Contrib Mineral Petrol (2014) 168:1061
DOI 10.1007/s00410-014-1061-z
ORIGINAL PAPER
Pre‑ and syn‑eruptive degassing and crystallisation processes
of the 2010 and 2006 eruptions of Merapi volcano, Indonesia
Katie Preece · Ralf Gertisser · Jenni Barclay · Kim Berlo ·
Richard A. Herd · Edinburgh Ion Microprobe Facility
Received: 30 December 2013 / Accepted: 2 September 2014
© The Author(s) 2014. This article is published with open access at Springerlink.com
Abstract The 2010 eruption of Merapi (VEI 4) was the
volcano’s largest since 1872. In contrast to the prolonged
and effusive dome-forming eruptions typical of Merapi’s
recent activity, the 2010 eruption began explosively, before
a new dome was rapidly emplaced. This new dome was subsequently destroyed by explosions, generating pyroclastic
density currents (PDCs), predominantly consisting of dark
coloured, dense blocks of basaltic andesite dome lava. A
shift towards open-vent conditions in the later stages of the
eruption culminated in multiple explosions and the generation of PDCs with conspicuous grey scoria and white pumice
Communicated by O. Müntener.
Electronic supplementary material The online version of this
article (doi:10.1007/s00410-014-1061-z) contains supplementary
material, which is available to authorized users.
K. Preece (*) · J. Barclay · R. A. Herd
School of Environmental Sciences, University of East Anglia,
Norwich NR4 7TJ, UK
e-mail: K.Preece@uea.ac.uk
R. Gertisser
School of Physical and Geographical Sciences, Keele University,
Keele, Staffordshire ST5 5BG, UK
K. Berlo
Department of Earth Sciences, University of Oxford, South Parks
Road, Oxford OX1 3AN, UK
Present Address:
K. Berlo
Department of Earth & Planetary Sciences, McGill University,
Quebec H3A 2A7, Canada
Edinburgh Ion Microprobe Facility
University of Edinburgh, West Mains Road, Edinburgh EH9 3JW,
UK
clasts resulting from sub-plinian convective column collapse.
This paper presents geochemical data for melt inclusions and
their clinopyroxene hosts extracted from dense dome lava,
grey scoria and white pumice generated during the peak of
the 2010 eruption. These are compared with clinopyroxenehosted melt inclusions from scoriaceous dome fragments
from the prolonged dome-forming 2006 eruption, to elucidate any relationship between pre-eruptive degassing and
crystallisation processes and eruptive style. Secondary ion
mass spectrometry analysis of volatiles (H2O, CO2) and light
lithophile elements (Li, B, Be) is augmented by electron
microprobe analysis of major elements and volatiles (Cl, S,
F) in melt inclusions and groundmass glass. Geobarometric
analysis shows that the clinopyroxene phenocrysts crystallised at depths of up to 20 km, with the greatest calculated
depths associated with phenocrysts from the white pumice.
Based on their volatile contents, melt inclusions have reequilibrated during shallower storage and/or ascent, at depths
of ~0.6–9.7 km, where the Merapi magma system is interpreted to be highly interconnected and not formed of discrete
magma reservoirs. Melt inclusions enriched in Li show uniform “buffered” Cl concentrations, indicating the presence
of an exsolved brine phase. Boron-enriched inclusions also
support the presence of a brine phase, which helped to stabilise B in the melt. Calculations based on S concentrations
in melt inclusions and groundmass glass require a degassing
melt volume of 0.36 km3 in order to produce the mass of SO2
emitted during the 2010 eruption. This volume is approximately an order of magnitude higher than the erupted magma
(DRE) volume. The transition between the contrasting eruptive styles in 2010 and 2006 is linked to changes in magmatic flux and changes in degassing style, with the explosive
activity in 2010 driven by an influx of deep magma, which
overwhelmed the shallower magma system and ascended
rapidly, accompanied by closed-system degassing.
13
1061
Page 2 of 25
Keywords Merapi · 2010 eruption · 2006 eruption ·
Melt inclusions · Volatiles · Degassing
Introduction
Arc volcanoes may erupt explosively or effusively and transitions between eruptive styles are common. Transitions
between effusive dome-forming and explosive (sub) plinian
eruptions have been related to changes in various factors,
including magma composition, volatile content, degassing
regime, changes in magma supply, ascent rate and overpressure, as well as crystallisation during ascent (e.g. Jaupart and Allègre 1991; Woods and Koyaguchi 1994; Eichelberger 1995; Villemant and Boudon 1998; Martel et al.
1998; Melnik and Sparks 1999, 2005; Scandone et al. 2007;
Ruprecht and Bachmann 2010). Deep magmatic influx
into a reservoir has been identified as an eruption trigger
in many previous studies (e.g. Murphy et al. 2000; Ridolfi
et al. 2008; Suzuki et al. 2013). Magmatic influx may in
turn affect magma ascent dynamics, influencing the rate of
magma ascent and whether ascent is sustained or pulsatory
in its nature (Wolf and Eichelberger 1997; Scandone et al.
2007). Previous studies have shown that the style of degassing (open vs. closed) is pivotal in determining eruptive
behaviour. During open-system degassing, exsolved volatiles are separated and lost from the melt through various
pathways such as through the vent and/or the conduit walls,
leading to effusive eruptive activity (e.g. Eichelberger et al.
1986; Melnik and Sparks 1999; Villemant et al. 2008). In
closed-system degassing, the exsolved volatiles remain
within the system, tending to result in increased overpressure, vesicularity and capacity for explosive eruption (Wilson et al. 1980). During magma ascent, processes of degassing and crystallisation, and the consequent changes to
magma rheology also influence eruptive style via complex
feedback mechanisms (Sparks 1997; Melnik and Sparks
1999, 2005).
Silicate melt inclusions trapped in phenocrysts potentially retain evidence about the pre-eruptive magma that
may not be preserved elsewhere, providing critical information about the processes operating during magmatic
evolution. For example, melt inclusions can preserve the
composition and dissolved volatile concentrations of a
pre-eruptive melt, and have previously been used to shed
light on minimum pressures of crystallisation and to supply information about exsolved fluids present during crystallisation [see Lowenstern (1995, 2003) and Kent (2008)
for reviews of the subject]. However, the interpretation of
melt inclusion volatile concentrations is challenging, both
analytically and due to potential changes in the initial composition by formation of boundary layer conditions (e.g.
Baker 2008) or by post-entrapment processes, including
13
Contrib Mineral Petrol (2014) 168:1061
crystallisation and diffusive loss through the crystal lattice
or cracks (e.g. Lowenstern 1995).
At Merapi, previous volcanological and petrological
work suggests that the eruptive style of past dome-forming and larger sub-plinian eruptions has been governed
by slow versus fast magma ascent rate and open- versus
closed-system degassing (Gertisser 2001; Gertisser et al.
2011) and that shallow-level dynamics control the eruptive behaviour (Gauthier and Condomines 1999). In addition, crustal carbonate assimilation and the resulting CO2
liberation has also been invoked to play a significant role
in governing the explosivity of eruptions (Chadwick et al.
2007; Deegan et al. 2010; Troll et al. 2012, 2013; Borisova et al. 2013). Secondary ion mass spectrometry (SIMS)
volatile data (CO2, H2O, F, Cl, S) of Merapi clinopyroxene
and amphibole-hosted silicate melt inclusions in lava and
scoria from unknown eruptive origins have recently been
published by Nadeau et al. (2013). The authors suggest that
CO2, liberated by crustal carbonate assimilation fluxed the
melt, thereby promoting CO2 enrichment of the melt and
H2O degassing. In addition, the magmatic volatile phase
exsolved into a H2O–Cl–F-rich brine and CO2–S-rich
vapour.
Focussing on the cataclysmic eruption of Merapi in
2010 and the previous eruption in 2006, this paper presents the hitherto most comprehensive set of melt inclusion data, including the first SIMS data for the 2006 and
2010 eruption products. Comparing data between these
eruptions is crucial to understand the recent pre-eruptive
magmatic system of Merapi. Data include measurements
of volatiles (H2O, CO2, Cl, S, F), light lithophile trace elements (B, Li, Be) and major element concentrations in a
comprehensive suite of clinopyroxene-hosted melt inclusions, gathered from stratigraphically controlled samples of
various stages of the 2010 Merapi eruption. These include
dense dome clasts, as well as grey scoria and white pumice,
from the subsequent sub-plinian stage (Fig. 1). These data
are complemented by analysis of the host clinopyroxene
phenocrysts as well as groundmass glass. The diverse and
stratigraphically controlled sample set, produced by rapidly changing eruptive behaviour in 2010, is used to shed
light on the magmatic contribution to the shifts in eruptive
style during a single eruptive period at Merapi. In addition,
comparison of the various 2010 samples to those originating from scoriaceous dome fragments produced during the
peak of the dome-forming 2006 eruption, elucidates preeruptive processes during the two most recent and contrasting eruptions of Merapi. Results of this work are beneficial
for hazard analysis at Merapi and possibly other domeforming volcanoes worldwide, providing insights into factors contributing to changes in eruptive behaviour, which,
as the 2010 eruption of Merapi demonstrated, may occur
rapidly and with few imminent warning signs.
Page 3 of 25
Contrib Mineral Petrol (2014) 168:1061
Date
Stage
Activity
23 Nov 2010
08 Nov 2010
7
05 Nov 2010
Summary of eruptive activity on 05 November 2010
Time Stage
8
1061
Activity
Samples in this study
Dome growth ceased. Declining
ash venting and deflation of new
dome
02:1104:21
6
White pumice
Sub-Plinian convective column (M11-50; M11-55)
collapse, generating PDCs rich
Grey scoria
in pumice and scoria
(M11-75)
00:1301:57
5
Retrogressive
gravitational
collapse of old summit domes
generating PDCs
00:0200:13
4
Paroxsymal explosion and
collapse of Stage 3 dome
generating PDCs up to 16 km
from summit
Rapid dome growth (35 m3 s-1)
producing a ~ 1.5x106 m3 in <12
hrs on 06 Nov.
Ash venting and lava fountains
456
Increasing SO2 emissions
3
Rapid dome growth (25m3 s-1) to ~
5x106 m3
29 Oct, 30 Oct, 1 Nov: Recurrent
gravitational dome collapse and
explosions
2
Relative quiescence
1
Beginning of eruption
(Explosions generating PDCs)
Rapid increase in seismicity and
deformation. Alert level raised to IV
(on scale of I - IV)
Non-eruptive. Increasing unrest
including seismicity, deformation,
heat flux and CO2 and SO2
degassing.
29 Oct 2010
26 Oct 2010
25 Oct 2010
31 Oct 2009
Merapi summit
26 Oct 2010 deposits
05 Nov 2010 deposits:
Stage 4 deposits
Stage 6 deposits (scoria flow)
Stage 6 deposits (pumice flow)
Dense dome clasts
(M11-27)
Fig. 1 Eruptive timeline of the 2010 eruption, with stages based on
Komorowski et al. (2013) and expanded section describing the paroxysmal eruptive activity of 5 November 2010 and samples used in this
paper. Deposit map showing the extent of deposits emplaced on 26
October 2010, as well as during Stage 4 and Stage 6 (modified after
Komorowski et al. 2013)
Background
of crystallisation at depths throughout the crust prior to
both eruptions (Costa et al. 2013; Preece et al. 2013; Troll
et al. 2013; Preece 2014). Geophysical data also suggest
the presence of multiple magma storage regions at Merapi.
Tilt and GPS data indicate an average source depth for
magma storage at 8.5 ± 0.4 km below the summit (Beauducel and Cornet 1999), broadly consistent with the depth
of an aseismic zone observed at >5 km below the summit,
thought to represent the presence of melt (Ratdomopurbo
and Poupinet 2000). In addition, an aseismic zone located
at 1.5–2.5 km depth below the summit is interpreted to be
a shallow ephemeral storage region, where magma is temporarily stored as it migrates from the deeper reservoir(s)
before eruption (Ratdomopurbo and Poupinet 2000). Shallow storage regions have also been proposed based upon
Bouguer gravity anomaly data (Saepuloh et al. 2010). The
volume of this shallow magma storage region has been estimated at ~1.6–1.7 × 107 m3 based upon (210Pb) in Merapi
fumarolic gas and magma residence times as determined
from (210Pb/226Ra) disequilibria (Gauthier and Condomines
1999; Le Cloarec and Gauthier 2003). Gas emissions at
Merapi magmatic system and volatiles
The plumbing system of Merapi is thought to consist of
multiple magma storage and crystallisation regions, ranging over almost the entire thickness of the crust (e.g. Beauducel and Cornet 1999; Ratdomopurbo and Poupinet 2000;
Gertisser 2001; Chadwick et al. 2007, 2013; Costa et al.
2013). Evidence for this comes from both petrological and
geophysical studies. Geobarometry of magmatic inclusions
indicates that crystallisation at Merapi occurs over a wide
range of depths (~2–45 km), with the majority occurring
at mid- to lower-crustal levels (12–18 km) (Chadwick et al.
2013). Mineral equilibria in lavas and pyroclastic rocks
also reveal a major magma storage region at mid- to lowercrustal levels (14–19 km) (Gertisser 2001), corroborated by
estimates of amphibole crystallisation depths (Preece et al.
2011; Nadeau et al. 2013). Petrological data, which elucidate the magma storage conditions prior to the 2006 and
2010 eruptions, demonstrate that there were multiple zones
13
1061
Page 4 of 25
Merapi are H2O-rich (~84–95 mol%), with lesser amounts
of CO2 (<10 mol%) and minor amounts of SO2 and H2S
(Le Guern et al. 1982; Zimmer and Erzinger 2003). Previous work has concluded that effusive volcanism at Merapi
is accompanied by open-system degassing, with most of
the degassing occurring within the conduit during ascent
or during magma residence in a shallow magma chamber
below the summit (Le Pennec et al. 2001; Le Cloarec and
Gauthier 2003). Evidence of a deeper gas and magma supply to this shallower system has been reported, with inputs
of deep, undegassed magma into the shallower, degassing
reservoir (Gauthier and Condomines 1999; Le Cloarec
and Gauthier 2003; Costa et al. 2013). For example, volcanic gases are enriched in S compared to modelled volatile phases, attributed to a deeper, reduced, mafic magma
supplying S to the shallow magmatic system (Nadeau et al.
2010, 2013). It is estimated that the magma degassing rate
is 40 times greater than the lava extrusion rate of the past
100 years would suggest (Allard et al. 2011). This suggests
that the magma storage region is large enough to accommodate substantial amounts of unerupted, degassing magma,
with continuous gas percolation through the magma system
(Allard et al. 2011). At Merapi, CO2 is thought to originate both from a mantle source and from crustal carbonate
assimilation. Isotopic data indicate a mantle-derived origin
for most volatiles at Merapi, with additional CO2 derived
from crustal contamination (Allard et al. 2011). The upper
crustal rocks around Merapi are comprised of a ~10-kmthick sequence of Cretaceous to Tertiary limestones, marls
and volcaniclastic rocks (van Bemmelen 1949; Hamilton
1979; Smyth et al. 2005). Calc-silicate xenoliths are commonly found within Merapi lavas, providing evidence for
the interaction of magma with crustal carbonate material
(e.g. Clocchiatti et al. 1982; Camus et al. 2000; Gertisser and Keller 2003; Chadwick et al. 2007; Deegan et al.
2010; Troll et al. 2012, 2013). Magma–carbonate interaction liberates CO2 through decarbonation reactions of crustal carbonates to the diopside and wollastonite assemblages
observed in the xenoliths, adding to the magmatic volatile
budget with the potential to sustain and intensify eruptions
at Merapi (Deegan et al. 2010; Troll et al. 2012, 2013).
Contrib Mineral Petrol (2014) 168:1061
activity and extrusion of a new dome after 5 November
(Surono et al. 2012; Komorowski et al. 2013; Pallister et al.
