Bull Volcanol (2004) 66:226–242
DOI 10.1007/s00445-003-0306-x
RESEARCH ARTICLE
Martin G. Wiesner · Andreas Wetzel ·
Sandra G. Catane · Eddie L. Listanco ·
Hannah T. Mirabueno
Grain size, areal thickness distribution and controls on sedimentation
of the 1991 Mount Pinatubo tephra layer in the South China Sea
Received: 16 April 2002 / Accepted: 16 June 2003 / Published online: 6 September 2003
Springer-Verlag 2003
Abstract The June 15, 1991 climactic eruption of Mt.
Pinatubo produced an extensive, largely co-ignimbritederived airfall ash layer on Luzon Island and across the
central South China Sea. The layer covers an area of
~4105 km2 with a volume of 5.5 km3. Near the coast of
Luzon, the deposit consists of two units: a normally
graded basal ash bed, unimodal in grain size, and a finergrained, internally structureless upper ash bed showing
grain size bimodality. With increasing distance from the
source, the coarse particle populations of the two units
merge and migrate towards a near-constant fine population (~11 mm); the distal region is covered by a fine-mode
dominated, virtually ungraded single ash layer. The
reversal of the winds from easterly directions at uppertropospheric and stratospheric levels to westerly directions in the middle and lower troposphere indicates that
both the coarse- and fine-mode components fell out from
high-altitude eruption clouds. The high-velocity upperlevel winds, however, would have transported finegrained ash particles far beyond the South China Sea,
which suggests that their settling was accelerated by
aggregation. The boundary between the units thus marks a
change from fallout of predominantly discrete pyroclasts
Editorial responsibility: R. Cioni
M. G. Wiesner ())
Institute of Biogeochemistry and Marine Chemistry,
University of Hamburg,
Bundesstrasse 55, 20146 Hamburg, Germany
e-mail: [email protected]
Fax: +49-40-428387081
A. Wetzel
Geological-Paleontological Institute, University of Basel,
Bernoullistrasse 32, 4056 Basel, Switzerland
S. G. Catane · E. L. Listanco
National Institute of Geological Sciences,
University of the Philippines,
Diliman, Quezon City 1101, Philippines
H. T. Mirabueno
Philippine Institute of Volcanology and Seismology,
C.P. Garcia Avenue, Diliman, Quezon City 1101, Philippines
to simultaneous fallout of aggregated fines and freely
falling, coarse-grained particles. The particle populations
composing the upper ash bed were almost completely
removed from the proximal areas by the upper-level
winds. At lower elevations, the counterclockwise circulation of a typhoon over the coastal area advected the ash
south and eastward, producing a thickness maximum in
the medial region (at about 160 km from source). The
strong displacement of fines, possibly aided by wind
turbulence, led to a break in bulk tephra thinning rates
close to the coastline. In the distal region, outside the
influence of the typhoon, southwest monsoonal winds
caused a distinct lobe axis inflection and thickness
asymmetry. Within this region, at about 420 km from
source, fallout of particle aggregates created a second
thickness maximum. Comparison of the field data with
previous experimental observations and tephra flux
records in the deep sea (Wiesner et al. 1995; Carey
1997; McCool 2002) implies that the transport of ash in
the water column was largely determined by vertical
density currents. Differences in the reaction of coarse and
fine particles to turbulence in the descending plumes
probably suppressed the segregation of fines but allowed
the coarser pyroclasts to maintain their initial order of
arrival at the sea surface. Considering typical fall rates of
convective plumes, modifications of the initial fallout
position of the particles by the South China Sea current
system are on the order of only a few kilometers. The
results suggest that convective sedimentation processes
ensure the preservation of atmospheric particle transport
directions, distances, and fallout modes in the deep sea.
Keywords Mt. Pinatubo · 1991 Tephra fallout · South
China Sea · Grain size · Thickness · Volume ·
Sedimentation processes
Introduction
The grain size and thickness of deep-sea ash layers
derived from subaerial eruptions are principally con-
227
Fig. 1 Thickness contours (in mm) of the June 15, 1991 Mt.
Pinatubo ash fall deposit in the South China Sea (isobaths in m) and
on Luzon Island (unlabeled isopach is 200 mm). On-land
thicknesses (layer C) were taken from Paladio-Melosantos et al.
(1996) and E. Listanco (unpublished data 1991). Marine core sites
are denoted as follows: circles R/V Sonne cruises SO-132 in 1998
and SO-140 in 1999 (this study); triangles SO-95 in 1994 (Wiesner
and Wang 1996); squares SO-114 and R/V Ocean Researcher 1
cruise OR-455, both in 1996 (Kuhnt et al. 1996; Wiesner et al.
1997; Wang 1999); pentagons SO-115 in 1996/1997 (Stattegger et
al. 1997); diamonds 1997 cruises of R/V Explorer (F. Siringan,
personal communication 1998); inversed triangles 1998 cruises of
R/V Xiangyanghong-14 (R. Chen, personal communication 1999);
cross 1990–1992 sediment trap mooring station SCS-C (Wiesner et
al. 1995). Filled symbols ash thickness measured; half-filled
symbols ash observed, thickness not reliable due to core disturbance
or postdepositional redistribution; open symbols no macroscopically visible ash. Numerals at core sites refer to those tephra sections
that were analyzed by sieving/pipetting and correspond with the last
two digits of the station labels given in Table 1 and Figs. 3, 4, 5,
and 6
trolled by eruptive column height, initial size range of
ejected particles, and prevailing wind strength (e.g.,
Kennett 1981). Despite the very low settling velocities
of fine-grained tephra in seawater (Cashman and Fiske
1991), distinct modifications of the fallout distribution by
ocean currents have rarely been observed (Ninkovich and
Shackleton 1975). Recent experimental studies suggest
that the major mechanism counteracting subaqueous
lateral advection is that ash particles can become
entrained into highly accelerated vertical density currents
(Carey 1997). The currents were shown to form as a result
of convective instabilities in the near-surface boundary
layer when the particle concentration becomes large; the
increase in solids concentration is a consequence of the
abrupt decrease in the settling velocity of tephra as it
crosses the air-sea interface. Support for this mechanism
comes from times-series particle flux studies in the deep
sea, which documented tephra sinking speeds several
orders of magnitude higher than predicted by Stokes’ law
(Wiesner et al. 1995).
Laboratory experiments have also shown that, apart
from particle loading, convective sedimentation depends
on turbulence and water-column stratification. Turbulent
motions can create large inhomogeneities in particle
concentration fields (e.g., Eaton and Fessler 1994),
initiating convective plume formation at even very low
initial volume fractions of solids (McCool 2002). Additionally, turbulence-induced preferential concentration
can cause the finer particles to be the first convectively
settling out from stratified polydisperse suspensions
228
(McCool 2002; McCool and Parsons 2002). Stratification
by temperature or salinity in turn can provide levels of
neutral buoyancy along which particle-laden currents
spread out laterally (Carey 1997; Parsons et al. 2001).
These factors may thus modify the initial order and
location of ash particles arriving at the sea surface.
The actual implications of convective sedimentation
for interpreting tephra deposits in terms of eruptive source
signal and atmospheric conditions still need to be assessed
(Manville and Wilson 2002). A necessary first step is to
obtain information on the character and dispersal patterns
of deep-sea ash layers for which the initial conditions are
largely known. Such an opportunity has been provided by
the 1991 eruptions of Mt. Pinatubo in the island of Luzon
(Fig. 1), which generated a series of stratospheric tephra
plumes that drifted across the South China Sea (e.g.,
Koyaguchi and Tokuno 1993; Lynch and Stevens 1996;
Holasek et al. 1996). Wiesner and Wang (1996) and
Wang (1999) have shown that the tephra falls formed a
widespread submarine deposit extending from Luzon
across the central South China Sea basin towards the shelf
break off central Vietnam. Attempts to trace out processproduct relationships failed, however, because of the lack
of sediment cores to reliably define downwind and acrosslobe grain size and thickness variations. This paper
presents a comprehensive set of new deep-sea core data
on the stratigraphy, distribution, and characteristics of the
ash layer, including an estimate of the tephra volume. The
results are compared with the timing and style of the
eruption, syn-eruptive wind vectors, and ocean currents.
The major processes controlling the tephra sedimentation
are discussed.
Sequence of Pinatubo eruptive events and products
The 1991 volcanic activities at Mt. Pinatubo began to
intensify on early June 12, when a brief sub-Plinian event
generated an eruption column of at least 19 km in height
(Hoblitt et al. 1996). Tephra falls were recorded mainly
southwest of the vent and produced a normally graded
layer of lapilli- to sand-sized andesite scoriae with minor
dacite pyroclasts (layer A; Paladio-Melosantos et al.
1996). Another three vertical eruptions occurred between
late June 12 and the early afternoon of June 14, followed
by numerous pyroclastic-surge-producing explosions
which continued for nearly 24 h (Hoblitt et al. 1996).
Ash clouds were lifted to altitudes of 5 to >24 km and
largely moved to the west and southwest (Lynch and
Stevens 1996; Holasek et al. 1996). The resulting fallout
deposits consist of a finely laminated layer of predominantly silt-sized tephra (layer B; Paladio-Melosantos et
al. 1996). Upward in the layer, the proportion of dacite to
andesite pyroclasts increases, reflecting the progressive
replacement of the limited volume of andesitic magma in
the conduit by dacitic magma (Hoblitt et al. 1996).
