Magmatic Differentiation Processes in Saucer-Shaped Sills: Evidence from the Golden Valley Sill in the Karoo Basin, South Africa

Home , Crystal mush

Magmatic differentiation processes in saucer-shaped sills: Evidence from the Golden Valley Sill in the Karoo Basin, South Africa

Christophe Y. Galerne* PGP—Physics of Geological Processes, University of Oslo, P.O. Box 1048 Blindern, 0316 Oslo, Norway, and Steinman Institute, Department of Geophysics, University of Bonn, Nussallee 8, D-53115 Bonn, Germany Else-Ragnhild Neumann Ingrid Aarnes PGP—Physics of Geological Processes, University of Oslo, P.O. Box 1048 Blindern, 0316 Oslo, Norway Sverre Planke PGP—Physics of Geological Processes, University of Oslo, P.O. Box 1048 Blindern, 0316 Oslo, Norway, and VBPR—Volcanic Basin Petroleum Research, Forskningsparken, Oslo, Norway

ABSTRACT 1. INTRODUCTION crystallization. The most recurrent geochemical profi les observed in such sills are I-, D-, and Analysis of compositional variations Bulk composition profi les of basic-ultrabasic S-shaped (Fig. 1; e.g., Gibb and Henderson, along profi les from tholeiitic sills provides sills have been investigated for a long time in 1992; Latypov, 2003a). The nomenclature is insights into syn- and post-emplacement order to understand the magmatic differentiation based on the variations in the whole-rock Mg#

magmatic differentiation processes. We processes associated with their emplacement and ( = cation proportion 100 × Mg/[Mg + Fetotal]) present here 18 whole-rock compositional profi les sampled from a saucer-shaped sill emplaced in the Karoo Basin (South Africa), I-shaped C-shaped D-shaped S-shaped the Golden Valley Sill. We show that different compositional profi le patterns pre- Top Top Top Top viously described in basic-ultrabasic sills may be found in different parts of a single saucer-shaped sill. The detailed examination of the mineral grain assemblage and compositions suggests that processes taking place in hundred-meter-thick sills relate to Sill thickness Sill thickness Sill thickness Sill thickness early and late fractional crystallization. Our observations in the Golden Valley Sill suggest that a signifi cant part of fractionation Base Base Base Base

takes place at a late stage of cooling when a Mg# Mg# Mg# Mg# crystalline skeleton or mush zone is formed. We show that porous fl ow of interstitial melt Uniform Positive segregation Negative segregation driven by forces related to the particular profile shape geometry (saucer-shaped) of the sill may Non-uniform profile shapes result in a post-emplacement compositional evolution. We propose that the process of Figure 1. Characteristic shapes of compositional profi les defi ned on the basis of Mg# (dif-

post-emplacement melt fl ow regionally ferentiation index, Mg# = molar 100 ×Mg/[Mg + Fetotal]) variations along vertical sections overprinted compositional patterns pro- across igneous intrusives. The I-shaped profi le is a uniform profi le showing no chemical duced by earlier crystal segregation from variation across the sill thickness. Nonuniform profi le patterns in relatively thin sheet intru- the cooling magma at fl uid-like stages dur- sions (100 m thick, e.g., Golden Valley Sill Complex) are characterized by D- and S-shaped ing the emplacement. compositional profi les also qualifi ed as reversed or inverse segregation. C-shaped compositional profi les are characteristics of large magmatic plutons, also qualifi ed as positive (i.e., expected) segregation.

*[email protected].

Geosphere; June 2010; v. 6; no. 3; p. 163–188; doi: 10.1130/GES00500.1; 18 ﬁ gures; 2 tables; 3 supplemental tables.

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

from fl oor to roof of the sills. The term I-shaped along the profi les. The profi les are classifi ed on 182.7 ± 0.3 Ma of a sill in the GVSC (Svensen profi le is used to describe sills showing little the basis of the established nomenclature, i.e., et al., 2007) implies that these sills are contem- fi eld, petrographic, and geochemical evidence I-, D-, and S-shaped profi les. Profi les that did poraneous with the Lesotho Flood basalts which of differentiation throughout the height of the not fi t into the established nomenclature were are dated to ca. 183 Ma (Duncan et al., 1997). sill (Mangan and Marsh, 1992; Marsh, 1996). termed X-shaped profi les. Mineral chemical Sill complexes in the Karoo Basin represent a The name D-shaped profi le was introduced to variations are illustrated for one profi le. The last signifi cant portion of the Jurassic Karoo igne- characterize sills with the least-differentiated section gives a review of the main results and ous province (Galerne et al., 2008; Marsh and composition at the sill center (i.e., highest discusses the various types of compositional Mndaweni, 1998). Saucer-shape sills are com- Mg#) and most evolved composition at the profi les found in the Golden Valley Sill in terms mon structures in sill complexes (Hansen and sill margins (i.e., lowest Mg#). Finally, an of cooling and crystallization processes. Cartwright, 2006; Hansen et al., 2004; Hansen S-shaped profi le shows S-shaped variations in et al., 2008; Polteau et al., 2008b; Thomson and Mg# upwards through the sill (Frenkel et al., 2. GEOLOGY AND GEOCHEMICAL Hutton, 2004). A saucer-shaped sill is character- 1988, 1989; Fujii, 1974; Gray and Crain, 1969; BACKGROUND OF THE GOLDEN ized by a fl at inner sill connected outward and Marsh, 1989). VALLEY SILL COMPLEX upward to a ring of discordant inclined sheets, D- and S-shaped profi les observed in rela- which are often terminated in a fl at outer sill. tively thin sills (100 m) pose a problem because The GVSC consists of a group of sills and The GVSC has intruded the Beaufort Group they are “inversed” from what is expected of dykes emplaced in an undeformed sequence of in the Karoo Basin sedimentary sequences. cooling magma. Large layered mafi c intru- sandstone and shale (Fig. 2). A concordia age of Like numerous sills in the Beaufort Group, the sions have typically mafi c margins and felsic cores (normal zoning) forming C-shaped geochemical profi les (Fig. 1; Skaergaard intrusion, 26°10 E 26°20 E 26°30 E 26°40 E Naslund, 1984). C-shaped zoning or composi- A' A tional profi le is interpreted as the result of in situ processes of fractional crystallization (Rice, Glen -31°80 N 1981). The lack of C-shaped profi les in rela- Sill tively thin sills is signifi cant and suggests that MV Sill (GS) processes occurring in a large magmatic body (MVS) are different from those in thin sills (100 m). An additional problem is that available data are Harmony generally restricted to one to three profi les in Sill a sill (e.g., Richardson, 1979). Consequently, Golden Valley Dyke (GVD)Golden (HS) we usually do not know if a single sill contains A -31°90 N different types of compositional profi les and Valley if there are variations in type of profi le along Sill the boundary of sills. Such information is, (GVS) N profiles however, of fi rst-order importance in order to P17 understand the magmatic differentiation pro- Morning cesses that occur in sills. P18 Sun (MS) Our strategy in this study was to document the vertical and lateral compositional and tex- Km -32°00 N tural variations within a single well-exposed 0 saucer-shaped sill in order to reveal the mag- 4 matic differentiation processes within this sill. B 26°S Our study object was the Golden Valley Sill o o 30026 Km 10 E SOUTH 26 20 E 26°30 E 26°40 E in the Golden Valley Sill Complex (GVSC), AFRICA Karoo Basin, South Africa (Aarnes et al., 2008; A Morning Glen Sill A' KAROO Galerne et al., 2008; Polteau et al., 2008a). The 30°S Sun (MS) Golden Valley (GS) BASIN excellent exposure of the Golden Valley Sill Sill (GVS) allowed systematic sampling along vertical pro- Harmony GVSC fi les all along the inclined sheets/outer sill. We Morning ? Sill (HS) 34°S GVD ? ? ? report here whole-rock analyses on 18 profi les 18°E22°E 26°E 27°E Sun (MS) (Fig. 2A; fi ve samples per profi le in average) together with mineral chemistry on representa- Figure 2. Summary of the geochemical architecture of the Golden Valley Sill Complex tive profi les. (GVSC) after Galerne et al. (2008). (A) Representation of the four different magma batches This paper has three main sections: The fi rst identifi ed in the GVSC. These are represented by different colors and are stratigraphically section presents the geology and geochemis- positioned in cross section A–A′ (Galerne et al., 2008). Profi le locations are indicated by try of the Golden Valley Sill followed by sam- brown and green cylinders; the colors indicate the geochemical signature of the Golden Val- pling strategy and methodology. The next sec- ley Sill and the Morning Sun sill, respectively. (B) Location of the GVSC in the Karoo Basin tion gives a systematic overview of signifi cant among the dolerite intrusives (represented in black). The location of the Lesotho Flood textural, petrographic, and chemical variations basalt remnant is indicated by a gray color.

164 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

GVSC’s main sill morphology is saucer-shaped. al., 2008). The 100-m-thick Golden Valley Sill istic of cooling features in the top contact of Four major saucer-shaped sills, all elliptical in is the best exposed sill in the GVSC. magmatic bodies can be observed in the upper shape, constitute the GVSC; the Golden Val- parts of the sill (Fig. 3B). In addition, sediment ley Sill, Glen Sill, Morning Sun, and Harmony 3. SAMPLING STRATEGY dykes (~40–50 cm thick) have been observed at Sill (Fig. 2). Additionally, a minor subcircular the surface of the exposed parts of the saucer- saucer-shaped sill, the MV Sill, is attached to A main objective in this study was to get an shaped sill. These observations suggest that the the northwestern boundary of the Golden Val- overview of the variation in type of profi le along top of the sill is located close to the original ley Sill. A major dyke (15 m thick and 17 km the border of the sill rather than the fi ner varia- roof. In many profi les the bottom contact with long, strikes NW-SE), the Golden Valley Dyke, tions in the sill structure. We therefore collected the host rock is exposed (example in Fig. 3C), has also been identifi ed in the western part of the samples in 18 profi les along the boundary of the but in several profi les the bottom contact was GVSC (Fig. 2). Golden Valley Sill. In each profi le rock samples covered by debris. However, extrapolating the The dolerites of the GVSC are basalts to were taken every 5–10 m, from the chilled bot- slope of the country-rock beneath the debris

basaltic andesites with TiO2 < 2 wt% and Zr < tom contact to the roof (Fig. 3). Some samples toward the cliff shows that the bottom contact is 180 ppm, which is characteristic of the low were taken more than 10 m from the next not more than 10 m below the lowermost expo- Ti-Zr (LTZ) basalts of the Karoo-Ferrar igne- because fresh samples could not be obtained sures. We thus believe that most of the original ous province (e.g., Cox et al., 1967; Erlank in between, or because some parts of the cliff thickness of the Golden Valley Sill is preserved et al., 1988; Marsh et al., 1997; Sweeney et in between were inaccessible to sampling. and that we have sampled the full thickness of al., 1994). A statistical analysis of the GVSC Because of these sampling problems some pro- the sill along profi les. geochemical data set (Galerne et al., 2008) fi les are represented by only three or four sam- The coordinates (latitude and longitude) showed that the saucer-shaped sills exhibit ples (e.g., P8, P9). The maximum number of acquired by GPS during sampling are reported three distinct geochemical signatures (shown samples along a single profi le is ten (P11, P15). with the whole-rock analyses in Tables 1 and 2 (and in different colors in Fig. 2). The identical geo- This means that the fi ner variations in the sill Supplemental Table 11). The distances between chemical signature, in addition to the physical structure will be missed and the classifi cation of the samples were measured directly using a deca- connection of the MV Sill and the Golden Val- some profi les may be dubious. However, in spite meter, and controlled by GPS. These measure- ley Sill, lead the authors to suggest that the MV of these limitations we found a systematic distri- ments were positioned with respect to the bottom Sill resulted from lateral overfl ow of the major bution of profi le types along the boundary of the contact, and corrected to give the real distance Golden Valley Sill. Also the Glen Sill has the Golden Valley Sill. We regard this result as very same geochemical signature as the Golden important and an indication that our sampling Valley Sill (beige in Fig. 2). The different geo- strategy served our main purpose. 1Supplemental Table 1 is a Microsoft Word docu- chemical signatures imply that at least three The roof of the sill is usually removed by ment. If you are viewing the PDF of this paper or read- ing it offl ine, please visit http://dx.doi.org/10.1130/ separate magma pulses were involved in the erosion and thus is not available for sampling. GES00500.S1 or the full-text article on www.gsapubs formation of the sills in the GVSC (Galerne et However, the horizontal entablature character- .org to view Supplemental Table 1.

A B HoHorizontalorizontta entablatureenntat blb ataturure IgIgneousgneouous sissilllll 100 m IgIgneousgneeouo s sillsisill HoHHostostst rrockocock 1 m

Outer-sill z C Igneous sill Profile y LoLower ccontact owewer coc x ontntatacact

Inclined ShShalea e Sample: -sheet 2 m Inner-sill

Figure 3. Sampling strategy of compositional proﬁ les on saucer-shaped sills from the GVSC. (A) Illustration of typical sampled cliffs in the Golden Valley Sill incline sheets cutting through undeformed host rock made of horizontal layered sandstone. The white dashed line indicates the extrapolated contact located under debris where bushes preferentially grow. (B) Typical horizontal entablature observed at the top of sill, characteristic of cooling features in the top contact of magmatic bodies. (C) Directly observed bottom contact of a horizontal inner sill with host metamorphic shale.