2013) (Fig. 1). The 2010 eruption chronology and deposits have previously been documented in detail (e.g. Surono
et al. 2012; Pallister et al. 2013; Charbonnier et al. 2013;
Komorowski et al. 2013). Komorowski et al. (2013) recognised eight stages of the 2010 eruption, which will be
referred to throughout this paper (Fig. 1): Stage 1: unrest
and magmatic intrusion (31 October 2009–26 October
2010); Stage 2: initial explosions (26 October 2010); Stage
3: recurrent rapid dome growth and destruction (29 October–4 November 2010); Stage 4: paroxysmal dome explosions and collapse (5 November 2010); Stage 5: retrogressive summit collapse (5 November); Stage 6: sub-plinian
fountain collapse (5 November 2010); Stage 7: rapid dome
growth with alternating effusive and explosive activity (5–8
November 2010); Stage 8: declining ash venting and degassing (8–23 November 2010).
In contrast to 2010, previous episodes of volcanic activity at Merapi over the last century were frequently characterised by prolonged dome extrusion and subsequent gravitational collapse to produce block-and-ash flows (BAFs)
or “Merapi-type nuées ardentes” (e.g. Andreastuti et al.
2000; Newhall et al. 2000; Voight et al. 2000; Gertisser
et al. 2012a; Surono et al. 2012). The previous eruption in
2006 is a well-characterised extrusive, dome-forming eruption at Merapi (Charbonnier and Gertisser 2008; Gertisser
et al. 2012b; Preece et al. 2013; Ratdomopurbo et al. 2013).
Although the 2006 eruption displayed typical Merapi
dome-forming activity, peak dome extrusion rates reached
3.3 m3 s−1 (Ratdomopurbo et al. 2013), which is high compared to other recent Merapi eruptions. For example, it is
an order of magnitude higher than peak dome extrusion
rates in 1994 (0.32 m3 s−1) (Ratdomopurbo 1995; Hammer et al. 2000). Throughout the >3 month long eruption,
BAFs were generated by almost daily gravitational dome
collapse. The peak of activity on 14 June 2006 consisted of
multiple phases of dome collapse (Charbonnier and Gertisser 2008; Lube et al. 2011; Gertisser et al. 2012b) producing BAFs that travelled up to 7 km from the summit (Charbonnier and Gertisser 2008).
The 2010 and 2006 eruptions of Merapi
Methodology
In 2010, Merapi volcano had its largest eruption (VEI 4)
since 1872 (e.g. Surono et al. 2012). In contrast to recent
prolonged and effusive dome-forming eruptions at Merapi,
such as the previous eruption in 2006 (Charbonnier and
Gertisser 2008; Preece et al. 2013; Ratdomopurbo et al.
2013), the 2010 eruption began explosively and a new
lava dome grew in the newly formed crater prior to explosive destruction of this dome during the peak of the eruption on 5 November 2010, followed by further explosive
13
Samples and sample preparation
Various samples from the different stages of the 2010 eruption were analysed, in particular, samples of dense dome
material emplaced on 5 November (Stage 4 of Komorowski
et al. 2013), as well as grey scoria and white pumice clasts
from PDC deposits emplaced by subsequent convective
fountain collapse (Stage 6 of Komorowski et al. 2013).
Contrib Mineral Petrol (2014) 168:1061
Clinopyroxene-hosted silicate melt inclusions were analysed in all samples, and groundmass glass was analysed in
all but the dense dome samples, as the groundmass was too
crystalline to allow for accurate glass analysis. For comparison, clinopyroxene-hosted melt inclusions and groundmass glass were also analysed from scoriaceous dome fragments from the block-and-ash flows emplaced at the peak
of the 2006 eruption on 14 June 2006 (Lobe 1 of Charbonnier and Gertisser 2008; Preece et al. 2013).
Grain mounts were prepared by crushing rock samples
for a few seconds in an agate mill, before sieving to different size fractions to separate larger pyroxene phenocrysts.
Several hundred clinopyroxene crystals were picked by
hand under a binocular microscope from >1-mm and
>500-μm-sieve fractions and mounted into low-volatility
Struers EpoFix epoxy resin. Each grain mount was polished by hand with aluminium polishing solution down to
0.3 μm, until melt inclusions were exposed. Aluminium
solution was used for polishing in order to avoid carbon
and boron contamination, which may occur with diamond
solutions. Inclusions were only studied after polishing,
with the proviso that some petrographic information may
have been lost during polishing. However, this method has
the advantage that many melt inclusions can be analysed
quickly, giving an extensive overview of the melt inclusion
population (cf. Humphreys et al. 2008).
Analytical methods
Melt inclusions in gold-coated samples were analysed by
secondary ion mass spectrometry (SIMS) for isotopes of
volatiles (1H+ and 12C+) and of light lithophile elements
(7Li+, 9Be+, 11B+) using the Cameca ims-4f ion microprobe
at the NERC Ion Microprobe Facility at the University of
Edinburgh (UK). A subset of these inclusions were subsequently analysed for 12C+ at high resolution in order to
minimise the interference of 24Mg2+ on 12C+. All 2010 and
2006 CO2 data in this paper are based on the high-resolution 12C+ results. Analyses were performed using a primary
16 −
O beam and positive secondary ion beam with an accelerating voltage of 4.5 kV. Energy filtering with a 75 ± 20 V
offset, or a 50 V offset for high-resolution 12C+ measurements, was used with the purpose of minimising the transmission of unwanted molecular species. Surface contamination was eliminated by performing a 50-μm-diameter, 10
nA raster of the sample surface for 3 min prior to analysis. Results are based on the last 10 cycles, with the first
5 disregarded, or, for high-resolution 12C+ measurements,
based on the last 8 with the first 4 disregarded, in order to
abate effects of potential surface contamination and allow
time for beam stabilisation. For all analyses, 30Si+ was used
as an internal standard and corrected with SiO2 contents as
measured by electron microprobe analysis. H2O and CO2
Page 5 of 25
1061
were measured using 1H+ and 12C+, respectively, calibrated
using dacitic to rhyolitic glass standards with known H2O
(up to 4.32 wt%) and CO2 (up to 10,380 ppm) concentrations and using working curves as described in Blundy
and Cashman (2008). Standards that were used include
the following experimental glasses: Sisson 51, Sisson 56,
Sisson 59, RB480, Lipari, STHS, BF147 and MC84r. CO2
backgrounds were measured to be <5 ppm using CO2-free
standards. Detection limits were typically ~10 ppm for CO2
and ~100 ppm for H2O. Analytical uncertainties based on
measurements of the glass standards are <10 % (relative)
for H2O and ~10–15 % (relative) for CO2.
Selected melt inclusions in the same polished epoxy
grain mounts were also analysed for H2O and CO2 using
attenuated total reflectance micro-Fourier transform infrared spectroscopy (ATR micro-FTIR) at the USGS, Menlo
Park, California. A Ge accessory crystal was used to measure evanescent wave absorption at 3,450 cm−1 (representing total H2O), at 2,350 cm−1 (representing molecular CO2)
and at 1,400–1,500 cm−1 (representing carbonate) using
the methods of Lowenstern and Pitcher (2013).
Each studied melt inclusion and host crystal was viewed
with the SEM and analysed by electron probe only after
the SIMS analysis in order to avoid possible contamination of C with the carbon coating. Major elements as well
as Cl, F and S in the melt inclusions and groundmass glass
were measured using Cameca SX100 electron microprobes
at The Open University and the University of Cambridge.
Glass was analysed using a defocussed beam diameter
of 5–10 μm, an accelerating voltage of 15–20 kV and a
4–20 nA beam current for major elements and a 10–20 nA
beam current for volatiles. Volatiles were analysed with
extended peak counting times. Na was always measured
first to minimise migration effects, and in-house natural
mineral standards were used for calibration. Major elements were measured in the clinopyroxene host crystals,
near to the site of each melt inclusion, as well as in other
clinopyroxene phenocrysts in 2010 and 2006 samples.
Although there is little zonation in the analysed clinopyroxene phenocrysts, if zoning was present then pyroxene analyses were made in the same zones in which melt inclusions
were situated. Pyroxenes were analysed using a 1–5 μm
beam diameter, a 15–20 kV accelerating voltage and a 15–
20 nA beam current. Detection limits were ~100–250 ppm
for major elements, ~100 ppm for S and Cl, and ~200 ppm
for F. Analytical uncertainties for major elements, based on
repeat analyses of natural mineral and rhyolitic glass standards, were in the order of 1–3 % (relative). Uncertainties
of volatile element determinations are estimated at <5 %
(relative) for S and Cl, and <15 % (relative) for F.
Back-scattered electron (BSE) images of all melt inclusions were acquired with a JEOL JSM 5900 LV SEM at the
University of East Anglia, using an accelerating voltage of
13
Page 6 of 25
8
20
White pumice
18
FeO* (wt. %)
16
14
6
4
2
12
a
10
b
0
9
3.0
8
2.5
7
CaO (wt. %)
2.0
1.5
1.0
6
5
4
3
2
c
0.0
0
7.5
7.5
6.5
6.5
5.5
4.5
3.5
50
5.5
4.5
3.5
2.5
e
2.5
d
1
K2O (wt. %)
Na2O (wt. %)
4
6
2006 scoria
Groundmass glass
2006
2010
Whole rock
2010
2006
6
0.5
f
1.5
55
60
65
70
SiO2 (wt. % )
20 kV and a working distance of 9 mm. All host crystals
were imaged using an accelerating voltage of 20 kV and
a working distance of 31–37 mm. Back-scattered electron
images were then analysed in order to disregard analyses
that were on cracks or occasional inclusions which contained any daughter crystals.
Whole-rock compositions were obtained from interior
portions of fresh samples, which were washed in MilliQ in a sonic bath for 15 min, dried overnight at >100 °C
and powdered in a tungsten carbide mill. Major elements
were analysed by X-ray fluorescence (XRF) using a Bruker
AXS S4 Pioneer at the University of East Anglia. Loss on
ignition (LOI) was carried out by heating in a furnace at
1,050 °C for 4 h.
All melt inclusion major element data were corrected for
the compositional effects of post-entrapment crystallisation
(PEC) of clinopyroxene at the melt inclusion–host interface. The corrections were performed by first calculating
the composition of the host clinopyroxene that should be
in equilibrium with the melt inclusion using an appropriate
13
Grey scoria
2010 Stage
Melt inclusions
Dense dome
MgO (wt. %)
Fig. 2 Major element variation
diagrams showing the composition of melt inclusions from
2010 dense dome material,
white pumice and grey scoria,
and 2006 scoriaceous dome
clasts, as well as groundmass
glass and whole-rock compositions from 2006 and 2010
samples. All melt inclusion
compositions are corrected
for PEC. All melt inclusion, groundmass glass and
whole-rock measurements are
normalised to 100 wt% on a
volatile-free basis. FeO* = all
iron reported as FeO
Contrib Mineral Petrol (2014) 168:1061
Al2O3 (wt. %)
1061
75
80
50
55
60
65
70
SiO2 (wt. % )
75
80
clinopyroxene-melt equilibrium model (Nielsen and Drake
1979). The equilibrium clinopyroxene was then added
in 0.1 wt% increments to the measured melt inclusion
composition until the equilibrium clinopyroxene composition becomes identical to that of the host, using the
reverse fractional crystallisation modelling function of the
Petrolog3 software (Danyushevsky and Plechov 2011). The
calculated degree of PEC is <10 % but typically <5 %.
Results
Major element geochemistry
Major element compositional variations in whole rocks,
melt inclusions and groundmass glass from 2010 and 2006
samples are shown in Fig. 2. Whole-rock compositions of
both the 2010 and 2006 eruptive products are high-K basaltic andesite and show a similar compositional range (Preece
et al. 2013; Preece 2014). SiO2 concentrations are between
Page 7 of 25
Contrib Mineral Petrol (2014) 168:1061
54.1 and 55.7 wt% for 2010 juvenile samples and between
55.2 and 56.1 wt% for the 2006 products. A comparison
of different 2010 lithologies reveals that the compositional
range of the dense dome material erupted on 5 November extends to slightly less evolved compositions (54.1–
55.6 wt% SiO2) than the white pumice (55.5–55.7 wt%
SiO2) and the grey scoria (55.1–55.7 wt% SiO2), although
the differences are marginal. There is a compositional gap
of >5 wt% SiO2 between the most evolved whole-rock
sample and the least evolved melt inclusion (Fig. 2). The
melt inclusions are mainly dacitic to rhyolitic in composition, with 63.1–72.4 wt% SiO2, with only two classed
as andesitic (61.7 and 62.3 wt% SiO2) when corrected
for PEC and normalised to 100 % on a volatile-free basis
(Table 1 and Electronic Supplementary Data). Both 2010
and 2006 melt inclusions cover a similar compositional
range. Melt inclusions from each 2010 lithology generally
span the entire observed compositional range, although
the most evolved melt inclusions (>70 wt% SiO2) come
exclusively from white pumice and grey scoria in the 2010
deposits. In comparison, 2006 clinopyroxene-hosted melt
inclusions presented in Nadeau et al. (2013) have a more
restricted compositional range between 64.2 and 69.8 wt%
SiO2, and in addition, one amphibole-hosted inclusion is
reported, which has 64.1 wt% SiO2. In 1998 dome samples,
Schwarzkopf et al. (2001) reported plagioclase-hosted melt
inclusions with 56.4–62.6 wt% SiO2 and Gertisser (2001)
reported average values between 63.2 and 66.1 wt% SiO2
for melt inclusions in different samples of the Merapi highK series rocks. Groundmass glass compositions represent
the most evolved compositions (66.8–76.4 wt% SiO2).
Overall trends of Al2O3, CaO, FeO* (total iron calculated
as FeO) and MgO correlate negatively with SiO2, and K2O
correlates positively with SiO2 (Fig. 2). Both TiO2 and
Na2O trends are inflexed, with TiO2 concentrations decreasing until ~67 wt% SiO2 before groundmass TiO2 concentrations increase with increasing SiO2, and Na2O correlating
positively until ~67 wt% SiO2, where Na2O concentrations
begin to decrease.
Volatile concentrations
H2O and CO2
SIMS analysis shows that the highest H2O concentrations
occur in melt inclusions from the 2010 grey scoria (up to
3.94 wt%) and white pumice (up to 3.91 wt%) (Table 1).
Melt inclusions from 2010 dense dome clasts are generally
more degassed, although the highest measured H2O in the
dome samples is 3.62 wt%. In comparison, melt inclusions
from the scoriaceous 2006 dome fragments contain up to
3.73 wt% H2O (Table 1) [see Electronic Supplementary
Data for full data set].
1061
CO2 concentrations in melt inclusions are generally
<200 ppm; although at high H2O concentrations, melt
inclusions in the white pumice have increased CO2 concentrations up to 695 ppm. Several melt inclusions appear
to have elevated CO2 concentrations (up to ~3,000 ppm)
at medium (~2 wt%) H2O contents, a feature particularly
associated with melt inclusions from the white pumice. The
maximum CO2 measured in the 2010 grey scoria and dense
dome melt inclusions is 146 ppm and 12 ppm, respectively,
and in melt inclusions from the 2006 dome scoria, the maximum is 158 ppm CO2 (Table 1), although one inclusion is
displaced towards higher CO2 concentrations (~2,400 ppm)
at <1 wt% H2O.
A selected group of 2010 melt inclusions were also
analysed for H2O and CO2 using ATR micro-FTIR. These
inclusions were primarily targeted for ATR micro-FTIR
analysis because results from the preceding SIMS analysis
suggested that these inclusions were enriched in CO2 (up to
~3,000 ppm) at medium (~2 wt%) H2O contents, relative
to the degassing trend displayed by other analysed inclusions (Fig. 3). ATR micro-FTIR results in this study reveal
that although measured H2O concentrations are similar to
the ones acquired by SIMS, the concentration of CO2 in the
inclusions was below the detection limit (200 ppm).
Comparison of the melt inclusions analysed in this
study with volatile data published by Nadeau et al. (2013)
from melt inclusions in clinopyroxene and amphibole phenocrysts and amphibole megacrysts from unknown Merapi
eruptions reveal some similarities. Apparent enrichment to
comparable elevated levels of CO2 occurs at a similar H2O
content. However, the highest H2O recorded by Nadeau
et al. (2013) is only ~2 wt%, even in amphibole megacrysthosted melt inclusions, which is lower than maximum H2O
concentrations recorded by this study.
