17
Provenance of late Carboniferous sandstones in the Pennine Basin (UK) from combined heavy mineral, garnet geochemistry and palaeocurrent studies C.R. Hallsworth a, , J.I. Chisholm b a HM Research Associates, West Mostard, Garsdale, Sedbergh, Cumbria LA10 5NT, UK b 4 Park Street, Loughborough, Leicestershire, LE11 2EG, UK Received 18 July 2007; received in revised form 7 November 2007; accepted 14 November 2007 Abstract Analyses of heavy mineral assemblages, garnet geochemistry and palaeocurrent directions may each independently provide useful information on the provenance of sandstones, but when such datasets are combined the number of options on the location of possible source areas can be greatly reduced. Also, variations in heavy mineral signatures through a stratigraphical section can be explained in terms of switching between source areas, rather than due to unroofing events in a single source area, if palaeocurrent directions change at the appropriate levels. Likewise, seemingly random changes in transport direction indicated by palaeocurrent variations through a short stratigraphical section can be shown by the heavy mineral data to be systematic and related to shifts between source areas. Illustrations of the mutual constraints provided by such combined datasets are drawn from late Carboniferous successions of the Pennine Basin, northern England, where the contributions of detritus from several distinct source terrains and sediment pathways are well documented [Hallsworth, C.R., Morton, A.C., Claoué-Long, J., Fanning, C.M., 2000. Carboniferous sand provenance in the Pennine Basin, UK: constraints from heavy mineral and detrital zircon age data. Sedimentary Geology 137, 147185; Chisholm, J.I., Hallsworth, C.R., 2005. Provenance of Upper Carboniferous sandstones in east Derbyshire: role of the WalesBrabant High. Proceedings of the Yorkshire Geological Society 55, 209233]. Examples are given of how mineralogically distinct and geographically separated provenances can be identified, how variations in heavy mineral suites can be used to identify whether variations in palaeoflow represent variations in provenance, and how localised mixing/recycling in incised channels can be inferred. The final example defines the regional extent of the widespread northerly-derived fluvial Yeadonian sandstones and provides evidence for the mixing of more local sediment supplies at the basin margins. © 2007 Elsevier B.V. All rights reserved. Keywords: Provenance; late Carboniferous; UK Pennine Basin; Heavy minerals; Palaeocurrents 1. Introduction 1.1. History We deal here with provenance studies relating to sandstones of late Carboniferous (Namurian and Westphalian) age in the Pennine Basin (Figs. 1 and 2), using data mainly from outcrop but extending the regional basis of the study with borehole data. Sedimentology of this succession has been summarised by Collinson (1988) for the Namurian and by Guion and Fielding (1988) for the coal-bearing part of the Westphalian. Sandstones are of fluvial and deltaic origin, having been deposited in a coastal plain setting that at times extended across most of northern Europe (Maynard et al., 1997). Northern provenance of most Namurian sandstones has been established for over a century (Sorby, 1859; Gilligan, 1920), and it was until recently supposed that most of the Westphalian sediment was brought in by the same major transport system (e.g. Cliff et al., 1991; Guion, 1992). However, large-scale inputs from other directions have been recognised, on the basis of palaeocurrents (Collinson and Banks, 1975; Chisholm, 1990), sandstone channel trends (Rippon, 1996), heavy mineral studies (Hallsworth, 1995, 1998, 2001), reworked palynomorphs (McLean and Chisholm, 1996), Available online at www.sciencedirect.com Sedimentary Geology 203 (2008) 196 212 www.elsevier.com/locate/sedgeo Corresponding author. E-mail address: [email protected] (C.R. Hallsworth). 0037-0738/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.sedgeo.2007.11.002

Provenance of late Carboniferous sandstones in the Pennine Basin (UK) from combined heavy mineral, garnet geochemistry and palaeocurrent studies

Embed Size (px)

Citation preview

Available online at www.sciencedirect.com

(2008) 196–212www.elsevier.com/locate/sedgeo

Sedimentary Geology 203

Provenance of late Carboniferous sandstones in the Pennine Basin (UK) fromcombined heavy mineral, garnet geochemistry and palaeocurrent studies

C.R. Hallsworth a,⁎, J.I. Chisholm b

a HM Research Associates, West Mostard, Garsdale, Sedbergh, Cumbria LA10 5NT, UKb 4 Park Street, Loughborough, Leicestershire, LE11 2EG, UK

Received 18 July 2007; received in revised form 7 November 2007; accepted 14 November 2007

Abstract

Analyses of heavy mineral assemblages, garnet geochemistry and palaeocurrent directions may each independently provide useful informationon the provenance of sandstones, but when such datasets are combined the number of options on the location of possible source areas can begreatly reduced. Also, variations in heavy mineral signatures through a stratigraphical section can be explained in terms of switching betweensource areas, rather than due to unroofing events in a single source area, if palaeocurrent directions change at the appropriate levels. Likewise,seemingly random changes in transport direction indicated by palaeocurrent variations through a short stratigraphical section can be shown by theheavy mineral data to be systematic and related to shifts between source areas. Illustrations of the mutual constraints provided by such combineddatasets are drawn from late Carboniferous successions of the Pennine Basin, northern England, where the contributions of detritus from severaldistinct source terrains and sediment pathways are well documented [Hallsworth, C.R., Morton, A.C., Claoué-Long, J., Fanning, C.M., 2000.Carboniferous sand provenance in the Pennine Basin, UK: constraints from heavy mineral and detrital zircon age data. Sedimentary Geology 137,147–185; Chisholm, J.I., Hallsworth, C.R., 2005. Provenance of Upper Carboniferous sandstones in east Derbyshire: role of the Wales–BrabantHigh. Proceedings of the Yorkshire Geological Society 55, 209–233]. Examples are given of how mineralogically distinct and geographicallyseparated provenances can be identified, how variations in heavy mineral suites can be used to identify whether variations in palaeoflow representvariations in provenance, and how localised mixing/recycling in incised channels can be inferred. The final example defines the regional extent ofthe widespread northerly-derived fluvial Yeadonian sandstones and provides evidence for the mixing of more local sediment supplies at the basinmargins.© 2007 Elsevier B.V. All rights reserved.

Keywords: Provenance; late Carboniferous; UK Pennine Basin; Heavy minerals; Palaeocurrents

1. Introduction

1.1. History

We deal here with provenance studies relating to sandstonesof late Carboniferous (Namurian and Westphalian) age in thePennine Basin (Figs. 1 and 2), using data mainly from outcropbut extending the regional basis of the study with borehole data.Sedimentology of this succession has been summarised by

⁎ Corresponding author.E-mail address: [email protected] (C.R. Hallsworth).

0037-0738/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.sedgeo.2007.11.002

Collinson (1988) for the Namurian and by Guion and Fielding(1988) for the coal-bearing part of the Westphalian. Sandstonesare of fluvial and deltaic origin, having been deposited in acoastal plain setting that at times extended across most ofnorthern Europe (Maynard et al., 1997). Northern provenanceof most Namurian sandstones has been established for over acentury (Sorby, 1859; Gilligan, 1920), and it was until recentlysupposed that most of the Westphalian sediment was brought inby the same major transport system (e.g. Cliff et al., 1991;Guion, 1992). However, large-scale inputs from other directionshave been recognised, on the basis of palaeocurrents (Collinsonand Banks, 1975; Chisholm, 1990), sandstone channel trends(Rippon, 1996), heavy mineral studies (Hallsworth, 1995, 1998,2001), reworked palynomorphs (McLean and Chisholm, 1996),

Fig. 1. Palaeogeography of Westphalian times, based on Guion (1992) and Hallsworth and Chisholm (2000). Areas of non-deposition are shown ornamented.Stratigraphical range of transport systems is shown in Fig. 2.