Geosphere, June 2010 165

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

TABLE 1. REPRESENTATIVE WHOLE-ROCK ANALYSES OF GOLDEN VALLEY SILL (GVS) AND MORNING SUN (MS) OF I-, S-, AND D-SHAPED COMPOSITIONAL PROFILES Sample K05-167 K05-168 K05-169 K05-170 K05-171 K05-32 K05-33 % Proﬁ les P 1 P 1 P 1 P 1 P 1 P 5 P 5 STD Latitude (N) -31.8237 -31.8239 -31.8240 -31.8241 -31.8242 -31.8984 -31.8984 Longitude (E) 26.2672 26.2674 26.2675 26.2674 26.2674 26.3144 26.3146 D.r.B. (m)7 142535488169 Height (nd) 14 29 52 73 100 100 85

SiO2 (wt%) 51.01 51.14 51.11 50.87 50.79 52.11 51.96

TiO2 (wt%) 0.99 1.05 1.06 0.97 1.05 0.72 0.92

Al2O3 (wt%) 15.47 14.93 14.91 15.49 15.22 15.86 15.19

Fe2O3t (wt%) 12.02 11.86 11.86 11.50 11.60 10.75 12.37 MnO (wt%) 0.18 0.18 0.18 0.18 0.18 0.18 0.20 MgO (wt%) 6.16 6.09 6.02 6.30 6.27 7.09 6.28 CaO (wt%) 10.51 10.32 10.35 10.37 10.56 10.99 10.43

Na2O (wt%) 2.58 2.71 2.55 2.61 2.52 2.27 2.32

K2O (wt%) 0.69 0.76 0.71 0.66 0.70 0.51 0.59

P2O5 (wt%) 0.15 0.16 0.16 0.16 0.17 0.11 0.11 SUM 99.76 99.20 98.91 99.11 99.06 100.59 100.36 Mg# 50.4 50.4 50.1 52.0 51.7 56.6 50.1 Cs (ppm) 0.2 0.5 0.2 0.2 0.2 0.3 0.3 0.085 Tl (ppm) b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0.064 Rb (ppm) 12 17 13 15 13 12 13 0.700 Ba (ppm) 203 216 211 198 207 161 193 1.273 Th (ppm) 1.6 1.5 1.5 1.3 1.3 1.3 1.4 1.527 U (ppm) 0.32 0.30 0.29 0.29 0.28 0.24 0.27 0.021 Nb (ppm) 8.1 8.2 8.1 8.7 8.1 7.0 7.5 1.973 Ta (ppm) 0.42 0.48 0.50 0.43 0.45 0.38 0.39 0.042 La (ppm) 10.9 11.2 11.8 11.4 11.9 9.4 10.2 0.361 Ce (ppm) 22.6 24.5 25.3 25.5 24.8 19.5 21.3 0.445 Pb (ppm) b.d. 4 4 6 4 5 7 2.503 Pr (ppm) 2.58 3.04 2.90 2.97 2.87 2.40 2.66 0.064 Mo (ppm) 0.4 0.4 0.2 1.5 0.2 1.0 1.3 1.421 Sr (ppm) 206 197 197 203 198 199 198 0.270 Nd (ppm) 13.4 14.7 14.8 15.6 15.2 11.1 12.6 0.212 Zr (ppm) 91 103 105 95 100 79 88 6.512 Hf (ppm) 2.35 2.51 2.47 2.50 2.52 2.18 2.24 0.127 Sm (ppm) 3.23 3.70 3.97 3.66 3.61 2.78 3.12 0.042 Eu (ppm) 1.13 1.05 1.13 1.16 1.11 0.90 1.00 0.042 Sn (ppm) 0.8 0.6 2.9 b.d. 0.5 0.5 0.5 0.700 Gd (ppm) 3.19 3.47 3.30 3.53 3.40 2.84 3.26 0.042 Dy (ppm) 3.88 4.00 4.16 4.15 4.25 3.17 3.72 0.021 Li (ppm) b.d. 3 b.d. 5 b.d. 6 8 0.267 Y (ppm) 26 27 27 28 27 23 26 0.148 Ho (ppm) b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0.636 Er (ppm) 2.17 2.34 2.27 2.49 2.16 1.91 2.29 0.021 Yb (ppm) 2.43 2.25 2.37 2.50 2.27 1.94 2.30 0.064 Lu (ppm) 0.35 0.34 0.36 0.39 0.35 0.29 0.32 0.021 Co (ppm) 32 34 34 32 34 35 38 0.814 Cr (ppm) 277 272 269 251 274 261 262 0.018 Cu (ppm) 103 106 109 99 105 70 98 2.546 Ni (ppm) 64 62 63 64 65 65 55 0.424 Sc (ppm) 37 37 37 37 38 39 41 1.026 V (ppm) 262 267 264 245 265 240 271 0.000 Zn (ppm) 91 94 94 89 92 79 96 1.697 Tb (ppm) 0.6 0.7 0.6 0.7 0.6 0.064 Tm (ppm) 0.3 0.4 0.4 0.3 0.3 0.636 (continued)

166 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

TABLE 1. REPRESENTATIVE WHOLE-ROCK ANALYSES OF GOLDEN VALLEY SILL (GVS) AND MORNING SUN (MS) OF I-, S-, AND D-SHAPED COMPOSITIONAL PROFILES (continued) Sample K05-34 K05-31 K05-35 K05-36 K05-37 K05-38 K05-39 % Proﬁ les P 5 P 5 P 5 P 5 P 5 P 5 P 5 STD Latitude (N) -31.8983 -31.8983 -31.8983 -31.8984 -31.8983 -31.8983 -31.8982 Longitude (E) 26.3147 26.3148 26.3148 26.3149 26.3153 26.3153 26.3153 D.r.B. (m) 60 46 38 25 15 9 5 Height (nd) 74 57 47 31 19 12 6

SiO2 (wt%) 52.29 52.33 51.30 51.47 51.38 51.71 51.99

TiO2 (wt%) 1.30 1.19 1.07 1.01 0.96 1.07 1.07

Al2O3 (wt%) 14.17 15.24 15.45 15.40 15.21 15.32 15.29

Fe2O3t (wt%) 13.63 13.01 11.81 11.81 11.96 11.79 12.03 MnO (wt%) 0.21 0.19 0.17 0.18 0.18 0.18 0.18 MgO (wt%) 5.35 5.71 5.91 6.19 6.24 6.08 6.06 CaO (wt%) 9.36 10.06 10.21 10.35 10.25 10.19 10.15

Na2O (wt%) 2.54 2.55 2.40 2.30 2.31 2.33 2.35

K2O (wt%) 0.86 0.82 0.72 0.65 0.65 0.70 0.70

P2O5 (wt%) 0.20 0.19 0.17 0.15 0.16 0.17 0.17 SUM 99.91 101.29 99.22 99.51 99.30 99.54 99.99 Mg# 43.7 46.5 49.8 50.9 50.8 50.5 49.9 Cs (ppm) 0.3 0.4 0.3 0.4 0.4 0.4 0.4 0.085 Tl (ppm) b.d. 0.2 b.d. b.d. 0.2 b.d. b.d. 0.064 Rb (ppm) 17 19 14 14 14 17 18 0.700 Ba (ppm) 264 255 220 203 206 214 214 1.273 Th (ppm) 2.2 1.9 1.8 1.8 1.6 1.7 1.6 1.527 U (ppm) 0.40 0.37 0.34 0.31 0.35 0.32 0.30 0.021 Nb (ppm) 10.8 10.3 8.7 8.7 8.2 10.2 9.8 1.973 Ta (ppm) 0.60 0.55 0.48 0.47 0.46 0.51 0.45 0.042 La (ppm) 14.7 14.4 12.3 11.7 11.7 13.0 12.7 0.361 Ce (ppm) 30.8 30.1 26.0 24.7 24.7 27.5 26.8 0.445 Pb (ppm) 7 7 8 8 7 7 7 2.503 Pr (ppm) 4.01 3.76 3.21 2.99 3.08 3.31 3.29 0.064 Mo (ppm) 1.3 1.4 1.3 1.2 1.3 1.1 1.3 1.421 Sr (ppm) 195 203 198 193 193 192 192 0.270 Nd (ppm) 18.3 17.4 15.1 14.6 14.3 15.3 15.6 0.212 Zr (ppm) 133 117 111 103 104 111 110 6.512 Hf (ppm) 3.64 3.19 2.91 2.82 2.71 2.79 2.92 0.127 Sm (ppm) 4.36 4.42 4.16 3.72 3.62 4.10 3.71 0.042 Eu (ppm) 1.24 1.20 1.09 1.01 1.09 1.10 1.07 0.042 Sn (ppm) 0.8 0.8 0.6 0.9 0.7 0.7 0.6 0.700 Gd (ppm) 4.42 4.27 3.82 3.52 3.65 3.75 3.82 0.042 Dy (ppm) 5.22 4.71 4.39 4.07 4.10 4.24 4.20 0.021 Li (ppm) 7 8 10 7 8 6 6 0.267 Y (ppm) 35 36 28 27 27 32 31 0.148 Ho (ppm) 1 1 b.d. b.d. b.d. b.d. b.d. 0.636 Er (ppm) 3.07 2.88 2.58 2.48 2.52 2.48 2.60 0.021 Yb (ppm) 3.27 3.06 2.76 2.57 2.58 2.57 2.66 0.064 Lu (ppm) 0.49 0.46 0.41 0.38 0.38 0.37 0.40 0.021 Co (ppm) 39 40 38 39 39 38 38 0.814 Cr (ppm) 110 176 241 271 285 246 262 0.018 Cu (ppm) 130 130 109 102 102 107 108 2.546 Ni (ppm) 40 52 61 64 67 62 62 0.424 Sc (ppm) 43 40 38 38 38 38 38 1.026 V (ppm) 303 282 260 261 252 268 269 0.000 Zn (ppm) 108 103 94 92 92 93 94 1.697 Tb (ppm) 0.064 Tm (ppm) 0.636 (continued)

Geosphere, June 2010 167

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

TABLE 1. REPRESENTATIVE WHOLE-ROCK ANALYSES OF GOLDEN VALLEY SILL (GVS) AND MORNING SUN (MS) OF I-, S-, AND D-SHAPED COMPOSITIONAL PROFILES (continued) Sample K04C-11 K04C-16 K04C-17 K04C-18 K04C-19 K04C-20 K04C-21 K04C-22 % Proﬁ les P 19 P 19 P 19 P 19 P 19 P 19 P 19 P 19 STD Latitude (N) -32.0011 -31.9983 -31.9986 -31.9989 -31.9989 -31.9989 -31.9988 -31.9988 Longitude (E) 26.2847 26.2797 26.2793 26.2790 26.2790 26.2791 26.2791 26.2792 D.r.B. (m) 75 73 47 4 6 20 32 42 Height (nd) 100 97 62 6 8 27 43 56

SiO2 (wt%) 50.76 50.97 50.21 50.29 50.87 50.47 50.47 50.68

TiO2 (wt%) 0.91 1.01 0.91 0.93 0.93 0.88 0.93 0.92

Al2O3 (wt%) 15.47 15.15 15.25 15.41 15.60 15.33 15.44 15.18

Fe2O3t (wt%) 10.74 11.35 10.69 10.82 10.83 11.02 10.92 11.15 MnO (wt%) 0.17 0.18 0.17 0.17 0.17 0.17 0.17 0.17 MgO (wt%) 7.02 5.93 6.72 6.99 7.07 7.51 7.15 7.23 CaO (wt%) 11.03 10.59 10.81 10.95 10.98 10.69 10.84 10.75

Na2O (wt%) 2.32 2.42 2.40 2.20 2.39 2.33 2.39 2.30

K2O (wt%) 0.37 0.45 0.43 0.41 0.41 0.40 0.46 0.44

P2O5 (wt%) 0.12 0.16 0.13 0.12 0.12 0.11 0.12 0.13 SUM 98.91 98.21 97.71 98.29 99.37 98.91 98.89 98.95 Mg# 56.4 50.9 55.5 56.1 56.4 57.4 56.5 56.2 Cs (ppm) 0.8 0.6 0.5 1.1 1.0 0.5 0.5 0.6 0.085 Tl (ppm) b.d. b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0.064 Rb (ppm) 7 8 9 8 8 8 9 7 0.700 Ba (ppm) 143 190 147 146 147 142 151 149 1.273 Th (ppm) b.d. 1.4 b.d. b.d. b.d. b.d. b.d. b.d. 1.527 U (ppm) 0.16 0.27 0.17 0.21 0.18 0.17 0.19 0.15 0.021 Nb (ppm) 5.1 7.7 5.0 5.4 5.8 7.0 5.7 5.1 1.973 Ta (ppm) 0.26 0.47 0.28 0.27 0.30 0.29 0.30 0.28 0.042 La (ppm) 7.8 10.6 7.3 7.4 8.0 7.9 7.3 6.7 0.361 Ce (ppm) 17.0 24.5 16.3 16.7 17.9 16.2 16.7 15.7 0.445 Pb (ppm) 5 5 2 2 6 7 4 7 2.503 Pr (ppm) 2.13 2.91 2.11 2.14 2.34 2.08 2.09 2.01 0.064 Mo (ppm) 0.7 0.9 1.1 0.9 1.0 0.9 0.8 0.7 1.421 Sr (ppm) 188 197 183 190 193 189 193 183 0.270 Nd (ppm) 10.8 16.0 11.2 10.4 12.5 11.5 10.9 10.4 0.212 Zr (ppm) 80 103 81 81 81 79 79 77 6.512 Hf (ppm) 2.05 2.59 1.83 1.90 2.06 2.11 1.94 1.76 0.127 Sm (ppm) 2.85 3.60 2.90 2.83 3.04 2.76 3.25 2.51 0.042 Eu (ppm) 0.97 1.15 0.91 0.94 1.10 0.98 1.00 0.96 0.042 Sn (ppm) 0.5 0.6 b.d. b.d. b.d. b.d. b.d. b.d. 0.700 Gd (ppm) 3.00 3.56 3.07 2.90 3.18 3.15 2.98 2.85 0.042 Dy (ppm) 3.79 4.00 3.79 3.81 3.98 3.75 3.81 3.76 0.021 Li (ppm) 8 2 b.d. 8 2 b.d. b.d. 13 0.267 Y (ppm) 25 27 24 25 25 23 24 24 0.148 Ho (ppm) b.d. b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0.636 Er (ppm) 1.97 2.26 2.08 1.99 2.15 1.94 2.04 1.90 0.021 Yb (ppm) 2.06 2.28 1.99 2.05 2.17 2.08 2.08 2.00 0.064 Lu (ppm) 0.33 0.35 0.31 0.34 0.32 0.28 0.31 0.30 0.021 Co (ppm) 34 33 35 34 35 36 35 36 0.814 Cr (ppm) 352 250 393 363 378 420 404 368 0.018 Cu (ppm) 100 108 98 101 100 97 98 97 2.546 Ni (ppm) 96 64 95 99 108 121 108 104 0.424 Sc (ppm) 35 37 33 35 34 33 34 33 1.026 V (ppm) 256 265 245 247 252 234 247 241 0.000 Zn (ppm) 86 94 87 86 87 88 86 87 1.697 Tb (ppm) 0.6 0.6 0.5 0.5 0.6 0.5 0.6 0.5 0.064 Tm (ppm) 0.3 0.4 0.3 0.3 0.3 0.3 0.3 0.3 0.636 Note: Abbreviations: n.d.—not determined; b.d.—below detection limit; D.r.B. (m)—distance with respect to bottom contact in meters; thickness (nd)—thickness normalized to 100. Global Position System gave accurate latitude and longitude for samples. Altitudes are reported in Galerne et al.(2008) and were used to control true measurement of distances between samples. However, GPS altitudes were not accurate enough for the precision suited for the present study and are thus not reported here. Standard detection limits (STD%) quoted as 3 times standard deviation measured on a blank solution.