In the early afternoon on June 15, shortly before 13:41
(GMT+8), the eruptive activity escalated into a 9-h-long
climactic phase (Holasek et al. 1996; Fig. 2). During the
Fig. 2 Velocity-height section of mean zonal (solid line) and
meridional (dashed line) winds on June 15, 1991, combined with
the timing and plume top altitudes (bold curved line) of the June
15–16 Pinatubo eruptions. Wind velocities were obtained from
ECMWF ERA-40 reanalysis data sets (for reference, see European
Center for Medium-Range Weather Forecasts 2000) and averaged
over the area 10.25–18.75N, 109.25–120.25E; positive (negative) values indicate westerly (easterly) flow for the zonal
component and northerly (southerly) flow for the meridional
component. Chronology of eruptive plume events was adapted from
Holasek et al. (1996)
first hour, the eruption column rapidly rose to about
40 km altitude and a massive umbrella plume intruded
between 25 and 35 km, expanding in all directions
(Koyaguchi and Tokuno 1993). The total column height
then gradually lowered by 7 km until 16:41 (Fig. 2), at
which time the dome-shaped overshoot of the plume top
reduced its height from about 17 to 3 km (Holasek et al.
1996). A large plume, re-attaining 35 km in altitude, was
sustained over the vent for the following 2–3 h (Fig. 2).
Subsequently, the eruption column continuously decreased in height, diminishing to pre-climactic levels
shortly after 22:41 (Fig. 2). During the first 4–5 h after the
onset of the eruption, the stratospheric umbrella plume
spread nearly symmetrically to about 750 km in diameter
(Holasek et al. 1996). The plume was then sheared out by
the upper-level winds and started to drift to the westsouthwest at altitudes in middle stratosphere down to the
upper troposphere (Self et al. 1996). By early June 16,
when its leading edge had approached the east coast of
Vietnam, the ash cloud covered the entire South China
Sea between 7 and 18N (Lynch and Stevens 1996;
Holasek et al. 1996). On Luzon Island, the climactic
fallout deposited a normally graded lapilli- to sand-sized
tephra bed (layer C; Paladio-Melosantos et al. 1996). The
layer contains white, phenocryst-rich and tan, phenocrystpoor dacite pumice, with the former (dominant) type
closely resembling the one present in the pre-climactic
deposits (Hoblitt et al. 1996; Pallister et al. 1996).
229
From June 16 onward, energetic ash emissions further
declined in both number and magnitude (Fig. 2) and
virtually stopped by the end of September (Pinatubo
Volcano Observatory Team 1991). Low-level post-climactic ash clouds were predominantly advected to the
northeast, while those that rose into upper troposphere or
penetrated the tropopause (~17 km) were carried to the
southwest. Ash falls related to this phase formed a
laminated, fine-sand- to silt-sized dacitic tephra bed (layer
D; Paladio-Melosantos et al. 1996).
On-land tephra thickness contours reflect both the
major atmospheric transport directions and changes in the
dispersive power of the eruptions. Layer A, B, and C ash
lobes trend southwestward, ever-widening from A to C,
whereas layer D stretches in a narrow band to the
northeast with a subsidiary lobe oriented to the southwest
(Paladio-Melosantos et al. 1996). Near-vent maximum
thicknesses of layers A, B, C, and D are 9, 46, 33, and
94 cm, respectively; along the lobe axes the thicknesses
decrease to 0.7, 1.5, 16, and 0.4 cm at the west coast of
Luzon over a distance of ~50 km (Koyaguchi and Tokuno
1993; Paladio-Melosantos et al. 1996). Based on an
extrapolation of the isopachs, the bulk volume of tephrafall deposits (layers A to D) was estimated at 3.8–4.8 km3
(1.8–2.2 km3 dense-rock equivalent, DRE) (PaladioMelosantos et al. 1996). Adding the dense-rock volume
of the June 15 pyroclastic flow deposits (2.1–3.3 km3;
Scott et al. 1996), the total magma volume would
approximate 3.9–5.5 km3. Holasek et al. (1996) calculated
a total magma volume of 4.8–7.8 km3 on the basis
eruption rates derived from plume altitude measurements
made with satellite and radar data.
Sampling and methods
Investigations in the South China Sea were conducted by
R/V Sonne (cruises SO-132, June–July 1998, and SO-140,
April–May 1999) at a total of 89 stations and water depths
between 500–4,500 m (Wiesner et al. 1998, 1999; see
Fig. 1). Sea-bottom cores were obtained by a Wuttke
standard box-corer (equipped with a 505060 cm spade
box) and a Wuttke multi-corer (equipped with 12 plastic
tubes each 65 cm in length and 9.5 cm in diameter). The
thickness of the ash layer was determined at various
points by direct observations on split cores (using a yardstick or caliper) and by a Northwest Metasystems microresistivity probe which was inserted into the cores
immediately after retrieval. The probe was similar to
the system described by Andrews and Bennett (1981) and
consisted of a 4-mm-wide pin with a <0.1-mm-thick edge
at its vertex, mounted on a rack-and-pinion gear, with a
calibrated scale of 0.1 mm accuracy. Resistivities varied
with particle grain-size and packing, and significant
breaks in values occurred at the ash-nepheloid layer
interface and at the boundary between the ash and
underlying siliciclastic sediments (clay) (for further
details, see Berner 2001). The probe was inserted in 1mm steps or less and the thickness values were directly
read from the scale. Measurements were carried out at
positions that visually showed no signs of bioturbation,
which was later cross-checked by inspection of X-ray
radiographs. At least nine positions per box-core and five
positions per multi-core-tube were selected. Thickness
values were averaged per core and the arithmetic mean of
all thickness averages per station was used for isopach
construction. Contouring was performed with a PC-based
program that was also used to calculate isopach areas. All
thickness data refer to uncompacted ash beds as the X-ray
radiographs did not exhibit any evidence for dewatering
of the deposits such as elutriation pipes, dish structures or
convolute bedding (Wiesner et al. 1998, 1999).
Tephra grain size spectra were obtained by wet-sieving
(<4.33f) and pipette analysis (>3.66f) at 1/3-phi intervals
(particle populations >9f were not further fractionated).
The overlapping range was used to reduce possible errors
due to misclassified material as sieve and hydraulic size
distributions are joined (Matthews 1991). This was
attained by subjecting the sieve fraction >4.33f to the
pipette method and starting the analyses at 3.66f. The
material hydraulically coarser than the sieve aperture it
passed through was then added to the quantity retained in
the appropriate interval of the wet-sieving method. Since
at a variety of stations the amounts of ash recovered from
the cores were insufficient to fully perform the sieving/
pipetting procedure, all ash samples were subjected to a
Galai CIS-100 Laser Particle Sizer. Analyses were
performed at steps of 2 and 12 mm over the size ranges
2–600 and 12–3,600 mm, respectively. Particle counts
were arithmetically combined on the basis of their size
range overlap and normalized to reflect a 100% volume
distribution between 2–3,600 mm. Maximum, median,
and mean grain sizes were calculated as f1, f50, and
(f16+f50+f84)/3, respectively; sorting was taken as
[(f84f16)/4]+[(f95f5)/6.6] (Folk and Ward 1957). For
the particle populations <9f, the correlation between the
mean grain diameter obtained by the Laser method (ML)
and sieving/pipetting (MS) was MS=1.13+1.03 ML
(r2=0.96) at ML=1–5f; the MS-intercept largely reflects
the bias of the former technique toward producing
coarser-grained size spectra relative to conventional
settling methods (Agrawal et al. 1991; Syvitski et al.
1991). The standard deviation of the mean grain diameter
ranged between €0.01f at 1.33f and €0.06f at 5.03f for
the Laser method (triplicate analyses per ash sample), and
between €0.14f at 2.30f and €0.18f at 6.24f for the
combined sieve/pipette method (duplicate analyses). Unless otherwise indicated, all grain size data presented in
the text and figures are those obtained by the latter
procedure.
Point counting of the modal constituents was performed with standard petrographic and stereo-microscopes. Counts were converted to weight percent by using
the densities for the Pinatubo glass and phenocrysts given
in Luhr and Melson (1996). Herein the major microscope
observations are described, while a detailed presentation
of the modal data will be given in Wiesner et al. (in
preparation). Bulk grain densities (Dg) of the ash samples
230
Fig. 3 Stratigraphy and grain size distributions of the submarine
Mt. Pinatubo ash fall deposit in four representative cores downwind
along the main dispersal axis (distance from source is given in
brackets; see inset for core locations). Open bars unit I; black bars
unit II; grey bars single ash layer
were measured gas-pycnometrically using a Micromeritics 1305 Multivolume Pycnometer. Bulk ash densities
(Db) were calculated as Db=Dg(1P/100), where P is the
total porosity (porosity values were taken from Haeckel et
al. 2001). Electron microprobe analyses were carried out
on a Philips Camebax 724. To obtain information on the
composition of the very fine particles, all fractions >9f
were analyzed by a Philips PW 3710 X-ray diffractometer. In the following sections, the terms proximal, medial
and distal are used to designate downwind ash deposits at
distances of <50 (on-land tephra), 50–340, and >340 km,
respectively, from the volcano.
sharp contact to unit II. Particle populations are unimodal
in grain-size with maximum clast and median diameters
around 0.4f and 2.2f, respectively, and a fine tail that
extends into the clay fraction (Fig. 3 and Table 1).
Downwind along the axis, the thickness decreases to
1 mm southeast of the Scarborough seamounts (station
132-13; Fig. 4a) over a distance of 250 km. This is
paralleled by the following textural changes: (1) maximum clast and coarse particle median grain sizes
decrease; (2) the percentage of fines increases producing
a secondary mode at around 6.5f; and (3) particle sorting
becomes poorer (Figs. 3, 5; Table 1).