F, S, Cl
Samples from 2010 show differences in F concentration,
depending upon lithology (Fig. 4). Groundmass glass in
white pumice samples contains 632–2,199 ppm F, with the
peak at between 1,000 and 1,250 ppm, whereas the groundmass glass in the grey scoria encompasses a larger range of
F (180–2,637 ppm), with a peak at relatively higher concentrations of 2,000–2,500 ppm (Fig. 4). Fluorine concentrations in 2010 melt inclusions also vary with lithology.
Those trapped in clinopyroxene phenocrysts from dense
dome clasts and those from grey scoria have a similar range
of F, containing up to 1,400 ppm and 1,480 ppm, respectively. White pumice-derived inclusions contain concentrations of F up to 2,390 ppm (Fig. 4). Melt inclusions from
2006 contain up to 2,050 ppm F, with the groundmass glass
containing up to 2,349 ppm. Fluorine concentrations in
the 2010 and 2006 melt inclusions and groundmass glass
13
13
Stage
4
4
4
4
4
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
Type
DD
DD
DD
DD
DD
GS
GS
GS
GS
GS
GS
GS
GS
WP
WP
WP
WP
WP
WP
WP
WP
WP
WP
WP
WP
WP
WP
DS
DS
DS
DS
DS
DS
DS
Sample
M11-27-5
M11-27-5
M11-27-5
M11-27-5
M11-27-5
M11-75
M11-75
M11-75
M11-75
M11-75
M11-75
M11-75
M11-75
M11-75
M11-55
M11-55
M11-55
M11-50
M11-50
M11-50
M11-50
M11-50
M11-50
M11-50
M11-50
M11-50
M11-50
M07-53
M07-53
M07-53
M07-53
M07-53
M07-53
M07-53
MI
x34-18
x34-19
x9-93
x2-103
x4-108
x5-44
x5-45
x5-46
x11-49
x4-50
x13-51
x30-53
x41-56
x25-5
x25-6
x52-7
x36-8
x11-59
x25-62
x9-64
x4-74
x27-77
x29-78
x29-79
x54-80
x71-81
x80-83
x79-25
x85-26
x79-27
x70-28
x48-86
x48-87
x6-88
SiO2
68.78
67.52
67.03
63.98
63.58
67.30
67.85
69.15
67.87
64.73
67.42
66.06
62.66
70.55
68.80
66.81
70.55
65.50
69.47
65.73
67.62
66.92
67.90
67.71
64.27
68.07
67.32
68.19
66.46
64.11
67.79
65.69
68.12
67.87
TiO2
0.34
0.44
0.50
0.35
0.42
0.38
0.40
0.50
0.48
0.42
0.43
0.50
0.54
0.29
0.29
0.44
0.36
0.38
0.33
0.45
0.37
0.36
0.36
0.39
0.28
0.36
0.37
0.39
0.40
0.78
0.42
0.41
0.51
0.33
Al2O3
16.81
15.97
16.73
14.91
14.91
16.04
16.82
16.60
16.25
15.89
16.34
16.48
16.45
14.14
14.41
15.34
14.21
16.83
13.72
15.86
16.00
15.81
15.34
16.09
15.24
15.47
16.22
16.24
16.01
16.23
14.95
16.35
16.20
16.28
FeO*
0.83
2.67
2.40
2.87
3.05
1.90
1.79
1.76
2.41
1.93
2.62
2.25
2.75
2.12
2.43
2.29
2.22
2.80
2.00
2.57
2.60
2.56
2.56
2.45
2.67
2.84
2.59
1.90
2.54
2.83
2.96
2.52
2.69
2.60
MnO
0.07
0.13
0.06
0.14
0.12
0.09
0.09
0.08
0.07
0.06
0.12
0.09
0.11
0.07
0.12
0.12
0.10
0.11
0.08
0.12
0.12
0.13
0.11
0.07
0.11
0.12
0.09
0.08
0.13
0.06
0.12
0.11
0.15
0.11
MgO
0.20
0.29
0.09
0.77
0.51
0.25
0.16
0.23
0.24
0.54
0.18
0.36
0.40
0.28
0.58
0.37
0.30
0.53
0.25
0.57
0.43
0.60
0.51
0.41
0.97
0.90
0.44
0.30
0.15
0.74
0.67
0.34
0.33
0.35
CaO
1.92
1.22
0.57
1.81
1.81
1.30
1.52
1.38
1.22
1.61
1.05
1.55
1.90
0.76
1.31
1.03
0.88
1.59
0.75
1.41
1.35
1.93
1.34
1.29
2.16
2.20
1.44
1.55
1.33
1.79
1.52
1.15
1.18
0.94
Na2O
4.87
5.13
6.52
5.42
4.08
4.79
4.98
5.40
6.04
4.70
6.22
4.62
4.32
4.26
4.42
4.85
4.54
5.26
3.91
4.36
4.80
4.65
4.92
5.06
5.00
4.75
5.00
4.98
6.89
6.70
4.66
5.07
4.73
5.02
K2O
5.12
6.31
5.82
6.27
6.33
5.22
4.93
4.90
5.29
5.13
5.18
5.46
4.63
5.84
6.00
5.82
5.65
5.57
6.09
5.32
5.63
4.69
5.44
5.61
5.26
5.43
5.56
4.64
5.40
6.00
5.86
5.97
6.17
6.33
P2O5
0.16
0.13
0.13
0.12
0.18
0.12
0.14
0.12
0.11
0.17
0.16
0.14
0.25
0.03
0.07
0.08
0.05
0.15
0.08
0.12
0.12
0.15
0.10
0.13
0.14
0.14
0.11
0.14
0.11
0.10
0.13
0.17
0.14
0.16
Total
99.10
99.81
99.85
96.64
95.32
97.39
98.68
100.12
99.98
95.18
99.72
97.51
94.01
98.34
98.43
97.15
98.86
98.72
96.68
96.51
99.04
97.80
98.58
99.21
96.10
100.28
99.14
98.41
99.42
99.34
99.08
97.78
100.22
99.99
80
50
90
110
160
140
135
95
85
20
90
275
245
95
285
95
75
10
85
S
210
105
110
40
10
180
445
90
60
315
45
205
180
45
970
1160
560
420
1060
1110
930
2390
830
810
710
370
570
260
1740
420
870
2050
1180
330
1400
1520
1140
800
F
760
450
1210
670
510
750
140
1480
Cl
3150
2690
2880
2770
2610
2860
2770
2620
2910
3180
2710
3470
2760
2440
3810
3850
2820
2670
5130
3220
2860
2770
3020
2900
2470
2660
2550
3050
2960
2910
2930
2780
3720
2640
3.0
5.2
2.9
0.1
3.1
5.3
5.1
4.6
PEC
0.1
6.0
17.0
0.3
4.7
6.4
9.7
5.5
6.9
3.4
7.5
6.1
7.7
4.4
3.1
4.2
3.9
7.4
4.8
3.7
4.7
1.5
2.5
3.4
2.2
Li
28
43
68
54
39
25
25
32
30
27
34
27
26
29
31
34
31
34
32
29
30
23
30
33
26
31
31
28
47
63
33
34
20
35
B
45
44
41
44
48
47
48
41
47
48
45
49
42
71
70
58
67
42
84
40
47
47
50
48
45
47
47
44
46
47
51
42
48
47
Be
1
2
1
1
2
2
2
1
2
1
2
1
1
2
2
2
2
1
3
1
2
2
2
2
2
2
1
1
2
1
2
1
2
2
H2O
0.78
0.40
1.13
0.77
0.22
3.39
2.79
3.94
1.08
3.83
1.27
3.66
3.08
1.83
1.62
2.04
2.37
2.11
1.19
3.79
1.92
3.28
2.46
2.31
1.91
1.97
1.88
3.33
1.13
1.93
0.65
0.83
0.42
0.82
7
9
1
51
31
15
193
158
3
534
146
97
109
146
18
65
86
5
CO2
Page 8 of 25
Stages 4 and 6 of the 2010 eruption refer to the chronology of Komorowski et al. (2013). Lithology types are as follows: dense dome (DD), grey scoria (GS), white pumice (WP),
scoriaceous dome clasts (DS). MI denotes melt inclusion number. PEC denotes degree of post-entrapment crystallisation corrected for (%). All major elements (in wt%) determined
by electron microprobe, FeO* denotes total iron reported as FeO, all values shown are uncorrected for PEC. Volatiles (S, F, Cl) in ppm also measured by electron microprobe. H2O
(in wt%), CO2, Li, B, Be (in ppm) measured by ion microprobe
Eruption
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
2006
2006
2006
2006
2006
2006
2006
Table 1 Representative analyses of melt inclusion compositions from the 2010 and 2006 eruptions of Merapi
1061
Contrib Mineral Petrol (2014) 168:1061
Page 9 of 25
Contrib Mineral Petrol (2014) 168:1061
200
100
MPa
MPa
400
10
m
ol
%
H
2O
500
0
M
Pa
200
ol%
50 m
Pa
600
300
MPa
M
50
%
m
ol
25
MPa
CO2 (ppm)
ol%
10
m
2 mol%
CO2 (ppm)
300
800
0
400
MPa
0
Pa
500
20
M
700
1200
400
0
1600
30
800
2400
25 m
ol%
900
2800
2000
1000
10 mol%
b
3200
2 mol%
a
1061
100
75 mol%
0
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
H2O (wt. %)
H2O (wt. %)
2010 Stage
Dense dome
4
Grey scoria
6
White pumice
6
2006 scoria
Trayem tephra
Breadcrust bomb
Amph. ph.-hosted (Nadeau et al. 2013)
Amph. mgxt.-hosted (Nadeau et al. 2013)
Cpx. ph.-hosted (Nadeau et al. 2013)
Model starting composition
Fig. 3 Melt inclusion H2O and CO2 concentrations from SIMS
analysis. a H2O versus CO2 measured with high-resolution SIMS.
All measured melt inclusions including the “high-CO2” inclusions
are shown, some of which were subsequently re-analysed with ATR
micro-FTIR. As a comparison to the 2010 and 2006 melt inclusions,
clinopyroxene-hosted melt inclusions from two older deposits are
also shown, namely a young, widespread basaltic breadcrust bombrich PDC deposit on Merapi’s southern flank (Newhall et al. 2000;
Gertisser et al. 2012a), tentatively ascribed to the VEI 4 1872 eruption, the last VEI 4 eruption prior to 2010 (Newhall et al. 2000), and
the pumiceous “Trayem tephra” produced during a prehistoric VEI
4 sub-plinian eruption (Gertisser 2001; Gertisser et al. 2012a). Note
that the breadcrust bomb and “Trayem tephra” CO2 values were
obtained by low-resolution measurements. In addition, previously
published data for amphibole phenocryst (Amph. ph.) amphibole
megacryst (Amph. mgxt.) and clinopyroxene phenocryst (Cpx. ph.)
hosted melt inclusions of unknown origins is also shown (Nadeau
et al. 2013). b Enlarged view of measured melt inclusions excluding the “high-CO2” set. All isobars (fine solid lines) and vapour isopleths (fine dashed lines) calculated with VolatileCalc (Newman and
Lowenstern 2002). Closed-system decompression degassing pathway
(bold solid line) with 1 % exsolved vapour (bold solid line) starting
at 1,000 ppm CO2 and 4.0 wt% H2O at 1,050 °C, and open-system
degassing pathway starting at the same conditions (bold dashed line)
calculated with VolatileCalc
are similar, although they extend to both higher and lower
concentrations compared to the 1994 Merapi dome-forming eruption (Gertisser 2001). In comparison, plagioclasehosted inclusions from the 1998 eruption appear to contain
less F, up to 700 ppm (Schwarzkopf et al. 2001), although
this may be an artefact of the small sample size reported.
Sulphur is present in low concentrations in all groundmass glass, with <100 ppm (detection limit) in 2010 glass
and up to 122 ppm in 2006 glass. Melt inclusions are generally more enriched in S than their respective groundmass, ranging up to 535 ppm S in 2010 products and up
to 345 ppm in inclusions from the 2006 eruption (Fig. 4).
Sulphur concentrations of the 2010 and 2006 melt inclusions and groundmass glass are similar to reported concentrations from the 1994 eruption (Gertisser 2001). A similar
range of 200–450 ppm S was also reported for plagioclasehosted melt inclusions from the 1998 eruption (Schwarzkopf et al. 2001).
Chlorine is more enriched in the melt inclusions than in
groundmass glass in both 2010 and 2006 samples (Fig. 4).
In melt inclusions, the Cl concentration ranges from 2,060
to 5,130 ppm in 2010 eruptive products, with concentrations most frequently between 2,500 and 3,000 ppm in all
sample types. The highest Cl concentrations are found in
melt inclusions hosted in clinopyroxene from white pumice samples. Chlorine concentrations in melt inclusions
from the 2006 samples show a comparatively narrow range
from 2,330 to 3,720 ppm. The groundmass glass concentrations of Cl range from ~900 ppm up to 3,550 ppm in samples from the 2010 eruption and are between ~1,000 and
2,920 ppm in those from 2006. The Cl concentrations in
the 2010 and 2006 melt inclusions and groundmass glass
are similar to those reported for the 1994 dome-forming
eruption (Gertisser 2001), although they extend to both
higher and lower values. In comparison, plagioclase-hosted
inclusions from recent eruptions contain a similar range of
13
Page 10 of 25
3000
2500
600
a
500
2000
S (ppm)
1500
1000
500
Cl (ppm)
5000
400
300
200
100
0
6000
b
0
3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
K2O (wt. %)
c
4000
Melt inclusions
3000
Dense dome
2000
White pumice
Grey scoria
2006 scoria
1000
2010 Stage
Fig. 4 Volatile contents in melt
inclusions and groundmass
glass plotted against differentiation indicator K2O wt%. a F
versus K2O, b S versus K2O, c
Cl versus K2O. Detection limits
are displayed as dashed lines
Contrib Mineral Petrol (2014) 168:1061
F (ppm)
1061
4
Groundmass
glass
6
Grey scoria
6
White pumice
2006 scoria
0
3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
K2O (wt. %)
between 3,010 and 3,320 pm Cl (Gertisser 2001), although
Schwarzkopf et al. (2001) report high concentrations of up
to 7,000 pm Cl in plagioclase-hosted inclusions from the
1998 eruption.
Light lithophile elements (B, Be and Li)
Concentrations of light lithophile elements, especially those
of B and Li, indicate differences within the melt inclusion
population. Beryllium concentrations are uniform, with
all measured melt inclusions containing 1–2 ppm, while
Li concentrations vary from 20–68 ppm. When plotted
against H2O and Cl (Fig. 5), divergent L-shaped trends
show enrichment of Li (>~35 ppm) in melt inclusions
from the 2010 dense dome material and the 2006 scoria,
which also show low H2O concentrations and ~2,700 ppm
Cl. There is no obvious correlation between Li and SiO2 or
K2O (Fig. 5). The overall range in B concentration is 35–
109 ppm, with enrichment (>~55 ppm) in melt inclusions
from white pumice that contain intermediate H2O concentrations (~1–2.5 wt%) and Cl up to ~5,000 ppm (Fig. 5).
Boron-enriched inclusions show a positive correlation with
SiO2 and K2O, but when plotted against Li, the data show a
divergent L-shaped trend (Fig. 5).
