Contrib Mineral Petrol (2013) 166:371–392 DOI 10.1007/s00410-013-0880-7
ORIGINAL PAPER
Small volume andesite magmas and melt–mush interactions at Ruapehu, New Zealand: evidence from melt inclusions Geoff Kilgour • Jon Blundy • Kathy Cashman Heidy M. Mader
•
Received: 1 November 2012 / Accepted: 12 April 2013 / Published online: 25 May 2013 Ó Springer-Verlag Berlin Heidelberg 2013
Abstract Historical eruptions from Mt. Ruapehu (New Zealand) have been small (\0.001 km3 of juvenile magma) and have often occurred without significant warning. Developing better modelling tools requires an improved understanding of the magma storage and transport system beneath the volcano. Towards that end, we have analysed the volatile content and major element chemistry of groundmass glass and phenocryst-hosted melt inclusions in erupted samples from 1945 to 1996. We find that during this time period, magma has been stored at depths of *2–9 km, consistent with inferences from geophysical data. Our data also show that Ruapehu magmas are relatively H2O-poor (\2 wt%) and CO2-rich (B1,000 ppm) compared to typical arc andesites. Surprisingly, we find that melt inclusions are often more evolved than their transporting melt (as inferred from groundmass glass compositions). Furthermore, even eruptions that are separated by less than 2 years exhibit distinct major element chemistry, which suggests that each eruption involved magma with a unique ascent history. Communicated by G. Moore.
Electronic supplementary material The online version of this article (doi:10.1007/s00410-013-0880-7) contains supplementary material, which is available to authorized users. G. Kilgour J. Blundy K. Cashman H. M. Mader School of Earth Sciences, University of Bristol, Wills Memorial Building, Bristol BS8 1RJ, UK G. Kilgour (&) Wairakei Research Centre, GNS Science, Taupo 3330, New Zealand e-mail:
[email protected] K. Cashman Department of Geological Sciences, 1272 University of Oregon, Eugene, OR 97403, USA
From these data, we infer that individual melt batches rise through, and interact with, crystal mush zones formed by antecedent magmas. From this perspective, we envision the magmatic system at Ruapehu as frequently recharged by small magma inputs that, in turn, cool and crystallise to varying degrees. Melts that are able to erupt through this network of crystal mush entrain (to a greater or lesser extent) exotic crystals. In the extreme case (such as the 1996 eruption), the resulting scoria contain melt inclusion-bearing crystals that are exotic to the transporting magma. Finally, we suggest that complex interactions between recharge and antecedent magmas are probably common, but that the small volumes and short time scales of recharge at Ruapehu provide a unique window into these processes. Keywords Andesite Volatile Melt inclusions Ruapehu Crystal mush Antecryst H2O CO2 Magma mixing
Introduction Magma erupted from andesitic volcanoes often records a complex history, and interactions between recharge and antecedent magmas and/or crystal mush zones are common (e.g., Nakamura 1995; Murphy et al. 1998; Nakagawa et al. 1999, 2002; Devine et al. 1998). Most studies that have highlighted this complexity, however, have focussed on moderate to large volume andesitic eruptions, where subtle, fine-scale interactions may be obscured by large recharge volumes. Mt. Ruapehu, New Zealand, a frequently active andesite volcano that has historically erupted very small magma volumes thus provides an interesting test case for imaging the complexity of subvolcanic magmatic processes. In addition, eruptions from Ruapehu have been
123
372
extremely difficult to predict (Sherburn et al. 1999), due to both the small volumes of magma involved in individual eruptions and the presence of an active hydrothermal system, which generates a rather noisy background seismic signal (Hurst 1998). Therefore, a secondary goal of our study is to improve our knowledge of both the magma plumbing and magma transport systems beneath Ruapehu. Pre-eruptive conditions of magma storage and recharge can be determined by analysing the compositions of coexisting crystals (phenocrysts and microlites) and glasses (groundmass and crystal-hosted melt inclusions). Together these data provide information on storage temperature and pressure, magma–magma interactions, and the mixing of magma and crystal mush zones (e.g., Roedder 1984; Blundy and Cashman 2005; Liu et al. 2006). When combined with airborne gas chemistry monitoring (e.g., Christenson et al. 2010), the volatile content of the glass phases also provides information on the fate of the main volatile components during degassing. We present volatile and major element chemistry of Ruapehu melt inclusions and groundmass glass from more than 50 years of eruptions (1945–1996). We use samples of scoria and lava from every historical magmatic eruption during this period to track changes in magma composition through time. Importantly, we provide new evidence that recent eruptions have been driven by small batches of recharge magma that have mingled with, entrained crystals from, and remobilised regions of shallow-stored, antecedent magma. We analysed representative samples (from GNS Science, New Zealand rock archives) from six magmatic eruptions at Ruapehu: 1945, 1969, 1971, 1977, 1995, and 1996. In this paper, we identify eruptions by year, not by the specific eruption date. This is particularly relevant for the 1995–1996 eruptive episode, as we were not able to distinguish between eruptions in September and October 1995 and the June and July 1996 eruptions (c.f. Nakagawa et al. 1999, 2002). Similarly, we were unable to identify a unique eruption date for the 1945 and 1971 samples. Thus, our samples are from eruptions (1) occurring between March and December 1945 [1945]; (2) on 22 June 1969 [1969]; (3) from April to July 1971 [1971]; (4) on 2 November 1977 [1977]; (5) from September to October 1995 [1995]; and (6) from June to July 1996 [1996] (Table 1). Geological background At 2,797 mASL, Mt. Ruapehu is the largest and most active andesitic stratovolcano in New Zealand. It is located at the southern end of the Taupo Volcanic Zone (TVZ) and the summit area is covered by small permanent glaciers and snowfields (Fig. 1). Ruapehu has been active for at least 200 ka (Gamble et al. 2003). Holocene activity has
123
Contrib Mineral Petrol (2013) 166:371–392
involved a series of overlapping craters (Hackett and Houghton 1989), with the most recent activity confined to the southernmost crater, currently occupied by the warm (15–40 °C) and acidic (0–1 pH) Crater Lake (Hackett, 1985). Historical activity has consisted primarily of frequent small phreatic (e.g., Kilgour et al. 2007, 2010) and phreatomagmatic eruptions through Crater Lake (Healy et al. 1978; Nairn et al. 1979; Houghton et al. 1987). Larger magmatic events occurred in 1945 (Oliver 1945; Reed 1945; Beck 1950; Gregg 1960), 1969 (Healy et al. 1978), 1975 (Houghton et al. 1987), and 1995–96 (Houghton et al. 1996; Bryan and Sherburn 1999; Nakagawa et al. 1999, 2002; Johnston et al. 2000). Eruptive activity typically involves surtseyan jets of lake water and steam accompanied by base surges and ballistic fall-out up to 2 km from the vent. These events are usually confined to the summit area (Houghton et al. 1987; Kilgour et al. 2010), but rare strombolian activity and more widespread sub-plinian to plinian ash falls also occur (Donoghue 1991; Pardo et al. 2011). Historical eruption narrative Volcanic activity at Ruapehu has been observed and recorded since c. 1850 (Gregg 1960; Hackett and Houghton 1989; Reed 1945). More than 40 eruptions have been reported since 1945, covering a range of eruptive styles and sizes (B. Scott pers. comm.). Our samples derive from magmatic eruptions that ejected juvenile scoria and ash onto the summit plateau and beyond. Eruptive activity at Ruapehu between March and July 1945 (Reed 1945) occurred prior to the installation of volcanic monitoring systems. The 1945 eruption initiated with explosive magmatic activity that produced high steam plumes and dispersed ash to c. 200 km from the vent (Johnston et al. 2000). A series of lava domes were then constructed and partially destroyed, presumably through mass wasting and sector collapse (Reed 1945; Beck 1950). The total volume of erupted magma is estimated at *0.1 km3 (Johnston et al. 2000). Between 1945 and the next magmatic eruption in 1969, a limited seismic network was installed and Crater Lake temperature measurements and chemical sampling were initiated. A moderately large eruption (0.9 9 106 m3) on 22 June 1969 ejected older lava lithics, lake sediments, and *5 vol % juvenile scoria (Healy et al. 1978). We have calculated the bulk volume of juvenile magma to be 4.5 9 104 m3, with a dense rock equivalent (DRE) of 1.7 9 104 m3. This event was preceded by only limited seismic precursors and a regular Crater Lake heating– cooling cycle considered to reflect normal activity (Healy et al. 1978).
Contrib Mineral Petrol (2013) 166:371–392
373
Table 1 Bulk XRF data of scoria from selected historical eruptions from Ruapehu Eruption year Sample number
1945 1945A
1969 22/5/69-1
1971 1971A
1977 1977A
1977 1977-8
1995 31195B
1995 31195A
1995 71195-04
1995 161195-34
1996a 130896
57.57
Major elements (wt %) SiO2
60.16
61.10
58.50
60.44
58.88
57.68
57.90
58.02
61.54
Al2O3
16.87
15.79
16.24
16.43
16.09
16.30
16.33
16.24
16.58
16.40
Fe2O3
5.80
6.22
6.94
7.65
7.17
7.36
7.33
7.25
6.36
7.43
MnO
0.09
0.09
0.10
0.07
0.10
0.11
0.11
0.11
0.08
0.12
MgO
3.55
4.00
5.04
3.62
4.73
5.45
5.42
5.31
3.61
5.37
CaO Na2O
6.02 3.47
6.10 3.36
7.31 3.30
7.08 2.55
7.30 3.11
7.75 3.25
7.74 3.25
7.67 3.22
6.30 2.88
7.87 3.30
K2O
1.67
2.15
1.54
1.39
1.48
1.36
1.38
1.43
1.64
1.32
TiO2
0.66
0.65
0.62
0.66
0.64
0.62
0.62
0.62
0.67
0.64
P2O5
0.14
0.16
0.14
0.14
0.14
0.13
0.14
0.14
0.14
0.10
LOI
1.49
0.20
0.05
-0.12
0.20
-0.04
-0.07
-0.08
0.00
0.00
Total
99.92
99.81
99.77
99.92
99.83
99.98
100.15
99.93
99.80
100.00
Trace elements (ppm) As
4
6
3
2
2
1
2
3
3
4
Ba
384
422
358
353
346
307
320
330
390
352
Ce
27
30
25
39
30
26
21
28
23
22
Cr
48
78
117
97
115
105
113
95
92
114
Cu
64
59
68
86
62
66
68
66
72
72
Ga
20
17
17
18
17
17
16
16
17
17
La
18
14
19
22
\5
\5
11
\5
\5
10
Nb Ni
\1 22
7 41
\1 57
\1 50
\1 57
\1 65
\1 62
\1 57
\1 45
4 59
Pb
11
13
12
12
13
13
12
10
14
10
Rb
60
83
54
51
53
48
48
50
56
47
Sc
19
19
23
22
24
24
28
21
21
26
Sr
280
243
261
252
258
262
263
259
269
264
Th
7
9
5
5
5
5
4
4
4
4
U
2
3
3
2
2
2
\1
2
2
1
V
152
153
176
184
178
182
189
175
173
196
Y
20
21
20
19
18
18
17
19
17
19
Zn
65
61
65
64
66
67
68
68
67
68
Zr
134
154
113
121
115
104
103
105
126
106
a
The 1996 sample is taken from Gamble et al. (1999)
During the 3 months prior to the 1971 eruption, Crater Lake temperatures rose from ca. 25 to 55 °C, and the abundance and amplitude of volcanic earthquakes (signalling magma movement) increased (Sherburn et al. 1999). There also appears to have been an increase in volcanic earthquakes over a few weeks before first eruption on 3 April 1971. The 1971 eruption was a small, phreatomagmatic event that was confined to the summit plateau. A phreatomagmatic eruption on 2 November 1977 was small compared to eruptions of 1945, 1969, and 1975 (samples of the 1975 scoria were not available for this study). The 1977 eruption occurred one to 2 weeks after the temperature of
Crater Lake increased from c. 19 to 30 °C, yet without any coincident volcanic earthquakes (Sherburn et al. 1999). The absence of volcanic earthquakes could reflect the small amount of fresh magma injected to shallow levels, as well as the rather sparse seismometer network installed at the time. Although limited field data exist for either eruption, they appear to have been similar in size to a well-characterised eruption in 2007 (based on photographs of the summit area after each eruption). Juvenile scoria was erupted during those events, and we have again assigned a value of 5 wt% of the bulk deposit as juvenile material. Therefore, if we take a similar bulk volume to that of the 2007 event (Kilgour et al.