Unit II is a finer-grained, virtually ungraded greyish
tan layer (Fig. 3), showing a southward displacement of
its isopachs relative to the position of the lobe axis of unit
I (Fig. 4b). The greatest near-coast thickness of the unit
occurs west of Lubang Island (27 mm at station 132-37;
Fig. 4b and Table 1). There, the particle populations are
poorly sorted (1.5f) and distinctly bimodal with modes at
around 3.6 and 6.8f, and a maximum clast size of about
2.8f (Table 1). Upsection in the layer, the coarse ash
component fines, whereas the fine mode remains nearly
constant (which typifies all coastal deposits of unit II;
Fig. 3). From station 132-37 towards the west-northwest
the thickness of the unit decreases, passes through a
minimum of 11 mm at 118.2E, and then increases again
to 29 mm at station 132-13 (Fig. 4b). Concurrently, the
Grain size and dispersal characteristics of the
submarine tephra
The submarine Pinatubo airfall deposit consists of two
units in the medial areas, referred to as unit I (base) and
unit II (top), whereas the distal region is covered by a
single ash layer. Unit I is a coarse-grained, ‘salt-andpepper’-colored tephra (Fig. 3) which displays a regular
west-southwesterly directed dispersal pattern up to the
Manila Trench; beyond the trench the lobe axis curves to
the west-northwest (Fig. 4a). In the on-axis core closest to
the coast (station 140-25, 88 km from the volcano), the
unit is 72 mm thick and distinctly normally graded with a
231
Table 1 Grain size data (in f-units) of the 1991 Mount Pinatubo
tephra layer and its stratigraphic units in the South China Sea. The
position of the core stations is shown in Fig. 1. For the formulas
Core station
number
Distancea
(km)
SO-140-25
88
SO-132-41
111
SO-132-37
134+
SO-132-29
213
SO-132-28
248
SO-132-13
SO-132-14
SO-140-04
SO-132-07
SO-140-03
SO-132-05
341
418
496
594
644
768d
used in calculating the statistical grain size parameters, refer to the
Sampling and Methods section
Stratigraphic
section
Thickness
(mm)
Maximum
clast size
Bulk ash
Unit II
Unit I
Bulk ash
Unit II
Unit I
Bulk ash
Unit II
Unit I
Bulk ash
Unit II
Unit I
Bulk ash
Unit II
Unit I
Bulk ash
Bulk ash
Bulk ash
Bulk ash
Bulk ash
Bulk ash
88
16
72
61
14
47
41
27
14
52
16
36
47
14
33
31
32
33
15
7
4
0.39
2.05
0.35
0.68
2.14
0.61
2.34
2.84
1.69
1.67
2.53
1.51
1.96
2.41
1.85
2.39
2.73
3.28
3.22
3.38
4.07
Median diameter
Fine
mode
Coarse
mode
6.49
6.46
6.75
6.82
6.76
6.79
6.76
6.56
6.56
6.47
6.56
6.62
6.48
6.76
6.65
6.59
6.60
5.83
6.15
2.22
3.29
2.19
2.44
3.52
2.37
3.90
4.30
3.64
3.14
4.08
2.95
3.40
3.78
3.31
3.69
4.02
4.22
4.44
-
Meanb
Sorting
>4fc
(wt%)
>5f
(wt%)
2.81
5.54
2.30
2.84
5.41
2.41
5.29
5.93
4.47
4.44
5.73
3.85
4.67
5.66
4.30
5.66
5.78
6.21
6.09
5.88
6.24
1.64
1.87
1.16
1.51
1.94
0.99
1.64
1.48
1.56
1.88
1.62
1.65
1.82
1.77
1.70
1.83
1.67
1.44
1.42
1.43
1.21
19.28
77.88
9.73
18.94
68.33
8.15
72.81
87.73
48.45
50.48
84.18
31.47
53.88
78.72
41.65
76.94
83.28
94.24
93.29
94.05
99.25
16.76
68.22
8.41
15.81
57.91
6.61
57.40
72.81
32.25
40.02
69.00
23.77
41.95
66.11
30.04
65.90
69.19
82.21
78.29
73.14
85.96
a
On axis downwind from vent; bmean (M) is correlated with the coarse mode median (MdC) by M=1.23+1.74 MdC (r2=0.90); cwt% of
particles >0f is >99.7 in all deposits; dProjected onto lobe axis (see Fig. 5)
maximum clast size and coarse-mode median shift to
coarser levels (with a relatively stationary fine mode) and
gradually merge with the ones of unit I (Fig. 5). As a
consequence, the contact between the units becomes
gradational (Fig. 3) and therefore the thickening of unit II
(thinning out of unit I) in the area of station 132-13 is not
well defined. Specific mineralogical or geochemical
signatures serving to distinguish between the two units
were not detected (see next section and Wang 1999) and,
hence stratigraphic correlation with the distal ash bed
cannot be established. This bed consists of a single,
internally structureless pale-grey layer, which is characterized by the dominance of the fine ash component
(Fig. 3). Figure 4 shows that the mean grain sizes of the
layer cover the same range of values as the majority of
unit II (~5.2–6.3f), while the respective values of unit I
plot at 2.5–5.1f. On the basis of these data and the
striking similarity in structural appearance (Fig. 3), the
distal deposits were combined with unit II (Fig. 4b). The
joined isopachs produce two discrete thickness maxima:
one off Lubang Island around station 132-37 (henceforth
referred to as the Lubang maximum) and a second one
centered southwest of the Scarborough seamount chain
(Fig. 4b). The second maximum (31–33 mm) stretches to
the west-northwest over a distance of 160 km, beyond
which the layer thins regularly towards the east coast of
Vietnam (Fig. 1). In the same direction, maximum clast
and coarse-mode median diameters continuously fine,
whereas the fine mode maintains relatively uniform
values (Fig. 5).
The thickness distribution of the total submarine
deposit defines a tephra lobe extending from Luzon
across the South China Sea from 11.5 to 16.5N and
111.5 to 120E (Fig. 1). Proximal to the coast, the
thickness contours are dominated by unit I and the lobe is
nearly symmetric about its dispersal axis which trends to
the west-southwest. At about 230 km from the volcano,
the axis swings to the west-northwest, accompanied by a
pronounced asymmetry of the lobe with a north-northeastward biased thickness distribution perpendicular to
the axis. South of the Macclesfield Bank, the orientation
of the axis becomes obscure because of the limited
amount of across-lobe thickness data (Fig. 1). Along the
dispersal axis, the total deposit thins exponentially from
88 mm near the coast to 35 mm at 118.2E (~280 km from
source). Farther downwind, the bulk ash isopachs do not
contour a discrete second thickness maximum but a
narrow, spur-like structure south of the Scarborough
seamounts. Beyond this structure (referred to as the
‘Scarborough spur’ in Fig. 5), the bulk deposit continuously thins to the 1-mm isopach at ~980 km from the vent
(Fig. 1). The bulk ash on-axis grain size distributions
exhibit a medially dominant coarse mode that migrates
toward and eventually merges with a relatively constant
fine mode with increasing distance away from the coast
(Figs. 5, 6). Beyond 640 km, the particle populations are
232
Fig. 4 Thickness contours (in mm) of a unit I (unlabeled isopach is
50 mm) and b unit II (*combined with the single ash layer to show
the total distribution of fine ash). Core site symbols are as in Fig. 1
but filled where the boundary between unit I and II is sharp, halffilled where boundary is gradual, and open where no boundary is
present. Numbers in italics at core sites are mean grain sizes (in funits) for the particle populations <9f. Values were obtained by the
Laser method and converted to sieve/pipette particle means on the
basis of the equation given in the section on Sampling and Methods
(bracketed values denote sites where contamination of the ash
section by particles from above and beneath was difficult to avoid
during sampling). Bold dashed line represents the bulk ash lobe
axis (see Fig. 1)
unimodal and tail off towards the 9f-level (Fig. 6).
However, the quantity of particles >9f (~10 wt%; Fig. 6)
is too large to be expressed as a continuing fining tail,
which suggests the existence of a second mode in the very
fine size range of the distal deposits. X-ray powder
diffractograms of the >9f fractions (not shown) were all
characterized by a large, unresolved baseline hump. This
typically occurs in the presence of substantial amounts of
amorphous silica (e.g., Poulsen et al. 1995), indicating
that these fractions are dominated by glass shards.
Fig. 5 Downwind variations in a thickness, b maximum clast size,
and c median diameter of the submarine Mt. Pinatubo tephra layer
and its stratigraphic units as a function of distance from source
along the bulk ash dispersal axis. Short-dashed thickness line in a
indicates diffuse contact between unit I and II. Bracketed data in b
and c refer to core sites located south of the lobe axis (their source
distance was orthogonally projected onto the center line; see Fig. 1
for core positions). Note that stations 132-37, -28, -29, -13, and -14
plot along a line which connects the thickness maximum of unit II
with the one of the distal ash layer (cf. Figs. 1, 4)
Correlation with the subaerial tephra fall deposits
Unit I contains white (and very rarely tan) pumice and
vitric shards totaling up to a glass content of 24–36 wt%.
The remaining fraction consists of 38–52 wt% plagioclase
(mean An=35 mol%); 20–23 wt% hornblende, occasionally rimmed by cummingtonite (mean Mg#=67); 3–5 wt%
233
Fig. 6 Bulk ash grain size distributions of the submarine Mt.
Pinatubo tephra layer at increasing distance downwind along the
main dispersal axis. Also shown is the percentage of ash in the >9f
size range as a function of distance from source. Size spectra of unit
I (open bars) and unit II (black bars) are stacked and normalized to
the bulk ash deposit on the basis of unit thickness and unit bulk
density (averaging 1.3 and 1.1 g/cm3, respectively); grey bars
single ash layer. Note that at station SO-132-13 unit I was too thin
(1 mm) to be analyzed separately by sieving/pipetting; also note
that the distal station SO-132-05 is positioned south of the center
line (cf. Figs. 1, 5, 6)
iron-titanium oxides; and <1 wt% biotite, quartz, zircon,
and lithic fragments. In unit II and the distal ash layer, the
content of glass is significantly increased (70–79 wt%),
while the phenocrysts essentially comprise the same
assemblage as in unit I (see also Wiesner et al. 1995 and
Wang 1999). Throughout the deposits, glass compositions
are rhyolitic (76–79% SiO2) and compatible to the highly
evolved matrix glasses of the dacitic ejecta deposited
during the June 1991 events on land (Pallister et al. 1996).
The absence of clinopyroxene (Mg#=75–85) and cooccurring low-silica glasses (68–73% SiO2) indicates that
layer A and the majority of layer B ash falls did not
contribute to the submarine tephra (see Pallister et al.