Clinopyroxene compositions
The major element composition of clinopyroxene host
crystals was measured alongside that of other pyroxene
13
phenocrysts within the same eruptive products to establish whether there are systematic compositional variations
between different eruptive products and whether host crystals are representative of the clinopyroxene population. Of
the 2010 lithologies, clinopyroxene from the dome samples have the widest compositional range (Wo40–50En33–
45Fs12–24) with the presence of crystals containing higher
proportions of Wo and Fs, and lower En components compared to the grey scoria (Wo42–47En38–44Fs12–15) and the
white pumice (Wo41–48En36–45Fs13–16) (Fig. 6). Analysed
clinopyroxene phenocrysts from the 2006 dome scoria are
similar in composition (Wo41–48En36–45Fs13–17) to those
from the 2010 eruption (Table 2). The magnesium number
[Mg# = 100 × Mg/(Mg + Fe2+)] of the 2006 pyroxenes
(Mg# = 76–83) also overlaps with the range of 2010 phenocrysts (Mg# = 61–86). The 2010 dome samples have
the largest range in Mg# (61–86), although most crystals
analysed from the dome lie between Mg# 71–84 (Table 2).
In comparison, grey scoria and white pumice samples contain clinopyroxene phenocrysts with Mg# = 77–83 and
Mg# = 75–84, respectively. Although the majority of clinopyroxene phenocrysts from the 2010 and 2006 samples
contain ~1–3 wt% Al2O3, the 2010 dome crystals display
the largest variation, with values ranging between 0.4 and
8.9 wt%. Other lithologies from the 2010 eruption also contain relatively high-Al2O3 clinopyroxene, with phenocrysts
from the white pumice containing 1.4–7.0 wt% Al2O3 while
those from the grey scoria contain 1.4–4.8 wt% Al2O3.
These values are similar to previous analyses of 2010 and
Contrib Mineral Petrol (2014) 168:1061
2006 Merapi clinopyroxenes, for which concentrations of
2–8 wt% and 1.5–6.5 wt% Al2O3 are reported (Costa et al.
2013). In comparison, the 2006 clinopyroxene samples
analysed here generally encompass a more restricted range,
with 1.4–3.8 wt% Al2O3, and a single analysis recording an
elevated Al2O3 content of 6.7 wt% (Fig. 6). The clinopyroxene phenocrysts hosting the melt inclusions that were
used for SIMS analysis sample the compositional range
found within the entire 2010 and 2006 clinopyroxene populations, although they do not sample the highest Al2O3
contents, as the analysed host crystals contain a maximum
of 5.6 wt% Al2O3 only (see Table 2 and Electronic Supplementary Data).
Discussion
Interpreting SIMS and ATR micro‑FTIR H2O and CO2 data
The scatter in melt inclusion H2O and CO2 concentrations in the SIMS data (Fig. 3a), with elevated CO2/H2O
compared to equilibrium degassing trends, has frequently
been noted in melt inclusions from subduction systems
(Atlas et al. 2006; Johnson et al. 2008; Vigouroux et al.
2008; Blundy et al. 2010; Berlo et al. 2012; Nadeau et al.
2013; Reubi et al. 2013). This feature is commonly interpreted as being the result of magma mixing (e.g. Atlas et al.
2006), complex degassing histories, involving CO2 fluxing
(Johnson et al. 2008; Vigouroux et al. 2008; Blundy et al.
2010; Nadeau et al. 2013) or non-equilibrium degassing
(Gonnermann and Manga 2005), or alternatively, due to
post-entrapment diffusive loss of H2O or H+ (Berlo et al.
2012; Reubi et al. 2013). At Merapi, elevated CO2/H2O
concentrations in amphibole and pyroxene-hosted melt
inclusions have previously been observed and interpreted
to reflect CO2 fluxing and the liberation of CO2-rich gas
via crustal carbonate assimilation (Nadeau et al. 2013). If a
significant amount of CO2 invades the system, the increase
in partial pressure of CO2 and corresponding decreasing
partial pressure of H2O results in increasing CO2/H2O in
the melt, shifting melt inclusion compositions isobarically
towards higher CO2 and lower H2O compositions. Crustal
assimilation at Merapi has been shown to have played a
part in the 2010 and 2006 eruptions of Merapi, evidenced
by elevated (compared to mantle values) whole-rock and
phenocryst δ18O values (Borisova et al. 2013; Troll et al.
2013). However, the question still remains as to whether
CO2 liberation from shallow (<10 km) crustal sediments
is likely to be recorded in melt inclusions. At the inferred
shallow depth (<10 km) and associated low pressure of
crustal CO2 liberation, it is unlikely that the CO2 is redissolved back into the melt. Instead, CO2 is more likely to
be lost through fumarolic activity or diffuse degassing
Page 11 of 25
1061
(Holloway and Blank 1994; Toutain et al. 2009; Troll
et al. 2012). Large increases of CO2/SO2, CO2/HCl and
CO2/H2O were detected in fumarolic gases in the months
leading up to the 2010 eruption, with a dramatic increase
in CO2 abundance from 10 mol% in September 2010 up
to 35–63 mol% on 20 October, interpreted to be due to a
progressive shift to degassing of a deeper magmatic source
(Surono et al. 2012). CO2 fluxing is a process that likely
takes place at Merapi and occurred prior to the 2010 and
2006 eruptions, both from depth and from the crust (Surono
et al. 2012; Borisova et al. 2013; Troll et al. 2013). However, isobaric dehydration due to CO2 fluxing should result
in melt inclusion compositions positioned along isopleth
lines on a H2O–CO2 plot and show a relatively continuous
range of H2O and CO2 contents (Reubi et al. 2013). The
Merapi melt inclusion data (Fig. 3) are inconsistent with
this expected trend, and so, although CO2 fluxing may be
occurring at Merapi, it is not likely that evidence of this
process is preserved in the studied set of melt inclusions.
Diffusive loss of H+ can displace melt inclusion compositions to high CO2 concentrations compared to degassing
trends. This process is often accompanied by Fe oxidation
and magnetite precipitation (Danyushevsky et al. 2002).
Magnetite daughter crystals were noted in some Merapi
melt inclusions, but none of these inclusions was chosen
for analysis. Significant diffusive loss of molecular H2O
is expected to result in crystallisation of daughter crystals
and decrepitation of the melt inclusion (Danyushevsky
et al. 2002), and has been shown experimentally to cause
the formation of shrinkage bubbles (Severs et al. 2007). No
decrepitation was observed and although, as noted above,
some Merapi melt inclusions contain small daughter crystals, none of these inclusions was chosen for analysis.
There is also no correlation between H2O and CO2 content
with calculated degree of PEC. Small bubbles were noted
occasionally in melt inclusions, indicating that H2O loss
may have occurred in a small percentage of inclusions. In
the presence of shrinkage bubbles, CO2 values in the melt
inclusion would represent minimum values, as CO2 preferentially partitions into the bubble, decreasing the concentration of CO2 in the melt with increasing dehydration (e.g.
Bucholz et al. 2013). Sub-micron bubbles, which may not
be detected during observation of the melt inclusions, could
result in high CO2 values, if sputtered during analysis.
However, 12C+ counts were recorded during each cycle of
SIMS analysis and counts always remained stable, indicating that CO2 concentrations are homogenous over the spatial extent of each analysis, precluding the possibility that
the high CO2 measurements are due to sub-micron bubbles. Diffusive loss of H2O has been shown experimentally
to occur on timescales of hours to days in olivine (Portnyagin et al. 2008; Chen et al. 2011; Gaetani et al. 2012). No
experimental data for H2O loss from clinopyroxene-hosted
13
Page 12 of 25
Contrib Mineral Petrol (2014) 168:1061
0
0
20
20
40
40
P (H2O) MPa
P (H2O) MPa
1061
60
80
100
120
4.5
4.0
4.0
3.5
3.5
3.0
3.0
H2O (wt. %)
H2O (wt. %)
4.5
2.5
2.0
Li-enrichment
1.5
c
2.0
1.5
d
0.0
74
74
72
72
70
70
68
66
64
B-enrichment
68
66
64
62
62
e
60
5500
5500
5000
5000
4500
4500
4000
Li-enrichment and
“buffered” Cl
3500
f
60
Cl (ppm)
Cl (ppm)
2.5
0.5
SiO2 (wt. %)
SiO2 (wt. %)
0.0
4000
3500
3000
3000
2500
2500
g
15
25
35
45
55
Li (ppm)
90
65
75
Dense dome
Grey scoria
80
White pumice
70
2010 Stage
2000
b
1.0
0.5
B (ppm)
100
140
1.0
4
6
6
2006 scoria
60
Divergent Li and B
50
40
i
30
15
13
80
120
a
140
60
25
35
45
55
Li (ppm)
65
75
2000
30
h
40
50
60
B (ppm)
70
80
90
inclusions, showing that Li is enriched in melt inclusions from the
2010 dense dome clasts and the 2006 scoriaceous dome clasts, and
B is enriched in white pumice melt inclusions. a, b Li and B versus
P (H2O), c, d Li and B versus H2O, e, f Li and B versus SiO2, g, h Li
and B versus Cl, i Li versus B
melt inclusions are available. However, Reubi et al. (2013)
determined that diffusion coefficients for natural pyroxene-hosted melt inclusions are several orders of magnitude smaller than experimentally determined H+ diffusion
coefficients (Hercule and Ingrin 1999; Stalder and Skogby
2003), and proved that pyroxene-hosted melt inclusions can
preserve H2O values close to entrapment values, even in
dome samples formed during low effusions rates (~0.6 m3
s−1). In addition, melt inclusions from slowly cooled samples are more prone to diffusive H2O loss (Hauri et al.
2002; Portnyagin et al. 2008; Lloyd et al. 2013). Therefore, although a small amount of diffusive H2O loss cannot be ruled out, and may possibly account for high CO2/
H2O in some melt inclusions from 2006 dome samples,
it is unlikely that such a process is the primary cause of
the apparently high CO2 values in the white pumice melt
10
inclusions, which are expected to be the fastest erupted and
cooled samples in this study.
Comparison of the SIMS data with the ATR microFTIR data collected in this study is essential to shed light
on the apparently high CO2/H2O melt inclusions in the
white pumice. When a subset of the high CO2/H2O inclusions (as measured by SIMS) was subsequently measured
by FTIR, no CO2 was detected (detection limit 200 ppm).
A possible explanation for the discrepancy could be due to
the fact that SIMS measures 12C+ and FTIR measures CO2.
Any 12C+ that is present in the melt inclusions as carbonate would not be recorded in the spectrum at 2,350 cm−1,
but in the range of 1,400–1,500 cm−1. However, this range
was also monitored during FTIR analysis, and no carbonate was detected (detection limit 500 ppm). The discrepancy between the SIMS and ATR micro-FTIR data sets
is potentially caused by heterogeneities in the distribution of CO2 in the melt inclusion. The sampling volume
of the two techniques is different, with SIMS sampling a
smaller region of the inclusion than FTIR. In this respect,
FTIR would not detect any heterogeneity, giving only an
average CO2 concentration of the inclusion, whereas SIMS
analysis will sample a smaller area within each inclusion,
6
2010 host cpx
9
Grey scoria
Other 2010 cpx
8
5
Other 2006 cpx
White pumice
4
6
6
2006 scoria
Al2O3 (wt. %)
7
Al2O3 (wt. %)
Dense dome
2006 host cpx
2010 Stage
◂Fig. 5 Light lithophile element (Li and B) concentrations in melt
6
5
4
3
2
4
3
2
1
1
a
0
39
41
43
45
47
Wo (mol. %)
49
b
0
51
39
41
43
45
Wo (mol. %)
47
49
6
10
9
5
8
Al2O3 (wt. %)
7
Al2O3 (wt. %)
Fig. 6 Clinopyroxene phenocryst compositions. a Al2O3
(wt %) versus Wo (mol%) (wollastonite end-member) of all
measured clinopyroxene phenocrysts in 2010 and 2006 products, including the phenocrysts
that host melt inclusions used
in this paper (white symbols),
as well as other phenocrysts
from the same eruptive products
(black symbols). b Al2O3 (wt %)
versus Wo (mol%) in the host
phenocrysts, divided into
eruptive stages. c Al2O3 (wt %)
versus Mg# of all measured
clinopyroxene phenocrysts in
2010 and 2006 products, including the host phenocrysts (white
symbols) that host melt inclusions used in this paper, as well
as other phenocrysts from the
same eruptive products (black
symbols). d Al2O3 (wt %) versus
Mg# in the host phenocrysts,
divided into eruptive stages
1061
Page 13 of 25
Contrib Mineral Petrol (2014) 168:1061
6
5
4
3
2
3
2
1
1
0
4
c
60
65
70
75
Mg #
80
85
d
0
90
70
72
74
76
78 80
Mg #
82
84
86
13
13
81.2
43.7
44.0
12.3
Mg#
Wo
En
Fs
14.8
43.6
41.6
78.6
4.000
0.029
0.228
13.9
40.8
45.3
80.8
4.000
0.027
0.864
0.012
0.185
0.777
0.080
0.000
0.019
0.161
1.874
100.12
0.38
21.78
14.07
0.40
8.56
0.00
0.67
3.68
50.59
x30-h1
M11-27-5
DD
4
2010
17.6
39.2
43.2
71.6
4.000
0.024
0.834
0.018
0.300
0.755
0.040
0.001
0.005
0.073
1.950
100.65
0.33
20.92
13.61
0.58
10.93
0.03
0.17
1.67
52.41
x34-h2
M11-27-5
DD
4
2010
14.2
43.1
42.7
76.9
4.000
0.026
0.816
0.019
0.249
0.825
0.023
0.001
0.010
0.078
1.952
100.07
0.37
20.55
14.93
0.61
8.77
0.05
0.35
1.79
52.66
x13-h1
M11-75
GS
6
2010
13.5
42.6
43.9
81.5
4.000
0.029
0.848
0.015
0.186
0.822
0.075
0.000
0.011
0.095
1.918
99.95
0.40
21.38
14.89
0.48
8.43
0.01
0.38
2.18
51.80
x11-h2
M11-75
GS
6
2010
13.7
43.2
43.1
79.8
4.000
0.026
0.825
0.015
0.210
0.827
0.053
0.001
0.014
0.113
1.915
100.13
0.37
20.83
15.00
0.47
8.52
0.02
0.52
2.60
51.80
x30-h1
M11-75
GS
6
2010
13.9
43.4
42.7
77.9
4.000
0.023
0.820
0.018
0.237
0.835
0.030
0.000
0.010
0.081
1.946
99.83
0.32
20.62
15.09
0.56
8.60
0.00
0.35
1.86
52.44
x41-h1
M11-75
GS
6
2010
13.8
40.5
45.7
79.4
4.000
0.025
0.865
0.009
0.199
0.767
0.064
0.000
0.020
0.181
1.870
100.52
0.35
21.88
13.93
0.30
8.50
0.01
0.71
4.17
50.68
x17-h1
M11-55
WP
6
2010
14.4
43.0
42.6
78.3
4.000
0.025
0.818
0.021
0.229
0.825
0.048
0.001
0.012
0.090
1.931
99.97
0.35
20.57
14.91
0.67
8.94
0.04
0.43
2.05
52.02
x44-h1
M11-55
WP
6
2010
14.4
38.5
47.1
82.7
4.000
0.024
0.891
0.008
0.153
0.728
0.121
0.000
0.025
0.246
1.804
98.95
0.33
22.16
13.01
0.26
8.71
0.00
0.88
5.56
48.04
x20-h1
M11-50
WP
6
2010
14.5
41.1
44.4
75.9
4.000
0.027
0.838
0.017
0.246
0.775
0.027
0.000
0.016
0.141
1.913
100.44
0.37
21.14
14.05
0.54
8.83
0.01
0.57
3.23
51.71
x71-h2
M11-50
WP
6
2010
2006
14.9
43.4
41.7
77.9
4.000
0.024
0.803
0.023
0.237
0.835
0.049
0.000
0.011
0.083
1.935
99.99
0.34
20.18
15.09
0.72
9.23
0.00
0.38
1.89
52.15
x70-h2
M07-53
DS
14 Jun
2006
14.5
41.4
44.1
77.0
4.000
0.025
0.836
0.018
0.235
0.785
0.040
0.000
0.015
0.136
1.909
99.89
0.35
20.97
14.16
0.58
8.84
0.02
0.55
3.11
51.32
x13-h1
M07-53
DS
14 Jun
2006
14.3
43.1
42.6
76.3
4.000
0.025
0.816
0.023
0.256
0.824
0.017
0.000
0.010
0.071
1.958
99.83
0.34
20.49
14.87
0.72
8.78
0.00
0.37
1.61
52.65
x28-h1
M07-53
DS
14 Jun
2006
14.2
42.2
43.6
76.2
4.000
0.025
0.829
0.017
0.252
0.803
0.018
0.001
0.014
0.104
1.937
99.85
0.35
20.80
14.49
0.54
8.67
0.02
0.50
2.38
52.10
x48-h2
M07-53
DS
14 Jun
Page 14 of 25
Stages 4 and 6 of the 2010 eruption refer to the chronology of Komorowski et al. (2013). Lithology types are as follows: dense dome (DD), grey scoria (GS), white pumice (WP), scoriaceous dome clasts (DS). FeO* denotes total iron reported as FeO. Mg# = 100 × Mg/(Mg + Fe2+). Clinopyroxene end-member compositions were calculated in mol% from the ferric structural formula as follows: wollastonite (Wo) = 100 × Ca/(Ca + Mg + Fe*), enstatite (En) = 100 × Mg/(Ca + Mg + Fe*) and ferrosilite (Fs) = 100 × Fe*/(Ca + Mg + Fe*)
where Fe* denotes total iron (Fe2+ and Fe3+). Ferric iron was calculated assuming perfect stoichiometry
4.000
Sum
0.800
0.842
0.027
Ca
Mn
Na
0.023
0.196
0.014
Fe2+
0.056
Mg
0.839
0.041
0.847
Fe3+
0.001
0.010
0.009
0.000
Ti
Cr
0.076
1.944
0.079
Si
1.938
100.15
0.41
20.16
15.19
0.73
9.18
0.05
0.36
Al
Ferric Form
0.37
99.59
Total
0.46
MnO
Na2O
7.66
FeO*
15.33
0.00
Cr2O3
21.20
0.33
TiO2
MgO
1.80
Al2O3
CaO
52.33
52.44
1.74
x12-h1
x21-h1
M11-27-5
DD
4
2010
SiO2
Cpx no.
