions to Mineralogy and Petrology, Implications from cold seeps, fractures and injected v. 102, p. 347–366, doi: 10.1007/BF00373728. . Part I: The fl uid fl ow system: Journal of χ Delaney, P.T., 1982, Rapid intrusion of magma into wet Geochemical Exploration, v. 78, p. 293–296, doi: K = , (A6) rock: Groundwater fl ow due to pore pressure 10.1016/S0375-6742(03)00046-3. H μβ increases: Journal of Geophysical Research, v. 87, McElwain, J.C., Wade-Murphy, J., and Hesselbo, S.P., p. 7739–7756, doi: 10.1029/JB087iB09p07739. 2005, Changes in carbon dioxide during an oceanic Duncan, R.A., Hooper, P.R., Rehacek, J., Marsh, J.S., and anoxic event linked to intrusion into χ where KH is the hydraulic diffusivity, is matrix Duncan, A.R., 1997, The timing and duration of the : , v. 435, p. 479–482, doi: 10.1038/ permeability, and µ is viscosity of fl uid. This modi- Karoo igneous event, southern Gondwana: Journal nature03618. fi ed hydraulic diffusion equation is similar to that of of Geophysical Research, v. 102, p. 18,127–18,138, Mourgues, R., and Cobbold, P.R., 2003, Some tectonic Delaney (1982). The fi rst part on the right side of doi: 10.1029/97JB00972. consequences of fl uid overpressures and seepage equation A5 describes the pressure diffusion (simi- Galerne, C.Y., Neumann, E.R., and Planke, S., 2008, forces as demonstrated by sandbox modelling: Tec- Emplacement mechanisms of sill complexes: Infor- tonophysics, v. 376, p. 75–97, doi: 10.1016/S0040 lar to heat conduction equation A1); the second part mation from the geochemical architecture of the -1951(03)00348-2. describes the development of pressure anomalies due Golden Valley Sill Complex, South Africa: Journal Nichols, R.J., Sparks, R.S.J., and Wilson, C.J.N., 1994, to changes in temperature. The initial overpressure is of Volcanology and Geothermal Research, v. 177, Experimental studies of the fl uidization of lay- zero, because the fl ow only depends on the evolving p. 425–440, doi: 10.1016/j.jvolgeores.2008.06.004. ered sediments and the formation of fl uid escape pressure anomalies. Ganino, C., and Arndt, N.T., 2009, Climate changes caused structures: , v. 41, p. 233–253, doi: by degassing of sediments during the emplacement 10.1111/j.1365-3091.1994.tb01403.x. ACKNOWLEDGMENTS of large igneous provinces: Geology, v. 37, p. 323– Norton, D., Taylor, H.P., and Bird, D.K., 1984, The geom- 326, doi: 10.1130/G25325A.1. etry and high-temperature brittle deformation of Hanson, R.B., 1992, Effects of fl uid production on fl uid-fl ow the Skaergaard Intrusion: Journal of Geophysical This study was supported by a PetroMaks grant during regional and contact metamorphism: Journal Research, v. 89, p. 178–192. from the Norwegian Research Council to Svensen. of Metamorphic Geology, v. 10, p. 87–97, doi: Peterson, G.L., 1968, Flow structures in sandstone dikes: We thank Goonie Marsh and Luc Chevallier for dis- 10.1111/j.1525-1314.1992.tb00073.x. Sedimentary Geology, v. 2, p. 177–190, doi: cussions during our fi eld trips to South Africa, in par- Hanson, R.B., 1995, The hydrodynamics of con- 10.1016/0037-0738(68)90024-9. ticular Goonie for showing us the Waterdown Dam tact metamorphism: Geological Society of Philpotts, A.R., and Carroll, M., 1996, Physical properties America Bulletin, v. 107, p. 595–611, doi: of partly melted tholeiitic : Geology, v. 24, locality, and Dirk Liss for the company and assistance 10.1130/0016-7606(1995)107<0595:THOCM> p. 1029–1032, doi: 10.1130/0091-7613(1996)024 during sampling of the sediment dikes. Else-Ragnhild 2.3.CO;2. <1029:PPOPMT>2.3.CO;2. Neumann and the Golden Valley Study Group at PGP Harms, J.C., 1965, Sandstone dikes in relation to Laramide Podladchikov, Y.Y., and Wickham, S.M., 1994, Crystalliza- (Physics of Geological Processes, University of Oslo) faults and stress distribution in the south- tion of hydrous —Calculation of associated contributed with valuable input to the project. We also ern Front Range, Colorado: Geological Soci- thermal effects, volatile fl uxes, and isotopic altera- thank Joe Cartwright and an anonymous referee for ety of America Bulletin, v. 76, p. 981–1002, doi: tion: Journal of Geology, v. 102, p. 25–45, doi: critical comments. 10.1130/0016-7606(1965)76[981:SDIRTL]2.0.CO;2. 