1996 and Hoblitt et al. 1996 for the composition of the
pre-climactic deposits). Discrete ash beds underlying unit
I were not observed. Glass shards disseminated in the first
2–3 mm of the pelagic sediments beneath the base of the
unit were found to be almost entirely andesitic (53–58%
SiO2). These shards are identical to those reported by
Yang and Fan (1990) in the pre-Pinatubo surface
sediments of the eastern and central South China Sea
(55–59% SiO2). However, among the mineral phases
present few juvenile phenocrysts were detected, consisting of clinopyroxene (Mg#=82), plagioclase (An37), and
hornblende (Mg#=68), all surrounded by clear vesiculated
glass (69–76% SiO2). Pre-climactic ash plumes must
therefore have crossed the eastern South China Sea,
which is also seen from satellite images (Lynch and
Stevens 1996; Oswalt et al. 1996).
Late B ash falls produced in the morning and early
afternoon of June 15 were devoid of clinopyroxene
(Hoblitt et al. 1996) and hence are indistinguishable in
composition from the submarine tephra. Yet, the dispersal
of the June 12–14 tephra falls in combination with the
following observations suggest that, if at all, only small
quantities of the late B ash settled out over the offshore
stations: (1) the injection heights of the ash did not exceed
the plume altitudes attained during the June 12–14
eruptions (Holasek et al. 1996); (2) the strength and
direction of the high-altitude winds did not vary significantly during the pre-climactic phase, while the low- and
mid-level winds shifted to more westerly directions on
June 15 (see Gibson et al. 1997 for data sets); and (3) the
late B particle populations were similar in grain size to
those ejected during the preceding layer B eruptive events
(Koyaguchi and Tokuno 1993; Paladio-Melosantos et al.
1996) and, hence, should not have been wind-transported
to greater distances. Nonetheless, some of this ash could
have been carried to great heights and advected far to the
west as the early paroxysmal eruption column rose
through the previous plume (Holasek et al. 1996).
Regardless of this complexity, the base of unit I marks
the onset of the climactic fallout because those particle
sizes composing layer B (>4f; see Koyaguchi and Tokuno
1993) are almost absent (Fig. 3).
The post-climactic ash falls plot in the range of unit II
in terms of both grain size and composition (PaladioMelosantos et al. 1996; Pierson et al. 1996), but are
unlikely to have contributed significantly to the unit. This
is deduced from two observations: (1) the dispersal
pattern of layer D is inconsistent with the vast areal extent
and southward displacement of unit II; and (2) the rate of
thinning of the layer to the west is distinctly higher and its
thickness along the coast of Luzon is much lower than the
ones of layers A and B (see section on Sequence of
Eruptive Events and Products). Taken collectively, the
above data lead to the conclusion that the bulk of the
deep-sea tephra is the product of the climactic eruption
phase. Accordingly, the isopachs of the submarine ash
layer were combined with the ones of layer C (Fig. 1).
The course of the thickness contours of the two units
suggests that the lower bed extends onto Luzon Island,
whereas the upper bed appears to be largely restricted to
the South China Sea (Fig. 4). In the climactic fall deposits
on land, the clear textural bipartition is lacking (PaladioMelosantos et al. 1996). Around the volcano, the normal
grading of layer C is interrupted by thin pyroclastic flow
and ash cloud deposits (Scott et al. 1996). This was
initially taken as evidence that during the 9-h climactic
234
Fig. 7 Wind vector fields on June 15, 1991 at 14:00 local time
(GMT+8) in a the lower stratosphere (70 hPa, ~19 km), b the upper
troposphere (250 hPa, ~11 km), c the middle troposphere (600 hPa,
~4.5 km), and d the lower troposphere (925 hPa, ~0.8 km). Vectors
were calculated from ECMWF ERA-15 reanalysis data sets (for
reference see Gibson et al. 1997). Scales of velocity vectors are 10
and 20 m/s. Pinatubo tephra thickness contour intervals (in mm) are
as in Fig. 1
phase Plinian pumice fall and (co-)ignimbrite activity
were at least partly concurrent (Scott et al. 1996).
Integrating new stratigraphic data with earlier eyewitness
accounts, Rosi et al. (2001) arrived at the following
results: (1) layer C accumulated during the first 3 h of the
eruption; and (2) a large part of pyroclastic flow activity
occurred thereafter, unaccompanied by ash falls on
Luzon. The authors considered that the large plume
above the vent following the initial 3-h peak (Fig. 2) was
co-ignimbrite. This implies that downwind-displaced ash
falls from this high-altitude plume may have generated
unit II, while unit I represents the submarine counterpart
of layer C. Recently, Dartevelle et al. (2002) inferred that
even layer C was substantially derived from co-ignimbrite
clouds, with lesser Plinian input than previously thought,
based on two findings: (1) the distinctly finer-grained
nature of the layer compared to fall deposits produced by
Plinian eruptions of similar intensity; and (2) the strong
water enrichment in the ash plumes (indicated by satellite
data), which was not expected for a Plinian column.
Atmospheric controls on the tephra dispersal
During the climactic eruption, the winds at stratospheric
and upper tropospheric levels were vectored to azimuths
between 250 and 290 at speeds of 15–32 m/s; maximum
wind velocities were attained at heights between 30 and
35 km (Figs. 2, 7). At lower levels, a drastic departure
from the climatological mean summer wind regime was
induced by Typhoon Yunya which approached eastern
Luzon in the early morning of June 15 (Oswalt et al.
1996). Upon landfall, Yunya downgraded to a tropical
storm and passed about 75 km northeast of Mt. Pinatubo
during the initial stage of the climactic eruption. The
storm then tracked north-northwestward, entered the
South China Sea at approximately 20:00, and finally
dissipated west of northern Luzon in the early morning of
June 16 (Oswalt et al. 1996). Its counterclockwise
circulation was about 500 km across, extending over
Luzon, Mindoro and the eastern South China Sea from the
surface up to the 400 hPa (~7.6 km) level, with wind
235
velocities of up to 12 m/s (UN Economic and Social
Commission for Asia and the Pacific 1992; Figs. 2, 7).
Outside the influence of Yunya’s circulation, westerly
winds (1–8 m/s) prevailed at mid-tropospheric levels; in
the lower troposphere winds were mainly directed to the
northeast (southwest monsoon) at speeds of 10–13 m/s
(Fig. 7).
In Fig. 7a–d, representative wind vector fields illustrating the major changes in the flow regime with altitude
were superimposed on the thickness contours of the bulk
ash deposit. The plots show a clear congruence of the
stratospheric and upper tropospheric wind directions with
the lobe axis of the medial bulk ash deposit and hence
with the lobe axis of unit I (cf. Figs. 4a, 7a, b). The
southward shift of unit II, on the other hand, is in line with
the northwesterly-to-westerly flow components of Yunya’s outer circulation off Luzon (cf. Figs. 4b, 7c, d).
Figure 7c and d further reveal that the western boundary
of Yunya’s flow system roughly coincides with the
inflection point of the bulk lobe axis. Beyond that point,
both the orientation of the centerline and the lobe
asymmetry indicate that ash falls released from the
west-southwestward drifting plume were strongly redistributed by the southwest monsoon (Fig. 7d).
For the particle populations that settled out over the
South China Sea, Wiesner et al. (1995) have shown that
the presence of the ash components finer than about 5f is
unexpected. If these particles fell freely in the June 15
wind profile from the observed height of the major
eruption plumes, they would have arrived at sea level at
distances of > 2,500 km downwind from the volcano. The
calculations presented by the authors suggest that the finemode particles should have fallen out from lower
tropospheric levels in order to accumulate in the offshore
regions individually. At these levels, however, fine and
coarse pyroclasts would have been transported in the
opposite direction (Fig. 7c and d). Co-settling of coarse,
low-density and fine, high-density individual pyroclasts
can be considered to have been, if at all, of minor
significance in producing the grain size bimodality. First,
throughout the tephra bed, discrete phenocrysts of both
plagioclase and hornblende are present in both modes (see
also Wiesner et al. 1995 and Wang 1999 for the modal
composition of the tephra). Secondly, in unit II and the
distal ash, the coarse mode is dominated by pumice shards
(70%) while the fine mode is largely composed of
bubblewall shards (73%), and the density contrast
between these types of shards is small (2.46 and 2.34 g/
cm3, respectively; Fisher 1965). Furthermore, the limited
westward dispersal of the pre- and post-climactic ash falls
argues against the possibility that the widespread occurrence of bimodal spectra is due to coarse particles
overtaking fines from earlier eruptions. Consequently,
both the coarse and fine ash must have been derived from
high altitude (above Yunya’s flow system), and their codeposition requires a process accelerating the settling of
fines. Observational and numerical evidence exist for
several historical eruptive events that this process is
subaerial particle aggregation (Carey and Sigurdsson
1982; Brazier et al. 1983; Cornell et al. 1983; Varekamp
et al. 1984). Vapor condensation, electrostatic attraction,
physical interlocking, and turbulence have been proposed
as the principal mechanisms inducing aggregation (Sorem
1982; Gilbert et al. 1991; Riley et al. 2001). Basically two
types of clusters are produced: (1) single-grain clusters
composed of a large particle with adhering fines (settling
at grain-specific rates of the largest particle); and (2)
multiple-particle clusters composed of fine ash and a few
larger grains (settling at aggregate-specific rates) (Schumacher 1994). The boundary between the units can thus
be explained by a change from fallout of predominantly
single-grain clusters to simultaneous fallout of aggregated
fines and freely falling, coarse-grained particles. Multiparticle aggregates then became an increasingly important
component of the two units as the individually settling
particles attained lower terminal velocities. This is
suggested from the general increase in the proportion of
particles >5f with the fining of the coarse-mode median
downwind along the lobe axis and southward to Lubang
Island (Table 1 and Figs. 3, 6).