197C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

and radiometric dating of detrital minerals (Glover et al., 1996;Leng et al., 1999; Hallsworth et al., 2000; Evans et al., 2001;Dahlgren and Corfu, 2001). However, it is the combination ofpalaeocurrent and heavy mineral studies (Chisholm et al., 1996;Hallsworth and Chisholm, 2000; Chisholm and Hallsworth,

2005) that has proved most useful in the elucidation of thecomplexities of provenance, as we illustrate here.

Palaeocurrent data and sandstone lithology have previouslybeen used to establish that the Rough Rock sandstones of lateNamurian (Yeadonian) age have a northern provenance, with

Fig. 2. Generalised late Carboniferous section for the Yorkshire–Derbyshire partof the Pennine Basin (Fig. 6), based on Hallsworth and Chisholm (2000) andChisholm and Hallsworth (2005). Variations of provenance are characterised byheavy mineral suites shown in Table 2; transport paths are shown in Fig. 1.Stratigraphical extent of intervals described in text are shown by vertical bars.BB is Better Bed Coal; 80Y is 80 Yard Coal (Greenmoor Rock, Elland Flags,Grenoside Sandstone and Bole Hill Sandstone lie between these coals); RR isRough Rock and Rough Rock Flags, plus equivalents; HF is Haslingden Flags.

198 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

some dilution from a more local source on the southern marginof the basin (Bristow, 1988). Our work on the same stratigraphicinterval combines a regional study of heavy minerals with thepalaeocurrents and, where possible, extends it out towards the

Table 1Classification of late Carboniferous rocks in the Yorkshire–Derbyshire part of theboundaries from Holliday and Molyneux, 2006)

The formation and group boundaries marked by asterisk (⁎) are notably diachronou

basin margins using borehole data. This is a good example ofthe benefit of combining data sets and also shows the limitationsof using the individual forms of data alone.

1.2. Stratigraphy

The Namurian andWestphalian succession in the Yorkshire–Derbyshire part of the Pennine Basin varies between about1.5 and 3.5 km in thickness (Fig. 2), with no major breaks insequence (Smith et al., 1967; Lake, 1999; Waters, 2000). Thestratigraphy is well constrained by widespread marine markerhorizons, non-marine bivalve and miospore zonations, andworked coal seams. A summary of current nomenclature, toformation level, is given in Table 1.

1.3. Heavy mineral analysis

Heavy mineral assemblages are sensitive indicators of sedi-ment provenance. The composition of heavy mineral assem-blages is not, however, entirely controlled by the source rockmineralogy. Variations in hydraulic conditions during deposi-tion can alter the relative abundance of minerals of differentdensities and grain size, while diagenesis and weathering candissolve susceptible minerals. These processes are discussed indetail by Morton (1985a) and Morton and Hallsworth (1999). Itis crucial for accurate provenance studies that these factors areaccounted for and two complementary approaches have beenrecommended to determine the provenance-related features ofheavy mineral assemblages (Morton and Hallsworth, 1994).The first relies on the ratios of the abundance of minerals withsimilar hydraulic and diagenetic behaviour, such as monazite

Pennine Basin (based mainly on Powell et al., 2000, with international stage

s.

199C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

and zircon. These ratios are determined from the very fine grainsize fraction (63–125 µm), which minimises any difference inhydraulic behaviour. The second approach concentrates on thevarietal characteristics of a single mineral group, as revealed, forexample by garnet geochemistry. Both approaches have beenused to determine provenance characteristics of Namurian–Westphalian sandstones in the Pennine Basin.

The heavy mineral ratios that have proved most useful hereare monazite:zircon (MZi), and chrome spinel:zircon (CZi).These compare the abundances of very stable minerals withsimilar densities; they are thusminimally affected by the changesin hydraulic conditions during sedimentation and are sensitiveprimarily to the variations in provenance. The garnet:zircon(GZi) ratio is also useful, but garnet is less stable and can dissolveunder both deep burial and extreme weathering conditions. Deepburial is likely to have had some effect on the amount of garnet,given that themaximumdepth of burial for the basal Namurian inthe East Pennine coalfields would have been about 3000 m(Fraser et al., 1990). Garnet dissolution is active at these depthsin Paleocene sandstones in the Central North Sea (Morton andHallsworth, 1999) andmay have been accentuated in the PennineBasin if heat flow was higher there. This possibility is supportedby an observed increase in the degree of surface etching of garnetgrains in the lower part of the sequence (Namurian and earlyWestphalian). Wide variations in the apatite:tourmaline ratio(ATi) suggest that apatite has been variably dissolved by contactwith acidic waters during floodplain storage or at present dayoutcrop and this process is also likely to have caused variabledissolution of garnet. High GZi values can therefore beconsidered a representative of provenance, but low GZi valuesmay be a result of either provenance or garnet dissolution. Thepresence of low GZi in many sandstones has reduced the effec-tiveness of garnet geochemistry as a provenance tool. However,it has still proved useful and has been used to identify thepresence of additional detrital supplies not obvious from theheavy mineral suite. Further information about the use of garnetgeochemistry in identifying and characterising different sourceareas is given in Morton (1985b) and Morton et al. (2004).

Using these techniques a number of different heavy mineralsuites have been identified in the Pennine Basin (Hallsworth andChisholm, 2000; Chisholm and Hallsworth, 2005). The maindistinguishing features of these suites are summarised in Table 2.

Table 2Definition of heavy mineral suites

Heavy mineral suites Heavy mineral ratios

MZi CZi GZi

Mexborough 1–10 (mostly N5) 1–24 (mostly 1–13) OriginallClifton mostly b2 34–66 OriginallGreenmoor mostly b2 4–25 OriginallBarnsley mostly b2 0–4 OriginallElland [Westphalian] 2–13 b0.5 OriginallMillstone Grit [Namurian] 5–19 b0.5 Originall

Py is pyrope, gr is grossular and sp is spessartine.⁎ Some sandstones with these heavy mineral suites may have low GZi: this is ppresent day outcrop.

Heavy mineral suites derived from the north are described asMillstone Grit suite (in the Namurian) or Elland suite (in theWestphalian). Both suites are characterised by high MZi andvery low CZi, but there was an overall decrease in the abundanceof monazite supplied by the northern transport system throughtime, with the Namurian sandstones having overall higher MZi(5–19) than the Westphalian (MZi 2–13). Westerly derivedsandstones, which are identified by very low MZi and GZi, areclassified as Greenmoor, Barnsley or Clifton suites dependingon their CZi content. Southerly derived sandstones have a rangeof MZi and CZi values, along with distinctive garnet geo-chemistry, and are described as having a Mexborough suite.Deviations from these main suites are found towards the basinmargins, owing to mixing with locally derived detritus.

1.4. Palaeocurrent data

All the palaeocurrent measurements were made on cross-bedding foresets, and the procedure followed was the same inboth the studies cited (Hallsworth and Chisholm, 2000;Chisholm and Hallsworth, 2005). At each locality all accessibleforeset dips were recorded, and a stereographic correction wasapplied for any tectonic dip considered to be 5 degrees or more.Corrected foreset dips less than 10 degrees were then rejected,on the ground that low-angle depositional dips are present inmany of these sandstones, and the relationship of these topalaeoflow direction is not known. Sandstones with less than 5readings in total were omitted. Vector mean values werecalculated for each locality where measurements were made,and for all the measurements from each sandstone unit. Thelatter figures are those used in the first two examples describedbelow. The third example uses individual foreset directions andlocality means, together with data from Bristow (1988).

The great majority of late Carboniferous sandstones in thePennine Basin are of fluvial origin (Collinson, 1988; Guion andFielding, 1988; Guion et al., 1995) and the palaeocurrentdirections shown by cross-bedding relate to fluvial subenviron-ments, mainly channels, crevasse splays, and mouthbars ofdeltas.