10.1086/629646. Hersum, T., Hilpert, M., and Marsh, B., 2005, Permeabil- Polteau, S., Ferre, E.C., Planke, S., Neumann, E.R., and ity and melt fl ow in simulated and natural partially Chevallier, L., 2008a, How are saucer-shaped sills REFERENCES CITED molten basaltic magmas: Earth and Planetary Sci- emplaced? Constraints from the Golden Valley Sill, ence Letters, v. 237, p. 798–814, doi: 10.1016/j.epsl South Africa: Journal of Geophysical Research, Aarnes, I., Podladchikov, Y.Y., and Neumann, E.R., 2008, .2005.07.008. v. 113, B12104, doi: 10.1029/2008JB005620. Post-emplacement melt fl ow induced by thermal Hersum, T.G., Marsh, B.D., and Simon, A.C., 2007, Contact Polteau, S., Mazzini, A., Galland, O., Planke, S., and Malthe- stresses: Implications for differentiation in sills: partial melting of granitic , melt segre- Sorenssen, A., 2008b, Saucer-shaped intrusions: Earth and Planetary Science Letters, v. 276, p. 152– gation, and re-injection as dikes into Ferrar dolerite Occurrences, emplacement and implications: Earth 166, doi: 10.1016/j.epsl.2008.09.016. sills, McMurdo Dry Valleys, Antarctica: Journal and Planetary Science Letters, v. 266, p. 195–204, Ai, H.A., and Ahrens, T.J., 2004, Dynamic tensile strength of Petrology, v. 48, p. 2125–2148, doi: 10.1093/ doi: 10.1016/j.epsl.2007.11.015. of terrestrial rocks and application to - petrology/egm054. Retallack, G., and Jahren, A.H., 2008, release ing: Meteoritics & Planetary Science, v. 39, p. 233– Jamtveit, B., Svensen, H., Podladchikov, Y., and Planke, from of coal during Late Perm- 246, doi: 10.1111/j.1945-5100.2004.tb00338.x. S., 2004, Hydrothermal vent complexes associated ian extinction events: Journal of Geology, v. 116, Caddick, M.J., and Thompson, A.B., 2008, Quantifying with sill intrusions in sedimentary basins, in Breit- p. 1–20, doi: 10.1086/524120. the tectono-metamorphic evolution of pelitic rocks kreuz, C., and Petford, N., eds., Physical geology of Ross, P.-S., and White, J.D.L., 2005, Unusually large clas- from a wide range of tectonic settings: Mineral high-level magmatic systems: Geological Society tic dykes formed by elutriation of a poorly sorted, compositions in equilibrium: Contributions to of London Special Publication 234, p. 233–241. coarser-grained source: Geological Society of Lon- Mineralogy and Petrology, v. 156, p. 177–195, doi: Johnson, M.R., Van Vuuren, C.J., Visser, J.N.J., Cole, D.I., don Journal, v. 162, p. 579–582, doi: 10.1144/0016 10.1007/s00410-008-0280-6. H. de V. Wickens, Cristie, A.D.M., and Roberts, -764904-127. Catuneanu, O., Hancox, P.J., and Rubidge, B.S., 1998, D.L., 1997, The foreland Karoo basin, South Africa, Rowsell, D.M., and De Swardt, A.M.J., 1976, Reciprocal fl exural behaviour and contrasting stra- in Selley, R.C., ed., African basins: Sedimentary in Cape and Karoo sediments, South Africa, and tigraphies: A new basin development model for the Basins of the World Volume 3: Amsterdam, Else- its bearing on their hydrocarbon potential: Geo- Karoo retroarc foreland system, South Africa: Basin vier, p. 269–318. logical Society of South Africa Transactions, v. 79, Research, v. 10, p. 417–439, doi: 10.1046/j.1365 Jolly, R.J.H., and Lonergan, L., 2002, Mechanisms and con- p. 81–145. -2117.1998.00078.x. trols on the formation of sand intrusions: Geologi- Rozhko, A.Y., Podladchikov, Y.Y., and Renard, F., 2007, Chevallier, L., and Woodford, A., 1999, Morpho- cal Society of London Journal, v. 159, p. 605–617, Failure patterns caused by localized rise in pore- and mechanism of emplacement of the dolerite doi: 10.1144/0016-764902-025. fl uid overpressure and effective strength of rocks: rings and sills of the western Karoo, South Africa: Jourdan, F., Feraud, G., Bertrand, H., Kampunzu, A.B., Geophysical Research Letters, v. 34, L22304, doi: South African Journal of Geology, v. 102, p. 43–54. Tshoso, G., Watkeys, M.K., and Le Gall, B., 2005, 10.1029/2007GL031696. Cobbold, P.R., and Rodrigues, N., 2007, Seepage forces, Karoo large igneous province: Brevity, origin, Rubin, A.M., 1993, Tensile fracture of rock at high confi ning important factors in the formation of horizon- and relation to mass extinction questioned by new pressure. Implications for dike propagation: Journal tal hydraulic fractures and bedding-parallel 40Ar/39Ar age data: Geology, v. 33, p. 745–748, doi: of Geophysical Research, v. 98, p. 15,919–15,935, fi brous veins (‘beef’ and icone-in-cone): Geofl u- 10.1130/G21632.1. doi: 10.1029/93JB01391. ids, v. 7, p. 313–322, doi: 10.1111/j.1468-8123 Kokelaar, B.P., 1982, Fluidization of wet sediments during Schiffman, P., Bird, D.K., and Elders, W.A., 1985, Hydro- .2007.00183.x. the emplacement and cooling of various igneous thermal mineralogy of sandstones Connolly, J.A.D., 1997, Devolatilization-generated fl uid bodies: Geological Society of London Journal, from the Colorado in the Cerro Pri- pressure and deformation-propagated fl uid fl ow v. 139, p. 21–33, doi: 10.1144/gsjgs.139.1.0021. eto geothermal system, Baja California, Mexico:

Geosphere, June 2010 223

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021 Svensen et al.