The secondary thickening in the distal region and the
dominance of the fine mode in the associated deposits are
strikingly similar to the dispersal characteristics of two
on-land ash layers: the 1980 Mount St. Helens and the
1991 Mount Hudson tephra (Sarna-Wojcicki et al. 1981;
Scasso et al. 1994). For the Mount St. Helens tephra,
Carey and Sigurdsson (1982) numerically reproduced the
second thickness maximum by allowing all ash particles
>4f to fall as aggregates; the terminal settling velocity of
the aggregates was assumed to correspond to the one
attained by those particles that compose the coarse-mode
median of the deposits at the maximum. In the Pinatubo
case, the co-deposited coarse mode at the second
thickness maximum southwest of Scarborough has a
median diameter of around 4.1f (Table 1). Notably,
however, in the area of the Lubang maximum the coarsemode median of unit II is at about the same level (4.3f;
Table 1). In both cases, the coarse particle populations
were found to be dominated by glass shards and
plagioclase. Hence, if the coarse-mode median is indicative of the settling velocity class of the fine-mode
particles, then the two thickness maxima have to be
attributed to fallout of ash aggregates of similar settling
behavior. Numerical models have shown that such a case
requires a considerable fraction of the total mass of
particles having the same settling velocity to be present
across the layer where the wind strength is highest
(Armienti et al. 1988). Accordingly, the ash clusters may
have fallen out from different (but great) heights, with a
large proportion of aggregated fines transported at levels
of the easterly (stratospheric) wind maximum.
The dispersal effects of the wind regime in the middle
and lower troposphere can be evaluated on the basis of the
lateral variations in particle mean diameter (note that the
mean diameter is positively correlated with the coarsemode median; see Table 1). In unit I, the particle mean
regularly shifts to finer levels not only downwind along
the lobe axis but also from the axis towards the northern
236
and southern lobe margins (Fig. 4a). In contrast, the
particle means of the upper unit tend to fine from north to
south across the lobe and from west to east towards
Lubang Island (Fig. 4b). These trends indicate that (1) the
fallout of the coarse particles generating the lower unit
was not significantly affected by Yunya’s flow system;
(2) the finer-grained particles and co-settling ash aggregates were sorted by size or density by the storm’s
crosswinds; and (3) part of this ash was transported back
toward the coast by Yunya’s westerly components. The
position of the Lubang thickness maximum thus reflects
overall lower effective settling velocities of the particle
populations of unit II at all wind levels compared to unit I.
In the distal region, the mean diameter varies irregularly
across the lobe despite the strong northeasterly winds (cf.
Figs. 4b, 7d). Considering the dominance of the finemode in that region, the likely reason is that the bulk of
the particles were transported in aggregates; any winnowing would therefore not significantly change the
granulometry of the deposits left behind.
Further information on the atmospheric transport of the
ash is obtained from the plot of the logarithm of thickness
versus square root of isopach area in Fig. 8. The bulk ash
isopach data 10 mm yield two straight best-fit lines
which intercept at a thickness of 77 mm (area1/276 km);
the thickness half-distances are 33 km for the proximal
segment and 98 km for the medial-to-distal segment. This
denotes a significant change in dispersal mechanism over
the coastal area of western Luzon. The break-in-slope is
too far from source to be explained by the transition from
column margin to umbrella cloud sedimentation. The
distance between the column corner and the central plume
axis is typically less than 30% of the column height (e.g.,
Sparks et al. 1992). In the Pinatubo case, the height of the
climactic column would imply that this transition occurs
at distances of 8–11 km from the volcano. Such a break in
thinning rates has not been observed, but data on nearvent localities where this feature would be expected to be
seen are scarce (Paladio-Melosantos et al. 1996). The two
segments are, therefore, related to fallout from the
umbrella cloud.
Bonadonna et al. (1998) have shown that breaks in
thinning rates at greater distances from source can be a
consequence of the change in sedimentation law from
intermediate to low Reynolds number. The deposits
comprising the proximal segment (i.e., layer C) are
unimodal in grain size, with a median diameter ranging
from 1 to 1f (Paladio-Melosantos et al. 1996). Such
particles would settle at intermediate Reynolds numbers
from the height of the ash cloud to sea level (Bonadonna
et al. 1998); particles >4f in size, which are expected to
fall at low Reynolds numbers, account for less than 5% of
the particle populations on land (Paladio-Melosantos et al.
1996). Close to the break-in-slope, in the bulk ash deposit
at station 140-25 (88 km from source), the proportion of
particles >4f is about 19% and then increases to >50%
farther downwind (Table 1). However, since more than
78% of the >4f-fractions are finer than 5f (Table 1), the
majority of the low Reynolds number particles did not
Fig. 8 Logarithm of thickness versus square root of isopach area
for the June 15, 1991 Mt. Pinatubo ash-fall deposit along with two
straight-line approximations for the isopach ranges 30–1 cm (T1
and T2) and 30–0.1 cm (T1 and T3) (see Fig. 1 for data source).
Open symbols denote isopachs that could not be closed with
confidence. Inset lists the regression equations for T1, T2, and T3.
Numerals in boxes are bulk volumes in km3 calculated for the two
straight-line segments T1-T2 and T1-T3 according to the method of
Fierstein and Nathenson (1992). Tmax is the extrapolated thickness
at the vent; bt1, bt2, and bt3 are the thickness half-distances for the
segments T1, T2, and T3. Dotted lines mark the point of interception
of the regression lines T1 and T2, with Tip=7.7 cm and A1/2ip=
76.3 km
settle individually. Assuming again that the co-occurring
coarse-mode median provides an approximation for
aggregate settling behavior, neither unit I nor unit II
exhibit median values plotting in the low Reynolds
number regime (except at the Lubang thickness maximum); only beyond about 450 km on-axis from source do
the median diameters clearly increase above the 4f-level
(Table 1). This transition is too large to explain the breakin-slope in terms of particle Reynolds number.
A first-order contribution to the break in thinning rates
is the almost complete removal of the fine ash from the
proximal areas and fallout beyond the coastline. The
distal overthickening of this ash can be assumed to have
provided only a second-order contribution as this process
causes a decrease in tephra dispersal (e.g., Hildreth and
Drake 1992). Another factor known to lower thinning
rates is atmospheric turbulence, causing finer particles to
be sustained aloft longer and wind-dispersed to greater
distances (Fierstein and Hildreth 1992; Hildreth and
Drake 1992). Aggregated ash may react similarly because
of the generally highly porous nature and, consequently,
low density of such clusters (Sorem 1982; Varekamp et
al. 1984; Schumacher 1994). To approximate the segment
of unit I and of the combined deposit of unit II and the
distal ash (not shown), their isopachs 10 mm were
closed by a straight line at their eastern termination. This
yields decay constants of 0.004 km1 for the combined
deposit and 0.010 km1 for unit I. The flatter slope
237
indicates that turbulence may have indeed influenced the
dispersal of the fine ash.
Subaqueous tephra sedimentation
Information on the current systems that prevailed in the
South China Sea in June 1991 is not available. Possible
effects of subaqueous advection on the tephra distribution
have, therefore, to be evaluated on the basis of long-term
monthly means and modeled data. The passage of Yunya
is not expected to have caused significant anomalies in
upper-ocean flow because of the storm’s relatively low
wind speed and short residence time (<1 day) over the
South China Sea (see Behara et al. 1998 for oceanic
response to typhoon forcing). Fig. 9a shows that in June
the surface currents generally flow to the northeast and
east-northeast at 0.18–0.22 m/s, forced by the southwestmonsoon and Ekman transport. Below the Ekman layer,
the essential feature during the summer season is an
anticyclonic gyre centered at 11N, 113E; off Vietnam,
currents are vectored to the north-northeast, then turn to
the east at 13N and finally recurve into a southsoutheasterly flow in the eastern South China Sea
(Fig. 9b). The gyre gradually diffuses downdepth, paralleled by an exponential decay in maximum current
velocities from 0.1 m/s at 50 m to less than 0.001 m/s
below 2,000 m water depth (Shaw and Chao 1994; Chao
et al. 1996). In Fig. 9 two representative current vector
fields (surface and 900 m water depth) were superimposed on the thickness contours of the bulk ash deposit.
The plots reveal that if particle residence times were
sufficiently long, the flow regime could have contributed
to both northeastward advection of fines in the distal
region and southward advection off Luzon.
An approximation of the mean settling velocity
attained by the tephra components in the water column
was obtained by Wiesner et al. (1995). The authors
recorded distal fallout of Pinatubo ash in time-series
sediment traps that were fixed to a bottom-tethered
mooring array deployed in the central South China Sea at
4,300 m water depth; the trap depths were 1,200 and
3,700 m below sea level (see Fig. 1 for trap system
position). Settling rates were determined from a trajectory
analysis of the arrival of the main ash cloud over the site
and the timing of ash interception by the traps. The data
indicated sinking speeds of about 2.2 cm/s, or two orders
of magnitude greater than predicted by Stokes’ Law for a
particle corresponding in size to the median diameter of
the ash collected by the traps (5.8f). To explore the
underlying mechanism, settling experiments analogous to
the fallout at the trap site were performed by Carey
(1997). Particles 20–180 mm in diameter and a typical
distal fallout mass flux rate of 0.17 g cm2 h1 (constrained by measurements from the 1980 eruption of
Mount St. Helens) were used in the simulations. Three
important observations were made: (1) subaerial multiparticle ash aggregates broke apart when falling onto the
water surface; (2) because of the abrupt decrease in the
Fig. 9 Current velocity vectors in the South China Sea at a the
surface (derived from June mean ship drift data; Levitus 1982) and
b 900 m water depth (re-gridded from data modeled for the
southwest-monsoon period by Chao et al. 1996). Scales of velocity
vectors are 0.2 and 0.004 m/s. Pinatubo tephra thickness contour
intervals (in mm) are as in Fig. 1
settling of tephra from that in air to water, particle
concentrations dramatically increased just below the
surface, causing convective instabilities in the boundary
layer; and (3) the instabilities initiated the descent of a
series of ash-laden plumes or vertical density currents
which settled at speeds identical to those recorded by the
traps.