DD
M11-25-5
Type
Sample
4
2010
Stage
Eruption
Table 2 Representative clinopyroxene compositions from the 2010 and 2006 eruptive products of Merapi
1061
Contrib Mineral Petrol (2014) 168:1061
Contrib Mineral Petrol (2014) 168:1061
possibly enabling the detection of any heterogeneity. The
idea of heterogeneous distribution of CO2 in the melt inclusions is supported by the analysis of one melt inclusion that
was large enough to be analysed by SIMS in two separate
areas, with one measurement of 1,167 ppm CO2 and one
of 146 pm CO2. A low CO2 inclusion from the 2006 dome
scoria was large enough to analyse twice and both analyses
yielded similar results, although other melt inclusions were
too small to obtain multiple analyses. Possible causes for
the heterogeneous distribution of CO2 in the white pumice
melt inclusions may be linked to a post-entrapment process. For example, potential heterogeneity may be linked to
the proximity of the SIMS analysis spot with a bubble not
in the plane of view. If CO2 diffuses into the bubble, this
may possibly result in heterogeneous distribution of CO2
within the inclusion. Although no bubbles were observed
in these inclusions, it is possible that they were not in the
plane of view or were removed during the polishing process. Another possibility may be that CO2 diffusion, unlike
the relatively fast diffusion of H2O (e.g. Zhang et al. 1991),
cannot keep up with host crystallisation after entrapment,
potentially becoming highly concentrated in some areas
and resulting in heterogeneity. This hypothesis may help to
explain why apparent CO2 enrichment is only seen in rapidly ejected white pumice samples and not in those from
the 2010 dome material, which extruded more slowly,
potentially allowing time for re-equilibration of the CO2 in
the inclusion. Although beyond the scope of this work, heterogeneous CO2 distribution in natural silicate melt inclusions is an important issue that demands further work to
understand fully. If the high CO2/H2O ratios in these inclusions are due to a post-entrapment process, then other melt
inclusions showing the same trend need to be interpreted
with caution. In the light of this, the apparently elevated
CO2 values are reported here but have not been used in the
reconstruction of magmatic degassing pathways and equilibration pressures. The remaining CO2 measurements are
retained as they reproduce modelled trends for equilibrium
degassing, thereby giving additional confidence in results.
Clinopyroxene crystallisation and melt inclusion
entrapment
The crystallisation pressures of melt inclusion-hosting
clinopyroxene phenocrysts were calculated using three
different barometers (Fig. 7). The first is a clinopyroxenemelt model, which is based on the Al partitioning between
clinopyroxene and liquid, calibrated for hydrous systems
by Putirka (2008, Eq. 32c). The model requires an input
of H2O concentration (4 wt% used, as determined by the
maximum measured H2O concentration in melt inclusions) and an estimated temperature, calculated using the
Putirka et al. (2003) thermometer (range between 1,025
Page 15 of 25
1061
and 1,058 °C). The andesitic to rhyolitic compositions of
the melt inclusions and groundmass glass are not in equilibrium with the clinopyroxene host (e.g. the Mg/Fe equilibrium criterion of Putirka (2008) is not met). Therefore,
a basaltic andesite bulk rock composition with ~55.5 wt%
SiO2 was used, which is representative of the composition
of the 2010 and 2006 products and satisfies the equilibrium criteria. The model has a standard error of estimate
(SEE) of ± 150 MPa and has recently yielded satisfactory
results for similar Indonesian volcanic systems (Dahren
et al. 2012; Jeffery et al. 2013). The second barometer
is that of Putirka (2008), which is a recalibrated version
of Nimis (1995) suitable for hydrous systems and based
only upon the clinopyroxene composition. The recalibration aims to remove the systematic error associated with
the Nimis (1995) barometer, which can yield low pressure estimates, at the cost of requiring an H2O estimate in
addition to the temperature input already needed. The SEE
is ± 260 MPa (Putirka 2008). As an additional test, a second clinopyroxene barometer was employed using CpxBar
(Nimis 1999), an Excel spreadsheet based upon the pressure-dependent clinopyroxene unit cell arrangement. Calculations for the Merapi clinopyroxenes utilised the mildly
alkaline series calibration (standard error: 200 MPa), with
a temperature input of 1,050 °C. The range of 2010 clinopyroxene (n = 57) crystallisation pressures obtained by all
three models is within error of each other: 80–510 (±150)
MPa (Putirka 2008, Eq. 32c), −60 to 430 (±260) MPa
(Putirka 2008, Eq. 32b) and 20–540 (±200) MPa (Nimis
1999), equivalent to depths of 2.9–18.6 (±5.5) km, 0–15.7
(±9.5) km and 0.7–19.7 (±7.3) km, respectively (Fig. 7),
assuming an average crustal density of 2,800 kg/m3, as
also used in depth calculations at Merapi by Costa et al.
(2013). Taking into account the pressure range obtained
by all three models, crystallisation pressures of the 2006
clinopyroxenes (n = 15) lie within a similar, albeit more
restricted range of 120–420 MPa, equivalent to depths of
4.4–15.3 km. One striking feature of the data is that the
“deepest” crystals are from the white pumice samples,
with 80 % of those crystallised at P > 300 MPa using
the Putirka (2008) models (Eq. 32b and 32c), and nearly
80 % of the those crystallised at P > 400 MPa using the
Nimis (1999) model, originating from the white pumice
(Fig. 7). This is consistent with results from amphibole
barometry (Preece 2014), which also suggest that the
“deepest” amphibole crystals are from white pumice samples. In comparison, Costa et al. (2013) proposed that the
2010 magma was stored within a multi-depth plumbing
system comprising: (1) a deep reservoir at 30 (±3) km or
~800 MPa, as evidenced by some amphiboles and high-Al
clinopyroxene, (2) a reservoir at intermediate depths of 13
(±2) km or 300–450 MPa, where other amphiboles, highAl cpx and high-An plagioclase grew and (3) a shallow
13
Page 16 of 25
26
26
a
Nimis (1999)
24
22
22
20
20
18
16
14
12
10
8
18
16
14
12
10
8
6
6
4
4
2
2
0
0
0
26
0
50 100 150 200 250 300 350 400 450 500 550
Putirka (2008 Eq. 32b)
24
50 100 150 200 250 300 350 400 450 500 550
18
b
Dense dome
16
22
Grey scoria
White pumice
14
20
18
Number of Analyses
Number of Analyses
c
Putirka (2008 Eq. 32c)
24
Number of Analyses
Number of Analyses
Contrib Mineral Petrol (2014) 168:1061
16
14
12
10
8
6
2010 Stage
1061
4
d
6
6
2006
12
10
8
6
4
4
2
2
0
0
0
50 100 150 200 250 300 350 400 450 500 550
Pressure (MPa)
0
25
50
75 100 125 150 175 200 225 250 275
Pressure (MPa)
Fig. 7 Histograms to show pressure (MPa) of clinopyroxene host
crystallisation. a Calculated with CpxBar (Nimis 1999) using the MA
(mildly alkaline) calibration at 1,050 °C. b calculated using Equation 32b of Putirka (2008), one value of minus pressure not plotted.
c Calculated using Equation 32c of Putirka (2008). d Histogram of
last re-equilibration pressures (MPa) of melt inclusions from different
2010 stages and from 2006, calculated using VolatileCalc (Newman
and Lowenstern 2002) at 1,050 °C, based on melt inclusion H2O and
CO2 concentrations
reservoir at <10 km (~100 MPa) where extensive crystallisation produced low-Al cpx, orthopyroxene and plagioclase. The clinopyroxenes analysed in this study therefore
appear to have crystallised in a depth range consistent with
the intermediate and shallow magma storage zones proposed by Costa et al. (2013).
The wide variation of H2O concentrations in melt inclusions suggests that some inclusions have ruptured on timescales that were long enough to allow the escape of H2O,
but not to allow chemical re-equilibration with the groundmass melt (Blundy and Cashman 2005) (Fig. 8). This is
consistent with the fact that the melt inclusions with the
lowest H2O concentration are from the densest (dome)
samples, which have experienced a protracted cooling history at low pressure in the dome. When focussing only on
the scoria and pumice samples, the general trend is for
lower SiO2 and K2O melt inclusions to have higher H2O
concentration than the higher SiO2 and K2O melt inclusions, which is consistent with vapour-saturated crystallisation in response to decompression (Fig. 8). Individual melt
inclusion entrapment pressures, or the last pressures of reequilibration assuming vapour saturation, were calculated
using the Papale H2O–CO2 model (Papale et al. 2006), and
using VolatileCalc1.1 (Newman and Lowenstern 2002),
with both sets of calculations carried out at a temperature of
1,050 °C based upon the clinopyroxene-melt thermometry
13
4.5
Dense dome
4.5
Decompression vapour sat.
crystallisation
4.0
3.5
3.5
3.0
3.0
2.5
2.0
1.5
1.0
0.5
a
0.0
61
Grey scoria
4.0
H2O (wt. %)
H2O (wt. %)
Fig. 8 H2O versus indicators
of magmatic differentiation.
a SiO2, b K2O. Melt inclusions last equilibrated during
decompression vapour-saturated
crystallisation, with evidence
that some melt inclusions ruptured during ascent
White pumice
6
6
2006 scoria
2.5
2.0
1.5
b
0.5
0.0
65
4
1.0
‘Ruptured’
inclusions
63
1061
2010 Stage
Page 17 of 25
Contrib Mineral Petrol (2014) 168:1061
67
69
SiO2 (wt. %)
and previous thermometry of the 2010 products (Costa
et al. 2013). Pressure values calculated with each model
are similar. Minimum equilibration pressures, assuming
volatile saturation, range from <5 to 265 MPa using VolatileCalc. This pressure range is similar to that previously
reported for Merapi pyroxene-hosted melt inclusions,
which found the majority were trapped below 266 MPa
(Nadeau et al. 2013). However, as stated above, the lowest
values probably reflect the fact that some melt inclusions
have ruptured during ascent. To get a better estimate of the
lowest entrapment pressure, the most evolved inclusion that
sits on the decompression trend (Fig. 8) can be used, which
suggests equilibration at ~16 MPa. Therefore, melt inclusion equilibration depths range between ~0.6 and 9.7 km.
In summary, there is an overlap between clinopyroxene crystallisation and melt inclusion equilibration at the
lower end of the pressure range, although clinopyroxene
crystallisation continues to greater depths, not reflected in
the melt inclusion population. Clinopyroxene phenocrysts
crystallised at depths up to ~20 km, with calculated minimum melt inclusion equilibration occurring between 0.6
and 9.7 km. The clinopyroxene barometry results are in
agreement with previous clinopyroxene barometry carried on recent eruptive products from Merapi (Gertisser
2001; Chadwick et al. 2013). Clinopyroxene crystallisation depths are also in accord with other geophysical and
petrological evidence concerning the magmatic plumbing system of Merapi (e.g. Beauducel and Cornet 1999;
Gertisser 2001; Costa et al. 2013). The calculated melt
inclusion equilibration pressures complement those previously reported for Merapi (Nadeau et al. 2013). The discrepancy between the higher crystallisation pressure of
the host clinopyroxene phenocrysts and the equilibration
pressure of the melt inclusions can be explained in several ways. One possible explanation is that the melt was
not vapour-saturated at the time of entrapment, therefore
recording seemingly lower pressures than those of actual
entrapment. This scenario is not considered likely as it is
71
73
4.0
4.5
5.0
5.5
6.0
6.5
7.0
K2O (wt. % )
not consistent with the trends of vapour-saturated crystallisation (Fig. 8), nor with the presence of exsolved brine
(see below), and is contradictory to previous evidence of
volatile-saturated melt inclusions at Merapi (Nadeau et al.
2013). Another interpretation is that the inclusions are not
primary but secondary, i.e. not trapped within hosts during initial crystallisation, but instead during subsequent
partial dissolution, as previously observed for Merapi
melt inclusions in amphibole and pyroxene hosts (Nadeau
et al. 2013). However, there is no petrographic evidence
for dissolution of the clinopyroxene host crystals that
were analysed for this study. Therefore, the most likely
explanation for the pressure difference is due to some
melt inclusions re-equilibrating with the magma after
initial entrapment. This is consistent with the vapoursaturated crystallisation trends, as shown in Fig. 8. The
melt inclusions re-equilibrated over a range of pressures
<265 MPa (~<10 km), indicating either that there was
enough time for re-equilibration during ascent, or, more
likely, that magma was stored at this pressure range for a
period of time before eruption. The fact that the maximum
melt inclusion re-equilibration depth of ~10 km matches
that of previous melt inclusion studies (Nadeau et al.
2013), hints at the possibility that the magma has stalled
at this depth. This depth is broadly consistent with geophysical (Beauducel and Cornet 1999; Ratdomopurbo and
Poupinet 2000) and petrological (Chadwick et al. 2013;
Costa et al. 2013) evidence for a magma storage region
beneath Merapi. Seismic observations prior to the 2010
eruption detected heightened activity at the beginning of
September 2010, with tremors at focal depths of 2.5–5 km
depth until 17 October 2010, when activity became shallower and focal depths were <1.5 km (Budi-Santoso et al.
2013; Jousset et al. 2013). This is consistent with ascent
of a magma body over 1.5 months prior to eruption (BudiSantoso et al. 2013) with the depths correlating with those
recorded by many of the melt inclusions and shallower
crystallised clinopyroxenes.
13
Page 18 of 25
Using VolatileCalc (at 1,050 °C), trends for open- and
closed-system degassing were calculated and compared to
the Merapi melt inclusion data. In the open-system model,
exsolved volatiles are removed from the system, whereas
for closed-system runs, the exsolved vapour remains within
the system, acting as a buffer on the remaining melt, including those volatiles that remain in solution in the depressurising magma. Closed-system models fit the data better
than open-system ones, with the overall best fit having an
initial starting point of 4.0 wt% H2O, 1,000 ppm CO2 and
1 % exsolved vapour (Fig. 3b). This closed-system trend is
mainly defined by white pumice inclusions, indicating that
closed-system degassing occurred prior to the sub-plinian
explosive phase of the 2010 eruption, with the exsolved
volatiles staying within the system, thus likely leading to
increased overpressure and increased explosive activity.