The stratigraphic variations in provenance that occurredduring the deposition of Namurian and Westphalian sandstonesin Yorkshire and Derbyshire, as identified by combined heavy

Garnet chemistry Palaeocurrents Provenance

y high ⁎ Multiple components (see Fig. 5) From SE and E Southeasterny low [garnet rare or absent] From W Westerny low [garnet rare or absent] From W Westerny low [garnet rare or absent] From W Westerny high⁎ Py-rich, gr-poor garnets From N and E Northerny high⁎ Py-rich, gr-poor garnets From N and E Northern

robably due to dissolution of garnet during sediment transport and possibly at

Fig. 3. Crossplots relating heavy mineral ratios to palaeocurrent vectors for Westphalian-age sandstones of differing provenance in Yorkshire. Data from Hallsworthand Chisholm (2000). Heavy mineral values are for individual samples from a particular stratigraphic unit; palaeocurrent directions are averages of all readings fromthe unit, expressed as degrees clockwise from north. a, early to mid Westphalian sandstones: northern provenance (black symbols) contrasts with western provenance(open symbols). b, late Westphalian sandstones: southeastern provenance.

200 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

mineral and palaeocurrent data sets (Hallsworth and Chisholm,2000; Chisholm andHallsworth, 2005), are summarised in Fig. 2.

2. Identifying mineralogically distinct and geographicallyseparated sediment sources: Pennine Coal Measures Group

The first example derives from a study of heavy mineralassemblages and palaeocurrent directions for a set of Westpha-lian sandstones in the Yorkshire part of the Pennine Basin(Hallsworth and Chisholm, 2000).

Early to mid-Westphalian sandstones (Fig. 2, Langsettian andDuckmantian) display a simple interplay between two miner-

alogically distinct and geographically separated sedimentsources. Fig. 3a shows crossplots of heavy mineral valuesagainst palaeocurrent directions. The data points are pattern-coded for heavy mineral suites that were defined and named byHallsworth and Chisholm (2000) on the basis of overall mineralassemblages. The crossplot for monazite abundance shows acorrelation between higher MZi values, characteristic of theElland suite, and transport from the north and east. Lower MZivalues (mostly b2), characteristic of the Greenmoor, BarnsleyandClifton suites, correlate with transport from thewest. Chromespinel is consistently rare or absent from the northerly derivedElland sandstones but is present in variable amounts in the

Fig. 4. MZi–CZi crossplot comparing ranges of values recorded from different heavy mineral suites in Westphalian sandstones in Yorkshire. Data from Hallsworth andChisholm (2000).

201C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

westerly derived sandstones. The Barnsley has low CZi (0–4),the Greenmoor moderate CZi (4–25) and the Clifton very highCZi (34–66). The garnet abundance plot is complicated by thetendency for grains to dissolve either during periods of weath-ering during transport or after deposition (Morton and Halls-worth, 1999), so low GZi values can occur in any heavy mineralsuite. However, with that proviso, high values are characteristicof the Elland suite, and can be seen to correlate with transport

Fig. 5. Typical garnet compositions from Elland and Mexborough suites, and Bolealmandine plus spessartine (AS), pyrope (P) and grossular (G) end members. Open cgarnets with less than 5% spessartine. Each ternary plot represents a single garnet p

from the north and east. Westerly derived sandstones haveconsistently low garnet levels.

The mineralogy of sandstones deposited in the late Westpha-lian (Fig. 2, Bolsovian and latest Duckmantian), displays a rangeof MZi (1–10) and CZi (1–24), and generally has abundantgarnet. This mineralogy does not match either the northerly-derived (Elland) suite or the westerly-derived (Barnsley, Green-moor or Clifton) suites, and has been categorised separately, as

Hill Sandstone. Analyses expressed in terms of the relative abundances of theircles represent garnets with greater than 5% spessartine, filled circles representopulation determined by electron microprobe analysis of 50 grains.

Fig. 6. Outcrop geology of east Pennine area. ABCD is line of section in Fig. 7.Main towns: Ba Barnsley, Br Bradford, Ch Chesterfield, Db Derby, Ha Halifax,Hd Huddersfield, L Leeds, N Nottingham, S Sheffield, W Wakefield.

202 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

theMexborough suite (Fig. 3b). As shown on Fig. 4, the range ofMZi and CZi recorded by many of the Mexborough samplescould suggest that they were derived by a mixing of the northernand the western derived sediment either through recycling ormixing of transport systems. However, this possibility isdisproved by the garnet geochemistry (Fig. 5), which clearlydifferentiates the Mexborough suite from the earlier northerlyderived Elland-suite sandstones (westerly derived detritus isgenerally devoid of garnet). However, since it is not realisticallypossible to analyse garnets from every sample, analysis ofpalaeocurrent direction is very important in establishing the

Fig. 7. Summary of heavy mineral and palaeocurrent data for the early Westphalian sYorkshire and Derbyshire. Top, Generalised stratigraphic section based on outcrop a2000, Fig. 5). Line of section ABCD and borehole sites are shown in Fig. 6. Centre, Hand Chisholm (2000) and Chisholm and Hallsworth (2005). Bottom, Cross-bedding reand Chisholm and Hallsworth (2005).

presence of Mexborough-suite sandstones. These sandstoneswere deposited by currents flowing mainly towards the north-west and represent a major change in provenance (Fig. 3b).

The palaeocurrent data constrain the heavy mineral results,showing that the Greenmoor, Barnsley and Clifton suites werederived from a region that lay to thewest of the Pennine Basin, theElland suite was transported into the basin from the north and east,whereas the Mexborough suite was transported from the south-east. The heavy mineral values alone would have given much lessprecise indications of provenance. For example, chrome spinel,which is an important component of the Greenmoor, Barnsley andClifton suites, could have been derived from the ultramafic rocksin the Rhenish massif (Press, 1986) to the southeast, or from theGullfjellet and Karmøy ophiolites of Southern Norway (Stephenset al., 1985; Rankin et al., 1988) but the palaeocurrent evidence, ofconsistent transport from the west, effectively excludes theseprovenances. The location of the western source has not beenidentified, but possibilities have been discussed by Hallsworthet al. (2000) and Hallsworth and Chisholm (2000). There are anumber of ultramafic rocks to thewest that are potential sources ofthe chrome spinel bearing detritus. Small local sources, such asthe Mona Complex of Anglesey, are considered unlikely to havebeen the main suppliers of chrome spinel because of the large sizeand scale of the Westphalian river systems involved (Rippon,1996). Furthermore, zircon age data rule out significantcontributions from theMona Complex, since the westerly deriveddetritus lacks late Precambrian to early Cambrian zircons thatwould be associated with derivation from this complex (Halls-worth et al., 2000). Possible ultramafic complexes further afieldinclude large masses of ophiolitic material inferred to have beenemplaced across Ireland during the Ordovician collision of anoceanic arc with the Laurentian continent (Dewey and Mange,1999), or similar slices of oceanic crust obducted farther west inNewfoundland, during the same tectonic episode (Skehan, 1988).The detrital zircon age spectrum from the Clifton Rock, one of thewestern-derived sandstones, indicates that the detritus wasderived from a Laurentian–Baltican terrain. However, the patternof orogenic events in Labrador andNewfoundland does notmatchthe zircon age spectrum exactly, since the Caledonian intrusives inthis area appear to be too young to have sourced the Caledonianzircons in the Clifton Rock (Hallsworth et al., 2000). Althoughexact identification of the original source has so far proveddifficult, the rarity of garnet and of zircons with high uranium andthorium contents suggest that the majority of the westerly derivedsediment is polycyclic. Garnet, although comparatively stable, issubject to dissolution during both surficial weathering and burialdiagenesis (Morton and Hallsworth, 1999), and is thereforeunlikely to survive repeated recycling episodes. Zircons with highuranium and thorium contents are more liable to become meta-mict (i.e. to have their crystal structure compromised). Suchmetamict zircons are likely to become depleted during sediment

uccession between the 80 Yard (Upper Band) and Better Bed (Kilburn) coals innd borehole data (redrawn from Chisholm, 1990, Figs. 3 and 9; and Chisholm,eavy mineral ratios for the four sandstones shown above. Data from Hallsworthadings for the same four sandstones. Data from Hallsworth and Chisholm (2000)

203C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

Fig. 8. Generalised stratigraphy of the Yeadonian Substage of the Namurian inthe central part of the Pennine Basin, between Leeds and the Rossendale sub-basin (Fig. 12), showing the relationship between northerly- and westerly-derived sandstones. Based on McLean and Chisholm, 1996, Fig. 2.