123
374
Contrib Mineral Petrol (2013) 166:371–392
a
b
Fig. 1 a Location map of Mt Ruapehu at the southern end of the Taupo Volcanic Zone. b Photograph of Ruapehu’s summit plateau and Crater Lake. This photograph is taken from the South towards the
North with Ngauruhoe in the background (GNS Science archive image). Approximate locations for the two main vents (North and Central) beneath the lake (Christenson et al. 2010) are marked
2010), the amount of magma erupted in each eruption in 1971 and 1977 is of the order 1 9 104 m3 (DRE). A period of relative quiescence occurred from 1988 to 1994. However, volcanic tremor (c. 2 Hz) at Ruapehu rose to, and was maintained at, high levels starting in the early 1990s. During this time period, Crater Lake heating cycles were often punctuated by steam explosions, although there were no signs of significant (magmatic) eruptive activity (Sherburn et al. 1999). A period of rapid heating of Crater Lake occurred in November 1994. During this time, the lake temperature increased from c. 19 to 50 °C in 1 month, and numerous phreatic eruptions occurred within Crater Lake. However, this activity was not interpreted as indicative of magmatic injection to shallow levels (Christenson 2000). The lake heating cycle appeared to be over by April 1995, but was immediately followed by another heating event. More phreatic eruptions occurred from April to early July. By this time, a steady increase in the Mg/Cl ratio of the lake water suggested that magma was being injected to shallow levels and interacting with the hydrothermal system (Christenson 2000). Moderate levels of seismicity and Crater Lake temperatures, combined with and the absence
of deformation determined from theodolite levelling surveys (B. Scott 2011 pers. comm.), indicated that the amount of magma driving the Mg/Cl ratio changes was very small. On 18 September 1995, a small phreatomagmatic eruption occurred with few seismic precursors (Bryan and Sherburn 1999); following this event, a large phreatomagmatic eruption developed. A lull in activity from November 1995 to June 1996 allowed direct measurements of the main fumaroles within the inner crater (Christenson 2000). Tremor increased to pre-September 1995 levels on 15–16 June 1996 and on 17 June, pulsating, phreatomagmatic eruptions graded into a more continuous eruption, with plumes reaching 8.5 km asl (Prata and Grant 2001). The 1995–1996 eruptions were approximately two orders of magnitude larger than the eruptions of the 1960s and 1970s. Tephra dispersal mapping suggests DRE volumes for the 11–14 October 1995 and the 17–18 June 1996 eruptions of 3 9 107 m3 and 6 9 106 m3, respectively (Cronin et al. 1998; Fig. 2). Between 1996 and 2007, Ruapehu remained relatively quiet except for one small phreatic event on 4 October 2006 (Kilgour et al. 2007; Mordret et al. 2010). Then on 25 September 2007, after *9 min of precursory seismic
123
Contrib Mineral Petrol (2013) 166:371–392
375
Fig. 2 Variations in eruption volume, magmatic temperature, and mineral chemistry between historical eruptions at Ruapehu. Dense rock equivalent (DRE) volumes for 1945 (Johnston et al. 2000), 1969 (Healy et al. 1978), 1971 and 1977 (assuming a similar volume erupted in 2007 from Kilgour et al. (2010)), and 1995–1996 (Cronin et al. 1998). Magmatic temperatures were calculated using the plagioclase-liquid (open triangles) and clinopyroxene-liquid (filled
circles) geothermometers of Putirka (2008), and Fe–Ti oxides (crosses) using LePage (2003) for 1969 only. Plagioclase core, rim, and microlite (Plag C, R, M, respectively) anorthite content (An %), and clinopyroxene (cpx) and orthopyroxene (opx) magnesian number (Mg#) are also shown. Plagioclase rim MgO content is expressed as a range (open rectangles). Average values for plagioclase (An) and pyroxene (Mg#) are denoted by an ‘‘x’’
signals (Jolly et al. 2010), a short-lived phreatic eruption from Crater Lake created a northerly directed blast that deposited ballistic blocks and surtseyan jets (Kilgour et al. 2010). Mapping of the deposits within the summit plateau yielded a total volume of *3 9 105 m3. The presence of juvenile magma in the 2007 eruption is equivocal and so we have not included these samples in our study.
The synthesis of data from historical eruptions presented above shows that recent eruptions of Ruapehu are small but frequent, and may occur without warning. The volumes of magma involved in priming the magmatic system for eruption also appear to be small. Here, we use detailed compositional analyses of samples from these eruptions, in the context of this eruptive narrative, to improve our
123
376
understanding of the physical conditions of magma storage, recharge, and eruption at Ruapehu.
Methods All samples were analysed for bulk rock chemistry by X-ray fluorescence (XRF) for major and trace elements at SpectraChem, Wellington (NZ). Samples were then prepared for the analysis of phase compositions by lightly crushing individual scoria clasts for each eruption (we combined two scoria clasts from 1995) and hand-picking phenocrysts for mounting in epoxy resin onto glass slides. The preparation of crystal-separate thin sections, rather than grain mounts, minimised the carbon background from excess epoxy during secondary ion mass spectrometry (SIMS) analysis of volatiles. Each crystal separate was polished to ca. 100 lm thickness to expose melt inclusions. We obtained back-scattered electron (BSE) images using a Hitachi S-3500 N SEM at the University of Bristol at 15 kV and at a working distance of *15 mm. These images were used to map melt inclusions trapped in plagioclase, orthopyroxene, and clinopyroxene phenocrysts. Approximately, 70 % of the melt inclusions were larger than 25 lm, the minimum spot size of SIMS analyses. SIMS analyses of the volatiles dissolved in melt inclusions were carried out on Au-coated grain mounts using a Cameca IMS-4f ion microprobe at the University of Edinburgh. We constructed working curves 1H/30Si versus H2O and 12C/30Si versus CO2 (e.g., Blundy and Cashman 2008) using a total of nine rhyolitic glass standards that range from 0.15 to 4.1 wt% H2O and 0–2,860 ppm CO2. Standards were run at three intervals during each day to account for drift in the analyses. 1H was analysed at low mass resolution. For 12C, interference from 24Mg2? is significant at the relatively high MgO contents (2.5 wt%) of Ruapehu melt inclusions. Separation of the Mg and CO2 spectra thus required us to conduct our SIMS measurements at high mass resolution, and first analyse melt inclusions for CO2 followed by H2O. We did not analyse the 1945 lava sample via SIMS due to concerns over H2O diffusion out of the melt inclusions in such slowly cooled samples (e.g., Hauri 2002). Electron probe micro-analysis (EPMA) can damage hydrous silicate glass. For this reason, we measured the major element composition of the inclusions by EPMA after SIMS analysis was completed (e.g., Blundy et al. 2010). EPMA was conducted using a CAMECA SX-100 five-spectrometer wavelength dispersive spectrometry (WDS) instrument at the University of Bristol. Melt inclusions and groundmass glasses were analysed using a 15-kV accelerating voltage, 4 nA beam current with a defocused 10 lm beam, with K and Na being analysed first
123
Contrib Mineral Petrol (2013) 166:371–392
to reduce the effects of alkali migration (e.g., Humphreys et al. 2006). Plagioclase and pyroxene crystals were analysed using a 20-kV accelerating voltage, 10 nA beam current and a focussed beam. Calibration used a selection of mineral and oxide standards. Data were reduced using the ZAF procedure.
Results Petrography of Ruapehu samples Samples available for this study include a lava sample from 1945 and scoria from 1969, 1971, 1977, 1995, and 1996. The scoriae are vesicular, porphyritic, and microlite-rich (with the exception of 1969), similar to pre-historic Ruapehu scoria (Hackett 1985; Graham and Hackett 1987; Gamble et al. 1999). Here, we use the term phenocryst to signify a crystal that is significantly larger (*1–2 mm) than microlites present in the groundmass, independent of its origin. We use antecryst to mean a crystal that demonstrably grew within a different magma than its current host (i.e., exotic), and cognate to mean a crystal that grew from and erupted with its host magma. Phenocrysts of plagioclase dominate the mineral assemblage with lesser amounts of clinopyroxene and orthopyroxene (Table 2). However, there is no clear crystallisation sequence; pyroxene is often found within plagioclase phenocrysts, while plagioclase inclusions are also seen in pyroxene crystals. The implication is that these three phases precipitated cotectically. Of the minor phases, magnetite is present in all samples as small blocky grains up to 30 lm across. Ilmenite is absent in all samples except in 1969 scoria, in agreement with previous observations (Nakagawa et al. 1999; Price et al. 2012). Hornblende is absent in all samples; this is also the case for all but one lava exposed on the edifice of the volcano (Hackett 1985). Microlites (crystals \100 lm across) of plagioclase, clinopyroxene, orthopyroxene, magnetite, and rare ilmenite
Table 2 Representative crystallinity of Ruapehu scoria Eruption date Sample number
1945 1945F
1969 1969A
1971 1971A
1977 1977A
1995 031195B
1996a 57536
Phenocrysts (%)
40
29
33
31
26
36
Groundmass (%)
59
42
50
49
54
39
1
29
17
20
21
25
Vesicle (%) Phenocrysts Plagioclase
29
20
23
22
18
19
Clinopyroxene
6
4
5
4
3
8
Orthopyroxene
5
5
5
4
5
9
a
The 1996 sample is taken from Gamble et al. (1999)
Contrib Mineral Petrol (2013) 166:371–392
are present in the groundmass of most samples (c.f. Nakagawa et al. 1999). The bulk compositions of scoria and lava samples from 1945 to 1996 have been reported by Gamble et al. (1999) and Nakagawa et al. (1999) and pre-historical compositions by Price et al. (2012). We have conducted further bulk rock XRF analyses on samples from historical eruptions, which agree well with data from Gamble et al. (1999) (Table 1). There are no discernible systematic changes in bulk chemistry, which ranges from *57 to 64 wt% SiO2, mineral content or abundance, or isotopic composition through time (Gamble et al. 1999). This chemical monotony attests to a genetic link among all Ruapehu magmas. Interestingly, scoria erupted during the 1995–1996 eruptions span the entire range of the historical record (*58–62 wt% SiO2). Mineralogy Plagioclase phenocrysts are up to 4 mm across and are generally zoned from a calcic core (An55–82) to a sodic rim (An52–65) (Fig. 2). The anorthite (An) content of phenocryst rims is similar to that of plagioclase microlites (An50–65). Rare plagioclase phenocrysts are reversely zoned from *An52 to * An60. The MgO content of plagioclase is commonly used to monitor mafic inputs to the magmatic system (e.g. Hattori and Sato 1996). There is not much variation in MgO content among our samples, except for 1971 and 1995 samples, which show elevated MgO in some plagioclase phenocryst rims (up to 0.15 wt%; see also Nakagawa et al. 1999). Clinopyroxene phenocrysts are subhedral to euhedral and B3 mm in length. All analysed samples show the same
a
Fig. 3 a BSE image of a clinopyroxene phenocryst from the 1995 eruption of Ruapehu. Oscillatory zoning is common throughout all eruptions. The greyscale images highlight zones of relatively higher Fe (light grey) and Mg (dark grey) zones. Note the more diffuse boundary at the core compared to the sharp boundary at the rim.