Inclusions from other stages of the eruption have re-equilibrated after the majority of the CO2 was degassed.
Evidence of pre‑ and syn‑eruptive crystallisation
and degassing processes from F, S and Cl concentrations
Fluorine generally behaves as an incompatible element,
although may be incorporated into certain minerals present
in the Merapi mineral assemblage. For example, amphibole from the 2010 eruption has been found to include up
to 0.6 wt% F and apatite contains 2.8–5.4 wt% F (Preece
2014). Fluorine concentrations in melt inclusions and
groundmass glass (Fig. 4) are similar and within the typical
range for subduction related magmas (Aiuppa et al. 2009
and references therein). The results are compatible with the
fact that F is highly soluble in silicate melts (e.g. Carroll
and Webster 1994), and generally partitions in favour of
the melt relative to the vapour phase; therefore, it is often
not significantly extracted from the melt by degassing (e.g.
Balcone-Boissard et al. 2010).
Melt inclusions from the 2010 eruption contain a range
of S from below the detection limit (~100 ppm) up to
535 ppm. Total SO2 emissions for the entire 2010 eruption
were calculated at 0.44 Tg, based on satellite observations
(Surono et al. 2012). Assuming that the maximum S concentration in the melt inclusions (535 ppm) represents the
pre-eruptive volatile content of the melt, and the S concentration in the groundmass glass (<100 ppm) indicates the
post-eruptive volatile content, it is possible to calculate the
mass and volume of magma needed to produce the 0.44 Tg
of SO2 that were emitted during the eruption (Self et al.
2004). The approximate phenocryst content of the 2010
dome samples is ~40 vol%, and taking into account the
fraction of felsic to mafic minerals in the rocks, the proportion of phenocrysts is ~45 wt% (i.e. the melt or groundmass
13
occupies 55 wt%). Results indicate that the mass of degassing magma in the 2010 eruption was 9.2 × 1011 kg, and
taking into account a melt density of 2,550 kg m−3, calculated from the composition of a typical dome clast using
the software Pele (Boudreau 1999), the corresponding melt
volume is 0.36 km3. In comparison, it has been estimated
that between 0.03 and 0.06 km3 of magma (DRE) was
erupted during the 2010 eruption (Surono et al. 2012), an
order of magnitude less than the calculated total volume of
degassing melt in the system, indicating either that a large
proportion of melt did not erupt and remained in the system
or that extra sources of volatiles (S) contributed to the volatile budget of the eruption.
The Cl content of the groundmass glass is generally
lower than the concentration in the melt inclusion population, indicating Cl exsolution at low pressures during syneruptive degassing (Fig. 9). The Cl concentrations in the
melt inclusions are similar to other arc magmas (Aiuppa
et al. 2009 and references therein). Melt inclusion Cl concentrations range between 2,060 and 5,130 ppm, although
nearly 90 % of the data in this study range between 2,400
and 3,400 ppm (Fig. 9), with the higher concentrations
only observed in white pumice melt inclusions. The simplest explanation for a relatively limited range of Cl in the
melt inclusions, regardless of variations in H2O, could be
that the magma underwent decompression degassing with
preferential loss of H2O and no apparent loss of Cl (Webster et al. 2010; Mann et al. 2013). However, an alternative
Dense dome
4.5
Grey scoria
4.0
White pumice
2010 Stage
Degassing history (H2O and CO2)
Contrib Mineral Petrol (2014) 168:1061
4
6
6
2006 scoria
3.5
3.0
H2O (wt. % )
1061
2.5
2.0
1.5
1.0
0.5
0.0
0
1000
2000
3000 4000
Cl (ppm)
5000
6000
Fig. 9 Concentration of Cl versus H2O in melt inclusions. Melt
inclusions from the 2010 samples as well as from the 2006 scoriaceous dome show constant Cl concentrations with variable H2O contents, providing evidence that the Cl concentration in the melt was
“buffered” by a hydrochloride phase. The only inclusions that deviate
to higher Cl concentrations are several from the white pumice
Contrib Mineral Petrol (2014) 168:1061
explanation is that the silicate melt was in equilibrium with
a magmatic hydrosaline chloride liquid ± vapour (e.g.
Lowenstern 1994; Shinohara 1994; Webster 1997, 2004).
Chlorine concentrations in the silicate melt are “buffered”
and so remain constant as Cl reaches its solubility limit
in equilibrium with the liquid phase(s). Similar indirect
geochemical evidence for the presence of a hydrosaline
chloride liquid from melt inclusion data has been reported
at Mount Hood (Koleszar et al. 2012), the Bandelier
Tuff, Valles Caldera (Stix and Layne 1996) and Vesuvius
(Signorelli et al. 1999; Fulignati and Marianelli 2007). The
Cl concentrations in the Merapi melt inclusions are similar
to saturation values for felsic melts determined experimentally (e.g. Métrich and Rutherford 1992; Webster 1997) as
well as the values recorded in natural felsic melt inclusions
interpreted to have been in equilibrium with a Cl-rich liquid (Stix and Layne 1996; Koleszar et al. 2012). This evidence for the presence of a saline liquid or “brine” phase is
corroborated by the Li and B data (see section below).
Li and B enrichment
Lithium is enriched in melt inclusions from the 2010 dense
dome material and the 2006 scoria, which contain relatively low concentrations of H2O. It displays an L-shaped,
divergent trend when plotted against H2O and Cl and displays no correlation with SiO2 or K2O (Fig. 5). Enrichment of B occurs in melt inclusions from white pumice
samples, which have mid-range H2O content, with variable
Cl, including elevated contents. These contrasting features
indicate that Li and B enrichment are produced by different
processes, corroborated by the decoupled relationship of Li
versus B (Fig. 5) as well as the fact that enrichment occurs
in distinct and separate lithological types.
Lithium-enriched inclusions, only detected in 2010
dense dome samples and 2006 scoriaceous dome samples,
have last equilibrated at near-surface depths, with the maximum Li enrichment in 2010 inclusions at PH2 O = 14.5 MPa
(0.5 km depth) and PH2 O = 36.5 MPa (1.3 km depth) for
maximum Li in 2006. This implies that Li enrichment is
a pressure-mediated process. Lithium enrichment in early
erupted dome samples, detected in melt inclusions equilibrated at specific H2O concentrations, has also been noted
at Mount St. Helens (Berlo et al. 2004; Blundy et al. 2008)
and at Shiveluch (Humphreys et al. 2008). This has been
attributed to pre-eruptive, vapour-mediated transfer of
Li, derived from deeper within the magma transport system. Alternatively, Li-rich inclusions may be produced
via re-equilibration with a Li-rich brine phase (Kent et al.
2007). An alkali-rich aqueous vapour can be produced via
the degassing of an H2O-saturated magma body. Buoyant
upward migration of this vapour results in its exsolution to
produce two phases: a dense alkali-rich brine and a lower
Page 19 of 25
1061
density H2O-rich vapour (Kent et al. 2007). The lower
density H2O-rich vapour is subsequently lost via degassing and the alkali-rich brine re-equilibrates with the melt,
resulting in Li-enriched melt inclusions. Gas measurements
and mass balance models (Nadeau et al. 2013) also predict
the presence of a brine phase below Merapi and suggests
that the volatile phase is exsolved as a single supercritical fluid below at least 5 km depth and exsolves directly
as vapour + brine phases above this critical depth. In the
Merapi melt inclusions, the Li-enriched inclusions all contain similar Cl concentrations (Fig. 5), further indicating
that Li-enrichment occurred during “buffering” of the melt
Cl concentrations, related to equilibration of the melt with
the brine phase. The inflexed trend in SiO2 versus Na2O,
with decreasing Na2O in the groundmass glass (Fig. 2),
may also be attributed to Na partitioning into the exsolving
fluid phase during volatile-saturated crystallisation (Blundy
et al. 2008).
The behaviour of B in magmatic and hydrothermal systems is complex and previous studies have reached differing conclusions as to whether B partitions preferentially
into the melt or into the aqueous phase(s) (e.g. Pichavant
1981; Webster et al. 1989; Heinrich et al. 1999; Hervig
et al. 2002; Schatz et al. 2004). Recent studies of natural
samples have shown that transient B depletion and enrichment may be related to changing degassing regimes
(Menard et al. 2013; Vlastélic et al. 2013) or alternatively
enrichment may occur due to crystallisation and magmatic fractionation (Wright et al. 2012). In the Merapi melt
inclusions, the correlation of B with SiO2 indicates that B
behaves as an incompatible element. The enrichment of
B with crystallisation shows that B remained in the melt,
suggesting a low vapour-melt partition coefficient, as also
observed, for example, in melt inclusions from Crater Lake
(Wright et al. 2012). One possible control on the partitioning behaviour of B is the NaCl content of the brine and
vapour, as higher B solubility is expected in melts which
are in equilibrium with saline brines (Schatz et al. 2004).
Therefore, the enrichment of B in the Merapi melt may be
due to the presence of the brine which acted to increase
B solubility in the melt, allowing the incompatible B to
become enriched in the melt during crystallisation.
Driving forces behind the 2010 eruption and comparison
with the 2006 eruption
In many respects, the contrasting behaviour of the 2010
and 2006 eruptions is not reflected in the investigated eruptive products. Whole rock, melt inclusion and groundmass
compositions are all similar in terms of major elements
and volatile (H2O, CO2, F, Cl and S) concentrations are
predominantly similar also. Whole-rock compositions are
similar for the 2010 and 2006 erupted lavas and remained
13
1061
Page 20 of 25
Contrib Mineral Petrol (2014) 168:1061
nearly constant throughout the duration of each eruption.
This indicates that bulk magmatic composition cannot be
a major factor in the change in behaviour between the two
basaltic andesite eruptions. However, there are some differences between the products, particularly in melt inclusions
originating from the white pumice, which is unique to the
2010 eruption. The white pumice clasts show numerous
distinctive features when compared to other 2010 samples
and reveal evidence of an increase in deep magma supply,
highlighted by an abundance of clinopyroxene phenocrysts
that crystallised at >11–15 km depth (>300–400 MPa)
(Fig. 10). This is in agreement with other petrological and
monitoring data (Surono et al. 2012; Costa et al. 2013;
Jousset et al. 2013). Increasing fumarole gas temperatures
and ratios of CO2/SO2, CO2/HCl and CO2/H2O in months
prior to the 2010 eruption, were interpreted to be due to
the degassing of a deep magmatic source (Surono et al.
2012). Costa et al. (2013) proposed that the 2010 eruption
was preceded by an influx of deeper, hotter, more volatilerich magma that was up to 10 times more voluminous than
that in 2006. Melt inclusions from grey scoria and white
pumice clasts also preserve evidence of closed-system
conditions during ascent, with higher H2O concentrations
compared to the dome melt inclusions. Volatile concentrations in melt inclusions from white pumice clasts fit with
5
Exsolution of brine
phase with enrichment
of Li and B in melt
inclusions
Crustal carbonates and
marls
Melt inclusion
re-equilibration
Depth (km)
10
Major zone of cpx
crystallisation
T = 1025 - 1058 °C
15
20
Magma influx from the
deep magma system
Deeper magma system
25
Fig. 10 Schematic diagram of the Merapi magmatic plumbing system as revealed by 2010 and 2006 melt inclusions and clinopyroxene
host phenocrysts
13
modelled concentrations predicted for a closed-system
in equilibrium with 1 % exsolved volatiles. A deep influx
of magma would have caused increased overpressure
and faster magma ascent. Closed-system degassing and
fast magma ascent rates, as a result of the deep influx of
magma, were key driving forces behind the explosive
nature of the 2010 eruption. Closed-system degassing
likely sustained explosive behaviour, generating a subplinian convective column, which collapsed to produce the
scoria- and pumice-rich PDCs during Stage 6 of the 2010
eruption. The continuous range of melt inclusion re-equilibration at depths between 9.7 and 0.6 km suggests that the
melt region, at these depths at least, is interconnected and
not formed by isolated chambers (Fig. 10).
Silicate melt from early-erupted products (i.e. dome
clasts) of both 2010 and 2006 was saturated in Cl, preserving evidence of the presence of a hydrosaline fluid or
“brine” phase within the magmatic system prior to each
eruption. Further evidence for this comes from the fact
that these same melt inclusions are enriched in Li, indicating similar shallow-level processes operating prior to the
dome-building stage of each eruption, in agreement with
Nadeau et al. (2013). White pumice melt inclusions are
enriched in B, which was stabilised in the melt rather than
degassing, due to the presence of the brine phase. It acted
as an incompatible element, thereby becoming enriched in
the melt during fractional crystallisation.
Although the melt inclusions and their clinopyroxene
hosts in this study have shed light on the magmatic system and magmatic contributions to the eruptive behaviour
of Merapi during 2010 and 2006, other factors are likely
to have contributed to the eruptive dynamics. Shallowlevel degassing and crystallisation of microlites, likely
not reflected in melt inclusion data, have been associated
with increasing overpressure, changing degassing regimes
and increasing explosivity in 2010 (Preece 2014). Previous workers (e.g. Chadwick et al. 2007; Deegan et al. 2010;
Troll et al. 2012, 2013; Borisova et al. 2013) have proposed
that CO2 liberation via crustal carbonate assimilation has
the potential to sustain and intensify eruptions at Merapi.
Monitoring data of the 2010 eruption indicate large CO2
emissions prior to the 2010 eruption (Surono et al. 2012),
and previous melt inclusions have been interpreted to
reflect CO2 fluxing at Merapi (Nadeau et al. 2013). However, in contrast to previous studies (Nadeau et al. 2013),
the CO2 data and trends collected during this study cannot be reliably attributed to CO2 fluxing due to the apparent heterogeneous distribution of CO2 in some of the melt
inclusions. The peak of the 2010 eruption was preceded
by a regional tectonic earthquake. On 4 November 2010,
a M4.2 earthquake occurred ~200 km south of Merapi,
at 23:56 local time, just prior to the climactic phase of 5
November. Several authors have suggested a probable
Page 21 of 25
Contrib Mineral Petrol (2014) 168:1061
connection between regional seismic activity and the intensification of eruptive activity at Merapi (e.g. Walter et al.
2007; Surono et al. 2012; Troll et al. 2012; Jousset et al.
2013). Although Jousset et al. (2013) stress that the main
factor causing over-pressurisation at Merapi in 2010 was
related to magmatic influx and ascent, it is possible that
with the magmatic system already at a critical pressurised
stage, a small external force can affect the gas phase, promoting rapid degassing, fragmentation and eruption of an
already unstable system (e.g. Brodsky et al. 1998; Davis
et al. 2007; Walter et al., 2007).
The melt inclusions and clinopyroxene hosts shed light
on the magmatic plumbing system prior to the two most
recent eruptions of Merapi and reveal an influx of deeper
magma during the 2010 eruption, which was a key factor
in determining the eruptive dynamics of the cataclysmic
2010 eruption. However, in addition to this, the influence of
factors such as shallow-level degassing and crystallisation
feedback mechanisms, carbonate assimilation and regional
earthquakes should not be ignored.
