Sedimentology (2007) 54, 1163–1189
doi: 10.1111/j.1365-3091.2007.00877.x
Widespread transport of pyroclastic density currents from a large silicic tuff ring: the Glaramara tuff, Scafell caldera, English Lake District, UK R. J. BROWN* 1 , B. P. KOKELAAR and M. J. BRANNEY* *Department of Geology, University of Leicester, University Road, Leicester LE1 7RH, UK (E-mail:
[email protected]) Earth and Ocean Sciences Department, University of Liverpool, Liverpool L69 3GP, UK ABSTRACT
The Glaramara tuff presents extensive exposures of the medial and distal deposits of a large tuff ring (original area >800 km2) that grew within an alluvial to lacustrine caldera basin. Detailed analysis and correlation of 21 sections through the tuff show that the eruption involved phreatomagmatic to magmatic explosions resulting from the interaction of dacitic magma and shallow-aquifer water. As the eruption developed to peak intensity, numerous, powerful single-surge pyroclastic density currents reached beyond 8 km from the vent, probably >12 km. The currents were strongly depletive and deposited coarse lapilli (>5 cm in diameter) up to 5 km from source, with only fine ash and accretionary lapilli deposited beyond this. As the eruption intensity waned, currents deposited fine ash and accretionary lapilli across both distal and medial regions. The simple wax–wane cycle of the eruption produced an overall upward coarsening to fining sequence of the vertical lithofacies succession together with a corresponding progradational to retrogradational succession of lithofacies relative to the vent. Various downcurrent facies transitions record transformations of the depositional flow-boundary zones as the depletive currents evolved with distance, in some cases transforming from granular fluid-based to fully dilute currents primarily as a result of loss of granular fluid by deposition. The tuff-ring deposits share several characteristics with (larger) ignimbrite sheets formed during Plinian eruptions and this underscores some overall similarities between pyroclastic density currents that form tuff rings and those that deposit large-volume ignimbrites. Tuff-ring explosive activity with such a wide area of impact is not commonly recognized, but it records the possibility of such currents and this should be factored into hazard assessments. Keywords Accretionary lapilli, caldera, phreatomagmatism, pyroclastic density current, tuff ring, volcanic hazards.
INTRODUCTION Hydrovolcanic explosive eruptions in aquifer, lacustrine and shallow marine settings are common and have caused numerous fatalities; associated hazards include pyroclastic density 1
Present address: Department of Earth Sciences, Wills Memorial Building, University of Bristol, Queen’s Road, Bristol BS8 1RJ, UK.
currents, ballistic-block and tephra fallout, lahars, seiches and tsunamis (Tanguy et al., 1998; Mastin & Witter, 2000). Pyroclastic density currents generated during this type of activity typically result from transient fountaining of tephra ejected in upward-directed jets that are either single pulses or unsteadily sustained for only a short duration. In most tuff ring-forming eruptions, the bulk of the tephra falls from heights of <1 km and the resultant currents have runout distances of
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists
1163
1164
R. J. Brown et al.
between 0Æ5 and 3 km. A few such eruptions have involved currents that travelled much further, e.g. >6 km in the Rotomahana-Waimangu 1886 eruption (Nairn, 1979), 6 km at Taal Volcano (Moore et al., 1966; Moore, 1967) and 7 km at Monte Guardia volcano, Lipari (Colella & Hiscott, 1997). These runout distances indicate that unsteady pyroclastic fountaining can involve transient high mass fluxes (e.g. Bursik & Woods, 1996). Such extensive currents represent a considerable volcanic hazard. There are numerous accounts of basaltic and trachybasaltic tuff ring-forming eruptions (e.g. Waters & Fisher, 1971; Kienle et al., 1980; Yamamoto et al., 1991; Sohn, 1996), but relatively few due to silicic magma (e.g. Sheridan & Updike, 1975; Brooker et al., 1993; Colella & Hiscott, 1997). The mechanisms by which magma and water come together to interact are thought to be diverse, including static contact, direct venting into clear water, ascent through a water-saturated slurry, ascent through a lithified permeable aquifer, and ascent along a flooded fault (Kokelaar, 1982, 1983, 1986; Morrissey et al., 2000; White & Houghton, 2000). Where the supply of water is limited, as in eruption through an aquifer as opposed to through open water, explosivity may operate near maximum thermodynamic efficiency and, hence, thrust tephra to higher altitudes at greater velocity than when excess water quenches the system. Aquifer-related eruptions (‘Taalian’ eruptions; Kokelaar, 1986) are commonly more powerful than eruptions through standing water or wet slurry (Surtseyan type), as is manifest in the relatively low profiles of the associated tuff rings and wide distribution of the associated tephra (e.g. Wohletz & Sheridan, 1983). However, the runout distance of pyroclastic currents is not simply related to the height from which tephra falls to the ground. While increased potential energy converts to greater kinetic energy, this can increase mixing of the current with air which, in turn, can promote elutriation and loss of mass from the current by lofting (e.g. Bursik & Woods, 1996). Furthermore, while increased mass flux of erupted magma might be expected to relate to greater runout, in hydrovolcanoes this can cause changes in fountain height according to changes in the mixing with water and, hence, fragmentation and the efficiency of the steam explosivity. In some cases of steady supply of magma at high mass flux through wet media, a significant proportion of the fragmenting magma may not be able to interact with the water, being physically isolated from it by the outer parts of the flow where mixing and vapour
formation do occur (Kokelaar, 1983, 1986; Kokelaar & Busby, 1992; Kokelaar & Ko¨niger, 2000). In this paper, the Glaramara tuff, which recorded a powerfully explosive hydrovolcanic eruption fuelled by dacitic magma that interacted with shallow-level aquifer water near the margin of a large, partly flooded caldera, is described. Modern concepts of pyroclastic sedimentation (e.g. Branney & Kokelaar, 2002; Choux & Druitt, 2002) were applied to reconstruct the eruptive history and to understand the flow and depositional behaviour of pyroclastic density currents that apparently had considerable lateral reach. The tuff is well exposed in a glacially scoured and dissected mountainous terrain that allows partial three-dimensional (3D) reconstruction of its lithofacies architecture. There are excellent records of distal currents and associated deposition, which are rarely preserved in other tuff rings because of erosion or sedimentation into water. The eruption generated numerous, closely successive currents that swept across the ground >8 km from the vent, and a low-aspectratio tuff ring was constructed. It has been inferred that more than 0Æ2 to 0Æ5 km3 of tephra was deposited in an area of >800 km2, mainly by pyroclastic density currents. All of the ejecta in the ‘ring’, even the thin distal deposits, are here termed Glaramara tuff to avoid semantic issues such as those regarding ‘ring’ versus ‘apron’ deposits.
Geological Setting Scafell caldera volcano lies within the Lake District National Park, NW England (Fig. 1; Branney & Kokelaar, 1994; Millward et al., 2000). The rocks form part of the 6 km thick Ordovician Borrowdale Volcanic Group, which records several overlapping and subsiding volcanic fields at a weakly extensional or transtensional Caradoc continental margin, beneath which there was active subduction of oceanic lithosphere. Scafell caldera volcano is deeply incised and particularly well exposed, revealing the caldera floor and caldera fill, which includes silicic domes, caldera lake deposits and bounding caldera faults. The volcano has been intensively studied and its subsidence history is well-known (Branney, 1988a,b, 1991; Branney & Soper, 1988; Davis, 1989; Branney & Kokelaar, 1994; Millward et al., 2000; Brown, 2001; Kokelaar et al., 2007). Pre-caldera basaltic to andesitic volcanism produced low-profile shields and sill complexes within subsided depressions (Petterson et al., 1992; e.g. Lingcove Formation; Fig. 2). Subsidence of the 17 km diameter Scafell caldera was
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera
1165
Fig. 1. Simplified geological map (modified from Branney & Kokelaar, 1994) of the Glaramara tuff showing the localities of measured sections. Coordinates are from the Ordnance Survey National Grid. Inset: simplified geological map of the Lake District, with position of study area marked by box.
accompanied by voluminous andesitic to rhyolitic explosive eruptions (tens to hundreds of km3) that deposited a succession of andesitic to rhyolitic welded ignimbrites intercalated with thin layers of phreatomagmatic tuff (Fig. 2; Whorneyside and Airy’s Bridge Formations; Davis, 1989; Branney, 1991; Branney et al., 1992; Branney & Kokelaar, 1994). Following caldera collapse, flooding of the caldera was associated with deposition of more than 500 m of lacustrine pyroclastic and sedimentary rocks (Lingmell Formation; Seathwaite Fells Formation and Sprinkling Tarn Formation; Kneller & McConnell, 1993; Millward et al., 2000; Brown, 2001; Kokelaar et al., 2007). Present-day exposure shows that the basin occupied more than
100 km2, but the basin margins were complex, involving downsag, and in some areas were probably gently sloping rather than bounded by steep scarps (Kokelaar et al., 2007). Available evidence suggests that inundation was by freshwater although possible connections to the sea are not excluded. Preservation of the 6 km thickness of the entire volcanic succession (Borrowdale Volcanic Group) was accomplished largely by volcanotectonic subsidence (Branney & Soper, 1988) and, with the very high eustatic stands of sea-level prevalent during Caradoc times, variously estimated at up to 600 m above the present present (Hallam, 1984; Ross & Ross, 1992; Fortey et al., 1995), it is readily conceived that the caldera may have lain close to sea-level.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1166
R. J. Brown et al.
Fig. 2. Simplified vertical succession of the middle part of the Borrowdale Volcanic Group (modified from Branney & Kokelaar, 1994).
A predominantly non-marine setting is favoured in this study, because no marine fauna is preserved in the sedimentary rocks, no marine or
near-shore sedimentary facies have been recognized, and the exceedingly sparse trace fossils suggest freshwater (Branney, 1988b; Molyneux,
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera 1988; Johnson et al., 1994). The final infilling of the basin was by lacustrine and alluvial sediments, succeeded mainly by subaerial pyroclastic deposits (Sprinkling Tarn Formation; Fig. 2), including the Glaramara tuff (Glaramara Tuff Member), which is the focus of this paper. The succession is overlain by extensive ignimbrites from another centre located to the east (Lincomb Tarns Formation; Kneller & McConnell, 1993).