168 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

TABLE 2. WHOLE-ROCK XRF OF SELECTED D-SHAPED PROFILE DATA (AARNES ET AL., 2008) Sample K04A-12 K04A-13 K04A-14 K04A-15 K04A-16 K04A-17 % Proﬁ les P 14 P 14 P 14 P 14 P 14 P 14 STD D.r.B. (m)65473930181 Height (nd) 86 62 51 39 24 1

SiO2 (wt%) 52.31 52.02 51.44 51.57 52.03 51.82 0.09

TiO2 (wt%) 0.97 0.92 0.93 0.95 0.96 1.01 0.80

Al2O3 (wt%) 15.28 16.20 15.57 15.36 16.09 15.87 0.44

Fe2O3t (wt%) 11.80 10.99 11.41 11.67 11.16 11.66 0.11 MnO (wt%) 0.19 0.17 0.18 0.18 0.18 0.18 0.00 MgO (wt%) 6.42 6.65 6.98 7.13 6.58 6.44 0.35 CaO (wt%) 10.37 10.86 10.89 10.81 10.76 10.64 0.15

Na2O (wt%) 2.18 2.23 2.14 2.12 2.26 2.18 0.40

K2O (wt%) 0.71 0.64 0.60 0.61 0.64 0.68 0.00

P2O5 (wt%) 0.16 0.15 0.13 0.13 0.15 0.15 6.29 L.O.I. 0.19 0.11 0.00 0.02 0.09 0.15 0.00 SUM 100.58 100.94 100.27 100.55 100.90 100.78 0.06 Mg# 51.9 54.5 54.8 54.8 53.9 52.2 Rb (ppm) 16 14 15 16 15 17 4.09 Nb (ppm) 7768676.42 Sr (ppm) 198 207 198 196 206 204 0.80 Zr (ppm) 101 90 89 87 88 96 1.68 Y (ppm) 26 24 21 22 24 25 4.25 Co (ppm) 48 45 48 49 46 45 1.20 Cr (ppm) 269 346 362 384 341 296 1.70 Cu (ppm) 98 83 92 95 91 102 0.86 Ni (ppm) 63 77 81 84 76 69 1.26 V (ppm) 280 273 274 286 286 277 1.80 Zn (ppm) 94 85 89 92 89 90 1.65

of the sample normal to the dip direction of the the Golden Valley Sill. The proﬁ les P17 and Th, Tl, U, Y), and the rare earth elements (REE) sill bottom contact. These values are reported in P18 were aimed to capture compositional vari- using ICP-MS. The instrument used was a Per- Tables 1 and 2 and Supplemental Tables 1, 22 and ations in the two overlying sills (Golden Valley kin Elmer Elan 5000 and the instrument was 33 under the acronym D.r.B. for distance with Sill and Morning Sun). In order to investigate calibrated with natural and synthetic standards. respect to the bottom contact. The corresponding more precisely the compositional variations We also dissolved 0.2 g of powdered sample

values in height normalized to 100 for each sam- along profi les from the underlying Morning in 6 ml of HF and HClO4 (2:1 mixture). This ple in a given profi le are also reported in Tables 1 Sun saucer-shaped sill, three additional profi les was then evaporated to dryness, cooled, and

and 2 and in Supplemental Table 1 (see footnote from the Morning Sun South segment are pre- dissolved in 20 ml of 10% HNO3. This solution 1), and are illustrated in Figures 4–7. sented (green cylinders, Fig. 2A). was analyzed by ICP-AES for Fe, Mg, Ca, Na, Galerne et al. (2008) showed that the south- K, Ti, P, Mn, Ba, Co, Cr, Cu, Li, Ni, Pb, Sc, Sr, ern part of the Golden Valley Sill has a compli- 4. ANALYTICAL TECHNIQUES V, and Zn. The analytical precision is 1% for cated relationship with the underlying Morning Si, Al, Fe, Mg, and Ca and 2% for Na, K, Ti, Sun sill. Based on fi eld observations and differ- Powders for geochemical analyses were P, and Mn. Detection limits for measured trace ences in geochemistry the authors showed that prepared from 30 g of fresh rock using a stone elements are reported in Table 1. The rest of the profi les 17 and 18 (Figs. 2, 7, and 8) from the mill. Eighty-eight samples were analyzed by samples (47) were analyzed for major and trace Golden Valley Sill west limb were composite. inductively coupled plasma–atomic emis- elements (Rb, Sr, Nb, Zr, Y, Co, Cr, Cu, Ni, V, The lower part (10 m above fl oor contact) of sion spectrometry (ICP-AES) and inductively Zn; Table 2) by standard X-ray fl uorescence these profi les corresponds to the continuation coupled plasma–mass spectrometry (ICP-MS) (XRF) techniques at the University of Bergen of the Morning Sun sill northwestward below at the University of London, Royal Holloway. (Norway). Standard deviations are reported We weighed 0.2 g of powdered sample into a in Table 2. Tables 1 and 2 present representa-

graphite crucible and added 1.0 g of LiBO2. The tive analyses of characteristic selected profi les 2Supplemental Table 2 is a Microsoft Excel spread- powders were carefully mixed and fused at 900 from the Golden Valley Sill and Morning Sun sheet. If you are viewing the PDF of this paper or read- ºC for 20 min. The resulting mixture was dis- sill. Supplemental Table 1 (see footnote 1) pro- ing it offl ine, please visit http://dx.doi.org/10.1130/ GES00500.S2 or the full-text article on www.gsapubs solved in 200 ml of cold 5% nitric acid. Ga was vides the vertical position and group of samples .org to view Supplemental Table 2. added to the fl ux to act as an internal standard. belonging to profi les for whole-rock analyses This solution was then analyzed for Si, Al, and published by Galerne et al. (2008). 3Supplemental Table 3 is a Microsoft Excel spread- Zr by ICP-AES using a Perkin Elmer Optima The minerals were analyzed for major ele- sheet. If you are viewing the PDF of this paper or read- 3300R. The instrument was calibrated with ments using a CAMECA SX100 electron ing it offl ine, please visit http://dx.doi.org/10.1130/ GES00500.S3 or the full-text article on www.gsapubs natural and synthetic standards. The solution microprobe (EMP) at the University of Oslo. .org to view Supplemental Table 3. was also used to analyze for (Cs, Nb, Rb, Ta, The instrument was fi tted with an integrated

Geosphere, June 2010 169

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

energy-dispersive spectrometer and fi ve wave- 5. TEXTURE AND PETROGRAPHY tion, see Sec. 3), two types of textural varia- length-dispersive crystal spectrometers. Accel- tions from bottom to top in the sill have been erating voltage was 15 kV and counting times 5.1. Texture consistently observed and identifi ed. Both types were 20–30 s. Minerals were analyzed using exhibit fi ne-grained texture at the sill margins a beam current of 15 nA and a focused beam Like the other sills in the GVSC, the Golden (within the uppermost and lowermost 5 m; Fig. with 5 μm diameter. The detection limit of the Valley Sill is a nonlayered intrusion, but does 4A). The fi rst textural type is homogeneously analysis is on average 0.05 wt% (Supplemental present some textural variations. Also fi ner varia- medium grained over the sill height (0.2–0.8 Tables 2 and 3 [see footnotes 2 and 3]). tions may exist (subjected to sampling resolu- mm; Fig. 4B, log A). This type of textural log

pl Upper C sill 100 margin cpx 90 cpx K04-C21cpx 80 cpx 0 1 mm 70

pl B 60 cpx Sill cpx 50 K04-C19 Texture center legend 40

0 1 mm 30 Height normalized (meter)

20 A pl ol 10 K04-C18 Bottom pl sill 0 margin A B 0 1 mm Charactristic ideal logs Figure 4. Example of typical textural variations in profi les from the Golden Valley Sill. (A) The sill margins are characterized by a porphyritic texture dominated by aphanitic subhedral plagioclase grains and scattered clusters of large zoned plagioclase laths (1 mm). (B) Average, medium-grained dolerite is found along the overall thickness of the sill. There is a tendency for clinopyroxene to form oikocrysts. (C) Medium- to coarse-grained, well-developed ophitic texture found in the central part of some profi les (log B). The logs A and B show simplifi ed textural variations with the upwards textural changes (A)-(B)-(A) and (A)-(B)-(C)-(B)-(A), respectively. Log A is the most common in the Golden Valley Sill. Log B is based on the profi les P5, P14, and P19 found in the central part of the Golden Valley east limb and in profi le P19 from Morning Sun south limb.

Figure 5. Whole-rock compositional profi les from North to South of the Golden Valley Sill east limb. Profi les are represented in normalized height to 100 on the left scale and in true thickness on the right scale. Mg# is used to characterize the profi le shapes and behave similarly to

compatible major elements (i.e., CaO [5 × wt%]). TiO2 ([wt%]) and P2O5 ([10 × wt%]) show similar trends and together behave opposite to compatible major elements in differentiated profi les. Strongly compatible (Ni [ppm], Cr [ppm/4.5]) and incompatible elements (Zr [ppm], Y[4 × ppm]) are represented in the third and fourth columns, respectively; they follow the general trends formed by the major compatible and incompatible elements. Characterization of profi le into type of shape is described in the text and is reported in Figures 5–7 in the col- umn of major strongly compatible elements (Mg#, CaO). X-shape designates complex profi le shapes.

170 Geosphere, June 2010

40 50 I-shaped 20 GVS P2 0 0 100 60

50 I-shaped (?) 40 20 GVS P3 0 0 100 40 X-shaped 30 50 20 10 GVS P4 0 0 100 80 60 S-shaped 50 40 20 GVS P5 Height (normalized) 0 0

100 100 (m) bottom the to Distance 80 60 50 40 X-shaped 20 GVS P6 0 0 100 150

100 50 X-shaped 50 GVS P7 0 0 100 30

20 50 I-shaped (?) 10 GVS P8 0 0 100 25 20 50 I-shaped (?) 15 10 GVS P9 5 0 0 40 45 50 55 60 0.5 1.0 1.5 2.0 2.5 50 100 150 50 100 150 wt% wt% ppm ppm

Geosphere, June 2010 171

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

100 Mg# Ni Zr 30 5*CaO Cr/4.5 4*Y 20 50 X-shaped 10 TiO GVS P10 2 10*P2O5 0 0 100 60

50 X-shaped 40 20 GVS P11 0 0 100 60

40 50 X-shaped 20 GVS P12 0 0 100 25 20 15 50 D-shaped 10 GVS P13 5 Height (normalized) 0 0 100 (m) bottom the to Distance 60

50 D-shaped 40 20 GVS P14 0 0 100 100

50 50 X-shaped GVS P15 0 0 100 20 15 50 X-shaped (?) 10 GVS P16 5 0 0 40 45 50 55 60 0.5 1.0 1.5 2.0 2.5 50 100 150 50 100 150 wt% wt% ppm ppm

Figure 6. Whole-rock compositional profi les from North to South of the Golden Valley Sill west limb. Diagrams described in Figure 5. With the exception of two D-shaped profi les in the central part (P13 and P14), the compositional profi les in the Golden Valley Sill west limb are all complex and do not fi t the established nomenclature. As mentioned in the text the profi le P15 is not a complete profi le and thus is not further considered in terms of profi le shaped to magmatic differentiation.

172 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

is common, particularly at the Golden Valley plagioclase laths as well as small olivine grains. pyroxene can also be found in plagioclase-rich Sill east limb at its south and northern ends. In Many clinopyroxene oikocrysts are zoned. The domains as small interstitial grains (0.8–1.0 mm) the second type of textural variations, the fi ne- Fe-enriched rims contain closely spaced exsolu- or in groups of several grains (1.0–2.0 mm). grained texture at the sill margins shifts through tion lamellae of pigeonite. Exsolution lamellae of In detail, the second type of texture (log B), medium-grained zones, to coarser-grained zones clinopyroxene may also be present in pigeonite. with large pyroxene oikocrysts, is often not near the center of the sill (Fig. 4C, log B). This The clinopyroxene oikocrysts are surrounded by well-developed (e.g., P14, Golden Valley Sill coarsening is partly made up by the progressive large plagioclase laths that increase in size from west limb). The clearest expression of this tex- development of clinopyroxene oikocrysts form- the sill margins toward the central parts where tural pattern has been observed in the profi le P5 ing a “clinopyroxene-rich domain.” The large they form “plagioclase-rich domains” domi- from the Golden Valley Sill east limb. The log pyroxene oikocrysts (0.5–2 mm up to 1 cm nated by large euhedral plagioclase and clusters B–type profi le is asymmetrical in the profi le P5 in diameter) enclose randomly oriented small (both can attain 6 mm) and Fe-Ti oxides. Clino- with the coarser-grained part halfway to the sill

Golden Valley Sill South West segment 50 100 Mg# TiO Ni Zr 2 40 GVS P17 5*CaO 10*P2O5 Cr/4.5 4*Y 30 50 I-shaped 20 10 0 0 100 GVS P18 100 Height (normalized)

50 I-shaped (?) Distance tothe bottom (m) 50

0 0 40 45 50 55 60 0.5 1.0 1.5 2.0 2.5 50 100 150 50 100 150 wt% wt% ppm ppm Morning Sun sill South segment

Mg# TiO2 Ni Zr 5*CaO 10*P2O5 Cr/4.5 4*Y 80 100 60

50 D-shaped 40 20 MS P19 0 0 100 80

X-shaped 60 50 (?) 40 20 MS P20

Height (normalized) 0 0 100 25 (m) bottom tothe Distance 20 15 50 X-shaped 10 MS P21 5 0 0 40 45 50 55 60 0.5 1.0 1.5 2.0 2.5 50 100 150 50 100 150 wt% wt% ppm ppm

Figure 7. Whole-rock compositional profi les from the southern region of the Golden Valley Sill west limb (P17 and P18) and Morning Sun south limb from west to east (P19–P21). Diagrams described in Figure 5. Two of the profi les, P17 and P18, from the Golden Valley Sill west limb indicate that these profi les are composite. The lower parts of the profi les are characterized by the Morning Sun chemical signature, whereas the upper part is characterized by the Golden Valley Sill chemical signature (chemical signature established by Galerne et al. [2008]).