377
compositional range (with a mean of Wo*42; En*43; Fs*15; Fig. 2). Oscillatory zoning of clinopyroxenes is evident in all Ruapehu scoria (e.g., Nakagawa et al. 1999, 2002). Normal zoning (Mg#53–77) is most common, although many clinopyroxene crystals (both phenocrysts and microlites) preserve a very thin (2–5 lm) Mg-rich outermost rim (Fig. 3). Melt inclusions are common as both small (\10 lm), glassy inclusions forming concentrically to the growth pattern and large ([50 lm), isolated inclusions near the core of the crystal. Orthopyroxene phenocrysts are euhedral to subhedral, relatively unzoned, and B4 mm in length, with a composition of Wo*3; En*60; Fs*37—enstatite (Fig. 2). Rare zoned (both normal and reverse; Mg#36–50) orthopyroxene crystals are found in all historical samples. Orthopyroxenehosted melt inclusions are less common than in clinopyroxene and are present as small (\ 10 lm) inclusions on the margins of the crystal. Larger ([30 lm) melt inclusions are uncommon in phenocrysts from all eruptions. Major element chemistry of groundmass glass The groundmass glasses of Ruapehu scoria span a wide compositional range from 58 to 78 wt% SiO2 (Table 3; Figs. 4a, 5, Supplementary Table 1). Glass compositions from individual Ruapehu eruptions can be distinguished from each other by means of major element binary plots. The least evolved glasses are found in 1995 and 1996 scoria (Fig. 5). Groundmass glasses from 1945 lava and 1969 scoria are significantly more evolved than all other historical samples; in the case of 1945, this is possibly due to slower cooling of the lava sample. Conversely, glasses from the 1971 and 1977 eruptions extend to lower SiO2
b
b Dashed black lines denote the boundary between relatively diffuse Fe- and Mg-rich zones. Arrow points to the *3 lm wide, dark grey rim (Mg-rich) on the outermost margin of the crystal, elsewhere interpreted as late-stage mixing (e.g., Saunders et al. 2012). Groundmass glass (gl) is shown in (b)
123
378
Contrib Mineral Petrol (2013) 166:371–392
Table 3 Average major element composition of groundmass glass from historical Ruapehu eruptions Major Element (wt %)
1945
1969
1971
1977
1995
1996
SiO2
77.28 (0.91)
72.07 (1.24)
68.88 (2.08)
67.93 (1.07)
63.2 (1.24)
62.06 (1.23)
Al2O3 FeO
12.41 (0.7) 1.82 (0.35)
13.32 (0.91) 3.2 (0.52)
14.36 (1.34) 5.09 (0.89)
13.95 (0.43) 6.07 (0.39)
15.34 (0.59) 6.47 (0.55)
15.68 (1.2) 4.61 (3.24)
MnO
0.02 (0.01)
0.05 (0.02)
0.07 (0.02)
0.09 (0.02)
0.11 (0.04)
2.48 (3.37)
MgO
0.11 (0.13)
0.49 (0.32)
0.82 (0.53)
1.22 (0.17)
2.5 (0.61)
2.45 (0.48)
CaO
1.03 (0.37)
2.06 (0.51)
3.45 (0.69)
3.9 (0.3)
5.58 (0.6)
5.85 (0.66) 3.36 (0.65)
Na2O
1.97 (0.42)
3.83 (0.31)
2.88 (0.84)
2.33 (0.67)
3.18 (0.61)
K2O
4.47 (0.45)
4.15 (0.3)
3.01 (0.31)
3.05 (0.16)
2.45 (0.26)
2.36 (0.45)
TiO2
0.88 (0.23)
0.68 (0.08)
1.17 (0.18)
1.19 (0.05)
0.99 (0.08)
1.01 (0.11)
P2O5 Number of analyses
0.19 (0.05) 18
0.13 (0.05) 36
0.25 (0.06) 17
0.25 (0.04) 31
0.17 (0.08) 28
0.2 (0.04) 45
Standard deviations are given in parentheses. The full dataset can be found in Supplementary Table 1
with little overlap in major element composition with the 1969 glasses. This offset is most clearly seen in Na2O (Fig. 4a). Therefore, although some of the eruptions analysed here were less than 2 years apart (1969 and 1971) their groundmass glass composition is distinct (Fig. 4a–d). A further striking feature of the groundmass glasses is the wide range in SiO2 content (4–8 wt%) within an individual eruption, attesting to the heterogeneous nature of groundmass glasses from Ruapehu (Fig. 4a–d). For instance, the SiO2 content of groundmass glasses erupted in 1969 ranges from 69 to 74 wt%, 1971 from 65 to 73 wt%, 1977 from 66 to 70 wt%, and 1995–96 from 59 to 66 wt%. The latter are less evolved, however, than the range of 62–70 wt% SiO2 reported for the 1995–1996 eruptions by Nakagawa et al. (1999); we are not able to explain this discrepancy. For all eruptions, the groundmass glass exhibits a linear trend towards the bulk rock composition in most binary major element plots (Fig. 4a–d). A notable exception is Na2O (Fig. 4a), where linear trends of negative slope do not extrapolate to the bulk rock composition. Major element chemistry of melt inclusions The major element chemistry of some pyroxene- and plagioclase-hosted melt inclusions is presented in Table 4 (for the full dataset, refer to Supplementary Table 2) and plotted in Fig. 4e–h. The major element composition of pyroxene- and plagioclase-hosted inclusions is similar in all samples and covers the same range as the groundmass glass (i.e., 60 to 73 wt% SiO2). Unlike the groundmass glass compositions, however, plagioclase- and pyroxenehosted melt inclusions from individual eruptions do not form distinct clusters in major element plots, but instead exhibit significant overlap between eruptions (Fig. 4e–h). Overall, however, the most evolved inclusions are from the earliest eruption analysed (1945), and the least
123
evolved inclusions are from the 1995–1996 eruptions (Fig. 5). Volatile content of melt inclusions Our SIMS analyses (CO2, H2O, Li, Be, B, F, Cl) of melt inclusions from historical eruptions at Ruapehu (Table 4; Fig. 6) are the first direct measurements of the volatile content of Ruapehu magmas. Most Ruapehu melt inclusions have H2O contents of 1–1.5 wt%, with a small number of inclusions exceeding 2.5 wt% (Fig. 7), which is relatively dry compared to similar andesitic systems in arc settings (e.g., Blundy et al. 2010; Devine et al. 1998; Portnyagin et al. 2007; Wallace 2005). According to the literature, the only arc andesite magma with a similarly low H2O content is from the 1994 to 1998 eruption of Popocate´petl (Mexico), with a range of 0.8–3.02 wt% H2O (Atlas et al. 2006). The low H2O contents have implications for magma evolution, phase relations, and transport properties. Ruapehu melt inclusions range in CO2 from c. 25 to 1,059 ppm, with most inclusions having less than c. 600 ppm CO2 (Table 4; Fig. 7). These CO2 values are significantly higher than most intermediate subduction zone magmas (Wallace 2005) but are again similar to those from Popocate´petl, where melt inclusions preserve B1,458 ppm CO2 (Atlas et al. 2006). Melt inclusions preserve elevated halogen contents with up to 2,069 ppm F and 1,342 ppm Cl (Table 4; Fig. 6d, g, h). Average halogen values in melt inclusions are c. 955 ± 175 ppm F and 659 ± 128 ppm Cl; average groundmass glass values are 838 ± 138 ppm F and 515 ± 88 ppm Cl. These data suggest only limited degassing at very low pressures, consistent with experiments on basaltic bulk compositions that show that the significant Cl loss requires very low pressures, after most
Contrib Mineral Petrol (2013) 166:371–392
379
a
e
b
f
c
g
d
h
Fig. 4 Major element plots of groundmass glass (a–d) and phenocryst-hosted melt inclusion compositions (e–h) from 1945 to 1996. Bulk rock XRF data are shown as the open black ellipses. All data are re-calculated to anhydrous values. Note the distinction between each
eruption is clear in the groundmass glass, yet the melt inclusions are relatively tightly clustered with significant overlap. Samples are from the same suite of samples as Fig. 2 and Table 1
of the initial H2O, SO2, and CO2 has already exsolved (Lesne et al. 2012). Li concentrations in melt inclusions vary between 22 and 80 ppm, with most of the data confined to between 40 and 70 ppm for all eruptions (Table 4; Fig. 6). Matrix glasses contain 26–50 ppm Li. Li contents of melt inclusions and
groundmass glasses tend to increase with increasing H2O (Fig. 6a), suggesting that Li partitions modestly into the vapour phase during degassing. Be concentrations are low, from 1 to 5 ppm, and show no correlation with H2O (Fig. 6b). Be concentrations in the matrix glass are similar to the lowest value in melt inclusions of *1 ppm. B contents
123
380
Contrib Mineral Petrol (2013) 166:371–392
Fig. 5 Plot of K2O versus SiO2 showing the evolution of groundmass glasses and melt inclusions to less-evolved compositions with time. All data have been re-calculated to anhydrous values. Symbols are the same as in Fig. 4
of most melt inclusions range from *30 to *80 ppm and are not correlated with H2O (Fig. 6c). Matrix glasses range from *30 to *50 ppm B, with some higher values
123
recorded in groundmass glass from the 1969 eruption. Evidently, the behaviour of B and H2O is decoupled during magmatic degassing.