The major difficulty in discussing convective transport-product relationships is the lack of experimental data
on the textural and structural characteristics of ash beds
produced by this process. The ash plumes intercepted by
the traps were characterized by a coarse mode (at 3.5f)
that graded toward a near-constant fine mode (at 6.5f)
over the course of the fallout; both modes were composed
of glass shards, hornblende and plagioglase (Wiesner et
al. 1995). Convective settling must, therefore, have
inhibited segregation of the fine particles released from
the subaerial aggregates, but allowed sorting by size of
the coarse particles. The cause for these contrasting trends
may lie in differences in the reaction of particles to
238
turbulence. Convection plumes that grow out of the
boundary layer produce large-scale turbulent motions
causing the plumes to meander, swirl, and merge (e.g.,
Hoyal et al. 1999; Maxworthy 1999; Parsons et al. 2001).
When these motions are faster than the Stokes settling of
the particles entrained, the particles will simply follow the
local velocity fluctuations and if the solid-liquid suspensions are well mixed by the turbulence, the particles
should be randomly distributed (Eaton and Fessler 1994).
The mean settling velocity for the fine-mode components
of the ash in seawater is ~0.01 cm/s. Since this value is
negligible compared to the average of the velocity
fluctuations en route (i.e., to the speed of the vertically
traveling plumes of ~2 cm/s), a deposit of unsorted ash is
likely to be generated. The coarse-mode components
clustering around 3.5f settle at ~0.7 cm/s and hence on
the order of magnitude of the plume fall rate. Such
particles will not have sufficient time to respond to the
fluctuations and will largely follow the mean flow (e.g.,
McCool 2002). The coarser pyroclasts may thus have
maintained their initial order of arrival at the sea surface
up to the point of attaining the Stokes settling velocity
class of the individual fine mode particles.
Aggregation is another process which could have
contributed to the vertical transport of the ash particles.
Wiesner et al. (1995) assumed that in this case physical
mechanisms of aggregation must have substantially
dominated over biological processes because of the
near-absence of organic matter in the descending plumes
(<0.002 wt%). The efficiency of physically coagulated,
multi-particle aggregates, however, is questioned by
shear-flow and differential-settling controlled flocculation
experiments performed by Lick et al. (1993). The authors
used organic-free suspensions of river-borne particles in
the range of very fine silt at concentrations of 0.001 to
0.2 g/L seawater. The results indicate that the time needed
to generate flocs attaining settling velocities close to
2 cm/s is on the order of weeks. Furthermore, for these
flocs to remain stable (at sizes on the order of millimeters), very low concentrations (0.001 g/L) are required.
Irrespective of the problem that solids concentrations at
the trap site cannot be quantified, the rate of aggregation
is inconsistent with the timing of ash interception
(<3 days; Wiesner et al. 1995). Moreover, the ash
particles should flocculate at slower rates owing to their
larger size. Increasing the particle concentration to 0.2 g/
L was found to reduce the time for the floc diameter to
reach its maximal value (Lick et al. 1993). However, this
would lead to an even greater discrepancy to the trap data
because as the concentration increases both floc size and
floc settling velocity significantly decrease (due to
enhanced aggregate collision rates) (Lick et al. 1993;
Teeter 2002). At concentrations of ~0.4 to 1 g/L, the role
of flocculation as a means to accelerate the primary
particles becomes insignificant; particle interactions begin
to hinder settling and dense suspensions are generated
which sink convectively even if flocs are still present
(Parsons et al. 2001; Teeter 2002; McCool 2002).
Information on the settling behavior of glass shards over
this critical range of concentrations (and up to 6 g/L) has
been presented by Wang (1999), Parsons et al. (2001),
and McCool (2002). Settling and flocculation experiments
were performed with shards in the size range of the fineash components of the Pinatubo tephra in both fresh- and
seawater. In any of the experiments, the shards did not
flocculate but settled convectively; high solids concentration, shard size, and the timing of convective plume
formation (on the order of seconds) were proposed to
have inhibited floc growth. Collectively, the experimental
results and field observations imply that convective
sedimentation was the major mode of ash transport at
the trap site. The striking similarity in the grain size
characteristics of the ash to unit II and the distal tephra
layer then indicates that these deposits were produced by
the same process.
The almost perfect normal grading of the coarse
particles in the near-coast deposits of unit I suggests that
the majority of the pyroclasts settled through the water
column individually. The small amounts of fine ash
particles present throughout the unit (~7–8%; Fig. 3 and
Table 1) may have been dragged down in the wakes of
larger grains. Alternatively, in analogy to the formation of
unit II, convective transport may have contributed to their
co-deposition. In the lower unit, the median grain size of
the coarse particle populations is 2.2f (Table 1). For the
2.2-f modal constituents (pumice, feldspar, and hornblende), the terminal settling velocities in seawater plot
between about 1 and 2.5 cm/s. This range largely accords
with typical fall rates of convective plumes (1–2 cm/s)
measured experimentally over a wide range of particle
concentrations (McCool 2002; see also Bradley 1965 and
Carey 1997). Hence, if solids concentrations or turbulent
mixing in the near-surface waters were sufficient to
initiate vertical density currents (see Introduction), the
fine ash could have co-settled with a large part of the
coarse particles. Nonetheless, farther downwind, convective transport must have become predominant, generating
the near-constant fine-mode in the lower unit as the
coarse mode fined and finally merged with the one of unit
II.
During the southwest monsoon period the South China
Sea is distinctly stratified, with a thermocline at ~25 m
and haloclines at ~10,200, and 600 m water depths (Haupt
et al. 1994). These may have provided levels of neutral
buoyancy along which density currents are deflected
laterally (Carey 1997). Based on the data obtained so far,
it is not possible to assess the consequences of this
stratification (or plume spreading on the seafloor) for the
final depositional position of the particles. An estimate
can be given for the current-induced modification of the
initial sea surface position of a particle that becomes
entrained into vertical density currents. The following
conditions are considered: (1) average plume sinking rates
of 2 cm/s; (2) maximum current velocities of 0.22 m/s at
the surface, 0.1 m/s at 50 m, 0.007 m/s at 900 m, and
0.001 m/s at 2,000 m (see Chao et al. 1996); and (3) a
linear velocity decrease between these levels. Then the
maximum horizontal, unidirectional displacement of the
239
ash particles would only be about 3.3 km until settling out
at the deepest core site (4,500 m water depth). This small
distance strongly indicates that the fallout distribution and
lobe axis inflection reflect the atmospheric conditions.
Volume of the Pinatubo eruptive products
Integration of the log thickness versus area1/2 plot to
infinity yields a total volume of 5.5 km3 for the combined
deposit of layer C and the submarine ash layer (Fig. 8).
This estimate reduces to 5.2 km3 when the areas of the
thinnest two isopachs (1 and 5 mm), closed by extrapolation, are included (Fig. 8). The ‘lost’ volume beyond
the 10-mm contour line is about 18% of the total,
providing confidence that the data measured to this
isopach give a reasonably good estimate. PaladioMelosantos et al. (1996) estimated the volume of layer
C at 1.1 km3, i.e., about 80% of the tephra volume fell out
over the South China Sea. The average bulk ash density
was calculated at 1.1 g/cm3 for both layer C (PaladioMelosantos et al. 1996) and the offshore deposits. Using a
density of 2.4 g/cm3 appropriate to dacite magma, a prevesiculation volume of 2.5 km3 DRE is obtained. The
total magma volume erupted then approximates 4.8–
6.0 km3 when the dense-rock volumes of layers A, B, and
D (~0.21 km3; Paladio-Melosantos et al. 1996) and that of
the June 15 pyroclastic flows (2.1–3.3 km3; Scott et al.
1996) are added. This is in good agreement with the
estimate given by Holasek et al. (1996) (4.8–7.8 km3
DRE) and supports the reliability of volume calculations
based on numerical models using remote sensing data.
The total mass of the climactic tephra fall and pyroclastic
flow deposits is 11–141012 kg, which translates to a
value of around 6.1 for the eruption magnitude index
(defined as log10[erupted mass] –7; e.g., Pyle 2000).
Conclusions
Tephra falls from the June 1991 climactic eruption of Mt.
Pinatubo produced a discrete ash layer on Luzon Island
and across the central South China Sea, covering an area
of ~4105 km2. Near the coast of Luzon, the deposit
consists of a normally graded unit, unimodal in grain size,
which is covered by a finer-grained, internally structureless unit showing grain size bimodality. The basal ash bed
correlates with the subaerial tephra fall deposits (layer C),
whereas the upper unit does not appear to have an on-land
counterpart. Based on recent studies (Rosi et al. 2001;
Dartevelle et al. 2002), the two units may have been
derived from co-ignimbrite-dominated ash plumes generated by two strong eruptive pulses during the climactic
phase.
With increasing distance from the volcano, the unit
distinction fades as the coarse particle populations of the
lower and upper ash bed merge and migrate towards a
stationary fine particle population; the distal region is
covered by a fine-mode dominated, virtually ungraded
single ash layer. The westerly azimuths of the wind
vectors at upper-tropospheric and stratospheric levels and
easterly azimuths in the middle and lower troposphere
indicate that both the coarse- and fine-mode components
fell out from high-altitude eruption clouds. Trajectory
estimates (Wiesner et al. 1995) suggest, however, that the
high-altitude winds would not allow the fine ash particles
to settle out individually over the South China Sea. The
mechanism that accelerated the settling of fines is
assumed to be particle aggregation, in analogy to
conclusions drawn from other tephra layers displaying
the same downwind grain size bimodality characteristics
(e.g., Brazier et al. 1983). The boundary between the units
can thus be explained by a change from fallout of
predominantly discrete pyroclasts to simultaneous fallout
of aggregated fines and freely falling, coarse-grained
particles.