Terminology Lithofacies are defined using non-genetic terminology based upon internal sedimentary structure, grain size, sorting and composition (modified after Sohn & Chough, 1989). (These lithofacies are described and interpreted in Table 1, with an explanation of the abbreviations after Branney & Kokelaar, 2002.) Lateral lithofacies associations were identified by tracing marker horizons across the region. All the lithofacies of the Glaramara tuff are lithified (i.e. tuff and lapilli tuff; ‘ash’ in ash aggregates and ashfall indicates that ash was aggregated/deposited). Evidence for high-temperature welding (e.g. welded shards, matrix spherulites or perlite, rheomorphic deformation) is absent, and the terms ‘fiamme’ and ‘eutaxitic’ are used non-genetically in this paper; ‘fiamme’ are flattened, originally pumiceous clasts, and ‘eutaxitic’ refers to the streaky fabric that fiamme and similarly deformed smaller particles impart to the rocks. Fiamme in the Glaramara tuff are composed chiefly of chlorite and epidote and they represent former pumice lapilli altered during diagenesis to a clay framework that collapsed during burial; similar fiamme of the same origin also occur in aqueous sedimentary lithofacies in the Scafell caldera (e.g. Branney & Sparks, 1990). The Glaramara tuff preserves a range of original ash aggregate types and the term ‘accretionary lapilli’ is reserved only for those with an internal structure of concentric laminations; ash aggregates that lack an internal concentric lamination are referred to as ‘pellets’ and those that have a single rim of fine-grained ash around a massive core as ‘coated pellets’ (see below).
GLARAMARA TUFF The Glaramara tuff comprises massive, stratified and cross-bedded tuffs, lapilli tuffs and ash aggregate-bearing tuffs, all emplaced subaerially. It is named after the craggy summit of Glaramara where
1167
it is particularly well exposed (Fig. 1; Kneller & McConnell, 1993). It is dacitic in composition (Millward et al., 2000) and juvenile clasts typically have 2 to 10% plagioclase phenocrysts. Volcanic glass has been altered to cryptocrystalline quartz– feldspar. The tuff crops out over approximately 80 km2. The thickest and coarsest deposits (and hence the most proximal by inference) occur in the north, where they reach 7 m thick (at Black Wall and Coombe Head; Fig. 1) and are thought to be within 2 km of the inferred vent location (see below) at the northern margin of Scafell caldera (Fig. 1). The tuff is continuously exposed towards the south-west (between Glaramara and Scafell; Fig. 1). To the south-east, more widely spaced exposures show that the unit thins and becomes finer grained progressively southwards. The Glaramara tuff conformably overlies coarse sandstones, gravel conglomerates and finegrained breccias that record alluvial reworking of an ignimbrite plain by unconfined floods (Brown, 2001). The tuff is overlain in the north by ignimbrite and sandstone, whereas in the south it is unconformably overlain by lacustrine siltstones and sandstones. Previously, the Glaramara tuff was considered to extend much farther eastwards (Millward et al., 2000), but these eastern strata lie at stratigraphically higher levels and were deposited during later eruptions (Brown, 2001). Detailed analysis of 21 sections through the Glaramara tuff (Fig. 1-G1 to G21) has facilitated its division into three laterally persistent lithostratigraphic units (Units 1 to 3 in Fig. 3), which are described and interpreted below.
Unit 1 Unit 1 comprises fine-grained tuff containing abundant lapilli-sized ash aggregates. It is up to 120 cm thick in northern regions and thins systematically southwards to several decimetres around Scafell (Fig. 4). There are three subunits (1a–c). Subunits 1a and 1c enclose coarser-grained subunit 1b, and comprise massive and diffusebedded, accretionary lapilli-bearing tuff (mTa, dbTa; Table 1) with subordinate stratified layers (sT; xsT). Subunits 1a and 1b each reach thicknesses of 60 cm in the north and thin to <10 cm southwards. Low-angle cross-stratification is most common in subunit 1a (Fig. 5A), and small stoss-erosional bedforms occur in northern outcrops (e.g. log G2; Fig. 4). Internal stratigraphies of subunits 1a and 1c are complicated by nonuniform lateral thickness variations and scour
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1168
R. J. Brown et al.
Table 1. Summarized description and interpretation of lithofacies in the Glaramara Tuff. Lithofacies
Description
Interpretation
Massive lapilli tuff (mLT); eutaxitic (emLT); lithic-rich (lmLT)
Lithology: angular to sub-angular lithic lapilli (<2–20%), recess-weathered fiamme lapilli (<40 mm long; aspect ratios of 1:4–1:20) and rare lithic blocks (40 cm) matrix-supported in poorly sorted altered vitric-crystal tuff matrix; lithic clasts (<40 cm diameter, typically 5–15 cm) comprise dense dacite and rare sub-rounded sandstone intraclasts Structure: massive to diffuse stratified (10–30 cm); non-graded; stratification plane-parallel or discontinuous defined by fiamme or lithic concentrations; inverse-graded fiamme (if) in emLT. Geometry: dm-thick, metres-wide lenses; tabular bodies
Massive nature, poor sorting and absence of tractional structures indicate rapid progressive aggradation from a high-concentration fluid escape-dominated depositional flow-boundary of a PDC; clast population comprises primary juvenile material; diffuse bedding results from local current unsteadiness, non-uniformity and the development of granular-flow dominated flow-boundaries; poorly vesicular dacite clasts represent juvenile material or explosively disrupted lava
Diffuse-bedded lapilli tuff (dbLT); lithic-rich (ldbLT)
Lithology: similar to massive lapilli tuff (mLT) Structure: discontinuous, sub-parallel diffuse bedding defined by cm-thick fiamme and lithic horizons, by laterally impersistent planar scour surfaces or by thin (<1 cm) fine-grained layers Geometry: individual beds persistent over 2–10 m before dying out or bifurcating. These lithofacies pass laterally and vertically into mLT, lmLT, bLT, ldbLT and sLT; decimetre-thick packages interbedded with mLT and sLT
General interpretation same as for massive eutaxitic lapilli tuff (emLT); the diffuse bedding results from current unsteadiness, perhaps in granular flow-dominated flow boundaries, while the discontinuous nature and lateral thickness variations reflect current non-uniformity; low-angle truncations and basal scours indicate brief periods of erosion and impingement of turbulent eddies on substrate
Bedded lapilli tuff (bLT)
Lithology: similar to massive lapilli tuff, but lacks coarse lithic lapilli Structure: well bedded, internally massive lapilli tuff; beds typically <20 cm thick with parallel to sub-parallel boundaries; discontinuous, low-angle cross-stratification overlies low-relief basal scour surfaces; upper few centimetres is normally graded while fiamme are inverse graded and concentrated in upper parts of beds; scour surfaces persistent over decimetres to metres Geometry: occurs as discrete tabular beds, and in stacked sets up to 1 m thick; passes vertically and laterally into dbLT and mLT
General interpretation as for emLT (see above); normal-grading and inverse-graded fiamme record waning of the current with time; stratification records local tractionsedimentation, while scours indicate unsteadiness or impingement of turbulent eddies on the substrate; resembles many modern pumiceous ignimbrites with inverse-graded pumice lapilli (e.g. Sparks et al., 1973); similar lithofacies in lower units of the Borrowdale Volcanic Group interpreted as ignimbrites (see Branney, 1988a,b; Davis, 1989).
Stratified lapilli tuff (sLT)
Lithology: alternating eutaxitic poorly sorted fines-rich layers (with matrixsupported fiamme lapilli, <1 cm) and better-sorted non-eutaxitic crystal-lithic layers (matrix- to clast-supported angular lithic lapilli and plagioclase crystals) Structure: parallel to sub-parallel strata <1–3 cm thick in sets 5 cm to <1 m thick; very low-angle truncations are common; individual strata are generally persistent for several metres Geometry: passes laterally and vertically into dbLT, bLT and mLT
Well-developed stratification and well-sorted layers indicate tractional sedimentation from the low-concentration flow boundary zone of a PDC, allowing efficient elutriation and winnowing of lowdensity pumice and fine ash; poorly sorted eutaxitic layers may record individual pulses or surges of the current (e.g. Sohn & Chough, 1989).
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera
1169
Table 1. (Continued) Lithofacies
Description
Interpretation
Cross-stratified lapilli tuff (xsLT)
Lithology: similar to stratified lapilli tuff (sLT) Structure: low-angle cross-stratification; millimetre–centimetre thick strata similar to those described in sLT; strata comprise planar or weakly asymptotic foreset laminae that dip 10 and are truncated on their stoss-sides; stoss-side laminae rarely preserved; low-angle truncations are common; bedform amplitude 10–20 cm, with crests truncated by undulatory erosion surfaces persistent over metres Geometry: centimetre- to decimetre-thick packages that pass vertically into sLT and dbLT
Deposited from traction-dominated flow-boundaries of fully dilute PDCs; low-angle truncations record erosion due to current unsteadiness, turbulent eddies or the passage of successive currents; stoss-erosional bedforms similar to those described by Fisher & Waters (1970), Schmincke et al. (1973); Cole (1991) and Allen (1984); upcurrent-dipping strata resemble chute-and-pool structures described by Schmincke et al. (1973), Allen (1984) and Cole (1991) and record the plastering of moist ash against topographic irregularities (Waters & Fisher, 1971)
Massive tuff (mT); massive ash aggregate tuff (mTa)
Lithology: Scattered feldspar chips in an irresolvable quartz–feldspar granular matrix; rare lithic and fiamme lapilli; mTa contains matrix to clast-supported ash aggregates Structure: internally massive units Geometry: tabular and lenticular beds (2–15 cm thick); drapes bedforms in lithofacies xsLT; passes vertically into bTa, dbTa and sT
Massive nature and absence of tractional structures indicates deposition from fluid escape-dominated flow boundaries of PDCs
Deposited from traction-dominated Low-angle cross-stratified tuff (xsT) Lithology: as sLT flow-boundaries of fully dilute PDCs stratified tuff (sT) Structure: low-angle cross-stratification (xsT) or planar stratification (sT); individual laminae discontinuous over 10–50 cm; pinch-and-swell structures and low-angle truncations common; ripple cross-lamination rare (wave-length of 9 cm, amplitude of 8 mm) Geometry: packages 2–15 cm thick; interstratified with mTa, sLT and dbLT Diffuse-bedded ash aggregate lapilli Lithology: as mTa; ash aggregates vary tuff (dbTa) from small (2–5 mm diameter) ash pellets to large (>10 mm) coated pellets and accretionary lapilli (see text); rim fragments common; rare fiamme Structure: diffuse bedding (2–15 cm thick) defined by ash aggregate abundance; pinch-and-swell structures common; laterally variable proportions of ash aggregates, fragments and matrix; beds variably persistent (from several metres to >7 km) Geometry: occurs as sets several dm thick; passes laterally and vertically into sLT and mT
Abundant accretionary lapilli indicate the presence of moisture in ash-rich clouds; pinch-and-swell structure indicates deposition from ash-rich PDCs; laterally persistent ash aggregates horizons may reflect fallout from umbrella clouds or co-ignimbrite ash clouds; fragmented aggregates record impact-fracturing of lithified ash aggregates during transport or on impact
Scheme follows that devised by Branney & Kokelaar (2002). Composition: T, tuff; LT, lapilli tuff; l, lithic-rich; a, ash aggregate-bearing; f, fiamme. Structure: m, massive; db, diffuse-bedded; b, bedded; s, stratified; xs, cross-stratified; i, inverse-graded; e, eutaxitic.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
B
Fig. 3. Measured sections through the Glaramara tuff. (A) Summary graphic logs through the Glaramara tuff illustrating the lateral thickness variations of the units. The tuff-ring exhibits a gross thinning and fining from north to south. Local palaeotopographic highs account for the thinner sequences around Coombe Door (logs G2–G4). Downcurrent facies transitions (DFT 1–4; see Fig. 10) are marked by shaded areas. These transitions are inferred to extend southwards from Coombe Head (log G5), the most proximal outcrop. For lithofacies codes, descriptions and interpretations see text and Table 1. (B) Fence diagram illustrating the distribution in space of the Glaramara tuff and the gross thinning towards the south.