1.27
1.41
1.68
0.73
1.54
1.90
2.19
2.15
0.96 0.93
0.94
0.85
0.99
0.96
0.98
0.98
0.71
0.95
0.83
0.77
1.36
1.33
1.20
1.29 1.00
1969py-09 inc 01
1969py-09 inc 02
1969py-09 inc 04
1969py-09 inc 05
1969py-09 inc 06
1969py-12 inc 04
1969py-12 inc 05
1969py-12 inc 06
1969pl-05 inc 01 1969pl-05 inc 02
1969pl-05 inc 03
1969pl-10 inc 01
1969pl-11 inc 01
1969pl-11 inc 02
1969pl-11 inc 03
1969pl-11 inc 04
1969pl-11 inc 05
1969pl-11 inc 06
1969pl-11a inc 06
1969pl-12 inc 01
1971py-02 inc 01
1971py-02 inc 03
1971py-03 inc 01
1971py-03 inc 02 1971py-07 inc 01
1.75
0.97
1969py-08 inc 01
1971py-08 inc 01
1.59
1969py-07 inc 07
2.18
1.58
1969py-07 inc 06
1.96
1.32
1969py-07 inc 04
1971py-07 inc 03
1.52
1969py-07 inc 03
1971py-07 inc 02
H2O (wt%)
530
587
321
850 90
530
155
169
350
309
216
216
279
292
303
66
247
39
176 343
185
104
79
360
33
503
286
250
84
176
1059
484
766
CO2 ppm
55
38
36
32 24
35
39
42
43
47
50
43
52
53
51
57
54
57
51 57
44
43
79
44
44
59
44
38
46
43
37
41
43
Li ppm
3
4
3
2 2
2
2
4
2
2
2
2
2
2
2
2
2
2
2 2
4
4
4
2
2
2
2
2
4
2
2
2
2
Be ppm
63
49
52
52 47
56
53
57
57
59
66
53
49
47
53
63
63
78
67 74
61
58
79
58
66
48
66
58
70
62
53
63
61
B ppm
1287
908
963
1382 717
1411
816
737
1095
999
1001
978
1015
1076
924
844
1306
1404
1001 1507
1281
1213
1046
705
839
2005
810
733
673
883
840
876
803
F ppm
885
458
491
599 371
699
515
584
599
586
434
421
515
594
553
597
821
999
685 997
670
614
603
824
422
1131
1037
846
304
767
607
692
538
Cl ppm
Secondary ion mass spectrometry (SIMS) data
Sample number
Melt inclusions
68.51
68.95
69.38
68.66 69.96
68.47
67.71
68.34
69.59
70.17
70.33
70.33
70.69
69.90
70.40
70.65
68.17
66.71
68.64 67.16
68.07
68.21
68.71
68.33
72.69
68.84
68.64
69.86
73.18
70.01
69.44
70.16
70.13
SiO2
1.27
1.01
1.05
0.70 1.16
0.83
1.13
1.07
1.11
0.90
0.94
0.94
0.73
1.05
0.87
0.90
1.57
1.85
1.57 1.98
0.63
0.70
0.98
0.94
0.67
0.75
0.95
0.75
0.49
0.86
0.78
0.82
0.87
TiO2
3.41
3.61
3.16
4.17 3.59
3.97
3.91
3.87
3.06
3.88
3.84
3.84
4.05
4.22
4.06
3.92
3.87
3.57
3.12 3.21
4.54
4.29
4.66
3.31
3.74
3.88
3.19
3.23
3.57
3.63
3.78
3.75
3.74
Na2O
13.97
15.03
14.73
14.33 14.24
14.81
14.30
14.43
12.64
13.44
13.36
13.36
13.39
13.60
13.52
13.32
13.00
12.64
12.71 12.81
15.65
15.88
14.46
14.49
12.82
14.64
14.66
14.59
12.70
14.53
14.47
14.11
14.39
Al2O3
4.52
3.99
4.43
4.39 4.28
4.55
5.67
5.10
5.43
4.12
4.28
4.28
4.13
3.85
4.04
4.12
5.33
6.45
5.72 6.16
3.35
3.49
3.95
4.57
3.31
4.41
3.96
3.72
3.55
3.56
4.03
3.62
3.66
FeO
4.40
2.91
3.12
5.01 3.69
4.53
3.61
3.92
4.12
4.12
4.04
4.04
4.17
4.30
4.14
4.19
3.73
3.64
3.75 3.56
3.68
3.41
3.87
4.44
4.51
3.19
4.52
4.58
4.46
3.94
3.52
4.05
3.60
K2O
3.00
3.50
3.24
2.05 2.44
2.14
2.72
2.61
2.65
2.49
2.31
2.31
2.13
2.30
2.22
2.12
2.63
3.42
3.01 3.44
2.94
2.90
2.54
3.03
1.67
3.20
3.10
2.56
1.54
2.76
3.04
2.75
2.86
CaO
Electron probe micro-analysis (EPMA) data (wt%)
0.78
0.92
0.77
0.60 0.48
0.57
0.80
0.65
1.28
0.81
0.73
0.73
0.62
0.66
0.65
0.66
1.55
1.53
1.33 1.49
1.03
1.04
0.79
0.72
0.43
0.95
0.85
0.64
0.41
0.71
0.83
0.68
0.69
MgO
0.00
0.00
0.00
0.01 0.03
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.02
0.00
0.00
0.02
0.02
0.00 0.00
0.02
0.00
0.01
0.03
0.00
0.00
0.02
0.01
0.00
0.00
0.00
0.00
0.00
Cr2O3
0.05
0.05
0.10
0.00 0.11
0.05
0.11
0.00
0.07
0.02
0.15
0.15
0.09
0.04
0.06
0.10
0.09
0.11
0.11 0.12
0.08
0.07
0.01
0.08
0.16
0.08
0.08
0.04
0.10
0.00
0.07
0.05
0.04
MnO
1001953
1001781
1001273
1002758 100 567
1001996
1001054
1001358
1001669
1001454
1001268
1001112
1001411
1001446
1001449
100 537
1001522
100 806
1001336 1001750
1001013
100 789
100 729
1001547
100 303
1001730
1001235
1001085
100 529
100 878
1002682
1001686
1002159
Total
Pressure (bars)
0.26
0.38
0.43
0.12 0.34
0.15
0.28
0.23
0.09
0.11
0.15
0.11
0.14
0.14
0.14
0.30
0.12
0.22
0.16 0.13
0.53
0.67
0.55
0.26
0.33
0.26
0.28
0.26
0.29
0.41
0.18
0.19
0.19
XH2O
Pyroxene
Pyroxene
Pyroxene
Pyroxene Pyroxene
Pyroxene
Pyroxene
Pyroxene
Plagioclase
Plagioclase
Plagioclase
Plagioclase
Plagioclase
Plagioclase
Plagioclase
Plagioclase
Plagioclase
Plagioclase
Plagioclase Plagioclase
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Host mineral
Table 4 Volatile content (SIMS) and major element chemistry (EPMA) of phenocryst-hosted melt inclusions from historical Ruapehu scoria. Major elements are recalculated to anhydrous values
Contrib Mineral Petrol (2013) 166:371–392 381
123
H2O (wt%)
2.03
1.72
1.66
1.36 1.41
1.47
1.58
2.48
1.74
1.45
1.42
1.43
1.42
1.56
1.53
1.30
1.52
1.39
1.56
1.63 0.89
1.28
1.29
1.42
1.33
1.42
1.49
1.13
0.45
1.25
1.89
1.64
1.77
1.45
1971py-08 inc 02
1971py-08 inc 03
123
1971py-08 inc 04
1971py-09 inc 01 1977py-01 inc 02
1977py-01 inc 03
1977py-01 inc 04
1977py-04 inc 01
1977py-05 inc 02
1977py-06 inc 01
1977py-06 inc 02
1977py-06 inc 03
1977py-06 inc 04
1977py-07 inc 01
1977py-08 inc 03
1995py-02 inc 01
1995py-03 inc 02
1995py-05 inc 04
1995py-05 inc 05
1995py-06 inc 05 1995py-07 inc 06
1995py-08 inc 02
1995py-09 inc 01
1995py-09 inc 02
1995py-09 inc 03
1995py-10 inc 01
1995py-10 inc 02
1995py-10 inc 04
1996py-01 inc 01
1996py-02 inc 01
1996py-02 inc 04
1996py-03 inc 01
1996py-03 inc 02
1996py-04 inc 01
897
626
379
274
471
115
133
358
201
376
646
404
546
190 1043
63
729
282
292
831
501
90
242
62
65
377
486
271
106
538 314
436
380
360
CO2 ppm
40
33
34
38
42
30
26
35
35
39
41
40
36
36 41
40
39
35
29
43
37
43
42
43
44
47
71
41
41
41 42
46
43
54
Li ppm
4
3
4
1
4
3
4
5
2
2
3
2
1
1 2
1
3
1
2
2
4
4
3
4
2
2
4
3
2
3 2
2
2
2
Be ppm
57
57
54
52
55
37
43
68
63
51
53
51
64
46 38
41
50
45
56
62
53
61
60
63
60
44
34
57
63
62 65
62
57
60
B ppm
935
1133
1038
1435
1313
800
1065
1217
1032
910
943
971
1103
1394 1036
832
725
951
1094
954
1221
948
1013
970
1198
1218
825
1272
837
752 1010
901
957
910
F ppm
821
679
686
1216
872
479
556
1043
888
995
1207
1092
727
882 882
640
753
614
649
844
809
785
766
849
807
571
675
954
465
510 617
734
670
687
Cl ppm
Secondary ion mass spectrometry (SIMS) data
Sample number
Melt inclusions
Table 4 continued
65.78
66.70
67.23
64.90
66.26
64.86
68.32
67.63
68.32
64.94
62.80
65.65
60.57
63.09 60.14
63.32
64.14
66.14
64.63
66.95
67.13
65.98
64.20
66.34
67.76
63.76
66.20
65.84
69.06
70.48 69.03
68.68
69.63
69.04
SiO2
1.12
0.95
0.94
0.80
0.80
1.18
0.92
0.88
0.92
0.85
1.44
0.72
0.87
0.80 0.94
0.96
1.16
1.00
1.22
1.00
0.85
0.94
0.92
1.04
1.02
0.74
0.95
1.01
1.03
0.76 1.02
1.18
0.76
1.10
TiO2
3.69
3.54
3.26
2.90
3.89
3.20
3.53
3.57
3.53
3.84
3.68
3.72
6.50
3.88 4.19
4.00
3.79
4.26
4.00
3.67
3.75
3.43
3.25
3.54
3.75
4.09
3.99
4.41
3.74
3.49 3.63
3.08
3.11
3.85
Na2O
14.46
13.85
14.05
14.49
14.57
13.99
14.18
14.43
14.18
14.90
13.50
15.10
14.43
15.58 15.99
15.25
13.78
14.99
14.29
14.40
13.37
12.74
11.18
12.94
14.12
14.97
15.95
15.18
13.53
15.37 13.80
13.72
13.79
14.56
Al2O3
5.79
5.75
5.46
5.12
4.88
6.28
4.75
4.63
4.75
6.12
8.57
5.36
5.89
6.51 7.46
6.67
8.01
5.87
6.65
5.14
5.67
5.65
6.37
5.77
5.07
5.29
4.16
4.76
4.85
3.01 4.64
5.09
4.59
4.09
FeO
3.61
3.73
3.78
4.68
3.86
3.34
3.84
4.10
3.84
3.07
2.81
3.12
4.34
2.85 2.28
2.21
3.11
2.60
3.06
3.85
3.26
3.24
2.94
3.21
3.58
2.78
3.19
3.40
2.98
3.51 3.05
4.93
5.02
3.84
K2O
4.05
3.93
3.87
4.73
3.90
4.85
3.24
3.44
3.24
4.51
5.00
4.43
5.25
5.03 6.26
5.14
4.19
3.75
4.23
3.66
4.08
5.12
6.79
4.74
3.40
5.53
3.69
3.63
3.53
2.48 3.51
2.41
2.33
2.71
CaO
Electron probe micro-analysis (EPMA) data (wt%)
1.34
1.36
1.29
2.24
1.68
2.17
1.08
1.21
1.08
1.55
1.93
1.68
2.07
2.10 2.55
2.32
1.54
1.29
1.71
1.27
1.75
2.72
4.24
2.25
1.19
2.73
1.31
1.53
1.22
0.83 1.23
0.70
0.68
0.70
MgO
0.00
0.00
0.01
0.00
0.00
0.01
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00 0.00
0.00
0.00
0.00
0.02
0.01
0.01
0.00
0.02
0.02
0.00
0.01
0.02
0.02
0.00
0.02 0.02
0.00
0.00
0.00
Cr2O3
0.10
0.12
0.08
0.07
0.09
0.06
0.07
0.04
0.07
0.16
0.20
0.18
0.05
0.10 0.13
0.05
0.18
0.05
0.12
0.00
0.07
0.16
0.08
0.15
0.09
0.05
0.05
0.07
0.02
0.00 0.06
0.04
0.07
0.06
MnO
1002551
1002131
1001561
1001354
1001736
100 821
100 782
1001451
1001029
1001553
1002406
1001481
1002330
1001133 1002599
100 691
1002538
1001326
1001504
1002477
1001911
100 984
1001628
100 844
100 715
1001513
1001741
1001298
100 785
1001446 1001274
1001876
1001654
1001503
Total
Pressure (bars)
0.17
0.25
0.29
0.39
0.19
0.11
0.31
0.27
0.33
0.22
0.17
0.23
0.11
0.37 0.1
0.53
0.16
0.30
0.22
0.18
0.22
0.33
0.22
0.37
0.43
0.31
0.46
0.31
0.41
0.25 0.27
0.24
0.27
0.38
XH2O
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene
Pyroxene Pyroxene
Pyroxene
Pyroxene
Pyroxene
Host mineral
382 Contrib Mineral Petrol (2013) 166:371–392
1.41
1.38
1.65
1.50 1.39
1.48
1.53
1996py-04 inc 02
1996py-06 inc 01
1996py-08 inc 01
1996py-09 inc 01 1996py-09 inc 02
1996py-09 inc 03
1996py-09 inc 04
0.07
0.34
0.21
0.40
0.83
0.61
0.44
0.17
0.31
0.88
0.23
0.17 0.13
0.15
0.36
0.42
0.21
1969-09 gl01