204 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

recycling since they are more readily dissolved during weatheringcompared with the non-metamict grains (Balan et al., 2001), andare also more likely to be mechanically unstable. It is thereforelikely that the sediment was derived largely from pre-existing,mature sandstones (Hallsworth et al., 2000), which may haveformed extensive outcrops in the western source area.

Garnet geochemistry provides the main indication of thelithologies contributing detritus to the northerly-derived Ellandsuite sandstones. Low-grossular, high pyrope almandines areonly formed in high-grade (granulite facies) metasediments orcharnockites. Monazite could also have been derived fromthe high-grade rocks, or from granites intruded into them.Radiometric ages of detrital zircons and monazites in the Ellandsuite sandstones (Cliff et al., 1991; Hallsworth et al., 2000;Evans et al., 2001) suggest a source location in the Caledonianorogen, between Laurentia and Baltica. The heavy mineralsuites do not show any difference in provenance betweensandstones entering the Pennine Basin from the north or theeast. This suggests that differences in palaeocurrent reflectdifferences in transport path rather than differences in sourcearea.

A combination of heavy mineral data and garnet geochem-istry (Fig. 5) was required to show that the Mexborough suitesandstones were derived from a heterogenous source terraincomprising low to moderate grade metasediments, high gradebasic gneisses, granites and ultramafic rocks. This is discussedmore fully by Hallsworth and Chisholm (2000). Although theselithologies are present in Scandinavia, to the northeast, thepalaeocurrent evidence indicates that the source area lay to the

south, probably within the uplifting Variscan orogenic belt. Thishas been confirmed by zircon ages in the Mexborough suite,which are either Carboniferous or Late Precambrian to earlyPalaeozoic, diagnostic of derivation from the Variscan high-lands (Hallsworth et al., 2000). Radiometric dating of detritalmonazites from the Dalton Rock, a member of the Mexboroughsuite, also shows the presence of Variscan-age rocks in thesource area (Evans et al., 2001).

3. Identifying whether variations in palaeoflow directionrepresent variations in provenance: Pennine Lower CoalMeasures Formation

The second example illustrates how the northern and westernsediment sources interacted during a short stratigraphic intervalof early Westphalian age in the eastern part of the Pennine Basin(Figs. 2 and 6). A study by Chisholm (1990) identified thecontrasting transport directions in this interval and linked theseto lithological differences, but the absence of other evidencemade it possible to argue that the sediment had arrived in thebasin by different pathways from the same northern source area.The heavy mineral data (Hallsworth and Chisholm, 2000;summarised above) provide the missing evidence, and showthat the sediment transported from different directions doesindeed derive from different source areas.

Fig. 7 summarizes published information on heavy mineralsand transport directions for this stratigraphic interval. At the topis a north–south stratigraphic section based on borehole andoutcrop information summarised by Chisholm (1990). Nomen-clature of sedimentary cycles is that of Waters et al. (1996).Heavy mineral samples are numbered 1–21, and within eachsedimentary cycle the sample numbers increase from north tosouth. The heavy mineral data can be identified in the sourcepublications (Hallsworth and Chisholm, 2000; Chisholm andHallsworth, 2005) by stratigraphic level and geographicallocation. Below the stratigraphic section are plotted the threesignificant heavy mineral ratios identified in the first exampledescribed above (Fig. 3a): Elland Flags sandstones and mostGrenoside Sandstones have high MZi and GZi combined withlow CZi, whereas the Greenmoor Rock has low MZi and GZiwith variable but generally high CZi. Palaeocurrent summaries,at the bottom of the diagram, show that Elland Flags sandstoneswere transported from the north, Grenoside Sandstones from theeast, and Greenmoor Rock sandstones from the west, all inagreement with the wider conclusions reached from the largerdatabase (Fig. 3a). Switching between major transport systemsseems to have been associated, in each instance, with a floodingevent that affected the whole Pennine region (Chisholm, 1990).

The Bole Hill Sandstone, which appears to occupy an incisedchannel in the northerly derived Grenoside cycle (see Fig. 7), hasbeen transported from the west and as such could be expected tohave a westerly derived Barnsley, Greenmoor or Clifton typeheavy mineral suite with very low MZi and GZi. The CZi isvariable in the westerly-derived sediments but individual sand-stone units tend to have a relatively restricted range of values.TheBoleHill Sandstone, with its wide range of CZi (0.5–28) andthe ranges of MZi and GZi that are higher than usually recorded

Fig. 9. Regional map of the Pennine Basin showing the heavy mineral sample sites (details are given in Table 3) and palaeocurrent data for Yeadonian sandstones.Open arrows represent the palaeocurrent directions recorded in the present study; solid arrows represent the palaeocurrent directions taken from Bristow (1988).

205C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

in westerly-derived suites, is therefore anomalous. This suggeststhat the westerly-derived sediment was mixed with sedimentfrom other sources. The garnets (see Fig. 5) include a group oflow grossular, high-pyrope almandines, which are typical of thenorthern provenance (see Fig. 5 and example 1), suggesting thatnortherly-derived detritus contributed to the BoleHill Sandstone.A northerly component would also explain the raised MZi. The

additional sediment could have come from erosion of oldernorthern-derived Carboniferous sands, or could have beensupplied via a contemporaneous tributary flowing from thenorthern source area.

The mineralogy also provides evidence of a third source ofdetritus. This has been identified by a second group of garnets,which comprise low pyrope, spessartine-rich almandines (see

Table 3Locality details for Yeadonian heavy mineral samples

Locationnumber

Location Grid reference No. ofsamples

1 Throckley Borehole NZ 1456 6762 42 Rowlands Gill Borehole NZ 1664 5815 43 Woodland Borehole NZ 0909 2769 54 Mousegill Beck NY 8371 1239 55 River Greta SD 6170 7223

to 6468 71841

6 Waters Farm Borehole SD 7537 6763 37 Winksley Borehole SE 2507 7151 48 Farnham Borehole SE 3469 5996 59 M65 motorway cut, Stanworth

(east) (UH)SD 6419 2448 1

10 M65 motorway cut, Stanworth (west) SD 6402 2450 111 Old quarry near Heap Clough,

Haslingden (LH)SD 7642 2356 3

12 Greens Clough SD 892 260 313 Greens Clough, lip of waterfall (LH) SD 8938 2615 114 Heys Brittania Quarry, Facit (UH) SD 8846 2012 115 Green Road, Baildon SE 1500 3900 216 Apperley Lane Quarry SE 1970 3920 217 Poplar Road, Shipley SE 1520 3660 218 Beaumont Park, Huddersfield SE 1278 1458 219 Askern No.1 Borehole SE 5652 1502 420 Butterwick Borehole SE 8421 0563 121 Corringham No 1 Borehole SK 8940 9276 422 Coombes Rocks SK 0194 9200 123 Old quarry, Delf Hill, Foldrings SK 2968 9392 224 A6 Bypass Borehole 603

(Whaley Bridge)SJ 9957 8474 4

25 A6 Bypass Borehole 641(Whaley Bridge)

SK 0025 8465 2

26 Old quarry, White Edge Moor,near Nether Padley

SK 2617 7856 2

27 Old quarry near Longshaw Estate SK 2650 7860 128 Old quarry, Reeve Edge SK 0128 6966 129 Old quarry, Biddulph Park SJ 9016 6165 130 Bald Stone SK 0154 6382 131 Rock End, Biddulph SJ 8996 5603 132 Llangollen canal aqueduct (Gwespyr

Sandstone, formerly Aqueduct Grit)SJ 2710 4210 2

33 Wetley Rocks SJ 9660 4920 234 Old quarry, Noonsun Common SK 0158 4974 135 Old quarry, Gimmershill SK 0332 4725 136 Bullbridge rail cut SK 359 521 337 Old quarry, Pinchom's Hill, Belper SK 3595 4691 238 Old quarry near Horsley Park Farm,

CoxbenchSK 3779 4331 1

39 Old quarry, Castle Farm, Coxbench SK 3740 4320 640 Carvers Rock SK 3320 2270 741 Melbourne Quarry SK 3825 2495 242 Melbourne (High Wood) Borehole SK 3820 2374 1

LH Lower Haslingden Flags, UH Upper Haslingden Flags. All others are RoughRock, Rough Rock Flags or their presumed lateral equivalents.