Geosphere, June 2010 173

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

center (Fig. 4C). A similar asymmetrical texture found as crystal aggregates (clusters; Fig. 4A) of Golden Valley Sill (Fig. 5), the west limb (Fig. variation has been observed in proﬁ le P19 from larger subhedral grains (2.0–3.6 mm) that form 6), and the southern part of the Golden Valley the Morning Sun South limb. a plagioclase-phyric texture. These plagioclase Sill together with the Morning Sun south limb clusters are present all the way from the chilled (Fig. 7). These ﬁ gures report the variations of 5.2. Petrography margins toward the center of the sill and are usu- strongly compatible elements (i.e., Mg#, CaO, ally made of zoned crystals. Large plagioclase Ni, and Cr) and incompatible elements (i.e.,

The main phases in the Golden Valley Sill single laths found in the coarser regions are also TiO2, P2O5, Zr, and Y). TiO2 and P2O5 are incom- are plagioclase, clinopyroxene, Fe-Ti oxides, usually zoned. Olivine is usually subhedral and patible in these rocks as mineral phases in which and olivine. Pigeonite is present in small pro- is found as single grains and grain clusters, as these elements are compatible (i.e., Fe-Ti oxides portions (<2%). Accessory apatite, pyrite, bio- inclusion in clinopyroxene oikocrysts, and occa- and apatite) only appear late on liquidus. In Fig- tite (<0.5%), and very rare fl uorine and sphene sionally in plagioclase clusters. Olivine can be ures 5–7 the chemical components are plotted may be present. The relative proportions of partly altered and replaced by minerals such as in different scales in order to better compare the plagioclase:clinopyroxene:Fe-Ti oxides:olivine iddingsite, carbonate, and serpentine. trends they form. In order to simplify the com- are roughly 50:40:5:5. The matrix is dominantly parison between the different profi les the ver- made of euhedral plagioclase laths (0.5 mm), 6. CHARACTERIZATION OF tical height has been normalized to 100 (Figs. larger euhedral plagioclase grains may occur COMPOSITIONAL PROFILES 5–9). The distance of the samples with respect (6 mm). Also found in the matrix are euhedral to the bottom contact (D.r.B.) and their con- to subhedral olivine grains (0.4–2.4 mm) and Figures 5–7 show compositional profi les versions in height normalized to 100 (nd), are anhedral clinopyroxene. Plagioclase is also from north to south along the east limb of the reported in Tables 1 and 2 and in Supplemental

100 NW SES P17 GVS GVS MS Roof zone GVS A West limb East Chilled (Segment 2) limb GVGVSVS 50 contact P17 Golden Valley Sill MS

Height (normalized) (GVS) 0 40 45 50 55 60 65 D BasBBasalasala zonzzoneone 100 S N P18 RooRooff zzoneone N GVGVSVS GVS BasBasalal zonezone 50 West limb C (Segment 1)

P18 Morning Sun

Height (normalized) 0 MS sill (MS) 40 45 50 55 60 65 P19 Mg# P20 GVS WeWestst limlimbb RooRoofof zonezonene B (Segment(Segment 1)1) MSS MS P21 SouthSouth segssegmentmennt South segment 1 km NW BasBasalal zonzonee SEE

100 100 100 P19 P20 P21

50 50 50

Height (normalized) 0 0 0 40 45 50 55 60 65 40 45 50 55 60 65 40 45 50 55 60 65 Mg# Mg# Mg#

Figure 8. (A) Detailed geological map of the southern region of the GVSC showing the relationship between the Golden Valley Sill and its underlying neighbor MS saucer-shaped sill. (B) View in cross section of the MS south segment propagating underneath the Golden Valley Sill west limb. (C) Illustration of the inferred contact between the Morning Sun sill and Golden Valley Sill (no chilled contact, P18). (D) Illustration of the chilled contact between the Morning Sun and Golden Valley Sill saucer-shaped sills (P17, 10 m from the ﬂ oor contact).

174 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

Table 1. True distance from the bottom contacts TiO2 = 0.97–1.17 wt%, Zr = 91–109 ppm, and in a symmetrical manner forming a “D-shaped” are given for each profi le on the right side of the Y = 25–31 ppm. profi le on the basis of strongly compatible ele- diagrams (Figs. 5–7, 9). Profi les at the southern end of the Golden ments (Mg#, CaO, Ni, and Cr). It varies from Characterization of profi les into type of shape Valley Sill west limb, P17 and P18, are com- lower values at the sill margins (e.g., P14; Figs. was based on consistent variations of strongly posite. They show a chilled contact between 6 and 9: Mg# = 51.9–52.2, Ni = 63–69 ppm) to compatible elements (i.e., Mg#, CaO, Ni, and the lowermost part, which has been identifi ed higher values at the sill center (e.g., P14; Figs. 6 Cr). Profi les that corresponded to established as the Morning Sun, and the upper part which and 9: Mg# = 54.8, Ni = 84 ppm). Incompatible nomenclature, i.e., I-, D-, and S-shaped have is part of the Golden Valley Sill (Galerne et al., elements show a clear opposite behavior, chang- been marked in Figures 5–7. Those that did not 2008). The upper parts of these profi les, which ing from higher values at the sill margins (e.g.,

fi t in any established nomenclature are marked as represent the Golden Valley Sill, are I-shaped P14; Figs. 6 and 9: P2O5 = 0.16–0.15 wt%, Zr = X-shaped profi les (Figs. 5–7). For profi les with (Figs. 7 and 8). 96–101 ppm) to lower values at the sill center

few samples (e.g., P9) or large distances between (e.g., P14; Figs. 6 and 9: P2O5 = 0.13 wt%, Zr = some of the samples (P3, P8, P9, P16, P18, and 6.2. Complex Profi les 87 ppm). Profi le 13 shows more restricted but P20) the low resolution makes the classifi cation similar variations (Figs. 6 and 9). questionable. For these profi les the classifi cation In the Golden Valley Sill D-, S-, and X-shaped Another example of a D-shaped profi le is P19 problem is acknowledged by a question mark profi les comprise P4–P7 and P10–16. The wid- in the underlying Morning Sun sill (Figs. 7 and (Figs. 5–7). Figure 9 presents the most charac- est variations are found in the S-shaped profi le 8) which shows Mg# = 55.8–55.4 at the roof teristic I-, D-, and S-shaped profi les identifi ed in P5 from the central part of the Golden Valley and fl oor margins, respectively, and Mg# = 57.4 the Golden Valley Sill. These will be used as ref- Sill east limb: Mg# = 43.7–56.6, CaO = 9.4– at the sill center. However, this profi le is asym-

erence profi les for further interpretation. 11.0 wt%, TiO2 = 0.72–1.30 wt%, P2O5 = 0.11– metrical with the most mafi c composition well 0.20 wt%, Ni = 40–67 ppm, Cr = 110–285 ppm, below the center of the sill (Fig. 7). Additionally, 6.1. I-Shaped Profi les Zr = 79–133 ppm, and Y = 23–36 ppm (Figs. a reversal in the composition is observed in the 5 and 9). The S-shape of this profi le is also last meters from the roof contact (Mg# = 50.8– The profi les P1–P3, P8, and P9 (Fig. 5) are refl ected in other chemical parameters, with 56.4). Unlike in a typical D-shaped profi le of classifi ed as I-shaped. Each of these profi les positive correlations between Mg# and compat- the Golden Valley Sill, CaO shows a relatively shows homogeneous compositions. However, ible elements and negative correlations between constant concentration between 10.7 and 11.0 the compositions vary somewhat from profi le Mg# and incompatible elements. wt%. The strongly compatible trace elements to profi le: e.g., Mg# = 49.4–52.6, CaO = 10.3– The profi le P14 presents signifi cant variations (Ni, Cr), however, confi rm the D-shaped behav- 10.7 wt%, Ni = 62–69 ppm, Cr = 251–301 ppm, in concentration along the thickness of the sill ior of the profi le, and the incompatible major

Major element Trace element CompatibleIncompatible Compatible Incompatible 50 100 Mg# Ni 5*CaO Cr/4.5 40 I-shape 30 50 20

TiO2 Zr 10 GVS P1 10*P O 4*Y 0 2 5 0 100 60 D-shape 50 40 20 GVS P14 Height (normalized) 0 0 100 80 (m) bottom the to Distance 60 S-shape 50 40 20 GVS P5 0 0 40 45 50 55 0.5 1.0 1.5 2.0 2.5 50 100 100 150 wt% wt% ppm ppm

Figure 9. Compatible and incompatible element behavior in selected I-, D-, and S-shaped compositional proﬁ les from the Golden Valley Sill.

Geosphere, June 2010 175

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

elements show a clear negative correlation with tion associated with clinopyroxene-rich and The chemistry of plagioclase clusters and pla-

the strongly compatible elements (P2O5, Y). plagioclase-rich domains (Figs. 11B and 11C). gioclase laths in the two distinct textural domains The remaining complex profi les are called Pigeonite presents a more scattered composi- are represented in Figures 11D–11F. The rela- X-shaped profi les as they do not fi t to any tion and does not show any particular trend tively high anorthite values at the core of the

particular end-member type of shapes, i.e., I, (Supplemental Tables 2 and 3 [see footnotes clusters (An60-An65) are regarded as a represen- D, and Sshape. Unlike the end-member type 2 and 3]). Olivine in both domains indicates a tative composition of early fractionated plagio- of shapes, some of the X-shaped profi les do general decrease in Mg# inward in the sill (Fig. clase, before the magma is emplaced and forms not show clear negative correlations between 11B). Olivine is most magnesium-rich in clino- the Morning Sun saucer-shaped sill (CaO = strongly compatible and incompatible elements pyroxene-rich domains (Fig. 11B) with Mg# 16–17 wt%; Fig. 11D). However, the rims are (e.g., P4, P7, and P10; Figs. 5 and 6). Other varying from 80 at the fl oor to 51 in the center much less CaO-rich (CaO = 12–13 wt%; Fig. X-shaped profi les, however, show positive cor- of the sill. Olivine in a plagioclase-rich domain 11D) and have compositions similar to plagio- relations between strongly compatible elements shows Mg# values from 80 at the fl oor to 31 in clase laths in the plagioclase-rich domain (Fig. and Mg#, and negative correlations between the center of the sill (Fig. 11B). The clinopyrox- 11E). Finally, the plagioclase found as inclu- Mg# and incompatible elements (e.g., P6, P11, ene forming large oikocrysts is homogeneous sions in clinopyroxene-rich domains shows and P15; Figs. 5 and 6). with Mg# = 80 along the D-shaped region of a D-shaped profi le with respect to CaO, rang- The profi le P15 presents the widest range of the profi le (Fig. 11C). Clinopyroxene from the ing from CaO = 12.5 wt% at the fl oor and near variation for the Golden Valley Sill west limb plagioclase-rich domain is more iron-rich with the upper part of the D-shaped region to 17 with Mg# = 49.3–54.6, CaO = 10.2–11.1 wt%, Mg# values similar to those of the whole-rock wt% approximately in the central part of the

Ni = 57–82 ppm, Cr = 247–355 ppm, TiO2 = chemistry (56–58). D-shaped region (Fig. 11F). These are similar

0.92–1.10 wt%, P2O5 = 0.14–0.17 wt%, Zr = 82–101 ppm, and Y = 24–28 ppm. The profi le P15, sampled on the inclined part of the Golden Valley Sill west limb, is located eastward from the profi le P14 location. It presents a shape 50 that could be described as an inverse S shape where compatible and incompatible elements present perfect mirror images of one another. 40 Pigeonite However, the bottom contact of this profi le was not confi dently identifi ed (the only profi le 30 reported with this problem). Thus the profi le P15 cannot be further interpreted in terms of

signifi cant profi le shape to the magmatic dif- (wt %) FeO* 20 olivine ferentiation process. Whole -rock 7. MINERAL CHEMISTRY IN 10 A D-SHAPED PROFILE plagioclase clinopyroxene 0 The mineral analyses presented here are aimed at characterizing the chemical varia- Cpx-rich d. tions along the well-defi ned D-shaped profi le clinopyroxene ol cpx pg pl P19 in more detail than obtained from whole- 20 rock chemistry. The profi le P19 also shows the Core clearest textural variations defi ned by the log B Rim type (Fig. 4). We systematically probed plagio- Pl-rich dom. clase, clinopyroxene, pigeonite (when present), ol cpx pg pl and olivine in the texturally clinopyroxene-rich Core plagioclase and plagioclase-rich domains. The result indi- Rim cates that: (1) the clinopyroxene-rich domain 10 whole contains Mg-rich clinopyroxene and olivine CaO (wt %) -rock and CaO-rich plagioclase (Fig. 10), and (2) the plagioclase-rich domains show wider variations Pigeonite in olivine chemistry and extend to much lower olivine Mg# values (Fo80-Fo30); clinopyroxene is more iron-rich, and plagioclase shows a wider range 0 in CaO content (6–18 wt%; Fig. 10; Supple- 0 1020304050 mental Tables 2 and 3 [see footnotes 2 and 3]). MgO (wt %) Variations in mineral chemistry along profi le P19 are shown in Figure 11A. In Mg# versus Figure 10. Plots of FeO and CaO vs. wt% MgO for plagioclase, clinopyrox- normalized height only olivine and clinopy- ene, olivine, and pigeonite in the D-shape profi le P19 from the Morning Sun roxene show signifi cant compositional varia- south limb. Data are reported in Supplemental Table 2.