1969-10 gl01
1969-10 gl02
1969-07 gl01
1969-12 gl01
1969-04 gl01
1969-12 gl08
1969-14 gl06
1977-07 gl03
1977-08 gl05
1995-06 gl01
1995-08 gl01 1995-08 gl02
1995-07 gl01
1995-07 gl02
1996-06 gl01
1996-10 gl01
Groundmass glass
H2O (wt%)
442
2
199
154
419 917
177
79
1005
10
22
7
45
0
382
0
245
87
167
181 259
279
755
603
CO2 ppm
31
32
32
31
34 32
33
40
38
32
41
44
42
50
46
33
26
35
36
37 35
74
26
40
Li ppm
1
1
1
1
1 2
1.6
1.8
1.7
3.5
4.5
2.1
2.0
2
2
2
1
4
4
4 4
1
3
4
Be ppm
36
40
42
42
42 41
41
54
48
43
57
73
63
69
67
37
21
50
52
49 53
66
55
54
B ppm
1081
930
860
918
816 812
727
951
595
431
707
502
852
466
549
514
233
1089
1061
942 944
2069
710
1163
F ppm
473
667
641
538
581 479
648
590
414
197
351
239
537
270
321
294
129
528
588
538 584
1342
688
703
Cl ppm
Secondary ion mass spectrometry (SIMS) data
Sample number
Melt inclusions
Table 4 continued
60.79
61.15
63.58
62.64
62.54 63.79
63.89
68.10
66.80
72.67
74.15
72.82
72.89
72.62
73.29
71.23
71.64
65.97
65.73
66.15 66.51
66.26
67.97
64.77
SiO2
1.04
1.03
1.01
0.92
0.97 1.08
1.02
1.24
1.21
0.70
0.52
0.73
0.77
0.68
0.59
0.74
0.67
1.02
1.14
1.21 1.13
0.63
0.86
1.07
TiO2
3.37
2.42
2.10
3.49
3.90 3.63
2.16
2.09
3.24
3.32
3.35
3.85
3.45
4.06
3.92
4.38
3.77
4.09
4.17
3.49 3.91
2.27
3.49
3.59
Na2O
15.40
15.87
15.65
15.89
15.30 15.03
15.34
13.59
13.91
13.44
13.03
12.68
13.25
12.53
12.52
12.71
12.93
14.67
13.87
14.23 13.84
14.25
13.66
14.90
Al2O3
7.92
7.42
6.55
6.48
6.74 6.40
6.99
6.29
6.30
3.06
2.64
3.35
3.18
3.19
3.04
3.74
3.81
5.87
6.30
5.96 5.98
6.23
5.16
6.47
FeO
2.90
2.89
2.53
2.36
2.09 1.91
2.73
3.13
3.04
3.95
4.40
4.17
4.00
4.52
4.46
4.37
4.10
2.87
3.17
3.01 3.15
4.99
4.23
3.65
K2O
5.34
6.07
5.70
5.51
5.88 5.83
5.17
3.91
4.00
2.22
1.48
1.72
1.91
1.78
1.62
1.95
1.98
3.87
3.90
4.17 3.83
3.77
3.35
4.10
CaO
Electron probe micro-analysis (EPMA) data (wt%)
2.85
2.84
2.50
2.39
2.30 1.99
2.40
1.22
1.17
0.46
0.34
0.43
0.39
0.47
0.43
0.65
0.89
1.49
1.51
1.62 1.49
1.44
1.22
1.39
MgO
0.00
0.01
0.01
0.01
0.01 0.00
0.02
0.02
0.00
0.00
0.00
0.00
0.04
0.00
0.01
0.00
0.04
0.00
0.00
0.01 0.04
0.02
0.00
0.00
Cr2O3
0.16
0.10
0.16
0.09
0.09 0.09
0.07
0.09
0.11
0.04
0.04
0.05
0.04
0.05
0.02
0.06
0.02
0.13
0.16
0.09 0.12
0.11
0.06
0.02
MnO
100
100
100
100
100 100
100
100
100
100
100
100
100
100
100
100
100
100 759
1001137
1001070 1001351
1001444
1002359
1002132
Total
Pressure (bars)
0.46
0.31
0.36 0.26
0.32
0.17
0.19
XH2O
Pyroxene
Pyroxene
Pyroxene Pyroxene
Pyroxene
Pyroxene
Pyroxene
Host mineral
Contrib Mineral Petrol (2013) 166:371–392 383
123
384 Fig. 6 Trace element (Li, Be, B, Cl) variation with H2O measured by SIMS. a Li increases with increasing H2O. Be b and B c show no correlation with H2O. d A weak, positive correlation exists between Cl and H2O. (e–h) Trace element (Li, B, F, Cl) content versus pressure plots. Cl degasses at low pressure, while the other trace elements show no change with pressure
Contrib Mineral Petrol (2013) 166:371–392
a
e
b
f
c
g
d
h
Discussion Magmatic storage conditions We have determined magmatic temperatures for the historical Ruapehu eruptions using several different geothermometers. The absence of touching Fe–Ti oxide pairs in all samples (except for disparate pairs in the 1969 sample) precludes us from using the method of Lindsley and Anderson (1983); for this reason, we have used the plagioclase-liquid (Putirka 2008), clinopyroxene-liquid (Putirka 2008), orthopyroxene-liquid (Putirka 2008), and twopyroxene (Lindsley and Frost 1992) geothermometers (assuming a H2O content of *1.5 wt%). In all geothermometer calculations, we have assumed a pressure of 250 MPa, which is based on the volatile data (see below)
123
and geophysical interpretations (Ingham et al. 2009; Rowlands et al. 2005). While these methods yield different absolute temperatures, the trend in the data from one eruption to another is consistent (Fig. 2). All historical eruptions plot between 910 and 1,080 °C, with the bulk of the data clustering between 950 and 1,050 °C. There is a general increase in magmatic temperature from the earliest sample analysed (1945) to the latest scoria sample (1996) (Fig. 2). The relatively low temperature for 1945 may reflect slow cooling and re-equilibration of this lava sample. Temperature estimates for the 1995–1996 eruptions occupy a limited range between 1,000 and 1,080 °C with no apparent clustering towards the high or low estimates. These data contrast with those of Nakagawa et al. (1999), who found two separate populations of clinopyroxene– orthopyroxene pairs, one that yielded temperatures of
Contrib Mineral Petrol (2013) 166:371–392
385
a
b
Fig. 7 Volatile contents as measured by SIMS against modelled degassing histories for pyroxene-hosted melt inclusions from Ruapehu. a H2O versus CO2 content of melt inclusions plotted with calculated isobars (dashed grey lines) and vapour isopleths in mol % H2O (dashed black lines). Isopleths and isobars were calculated using Papale et al. (2006). Illustrative closed-system degassing curves (curved black lines) are plotted from three starting compositions (Curve 1–1.28 wt% H2O and 1,060 ppm CO2, Curve 2–1.75 wt% H2O and 800 ppm CO2, and Curve 3–2.18 wt% H2O and 700 ppm CO2). b Plot of calculated XH2O versus saturation pressure showing
the coherence of all melt inclusions between closed-system degassing paths 1 and 3. Significant H2O-loss or CO2-fluxing would result in the inclusion population following a systematic decrease in XH2O, with a slight decrease in saturation pressure (arrowed line). Rare inclusions that fall below degassing curve 1 in (b) probably did lose significant H2O. Closed-system degassing of distinct magmas with a similar volatile content can explain the observed variations in H2O versus CO2 and XH2O versus pressure space. Average propagated errors in the SIMS analyses are shown
*1,000 °C and another with temperatures of 1,000– 1,200 °C. The high temperatures, which were calculated using the method of Lindsley and Anderson (1983), appear unreasonably high for the andesite composition of the Ruapehu ejecta. For instance, the liquidus temperature of the 1995–1996 magma is approximately 1,150 °C (using Danyushevsky and Plechov, 2011) which constrains the maximum phenocryst temperature. Also, the updated geothermometers of Putirka (2008) produce a more homogeneous temperature than the Lindsley and Anderson (1983) method. Therefore, for the purposes of this paper, we have used the Putirka (2008) geothermometers throughout. The volatile contents of melt inclusions are similar for all analysed eruptions (Figs. 6, 7). They contain
B1,000 ppm CO2 and *1.5 wt% H2O, which suggests a minimum (volatile saturation) trapping pressure of *50–270 MPa (at between 920 and 1,030 °C using the calculation of Papale et al. 2006). This is in agreement with estimates from the phenocryst melt and two-pyroxene geobarometers of Putirka (2008). If we assume a crustal density of 2,600 kg/m3 and volatile saturation, this pressure range suggests that the magma storage region beneath Ruapehu extends from *2 to 9 km below the volcano. Our data show that Ruapehu magmas are relatively dry compared to other arc andesites. While the phase equilibria of Moore and Carmichael (1998) appear consistent with a dry, andesitic magma, it has been proposed that phenocryst-hosted melt inclusions are able to rapidly hydrate or
123
386
de-hydrate due to H? diffusion (e.g., Gaetani et al. 2012) and rapid equilibration with the surrounding melt. Lloyd et al. (2013) also suggest that melt inclusions within large scoria or lapilli clasts are prone to significant dehydration. Clearly, the H2O content of our melt inclusions will have a significant effect on the pressure determinations. In order to assess our melt inclusion measurements, we plotted H2O versus CO2 (Fig. 7a) and the calculated XH2O (mol fraction H2O of the vapour) against saturation pressure (Fig. 7b). We then compared our data to modelled open or closed-system degassing profiles (Papale et al. 2006). Our data are weakly scattered around a mean of 1.5 wt% H2O and range from 50 to 1,000 ppm CO2. While these data could record a complex interplay between CO2-fluxing, crystallisation, and H2O loss (e.g., Blundy and Cashman, 2008, Spilliaert et al. 2006), we consider a simpler interpretation. In the XH2Ovapor versus pressure diagram, all the data appear to follow a relatively simple closed-system degassing profile wherein XH2O increases with decreasing pressure (Fig. 7b). There is some scatter in the data, which may reflect different initial volatile contents or degassing trajectories. However, the data are not consistent with significant diffusive loss of H2O from the melt inclusions. If