Conclusion
Using melt inclusions and their clinopyroxene hosts, this
study has revealed information about the pre-2010 and
pre-2006 Merapi magma system, and key factors that contributed to the cataclysmic events of 2010. Despite the
contrasting eruptive behaviour, the 2010 and 2006 eruptive products are remarkably similar in terms of bulk rock
and melt inclusion major element concentrations. In addition, volatile contents of both melt inclusions and groundmass glass are also similar. The continuous range of melt
inclusion re-equilibration at depths <10 km suggests that
the melt region, at these depths at least, is interconnected
and not formed by isolated chambers. Dome fragments
from both eruptions reveal evidence of an exsolved brine
phase prior to eruption, with melt inclusions from these
clasts enriched in Li and with “buffered” Cl concentrations. In addition, melt inclusions from the 2010 white
pumice formed during explosive activity and which equilibrated during closed-system degassing, are enriched in B,
which was stabilised in the melt due to the presence of
the brine phase. Clinopyroxene host phenocrysts from the
white pumice crystallised at greater depths (up to 20 km)
compared to those erupted during other stages of the 2010
eruption or those from 2006. The transition from effusive
dome-forming to explosive (sub-plinian) behaviour during the 2010 eruption was triggered by an influx of magma
from depth which increased the overpressure and “overwhelmed” the system. This was further increased by a
closed-system degassing regime, with exsolved volatiles
staying in the system. Careful melt inclusion analysis,
1061
utilising various complementary techniques including
SIMS and ATR micro-FTIR revealed heterogeneous CO2
distribution in some melt inclusions, thought to have
formed during post-entrapment modifications, and highlights the need for caution in melt inclusion studies. This
work emphasizes the influence of magmatic flux, and magmatic degassing in controlling the eruptive style at Merapi
volcano. Variations in these factors will serve to modulate
future activity, controlling whether eruptions will be effusive and dome-forming or more explosive.
Acknowledgments We are grateful to staff at the Edinburgh Ion
Microprobe Facility, especially to Cees-Jan de Hoog, for technical assistance during SIMS analysis as well as valuable suggestions
regarding data interpretation, and to Richard Hinton for advice. Many
thanks to Jake Lowenstern (US Geological Survey, Menlo Park) for
kindly carrying out ATR micro-FTIR analysis, as well as for useful
discussions. Thanks go to Andy Tindle (The Open University), Chiara Petrone and Iris Buisman (University of Cambridge) for help with
electron microprobe analyses, as well as to Bertrand Lézé (University
of East Anglia) for technical support with the SEM. Peter Greatbatch
and David Wilde (Keele University) are acknowledged for assistance in polishing resin blocks. Adam Jeffery (Keele University) is
acknowledged and thanked for valuable advice regarding thermobarometric techniques. We thank Jean-Christophe Komorowski (Institut
de Physique du Globe de Paris, Université Paris Diderot) for kindly
sharing a base map for Fig. 1 and for thought-provoking discussions
in the field. Sylvain Charbonnier and Aurélie Germa (University of
South Florida) are thanked for their assistance in collecting samples
in the field. We appreciate the constructive reviews by Val Troll and
an anonymous reviewer, as well the Editor, Jon Blundy, whose comments helped to clarify the paper. This work has been supported by
the Natural Environment Research Council (NERC) through Urgency
grant NE/I029927/1, as well as a NERC studentship to KP (grant
number NE/H524506/1).
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s)
and the source are credited.
References
Aiuppa A, Baker DR, Webster JD (2009) Halogens in volcanic systems. Chem Geol 263:1–18
Allard P, Métrich N, Sabroux J-C (2011) Volatile and magma supply
to standard eruptive activity at Merapi volcano, Indonesia. Geophys Res Abstr 13:EGU2011-13522
Andreastuti SD, Alloway BV, Smith IEM (2000) A detailed
tephrostratigraphic framework at Merapi volcano, Central Java,
Indonesia: implications for eruption predictions and hazards
assessment. J Volcanol Geoth Res 100:51–67
Atlas ZD, Dixon JE, Sen G, Finny M, Lillian A, Pozzo MD (2006)
Melt inclusions from Volcán Popocatépetl and Volcán de
Colima, Mexico: melt evolution due to vapor-saturated crystallization during ascent. J Volcanol Geoth Res 153:221–240
Baker DR (2008) The fidelity of melt inclusions as a record of melt
composition. Contrib Miner Petrol 156:377–395
Balcone-Boissard H, Villemant B, Boudon G (2010) Behaviour of
halogens during the degassing of felsic magmas. Geochem
Geophys Geosyst 11:Q09005
13
1061
Page 22 of 25
Beauducel F, Cornet FH (1999) Collection and three- dimensional
modeling of GPS and tilt data at Merapi volcano, Java. J Geophys Res 104:725–736
Berlo K, Blundy J, Turner S, Cashman K, Hawkesworth C, Black S
(2004) Geochemical precursors to volcanic activity at Mount St.
Helens, USA. Science 306:1167–1169
Berlo K, Stix J, Roggensack K, Ghaleb B (2012) A tale of two magmas, Fuego, Guatemala. Bull Volc 74:377–390
Blundy J, Cashman K (2005) Rapid decompression-driven crystallisation recorded by melt inclusions from Mount St. Helens volcano. Geology 33:793–796
Blundy J, Cashman K (2008) Petrologic reconstruction of magmatic
system variables and processes. Rev Mineral Petrol 69:179–239
Blundy J, Cashman KV, Berlo K (2008) Evolving magma storage
conditions beneath Mount St. Helens inferred from chemical
variations in melt inclusions from the 1980–1986 and current
(2004–2006) eruptions. In: Sherrod DR, Scott WE, Stauffer PH
(Eds) A volcano rekindled: the renewed eruption of Mount St.
Helens, 2004–2006. US Geol Surv Prof Pap 1750, pp. 755–790
Blundy J, Cashman KV, Rust A, Witham F (2010) A case for CO2rich arc magmas. Earth Planet Sci Lett 290:289–301
Borisova AY, Martel C, Gouy S, Pratomo I, Sumarti S, Toutain J-P,
Bindeman IN, de Parseval P, Metaxian J-P, Surono (2013)
Highly explosive 2010 Merapi eruption: evidence for shallowlevel crustal assimilation and hybrid fluid. J Volcanol Geoth Res
261:193–208
Boudreau AE (1999) PELE-a version of the MELTS software program for the PC platform. Comput Geosci 25:201–203
Brodsky EE, Sturtevant B, Kanamori H (1998) Earthquakes, volcanoes, and rectified diffusion. J Geophys Res Solid Earth
103:23827–23838
Bucholz CE, Gaetani GA, Behn MD, Shimizu N (2013) Post-entrapment modification of volatiles and oxygen fugacity in olivinehosted melt inclusions. Earth Planet Sci Lett 374:145–155
Budi-Santoso A, Lesage P, Dwiyono S, Sumarti S, Surono, Subandriyo, Jousset P, Metaxian J-P (2013) Analysis of the seismic
activity associated with the 2010 eruption of Merapi Volcano,
Java. J Volcanol Geoth Res 261:153–170
Camus G, Gourgoud A, Mossand-Berthommier P-C, Vincent P-M
(2000) Merapi (Central Java, Indonesia): an outline of the
structural and magmatological evolution, with special emphasis to the major pyroclastic events. J Volcanol Geoth Res
100:139–163
Carroll MR, Webster JD (1994) Solubilities of sulfur, noble gases,
nitrogen, chlorine and fluorine in magmas. In: Carroll MR, Holloway JR (eds) Volatiles in magmas. Rev Mineral Geochem
30:231–279
Chadwick JP, Troll VR, Ginibre C, Morgan D, Gertisser R, Waight
TE, Davidson JP (2007) Carbonate assimilation at Merapi Volcano, Java, Indonesia: insights from crystal isotope stratigraphy.
J Petrol 48:1793–1812
Chadwick JP, Troll VR, Waight TE, van der Zwan FM, Schwarzkopf
LM (2013) Petrology and geochemistry of igneous inclusions in
recent Merapi deposits: a window into the sub-volcanic plumbing system. Contrib Miner Petrol 165:259–282
Charbonnier SJ, Gertisser R (2008) Field observations and surface
characteristics of pristine block-and-ash flow deposits from the
2006 eruption of Merapi volcano, Java, Indonesia. J Volcanol
Geoth Res 177:971–982
Charbonnier SJ, Germa AM, Connor CB, Gertisser R, Preece K,
Komorowski J-C, Lavigne F, Dixon TH, Connor LJ (2013)
Evaluation of the impact of the 2010 pyroclastic density currents at Merapi volcano from high-resolution satellite imagery
analysis, field investigations and numerical simulations. J Volcanol Geoth Res 261:295–315
13
Contrib Mineral Petrol (2014) 168:1061
Chen Y, Provost A, Schiano P, Cluzel N (2011) The rate of water
loss from olivine-hosted melt inclusions. Contrib Miner Petrol
162:625–636
Clocchiatti R, Joron JL, Kerinec F, Treuil M (1982) Quelques données préliminaires sur la lave du dôme actuel du volcan Mérapi
(Java, Indonésie) et sur ses enclaves. CR Acad Sci Paris
295:817–822
Costa F, Andreastuti S, de Maisonneuve CB, Pallister JS (2013) Petrological insights into the storage conditions, and magmatic
processes that yielded the centennial 2010 Merapi explosive
eruption. J Volcanol Geoth Res 261:209–235
Dahren B, Troll VR, Andersson UB, Chadwick JP, Gardner MF, Jaxybulatov K, Koulakov I (2012) Magma plumbing beneath Anak
Krakatau volcano, Indonesia: evidence for multiple magma
storage regions. Contrib Miner Petrol 163:631–651
Danyushevsky LV, Plechov P (2011) Petrolog 3: integrated software
for modeling crystallization processes. Geochem Geophys Geosyst 12:Q07021
Danyushevsky LV, McNeill AW, Sobolev AV (2002) Experimental
and petrological studies of melt inclusions in phenocrysts from
mantle-derived magmas: an overview of techniques, advantages
and complications. Chem Geol 183:5–24
Davis M, Koenders MA, Petford N (2007) Vibro-agitation of chambered magma. J Volcanol Geoth Res 167:24–36
Deegan FM, Troll VR, Freda C, Misiti V, Chadwick JP, McLeod CL,
Davidson JP (2010) Magma-carbonate interaction processes
and associated CO2 release at Merapi volcano, Indonesia:
insights from experimental petrology. J Petrol 51:1027–1051
Eichelberger JC (1995) Silicic volcanism: ascent of viscous magmas
from crustal reservoirs. Annu Rev Earth Planet Sci 23:41–63
Eichelberger JC, Carrigan CR, Westrich HR, Price RH (1986) Nonexplosive silicic volcanism. Nature 323:598–602
Fulignati P, Marianelli P (2007) Tracing volatile exsolution within
the 472 AD “Pollena” magma chamber of Vesuvius (Italy) from
melt inclusion investigation. J Volcanol Geoth Res 161:289–302
Gaetani GA, O’Leary JA, Shimizu N, Bucholz CE, Newville M
(2012) Rapid reequilibration of H2O and oxygen fugacity in
olivine-hosted melt inclusions. Geology 40:915–918
Gauthier P-J, Condomines M (1999) 210Pb-226Ra radioactive disequilibria in recent lavas and radon degassing: inferences on magma
chamber dynamics at Stromboli and Merapi volcanoes. Earth
Planet Sci Lett 172:111–126
Gertisser R (2001) Gunung Merapi (Java, Indonesien): Eruptionsgeschichte und Magmatische Evolution eines Hochrisiko-Vulkans.
Ph.D. Thesis, Universität Freiburg, Germany
Gertisser R, Keller J (2003) Trace element and Sr, Nd, Pb, and O isotope variations in medium-K and high-K volcanic rocks from
Merapi volcano, Central Java, Indonesia: evidence for the
involvement of subducted sediments in Sunda Arc magma genesis. J Petrol 44:475–489
Gertisser R, Charbonnier SJ, Troll VR, Keller J, Preece K, Chadwick
JP, Barclay J, Herd RA (2011) Merapi (Java, Indonesia): anatomy of a killer volcano. Geol Today 27:57–62
Gertisser R, Charbonnier SJ, Keller J, Quidelleur X (2012a) The geological evolution of Merapi volcano, Central Java, Indonesia.
Bull Volc 74:1213–1233
Gertisser R, Cassidy NJ, Charbonnier SJ, Nuzzo L, Preece K (2012b)
Overbank block-and-ash flow deposits and the impact of valleyderived, unconfined flows on populated areas at Merapi volcano, Java, Indonesia. Nat Hazards 60:623–648
Gonnermann HM, Manga M (2005) Nonequilibrium magma degassing: results from modeling of the ca. 1340 A.D. eruption of
Mono Craters, California. Earth Planet Sci Lett 238:1–16
Hamilton W (1979) Tectonics of the Indonesian region. US Geol Surv
Prof Pap 1078:1–345
Contrib Mineral Petrol (2014) 168:1061
Hammer JE, Cashman KV, Voight B (2000) Magmatic processes
revealed by textural and compositional trends in Merapi dome
lavas. J Volcanol Geoth Res 100:165–192
Hauri E, Wang J, Dixon JE, King PL, Mandeville C, Newman S
(2002) SIMS analysis of volatiles in silicate glasses 1. Calibration, matrix effects and comparisons with FTIR. Chem Geol
183:99–114
Heinrich CA, Günther D, Audétat A, Ulrich T, Frischknecht R
(1999) Metal fractionation between magmatic brine and vapor,
determined by microanalysis of fluid inclusions. Geology
27:755–758
Hercule S, Ingrin J (1999) Hydrogen in diopside: diffusion, kinetics of extraction-incorporation, and solubility. Am Mineral
84:1577–1587
Hervig RL, Moore GM, Williams LB, Peacock SM, Holloway JR, Roggensack K (2002) Isotopic and elemental partitioning of boron
between hydrous fluid and silicate melt. Am Mineral 87:769–774
Holloway JR, Blank JG (1994) Application of experimental results
to C–O–H species in natural melts. Rev Mineral Geochem
30:187–230
Humphreys MCS, Blundy JD, Sparks RSJ (2008) Shallow-level
decompression crystallisation and deep magma supply at Shiveluch Volcano. Contrib Miner Petrol 155:45–61
Jaupart C, Allègre C (1991) Gas content, eruption rate and instabilities of eruption regime in silicic volcanoes. Earth Planet Sci
Lett 102:413–429
Jeffery A, Gertisser R, Troll VR, Jolis EM, Dahren B, Harris C, Tindle
AG, Preece K, O’Driscoll B, Humaida H, Chadwick JP (2013)
The pre-eruptive magma plumbing system of the 2007–2008
dome-forming eruption of Kelut volcano, East Java, Indonesia.
Contrib Miner Petrol 166:275–308
Johnson ER, Wallace PJ, Cashman KV, Delgado Granados H, Kent
AJR (2008) Magmatic volatile contents and degassing-induced
crystallization at Volcán Jorullo, Mexico: implications for melt
evolution and the plumbing systems of monogenetic volcanoes.