176 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills Rim Core Cpx-rich CaO Mg# Rim Core le of CaO (wt%) vs. Pl-rich le of Mg# vs. normal- Mineral chemistryMineral Whole-rock chemistryWhole-rock Plagioclase cluster Domains ole-rock compositional proﬁ ole-rock ain (E), and plagioclase laths from clinopyroxene- ain (E), and plagioclase laths from hole-rock compositional proﬁ hole-rock Mg# CaO (wt.%) Clinopyroxene Plagioclase laths 50 60 70 80 le from the MS saucer-shaped sill (P19). The mineral analyses (olivine, sill (P19). the MS saucer-shaped le from 10 11 12 13 14 15 16 17 18 BC Olivine Mg# CaO (wt.%) Plagioclase laths 40 50 60 70 80 90 10 11 12 13 14 15 16 17 DEF A MS P19 MS P19 Mg# CaO (wt.%) Plagioclase cluster D-shaped part 10 11 12 13 14 15 16 17 50 52 54 56 58 60 0 0

80 60 40 20 80 60 40 20

100 100 Height normalized Height Height normalized Height Figure 11. Mineral chemistry compared with whole-rock chemistry along a D-shaped compositional proﬁ with whole-rock Mineral chemistry compared 11. Figure domains. (A) Plot of the w both plagioclase-rich and clinopyroxene-rich from plagioclase clusters and laths) are clinopyroxene, rich domain (F). ized height as a reference frame to compare with the mineral chemistry of olivine (B) and clinopyroxene (C). (D) Plot of the wh with the mineral chemistry of olivine (B) and clinopyroxene frame to compare ized height as a reference plagioclase-rich dom (D), plagioclase laths from with the mineral chemistry of plagioclase cluster normalized height to compare

Geosphere, June 2010 177

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

values to those found in the cores of the plagio- limb becomes distinct from the Golden Valley (Fujii, 1974; Gray and Crain, 1969; Marsh, clase clusters (Fig. 11D). Sill west limb. The westernmost Morning Sun 1989) and/or new in situ grown minerals (Fren- profi le is D-shaped (P19; Figs. 7 and 8). kel et al., 1988, 1989) may result in complex 8. GEOGRAPHICAL DISTRIBUTION OF compositional profi le shapes. COMPOSITIONAL PROFILES 8.3. Complex or X-Shaped Profi les These processes refer to quite different stages of crystallization of igneous sills: from Figure 12 shows the distribution of the main Complicated types of shapes (X-shaped pro- a liquid state where convection will dominate, compositional profi les (Mg# versus normalized fi les) occur in four regions of the Golden Valley to near solidus conditions where mush-related height) around the Golden Valley Sill and along Sill (Fig. 12), that is between I-shaped and D- processes will be dominant. Before comparing the Morning Sun south segment. It shows that: or S-shaped profi les along both limbs, includ- the Golden Valley Sill profi les with theoretical (1) all types of compositional profi les, I-, D-, and ing the MV Sill. The widest range of variation profi les based on the existing models outlined S-shaped, except for C-shaped, are present in in the MV Sill occurs in the profi le P11, which above, we present a review of the main fi ndings. the single Golden Valley Sill; and (2) differently has the most complicated shape observed. shaped profi les are systematically distributed Yet, this profi le also shows a negative correla- 9.1. Review of Main Results along this elliptical sill; I-shaped profi les are tion between some of the strongly compatible more abundant at, and close to, the northern and (Mg#, CaO) and incompatible (Zr, Y; Fig. 6) Our documentation of compositional pro- southern tips of the Golden Valley Sill, whereas elements. The profi le P10 at the very northern fi les in a single saucer-shaped sill shows the D and Sshapes are restricted to the central parts end of the Golden Valley Sill west limb presents following. of the western and eastern fl anks (Fig. 12). an interesting pattern with a gradual decrease in (1) A variety of compositional profi les, I-, D- strongly compatible elements toward the upper and S-shaped (Fig. 9), have been found at dif- 8.1. I-Shaped Profi les part of the profi le (e.g., Mg#, Figs. 6 and 12). ferent locations around a single, ~100-m-thick Incompatible elements, on the other hand, do tholeiitic saucer-shaped sill (Figs. 5–7, 12). With the exception of the northern part of the not show any particular variations along the sill Additionally, there is a systematic distribution of Golden Valley Sill west limb all I-shaped profi les height (Fig. 6). the various types of profi les around the Golden are located at the northern and southern ends Valley Sill (Fig. 12). I-shaped compositional of the Golden Valley Sill where the curvature 9. DISCUSSION profi les are observed at the northern and south- is strongest (Fig. 12). The absence of I-shaped ern parts of the sill. D and S shapes and more profi les along the northern part of the west limb Only one explanation has so far been pro- complex compositional profi les are observed in may be due to the presence of the minor MV Sill, posed for the origin of sills with I-shaped com- the central regions of the conjugate limbs. An which is a direct continuation of the Golden Val- positional profi les. It consists of the injection of exception to this general pattern is profi les in ley Sill (Figs. 5 and 12). Although some varia- a phenocryst-poor magma; crystallization and the area of the minor saucer-shaped sill, the MV tions along single profi les do occur (e.g., P1, P2, crystal growth predominantly occurs within the Sill, in the northwestern part of the Golden Val- and P17), most components are constant along solidifi cation front. This will prevent any evo- ley Sill, where complicated compositional pro- the sill height. The number of samples in pro- lution of the magma because the interstitial liq- fi les break the symmetry in the Golden Valley fi les P8 and P9 are low (four and three samples, uid will be captured, inhibiting differentiation Sill (Figs. 6 and 12). respectively) for this part of the sill that is signifi - (Mangan and Marsh, 1992; Marsh, 1996). In (2) There is a consistent negative correlation cantly thinner (P9 = 28 m). Thus the low number contrast, there is a wide range of processes that between strongly compatible and incompatible of samples makes the classifi cation of these pro- can explain the formation of D- and S-shaped elements in profi les that show signifi cant varia- fi les as I-shaped uncertain. profi les. Examples range from multiple or pro- tions (i.e., D- and S-shaped profi les and some longed continuous magma infl uxes (S-shaped: of the X profi les, Figs. 5–9). These observations 8.2. D- and S-Shaped Profi les Gorring and Naslund, 1995), a convective fl ux are consistent with a fractional crystallization of refractory components within the crystal- process. Profi les collected along the central parts of liquid mush of the boundary layer during in (3) No textural variations were observed the western and eastern limbs in the Golden situ differentiation or compositional convection along I-shaped profi les from the North and Valley Sill show considerable compositional (S-shaped: Tait and Jaupart, 1996), and gravita- South region of the conjugated limb’s tips of the variations giving rise to D-, S-, and X-shaped tional settling (S-shaped: Frenkel et al., 1989). Golden Valley Sill (Fig. 4, log A). profi les (Fig. 12). D-shaped compositional pro- Soret fractionation (Latypov, 2003a, 2003b; (4) Textural analysis of the D- and S-shaped fi les are found at the center of the Golden Valley Latypov et al., 2007) was recently proposed profi les (P14, P5) in the central region of the Sill west limb, whereas the S-shaped and other to explain the origin of the often observed Golden Valley Sill limbs, and one profi le from complex profi les are found at the center of the marginal reversals in layered intrusions. This the Morning Sun south limb (P19), show clear Golden Valley Sill east limb. On the east limb model was combined with the in situ crystal- textural variations (Fig. 4, log B) that correlate the S-shaped profi le P5 is located near the high- lization in thermal boundary layers of Tait and with the chemical variations. In these profi les est parts of the inclined sheet, whereas, on the Jaupart (1996; Latypov et al., 2007). Another the central part of the D-shaped region exhib- west limb a particularly well-defi ned D-shaped process recently proposed to explain the forma- its two distinct textural domains. One domain is profi le (P14) is located on the fl at outer sill. tion of D-shaped profi les and marginal rever- characterized by circular ophitic clinopyroxene. Finally, the Morning Sun south limb has been sal is the post-emplacement melt fl ow induced The second domain surrounds the fi rst one and followed from a position directly below the by thermal stresses (Aarnes et al., 2008), mostly contains plagioclase and Fe-Ti oxides. Golden Valley Sill west limb’s southern region which is further detailed in Sec. 9.2.1. Finally, (5) The profi le with the clearest textural varia- (Figs. 7 and 8), down to progressively lower gravity-induced settling of crystals present tion is the D-shaped profi le P19 in the Morn- altitudes. In this part, the Morning Sun south in the magma at the time of the emplacement ing Sun sill. EMP analyses of the two different

178 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

P1 P2 P3 ? P10

P11 X-shaped P12

I-shaped P5 X-shaped

o GGlenlen -31 80 N MV Sillillll X-shaped P13 Silll (MVS) (GS)(GGS)GSS) L1 P6 S-shaped HarmonyHHarm Golden Sill D-shaped o Valley -31 90 N P14 GoldenG ValleyVall Dyke (GVD) (HS) Sill Vall X-shaped alleyaalleal Dyke ( ll (GVS) N P7

ke ( MorningMor GVD)GVD Sun (MS(MS)MS)

P16 o X-shaped -322 00 N ? TarkastadTar stasta I-shaped P8

I-shaped Km ?

o o 266o 101 E 26 20 E 266 330 E 266o 404 E 0 4

P17 P18 ? P9 ?

D-shaped GVS West limb GVS East limb P19 100 Compositional range of

GVS 50

MS MS South limb Mg# Height (normalized) 0 40 50 60

Figure 12. Summary of the systematic variations of compositional profi le shapes defi ned on the basis of Mg# along the Golden Val- ley Sill and Morning Sun south segment. The fi gure shows a symmetrical distribution of profi le shapes from north to south on both Golden Valley Sill east and west limbs. The Golden Valley Sill presents I-shaped profi les at the northern and southern parts of both limbs. The central parts of the east and west limbs present S- and D-shaped profi les, respectively. The irregularity of the I-shapes on the northern end part of the Golden Valley Sill west limb corresponds to the MV Sill minor saucer-shaped sill that is the smallest of the Golden Valley Sill Complex saucers.

Geosphere, June 2010 179

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

textural domains showed that clinopyrox- with a thickness of 100 m or more (Barriere, convection has stopped. In this regime melt/ ene-rich domains contain Mg-rich minerals, 1976). Thus fl ow segregation involving a Bag- solid separation can only occur through fl ow of whereas plagioclase-rich domains contain more nold effect is unlikely to be a major mechanism the remaining melt fraction through the porous Fe-rich minerals (Figs. 11B and 11C). Similar, in the presently studied sill. crystal framework (Aarnes et al., 2008). Below, but less evident, results have been found in the There is no evidence of multiple injections various magmatic processes of melt-crystal seg- high-Mg# part of the S-shaped profi le P5 (Sup- in the Golden Valley Sill. The samples are all regation leading to fractional crystallization are plemental Table 3 [see footnote 3]). homogeneous, showing no abrupt changes in examined at the early and late stage of crystal- the texture. Furthermore, a structural and geo- lization. The resulting theoretical profi les are 9.2. Compositional Profi les and Processes in chemical investigation of the whole Golden Val- compared to the compositional profi les obtained the Golden Valley Sill ley Sill Complex indicated that each individual in the Golden Valley Sill. saucer-shaped sill resulted from a single impulse All together, these results suggest that the of magma, corresponding to an individual geo- 9.2.1. Compositional Profi les and Processes at magmatic differentiation processes may differ chemical signature amongst the six distinct Early Stage of Crystallization in different parts of a single cooling sill to the magma batches geochemically discriminated Gravitational ordering or segregation is his- extent that contrasting compositional profi les (Galerne et al., 2008). torically the fi rst mechanism postulated to result are produced (I-, D-, and S-shaped, and more Fractional crystallization embraces a wide in fractional crystallization (Bowen, 1915, complex). A recent in-depth analysis of avail- range of processes that occur in a cooling mass 1928). It is based on the idea that mineral phases able mechanisms to explain the formation of the from the earliest to the latest stages of crys- in magma chambers will settle or be buoyant various types of compositional profi les conven- tallization. It corresponds to any process that with respect to the ambient melt. This will result tionally classifi ed as I-, D-, and S-shaped in sills prevents a solid and a melt originally at equi- in compositional evolution. Profi les expected to has been made by Latypov (2003a, 2003b). In librium, to continuously reequilibrate during form as the result of this process are sketched in these papers the author pointed out the presence physicochemical changes (e.g., cooling). This Figure 13. The expected profi les will be more of marginal compositional reversals in terms of leads to chemical changes. In detail, fraction- enriched in Mg# toward the base of the pro- modal, phase, and cryptic layering representation (i.e., segregation) processes may diffi le as heavy minerals such as olivine will tend ing a mirror image of the large Layered Series fer (e.g., see review by Latypov [2003b] and to sink. According to this model, plagioclase (C-shaped in our nomenclature). A number of references therein). During the early stages of would be slightly more buoyant and collects distinctive features of marginal compositional magma crystallization, newly grown crystals in the upper part of the magma body where the reversals have been described by Latypov (or crystals brought through the feeding con- CaO concentration will consequently be high- (2003a). All together, it appears that the avail- duct) may be segregated by processes such as est. This will most likely result in asymmetrical able processes reviewed in Latypov (2003a, convective fractionation (e.g., Sparks et al., but linear profi les (Fig. 13). 2003b, and references therein) briefl y men- 1984), or crystal settling (Wager and Brown, The profi le P10 at the north end of the Golden tioned above are feasible, under some circum- 1968; Wager et al., 1960). Thus early formed Valley west limb presents an asymmetrical linear stances. However, they are unlikely to represent minerals may collect in the calm part of a con- profi le for strongly compatible elements, partic- the dominant explanation for a single sill yield- vecting magma body or at the cooling margins. ularly Mg#, Ni, and Cr (Fig. 6). Mg# in the pro- ing differently shaped compositional profi les During late stages of crystallization the magma fi le P10 is higher at the base and lower near the (I-, D-, and S-shaped), such as the Golden Val- body consists of a continuous crystal mush and top of the sill. This suggests that early formed ley Sill. Furthermore, they are unable to explain the formation of basal and top reversals that form D- and S-shaped profi les. Calm cooling mass Petrographic observations from the Golden Valley Sill show that two types of textural logs can be distinguished (Fig. 4). Although log B Result in Mg# indicates a coarsening inward, the sill marked g CaO, Sr by the development of large oikocryst clinopy- asymmetrical Height roxene, the Golden Valley Sill is characterized profiles by relatively homogeneous medium-grained texture over its height (log A). Furthermore, we show that numerous profi les indicate I-shaped Differentiation index for compositional profi les associated with homoge- Crystal bouyancy (pl) -Heavy minerals (e.g. ol, cpx) Mg# neous textural profi les (Fig. 4, log A; Fig. 12). Crystal settling (ol, cpx) -Light minerals (e.g. pl) CaO, Sr The only available mechanism proposed by Mangan and Marsh (1992) and Marsh (1996) to Figure 13. Effect of gravitational ordering on whole-rock compositional profi les. Early explain the formation of I-shaped compositional nucleated minerals (i.e., ol, cpx, pl on liquidus) will order according to their relative profi les involves the injection of phenocryst- density. Heavy minerals will tend to settle down whereas light minerals will be buoyant. poor magma, thus implying that the Golden Val- This segregation prevents any further reequilibration with the melt from which minerals ley Sill was formed through the injection of such grew inducing whole-rock compositional changes along profi les. Heavy, early formed phenocryst-poor magma. Flow inducing crystal minerals such as olivine and clinopyroxene will settle down thereby enriching the lower segregation and concentration of larger pheno- part of profi les in Mg#. On the contrary, light minerals such as plagioclase will be buoy- crysts in the center of a fl ow channel has been ant thereby enriching the upper part of profi les in CaO, Sr. Complications may come shown to be particularly inoperative for sills from using CaO if clinopyroxene is on the liquidus, thus Sr may be used instead.