the inclusions had dehydrated significantly, we would expect a large number (if not all) of the melt inclusions to record low XH2Ovapor and anomalously low pressures, inconsistent with other independent data. To illustrate the effect of diffusive H2O loss, we took one melt inclusion composition and progressively reduced its H2O content by 0.5 wt%, from 2.5 wt%. At each point, we calculated the XH2O and saturation pressure. The result provides a vector for which melt inclusions would trend towards given significant H2O loss (Fig. 7b). From this, we can see that Ruapehu melt inclusions do not exhibit significant H2O loss. To explain the H2O and CO2 data, we first must consider that the bulk chemistry of Ruapehu magmas is broadly similar (Table 1), with relatively small variations between eruptions. This indicates that all eruptions are derived from a similar parental magmatic system (e.g., Gamble et al. 1999). Therefore, we considered the degassing trajectories of three magmas with broadly similar major element chemistry and variable initial H2O and CO2. Most of the data plot between the two bounding closed-system degassing curves (curve 1–1.28 wt% H2O and 1,060 ppm CO2; curve 3–2.18 wt% H2O and 700 ppm CO2). Based on these data, we can conclude that Ruapehu magmas have a H2O content of up to 2.18 wt% and a CO2 content of at least 700 ppm. As stated above, ilmenite is only observed within scoria from 1969 (magnetite is noted in all scoria). In order to calculate equilibrium magmatic temperature and oxygen fugacity, it is best to analyse touching ilmenite–magnetite
123
Contrib Mineral Petrol (2013) 166:371–392
pairs. However, as we were unable to find touching pairs in the 1969 sample, we tested all possible combinations of disparate ilmenite and magnetite compositions for equilibrium using the method of Bacon and Hirschmann (1988). Only equilibrium pairs were used to determine the fugacity and temperature of the 1969 magma using ILMAT (LePage 2003). We calculated an oxygen fugacity of log fO2 -11.20, equivalent to the Ni-NiO oxygen buffer (NNO) and temperature c. 939 °C (O’Neill and Pownceby 1993; Frost 1991), which is *20 °C higher than the average of the plagioclase-liquid and pyroxene-liquid geothermometers. The remarkably similar mineralogy and phenocryst compositions suggest that the oxygen fugacity of all historical Ruapehu magmas lies close to NNO, similar to prehistorical magmas (Price et al. 2012). To further assess the consistency of our pressure–temperature estimates, we compare our data to high temperature and pressure melting experiments on andesites. The bulk composition of Ruapehu magmas is similar to that of Volca´n Colima, Mexico (Moore and Carmichael 1998), thus the starting compositions for hydrous phase equilibria experiments from that volcano provide a reasonable comparison to natural Ruapehu samples, except that the Colima experiments were run under H2O-saturated conditions, that is PH2O = Ptot. Because Ruapehu magmas are H2O-poor (PH2O \ Ptot), a pressure correction must be applied to the experimental data. This correction is relatively straightforward because of the negligible effects of CO2 on phase equilibria in silicate systems at low pressures. If we take the most volatile-rich melt inclusion of *1,000 ppm CO2 and 1.5 wt% H2O, we calculate a saturation pressure of *270 MPa, with an XH2O of *12–16 % at a temperature range of between 915 and 1,030 °C (Papale et al. 2006). This equates to a PH2O of *32–43 MPa. If we assume that the addition of CO2 simply increases the Ptot, without effecting phase relations, then we can use this value of PH2O to match our data to the experiments. Using this correction, Ruapehu magmas plot near the 2 wt% H2O isopleth, in a region mostly outside of the hornblende stability field, but with plagioclase, orthopyroxene, clinopyroxene, and magnetite stable, in accord with the observed phenocryst populations. The experimental equilibrium plagioclase composition is *An60–65, consistent with the measured composition of plagioclase rims (Fig. 8). From this comparison, we conclude that the mineralogy, volatile content, and temperature of Ruapehu magma are consistent with the andesite phase equilibria of Moore and Carmichael (1998) at PH2O of *40 MPa. Interaction between magma and crystal mush A compositional comparison of melt inclusions, groundmass glass, and bulk rock can be used to determine the
Contrib Mineral Petrol (2013) 166:371–392
extent to which melt inclusions and their phenocryst hosts are in chemical equilibrium with the host magma. However, we must first consider whether any of the melt inclusions have been modified by post-entrapment crystallisation. Daughter minerals are absent from all melt inclusions analysed. To evaluate the extent of melt inclusion crystallisation onto the host crystal, we plotted
Fig. 8 Phase diagram of an andesite of similar bulk composition to Ruapehu (Volcan Colima), in terms of PH2O versus temperature, redrawn from Moore and Carmichael (1998). Grey box represents the calculated (using Papale et al. 2006) PH2O conditions of H2O-poor, CO2-rich historical Ruapehu magmas. Magmatic temperatures were determined by crystal-melt and Fe–Ti oxide geothermometry. Ruapehu magmas occupy a PH2O-temperature space where the equilibrium phase assemblage consists of plagioclase (Plag), orthopyroxene (Opx), clinopyroxene (Aug), and magnetite (Mt). Hornblende (Hbl) is absent from Ruapehu due to the relatively low H2Osaturated pressure and high-temperature conditions of Ruapehu magmas. Sub-horizontal dashed lines are H2O concentration isopleths
387
separately MgO versus Al2O3 (Fig. 9) for plagioclase- and pyroxene-hosted inclusions, as these elements are differently compatible in pyroxene and plagioclase crystals. For example, pyroxene-hosted melt inclusions that crystallise on the host would result in a displacement to very low MgO content with little change in Al2O3, whereas crystallisation of plagioclase would lead to a decrease in Al2O3 and slight increase in MgO. These different trends are shown as vectors corresponding to 5 wt% crystallisation in Fig. 9. In general, the compositional overlap between plagioclase- and pyroxene-hosted melt inclusions suggests that post-entrapment crystallisation was limited. Specifically, compositional variations in the melt inclusions follow cotectic crystallisation trends; that is, the trends do not follow the vectors anticipated for post-entrapment crystallisation of the host mineral (Fig. 9). As melt inclusions do not appear to have experienced significant post-entrapment crystallisation, they can be used to examine variations in crystallisation (driven by cooling, decompression, and/or H2O loss) and magma mingling/mixing in the small magma batches produced by Ruapehu eruptions. It is useful to discuss these data in two groups: (1) 1945–1977 and (2) 1995–1996. We use the incompatible elements K2O and TiO2 as plotting parameters as these two components best illustrate the compositional differences between eruptions. 1945 to 1977 The major element chemistry of plagioclase- and pyroxene-hosted melt inclusions from 1945 plots along the same fractional crystallisation trend as the bulk rocks (Fig. 10). The melt inclusion compositions occupy a wider range than the groundmass glass. Under equilibrium conditions, the compositions of melt inclusions should lie between the bulk rock and the groundmass glass on a plot of two
Fig. 9 MgO v Al2O3 plot of plagioclase- and pyroxenehosted melt inclusions from the 1995 eruption of Ruapehu. These data show that the chemistry of each inclusion is largely independent of the host mineral. Plagioclase- and pyroxene-hosted melt inclusions that had crystallised after being trapped would trend away from their respective host crystal along the vectors shown as black arrows. The length of arrows approximates 5 % crystallisation of plagioclase (Plag) and pyroxene (Pyrox)
123
388
Contrib Mineral Petrol (2013) 166:371–392
nominally incompatible major elements, in this case K2O and TiO2 (e.g., Faure and Schiano 2005). Melt inclusions that fall outside of the equilibrium line may indicate that the crystals are exotic to the host magma, as appears to be the case for the 1945 data. The most likely source for exotic crystals at Ruapehu is crystal mush, such as that invoked by Nakagawa et al. (1999), because of both numerous crystal clots within scoria clasts and the presence of a high- and low-temperature signature from twopyroxene geothermometry. Together, these data suggest that the 1945 magma probably intersected, and interacted with, at least one crystal mush zone during ascent.