Mineralogical Magazine, v. 49, p. 435–449, doi: chemistry as a tracer of fl uid fl ow in a hydrothermal margin of Gondwanaland: Geological Society of 10.1180/minmag.1985.049.352.14. vent complex in the Karoo Basin, South Africa: America Memoir 184, p. 223–279. Smith, R.M.H., 1990, A review of stratigraphy and sedi- Geochimica et Cosmochimica Acta, v. 72, p. 4929– Vitanage, P.W., 1954, Sandstone dikes in the South Platte mentary environments of the Karoo Basin of South 4947, doi: 10.1016/j.gca.2008.07.023. area, Colorado: Journal of Geology, v. 62, p. 493, Africa: Journal of African Earth Sciences, v. 10, Svensen, H., Planke, S., Polozov, A.G., Schmidbauer, N., doi: 10.1086/626193. p. 117–137, doi: 10.1016/0899-5362(90)90050-O. Corfu, F., Podladchikov, Y.Y., and Jamtveit, B., Wagner, W., and Pruss, A., 2002, The IAPWS formulation Svensen, H., Planke, S., Malthe-Sorenssen, A., Jamtveit, 2009, Siberian gas venting and the end-Permian 1995 for the thermodynamic properties of ordinary B., Myklebust, R., Rasmussen Eidem, T., and Rey, environmental crisis: Earth and Planetary Sci- water substance for general and scientifi c use: Jour- S.S., 2004, Release of methane from a volcanic ence Letters, v. 277, p. 490–500, doi: 10.1016/j nal of Physical and Chemical Reference Data, v. 31, basin as a mechanism for initial Eocene global .epsl.2008.11.015. p. 387–535, doi: 10.1063/1.1461829. warming: Nature, v. 429, p. 542–545, doi: 10.1038/ Turcotte, D., and Schubert, G., 2002, Geodynamics (second Walther, J.V., and Orville, P.M., 1982, Volatile production nature02566. edition): Cambridge, UK, Cambridge University and transport in regional metamorphism: Contribu- Svensen, H., Jamtveit, B., Planke, S., and Chevallier, L., Press, 456 p. tions to Mineralogy and Petrology, v. 79, p. 252– 2006, Structure and evolution of hydrothermal vent Turner, B.R., 1972, Silica diagenesis in the Molteno sand- 257, doi: 10.1007/BF00371516. complexes in the Karoo Basin, South Africa: Geo- stone: Geological Society of South Africa Transac- Walton, M.S., and O’Sullivan, R.B., 1950, The intrusive logical Society of London Journal, v. 163, p. 671– tions, v. 75, p. 55–66. mechanics of a clastic dike: American Journal of 682, doi: 10.1144/1144-764905-037. Van Biljon, W.J., and Smitter, Y.H., 1956, A note on the Science, v. 248, p. 1–21. Svensen, H., Planke, S., Chevallier, L., Malthe-Sorenssen, occurrence of two sandstone dykes in a Karoo A., Corfu, F., and Jamtveit, B., 2007, Hydrother- dolerite sill near Devon, South-Eastern Transvaal: mal venting of greenhouse gases triggering Early Geological Society of South Africa Transactions, Jurassic global warming: Earth and Planetary Sci- v. 59, p. 135–139. ence Letters, v. 256, p. 554–566, doi: 10.1016/j Veevers, J.J., Cole, D.I., and Cowan, E.J., 1994, Southern .epsl.2007.02.013. Africa: Karoo Basin and , in Veevers, MANUSCRIPT RECEIVED 17 FEBRUARY 2009 Svensen, H., Bebout, G., Kronz, A., Li, L., Planke, S., Che- J.J., and Powell, C.M., eds., Permian–Triassic Pan- REVISED MANUSCRIPT RECEIVED 09 NOVEMBER 2009 vallier, L., and Jamtveit, B., 2008, Nitrogen geo- gean basins and foldbelts along the Panthalassan MANUSCRIPT ACCEPTED 28 DECEMBER 2009

224 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/211/3338840/211.pdf by guest on 30 September 2021