180 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

olivine may have settled down. Although CaO C-shaped profi le has been found in the Golden ally, Ni and Cr are also enriched in the center shows a similar tendency suggesting settling Valley Sill. However, the profi le P21 (Figs. 7 and even though their molecular weights are in the of clinopyroxene, a slight belly shape toward 8) in the Morning Sun sill may be regarded as an same range as Fe (Fig. 9). the bottom contact is observed. Sr, in contrast, asymmetrical C-shaped type. Unfortunately, its shows a negative correlation to CaO (table 2 in sampling resolution was too low to be ascertained 9.2.2. Compositional Profi les and Processes at Galerne et al., 2008). Sr is strongly compatible as a C-shaped type and was left unclassifi ed. This Late Stage of Crystallization with plagioclase and its relative depletion in the suggests that even if C-shaped profi les may exist Largely motivated by the negative correlation lowermost part of the profi le suggests a possible in ~100-m-thick sills, they are rare. This implies between compatible and incompatible elements effect of plagioclase buoyancy. The combined that convective fractionation is unlikely to be observed in D-shaped, S-shaped, and some effect of clinopyroxene settling and plagio- among the dominant processes resulting in differ- complex profi les, we suggest the examination clase buoyancy potentially results in the shape entiation in a sill such as the Golden Valley Sill. of a post-emplacement melt fl ow and separa- of the CaO profi le. This type of compositional Latypov (2003a, b) revived the hypothesis tion from the crystalline mush. This requires profi le is unique in the Golden Valley Sill. Thus of Soret fractionation to explain the marginal that crystallization has proceeded so far that a gravitational separation cannot be the dominant reversals producing D- and S-shaped compo- continuous porous crystal network has formed, mechanism for magmatic differentiation in the sitional profi les in basic-ultrabasic sills. Soret from which the remaining melt may fl ow from Golden Valley Sill, but may have occurred in the fractionation (thermal diffusion) is a process one region to another. northern end of the west limb. that causes heavy components (e.g., Fe) to In a schematic representation of a binary sys- An important alternative to settling and gravita- migrate toward the colder end of a thermal gra- tem with a complete solid solution (Fig. 14A), tional ordering is convective fractionation (Rice, dient, and the lighter components (e.g., Mg) the removal of a melt fraction at a constant 1981; Sparks et al., 1984). Both are complex pro- to migrate toward the hotter end. Although temperature implies a shift from the bulk com- cesses that explain the C-shaped profi les (in our Soret fractionation is a potential process that position toward higher values of Mg# for the nomenclature) observed in large magma cham- can explain D-shaped profi les, the geochemi- remaining drained region (Fig. 14B). The fl ow bers (e.g., Wager and Brown, 1968). It is based on cal data of the Golden Valley Sill show strong of the iron-rich melt fraction to other parts of the the consideration that the inward cooling from the evidence against it. If heavy elements migrate crystalline mush will change the bulk rock com- roof and bottom of a magma chamber will pref- toward the cooling margins, we would expect position toward lower Mg# values for the region erentially remove the phases that have the highest the same trend for all elements of differing where the fractionated melt is added (Fig. 14B). melting points from the melt through solidifi ca- molecular weight. In contrast, Ca increases Figure 14C sketched the case of a marginal melt tion. This leaves a melt relatively enriched in the toward the center of the sill and Na and Cr are fl ow from Aarnes et al. (2008). components that go into phases with low melting enriched toward the margin, although Ca is 9.2.2.a. Post-emplacement melt fl ow induced points at the solidifi cation front (Rice, 1981). No heavier than Na (D-shape in Fig. 9). Addition- by thermal stresses. Aarnes et al. (2008) showed

A B Drained region Supplied region C (source) (sink) Melt T fractionation Liquid (L) L + S C B M

C M Sill height Solid (S) Mg#

Sill L S L S L C C Sys Sys S Mg# Center (C) eq eq C C C Sys CeqCeq eq Ceq eq eq Ceq L S Sill C C Sys Sys Margin (M) eq eq New Ceq New Ceq Mass balance Mass balance at the sill center (C) at the sill margin (M) Figure 14. Schematic representation of the mass balance in a closed system of post-emplacement melt fl ow in a sill. (A) Initial system concentration at the sill center and margins represented in a T-X diagram (based on the example of olivine). Equilibrium concentrations of liquidus and solidus are shown by the temperature isotherms at the sill center (C) and the sill margin (M). (B) Representation of the marginal fl ow and subsequent mass balance in the T-X diagram. Departure of melt from the sill center causes an increase in Mg#; the addition of iron-rich melt at the sill margins yields a decrease in Mg#. (C) Compositional profi le from an originally homogeneous concentration over the sill height (pale blue straight line), resulting in a D-shaped profi le after melt fl ow from the sill center to the sill margins.

Geosphere, June 2010 181

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

through numerical modeling that thermally acti- tion (e.g., Φ = 0.5, Fig. 15B) can be written as: change in the temperature along profi les during vated stresses associated with the cooling and the fl ow event. sys L Φ S Φ crystallization of sills were suffi cient driving Ceq = Ceq · + Ceq · (1 – ). (1) All the profi les sampled along the Golden Val- forces to produce melt fractionation from the ley Sill are from either the transgressive sheets crystal-solid network originally of uniform com- To be considered as a continuous mush (the or the fl at outer sill. Hence we do not have any position throughout. The authors concluded that condition required for a porous fl ow to occur profi les in the fl at inner sill. The above process the thermal stresses induce large under-pressure in the entire sill) the higher porosity will be of post-emplacement melt fl ow implies the pres- near the cooling margins, where melt will be close to the random packing of crystals to form ence of a sink region in the saucer-shaped sill sucked in by porous fl ow. As a result, the margins a mush (crystallinity = 60%), i.e., 0.4 poros- where we expect to observe C-shaped profi les. will be enriched in incompatible elements and ity. This mush domain is defi ned as a comprise Fe, while the center will be relatively depleted between 90% < N < 55%, where N stands for in incompatible elements and enriched in Mg, crystallinity (Marsh, 2002). The mush zone is thereby producing a D-shaped profi le (Fig. 14C). bounded by the capture front that is defi ned as Figure 15. Schematic representation of the 9.2.2.b. A lateral source-sink model for post- the boundary between a crystal-free zone and a mass balance in the case of an open sys- emplacement melt fl ow in saucer-shaped sills. mush zone (i.e., 55% < N < 25%). tem post-emplacement melt fl ow in a sill. We present here a further step in developing the In the presently studied system we con- (A) Representation of the porosity (Φ), i.e., melt-separation model introduced by Aarnes et sider the stage where both solidifi cation fronts opposite of the crystallinity (N) at the stage al. (2008). Our aim is to develop a model that from the top and bottom of the sill meet in its when a continuous crystalline mush is cre- can explain the systematic geographical dis- mid-plane. Thus the maximum porosity at the ated. The starting moment of this stage tribution of compositional profi les around the sill center will lie within the range defi ned by occurs when both capture fronts from the Golden Valley Sill. The new model also takes the capture front (0.45 < Φ < 0.75), but close upper and bottom solidifi cation fronts meet into account the potential effect of the three- to 0.45 porosity that is the upper bound for the in the mid-plane of the sill. (B) Schematic dimensional (3-D) geometry of a saucer-shaped mush zone. Under the assumption that the start- representation of binary T-X phase dia- sill as compared to a horizontal sill, which is the ing composition over the sill height is homoge- gram with full solid solution between Fe basis for the model of Aarnes et al. (2008). neous, a phase rule similar to the above equation and Mg (Mg#) along the X axis. The further The 3-D geometry of a saucer-shaped sill has (1) can be applied to an entire profi le (Fig. 15C). assumed starting system composition is bal-

the potential to allow both vertical and lateral Speciﬁ cally for each isotherm (from T1 at the sill anced out between the liquidus and solidus

ﬂ ow, where distinct segments related to the center to T5 at the sill margin), the correspond- equilibrium concentrations for the isotherm

saucer-shaped geometry will act as source and ing phase proportion is given by the starting T0 and the solid melt proportion of 0.5. sink. We propose that a porous melt fl ow may porosity profi le (Φ ~0.45 at the sill center, and (C) Representation of the temperature and occur in a post-emplacement regime. Cooled 0.1 > Φ at the sill margins). porosity gradient having a Gaussian type from the outside, the inward progress of the top Figure 14 shows that melt which is extracted distribution with the highest value at the sill and bottom solidifi cation fronts in a horizontal from the center and supplied to the margins center and lowest value at the sill margins. sill will, with time, create a mid-plane poros- of the sill will generate increasing Mg# in the The corresponding representations for both ity region (Fig. 15A). The porosity profi le will melt-drained region and decreasing Mg# at the in a T-X diagram are highlighted from the have a similar Gaussian distribution across the melt-supplied region. The source-sink model above color fi elds; see inset B. The phase

sill height as temperature, because the crystal- implies that melt extraction will take place rule for each isotherm (T1 to T5) is written melt proportion (i.e., porosity) is controlled by from fl oor to roof in the melt-drained region below the phase diagram with, as starting cooling and crystallization (i.e., higher poros- of the sill, but will, because of the difference phase proportion, the porosity profi le (Φ, ity and temperature in the mid-plane of the sill in porosity (Fig. 15A), be most effective in left inset C) representing the original equi- and lower porosity and temperature near the the hot central region. The melt fraction avail- librium conditions for the given isotherms sill margins, Fig. 15C). Thus any forces acting able for extraction will decrease as permeabil- considered. Equilibrium (eq) is used here on the porous mush are likely to initiate melt ity and porosity decrease, toward the margins as opposed to the nonequilibrium process of fl ow and segregation from its equilibrium solid of the sill. This is represented in Figure 15D, nonequilibrium increment of melt extracted crystalline mush (Figs. 15B and 15C). This will where, for each mass balance at constant tem- (Φout) or supplied (Φin). (D) Similar mass

induce a bulk differentiation through melt seg- perature T1–5 and corresponding initial poros- balance as in Figure 14B for each isotherm Φ Φout regation from a “source region” which supplies ity , the melt fraction extracted ( ) is to (T1 to T5). The difference lay in the fact that melt to a “sink region” in the sill (Fig. 15D). be subtracted from the original melt fraction the amount of melt extract (Φout) along Such a system can be represented by a binary available in the source region, or to be added each isotherm will differ from one another. solid solution in a T-X diagram (Fig. 15B). This (melt fraction supplied: Φin) to the original Because the amount of melt (i.e., porosity) diagram displays the stability fi eld of a given melt fraction in the sink region; both melt frac- decreases toward the margins, the available phase (e.g., olivine) at solid (S) and liquid (L, tions can be assumed equal to each other (Φout amount of melt for advection correspond- i.e., melt) states and their onset representing ~Φin) but smaller than the original porosity ingly decreases. As a consequence of large the porosity (L+S). Readily, the liquidus rep- Φ. Applying the above equation (1) with the lateral advection, the Mg# increases at the resents the upper bound of the solid-liquid newly obtained melt fraction (ΦNew), the result- sill center will be greater as compared to mixture (Φ = 1) and the solidus represents the ing profi les will be D-shaped in the drained its margins, producing a D-shaped profi le lower bound or transition between the mix- region and C-shaped in the sink region (Fig. exceeding the original starting composition. ture and solid state (Φ = 0). Thus the mass 15D). The process described above assumes a The likely resulting profi le in the sink region balance at equilibrium for a given isotherm dominance of mechanical processes over diffu- will be a C-shaped profi le that is below the

(e.g., T0, Fig. 15B) and known phase propor- sive processes. By this, we imply no signiﬁ cant original starting composition.