The 1969 melt inclusions and groundmass glasses plot along a fractional crystallisation trend from the bulk rock composition (Fig. 10). Most inclusions are less evolved than the groundmass glass, which is to be expected if crystallisation continues in the melt after inclusions become trapped within a crystal. This suggests that the 1969 eruptions were driven by a small volume magma that was isolated from the larger magma storage region. In contrast, most melt inclusions from the 1971 eruption describe a fractionation trend that is different from the groundmass glass. The shift of exotic inclusions to high K2O at constant TiO2 requires that they crystallised from a
Fig. 10 Plots of TiO2 versus K2O (two incompatible elements) showing the composition of groundmass glass, melt inclusion and bulk rock XRF data for each Ruapehu eruption analysed. Cognate melt inclusions should fall on the same line (grey dashed line) as the bulk rock and groundmass glass (e.g., Faure and Schiano, 2005). The 1945 melt inclusions span a similar range in K2O to the groundmass glass, while the melt inclusions are displaced to lower TiO2. The crystals are therefore equivocally exotic. The 1969 melt inclusions are mostly less evolved than the groundmass glass, appear on a similar chemical trend to the whole rock, and we conclude that most melt inclusions (and hence the crystals) are cognate with the groundmass glass. The 1971 and 1977 melt inclusions are more evolved than the groundmass and are displaced either side of their respective equilibrium mixing lines from the groundmass glass and whole rock
compositions. Most of the 1971 and 1977 inclusions are considered exotic to the host melt. The 1995 melt inclusions have a larger compositional spread than the groundmass glass. Many inclusions are of a similar composition to the melt (cognate), while the more mafic and silicic end members are possibly exotic. The 1996 groundmass glass is less evolved than the melt inclusions. This implies that all of the 1996 melt inclusions and hence the entire population of inclusionbearing phenocrysts are exotic. The groundmass glass and melt inclusions are chemically distinct, but appear on the same mixing line; therefore, both melts must have had a similar parent composition and mineralogy. In all plots, black triangles represent the bulk rock composition for scoria from 1945 to 1996 (Gamble et al. 1999; this work)
123
Contrib Mineral Petrol (2013) 166:371–392
389
melt with a bulk composition that was distinct from the transporting magma (represented by the groundmass glass). Importantly, both exotic and cognate melt inclusions can be found within the same crystal, and without an obvious spatial distribution. However, there is evidence for some crystallisation within the transporting magma in the small number of 1971 inclusions that are chemically similar to the groundmass glass and could thus be genetically related to the host magma (i.e., cognate). The lack of overlap between either the melt inclusion populations or the groundmass glass compositions in 1971 and 1969 magmas is significant given the limited time between eruptions. These data suggest that prior to the eruption, the 1971 magma entrained crystals from a mush zone that was distinct from the 1969 magma. The exact date or size of eruptions that provided the antecrysts cannot be constrained, but we can discount crystals generated in 1945 based on their different melt inclusion compositions (Figs. 4, 5). As seen in the 1971 scoria, most of the phenocrysthosted melt inclusions in the 1977 scoria are displaced to a higher K2O content than the host glass and are therefore considered exotic (Fig. 10). Moreover, exotic melt inclusions from the 1971 and 1977 eruptions are not chemically similar, which suggests that the 1971 and 1977 magmas ascended through different parts of the subvolcanic system.
The groundmass glass and melt inclusions of the 1995 magma span a similar range in major element chemistry to earlier eruptions (Figs. 4, 5). The bulk of the melt inclusions from 1995 are more evolved than the groundmass glass, but appear to lie on a similar liquid line of descent, indicating that some of the crystals are probably cognate. The groundmass glass and melt inclusion compositions from 1995 plot near the bulk rock composition (Fig. 10), which suggests minimal crystallisation prior to eruption. Pyroxene- and plagioclase-hosted inclusions are both more- and less-evolved than the groundmass glass. Those that are more evolved are likely to be hosted by antecrysts derived from a crystal mush that is genetically linked to the historical eruptions. In contrast, a striking aspect of the
Fig. 11 Conceptual model for the magmatic system at Ruapehu. Small volume andesitic melts are residing as sills and dykes from \2 to *9 km depth, based on the volatile content of phenocryst-hosted melt inclusions (using Papale et al. 2006), combined with magnetotelluric soundings (Ingham et al. 2009) and seismic tomography (Rowlands et al. 2005). A hydrothermally altered boundary zone (grey diffuse boundary) exists on the margins of the dyke system
(Ingham et al. 2009). Small volume sills and dykes are distinct but closely spaced. When magma ascends, it interacts with a crystal mush zone/s (1 and 2), entraining some of those exotic crystals into the melt. On ascent, small defects in individual crystals allow for the ingress and then trapping of cognate melt inclusions (3). These cognate melt inclusions record the magmatic conditions of the new melt alongside those that of previous melt/s
As also seen in 1971, a small number of 1977 melt inclusions have a composition that is similar to that of the bulk rock. This suggests that less-evolved magma was introduced into the base of the andesitic magma storage region shortly before each eruption. Moreover, although the groundmass glass compositions of 1971 and 1977 scoria are similar, they are not identical (Fig. 4), and thus suggest that the two events involved two distinct magmas despite their extremely small volumes (e.g., Houghton et al. 1987). 1995 to 1996
123
390
1996 scoria is that most of the melt inclusions are more evolved than the hosting glass, which is similar in composition to that of the 1995 eruption (Fig. 10). For this reason, we consider that most, possibly all, of the melt inclusions analysed from the 1996 eruption are exotic to the transporting melt. More-evolved melt inclusion compositions may indicate that the mush from which antecrysts were entrained had cooled/crystallised significantly prior to interaction. There is substantial overlap in the compositions of plagioclase- and pyroxene-hosted inclusions from the 1995 and 1996 scoria; because some of the 1995 crystals are considered cognate, it seems likely that some of the phenocrysts from 1996 are antecrysts that originally grew in the 1995 magma (Fig. 10). That the groundmass glass from 1996 is distinctly more mafic than the melt inclusions suggests that the 1996 eruption involved a very crystalpoor, relatively mafic magma that entrained crystals from the partially crystalline 1995 magma and possibly from other parts of the magma storage region. Magma volumes and storage architecture The distinctive chemical signatures of groundmass glasses from individual Ruapehu eruptions suggest that each eruption tapped a slightly different magma. We know from field investigations and measurements of the eruptive deposits that the eruptive volumes were very small (between 1945 and 1996). The total volume of magma erupted between 1945 and 1996 is approximately 3.6 9 107 m3. This total magma volume estimate is dominated by the 1995–1996 eruptions and is very small in comparison with a single, moderately sized andesitic eruption (e.g., Bezymianny; Belousov et al. 2002, Colima; Saucedo et al. 2010). Based on the saturation pressure calculated from the H2O and CO2 content of phenocryst-hosted melt inclusions, we determined a magma storage depth of *2–9 km. This compares well to magnetotelluric (MT) data (Ingham et al. 2009) and seismic tomography (Rowlands et al. 2005) from Ruapehu. Ingham et al. (2009) observed a diffuse and weak low-resistivity anomaly that extends to *6 km, which they interpreted to be a dyke system. From *6 to more than 10 km and slightly east of the cone, a more intense low-resistivity anomaly (melt-bearing zone) was identified using both 2-D and 3-D inversions. The seismic tomography data also show a low velocity zone to the east of the cone from *3 to *9 km depth, although these data have been interpreted as a combination of crustal downwarping and the presence of thick volcaniclastic sediments (Rowlands et al. 2005). Our data are consistent with a magma storage region (possibly in the form of discrete, yet closely spaced sills and dykes) down to 9 or 10 km
123
Contrib Mineral Petrol (2013) 166:371–392
(Fig. 11). We suggest that prior to eruption, high-angle (c. 80˚) dykes (calculated from geophysical data) transport magma to the active vent beneath Crater Lake. These dykes probably pass through and interact with partial melt zones that may take the form of small sills (crystal mush zones) (Fig. 11). The sill-like nature of these bodies enables different eruptions to interact with mushes that were chemically and physically isolated from one another, as evinced by the melt inclusion data discussed above. The small volumes of these magma bodies beneath Ruapehu are unlikely to be readily imaged by geophysical techniques such as MT or seismic tomography. In fact, Ingham et al. (2009) use their MT data to suggest that it is unlikely that large volume magma bodies are able to accumulate in the shallow crust beneath Ruapehu. Although physically isolated at shallow depths, the various Ruapehu magmas are likely to be genetically linked at depth, and furthermore, it is possible that a temporal trend can be drawn based on our groundmass glass and major element chemistry data, whereby eruptions have become more mafic with time (since 1945). However, given that there are a number of eruptions for which we do not have samples or analyses (including 1975), we can only speculate that a temporal-chemical trend exists.
Conclusions Historical eruptions at Ruapehu (1945, 1969, 1971, 1977, 1995, and 1996) are characterised by very small volume magmas, each with a unique chemical composition and history. Volatile contents of melt inclusions and crystalmelt barometry have constrained the depth at which these magmas originated to be *2 to *9 km, which corresponds well to geophysical data. These small volume melts probably resided as distinct and closely spaced sills or dykes from 2 to 9 km. Before an eruption, magma was injected into the sill/dyke system leading to common magma-mush and magma–magma interaction. Most magmas interacted with crystal mush zones (at\*3 km depth) formed from antecedent magmas during ascent and eventually eruption. Due to their small volumes, Ruapehu magmas since 1945 show sensitivity to interaction between magmas and crystal mush zones that would be difficult to determine in larger magmatic systems. Therefore, data from Ruapehu offer a unique insight into the small-scale interactions that magmas experience on their ascent to eruption. We have shown that the chemical composition of phenocryst-hosted melt inclusions is often distinct from the groundmass glass. This implies that a significant proportion of the crystals are antecrysts; in some cases, antecrysts have incorporated rare melt inclusions from the new melt.
Contrib Mineral Petrol (2013) 166:371–392
In that respect, it is clear that the interpretation of magmatic processes at depth can only be achieved from cognate, rather than exotic, melt inclusions. Ruapehu magmas are low in H2O but CO2-rich compared to intermediate magmas from subduction settings elsewhere. This relatively low concentration of volatiles and the very small volumes of magma combine to account for the low explosivity and short duration of most eruptions at Ruapehu. While the largest eruptive episode (in 1995–1996) produced plumes to 20 km, the volatile content is similar to the smallest episode analysed (in 1971). Therefore, the controls on the size of eruptions at Ruapehu are not determined by the volatile content of magma alone. Acknowledgments This work was funded by the New Zealand Ministry of Science and Innovation (MSI) Geological Hazards Programme (GHZ) in the form of a PhD studentship to GK at the University of Bristol. Holly Goddard and Neville Orr are thanked for their assistance with sample preparation. We gratefully acknowledge support from NERC for access to the SIMS facility, Edinburgh, where Cees-Jan de Hoog provided expert guidance and patience. Stuart Kearns and Ben Buse are thanked for their support with EPMA and SEM analyses. JB is supported by ERC Advanced Grant ‘‘CRITMAG’’ and a Royal Society Wolfson Research Merit Award. KC acknowledges funding from the AXA Research Fund. Two anonymous reviewers are thanked for constructive and helpful reviews that helped us clarify and significantly improve the manuscript.