A
1170 R. J. Brown et al.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
Fig. 4. Unit 1 of the Glaramara tuff. Detailed graphic logs that illustrate the continuity of subunits and the general impersistence of individual ash aggregatebearing layers. Subunit 1b is picked out in grey as a marker horizon across the study area. Bed thicknesses and number decreases southwards in subunits 1a and 1b.
The Glaramara tuff, Scafell caldera 1171
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1172
R. J. Brown et al.
A
B
C
Fig. 5. Lithofacies of the Glaramara tuff ring. (A) Ash aggregate-tuff in subunit 1a at Black Wall (see Fig. 1, locality G1): stratified tuff (sT) with discontinuous strata and very low-angle truncations. Possible impacted coated pellet (arrowed). Lens cap is 58 mm in diameter. (B) Matrix-supported coated pellets with abundant debris (comminuted coated pellet rims). (C) Massive eutaxitic lapilli tuff (emLT) of subunit 1b. Note extremely poor sorting illustrated by fine matrix with accretionary lapilli (a), large lithic (l) and fiamme (f) lapilli (at Glaramara, see Fig. 1, locality G7). Coin is 2 cm in diameter.
surfaces; many layers are impersistent over metres to tens of metres, and the constituent lithofacies succeed each other with no preferred vertical sequence. Lapilli-sized ash aggregates are common in subunits 1a and 1c. Three original types are
recognized: (i) ‘Ash pellets’, comprising a structureless ash core, sometimes with a very thin (<0Æ5 mm) coating of fine ash; these are generally <6 mm in diameter and are only vaguely discernible in the outcrop. (ii) Larger (6 to 15 mm diameter) distinctly ‘coated’ pellets, each with a distinct layer, 0Æ5 to 1Æ5 mm thick, of fine-grained ash enclosing a central structureless ash core (Fig. 5A and B). (iii) Accretionary lapilli in which several concentric laminations of ash surround a central structureless ash core; the laminations are marked by abrupt and graded grain-size variations. Of these types, the ash pellets are the least abundant and occur either on their own or together with coated pellets, in matrix-supported and clast-supported layers. The coated pellets are the most common type of aggregates and form matrix- and clast-supported layers, 1 to 20 cm thick (Fig. 4). The matrix in these layers comprises fine tuff with abundant angular fragments of coated pellet rims (Fig. 5B). Accretionary lapilli typically occur sparsely within layers of coated pellets. At the base of subunit 1a, small-impact sags are present beneath some ash aggregates (e.g. log G16 in Fig. 4). Some of the layers with ash aggregates also contain fiamme (e.g. log G5 in Fig. 4). Subunits 1a and 1c thin towards the south, and the number of individual ash aggregate layers within them also decrease southwards (Fig. 4). The type and abundance of ash aggregates vary laterally within individual layers (Fig. 4). Some of the ash aggregate-bearing layers cannot be traced between exposures, and appear to have the form of broad lenses that are discontinuous over tens of metres. An exception is the uppermost layer in subunit 1c, which maintains its character laterally for 9 km (logs G15–G21 in Fig. 4). Subunit 1b is <20 cm thick and conformably overlies subunit 1a and it is present across the full extent of the Glaramara tuff outcrop (Fig. 4). At the base of subunit 1b, a thin (<2 cm), clast-supported crystal-lithic tuff layer (sT) exhibits slight local pinch-and-swell bedding, and no marked lateral grading accompanies its southward thinning. It is overlain by £18 cm of very poorly sorted, massive to diffuse-bedded eutaxitic lapilli tuff (emLT, edbLT), with matrix-supported to clast-supported lapilli in a fine-tuff matrix. The lapilli are angular to sub-rounded and irregularly shaped lithic clasts <10 cm in diameter, and recess-weathered fiamme (Fig. 5C). The lithic lapilli are mostly of aphyric and porphyritic dacite, and rare welded lapilli tuffs. The fiamme are sparsely porphyritic (2 to 10% plagioclase phenocrysts), £18 cm long, with
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera
Fig. 6. Maximum lithic and fiamme data for the Glaramara tuff. Note general north-to-south fining trend; outsized lithic blocks in subunit 2a at Coombe Head may have been ballistically emplaced. Inset on left indicates palaeocurrent direction of pyroclastic density currents as given by foreset laminae dips in dunes in subunit 2b. Data indicate eruption to the north of the study area.
aspect ratios of 1:3 to 1:8. In places the fiamme exhibit a crude inverse-grading (emLTif; e.g. log G11 in Fig. 4). The lithic clasts and fiamme decrease in size distally (southwards; Fig. 6). Local diffuse bedding is defined by discontinuous finegrained tuff layers, which have gradational to sharp and irregular boundaries and contain matrix-supported fiamme and sparse accretionary lapilli (e.g. log G2 in Fig. 4). In some localities (e.g. log G16 in Fig. 4), the matrix of the massive eutaxitic lapilli tuff is normally graded, and passes upwards into fine tuff with coated pellets (mTa). A second thin layer of well-sorted, crystal lithic lapilli tuff (sT), similar to that at the base, is the uppermost layer of subunit 1b.
Interpretation Subunits 1a and 1c are interpreted as the products of highly unsteady, non-uniform, ash-rich pyroclastic density currents. The pinch-and-swell bedding and impersistence of numerous beds
1173
indicate that the currents were non-uniform, and the complex vertical stratigraphy, with internal scour surfaces (Fig. 4), indicates that they were highly unsteady. The occurrence of both massive tuff beds and cross-stratified layers records deposition from currents whose flow boundaries varied between granular fluid-based and fully dilute (e.g. Branney & Kokelaar, 2002). The poorly sorted massive deposits with sparse coated pellets supported in a massive tuff matrix probably were deposited from a fluid escape-dominated flowboundary zone in which winnowing and tractional transport were limited by high concentrations of clasts, which dampened turbulence. The presence of ash aggregates indicates that moisture was present in the ash-laden eruption clouds (Gilbert & Lane, 1994). Bed impersistence, thickness changes and lateral variations in abundance of the ash aggregates indicate that the layers with ash aggregates are not simple fallout deposits, but were deposited from non-uniform pyroclastic density currents. The vertical changes between ash-aggregate, massive and cross-stratified layers indicate that the currents underwent rapid temporal changes, for example, in concentration and velocity. This study infers that ash pellets formed in dilute buoyant plumes rising above ground-hugging currents and in plumes above the eruption jets (see Self & Sparks, 1978; Schumacher & Schmincke, 1991; Schumacher & Schmincke, 1995). When the pellets could no longer be supported by turbulence within these plumes they settled into the density currents, where they acquired their coatings of fine ash (see Brown & Branney, 2004). The thicknesses of the fine-ash coatings presumably reflect the ashaccretion rate and the duration of transport within the ground-hugging currents. The accretionary lapilli are thought to have originated in the same way as the coated pellets, but with more protracted transport in one or more closely successive pyroclastic density currents during which the successive concentric laminations accreted. The abundant angular fragments of ash aggregates (Figs 4 and 5B) indicate very early lithification of the aggregates (e.g. De’Gennaro et al., 2000) within hot (>100C) currents, and that high-velocity impacts occurred between aggregates within turbulent parts of the currents (e.g. Brown & Branney, 2004). Few unambiguous ashfall layers containing ash aggregates have been recognized in subunits 1a and 1c (see Fig. 4), and it is inferred that most of the ash aggregates that formed fell into the pyroclastic density currents so that no pelletal fallout layers were formed.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1174
R. J. Brown et al.
This inference suggests that current activity, albeit highly unsteady, was more or less continuous and extended to distal locations. It seems less likely that protracted pauses occurred between successive currents and that ashfall layers formed in such pauses but were then repeatedly and completely eroded away. Diagenetic collapse of pumice clasts (to form fiamme, see Fig. 5C) has masked some primary deposit characteristics of subunit 1b. The lower and upper layers of well-sorted, crystal-lithic tuff within subunit 1b are interpreted as recording pyroclastic density currents with traction-dominated flow-boundary zones involving winnowing of fine ash; their thickness variations and local stratification (log G7 in Fig. 4) are not consistent with an ashfall origin. The somewhat enigmatic layer of poorly sorted massive eutaxitic lapilli tuff (emLT; subunit 1b; Fig. 4) is probably the deposit of a granular fluid-based density current. The large size of the fiamme indicates that the current had contained pumice lapilli and blocks, probably generated either during a period of magmatic explosivity when access of water to the eruptive conduit was temporarily restricted or when the eruptive mass flux was sufficiently high to prevent interior parts of the jet from physically interacting with the water. The fine-grained tuff horizons and the scattered accretionary lapilli, however, suggest that phreatomagmatic activity continued intermittently. The systematic southward decrease in maximum diameters of both lithic clasts and fiamme (Fig. 6) indicates that the current competence (particle transport capability) was depletive (Kneller & Branney, 1995), and that clast density was not a major factor in segregation. This result is in contrast to many granular fluid-based depletive currents in which pumices commonly segregate from lithic lapilli according to their lower density, and overpass to more distal locations (selective filtering, see Branney & Kokelaar, 2002; see also Choux & Druitt, 2002). This study concludes that pumice buoyancy and granular interactions played relatively minor roles in clast segregation.