182 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

A Top Crystallinity (N) Porosity (φ) B N > 90% φ < 0.1 T φ L 55% > N >90% Mush zone 0.45< <0.1 L + S T0 N < 55% capture front φ > 0.45

Sill height S 55% > N >90% Mush zone 0.45< φ <0.1 Mg# N > 90% φ < 0.1

Bottom L Sys S For C C φ = 0.5 Ceq eq eq T0 C Porosity gradient Temperature gradient Top Top φ < 0.1 L L + S T5 φ = 1 T1 T4 0.1< φ < 0.45 T2 φ = 0.5 T1 Sill φ < 0.45 T3 Sill center T0 φ = 0 S T4 T1 center 0.1< φ < 0.45 T5 T4 φ < 0.1 54 3 2 1 5 4 3 2 1 T5 Bottom L S C Sys C Bottom eq Ceq eq

D Drained region Supplied region (source) (sink)

L Sys S L Sys S C C C From C C C From eq eq eq 1 to 5 Melt eq eq eq 1 to 5 , φ T fraction T,φ Sys extracted Sys New Ceq Out In New Ceq New ( φ = φ ) < φ New In T, φ = φ φ Out T, φ = φ + φ

Mg# Mg#

Geosphere, June 2010 183

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

As none of the sampled region of the Golden saucer-shaped sill (outer sill and fl at inner sill, ated crystals may grow inside the saucer-shaped Valley Sill yield C-shaped profi les the only pos- respectively) may result in melt expulsion from sill. The hydrofracturing process associated with sible sink region in the saucer-shaped sill is the the high part and fl ow downward to low parts of the formation of the saucer-shaped sill (Galland nonsampled inner sill. Thus for the above model the saucer-shaped sill. et al., 2009) potentially combined with early to be correct we need to explain a melt fl ow related magmatic processes (e.g., unstable con- counteracting gravity, as the melt density should 9.3. Model Description for Lateral vection cells, crystal settling; Fig. 16A) eventu- be smaller than the solid density thereby being Source-Sink Model: The Case of a ally explains the even distribution of plagioclase relatively buoyant. Saucer-Shaped Sill clusters over the whole profi le sections all around A possible reason for a melt fl ow counteract- the saucer-shaped sill (Fig. 16A, 1a). Eventu- ing gravity is a pressure gradient triggering fl uid During the emplacement stage of a saucer- ally, in the thicker part of the saucer-shaped sill, fl ow. Specifi cally, a large enough downhill pres- shaped sill, the melt is actively supplied by settling of crystals behaving as effective dense sure gradient with a positive pressure in the high the feeding channel. Minerals can be brought plumes may occur (Brandeis and Jaupart, 1986; part and a negative pressure in the low part of a through the feeding channel and newly nucle- sketched inset in Fig. 16A, 1a).

A 1.a. Mixing phenomena

Stage 1: Emplacement Convection-like cells (mixing) Fully crystallized rock Mushy layer

1.b. Gravitational ordering

g Calm part Early stage of crystallization of cooling mass

B Stage 2: Post-emplacement Solidification front Mush zone Capture front Suspension zone Capture front Mush zone

Plagioclase cluster Late stage of crystallization Fully formed mush Preferential zone in mid-plane growth of newly formed Capture front plagioclase

Figure 16. Schematic representation of the processes occurring during the emplacement stage (A), and post- emplacement stage (B) of a saucer-shaped sill. This ﬁ gure is based on a centrally fed model for a saucer-shaped sill (e.g., Galland et al., 2009; Malthe-Sørenssen et al., 2004). (1a and 1b) These insets illustrate two important magmatic processes of early crystallizing magma; 1a illustrates some of the possible convective phenomena such as double-diffusion resulting in convective fractionation ultimately resulting in homogenization of the system through rollover. (1b) Illustration of possible gravitational ordering of crystals in calm parts of the cooling mass. The upper part of the mixing-like cell is highlighted so as to illustrate a possible combination of both stirring-like and gravitational ordering phenomena in the manner of a downward cold plume charged in an early high-temperature fractionated mineral. (B) Schematic representation of the solidiﬁ cation front inspired by considerations and results from Marsh (2002), Jerram et al., (2003), and Cashman (1993).

184 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

The post-emplacement regime refers to the grow inward, using the preexisting plagioclase when both solidifi cation fronts meet in the mid- stage where the saucer-shaped sill has reached clusters as initial building blocks. As defi ned by plane of the sill. At this stage the overall continu- its fi nal size and shape and is no longer supplied Jerram et al. (2003), the initial building blocks ous saturated porous mush is relatively buoyant by melt through the feeding channel. During consist of a mixed population of clusters (in our over the entire saucer-shaped sill. Specifi cally this stage, the sill progressively crystallizes. As case plagioclase clusters). The resulting crystal the buoyancy of the porous mush integrated the solidifi cation front progresses inward in the framework or mush will control and precondi- over the height between the higher outer sill and sill (Fig. 16B) a porosity gradient will arise with tion later textural responses (Cashman, 1993). lower inner sill will dynamically generate an the maximum porosity in the mid-plane region When all conditions required for the fl ow are overpressure ahead of the porosity mush located of the saucer-shaped sill (Figs. 15A and 16B). met, an adiabatic process (see Sec. 9.2.2.b) of in the high region of the saucer-shaped sill, i.e., The growth and nucleation rate of plagioclase porous melt fl ow-induced stresses on the mush the outer sill. This overpressure is caused by will be controlled by the cooling rate of the sill zone triggers a single fl ow event. We propose the buoyancy of the continuous mush zone in (Fig. 16B). Thus the solidifi cation front will that these conditions are met at the moment the mid-plane of the sill (Fig. 17A). Due to the

A Buoyancy-related pressure distribution

Overpressure Overpressure

Underpressure Underpressure B Melt flow pathway Drained region Melt Drained region path Melt Sink path region Formation Formation of Log B of Log B

C Resulting profiles I-shaped D-shaped S-shaped C-shaped Early stage ? related process

Melt flow path

Log A Mg# Log B Mg# Log B' Mg# Mg# Sink region Region unaffected by Drained region Region affected possibly resulting post-emplacement related with by melt flow path melt flow overpressure in C-shape

Figure 17. Schematic presentation of the effect of a buoyancy mush related post-emplacement melt fl ow in a saucer-shaped sill. (A) Dynamic pressure related to the crystalline mush buoyancy at the moment it is completed (i.e., when both the upper and the lower solidifi cation front meet in the mid-plane). Overpressure regions are generated in the elevated region of the sill and conjugated underpressure in the lower region of the sill. (B) The remaining melt fraction is expulsed from the overpressure region and is sucked in by the conjugated underpressure generated in the lower part of the sill. (C) Resulting compositional profi les: I-shape in regions not affected by melt fl ow; D-shape representing a drained region and S-shape are likely to represent the melt fl ow path in the lower part of the inclined sheet. C-shaped profi les are expected to be found in the sink region, i.e., inner sill.

Geosphere, June 2010 185

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

motion of the crystalline mush, an associated such a crystalline framework. This region that fractionation process that reveals the trace of dynamic underpressure region is expected to be sustains the compressive stresses applied on the the saturated porous mush zone. The proposed located at the transition between the inclined- porous network was previously believed to be mechanism is expected to be “instantaneous” sheet and the inner sill (Fig. 17A). Thus the melt diffi cult to identify at the scale of a thin section as compared to the period of time it takes for a is expected to fl ow downward following the (Philpotts and Carroll, 1996). Philpotts and Car- 100-m-thick sill to be fully crystallized. pressure gradient (Fig. 17B). The dynamically roll (1996), however, showed that in plagioclase- overpressured region will expel the saturated rich tholeiitic basalt with only 30% crystals, 10. INTERPRETATIONS melt fraction that will fl ow against gravity down a relatively strong network of crystals exists. to the dynamically underpressurized region They showed that the compressive strength of The symmetrical D-shaped profi les from where it will be sucked in (Fig. 17B). Finally, the mush increases dramatically from 336 to the Golden Valley Sill west limb P14 also because a limited amount of remaining melt 4333 Pa as the percentage of crystal increases show slight coarsening in the central region of fraction will be available, the melt fl ow will stop only from 33% to 37%. Like Marsh (2002), we the sill associated with increasing growth of as soon as it has been expelled from the over- emphasize that this result is relevant to plagio- interstitial clinopyroxene. These grains, how- pressure region. Thus the region that has been clase-rich basalts. It is, however, closely tied to ever, have not developed to the extent of the affected by melt expulsion and fl ow dries up and the specifi c phase equilibria of the considered circular oikocrysts observed in the D-shaped is expected to cool down faster. The feasibility magma and the crystallinity will vary somewhat profi le from the Morning Sun sill P19 and the of this process will be tested in the near future with magma type. S-shaped P5 profi le from the Golden Valley through a numerical code using the fully cou- Thus the presently described textural pattern Sill in the central region of the east limb. Thus pled set of equations developed by Tantserev et is likely to result from different local states of mineral chemistry differences between the two al. (2009) solving the scenario of differentiation pressures at a common temperature for the stud- textural domains could not be tested in the due to post-emplacement melt fl ow. ied magma. In the present model, this texture profi le P14. Nevertheless, published analyses We suggest that the plagioclase domain iden- pattern is interpreted as the result of the forced from Aarnes et al. (2008) from the same profi le tifi ed in our study represents the fossil trace of draining of the melt associated with the melt P14 indicate a homogeneous increase in Mg#

100 Range of olivine Top sill composition in profile P19 margin 100 80 cpx-domain pl-domain 60 90 Ol

40 80 ? Height normalized 1900 1890 20 Liquidus 70 0 40 50 60 70 80 90 60 1700 Solidus 100 50 T oC Liquid 80 40 1500 60 30 Height normalized (meter) 40

Height normalized Olivine 20 1300 20 Cpx 1205 10 0 50 60 70 80 Mg# Fa 20 40 60 80 Fo 0 Wt.% Forsterite Bottom sill Whole-rock Mg# Mineral Core margin chemistry chemistry Rim

Figure 18. Interpretation of the Morning Sun D-shaped profi le, P19. The textural log of the profi le P19 is presented on the left-hand side and corresponds to log B in Figure 4. However, the two textural domains were only identifi ed up to 70% of the total thickness. The contrast in olivine and clinopyroxene profi les between clinopyroxene and plagioclase domains is interpreted as the result of different stages of crystallization. That is, the widest range of compositional variations in olivine found in the upper part of the D-shaped region corresponds to early settlement of olivine from the roof section, whereas the symmetrical compositional profi le yielded by clinopyroxene forms the domain name after it is interpreted as the result of melt fl ow fractionation. Represented in the olivine T-X

diagram are the ranges of olivine composition in both textural domains. The upper boundary for both domains is Fo80, which con-

strains the original melt composition to ~Fo50. The lower boundary of the olivine composition in the clinopyroxene domain possibly indicates the T-P-X conditions at which the porous melt ﬂ ow fractionation occurred.

186 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Magmatic differentiation processes in saucer-shaped sills