206 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

Fig. 5). This type of garnet has not previously been identified inthe early part of the Westphalian. However, sediment shed fromthe Wales–Brabant High into the North Staffordshire Basinduring the early Namurian (Trewin and Holdsworth, 1972)contains similar garnets (Hallsworth, unpublished information),as do sediments shed from the same High into the WidmerpoolGulf in the latter stages of the Namurian (see next section). Thisdetrital source is also associated with variable amounts ofchrome spinel (Hallsworth, unpublished information) andvariable amounts of detritus from the Wales–Brabant Highcould therefore have helped create the wide range of CZirecorded in the Bole Hill Sandstone. The heavy mineral suitesand garnet geochemistry therefore point to an admixture ofwestern and northern-derived detritus as well as input from asource presumably located towards the south, on the Wales–Brabant High (Chisholm and Hallsworth, 2005). The widevariation in mineralogy between the samples suggests that thevarious components were not thoroughly mixed by the timethey reached the study area. This in turn favours the idea that thenortherly and the southerly derived components were added tothe main west-to-east flow at locations not far to the west of thepresent outcrop.

4. Identifying the regional extent of a large-scale fluvialsystem, and mixing at the basin margins: late Namuriansandstones

The third study also looks at the problems of identifying thepresence of mixing from different detrital supplies, but on aregional, rather than local, scale. Yeadonian (youngestNamurian) sandstones have been chosen to demonstrate thevalue of combining heavy mineral and palaeocurrent data,because this substage includes, at its top, the Rough Rock,which is a well-studied, laterally extensive sandstone in thePennine Basin. It has long been recognized that the feldspathicRough Rock and preceding Rough Rock Flags have a northernprovenance (Gilligan, 1920) and entered the Pennine Basinfrom both the north and east (Bristow, 1988, 1993). However,there is evidence of an increase in the quartz: feldspar ratiotowards the southern margin of the basin, which suggests thatthis part of the basin may also have received detritus from theadjacent Wales–Brabant High (Bristow, 1988). Earlier in theYeadonian there is evidence of detritus entering the PennineBasin from the west (Collinson and Banks, 1975; Bristow,1988; Maynard, 1992): the Haslingden Flags were deposited inthe area of the E–W trending Rossendale sub-basin (Figs. 8and 9), and are lithologically different to the northerly-derivedsediment, being greenish grey in colour and richer in mud clasts(McLean and Chisholm, 1996). There was therefore significantpotential for identifying mixing of different sediment supplysystems at the margins of the basin during the Yeadonian. Thestratigraphic relationships of Yeadonian sandstones are shownin Fig. 8.

Most previous studies of Yeadonian sandstone provenancehave been restricted to outcrops in the central and southern partsof the Pennine Basin (Figs. 1 and 9). To increase the regionalcoverage and extend the study into the margins of the basin, the

present work has included isolated outcrops of Yeadoniansandstones in North Wales and has supplemented the outcropdata set with borehole material. This has enabled the heavymineral study to establish provenance for isolated outcrops andsubsurface regions, by comparing them with sandstones fromthe more intensively studied areas. Unfortunately there are nopalaeocurrent data from the newly sampled areas, and the

Fig. 10. MZi–CZi crossplots demonstrating the Yeadonian data from different areas of the Pennine Basin.

Fig. 11. Garnet compositions from different Yeadonian sandstones. Top row shows the typical ‘northern’ garnet plots; bottom row shows garnets probably derivedfrom the Wales–Brabant High. Sample locality details are given in Table 3. Analyses expressed in terms of the relative abundances of the almandine plus spessartine(AS), pyrope (P) and grossular (G) end members. Open circles represent garnets with greater than 5% spessartine, filled circles represent garnets with less than 5%spessartine. Each ternary plot represents a single garnet population determined by electron microprobe analysis of 50 grains.

207C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

Fig. 12. Yeadonian palaeogeography and sediment transport paths inferred from heavy mineral data and palaeocurrent directions.

208 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

problems arising from the interpretation of heavy mineral datawithout palaeocurrent information emphasizes the importanceof combining both data sets. Details of the sample locations aregiven in Table 3 and shown on Fig. 9.

Work by Bristow (1988, 1993) has established that a widebody of northerly-derived fluvial sheet sandstones (herereferred to as the ‘north lobe’ of the Rough Rock) was spreadsouthwards across the central Pennine Basin. A second flow

209C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

from the north entered the Pennine Basin from the east, andadvanced along the northern margin of the Wales–Brabant High(Bristow, 1988; George, 2001). This flow led to the depositionof the Rough Rock ‘south lobe’ (Chisholm and Hallsworth,2005).

Data from the Yorkshire area, where the Rough Rock is welldeveloped, show that the main spread of Rough Rock (northlobe) contains a typical ‘Millstone Grit’ heavy mineral suite,having high MZi (Fig. 10a) and abundant high pyrope, lowgrossular garnets (Fig. 11, localities 12 and 15). As shown bythe garnet composition (Fig. 11, locality 32), the occurrence ofthis type of suite stretches southwestwards into Wales (Fig. 12),confirming that the high MZi Gwespyr Sandstone (formerAqueduct Grit) in the Wrexham area can be correlated with theRough Rock (Ramsbottom, 1974, fig. 39).

Fig. 13. Summary of heavy mineral and palaeocurrent data for the Rough Rocksandstone in quarries in the Coxbench area. Localities are shown in Table 3:locality 38 is top 5 m of section and locality 39 is remainder of section.Sandstone is mainly medium-grained and cross-bedded in sets up to c. 1 m thick,with some low-angle bedding. Arrows show the direction of individual cross-bed sets; those at the left of column relate to the beds sampled for heavy mineralanalysis. Sample A (MZi 4.9, CZi 0) has heavy mineral suite close to that fornorthern Millstone Grit, but samples B–G have MZi too low or CZi too high forthat suite. Garnet geochemistry of sample G is different from the northernMillstone Grit type (see Fig. 11, locality 39). Despite the mineral diversity,cross-bedding shows consistent palaeoflow from east to west.