Unit 2 Unit 2 conformably overlies subunit 1c; it consists of stratified, bedded and massive lapilli tuffs, interbedded with thin layers of coated pellets (mTa). This unit is up to 7 m thick in the north (Fig. 1) and thins to <30 cm southwards across 5 km (Fig. 3). It also thins abruptly to 1Æ2 m south-eastwards across a volcanotectonic fault
(at Coombe Door; Fig. 3) and it is divided into two subunits: 2a and 2b. Subunit 2a reaches 6 m thick (log G5 in Fig. 7) and comprises predominantly lapilli tuffs (Fig. 8A–C), variously massive (mLT), stratified (sLT), diffuse-bedded (dbLT) and bedded (bLT). In proximal, northern areas, the lower 60 cm comprises stratified lapilli tuff (sLT) with thin beds of massive eutaxitic lapilli tuff (emLT), which die out southwards (Fig. 7). These tuff beds are overlain by up to 3 m of decimetrethick, bedded (bLT) and massive lapilli tuffs (mLT, emLT) interbedded with subordinate stratified tuffs (sLT) in beds 3 to 15 cm thick. Eutaxitic layers, with fiamme up to several centimetres long, occur locally at the top of some of the massive beds (Fig. 7). These layers are overlain by massive to diffuse-bedded lithicrich lapilli tuffs (mlLT and dblLT), <1 m thick, comprising matrix-supported, non-vesicular angular lapilli and blocks of dacite, £40 cm in diameter (Fig. 8A and B). These lapilli tuffs exhibit rapid lateral variations in clast content, maximum clast size and bedding characteristics. The layer becomes thinner and finer grained towards the south, and the more proximal, massive to diffuse-bedded deposits pass distally into cross-stratified and dune-bedded facies (log G11; Fig. 7). Thin impersistent layers of tuff occur above and below the lithic-rich layer, and locally within it (e.g. log G8; Fig. 7); they contain scattered, small (<4 mm) pellets and are commonly cut by erosion surfaces. The lithic-rich layer grades upwards into <1Æ5 m of bedded and stratified tuffs and lapilli tuffs that resemble those at the base of the subunit. In places, layers of coarse fiamme and extensive thin layers (<3 to 80 mm) of coated pellets (mTa) occur interstratified within subunit 2a (Fig. 7) and are persistent across several kilometres (Figs 3 and 7). Two downcurrent facies transitions (DFT 1 and DFT 2) are recognized in this subunit (see below). Subunit 2b is finer grained than subunit 2a and comprises centimetre to decimetre-thick packages of stratified, cross-stratified and dune-bedded tuffs and lapilli tuffs (sT, xsT and xsLT), interbedded with thin beds of accretionary lapilli (mTa; Figs 8D and 9); it thins from 120 cm (at Glaramara) to <20 cm southwards (at Scafell; Fig. 8). Stoss-erosional dune forms with amplitudes of several decimetres are common (e.g. Fig. 8E and F). Sparse palaeocurrent data (12 measurements) from foreset laminae indicate transport towards the south (Fig. 6). Thin layers of both matrix-supported and clast-supported
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera
1175
Fig. 7. Selected graphic logs through subunit 2a showing downcurrent facies transitions and thickness changes. Various lithofacies are present including a distinctive lithic-rich horizon. Decrease in bed numbers and bed thicknesses towards the south is consistent with eruption north of the study area. See Fig. 1 for location of logged sections.
coated pellets (mTa) drape the dune forms (Fig. 8E and F) and thicken slightly in dune troughs (e.g. at Glaramara; Fig. 1). Towards the south, subunit 2b comprises <30 cm of diffusebedded pellet-rich tuff interbedded with stratified tuff layers a few centimetres thick (see logs G15 to 21; Fig. 9). The pellets have thick coatings of fine ash, similar to those in subunits 1a and 1c (see
above). Angular fragments of these coated pellets occur within the matrix (Fig. 9). In proximal, northern areas, the top of subunit 2b is defined as the uppermost occurrence of stratified or crossstratified tuffs and lapilli tuffs; this contact is difficult to trace southwards because of lateral lithofacies changes and the discontinuous nature of individual accretionary lapilli layers (Fig. 9).
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1176
R. J. Brown et al. the south, measuring 70 cm thick at Scafell. The southward fining is most evident in the lithic-rich horizon, which is a useful marker. Bedding
Lateral variations within Unit 2 are marked and four downcurrent facies transitions recognized (Fig. 10). Unit 2 generally thins and fines towards A
B
C
D
E
F D
G
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera
1177
Fig. 9. Selected graphic logs through subunit 2b showing continuity of thin coated pellet layers. Subunit thins toward the south. The subunit is characterized by stratified and cross-stratified lithofacies. Dips of foresets in dunebedded layers give a consistent southerly dispersal. See Fig. 1 for location of logged sections.
becomes more uniform towards the south and the lithological variation diminishes; southern exposures comprise mainly decimetre-thick bedded (bLT) and diffuse-bedded lapilli tuff (dbLT) in which bedding is marked by diffuse horizons of centimetre-scale and sub-centimetre-scale fiamme (Figs 7 and 8G). In the north, Unit 2 thins dramatically from 6 m to <2Æ4 m over a distance of 200 m (logs G7–G4;
Fig. 3). This occurs across a volcano-tectonic fault known to have been active before the Glaramara eruption (Kneller & McConnell, 1993).
Interpretation The lithofacies and their associations in Unit 2 (Fig. 10) record the dispersal of tephra by numerous pyroclastic density currents generated during powerful phreatomagmatic explosive activity.
Fig. 8. Lithofacies in Unit 2. (A, B) Massive lithic-rich lapilli tuff and diffuse-bedded lapilli tuff (lithofacies lmLT and dbLT) in subunit 2a at Coombe Head (see Fig. 1, locality G5). Lithic blocks are up to 40 cm in diameter and do not sit in impact sags, indicating lateral emplacement prior to deposition. Hammer is 40 cm long. (C) Diffuse bedded lithic rich lapilli tuff in subunit 2a at Glaramara (Fig. 1, locality G7). Lithic lapilli are weakly concentrated into sub-parallel discontinuous layers and lenses alternating with finer eutaxitic tuff. 13 cm exposed on rule. (D) Thin beds of massive ash aggregate-bearing tuff (mTa) interstratified with stratified and cross-stratified tuff (sT; xsT) in subunit 2b at Glaramara. Cross-stratified tuff defines small ripples with upcurrent-dipping strata (arrows). Bed in centre of photograph contains normal-graded coated pellets. Note small impact sags at base of upper coated pellet bed. Current from left to right. Coin is 2 cm in diameter. (E, F) Photograph and sketch of a stoss-erosional dune form in subunit 2b at Coombe Door (see Fig. 1, locality G3). Foresets build up from sub-horizontal strata and indicate dispersal to the south (current left to right). Metre-rule for scale. (G) Bedded lapilli tuff (bLT) at Scafell (see Fig. 1, locality G16). Beds overlie planar scour surfaces. Arrows point to the bases of diffuse-bedded fiamme horizons. Coin is 2 cm in diameter. 2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1178
R. J. Brown et al.
Fig. 10. Downcurrent facies transitions (DFT) in Unit 2 of the Glaramara tuff. Note that transitions represent only partial downcurrent sections as extreme proximal and distal parts of the tuff do not presently outcrop.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera Such deposits are typical of ‘Taalian’ explosivity, wherein magma and water interact within an aquifer and/or along flooded faults (see Kokelaar, 1986), and numerous closely spaced phreatomagmatic explosions generate short-lived, highly energetic single-surge type pyroclastic density currents that wax and wane rapidly (e.g. Sohn, 1996). Typically, the currents involve abundant fine-grained ash and form extensive dune fields proximally (e.g. the 1965 eruption at Taal; Moore, 1967). However, unlike the products at Taal, Unit 2 includes clear evidence (the fiamme in lithofacies edbLT and emLT) that the currents carried abundant pumice lapilli. Hence, a proportion of the erupting vesiculating magma was not fragmented by external water. The intercalated layers of fine tuff containing abundant ash aggregates (mTa; Figs 3, 7 and 9) record deposition predominantly by fallout from lofted plumes during quiescent intervals between successive currents, or from weak fully dilute pyroclastic density currents with low shear rates (fallout-dominated flow-boundary zone of Branney & Kokelaar, 2002), possibly the residual wakes of the highvelocity currents. The discontinuity of these layers mainly resulted from erosion by succeeding high-velocity currents. Similar successions containing intercalated thin fine-ash layers occur at many other hydrovolcanoes (e.g. Crowe & Fisher, 1973; Schmincke et al., 1973; Walker, 1984; Sohn & Chough, 1989; Chough & Sohn, 1990; Mastrolorenzo, 1994). The wide variety of sedimentary structures (massive beds, stratification, diffuse-bedding and dune forms; e.g. Fig. 8) reflects significant variations in the velocity and physical constitution of successive pyroclastic density currents, with concomitant variations in flow-boundary zone processes during deposition. Ultimately, the vertical lithofacies successions record changes in the dynamics of the eruption through time. The dominance of massive and diffuse-bedded lapilli tuffs in subunit 2a (Fig. 7) in proximal regions indicates deposition from flow-boundary zones in which particle concentrations were high enough to favour clast interactions (granular fluid-based pyroclastic density currents of Branney & Kokelaar, 2002). It is less easy to make inferences about the nature of the currents at higher levels. However, the rapid alternation of massive and stratified facies in the lower part of subunit 2a (Fig. 7) suggests that the flow-boundary zones rapidly evolved to tractional and this would be most readily achieved if the currents that deposited the massive beds had only thin
1179
granular-fluid parts (rather than being entirely composed of granular fluid). In this case, switches to a fully dilute state would involve relatively little change in just the lowermost part of the current, rather than a wholesale current transformation from dense to dilute. Horizons where fiamme are concentrated, commonly at the tops of massive beds (e.g. Fig. 8A), probably record waning flow conditions when pumice clasts were finally deposited at locations where, beforehand, similar pumices had overpassed as the result of their relative buoyancy and limited percolation tendency within the depositing granular fluid. This mechanism of producing upper pumice concentrations during waning flow conditions has been described in detail elsewhere (Branney & Kokelaar, 2002; Choux & Druitt, 2002; CarrascoNun˜ez & Branney, 2005). The cross-stratified tuff and dune bedforms (Figs 8E,F and 9) indicate deposition from fully dilute pyroclastic density currents, in which turbulence extended to the base of current and the lower flow boundary zone was dominated by tractional transport. The absence of intercalated fine-grained tuff layers or ash aggregate-bearing layers within the lower part of subunit 2a (Fig. 7) suggests that deposition occurred from a quasi-sustained (albeit unsteady) current, or from a series of currents that followed with practically no time for the settling of ash suspended in the atmosphere. In contrast, the intercalated accretionary lapillibearing layers within the upper part of subunit 2a (Fig. 7) indicate pauses between successive currents during this phase of the eruption. The observed thickening by 1 to 1Æ5 cm of the finegrained ash and pelletal layers in some dune troughs may have resulted from deposition by weak residual pyroclastic density currents or reworking by wind, or rolling of aggregates from dune crests into troughs. The lithic-rich horizon in the middle of subunit 2a (Fig. 4) records the climax of the eruption, during which powerful explosions generated widespread pyroclastic density currents that had the competence to transport coarse lithic clasts, presumably from the aquifer or vent walls above, to greater distances from source. The rapid vertical and downcurrent variations in bedding and grain size indicate that the currents were unsteady and non-uniform. The cross-stratification and dune bedforms in subunit 2b record weaker fully dilute currents in which the lower flow-boundary zones were dominated by traction. The marked thickness variations of Unit 2 across the Coombe Head volcanotectonic fault
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1180
R. J. Brown et al.
(Fig. 3) indicate that a west-facing fault scarp existed at the time of the eruption. The absence of pre-Glaramara sheet-flood deposits east of the fault and their presence west of it, suggest an elevated region that probably stood several metres above the surrounding reworked ignimbrite plain. The lack of marked thickness variations in Unit 1 tuffs (Fig. 3) suggests that the earlier ash-rich pyroclastic density currents were less affected by topography than those that deposited Unit 2, or that the fault scarp was active and increased during the Glaramara eruption.