The high-velocity upper-level winds almost completely removed the fine ash (aggregates and co-settling
individual particles) from the proximal areas. At lower
altitudes, the counterclockwise circulation of typhoon
Yunya advected the ash southward and back toward the
coast, producing both across-lobe sorting and a thickness
maximum in the medial region (at about 160 km from
source). The dispersal of the coarse ash forming the lower
unit was not significantly affected by the typhoon’s flow
system. The strong displacement of fines, possibly aided
by wind turbulence, led to a significant break in bulk
tephra thinning rates close to the coastline. In the distal
region, outside the influence of the typhoon, southwest
monsoonal winds created a distinct lobe axis inflection
and thickness asymmetry. Within this region, at about
420 km from source, fallout of particle aggregates
produced a second thickness maximum. Using the size
of the co-deposited coarse particles as a proxy for
aggregate settling behavior, it is suggested that the two
maxima were formed by ash clusters grouping into the
same settling velocity class. The second maximum then
implies that a large fraction of the total mass of these
aggregates was intruded at altitudes of the easterly
(stratospheric) wind maximum.
From a comparison of the field data with previous
experimental observations and tephra flux records in the
deep sea (Wiesner et al. 1995; Carey 1997; McCool
2002), it is concluded that the ash transport in the water
column was largely determined by vertical density
currents. With the onset of the fallout of aggregates and
co-settling coarse pyroclasts onto the sea surface, massive
amounts of slow settling individual particles were
released into the boundary layer. The particles then
became entrained into vertically traveling plumes that
settled at rates several orders of magnitude higher than
predicted by Stokes’ Law. Particles generating the lower
unit near the coast largely settled discretely, but with the
fining of the bulk populations farther downwind convective transport also became predominant. Re-aggregation
of particles in the water column is unlikely to have played
a role in enhancing the settling of the ash. Differences in
the reaction of coarse and fine particles to turbulence in
240
the descending plumes probably inhibited the segregation
of fines but allowed the coarser pyroclasts to maintain
their initial order of arrival at the sea surface. For the rate
of descent of convective plumes, modifications of the
initial fallout position of the particles by the South China
Sea current system are on the order of only a few
kilometers. Convective sedimentation thus appears to
ensure that both atmospheric particle transport directions
and distances, and the modes of subaerial particle settling
are preserved in the deep sea.
The calculated volume of the climactic tephra fallout
refines earlier total magma volume estimates for the
eruptive products generated in June–September 1991 to
4.8–6.0 km3 DRE. This ranks the Pinatubo eruption the
third largest of the twentieth century, following the
eruptions of Novarupta in 1912 and Santa Maria in 1902
(~13€3 and >8.5 km3 DRE, respectively; Fierstein and
Hildreth 1992; Williams and Self 1983).
Acknowledgements The authors thank the officers and crew of R/
V Sonne for their assistance in coring operations; Fernando
Siringan for providing core data obtained during the 1997 sampling
campaigns on board R/V Explorer; Ronghua Chen for releasing
unpublished core observations made on the 1998 cruises of R/V
Xiangyanghong-14; Birgit Schwinge and Peter Berner for assistance in grain-size analyses; Stefan Hagemann for access to
ECMWF data files; Hans-Rudolf Regg for assembling the microresistivity probe; Hajo Heye for the core photographs; Jeff Parsons,
Wayne McCool, William Dade, and Nick McCave for insightful
and stimulating discussions on convective sedimentation processes;
and the German Federal Ministry of Education and Research for
financial support (grant nos. 03G0132A and 03G0140A). A.Wetzel
acknowledges grant no. 2100-052256 from the Swiss Science
Foundation. Gerald Ernst and Steven Carey are thanked for their
helpful reviews. Editorial handling by Raffaello Cioni is gratefully
acknowledged.
References
Agrawal YC, McCave IN, Riley JB (1991) Laser diffraction size
analysis. In: Syvitski JPM (ed) Principles, methods, and
application of particle size analysis. Cambridge University
Press, Cambridge, pp 119–128
Andrews D, Bennett A (1981) Measurements of diffusivity near the
sediment-water interface with a fine-scale resistivity probe.
Geochim Cosmochim Acta 45:2169–2175
Armienti P, Macedonio G, Pareschi MT (1988) A numerical model
for simulation of tephra transport and deposition: applications
to May 18, 1980, Mount St. Helens eruption. J Geophys Res
93:6463–6476
Behara SK, Deo AA, Salvekar PS (1998) Investigation of mixed
layer response to Bay of Bengal cyclone using a simple ocean
model. Meteorol Atmos Phys 65:77–91
Berner P (2001) Widerstandsmessungen an unkonsolidierten marinen Sedimenten. Diplomarbeit, Departement Erdwissenschaften, Universitt Basel, pp 1–73
Bonadonna C, Ernst GGJ, Sparks RSJ (1998) Thickness variations
and volume estimates of tephra fall deposits: the importance of
particle Reynolds number. J Volcanol Geophys Res 81:173–
187
Bradley WH (1965) Vertical density currents. Science 159:1423–
1428
Brazier S, Sparks RSJ, Carey SN, Sigurdsson H, Westgate JA
(1983) Bimodal grain size distribution and secondary thickening in air-fall ash layers. Nature 301:115–119
Carey SN (1997) Influence of convective sedimentation on the
formation of widespread tephra fall layers in the deep sea.
Geology 25:839–842
Carey SN, Sigurdsson H (1982) Influence of particle aggregation
on deposition of distal tephra from the May 18, 1980 eruption
of Mount St. Helens Volcano. J Geophys Res 87:7061–7072
Cashman KV, Fiske RS (1991) Fallout of pyroclastic debris from
submarine volcanic eruptions. Science 253:275–280
Chao SY, Shaw PT, Wu SY (1996) Deep water ventilation of the
South China Sea. Deep-Sea Res 43:445–466
Cornell, W, Carey SN, Sigurdsson H (1983) Computer simulation
of transport and deposition of the Campanian Y5 ash. J
Volcanol Geotherm Res 17:89–109
Dartevelle S, Ernst GGJ, Stix J, Bernard A (2002) Origin of the
Mount Pinatubo climactic eruption cloud: implications for
volcanic hazards and atmospheric impacts. Geology 30:663–
666
Eaton JK, Fessler JR (1994) Preferential concentration of particles
by turbulence. Int J Multiphase Flow 20:169–209
European Center for Medium-Range Weather Forecasts (2000) The
ERA-40 project plan. European Center for Medium-Range
Weather Forecasts, Reading, UK, pp 1–60 (http://www.ecmwf.
int/research/era /Project/Plan)
Fierstein J, Hildreth W (1992) The plinian eruptions of 1912 at
Novarupta, Katmai National Park, Alaska. Bull Volcanol
54:646–684
Fierstein J, Nathenson M (1992) Another look at the calculation of
fallout tephra volumes. Bull Volcanol 54:156–167
Fisher RV (1965) Settling velocity of glass shards. Deep-Sea Res
12:345–353
Folk RL, Ward WC (1957) Brazos River bar, a study in the
significance of grain-size parameters. J Sediment Petrol 27:326
Gibson JKP, Kllberg P, Uppala S, Hernandez A, Nomura A,
Serrano E (1997) ERA description. ECMWF Re-Analysis
Project Report Series 1, European Center for Medium-Range
Weather Forecasts, Reading, UK, pp 1–77
Gilbert JS, Lane SJ, Sparks RSJ (1991) Charge measurements on
particle fallout from a volcanic plume. Nature 349:598–600
Haeckel M, van Beusekom J, Wiesner MG, Knig I (2001) The
impact of the 1991 Mount Pinatubo tephra fallout on the
geochemical environment of the deep-sea sediments in the
South China Sea. Earth Planet Sci Lett 193:151–166
Haupt B, Wiesner MG, Sarnthein M (1994) CTD profiles and
bottom water temperatures in the South China Sea. In:
Sarnthein M, Pflaumann U, Wang PX, Wong HK (eds)
Preliminary report on Sonne-95 cruise ‘Monitor Monsoon’ to
the South China Sea. Ber Rep Geol-Palontol Inst Univ Kiel
68:181–194
Hildreth W, Drake RE (1992) Volcn Quizapu, Chilean Andes.
Bull Volcanol 54:93–125
Hoblitt RP, Wolfe EW, Couchman MR, Pallister JS, Javier D
(1996) The preclimactic eruptions of Mount Pinatubo, June
1991. In: Newhall CG, Punongbayan RS (eds) Fire and mud:
eruptions and lahars of Mount Pinatubo, Philippines. Philippine
Institute of Volcanology and Seismology, Quezon City, and
University of Washington Press, Seattle, pp 457–511
Holasek RE, Self S, Woods AW (1996) Satellite observations and
interpretation of the 1991 Mount Pinatubo eruption plumes. J
Geophys Res 101:27635–27655
Hoyal DCJD, Bursik MI, Atkinson JF (1999) Settling-driven
convection: a mechanism of sedimentation from stratified
fluids. J Geophys Res 104:7953–7966
Kennett JP (1981) Marine tephrochronology. In: Emiliani C (ed)
The oceanic lithosphere. The sea. Wiley, New York, pp 1373–
1436
Koyaguchi T, Tokuno M (1993) Origin of the giant eruption cloud
of Pinatubo, June 15, 1991. J Volcanol Geotherm Res 55:85–96
Kuhnt W, Wiesner MG, and Shipboard Scientific Party (1996)
Cruise report, R/V Ocean Researcher 1 OR-455, Kaohsiung—
Keelong. Institute of Biogeochemistry and Marine Chemistry,
241
University of Hamburg, IBMC Library Ref No IIA 941, pp 1–
55
Levitus S (1982) Climatological atlas of the world ocean. NOAA
Professional Paper 13, US Government Printing Office, Washington, DC, pp 1–173
Lick W, Huang H, Jepsen R (1993) Flocculation of fine-grained
sediments due to differential settling. J Geophys Res 98:10279–
10288
Luhr JF, Melson WG (1996) Mineral and glass compositions in
June 15, 1991, pumices: evidence for dynamic disequilibrium
in the dacite of Mount Pinatubo. In: Newhall CG, Punongbayan
RS (eds) Fire and mud: eruptions and lahars of Mount Pinatubo,
Philippines. Philippine Institute of Volcanology and Seismology, Quezon City, and University of Washington Press, Seattle,
pp 733–750
Lynch JS, Stevens G (1996) Mount Pinatubo: a satellite perspective
of the June 1991 eruptions. In: Newhall CG, Punongbayan RS
(eds) Fire and mud: eruptions and lahars of Mount Pinatubo,
Philippines. Philippine Institute of Volcanology and Seismology, Quezon City, and University of Washington Press, Seattle,
pp 637–645
Manville V, Wilson CJN (2002) Vertical density currents: implications for the deposition and interpretation of deep-sea ash
beds. AGU Chapman Conference on explosive subaqueous
volcanism, Dunedin, New Zealand, January 21–25, 2002.