4.1. Widmerpool Gulf

The Rough Rock ‘south lobe’, which was deposited in theWidmerpoolGulf at the southernmargin of the PennineBasin, hasconsistent east-to-west flowing palaeocurrents (Figs. 9 and 12).Some of the samples in the south lobe have typical Millstone Gritassemblages (Hallsworth, 1998), with high MZi, low CZi(Fig. 10b) and abundant high pyrope, low grossular garnets(Fig. 11, locality 42), indicating detritus from the same source asthat supplying the main lobe. However, other sandstones from the‘south lobe’ (Fig. 10b) have lower overall MZi (1.5–4.8), manyhave small amounts of chrome spinel (CZi up to 2.4) and areassociated with a different garnet suite (Fig. 11, localities 39, 40and 41), indicating an input from another source. A detailedsection through the Rough Rock at Coxbench Quarry shows thefull range of mineralogical variations, all associated withconsistent palaeocurrents (Fig. 13). A similar range of mineralogy(Hallsworth, 1998), associated with consistent east-to-westpalaeocurrents, is also present around Melbourne on the southernmargin (Hathern Shelf) of the Widmerpool Gulf. In isolation, theheavy mineral data could have been taken to suggest variableinterplay between northern and western derived material, but thepalaeocurrent data show that supply from the west is unlikely.Garnet data from Coxbench and Melbourne Quarry (Fig. 11,localities 39 and 41) provide constraints on the origin of thesecond source. The assemblage is bimodal, comprising one groupof spessartine-rich, low pyrope, variable grossular garnets, andanother group of high pyrope, high grossular garnets. Thisassemblage is similar to those found in the Old Red Sandstone onthe Wales–Brabant High (Hallsworth et al., 2000), and suggeststhat variable local input from this adjacent landmass caused theobserved heterogeneity in mineralogy. Samples from the nearbyCarvers Rock (locality 40) contain both the northerly-derived andthe locally-derived garnet types, indicating that an intrabasinalmixing has taken place. The picture is further complicated by thepresence of some samples containing typicalMillstone Grit heavymineral suites (high MZi) in association with entirely locally-derived garnets. This could be explained by prolongedweatheringof northerly-derivedmaterial during transport prior to entering theWidmerpool Gulf, causing depletion of garnet, followed byaddition of sediment from the Wales–Brabant High, introducinglocally-derived garnets.

The data from the Rough Rock in the Widmerpool Gulf aretherefore compatible with the regional picture (Fig. 12; Bristow,1988; Maynard, 1992; Bristow, 1993; George, 2001), whichshows that the main lobe did not extend into the southeasternpart of the basin, but that a second lobe entered that area fromthe east. This lobe was ultimately fed from the north, but wasdiverted westwards along the south margin of the basin becauseof the existence of the Wales–Brabant High. The involvementof the northern source is diagnosed by the presence of typicalMillstone Grit heavy mineral suites identical to those of themain Rough Rock north lobe in Yorkshire. Additional materialwas added from the only remaining direction, namely theWales–Brabant High to the south, supporting the view ofBristow (1988) that sediment from a local source contributed tothe south lobe. The variability of the heavy mineral assemblages

210 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

and garnet compositions suggests incomplete mixing andhomogenisation between the northern and the southernsediment, in turn suggesting that the southern entry pointswere relatively local.

4.2. North Staffordshire sub-basin

The North Staffordshire sub-basin lies to the west of theWidmerpool Gulf, in the southwest part of the Pennine Basin(Fig. 12). The Rough Rock in this area shows palaeocurrentorientations ranging between north, west and south (Fig. 9).Bristow's palaeogeographical reconstruction (1988, fig. 11.11)suggested that this area was fed exclusively by the ‘south lobe’river. However, the heavy mineral data dispute this, because allthe samples here have a typical Millstone Grit (northern) heavymineral suite (Fig. 10c). There is no evidence to suggest thatsoutherly derived detritus, which is so prominent in the southlobe in the Widmerpool Gulf, entered the North Staffordshirearea at all. Heavy mineral data therefore indicate that the RoughRock in the North Staffordshire sub-basin is an extension of themain Rough Rock ‘north lobe’, rather than the ‘south lobe’.This conclusion could not be reached on the palaeocurrentevidence alone, since the variations in the flow orientationsare too wide to distinguish between the two river systems.Incidentally, there is no evidence for a gap in the original areaof Rough Rock deposition between the undoubted north lobeexposures around Buxton (Fig. 9, locality 28) and the NorthStaffordshire exposures (Fig. 9, localities 29–31, 33–35). Thegap shown by Bristow (1988) may relate to a gap in thepresent-day exposures.

4.3. Gainsborough Trough

The Gainsborough Trough lies to the south of the MarketWeighton High (Fig. 12). The only data in this area are fromboreholes and consequently palaeocurrent information islacking. During Yeadonian times the northwest–southeasttrending structural grain does not appear to have exerted anycontrol on the sedimentation, with deltas entering the area fromthe northeast (Steele, 1988; Fraser and Gawthorpe, 2003). TheAskern borehole, which lies at the western end of the Trough,has a range of MZi (3.2–7.4) approaching that of the typicalMillstone Grit suite, indicating predominantly northern deriva-tion (Fig. 10d). However, towards the east of the Trough(Butterwick and Corringham boreholes), MZi values are lower(1.3–2.9) and small amounts of chrome spinel are present (CZi0.5–3.8). Garnet geochemistry of a sample from the Butterwickborehole (Fig. 11, locality 20) proves the involvement of thenorthern supply system, but the lower MZi and higher CZi hereand at Corringham borehole (Fig. 10d) suggest dilution by amore local sediment supply. In the context of deltaic prograda-tion from the northeast (Steele, 1988), the most likely source fora local sediment supply would be the Market Weighton High orthe Mid North Sea High (Fig. 12). These highs are thought tohave subsided more slowly than the surrounding basins and tohave accumulated thin coverings of Carboniferous sediment(Collinson et al., 1993; Fraser and Gawthorpe, 2003). However,

the heavy mineral data suggest that there must have been, at thevery least, enough exposure to allow erosion and recycling ofpre-existing sediments. These could not have had the typicalNamurian ‘northern’ provenance and so were likely to have beenpre-Namurian in age. The only other possible area from whichthe Gainsborough Trough could have received locally derivedsediment would have been to the south, on the Wales–BrabantHigh. However, this is considered unlikely as it goes against theapparent direction of transport and because in the WidmerpoolGulf, Yeadonian sandstones derived from the Wales–BrabantHigh contain a distinctive garnet suite (Fig. 11, localities 39–42), which has not been found in the stratigraphically equivalentsandstones of the Gainsborough Trough.

4.4. Alston Block

Dilution of the main Rough Rock northern detrital supply isalso evident in the area of the Alston Block. The samples fromthe Alston Block area are all taken from boreholes (Table 3) andno palaeocurrent data are available. The main ‘north lobe’ ofYeadonian sediments would probably have been routed overthis area and, as expected, samples of this age from theWoodland and Throckley boreholes (Mills and Hull, 1968;Mills and Holliday, 1998) have a typical ‘northern’ MillstoneGrit heavy mineral suite (compare Fig. 10e with 10a). However,sandstones from the Rowlands Gill borehole have lower MZi(0.5–4.8) and one sample has a raised CZi of 2.4 (Fig. 10e),suggesting dilution of the northern sediment with other material.The sandstones in the Rowlands Gill borehole have beenassigned to the Marsdenian–Yeadonian (Mills and Holliday,1998, Fig. 12), and consequently they may or may not bedirectly equivalent to the Rough Rock. Nevertheless, the dataimply that there were additional localized detrital suppliesentering this region. Considering the palaeogeography of thearea (Fig. 12), the most likely source for such a supply is to thenorth, in the Southern Uplands of Scotland, or a localized highto the east, in the region of the Mid North Sea High.

4.5. Rossendale sub-basin

The Haslingden Flags were deposited on the western side ofthe basin in the early Yeadonian prior to the Rough Rock Flags(Fig. 8). As shown on Fig. 9, they have palaeocurrents thatindicate derivation from the west (Collinson and Banks, 1975;Bristow, 1988), and are lithologically different from northernderived sandstones (McLean and Chisholm, 1996). The UpperHaslingden Flags are distinctive in having high CZi with little orno monazite (see Fig. 10f) and would be assigned to theGreenmoor suite of the general classification (Table 2): there isno evidence to suggest any involvement of detritus from themonazite-bearing northern provenance. Although the earlierLower Haslingden Flags are lithologically similar, they arecharacterised by lower CZi and low but significant amounts ofmonazite (Fig. 10f). This range of mineralogy suggests somemixing of western and northern detritus. Because the palaeo-flow was consistently from the west, the most likely explanationis that the western transport system initially cut down through

211C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

and recycled underlying northerly derived Namurian sediments.After their appearance in the Haslingden Flags, sandstonesderived from the west became increasingly common through theearly and mid Westphalian (Fig. 2). The similarity of thelithologies and heavy mineral suites in these westerly derivedsandstones, and the common occurrence of reworked LowerPalaeozoic palynomorphs in them (McLean and Chisholm,1996; Chisholm et al., 1996), indicates that they all had a similarprovenance, as discussed in the first example described in thispaper. Further details are given by Hallsworth et al. (2000).