Unit 3 Unit 3 conformably overlies Unit 2 and is lithologically similar to subunits 1a and 1c (Fig. 3); it comprises a sequence of centimetre- to decimetrethick beds of accretionary lapilli-bearing tuff (mTa and dbTa) intercalated with stratified tuffs (sT and xsT). This unit thins south-west from 150 cm thick in the north (log G2; Fig. 11), thickening slightly to the south-east (towards Side Pike; Figs 1 and 11). In the north it overlies dune bedded lapilli tuffs of unit 2b, but the contact is more difficult to define in the south because of similarities between the distal deposits of unit 2b and those of unit 3 in this area (Fig. 10). There is no visible lateral grain-size grading. Rain-splash micro-bedding (e.g. Walker, 1981; Self, 1983), very fine-grained porcellaneous tuff, and load-and-flame structures occur in the upper parts of the unit at Glaramara (Fig. 1). The original top surface is absent, having been eroded, and the unit is unconformably overlain by later deposits (Brown, 2001). The types of ash aggregates in Unit 3 differ from north to south across the study area. In the north (logs G2 to G10 in Fig. 11), 90% of the beds containing ash aggregates are dominated by simple pellets, whereas in the south (logs G18, G20 and G21; Fig. 11) <20% of the beds are dominated by simple pellets and most beds are dominated by coated pellets. Accretionary lapilli with concentric laminations are rare throughout. Angular fragments of ash-aggregate pellet coatings are common in the matrix of the beds that contain the coated pellets (Fig. 8; see also subunits 1a and 1c). Despite the excellent continuity of exposure, correlation of ash-aggregate layers between logged sections is difficult: sharp bedding surfaces become diffuse or fade out laterally, and layers bifurcate, thicken or thin inconsistently (see Fig. 11). However, some of the ash aggregatebearing layers are sufficiently distinctive to be
traced for several kilometres, allowing the unit to be divided into packages: (i) a distinctive layer of clast-supported thick-rimmed coated pellets that persist distally more than 9 km with little thickness variation (logs G15 to G21 in Fig. 11); (ii) a 20 cm thick layer of stratified lapilli tuff (sLT) that thins to 1 cm across 5 km (logs G2 to G15 in Fig. 11); and (iii) a 20 cm thick package of crossstratified tuff (xsT) that traces 7 km westwards and locally includes two thin accretionary lapillibearing layers (logs G16 to G21; Fig. 11).
Interpretation The abundant fine-grained tuff with ash aggregates and the association with Unit 2, suggest that Unit 3 records a continuation of phreatomagmatic explosivity. Interpretations are similar to those for subunits 1a and 1c (Fig. 4). The cross-stratification, diffuse-bedding, non-uniform lateral thickness variations and impersistence of most layers (Fig. 11) indicate deposition from pyroclastic density currents that were both unsteady and depletive. The flow-boundary zones of the currents varied from relatively high-concentration types characterized by fluid-escape and granular interactions, to lower-concentration, traction-dominated types. Downcurrent thinning is consistent with depletive currents in which proximal deposition was more rapid, and possibly more sustained, than distal deposition. The close similarity of Unit 3 to the distal parts of subunit 2b (Fig. 9), suggests that Unit 3 deposits may originally have passed sourcewards into thicker beds of coarser lapilli tuff. The secondary thickening of Unit 3 in south-west regions (Fig. 11) may reflect the effect of local topography on passing currents. The general downcurrent increase in the thickness of ash coatings on the pellets, from dominance of small, simple pellets in proximal regions to larger thickly coated pellets in distal regions (>7 to 9 km from source; see below) suggests that thickness of ash coatings reflects transport duration: pellets deposited soon after they formed have little coating, whereas those that endured more prolonged transport downcurrent accreted thicker rims. This suggestion is consistent with our earlier inference that the rims accreted within the currents. Rain-impact micro-bedding indicates the local saturation of ash, possibly due to rain-flushing of ash clouds and deposition of ‘mud rain’ (e.g. Walker, 1981). This interpretation is consistent with the occurrences of load-andflame structures, which indicate deformation of wet, partly liquefied ash during the final stages of the Glaramara eruption.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
Fig. 11. Detailed logs through Unit 3, showing variable continuity of individual ash aggregate-bearing layers. Bed thickness and number decrease southwards. See text for lithofacies descriptions and interpretations. Inset shows location of logged sections.
The Glaramara tuff, Scafell caldera 1181
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
Fig. 12. Cartoon illustrating the down-current changes inferred to have occurred during the transport of single-surge, depletive pyroclastic density currents of the kind that deposited DFT 2 (Fig. 10). Repetitive phreatomagmatic blasts at source generate closely successive depletive density currents that travelled out from the vent, decelerated and lost capacity and competence with distance. Rapid deposition of coarse bedload in proximal regions and progressive ingestion of air resulted in downcurrent changes from fluid escape-dominated flow-boundaries near source to traction-dominated flow-boundaries at distance. Ash aggregates are inferred to have formed in ash plumes lofted above the successive currents; fallout through successive density currents resulted in the accretion of rims around the aggregates. Flow boundary illustrations modified from Branney & Kokelaar (2002).
1182 R. J. Brown et al.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera DISCUSSION
Downcurrent facies transitions The downcurrent lithofacies transitions shown by the Glaramara tuff [see downcurrent facies transitions (DFT) 1 to 4 in Fig. 10] are all accompanied by southward thinning and fining trends and are interpreted as proximal to distal transitions that are generally coincident with the palaeocurrent direction (see Fig. 6). The thickest and coarsest deposits, at Coombe Head (Fig. 1), are the most proximal. Each lithofacies transition records deposition from a single-surge, depletive current that travelled across mainly even ground and progressively lost capacity and competence as it decelerated with distance from source (Fig. 12). Similar inferences of small pyroclastic density currents are derived from tuff rings elsewhere (Sohn & Chough, 1989; Chough & Sohn, 1990; Lajoie et al., 1992; Colella & Hiscott, 1997; Vazquez & Ort, 2006). The proximal occurrence of numerous decimetre-thick beds of massive and diffuse-bedded lapilli tuff (DFT 1–3, Fig. 10) indicates that the pyroclastic density currents were initially characterized by rapid deposition from granular fluid-base currents with fluid escape-dominated and granular flow-dominated flow boundaries (Fig. 12); this is similar to the deposition of ignimbrite. Most currents, particularly radial ones, are strongly depletive where there is little topography or channelling (Kneller & Branney, 1995). Many of the coarse lithic blocks were probably transported predominantly as bed load, incompletely coupled with turbulent particulate fluid, which provided only partial support, so that they deposited rapidly. However, some of these blocks may have become more fully entrained, so that they travelled some distance within the granular fluid at low levels in the current (see Choux & Druitt, 2002). In contrast, the size and buoyancy of the large pumice clasts would have inhibited them from depositing through high-concentration flowboundary zones, in which case they would have become segregated and transported within a zone of neutral buoyancy (‘overpassing’; Branney & Kokelaar, 2002; Choux & Druitt, 2002). The scours and erosion surfaces within the massive and diffuse-bedded deposits indicate the periodic impingement of strong eddies onto the substrate. The downstream transitions from massive and diffuse-bedded lithofacies into stratified and cross-stratified lithofacies (DFT 2; Figs 7, 9 and 10) record proximal granular fluid-based currents
1183
with fluid-escape and granular flow-dominated deposition that transformed with runout into fully dilute currents with traction-dominated flow-boundary zones. Single-surge depletive currents are initiated with a finite amount of tephra that, with deposition and elutriation into a buoyant cloud during transport, progressively diminishes. As the current travels along the ground, the lower granular fluid component decreases in thickness as pyroclasts are lost by deposition more rapidly than they are replenished from overriding the more dilute parts of the current. The granular fluid component ultimately pinches out and downcurrent of this point the current is fully dilute and deposition is dominated by traction (Fig. 12). Thus, some granular fluidbased currents can change downstream to fully dilute currents, and the corresponding depositional flow-boundary zones become increasingly dominated by tractional processes (e.g. Sohn & Chough, 1989; Chough & Sohn, 1990; Colella & Hiscott, 1997; Branney & Kokelaar, 2002; Fig. 11). Unsteadiness in fully dilute pyroclastic density currents (e.g. Valentine, 1987; Sohn & Chough, 1989) may cause the deposition of numerous layers during the passage of a single current. Downstream lithofacies transitions from lapilli tuff (mLT or dbLT) to tuff (mT or dbT) as in DFT 1 and 3 (Fig. 10), indicate that the depositional flow-boundary zone of some granular fluid-based pyroclastic density currents remained at relatively high concentrations irrespective of the distance from source, inhibiting winnowing within lowermost zones. In this case, the currents seem to have maintained a flux of clasts towards the flow-boundary zone equal to, or greater than, the flux of clasts lost from the base by deposition. Such currents probably carried a large proportion of their load as granular fluid. Downstream transitions from cross-stratified lapilli tuff to cross-stratified tuff (DFT 4, Fig. 10) indicate that some depletive currents were tractional over long tracts, from near-proximal granular fluid-based currents to distal fully dilute currents. Cross-stratified lapilli tuff (xsLT) has been reported in several, mostly proximal, facies of ignimbrites (e.g. Mount St. Helens, Rowley et al., 1985; 15 June 1991 ignimbrite of Mount Pinatubo, Branney & Kokelaar, 2002; Cape Riva ignimbrite near Oia, Santorini, Druitt, 1985) and is a rather enigmatic lithofacies in terms of depositional conditions. The very poor sorting and matrix support indicate that the efficiency of particle segregation (e.g. by winnowing) in the flowboundary zone was significantly limited by high