Earth Planet Sci Lett 269:478–487
Jousset P, Budi-Santoso A, Jolly AD, Boichu M, Surono I, Dwiyono
S, Sumarti A, Hidayati S, Thierry P (2013) Signs of magma
ascent in LP and VLP seismic events and link to degassing: an
example from the 2010 explosive eruption at Merapi volcano,
Indonesia. Indonesia. J Volcanol Geoth Res 261:171–192
Kent AJR (2008) Melt inclusions in basaltic and related volcanic
rocks. In: Putirka KD, Tepley III FJ (eds) Minerals, inclusions
and volcanic processes. Rev Mineral Petrol 69:273–332
Kent AJR, Blundy J, Cashman KV, Cooper KM, Donnelly C, Pallister
JS, Reagan M, Rowe MC, Thornber CR (2007) Vapor transfer
prior to the October 2004 eruption of Mount St. Helens, Washington. Geology 35:231–234
Koleszar AM, Kent AJR, Wallace PJ, Scott WE (2012) Controls on
long-term explosivity at andesitic arc volcanoes: insights from
Mount Hood, Oregon. J Volcanol Geoth Res 219–220:1–14
Komorowski JC, Jenkins S, Baxter PJ, Picquout A, Lavigne F, Charbonnier S, Gertisser R, Preece K, Cholik N, Budi-Santoso A,
Surono (2013) Paroxysmal dome explosion during the Merapi
2010 eruption: processes and facies relationships of associated
high-energy pyroclastic density currents. J Volcanol Geoth Res
261:260–294
Le Cloarec M-F, Gauthier P-J (2003) Merapi Volcano, Central Java,
Indonesia: a case study of radionuclide behavior in volcanic
gases and its implications for magma dynamics at andesitic volcanoes. J Geophys Res 108(B5):2243
Le Guern F, Gerlach TM, Nohl A (1982) Field gas chromatograph
analyses of gases from a glowing dome at Merapi volcano, Java,
Indonesia, 1977, 1978, 1979. J Volcanol Geoth Res 14:223–245
Le Pennec J-L, Hermitte D, Dana I, Pezard P, Coulon C, Cochemé J-J,
Mulyadi E, Ollagnier F, Revest C (2001) Electrical conductivity
Page 23 of 25
1061
and pore-space topology of Merapi lavas: implications for the
degassing of porphyritic andesite magmas. Geophys Res Lett
28:4283–4286
Lloyd A, Plank T, Ruprecht P, Hauri E, Rose W (2013) Volatile loss
from melt inclusions in pyroclasts of differing sizes. Contrib
Miner Petrol 165:129–153
Lowenstern JB (1994) Chlorine, fluid immiscibility and degassing
in peralkaline magmas from Pantelleria, Italy. Am Mineral
79:353–369
Lowenstern JB (1995) Applications of silicate-melt inclusions to the
study of magmatic volatiles. In: Thompson JFH (ed) Magmas, fluids and ore deposits. Mineral Assoc Can Short Course 23:71–99
Lowenstern JB (2003) Melt inclusions come of age: volatiles, volcanoes, and Sorby’s legacy. In: De Vivo B, Bodnar RJ (eds)
Melt inclusions in volcanic systems: methods, applications and
problems. Developments in Volcanology, vol 5. Elsevier Press,
Amsterdam, pp 1–22
Lowenstern JB, Pitcher BW (2013) Analysis of H2O in silicate glass
using attenuated total reflectance (ATR) micro-FTIR spectroscopy. Am Mineral 98:1660–1668
Lube G, Cronin SJ, Thouret J-C, Surono (2011) Kinematic characteristics of pyroclastic density currents at Merapi and controls on
their avulsion from natural and engineered channels. Geol Soc
Am Bull 123:1127–1140
Mann CP, Wallace PJ, Stix J (2013) Phenocryst-hosted melt inclusions record stalling of magma during ascent in the conduit and
upper magma reservoir prior to vulcanian explosions, Soufrière
Hills volcano, Montserrat, West Indies. Bull Volc 75:1–16
Martel C, Pichavant M, Bourdier J-L, Traineau H, Holtz F, Scaillet
B (1998) Magma storage conditions and control of eruption
regime in silicic volcanoes: experimental evidence from Mt.
Pelée. Earth Planet Sci Lett 156:89–99
Melnik O, Sparks RSJ (1999) Nonlinear dynamics of lava dome
extrusion. Nature 402:37–41
Melnik O, Sparks RSJ (2005) Controls on conduit flow dynamics during lava dome building eruptions. J Geophys Res 110:B02209
Menard G, Vlastélic I, Rose-Koga EF, Piro J-L, Pin C (2013) Precise and accurate determination of boron concentration in silicate rocks by direct isotope dilution ICP-MS: insights into the
B budget of the mantle and B behavior in magmatic systems.
Chem Geol 354:139–149
Métrich N, Rutherford MJ (1992) Experimental study of chlorine
behaviour in hydrous silicic melts. Geochim Cosmochim Acta
56:607–616
Murphy MD, Sparks RSJ, Barclay J, Carroll MR, Brewer TS (2000)
Remobilization of andesite magma by intrusion of mafic
magma at the Soufriere Hills Volcano, Montserrat, West Indies.
J Petrol 41:21–42
Nadeau O, Williams-Jones AE, Stix J (2010) Sulphide magma as a
source of metals in arc-related magmatic hydrothermal ore fluids. Nat Geosci 3:501–505
Nadeau O, Williams-Jones AE, Stix J (2013) Magmatic-hydrothermal
evolution and devolatilization beneath Merapi volcano, Indonesia. J Volcanol Geoth Res 261:50–68
Newhall CG, Bronto S, Alloway B, Banks NG, Bahar I, del Marmol MA, Hadisantono RD, Holcomb RT, McGeehin J, Miksic JN, Rubin M, Sayudi SD, Sukhyar R, Andreastuti S, Tilling RI, Torley R, Trimble D, Wirakusumah AD (2000) 10,000
Years of explosive eruptions of Merapi Volcano, Central Java:
archaeological and modern implications. J Volcanol Geoth Res
100:9–50
Newman S, Lowenstern JB (2002) VOLATILECALC: a silicate meltH2O–CO2 solution model written in Visual Basic for excel.
Comput Geosci 28:597–604
Nielsen RL, Drake MJ (1979) Pyroxene-melt equilibria. Geochim
Cosmochim Acta 43:1259–1272
13
1061
Page 24 of 25
Nimis P (1995) A clinopyroxene geobarometer for basaltic systems based
on crystal-structure modelling. Contrib Miner Petrol 121:115–125
Nimis P (1999) Clinopyroxene geobarometry of magmatic rocks.
Part 2. Structural geobarometers for basic to acid, tholeiitic
and mildly alkaline magmatic systems. Contrib Miner Petrol
135:62–74
Pallister JS, Schneider DJ, Griswold JP, Keeler RH, Burton WC,
Noyles C, Newhall CG, Ratdomopurbo A (2013) Merapi 2010
eruption—chronology and extrusion rates monitored with satellite radar and used in eruption forecasting. J Volcanol Geoth
Res 261:144–152
Papale P, Moretti R, Barbato D (2006) The compositional dependence
of the saturation surface of H2O + CO2 fluids in silicate melts.
Chem Geol 229:78–95
Pichavant M (1981) An experimental study of the effect of boron on a
water saturated haplogranite at 1 Kbar vapour pressure. Contrib
Miner Petrol 76:430–439
Portnyagin M, Almeev R, Matveev S, Holtz F (2008) Experimental
evidence for rapid water exchange between melt inclusions in
olivine and host magma. Earth Planet Sci Lett 272:541–552
Preece KJ (2014) Transitions between effusive and explosive activity
at Merapi volcano, Indonesia: a volcanological and petrological
study of the 2006 and 2010 eruptions. Ph.D. Thesis. University
of East Anglia, UK
Preece K, Barclay J, Gertisser R, Herd R (2011) Petrological evidence
of magma storage, ascent and extrusion at Merapi volcano,
Java, Indonesia. Geophys Res Abstr 13:EGU2011-10518
Preece K, Barclay J, Gertisser R, Herd RA (2013) Textural and
micro-petrological variations in the eruptive products of the
2006 dome-forming eruption of Merapi volcano, Indonesia:
implications for sub-surface processes. J Volcanol Geoth Res
261:98–120
Putirka KD (2008) Thermometers and barometers for volcanic systems. Rev Mineral Geochem 69:61–120
Putirka KD, Mikaelian H, Ryerson F, Shaw H (2003) New clinopyroxene–liquid thermobarometers for mafic, evolved, and
volatile-bearing lava compositions, with applications to lavas
from Tibet and the Snake River Plain, Idaho. Am Mineral
88:1542–1554
Ratdomopurbo A (1995) Etude sismologique du volcan Merapi et
Formation du dome de 1994. PhD Thesis, Université Joseph
Fourier, Grenoble, France
Ratdomopurbo A, Poupinet G (2000) An overview of the seismicity
of Merapi volcano (Java, Indonesia), 1983–1994. J Volcanol
Geoth Res 100:193–214
Ratdomopurbo A, Beauducel F, Subandriyo J, Agung Nandaka IGM,
Newhall CG, Suharna Sayudi DS, Suparwaka H, Sunarta S
(2013) Overview of the 2006 eruption of Mt. Merapi. J Volcanol
Geoth Res 261:87–97
Reubi O, Blundy J, Varley NR (2013) Volatiles contents, degassing and crystallisation of intermediate magmas at Volcan de
Colima, inferred from melt inclusions. Contrib Miner Petrol
165:1087–1106
Ridolfi F, Puerini M, Renzulli A, Menna M, Toulkeridis T (2008)
The magmatic feeding system of El Reventador volcano (SubAndean zone, Ecuador) constrained by texture, mineralogy
and thermobarometry of the 2002 erupted products. J Volcanol
Geoth Res 176:94–106
Ruprecht P, Bachmann O (2010) Pre-eruptive reheating during
magma mixing at Quizapu volcano and the implications for the
explosiveness of silicic are volcanoes. Geology 38:919–922
Saepuloh A, Koike K, Omura M, Iguchi M, Setiawan A (2010) SARand gravity change-based characterization of the distribution
pattern of pyroclastic flow deposits as Mt. Merapi during the
past 10 years. Bull Volc 72:221–232
13
Contrib Mineral Petrol (2014) 168:1061
Scandone R, Cashman KV, Malone SD (2007) Magma supply,
magma ascent and style of volcanic eruptions. Earth Planet Sci
Lett 253:513–529
Schatz OJ, Dolejš D, Stix J, Williams-Jones AE, Layne GD (2004)
Partitioning of boron among melt, brine and vapor in the system
haplogranite–H2O–NaCl at 800 °C and 100 MPa. Chem Geol
210:135–147
Schwarzkopf LM, Schmincke HU, Troll VR (2001) Pseudotachylite on
impact marks of block surfaces in block-and-ash flows at Merapi
volcano, Central Java, Indonesia. Int J Earth Sci 90:769–775
Self S, Gertisser R, Thordarson T, Rampino MR, Wolff JA (2004)
Magma volume, volatile emissions, and stratospheric aerosols
from the 1815 eruption of Tambora. Geophys Res Lett 31:L20608
Severs MJ, Azbej T, Thomas JB, Mandeville CW, Bodnar RJ (2007)
Experimental determination of H2O loss from melt inclusions
during laboratory heating: evidence from Raman spectroscopy.
Chem Geol 237:358–371
Shinohara H (1994) Exsolution of immiscible vapor and liquid phases
from a crystallizing silicate melt: implications for chlorine and
metal transport. Geochim Cosmochim Acta 58:5215–5221
Signorelli S, Vaggelli G, Romano C (1999) Pre-eruptive volatile
(H2O, F, Cl and S) contents of phonolitic magmas feeding the
3550-year old Avellino eruption from Vesuvius, southern Italy. J
Volcanol Geoth Res 93:237–256
Smyth HR, Hall R, Hamilton J, Kinny P (2005) East Java: Cenozoic
basins, volcanoes and ancient basement. In: Proceeding, Indonesian Petroleum Association, Thirteenth Annual Convention &
Exhibition, pp 251–266
Sparks RSJ (1997) Causes and consequences of pressurisation in lava
dome eruptions. Earth Planet Sci Lett 150:177–198
Stalder R, Skogby H (2003) Hydrogen diffusion in natural and synthetic orthopyroxene. Phys Chem Miner 30:12–19
Stix J, Layne GD (1996) Gas saturation and evolution of volatile
and light lithophile elements in the Bandelier magma chamber between two caldera-forming eruptions. J Geophys Res
101:25181–25196
Surono Jousset P, Pallister J, Boichu M, Buongiorno MF, Budisantoso A, Costa F, Andreastuti S, Prata F, Schneider D, Clarisse L,
Humaida H, Sumarti S, Bignami C, Griswold J, Carn S, Oppenheimer C, Lavigne F (2012) The 2010 explosive eruption of
Java’s Merapi volcano—A ‘100-year’ event. J Volcanol Geoth
Res 241–242:121–135
Suzuki Y, Yasunda A, Hokanishi N, Kaneko T, Nakada S, Fujii T
(2013) Syneruptive deep magma transfer and shallow remobilization during the 2011 eruption of Shinmoe-dake, Japan—Constraints from melt inclusions and phase equilibria experiments.
J Volcanol Geoth Res 257:184–204
Toutain J-P, Sortino F, Baubron J-C, Richon P, Surono, Sumarti S,
Nonell A (2009) Structure and CO2 budget of Merapi volcano
during inter-eruptive periods. Bull Volc 71:815–826
Troll VR, Hilton DR, Jolis EM, Chadwick JP, Blythe LS, Deegan FM,
Schwarzkopf LM, Zimmer M (2012) Crustal CO2 liberation
during the 2006 eruption and earthquake events at Merapi volcano, Indonesia. Geophys Res Lett 39:L11302
Troll VR, Deegan FM, Jolis EM, Harris C, Chadwick JP, Gertisser
R, Schwarzkopf LM, Borisova AY, Bindeman IN, Sumarti S,
Preece K (2013) Magmatic differentiation processes at Merapi
volcano: inclusion petrology and oxygen isotopes. J Volcanol
Geoth Res 261:38–49
van Bemmelen RW (1949) The geology of Indonesia, vol 1A, general
geology. Government Printing Office, The Hague
Vigouroux N, Wallace PJ, Kent AJR (2008) Volatiles in high-K magmas from the western Trans-Mexican Volcanic Belt: evidence
for fluid fluxing and extreme enrichment of the mantle wedge
by subduction processes. J Petrol 49:1589–1618
Contrib Mineral Petrol (2014) 168:1061
Villemant B, Boudon G (1998) Transition from dome-forming to plinian eruptive styles controlled by H2O and Cl degassing. Nature
392:65–69
Villemant B, Mouatt J, Michel A (2008) Andesitic magma degassing
investigated through H2O vapour-melt partitioning of halogens
at Soufrière Hills volcano, Montserrat, (Lesser Antilles). Earth
Planet Sci Lett 269:212–229
Vlastélic I, Menard G, Gannoun A, Piro J-L, Staudacher T, Famin V
(2013) Magma degassing during the April 2007 collapse of Piton
de la Fournaise: the record of semi-volatile trace elements (Li,
B, Cu, In, Sn, Cd, Re, Tl, Bi). J Volcanol Geoth Res 254:94–107
Voight B, Constantine EK, Siswowidjoyo S, Torley R (2000) Historical eruptions of Merapi volcano, Central Java, Indonesia, 1768–
1998. J Volcanol Geoth Res 100:69–138
Walter TR, Wang R, Zimmer M, Grosser H, Luehr B, Ratdomopurbo
A (2007) Volcanic activity influenced by tectonic earthquakes:
static and dynamic stress triggering at Mt, Merapi. Geophys
Res Lett 34:L05304
Webster JD (1997) Chloride solubility in felsic melts and the role of
chloride in magmatic degassing. J Petrol 38:1793–1807
Webster JD (2004) The exsolution of magmatic hydrosaline chloride
liquids. Chem Geol 210:33–48
Webster JD, Holloway JR, Hervig RL (1989) Partitioning of lithophile trace elements between topaz rhyolite melt and H2O and
H2O + CO2 fluids. Econ Geol 84:116–134
Page 25 of 25
1061
Webster JD, Mandeville CW, Goldoff B, Coombs ML, Tappen C
(2010) Augustine volcano—the influence of volatile components in magmas erupted A.D. 2006 to 2,100 years before
present. In: Power JA, Coombs ML, Freymueller JT (eds) The
2006 eruption of Augustine volcano, Alaska. US Geol Surv Prof
Pap 1769, pp 383–423
Wilson L, Sparks RSJ, Walker GPL (1980) Explosive volcanic eruptions—IV. The control of magma properties and conduit geometry on eruption column behaviour. Geophys J R Astron Soc
63:117–148
Wolf KJ, Eichelberger JC (1997) Syneruptive mixing, degassing, and
crystallization at Redoubt Volcano, eruption of December, 1989
to May 1990. J Volcanol Geoth Res 75:19–37
Woods AW, Koyaguchi T (1994) Transitions between explosive and
effusive eruptions of silicic magmas. Nature 370:641–644
Wright HM, Bacon CR, Vazquez JA, Sisson TW (2012) Sixty thousand years of magmatic volatile history before the calderaforming eruption of Mount Mazama, Crater Lake, Oregon.
Contrib Miner Petrol 164:1027–1106
Zhang Y, Stolper EM, Wasserburg GJ (1991) Diffusion of water in
rhyolitic glasses. Geochim Cosmochim Acta 55:441–456
Zimmer M, Erzinger J (2003) Continuous H2O, CO2, 222Rn and temperature measurements on Merapi Volcano, Indonesia. J Volcanol Geoth Res 125:25–38
13
View publication stats