5. Conclusions

Three examples drawn from the late Carboniferous infill ofthe Pennine Basin show how datasets on heavy mineral suites,garnet chemistry and palaeocurrents can be combined todistinguish between sediments derived from different sourceareas. Different transport paths from a single source area canalso be recognised, and explanations for mixed mineralogiescan be proposed.

The first example applies the combined approach on a broadscale, to theWestphalian succession in a large sector of the basin,and reveals the major shifts of provenance that took place there.The main sediment supply systems to this part of the basin havedifferent mineralogies and are shown by the palaeocurrent datato have been derived successively from the north, the west, andthe southeast. A persistent split in the northern river system canbe inferred from the palaeocurrent directions, in that northern-derived sediment entered the basin at its eastern side as well asfrom the north.

The second example looks in detail at the interplay betweennorthern- and western-derived sediment during a short intervalin the early Westphalian. Palaeoflow variations, in themselvesof uncertain significance, are shown to correlate with miner-alogical variations that can only be explained by shifts ofprovenance. Mixed mineralogies in one bed transported fromthe west, the Bole Hill Sandstone, may be linked to incision intoearlier sandstones of northern origin, and to addition of materialfrom a subsidiary basin-margin source located to the south.

In the third example, the areal extent of the main northernriver system during the latest Namurian has been established byextending the heavy mineral dataset across a wider region thanthat covered by palaeocurrent data. Heavy mineral ratios showthat the characteristic signature of the northern-derived RoughRock of the central Pennines can be recognized over a largearea, notably in the Gwespyr Sandstone of North Wales and thelate Namurian sandstones of northeast England. Dilution of thisheavy mineral suite by sediment from subsidiary sources can beinferred at the northern and southern basin margins, and alsodownstream from an intrabasinal high situated to the northeastof the Gainsborough Trough. This implies that some part of theMid North Sea High or the Market Weighton High was emer-gent during the late Namurian. A further conclusion is that themain ‘north lobe’ branch of the river system extended into theNorth Staffordshire area, not the ‘south lobe’ as had previouslybeen thought. This would suggest that the south lobe did notextend beyond the Widmerpool Gulf.

As these examples show, heavy mineral provenance studiesare enhanced by linking the data with palaeocurrents, and like-wise provenance studies that depend primarily on palaeocur-rents are greatly improved when linked to information on theheavy minerals. Provenance studies based on outcrop can easilycombine the two approaches. However, palaeocurrent data fromborehole material are very limited, and under such circum-stances provenance studies have to rely on heavy mineral data,with the possibility of reduced precision.

Acknowledgements

We would like to thank Andy Morton for all his help andSarah Davies, Colin Waters, John Collinson, and an anonymousreviewer for their constructive comments which have consider-ably improved this paper.

References

Balan, E., Neuville, D.R., Trocellier, P., Fritsch, E., Muller, J.-P., Calas, G.,2001. Metamictization and chemical durability of detrital zircon. AmericanMineralogist 86, 1025–1033.

Bristow, C.S., 1988. Controls on the sedimentation of the Rough Rock Group(Namurian) from the Pennine Basin of northern England. In: Besly, B.M.,Kelling, G. (Eds.), Sedimentation in a Synorogenic Basin Complex: TheUpper Carboniferous Of Northwest Europe. Blackie, Glasgow and London,pp. 114–131.

Bristow, C.S., 1993. Sedimentology of the Rough Rock: a Carboniferous braidedriver sheet sandstone in northern England. In: Best, J.L., Bristow, C.S. (Eds.),Braided Rivers, 75. Geological Society Special Publication, pp. 291–304.

Chisholm, J.I., 1990. The Upper Band — Better Bed sequence (Lower CoalMeasures, Westphalian A) in the central and south Pennine area of England.Geological Magazine 127, 55–74.

Chisholm, J.I., 2000. Geology of the Huddersfield East area (SE 11 NE). BritishGeological Survey Technical Report, WA/00/01.

Chisholm, J.I., Hallsworth, C.R., 2005. Provenance of Upper Carboniferoussandstones in east Derbyshire: role of the Wales–Brabant High. Proceedingsof the Yorkshire Geological Society 55, 209–233.

Chisholm, J.I., Waters, C.N., Hallsworth, C.R., Turner, N., Strong, G.E., Jones,N.S., 1996. Provenance of Lower Coal Measures around Bradford, WestYorkshire. Proceedings of the Yorkshire Geological Society 51, 153–166.

Cliff, R.A., Drewery, S.E., Leeder, M.R., 1991. Sourcelands for theCarboniferous Pennine river system: constraints from sedimentary evidenceand U–Pb geochronology using zircon and monazite. In: Morton, A.C.,Todd, S.P., Haughton, P.D.W. (Eds.), Developments in SedimentaryProvenance Studies, 57. Geological Society of London, Special Publication,pp. 137–159.

Collinson, J.D., 1988. Controls onNamurian sedimentation in theCentral Provincebasins of northern England. In: Besly, B.M., Kelling, G. (Eds.), Sedimentationin a Synorogenic Basin Complex. Blackie, Glasgow and London, pp. 85–101.

Collinson, J.D., Banks, N.L., 1975. The Haslingden Flags (Namurian G1) ofsouth-east Lancashire; bar-finger sands in the Pennine Basin. Proceedings ofthe Yorkshire Geological Society 40, 431–458.

Collinson, J.D., Jones, C.M., Blackbourn, G.A., Besly, B.M., Archard, G.M.,McMahon, A.H., 1993. Carboniferous depositional systems in the South-ern North Sea. In: Parker, J.R. (Ed.), Petroleum Geology of NorthwestEurope: Proceedings of the 4th Conference. The Geological Society,London, pp. 677–687.

Dahlgren, S., Corfu, F., 2001. Northward sediment transport from the lateCarboniferous Variscan mountains: zircon evidence from the Oslo Rift,Norway. Journal of the Geological Society (London) 158, 29–36.

Dewey, J., Mange, M., 1999. Petrography of Ordovician and Silurian sedi-ments in the western Irish Caledonides: tracers of a short-lived Ordoviciancontinent-arc collision orogeny and the evolution of the LaurentianAppalachian–Caledonian margin. In: MacNiocall, C., Ryan, P.D. (Eds.),

212 C.R. Hallsworth, J.I. Chisholm / Sedimentary Geology 203 (2008) 196–212

Continental Tectonics, 164. Geological Society, London, Special Publica-tions, pp. 55–107.

Evans, J.A., Chisholm, J.I., Leng, M.J., 2001. How U–Pb detrital monazite agescontribute to the interpretation of the Pennine Basin infill. Journal of theGeological Society (London) 158, 741–744.

Fraser, A.J., Gawthorpe, R.L., 2003. An atlas of Carboniferous basin evolutionin northern England. Geological Society, London. Memoir, 28.

Fraser, A.J., Nash, D.F., Steele, R.P., Ebdon, C.C., 1990. A regional assessmentof the intra-Carboniferous play of Northern England. In: Brooks, J. (Ed.),Classic Petroleum Provinces, 50. Geological Society, London, SpecialPublications, pp. 417–440.

George, G.T., 2001. Late Yeadonian (Upper Sandstone Group) incised valleysupply and depositional systems in the South Wales peripheral forelandbasin: implications for the evolution of the Culm Basin and for the Silesianhydrocarbon plays of onshore and offshore UK. Marine and PetroleumGeology 18, 671–705.

Gilligan, A., 1920. The petrography of the Millstone Grit of Yorkshire.Quarterly Journal of the Geological Society of London 75, 251–294.