Programme and abstracts, 36
Matthews MD (1991) The effect of grain shape and density on size
measurement. In: Syvitski JPM (ed) Principles, methods and
application of particle size analysis. Cambridge University
Press, Cambridge, pp 22–33
Maxworthy T (1999) The dynamics of sedimenting surface gravity
currents. J Fluid Mech 392:27–44
McCool WW (2002) Sedimentation from hypopycnal flows. MSc
Thesis, School of Oceanography, University of Washington,
Seattle, pp 1–36
McCool WW, Parsons JD (2002) Evidence for widespread mixinginduced convective sedimentation from surface plumes. Eos
Trans AGU, 83(4), Ocean Sciences Meet Suppl, Abstract
OS11G–92
Ninkovich D, Shackleton NJ (1975) Distribution, stratigraphic
position and age of ash layer ’L’ in the Panama Basin region.
Earth Planet Sci Lett 27:20–34
Oswalt JS, Nichols W, O’Hara JF (1996) Meteorological observations of the 1991 Mount Pinatubo eruption. In: Newhall CG,
Punongbayan RS (eds) Fire and mud: eruptions and lahars of
Mount Pinatubo, Philippines. Philippine Institute of Volcanology and Seismology, Quezon City, and University of Washington Press, Seattle, pp 625–636
Paladio-Melosantos MLO, Solidum RU, Scott WE, Quiambao RB,
Umbal JV, Rodolfo KS, Tubianosa BS, Delos Reyes PJ, Alonso
RA, Ruelo HB (1996). Tephra falls of the 1991 eruptions of
Mount Pinatubo. In: Newhall CG, Punongbayan RS (eds) Fire
and mud: eruptions and lahars of Mount Pinatubo, Philippines.
Philippine Institute of Volcanology and Seismology, Quezon
City, and University of Washington Press, Seattle, pp 513–536
Pallister JS, Hoblitt RP, Meeker GP, Knight RJ, Siems DF (1996)
Magma mixing at Mount Pinatubo: petrographic and chemical
evidence from the 1991 deposits. In: Newhall CG, Punongbayan RS (eds) Fire and mud: eruptions and lahars of Mount
Pinatubo, Philippines. Philippine Institute of Volcanology and
Seismology, Quezon City, and University of Washington Press,
Seattle, pp 687–731
Parsons JD, Bush JWM, Syvitski JPM (2001) Hyperpycnal plume
formation from riverine outflows with small sediment concentrations. Sedimentology 48:465–478
Pierson TC, Daag AS, Delos Reyes PJ, Regalado MTM, Solidum
RU, Tubianosa BS (1996) Flow and deposition of posteruption
hot lahars on the east side of Mount Pinatubo, July-October
1991. In: Newhall CG, Punongbayan RS (eds) Fire and mud:
eruptions and lahars of Mount Pinatubo, Philippines. Philippine
Institute of Volcanology and Seismology, Quezon City, and
University of Washington Press, Seattle, pp 921–950
Pinatubo Volcano Observatory Team (1991) Lessons from a major
eruption: Mt. Pinatubo, Philippines. Eos Trans AGU 72:545–
555
Poulsen HF, Neuefeind J, Neumann H-B, Schneider JR, Zeidler
MD (1995) Amorphous silica studied by high energy X-ray
diffraction. Nucl Inst Meth B97, 162
Pyle DM (2000) Sizes of volcanic eruptions. In: Sigurdsson H,
Houghton BF, McNutt SR, Rymer H, Stix J (eds) Encyclopedia
of volcanoes, Academic Press, San Diego, pp 263–269
Riley CM, Shaw RA, Rose WI (2001) Turbulence-induced ash
aggregation in volcanic clouds. AGU 2001 Fall Meeting 82
(47), V42–1063
Rosi M, Paladio-Melosantos ML, Di Muro A, Leoni R, Bacolcol T
(2001) Fall vs. flow activity during the 1991 climactic eruption
of Pinatubo Volcano (Philippines). Bull Volcanol 62:549–566
Sarna-Wojcicki AM, Shipley S, Waitt RB jr, Dzurisin D, Wood SH
(1981) Areal distribution, thickness, mass, volume, and grain
size of air-fall ash from six major eruptions of 1980. In: Lipman
PW, Mullineaux DR (eds) The 1980 eruptions of Mount St.
Helens, Washington. US Geol Surv Prof Pap 1250:577–600
Scasso RA, Corbella H, Tiberi P (1994) Sedimentological analysis
of the tephra from the 12–15 August 1991 eruption of Hudson
volcano. Bull Volcanol 56:121–132
Schumacher R (1994) A reappraisal of Mount St. Helens’ ash
clusters – depositional model from experimental observation. J
Volcanol Geotherm Res 59:253–260
Scott WE, Hoblitt RP, Torres RC, Self S, Martinez MML, Timoteo
N (1996) Pyroclastic flows of the June 15, 1991 climactic
eruption of Mount Pinatubo. In: Newhall CG, Punongbayan RS
(eds) Fire and mud: eruptions and lahars of Mount Pinatubo,
Philippines. Philippine Institute of Volcanology and Seismology, Quezon City, and University of Washington Press, Seattle,
pp 545–570
Self S, Zhao J-X, Holasek RE, Torres RC, King AJ (1996) The
atmospheric impact of the 1991 Mount Pinatubo eruption. In:
Newhall CG, Punongbayan RS (eds) Fire and mud: eruptions
and lahars of Mount Pinatubo, Philippines. Philippine Institute
of Volcanology and Seismology, Quezon City, and University
of Washington Press, Seattle, pp 1089–1115
Shaw PT, Chao SY (1994) Surface circulation in the South China
Sea. Deep-Sea Res 41:1663–1683
Sorem RK (1982) Volcanic ash clusters: tephra rafts and scavengers. J Volcanol Geotherm Res 13:63–71
Sparks RSJ, Bursik MI, Ablay GJ, Thomas RME, Carey SN (1992)
Sedimentation of tephra by volcanic plumes. Part 2: controls on
thickness and grain-size variations of tephra fall deposits. Bull
Volcanol 54:685–695
Stattegger K, Kuhnt W, Wong HK, and Shipboard Scientific Party
(1997) Cruise report, R/V Sonne SO-115, Kota Kinabalu—
Singapore: sequence stratigraphy, late Pleistocene-Holocene
sea level fluctuations and high-resolution record of the postPleistocene transgression on the Sunda Shelf. Ber Rep Geol
Palont Inst Univ Kiel 86:1–211
Syvitski JPM, LeBlanc KWG, Asprey KW (1991) Interlaboratory,
interinstrument calibration experiment. In: Syvitski JPM (ed)
Principles, methods and application of particle size analysis.
Cambridge University Press, Cambridge, New York, pp 174–
193
Teeter AM (2002) Sediment transport in wind-exposed shallow,
vegetated aquatic systems. PhD Diss, Department of Oceanography and Coastal Studies, Louisiana State University, Baton
Rouge, pp 1–225
UN Economic and Social Commission for Asia and the Pacific
(1992) ESCAP/WMO Typhoon Committee Annual Review
1991, pp 1–125
Varekamp JC, Luhr JF, Prestegaard KL (1984) The 1982 eruptions
of El Chich
n Volcano (Chiapas, Mexico): character of the
eruptions, ash-fall deposits and gas phase. J Volcanol Geotherm
Res 23:39–68
Wang Y (1999) Fallout und Sedimentation der Pinatubo-Tephra
1991 im Sdchinesischen Meer. Dissertation, Fachbereich
Geowissenschaften, Universitt Hamburg, pp 1–140
242
Wiesner MG, Wang Y (1996) Dispersal of the 1991 Pinatubo
tephra in the South China Sea. In: Newhall CG, Punongbayan
RS (eds) Fire and mud: eruptions and lahars of Mount Pinatubo,
Philippines. Philippine Institute of Volcanology and Seismology, Quezon City, and University of Washington Press, Seattle,
pp 537–543
Wiesner MG, Wang Y, Zheng L (1995) Fallout of volcanic ash to
the deep South China Sea induced by the 1991 eruption of
Mount Pinatubo. Geology 23:885–888
Wiesner MG, Kuhnt W, and Shipboard Scientific Party (1997)
Cruise report, R/V Sonne SO-114, Manila—Kota Kinabalu.
Institute of Biogeochemistry and Marine Chemistry, University
of Hamburg, IBMC Library Ref No IIA 942, pp 1–55
Wiesner MG, Kuhnt W, and Shipboard Scientific Party (1998)
Cruise report, R/V Sonne SO-132, Singapore—Manila. Insti-
tute of Biogeochemistry and Marine Chemistry, University of
Hamburg, IBMC Library Ref No IIA 943, pp 1–120
Wiesner MG, Stattegger K, Kuhnt W, and Shipboard Scientific
Party (1999) Cruise report, R/V Sonne SO-140, Singapore–Nha
Trang–Manila. Ber Rep Inst Geowiss Univ Kiel 7:1–157
Williams SN, Self S (1983) The October 1902 plinian eruption of
Santa Maria volcano, Guatemala. J Volcanol Geotherm Res
16:33–56
Yang Y, Fan S (1990) Research on volcanic sediments and origin of
volcanic substance in the South China Sea during Late
Quaternary (in Chinese with English abstract). Trop Oceanol
9:52–60