in olivine composition at the center of the sill. The D-shaped profi les are interpreted as the over- and under-pressure in a crystalline porous Hence, mineral chemistry as well as whole- results of a post-emplacement related melt fl ow. mush. For a saucer-shaped sill the buoyancy of rock geochemistry show symmetrical profi les The Golden Valley Sill west limb is more devel- the porous mush zone located in the mid-plane and support a process of porous melt fl ow as a oped than its conjugate neighbor. It is formed of the sill triggers melt expulsion and fl ow from mechanism for differentiation in this region of by a continuous inclined-sheet to outer sill the overpressured outer-sill region, down to the the Golden Valley Sill. that is characterized by two D-shaped profi les underpressured region located near the transi- The mineral chemistry in the true D-shaped (e.g., P13 and P14). The P15 profi le located tion between the inner sill and inclined-sheet. part of the profi le P19 from the Morning Sun at the lower part of the inclined-sheet of the sill suggests that a more complex process has sill demonstrates a complicated pattern (Fig. ACKNOWLEDGMENTS occurred. Although whole-rock chemistry 6). Unfortunately, there are no exposed parts The authors acknowledge the careful reviews of and mineral compositional profi les from both of the Golden Valley Sill inner sill that allow Julian S. Marsh and Rais M. Latypov who improved textural domains are symmetrical, the olivine profi le sampling and would complete all the the original paper with their suggestions. We thank composition does not follow this general trend profi le shapes continuation to support our theo- Yuri Podladchikov for suggestions and constructive (Figs. 11 and 18). The olivine composition in retical model. However, our model predicts that discussions, and Timm John and Sergei Medvedev the two distinct textural domains shows a grad- C-shaped profi les may be found close to the for their suggestions and criticism of earlier versions of this paper. Muriel Erambert and John Nicholas ual decrease in Mg# from the bottom contact to transition between inclined-sheet and inner sill. Walsh are acknowledged for help with the analyti- the upper part of the D-shaped profi le (upper This may be confi rmed in a future study of the cal work. Thanks are due to the Norwegian Research boundary at 70 m height [nd]). Because olivine Morning Sun south limb, analyzing in greater Council (NFR) for fi nancial support through the has a higher liquidus temperature than clinopy- detail the vertical variations in the texture and project 159824/V30 “Emplacement mechanisms and magma fl ow in sheet intrusions in sedimentary roxene and plagioclase, the olivine profi le may chemistry. The profi le P21 indeed suggests an basins,” including a doctoral fellowship to Christophe be regarded as representing earlier magmatic asymmetrical C-shaped type of profi le. Further- Galerne. The work has also been supported by a Cen- processes than the post-emplacement melt fl ow more, the location of this profi le in respect to tre of Excellence grant from the Norwegian Research event. The asymmetry eventually suggests that the saucer geometry is at about the transition Council to P.G.P. early fractionated olivine was settling from the between the climbing-sheet and the inner sill. REFERENCES CITED upper region of the sill (Fig. 16A, 1a). This may The X-shaped and S-shaped profi les show explain the decrease in Mg# in the upper margin complexities in processes, possibly resulting Aarnes, I., Podladchikov, Y.Y., and Neumann, E.R., 2008, of the sill, observed, for example, in the upper from the combined effect of processes includ- Post-emplacement melt fl ow induced by thermal part of profi le P19. On the other hand, the sym- ing settling and post-emplacement melt fl ow at stresses: Implications for differentiation in sills: metrical profi le shown by the bulk, plagioclase, different stages. The regional preservation of the Earth and Planetary Science Letters, v. 276, p. 152– 166, doi: 10.1016/j.epsl.2008.09.016. and clinopyroxene compositional profi le are differently shaped profi les related to early mag- Barriere, M., 1976, Flowage differentiation—Limitation of consistent with a post-emplacement related melt matic processes indicates that different degrees Bagnold effect to narrow intrusions: Contributions fl ow through a continuous crystalline mush. of post-emplacement melt fl ow occurred within to Mineralogy and Petrology, v. 55, no. 2, p. 139– 145, doi: 10.1007/BF00372223. Finally, plagioclase clusters appear to be dis- the sill, with the most pronounced effect in the Bowen, N.L., 1915, The later stages of the evolution of tributed over the whole-thickness of the exam- central region of the conjugated limbs of the the igneous rocks: The Journal of Geology, v. 23, p. 1–89, doi: 10.1086/622298. ined profi les, including P19. The core compo- Golden Valley Sill. Bowen, N.L., 1928, The evolution of igneous rocks: Princ- sition in plagioclase clusters is signifi cantly eton, New Jersey, Princeton University Press, 332 p. different from their rim composition. As shown 11. CONCLUSIONS Brandeis, G., and Jaupart, C., 1986, On the interaction between convection and crystallization in cooling in Figure 11, CaO content of the rims is identi- magma chambers: Earth and Planetary Science Let- cal to the plagioclase laths (Figs. 11D and 11E). Our systematic analysis of the Golden Valley ters, v. 77, no. 3–4, p. 345–361, doi: 10.1016/0012 It suggests that the plagioclase clusters were Sill Complex along profi les indicates the fol- -821X(86)90145-7. Cashman, K.V., 1993, Relationship between Plagioclase brought through the feeding conduct and were lowing. (1) All types of compositional profi les crystallization and cooling rate in basaltic melts: homogeneously distributed across the sill. Dur- described in mafi c sills (I-, D-, S-, and possibly Contributions to Mineralogy and Petrology, v. 113, no. 1, p. 126–142, doi: 10.1007/BF00320836. ing the in situ post-emplacement crystallization also C-shaped) may form within a single sill. Cox, K.G., Macdonal, R., and Hornung, G., 1967, Geo- of the sill, plagioclase clusters eventually grew We also show that a systematic space distribu- chemical and petrographic provinces in Karroo rims of identical composition to the plagioclase tion of compositional, textural, and petrological basalts of Southern Africa: The American Miner- alogist, v. 52, no. 9–10, p. 1451–1474. laths formed in situ. variations along profi les laterally exists around Duncan, R.A., Hooper, P.R., Rehacek, J., Marsh, J.S., and Our model suggests that the differently shaped an example of a saucer-shaped sill, the Golden Duncan, A.R., 1997, The timing and duration of the profi les and their distribution around the Golden Valley Sill. (2) The formation of I-, D-, and Karoo igneous event, southern Gondwana: Journal of Geophysical Research. Solid Earth, v. 102, no. Valley Sill results from varying magmatic differ- S-shaped profi les in the GVSC results from a B8, p. 18127–18138, doi: 10.1029/97JB00972. entiation processes at different stages of the mag- shift in types of fractionation processes from Erlank, A.J., Duncan, A.R., Marsh, J.S., Sweeney, R.J., Hawkesworth, C.J., Milner, S.C., Miller, R., and matic evolution of the saucer-shaped sill. Simi- the early to the latest stages of cooling and Rogers, N.W., 1988, A laterally extensive geo- larly to Mangan and Marsh (1992) and Marsh crystallization. (3) The early stage of emplace- chemical discontinuity in the subcontinental Gond- (1996), the I-shaped profi les are interpreted to ment and crystallization of phenocryst-poor wana lithosphere, in Geochemical Evolution of the Continental Crust Conference Abstracts, Pocos de be the result of the early stage of crystallization magma induced compositional homogeniza- Caldas, Brazil, p. 1–10. associated with crystal growth predominantly tion that is later preserved during solidifi cation Frenkel, M.Y., Yaroshevsky, A.A., Ariskin, A.A., Barmina, occurring within the solidifi cation front. During in the post-emplacement stage and refl ected G.S., Koptev-Dvornikov, E.V., and Kireev, B.S., 1988, Dynamics of in situ differentiation of Mafi c the early stage of crystallization isolated gravi- in I-shaped profi les. (4) We propose that the Magmas: Moscow, Nauka, 216 p. (in Russian). tational ordering in the calm part of the cooling D- and S-shaped profi les were generated in a Frenkel, M.Y., Yaroshevsky, A.A., Ariskin, A.A., Barmina, G.S., Koptev-Dvornikov, E.V., and Kireev, B.S., magma at fl uidlike stages may develop, possibly post-emplacement porous melt fl ow regime. 1989, Convective cumulate model simulating the explaining the X-shaped profi le P10. This fl ow was driven by a buoyancy-related formation process of strifi ed intrusions, in Bonin,

Geosphere, June 2010 187

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021 Galerne et al.

B., Didier, J., Le Fort, P., Propach, G., Puga, E., and no. 9, p. 1619–1656, doi: 10.1093/petrology/ Earth and Planetary Science Letters, v. 266, no. 1–2, Vistelius, A. B., eds., Magma-Crust Interactions egg051. p. 195–204, doi: 10.1016/j.epsl.2007.11.015. and Evolution: Athens, Theophrastus, p. 3–88. Latypov, R.M., 2003b, The origin of marginal compositional Rice, A., 1981, Convective fractionation—A mechanism Fujii, T., 1974, Crystal settling in a sill: Lithos, v. 7, no. 3, reversals in basic-ultrabasic sills and layered intru- to provide cryptic zoning (macrosegregation), lay- p. 133–137, doi: 10.1016/0024-4937(74)90026-7. sions by Soret fractionation: Journal of Petrology, ering, crescumulates, banded tuffs and explosive Galerne, C.Y., Neumann, E.-R., and Planke, S., 2008, v. 44, no. 9, p. 1579–1618, doi: 10.1093/petrology/ volcanism in igneous processes: Journal of Geo- Emplacement mechanisms of sill complexes: egg050. physical Research, v. 86, p. 405–417, doi: 10.1029/ Information from the geochemical architecture Latypov, R.M., Chistyakova, S., and Alapieti, T., 2007, JB086iB01p00405. of the Golden Valley Sill Complex, South Africa: Revisiting problem of chilled margins associ- Richardson, S.H., 1979, Chemical variation induced by Journal of Volcanology and Geothermal Research, ated with marginal reversals in mafi c-ultramafi c fl ow differentiation in an extensive Karroo doler- v. 177, no. 2, p. 425–440, doi: 10.1016/j.jvolgeores intrusive bodies: Lithos, v. 99, p. 178–206, doi: ite sheet, southern Namibia: Geochimica et Cos- .2008.06.004. 10.1016/j.lithos.2007.05.008. mochimica Acta, v. 43, no. 9, p. 1433–1441, doi: Galland, O., Planke, S., Neumann, E.R., and Malthe- Malthe-Sørenssen, A., Planke, S., Svensen, H., and Jamtveit, 10.1016/0016-7037(79)90137-6. Sørenssen, A., 2009, Experimental modelling of B., 2004, Formation of saucer-shaped sills, in Breit- Sparks, R.S.J., Huppert, H.E., and Turner, J.S., 1984, The shallow magma emplacement: Application to sau- kreuz, C., and Petford, N., eds., Physical geology of fl uid-dynamics of evolving magma chambers: cer-shaped intrusions: Earth and Planetary Science high-level magmatic systems: Geological Society Philosophical Transactions of the Royal Society Letters, v. 277, no. 3–4, p. 373–383, doi: 10.1016/ [London] Special Publication, p. 215–227. of London Series A—Mathematical Physical and j.epsl.2008.11.003. Mangan, M.T., and Marsh, B.D., 1992, Solidifi cation front Engineering Sciences, v. 310, no. 1514, p. 511, doi: Gibb, F.G.F., and Henderson, C.M.B., 1992, Convection and fractionation in phenocryst-free sheet-like magma 10.1098/rsta.1984.0006. crystal settling in sills: Contributions to Mineral- bodies: The Journal of Geology, v. 100, no. 5, Svensen, H., Planke, S., Chevallier, L., Malthe-Sorenssen, ogy and Petrology, v. 109, no. 4, p. 538–545, doi: p. 605–620, doi: 10.1086/629611. A., Corfu, F., and Jamtveit, B., 2007, Hydrothermal 10.1007/BF00306555. Marsh, B.D., 1989, On convective style and vigor in sheet- venting of greenhouse gases triggering Early Juras- Gorring, M.L., and Naslund, H.R., 1995, Geochemical like magma chambers: Journal of Petrology, v. 30, sic global warming: Earth and Planetary Science reversals within the lower 100 M of the Palisades no. 3, p. 479–530. Letters, v. 256, no. 3–4, p. 554–566, doi: 10.1016/j Sill, New Jersey: Contributions to Mineralogy Marsh, B.D., 1996, Solidifi cation fronts and magmatic evo- .epsl.2007.02.013. and Petrology, v. 119, no. 2–3, p. 263–276, doi: lution: Mineralogical Magazine, v. 60, no. 398, Sweeney, R.J., Duncan, R.R., and Erlank, A.J., 1994, Geo- 10.1007/BF00307286. p. 5–40, doi: 10.1180/minmag.1996.060.398.03. chemistry and petrogenesis of Central Lebombo Gray, N.H., and Crain, I.K., 1969, Crystal settling in sills: A Marsh, B.D., 2002, On bimodal differentiation by solidifi - basalts of the Karoo Igreous Province: Journal of model for suspension settling: Canadian Journal of cation front instability in basaltic magmas, part 1: Petrology, v. 35, no. 1, p. 95–125. Earth Sciences, v. 6, no. 5, p. 1211–1216. Basic mechanics: Geochimica et Cosmochimica Tait, S.R., and Jaupart, C., 1996, The production of chemi- Hansen, D.M., and Cartwright, J.A., 2006, Saucer-shaped Acta, v. 66, no. 12, p. 2211–2229, doi: 10.1016/ cally stratifi ed and adcumulate plutonic igneous sill with lobate morphology revealed by 3D seismic S0016-7037(02)00905-5. rocks: Mineralogical Magazine, v. 60, no. 398, data: Implications for resolving a shallow-level sill Marsh, J.S., Hooper, P.R., Rehacek, J., Duncan, R.A., and p. 99–114, doi: 10.1180/minmag.1996.060.398.07. emplacement mechanism: Journal of the Geologi- Duncan, A.R., 1997, Stratigraphy in age of Karoo Tantserev, E., Galerne, C.Y., and Podladchikov, Y.Y., 2009, cal Society, v. 163, p. 509–523, doi: 10.1144/0016 Basalts of Lesotho and implications for correlations Multi-phase in multi-component porous visco- -764905-073. within the Karoo Igneous Province: Geophysical elastic media, in Ling, H. I., Smyth, A., and Betti, R., Hansen, D.M., Cartwright, J.A., and Thomas, D., 2004, 3D Monograph, v. 100, p. 247–272. eds., The Fourth Biot Conference on Poromechan- seismic analysis of the geometry of igneous sills Marsh, J.S., and Mndaweni, M.J., 1998, Geochemical ics: New York, DEStech Publications, Inc., p. 1179. and sill junctions relationships, in Davies, R. J., variations in a long Karoo dyke, Eastern Cape: Thomson, K., and Hutton, D., 2004, Geometry and growth Cartwright, J. A., Stewart, S. A., Lappin, M., and South African Journal of Geology, v. 101, no. 2, of sill complexes: Insights using 3D seismic from Underhill, J. R., eds., 3D Seismic Technology: p. 119–122. the North Rockall Trough: Bulletin of Volcanology, Application to the Exploration of Sedimentary Naslund, H.R., 1984, Petrology of the Upper Border Series v. 66, no. 4, p. 364–375, doi: 10.1007/s00445-003 Basins: Geological Society [London] Memoir, of the Skaergaard Intrusion: Journal of Petrology, -0320-z. p. 199–208. v. 25, no. 1, p. 185–212. Wager, L.R., and Brown, G.M., 1968, Layered igneous Hansen, D.M., Redfern, J., Federici, F., di Biase, D., and Philpotts, A.R., and Carroll, M., 1996, Physical prop- rocks: San Francisco, WH Freeman. Bertozzi, G., 2008, Miocene igneous activity in the erties of partly melted tholeiitic basalt: Wager, L.R., Brown, G.M., and Wadsworth, W.J., 1960, Northern Subbasin, offshore Senegal, NW Africa: Geology, v. 24, no. 11, p. 1029–1032, doi: Types of igneous cumulates: Journal of Petrology, Marine and Petroleum Geology, v. 25, no. 1, 10.1130/0091-7613(1996)024<1029:PPOPMT> v. 1, no. 1, p. 73–85. p. 1–15, doi: 10.1016/j.marpetgeo.2007.04.007. 2.3.CO;2. Jerram, D.A., Cheadle, M.J., and Philpotts, A.R., 2003, Polteau, S., Ferré, E.C., Planke, S., Neumann, E.R., and Quantifying the building blocks of igneous rocks: Chevallier, L., 2008a, How are saucer-shaped Are clustered crystal frameworks the foundation?: sills emplaced? Constraints from the Golden Journal of Petrology, v. 44, no. 11, p. 2033–2051, Valley Sill, South Africa: Journal of Geophysi- doi: 10.1093/petrology/egg069. cal Research, v. 113, no. B12104, p. 1–13, doi: Latypov, R.M., 2003a, The origin of basic-ultrabasic sills with 10.1029/2008JB005620. S-, D-, and I-shaped compositional profi les by in situ Polteau, S., Mazzini, A., Galland, O., Planke, S., and Mal- MANUSCRIPT RECEIVED 28 JANUARY 2009 crystallization of a single input of phenocryst-poor the-Sorenssen, A., 2008b, Saucer-shaped intru- REVISED MANUSCRIPT RECEIVED 27 OCTOBER 2009 parental magma: Journal of Petrology, v. 44, sions: Occurrences, emplacement and implications: MANUSCRIPT ACCEPTED 8 NOVEMBER 2009

188 Geosphere, June 2010

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/6/3/163/3338493/163.pdf by guest on 02 October 2021