Glover, B.W., Leng, M.J., Chisholm, J.I., 1996. A second major fluvialsourceland for the Silesian Pennine Basin of northern England. Journal ofthe Geological Society (London) 153, 901–906.

Guion, P.D., 1992. Westphalian. In: Cope, J.C.W., Ingham, J.K., Rawson, P.F.(Eds.), Atlas of Palaeogeography and Lithofacies. Geological Society ofLondon Memoir, 13, pp. 79–85.

Guion, P.D., Fielding, C.R., 1988. Westphalian A and B sedimentation in thePennine Basin, UK. In: Besly, B.M., Kelling, G. (Eds.), Sedimentation in aSynorogenic Basin Complex. Blackie, Glasgow and London, pp. 153–177.

Guion, P.D., Fulton, I.M., Jones, N.S., 1995. Sedimentary facies of coal-bearingWestphalian A and B north of theWales–Brabant High. In:Whateley, M.K.G.,Spears, D.A. (Eds.), European Coal Geology, 82. Geological Society SpecialPublication, pp. 45–78.

Hallsworth, C.R., 1995. Mineralogical evidence for variations in provenance ofthe Millstone Grit and Lower Coal Measures of the Bradford district. BritishGeological Survey Technical Report WH/95/200R.

Hallsworth, C.R., 1998. Stratigraphic variations in the Millstone Grit in theMelbourne area: implications for provenance. British Geological SurveyTechnical Report WA/98/53R.

Hallsworth, C.R., 2001. Mineralogical evidence for variations in provenance ofCarboniferous sandstones from Cumbria. British Geological SurveyTechnical Report, Stratigraphy Series. WH/1001/150.

Hallsworth, C.R., Chisholm, J.I., 2000. Stratigraphic evolution of provenancecharacteristics in Westphalian sandstones of the Yorkshire coalfield.Proceedings of the Yorkshire Geological Society 53, 43–72.

Hallsworth, C.R., Morton, A.C., Claoué-Long, J., Fanning, C.M., 2000.Carboniferous sand provenance in the Pennine Basin, UK: constraints fromheavy mineral and detrital zircon age data. Sedimentary Geology 137,147–185.

Holliday, D.W., Molyneux, S.G., 2006. Editorial statement: new official namesfor the subsystems, series and stages of the Carboniferous system.Proceedings of the Yorkshire Geological Society 56, 57–58.

Lake,R.D., 1999.TheWakefield district— a concise account of the geology.Memoirof the British Geological Survey 1 (50 000 Sheet 78, England and Wales).

Leng, M.J., Glover, B.W., Chisholm, J.I., 1999. Nd and Sr isotopes as clasticprovenance indicators in the Upper Carboniferous of Britain. PetroleumGeoscience 5, 293–301.

McLean, D., Chisholm, J.I., 1996. Reworked palynomorphs as provenanceindicators in the Yeadonian of the Pennine Basin. Proceedings of theYorkshire Geological Society 51, 141–151.

Maynard, J.R., 1992. Sequence stratigraphy of the Upper Yeadonian of northernEngland. Marine and Petroleum Geology 9, 197–207.

Maynard, J.R., Hofmann,W., Dunay, R.E., Bentham, P.N., Dean, K.P.,Watson, I.,1997. The Carboniferous of western Europe: the development of a petroleumsystem. Petroleum Geoscience 3, 97–115.

Mills, D.A.C., Holliday, D.W., 1998. Geology of the district around Newcastleupon Tyne, Gateshead and Consett. Memoir of the British GeologicalSurvey Sheet 20 (England and Wales).

Mills, D.A.C., Hull, J.H., 1968. The Geological Survey borehole at Woodland,Co. Durham (1962). Bulletin of the Geological Survey of Great Britain 28,1–37.

Morton, A.C., 1985a. Heavy minerals in provenance studies. In: Zuffa, G.G.(Ed.), Provenance of Arenites. Reidel, Dordrecht, pp. 249–277.

Morton, A.C., 1985b. A new approach to provenance studies: electronmicroprobe analysis of detrital garnets from Middle Jurassic sandstones ofthe northern North Sea. Sedimentology 32, 553–566.

Morton, A.C., Hallsworth, C.R., 1994. Identifying provenance-specific featuresof detrital heavy mineral assemblages in sandstones. Sedimentary Geology90, 241–256.

Morton, A.C., Hallsworth, C.R., 1999. Processes controlling the compositionof heavy mineral assemblages in sandstones. Sedimentary Geology 124,3–29.

Morton, A.C., Hallsworth, C.R., Chalton, B., 2004. Garnet compositions inScottish and Norwegian basement terrains: a framework for interpretation ofNorth Sea sandstone provenance. Marine and Petroleum Geology 21,393–410.

Powell, J.H., Chisholm, J.I., Bridge, D.McC., Rees, J.G., Glover, B.W., Besly,B.M., 2000. Stratigraphical framework for Westphalian to Early Permianred-bed successions of the Pennine Basin. British Geological SurveyResearch Report, RR/00/01.

Press, S., 1986. Detrital spinels from alpinotype source rocks in MiddleDevonian sediments of the RhenishMassif. Geologische Rundschau, vol. 75,pp. 333–340.

Ramsbottom, W.H.C., 1974. The Namurian of North Wales. In: Owen, T.R.(Ed.), The Upper Palaeozoic and Post-Palaeozoic Rocks of Wales.University of Wales Press, Cardiff, pp. 161–167.

Rankin, D.W., Furnes, H., Bishop, A.C., Cabanis, B., Milton, D.J., O'Brien, S.J.,Thorpe, R.S., 1988. Plutonism and volcanism related to the pre-Arenigevolution of the Caledonide–Appalachian orogen. In:Harris, A.L., Fettes, D.J.(Eds.), The Caledonian–Appalachian Orogen, 38. Geological Society,London, Special Publication, pp. 149–183.

Rippon, J.H., 1996. Sand body orientation, palaeoslope analysis and basin fillimplications in the Westphalian A–C of Great Britain. Journal of theGeological Society, London 153, 881–900.

Skehan, J.W., 1988. Evolution of the Iapetus Ocean and its borders in pre-Arenig times: a synthesis. In: Harris, A.L., Fettes, D.J. (Eds.), TheCaledonian–Appalachian Orogen, 38. Geological Society, London, SpecialPublication, pp. 185–229.

Smith, E.G., Rhys, G.H., Eden, R.A., 1967. Geology of the country aroundChesterfield, Matlock and Mansfield. Memoir of the Geological Survey ofGreat Britain, Sheet 112, England and Wales.

Sorby, H.C., 1859. On the structure and origin of the millstone-grit in SouthYorkshire. Proceedings of the Geological and Polytechnic Society of theWest Riding of Yorkshire, 3, pp. 669–675.

Steele, R.P., 1988. The Namurian sedimentary history of the GainsboroughTrough. In: Besly, B.M., Kelling, G. (Eds.), Sedimentation in a synorogenicbasin complex. Blackie, Glasgow and London, pp. 102–113.

Stephens, M.B., Furnes, H., Robins, B., Sturt, B.A., 1985. Igneous activitywithin the Scandinavian Caledonides. In: Gee, D.G., Sturt, B.A. (Eds.), TheCaledonide Orogen — Scandinavia and Related Areas. Wiley, New York,pp. 623–656.

Trewin, N.H., Holdsworth, B.K., 1972. Sedimentation in the Lower NamurianRocks of the North Staffordshire Basin. Proceedings of the YorkshireGeological Society 39, 371–408.

Waters, C.N., 2000. Geology of the Bradford district. Sheet description of theBritish Geological Survey, 1:50 000 Series Sheet 69 Bradford (England andWales).

Waters, C.N., Aitkenhead, N., Jones, N.S., Chisholm, J.I., 1996. LateCarboniferous stratigraphy and sedimentology of the Bradford area, andits implications for the regional geology of northern England. Proceedings ofthe Yorkshire Geological Society 51, 87–101.