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1184
R. J. Brown et al.
particle concentrations, yet the cross-stratification indicates that turbulence of the granular fluid, and tractional processes, nevertheless occurred. This study envisages particularly energetic turbulent sweeps through a basal granular fluid, which is consistent with restriction of the lithofacies to proximal regions, where the currents moved sufficiently rapidly so that turbulence affected even the lower granular fluid part of the current. The various flow-boundary zone transformations inferred from the Glaramara deposits relate to differences in the initial physical constitution of the erupted tephra that falls back to form the currents, for example, in the amount of potential and thermal energy and in the particle concentration, all of which vary considerably during phreatomagmatic explosivity (e.g. Kokelaar, 1986). Although the pyroclastic density currents deposited most of their coarser fraction within 5 km of source, they retained enough pyroclastic material to remain significantly denser than the atmosphere and continued to travel to >10 to 12 km from source (Fig. 12). In distal regions, the currents comprised only fine-grained ash and accretionary lapilli (see Fig. 8), although the occurrence here of both massive and stratified deposits indicates that particle concentrations in the flow-boundary zones of successive currents varied considerably.
Volcano location and morphology The inferred eruption style (i.e. repetitive phreatomagmatic explosions), the dominance of pyroclastic density current deposits over pyroclastic fall deposits, and the long, thin and wedge-like geometry of the preserved beds (Fig. 3) indicate that the Glaramara tuff outcrops represent subradial cross-sections through the south and southeastern flanks of an extensive tuff ring (see Fisher & Schmincke, 1984; Cas & Wright, 1987; Sohn, 1996). The Glaramara tuff ring thins and fines south-west over a distance of 5 km, with a concomitant decrease in the number and thickness of individual beds (Fig. 3). Coarse-grained flank deposits, best illustrated by Unit 2, fine outwards into a broad, tephra blanket that consisted predominantly of fine ash and ash aggregates. Palaeocurrent directions derived from foreset dips in stoss-erosional dune forms (e.g. subunit 2b; Fig. 6) indicate dispersal of pyroclastic density currents towards the south. By comparison with less dissected tuff rings (e.g. Heiken, 1971), we infer that the missing proximal parts of the Glaramara tuff ring probably dipped outwards
somewhat more steeply than the preserved medial and distal deposits, and may have been several tens of metres thick. By analogy to modern tuff rings it may be inferred that the vent lay at least several kilometres north of the most northerly outcrops in the study area (Fig. 1). Assuming a radial distribution of tephra from source, the Glaramara tuff ring may have originally covered an area of >800 km2, with a tentative erupted volume of 0Æ25 to 0Æ5 km3 (VEI 4). Comparable modern eruptions include the 1965 eruption of Taal Volcano, Philippines (e.g. Moore et al., 1966) and the 0Æ45 km3, 3Æ8 ka Astroni eruption, Campi Flegrei, Italy (Isaia et al., 2004).
Eruption dynamics The Glaramara eruption was triggered when rising dacitic magma encountered water in a shallow aquifer in the partially flooded Scafell caldera. The low profile of the Glaramara tuff and bedforms, the absence of slurry deposits, and the absence of evidence for moist ash accretion and soft sediment deformation structures, together, indicate that a large volume of water (i.e. a large body of surface water) did not flood the vent. The eruption commenced with a series of relatively weak phreatomagmatic explosions that generated low mass-flux currents that deposited most of their coarser clasts near to the vent and constructed a tephra blanket 6 to 8 km from the vent (Unit 1; Figs 3 and 12). The eruption intensity then waxed to its climax as successively more powerful explosions generated currents that were sufficiently competent to transport lithic lapilli to beyond 5 km and ash to more than 10 to 12 km from source (Unit 2). This waxing resulted in a progradation of facies away from the vent (Fig. 13). The Glaramara tuff is over a metre thick at its most distal exposure (Fig. 3) and presumably the ash layers extended significantly farther. Abundant pumice (preserved as fiamme) in the coarser-grained lithofacies suggests that some of the erupting magma did not fully interact with external water during the climactic phase of the eruption. Magma discharge rates exert a strong control over eruption style in phreatomagmatic eruptions (Houghton & Nairn, 1991; Houghton et al., 1996). It is inferred here that an increase in the mass flux of magma erupted during the waxing and climactic stages of the Glaramara eruption could have led to a decreased water:magma ratio overall, with incomplete phreatomagmatic fragmentation, perhaps towards the middle of the rapidly ascending flow. At the
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera
1185
Fig. 13. Time-distance plot showing the overall longitudinal architecture of the Glaramara tuff-ring. The upwardcoarsening to upward-fining sequence results from the inferred wax–wane dynamics of the eruption, which also produced the progradational to retrogradational sequence of facies. For clarity, the numerous thin flow-units that record the individual short-lived pyroclastic density currents, are not shown. Horizontal changes represent downcurrent lithofacies transitions resulting from the non-uniformity of the currents (see Fig. 10).
same time, for a large part of the eruptive jet there was highly efficient mixing and heat exchange between magma and water, with no excessive quenching, so that the explosivity was extremely powerful and capable of thrusting the tephra to considerable heights. This power along with increased mass in the fountain and current, could account for the extreme runout distances. The waning of the eruption was characterized by progressively weaker phreatomagmatic explosions, which produced a retrogradation of facies towards the vent (Fig. 13). Initial waning stages were characterized by pyroclastic currents with traction-dominated flow-boundary zones (subunit 2b). Prolonged pauses (seconds to minutes) between later currents allowed ash and accretionary lapilli to settle out from the atmosphere (subunit 2b). Weaker phreatomagmatic explosions in the closing stages of the eruption generated lesser pyroclastic density currents that mirrored the opening phase of the eruption, with deposition of ash and ash aggregates across the study area. An ash-rich eruption column and umbrella cloud probably developed during the persistent phreatomagmatic explosivity, as during the 1965 eruption at Taal Volcano (Moore, 1967), but deposits from this have not been recognized. The discontinuity of many layers in the distal ash deposits is not consistent with a simple fallout origin (see Figs 3 and 12), but any lofted plume may have been dispersed in a different sector.
A unified view of pyroclastic density currents Tuff rings conventionally were reported to be composed of ‘base-surge deposits’ (e.g. Moore, 1967; Crowe & Fisher, 1973; Chough & Sohn, 1990), and these were widely regarded to be distinct from ignimbrites (Fisher, 1979; Fisher et al., 1980; Walker, 1983; Walker & McBroome, 1983). However, since these early seminal works, ideas have changed and ‘pyroclastic flows’ and ‘pyroclastic (or base) surges’ are increasingly considered to be end-members of a spectrum of current types. Moreover, it is now recognized that the deposit of an individual current may exhibit characteristics of both end-members through time and space (e.g. Druitt, 1998; Branney & Kokelaar, 2002; Browne & Gardner, 2005; Sohn et al., 2005). Table 2 compares and contrasts typical features of ignimbrites and of density current deposits from tuff rings. Ignimbrites are characteristically less thinly bedded, more extensive and, in many cases, more voluminous than deposits of tuff rings. Yet, ignimbrites and tuff-ring successions have many lithofacies in common; significantly, mT, dsT, sT, xsT, mLT, dbLT, sLT and mlBr occur in both, as do low-angle antidune-like bedforms (see Walker et al., 1981; Rowley et al., 1985; Cole & Scarpati, 1993; Scott et al., 1996; Brown & Branney, 2004). This analysis of the Glaramara tuff demonstrates that lithofacies and lithofacies associations at tuff rings can be interpreted in the same way as the same lithofacies in ignimbrite
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1186
R. J. Brown et al.
Table 2. A summary of comparisons of some features in tuff-ring deposits with those in ignimbrites; both deposit types have many shared features and original processes.
Deposit feature or inferred process
Tuff-ring successions (e.g. this paper; Sohn & Chough, 1989; Colella & Hiscott, 1997; Vazquez & Ort, 2006)
Massive lapilli tuff (mLT) Diffuse bedded tuff (dsLT) Parallel-stratified tuff (//sT) Cross-stratified tuff (xsT) Proximal lithic breccias Interstratified fallout layers
Common, often subordinate Common Common, may dominate Present, typically abundant Common Common
Pellets and accretionary lapilli Downcurrent transitions from lithic-rich lithofacies to pumice-rich lithofacies Downcurrent transitions from massive lithofacies to stratified and cross-stratified lithofacies Pumice-rich tops (inverse grading) to beds (flow-units) Normal grading of lithic clasts Typical runout distance
Abundant in many tuff rings Present, less marked
Common, may dominate Common Common Common, typically subordinate Common May be present between ignimbrite flow-units Abundant in some ignimbrites Common
Common
Common in proximal regions
Common
Common
Common <5 km, some >10 km
Composition Current uniformity Flow-unit boundaries
Mafic or silicic (mafic common) Depletive currents common Abundant
Common Typically >5 km, some >80 km (e.g. Taupo ignimbrite) Mafic or silicic (silicic common) Depletive currents common Common in most sheets
sheets. Several lithofacies associations occur in both tuff rings (e.g. this study; Schmincke et al., 1973; Vazquez & Ort, 2006) and ignimbrite sheets (e.g. Neapolitan Yellow Tuff, Cole & Scarpati, 1993; Poris ignimbrite, Brown & Branney, 2004; Zaragoza ignimbrite, Carrasco-Nun˜ez & Branney, 2005); examples include vertical and lateral abrupt or gradational transitions from stratified to massive facies, downcurrent transitions from coarse proximal lithic-rich breccias to distal stratified tuffs, and the presence of wax–wane sequences recorded by an inverse to normal grading through sequences. Tuff rings and ignimbrites both may contain abundant accretionary lapilli (McPhie, 1986; De Rita et al., 2002; Freda et al., 2005). Given the same lithofacies and lithofacies associations, it is reasonable to infer that the pyroclastic density currents at tuff rings are not fundamentally different from those that deposit many ignimbrites (e.g. Sohn et al., 2005). Both involve fully dilute and granular fluid-based currents that may change from one to the other, both spatially and temporally, and in both the depositional flow-boundary zone can be dominated variously by direct fallout, traction, granular flow, and fluid escape (see Branney & Kokelaar,
Ignimbrite sheets (e.g. Wilson, 1985; Branney & Kokelaar, 2002; Brown & Branney, 2004)
2002). Currents in both situations are commonly depletive, other than where topography causes local accelerations or convergent flow. The predominance of massive lapilli tuff over stratified facies in ignimbrites suggests that in ignimbriteforming eruptions granular fluid-based currents predominate over fully dilute currents. This is a result of a higher mass flux in ignimbrite eruptions. Conversely, the abundance of stratified facies at tuff rings points to a predominance of fully dilute currents. The thinner layering and more abrupt and abundant vertical grain-size variations at tuff rings reflect the more punctuated nature of hydrovolcanic eruptions, producing highly unsteady, short-lived currents, in comparison with the more sustained currents that deposit thick ignimbrites. The shorter runout distance of many tuff ring-related currents reflects the substantially lower eruptive mass fluxes of tuff-ring explosions compared with those that form ignimbrites. Finally, fully dilute currents formed at tuff rings may loft rather less readily than fully dilute currents associated with ignimbrites. This is because they are cooler and more moist than fully dilute currents that form ignimbrites, and they advance across cooler deposits (not hot, thick ignimbrite sheets) and so tend to
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera develop less buoyancy. Hence, they may remain in contact with the ground for longer, with more extensive development of tractional fine ash layers. Despite the differences, there are clearly many similarities both in process and product between the currents produced during both tuffring explosivity and ignimbrite-forming eruptions.
1187
Leicester. The manuscript benefited greatly from thorough reviews by V. Manville, Y. K. Sohn and G. A. Valentine. We thank E. Johnson, D. Millward and D. Woodhall for useful discussions and access to BGS field data. J. Zalasiewicz and S. Self are thanked for comments on previous drafts and P. Cole is thanked for discussions in the field.
REFERENCES CONCLUSIONS The Glaramara tuff records a tuff ring-forming explosive eruption with a wide area of impact (>800 km2), and probable magnitude of VEI 4. The eruption was characterized by closely successive phreatomagmatic explosions that generated numerous pyroclastic density currents that were widespread, short-lived (e.g. single-surge) and depletive. The vertical lithofacies succession of the Glaramara tuff corresponds to a progradational to retrogradational migration of the lithofacies with respect to the source vent with time (Fig. 13). This overall architecture reflects waxing and then waning explosive energy. Downcurrent facies transitions record a progressive loss of capacity with distance and progressive dilution. Variations in the composition and structure of successive deposits record variations in the initial energetics and physical constitution of successive explosive jets, with concomitant variations in flow-boundary zone processes during deposition from the resultant currents. At times, the currents appear to have merged into a more sustained, but still highly unsteady current, but when sufficient pause occurred between them, fine ash and ash aggregates accumulated by fallout from residual ash plumes. There is a great deal of overlap in the physical constitution and depositional mechanisms of single-surge pyroclastic currents generated during tuff-ring explosivity and the larger volume, more sustained pyroclastic density currents of ignimbrite-forming eruptions. Tuff-ring explosivity on the scale of the Glaramara eruption is not reported widely in the literature. Extreme runout behaviour, to distances in excess of 12 km, needs to be considered during hazard evaluations, where silicic magmas may interact with ground water.
ACKNOWLEDGEMENTS RJB acknowledges an NERC-funded studentship (GT 4/97/144) awarded at the University of
Allen, J.R.L. (1984) Sedimentary Structures: Their Character and Physical Basis, Vol. II. Elsevier, Amsterdam, 663 pp. Branney, M.J. (1988a) Subaerial Explosive Volcanism, Intrusion, Sedimentation and Collapse in the Borrowdale Volcanic Group, SW Langdale, English Lake District. PhD thesis, University of Sheffield, Sheffield, 235 pp. Branney, M.J. (1988b) The subaerial setting of the Borrowdale Volcanic Group, English Lake District. J. Geol. Soc. London, 145, 887–890. Branney, M.J. (1991) Eruption and depositional facies of the Whorneyside Tuff Formation, Lake District: an exceptionally large-magnitude phreatoplinian eruption. Geol. Soc. Am. Bull., 103, 886–897. Branney, M.J. and Kokelaar, B.P. (1994) Volcanotectonic faulting, soft-state deformation and rheomorphism of tuffs during development of a piecemeal caldera, English Lake District. Geol. Soc. Am. Bull., 109, 507–530. Branney, M.J. and Kokelaar, B.P. (2002) Sedimentation of ignimbrite from pyroclastic density currents. Spec. Publ. Geol. Soc. London, 27, 143. Branney, M.J. and Soper, N.J. (1988) Ordovician volcano-tectonics in the English Lake District. J. Geol. Soc. London, 145, 367–376. Branney, M.J. and Sparks, R.S.J. (1990) Fiamme formed by diagenesis and burial-compaction in soils and subaqueous sediments. J. Geol. Soc. London, 147, 919–922. Branney, M.J., Kokelaar, B.P. and McConnell, B.J. (1992) The Bad Step Tuff: a lava-like rheomorphic ignimbrite in a calc-alkaline piecemeal caldera, English Lake District. Bull. Volcanol., 54, 187–199. Brooker, M.R., Houghton, B.F., Wilson, C.J.N. and Gamble, J.A. (1993) Pyroclastic phases of a rhyolitic dome-building eruption: Puketarata tuff-ring, Taupo Volcanic Zone, New Zealand. Bull. Volcanol., 55, 395–406. Brown, R.J. (2001) The Eruption Histories and Depositional Mechanisms of the Poris Ignimbrite (Tenerife, Canary Islands) and the Glaramara Tuff (Scafell Caldera, U.K.). PhD thesis, University of Leicester, Leicester, 171 pp. Brown, R.J. and Branney, M.J. (2004) Event stratigraphy of a caldera-forming ignimbrite eruption on Tenerife: the 273 ka Poris Formation. Bull. Volcanol., 66, 392–416. Browne, B.L. and Gardner, J.E. (2005) Transport and deposition of pyroclastic material from the 1000 A.D. calderaforming eruption of Volca´n Ceboruco, Nayarit, Mexico. Bull. Volcanol., 67, 469–489. Bursik, M.I. and Woods, A.W. (1996) The dynamics and thermodynamics of large ash flows. Bull. Volcanol., 58, 175– 193. Carrasco-Nun˜ez, G. and Branney, M.J. (2005) Progressive assembly of a massive layer of ignimbrite with normal-toreverse compositional zoning: the Zaragoza ignimbrite of central Mexico. Bull. Volcanol., 68, 3–20.
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
1188
R. J. Brown et al.
Cas, R.A.F. and Wright, J. (1987) Volcanic Successions: Modern and Ancient. Allen and Unwin, London, 528 pp. Chough, S.K. and Sohn, Y.K. (1990) Depositional mechanics and sequences of base surges, Songaksan tuff-ring, Cheju Island, Korea. Sedimentology, 37, 1115–1135. Choux, C.M. and Druitt, T.H. (2002) Analogue study of particle segregation in pyroclastic density currents, with implications for the emplacement mechanisms of large ignimbrites. Sedimentology, 49, 907–928. Cole, P.D. (1991) Migration direction of sand-wave structures in pyroclastic-surge deposits: implications for depositional processes. Geology, 19, 1108–1111. Cole, P.D. and Scarpati, C. (1993) A facies interpretation of the eruption and emplacement mechanisms of the upper part of the Neapolitan Yellow Tuff, Campi Flegrei, southern Italy. Bull. Volcanol., 55, 311–326. Colella, A. and Hiscott, R.N. (1997) Pyroclastic surges of the Pleistocene Monte Guardia sequence (Lipari Island, Italy): depositional process. Sedimentology, 44, 47–66. Crowe, B.M. and Fisher, R.V. (1973) Sedimentary structures in base-surge deposits with special reference to cross-bedding, Ubehebe Craters, Death Valley, California. Geol. Soc. Am. Bull., 84, 663–682. Davis, N.C. (1989) The Relationship between Ignimbrite Eruption and Caldera Collapse in the Borrowdale Volcanic Group of the Central Fells, English Lake District. PhD thesis, University of Sheffield, Sheffield, 153 pp. De’Gennaro, M., Cappelletti, P., Langella, A., Perrotta, A. and Scarpati, C. (2000) Genesis of zeolites in the Neapolitan Yellow Tuff: geological, volcanological and mineralogical evidence. Contrib. Min. Pet., 139, 17–35. De Rita, D., Giordano, G., Esposito, A., Fabbri, M. and Rodani, S. (2002) Large volume phreatomagmatic ignimbrites from the Colli Albani volcano (Middle Pleistocene, Italy). J. Volcanol. Geotherm. Res., 118, 77–98. Druitt, T.H. (1985) Vent evolution and lag breccia formation during the Cape Riva eruption of Santorini, Greece. J. Geol., 93, 439–454. Druitt, T.H. (1998) Pyroclastic density currents. In: The Physics of Explosive Eruptions (Eds J.S. Gilbert and R.S.J. Sparks), Geol. Soc. London Spec. Publ., 145, 145– 182. Fisher, R.V. (1979) Models for pyroclastic surges and pyroclastic flows. J. Volcanol. Geotherm. Res., 6, 305–318. Fisher, R.V. and Schmincke, H.-U. (1984) Pyroclastic Rocks. Springer-Verlag, Berlin, 472 pp. Fisher, R.V. and Waters, A.C. (1970) Base surge bedforms in maar volcanoes. Am. J. Sci., 268, 157–180. Fisher, R.V., Smith, A.L., Wright, J.V. and Roobol, M.J. (1980) Ignimbrite veneer deposits or pyroclastic surge deposits? Nature, 286, 912. Fortey, R.A., Harper, D.A.T., Ingham, J.K., Owen, A.W. and Rushton, A.W.A. (1995) A revision of Ordovician series and stages from the historical type area. Geol. Mag., 132, 15– 30. Freda, C., Gaeta, M., Karner, D.B., Marra, F., Renne, P., Taddeucci, J., Scarlato, P., Christensen, J.N. and Dallai, L. (2005) Eruptive history and petrologic evolution of the Albano multiple maar (Alban Hills, Central Italy). Bull. Volcanol., 68, 567–591. Gilbert, J.S. and Lane, S.J. (1994) The origin of accretionary lapilli. Bull. Volcanol., 56, 398–411. Hallam, A. (1984) Pre-Quaternary sea-level changes. Annu. Rev. Earth Planet. Sci., 12, 205–243.
Heiken, G.H. (1971) Tuff rings: examples from the Fort RockChristmas Lake Valley basin, south-central Oregon. J. Geophys. Res., 76, 5615–5626. Houghton, B.F. and Nairn, I.A. (1991) The 1976–1982 Strombolian and phreatomagmatic eruptions of White Island, New Zealand – eruptive and depositional mechanisms at a wet volcano. Bull. Volcanol., 54, 25–49. Houghton, B.F., Wilson, C.J.N., Rosenberg, M.D., Smith, I.E.M. and Parker, R.J. (1996) Mixed deposits of complex magmatic and phreatomagmatic volcanism: an example from Crater Hill, Auckland, New Zealand. Bull. Volcanol., 58, 59–66. Isaia, R., D’Antonio, M., Dell’Erba, F., Di Vito, M. and Orsi, G. (2004) The Astroni volcano: the only example of closely spaced eruptions in the same vent area during the recent history of the Campi Flegrei caldera (Italy). J. Volcanol. Geotherm. Res., 133, 171–192. Johnson, E.W., Briggs, D.E.G., Suthren, R.J., Wright, J.L. and Tunnicliff, S.P. (1994) Non-marine arthropod traces from the subaerial Ordovician Borrowdale Volcanic Group, English Lake District. Geol. Mag., 131, 395–406. Kienle, J., Kyle, P.R., Self, S., Motyka, R.J. and Lorenz, V. (1980) Ukinrek maars, Alaska, 1. April 1977 eruption sequence, petrology and tectonic setting. J. Volcanol. Geotherm. Res., 7, 11–37. Kneller, B.C. and Branney, M.J. (1995) Sustained high-density turbidity currents and the deposition of thick massive sands. Sedimentology, 42, 607–617. Kneller, B.C. and McConnell, B.J. (1993) The Seathwaite Fell Formation in the Central Fells. British Geological Survey Technical Report, WA/93/43. Kokelaar, B.P. (1982) Fluidization of wet sediments during the emplacement and cooling of various igneous bodies. J. Geol. Soc. London, 139, 21–33. Kokelaar, B.P. (1983) The mechanism of Surtseyan volcanism. J. Geol. Soc. London, 140, 939–944. Kokelaar, B.P. (1986) Magma-water interactions in subaqueous and emergent basaltic volcanism. Bull. Volcanol., 48, 275–289. Kokelaar, B.P. and Busby, C. (1992) Subaqueous explosive volcanism and welding of pyroclastic deposits. Science, 257, 196–201. Kokelaar, B.P. and Ko¨niger, S. (2000) Marine emplacement of welded ignimbrite: the Ordovician Pitts Head Tuff, North Wales. J. Geol. Soc. London, 157, 517–536. Kokelaar, B.P., Raine, P. and Branney, M.J. (2007) Catastrophic incursion of a large-volume, spatter-bearing pyroclastic density current into a caldera lake: Pavey Ark ignimbrite, Scafell caldera, England. Bull. Volcanol., in press. Lajoie, J., Lanzafame, G., Rossi, P.L. and Tranne, C.A. (1992) Lateral facies variations in hydromagmatic pyroclastic deposits at Linosa, Italy. J. Volcanol. Geotherm. Res., 54, 135–143. Mastin, L.G. and Witter, J.B. (2000) The hazards of eruptions through lakes and seawater. J. Volcanol. Geotherm. Res., 97, 195–214. Mastrolorenzo, G. (1994) Averno tuff-ring in Campi Flegrei (south Italy). Bull. Volcanol., 56, 561–572. McPhie, J. (1986) Primary and redeposited facies from a largemagnitude rhyolitic, phreatomagmatic eruption: Cana Creek tuff, Late Carboniferous, Australia. J. Volcanol. Geotherm. Res., 28, 319–350. Millward, D., Johnson, E.W., Beddoe-Stephens, B., Young, B., Kneller, B.C., Lee, M.K. and Fortey, N.J. (2000) Geology of
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
The Glaramara tuff, Scafell caldera the Ambleside District. Memoir of the British Geological Survey, Sheet 38 (England and Wales), London: The Stationary Office, 228 pp. Molyneux, S.G. (1988) Micropalaeontological evidence for the age of the Borrowdale Volcanic Group. Geol. Mag., 125, 541–542. Moore, J.G. and Peck, D.L. (1962) Accretionary lapilli in volcanic rocks of the western continental United States. J. Geol., 70, 182–194. Moore, J.G., Nakamura, K. and Alcaraz, A. (1966) The 1966 eruption of Taal Volcano. Science, 151, 955–960. Morrissey, M., Zimanowski, B., Wohletz, K. and Buettner, R. (2000) Phreatomagmatic fragmentation. In: Encyclopedia of Volcanoes (Ed. H. Sigurdsson), pp. 431–445. Academic Press, San Diego. Nairn, I.A. (1979) Rotomahana-Waimangu eruption, 1886: base surge and basalt magma. N. Z. J. Geol. Geophys., 22, 363–378. Petterson, N.G., Beddoe-Stephens, B., Millward, D. and Johnson, E.W. (1992) A pre-caldera plateau-andesite field in the Borrowdale Volcanic Group of the English Lake District. J. Geol. Soc. London, 149, 889–906. Ross, J.R.P. and Ross, C.A. (1992) Ordovician sea-level fluctuations. In: Global Perspectives on Ordovician Geology (Eds B.D. Webby and J.R. Laurie), pp. 327–335. Balkema, Rotterdam. Rowley, P.D., Macleod, N.S., Kuntz, M.A. and Kaplan, A.M. (1985) Proximal bedded deposits related to pyroclastic flows of May 18, 1980, Mount St Helens, Washington. Geol. Soc. Am. Bull., 96, 1373–1383. Schmincke, H.-U., Fisher, R. and Waters, A.C. (1973) Antidune and chute and pool structures in base surge deposits of the Laacher See area, Germany. Sedimentology, 20, 553– 574. Schumacher, R. and Schmincke, H.-U. (1995) Models for the origin of accretionary lapilli. Bull. Volcanol., 56, 626–639. Schumacher, R. and Schmincke, H.-U. (1991) Internal structure and occurrence of accretionary lapilli – a case study at Laacher See Volcano. Bull. Volcanol., 53, 612–634. Scott, W.E., Hoblitt, R.P., Torres, R.C., Self, S., Martinez, M.L. and Nillos, T.J. (1996) Pyroclastic flows of the June 15, 1991, climactic eruption of Mount Pinatubo. In: Fire and Mud: Eruptions of Pinatubo, Philippines (Eds C.G. Newhall and S. Punongbayan), pp. 545–570. Philippine Institute of Volcanology and Seismology, Quezon City. University of Washington Press, Seattle, WA. Self, S. (1983) Large-scale phreatomagmatic silicic volcanism: a case study from New Zealand. J. Volcanol. Geotherm. Res., 17, 433–469. Self, S. and Sparks, R.S.J. (1978) Characteristics of widespread pyroclastic deposits formed by the interaction of silicic magma and water. Bull. Volcanol., 41, 196–212. Sheridan, M.F. and Updike, R.G. (1975) Sugar Loaf Mountain Tephra - A Pleistocene rhyolitic deposit of base surge origin. Geol. Soc. Am. Bull., 86, 571–581.
1189
Sohn, Y.K. (1996) Hydrovolcanic processes forming basaltic tuff-rings and cones on Cheju Island, Korea. Geol. Soc. Am. Bull., 108, 1199–1211. Sohn, Y.K. and Chough, S.K. (1989) Depositional processes of the Suwolbong tuff-ring, Cheju Island (Korea). Sedimentology, 40, 837–855. Sohn, Y.K., Jeong, J.O. and Son, M. (2005) Long-runout pyroclastic surge on a Cretaceous alluvial plain, Republic of Korea. Terra Nova, 17, 13–24. Sparks, R.S.J., Self, S. and Walker, G.P.L. (1973) Products of ignimbrite eruptions. Geology, 1, 115–118. Tanguy, J.-C., Ribiere, C., Scarth, A. and Tjetjep, W.S. (1998) Victims from volcanic eruptions: a revised database. Bull. Volcanol., 60, 137–144. Valentine, G.A. (1987) Stratified flow in pyroclastic surges. Bull. Volcanol., 49, 616–630. Vazquez, J.A. and Ort, M.H. (2006) Facies variation of eruption units produced by the passage of single pyroclastic surge currents, Hopi Buttes volcanic field, USA. J. Volcanol. Geotherm. Res., 154, 222–236. Walker, G.P.L. (1981) Characteristics of two phreatoplinian ashes, and their water-flushed origin. J. Volcanol. Geotherm. Res., 9, 395–407. Walker, G.P.L. (1983) Ignimbrite types and ignimbrite problems. J. Volcanol. Geotherm. Res., 17, 65–88. Walker, G.P.L. (1984) Characteristics of dune-bedded pyroclastic surge bedsets. J. Volcanol. Geotherm. Res., 20, 281– 296. Walker, G.P.L. and McBroome, L.A. (1983) Mount St. Helens 1980 and Mount Pelee 1902 – flow or surge? Geology, 11, 571–574. Walker, G.P.L., Wilson, C.J.N. and Froggatt, P.C. (1981) An ignimbrite veneer deposit - the trail marker of a pyroclastic flow. J. Volcanol. Geotherm. Res., 9, 409–421. Waters, A.C. and Fisher, R.V. (1971) Base surges and their deposits: Capelinhos and Taal volcanoes. J. Geophys. Res., 76, 5596–5614. White, J.D.L. and Houghton, B.F. (2000) Surtseyan and related phreatomagmatic eruptions. In: Encyclopedia of Volcanoes (Ed. H. Sigurdsson), pp. 495–511. Academic Press, San Diego. Wilson, C.J.N. (1985) The Taupo eruption, New Zealand. II The Taupo Ignimbrite. Phil. Trans. Roy. Soc. London, 314, 229–310. Wohletz, K.H. and Sheridan, M.F. (1983) Hydrovolcanic explosions II: Evolution of basaltic tuff-rings and tuff cones. Am. J. Sci., 283, 385–413. Yamamoto, T., Soya, T., Suto, S., Uto, K., Takada, A., Sakaguchi, K. and Ono, K. (1991) The 1989 submarine eruption off eastern Izu Peninsula, Japan: ejecta and eruption mechanisms. Bull. Volcanol., 53, 301–308.
Manuscript received 17 August 2006; revision accepted 19 March 2007
2007 The Authors. Journal compilation 2007 International Association of Sedimentologists, Sedimentology, 54, 1163–1189