References Atlas ZD, Dixon JE, Sen G, Finny M, Martin-Del Pozzo AL (2006) Melt inclusions from Volca´n Popocate´petl and Volca´n de Colima, Mexico: melt evolution due to vapor-saturated crystallization during ascent. J Volcanol Geoth Res 153:221–240 Bacon CR, Hirschmann MM (1988) Mg/Mn partitioning as a test for equilibrium between coexisting Fe–Ti oxides. Am Mineral 73:57–61 Beck AC (1950) Volcanic activity at Mt. Ruapehu from August to December 1945. NZ Journ Sci Tech B31:1–13 Belousov A, Voight B, Belousova M, Petukhin A (2002) Powerful pyroclastic surge in the May 8–10, 1997 explosive eruption of Bezymianny volcano, Kamchatka, Russia. Bull Volcanol 64:455–471 Blundy JD, Cashman KV (2005) Rapid decompression-driven crystallization recorded by melt inclusions from Mount St. Helens volcano. Geology 33:793–796 Blundy JD, Cashman KV (2008) Petrologic reconstruction of magmatic system variables and processes. Rev Mineral Geochem 69:179–239 Blundy JD, Cashman KV, Rust A, Witham F (2010) A case for CO2rich arc magmas. Earth Planet Sci Lett 290:289–301 Bryan C, Sherburn S (1999) Seismicity associated with the 1995–1996 eruptions of Ruapehu volcano, New Zealand: narrative and insights into physical processes. J Volcanol Geoth Res 90:1–18 Christenson BW (2000) Geochemistry of fluids associated with the 1995–1996 eruption of Mt. Ruapehu, New Zealand: signatures and processes in the magmatic-hydrothermal system. J Volcanol Geoth Res 97:1–30 Christenson BW, Reyes AG, Young R, Moebis A, Sherburn S, ColeBaker JC, Britten KA (2010) Cyclic processes and factors
391 leading to phreatic eruption events: insights from the 25 September 2007 eruption through Crater Lake, Ruapehu. New Zealand. J Volcanol Geoth Res 191(1–2):15–32 Cronin SJ, Hedley MJ, Neall VE, Smith RG (1998) Agronomic impact of tephra fallout from the 1995 and 1996 Ruapehu Volcano eruptions, New Zealand. Environ Geol 34:21–30 Danyushevsky LV, Plechov P (2011) Petrolog3: integrated software for modelling crystallization processes. Geochem Geophys Geosys 12:Q07021. doi:10.1029/2011GC003516 Devine JD, Murphy MD, Rutherford MJ, Barclay J, Sparks RSJ, Carroll MR, Young SR, Gardner JE (1998) Petrologic evidence for pre-eruptive pressure–temperature conditions, and recent reheating, of andesite magma at Soufrie`re Hills Volcano, Montserrat, W.I. Geophys Res Let 25:3669–3672 Donoghue SL (1991) The Tufa Trig Formation: recent (0–1800 years B.P.) eruptives from Mount Ruapehu. Geological Society of New Zealand miscellaneous publication 59A:54 p Faure F, Schiano P (2005) Experimental investigation of equilibration conditions during forsterite growth and melt inclusion formation. Earth Planet Sci Lett 236:882–898 Frost BR (1991) Introduction to oxygen fugacity and its petrologic importance. Rev Mineral Geochem 25:1–9 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. doi:10.1130/G32992.1 Gamble JA, Wood CP, Price RC, Smith IEM, Stewart RB, Waight T (1999) A fifty year perspective of magmatic evolution on Ruapehu Volcano, New Zealand: verification of open system behaviour in an arc volcano. Earth Planet Sci Lett 170:301–314 Gamble JA, Price RC, Smith IEM, McIntosh WC, Dunbar NW (2003) 40 Ar/39Ar geochronology of magmatic activity, magma flux and hazards at Ruapehu volcano, Taupo Volcanic Zone, New Zealand. J Volcanol Geochem Res 120:271–287 Ghiorso MS, Sack RO (1995) Chemical mass transfer in magmatic processes. IV. A revised and internally consistent thermodynamic model for the interpolation and extrapolation of liquidsolid equilibria in magmatic systems at elevated temperatures and pressures. Contrib Mineral Petrol 119:197–212 Graham IJ, Hackett WR (1987) Petrology of andc-alkaline lavas from Ruapehu Volcano and related vents, Taupo Volcanic Zone, New Zealand. J Petrol 28:531–567 Gregg DR (1960) Volcanoes of Tongariro National Park. NZ DSIR Info Ser 28, 82 p Hackett WR (1985) Geology and petrology of Ruapehu volcano and related vents. PhD Dissertation, Victoria University, Wellington Hackett WR, Houghton BF (1989) A facies model for a Quaternary andesitic composite volcano: Ruapehu, New Zealand. Bull Volcanol 51:51–68 Hattori K, Sato H (1996) Magma evolution recorded in plagioclase zoning in 1991 Pinatubo eruption products. Am Mineral 81:982–994 Hauri EH (2002) SIMS investigations of volatiles in volcanic glasses, 2: abundances and isotopes in Hawaiian melt inclusions. Chem Geol 183:115–141 Healy J, Lloyd EF, Rushworth DEH, Wood CP, Glover RB, Dibble RR (1978) The eruption of Ruapehu, New Zealand on 22 June 1969. N. Z. DSIR Bull 224:1–80 Houghton BF, Latter JH, Hackett WR (1987) Volcanic hazard assessment for Ruapehu composite volcano, Taupo Volcanic Zone, New Zealand. Bull Volcanol 49:737–751 Houghton B, Neall V, Johnston D (1996) Eruption!. Penguin Books, Auckland, p 48 Humphreys MCS, Kearns SL, Blundy JD (2006) SIMS investigation of electron-beam damage to hydrous, rhyolitic glasses: implications for melt inclusion analysis. Am Mineral 91:667–679 Hurst AW (1998) Shallow seismicity beneath Ruapehu Crater Lake: results of a 1994 seismometer deployment. Bull Volcanol 60:1–9
123
392 Ingham MR, Bibby HM, Heise W, Jones KA, Cairns P, Dravitzki S, Bennie SL, Caldwell TG, Ogawa Y (2009) A magnetotelluric study of Mount Ruapehu volcano, New Zealand. Geophys J Int 179:887–904 Johnston DM, Houghton BF, Neall VE, Ronan KR, Paton D (2000) Impacts of the 1945 and 1995–1996 Ruapehu eruptions, New Zealand: an example of increasing societal vulnerability. Geol Soc Am Bull 112:720–726 Jolly A, Sherburn S, Jousset P, Kilgour G (2010) Eruption source processes derived from seismic and acoustic observations of the 25 September 2007 Ruapehu eruption, North Island, New Zealand. J Volcanol Geoth Res 191:33–45 Kilgour GN, Jolly AD, Sherburn S, Scott B, Miller C, Rae AJ (2007) The October 2006 eruption of Ruapehu Crater Lake. In: Mortimer N, Wallace L (eds) Joint GSNZ/NZGS Ann Conf GSNZ, Tauranga Kilgour G, Manville V, Della Pasqua F, Graettinger A, Hodgson KA, Jolly GE (2010) The 25 September 2007 eruption of Mount Ruapehu, New Zealand: directed ballistics, surtseyan jets, and ice-slurry lahars. J Volcanol Geotherm Res 191:1–14 Lepage LD (2003) ILMAT: an Excel worksheet for ilmenitemagnetite geothermometry and geobarometry. Comput Geosci 29:673–678 Lindsley DH, Anderson DJ (1983) A two-pyroxene thermometer. J Geophys Res 88:A887–A906 Lindsley DH, Frost BR (1992) Equilibria among Fe–Ti oxides pyroxenes, olivine and quartz: part 1. Theory. Am Mineral 77:987–1003 Liu Y, Anderson AT, Wilson CJN, Davis AM, Steele IM (2006) Mixing and differentiation in the Oruanui rhyolitic magma, Taupo, New Zealand: evidence from volatiles and trace elements in melt inclusions. Contrib Mineral Petrol 151:71–87 Lloyd AS, Plank T, Ruprecht P, Hauri EH, Rose W (2013) Volatile loss form melt inclusions in pyroclasts of differing size. Contrib Mineral Petrol 165:129–153 Moore G, Carmichael ISE (1998) The hydrous phase equilibria (to 3 kbar) of an andesite and basaltic andesite from western Mexico: constraints on water content and conditions of phenocryst growth. Contrib Mineral Petrol 130:304–319 Mordret A, Jolly AD, Duputel Z, Fournier N (2010) Monitoring of phreatic eruptions using interferometry on retrieved crosscorrelation function from ambient seismic noise. J Volcanol Geotherm Res 191:46–59 Murphy MD, Sparks RSJS, Barclay J, Carroll MR, Lejeune A-M, Brewer TS, MacDonald R, Black S, Young S (1998) The role of magma mixing in triggering the current eruption at the Soufriere Hills volcano, Montserrat, West Indies. Geophys Res Let 25(18):3433–3436 Nairn IA, Wood CP, Hewson CAY (1979) Phreatic eruptions of Ruapehu: April 1975. N Z J Geol Geophy 22:155–173 Nakagawa M, Wada K, Thordarson T, Wood CP, Gamble JA (1999) Petrological investigations of the 1995 and 1996 eruptions of Ruapehu volcano, New Zealand: formation of discrete and small magma pockets and their intermittent discharge. Bull Volcanol 61(1–2):15–31 Nakagawa M, Wada K, Wood PC (2002) Mixed magmas, mush chambers and eruption triggers: evidence from zoned clinopyroxene phenocrysts in andesitic scoria from the 1995 eruptions of Ruapehu volcano, New Zealand. J Petrol 43:2279–2303
123
Contrib Mineral Petrol (2013) 166:371–392 Nakamura M (1995) Continuous mixing of crystal mush and replenished magma in the ongoing Unzen eruption. Geol 23:807–810 Oliver RL (1945) Further activity of Mount Ruapehu, May-July 1945. NZ J Sci Tech B26:24–33 O’Neill HSC, Pownceby MI (1993) Thermodynamic data from redox reactions at high temperatures. I. An experimental and theoretical assessment of the electrochemical method using stabilized zirconia electrolytes, with revised values for the Fe– ‘‘FeO’’, Co– CoO, Ni–NiO and Cu–Cu20 oxygen buffers, and new data for the W–WO2 buffer. Contrib Mineral Petrol 114:296–314 Papale P, Moretti R, Barbato D (2006) The compositional dependence of the saturation surface of H2O ? CO2 fluids in silicate melts. Chem Geol 229(1–3):78–95. doi:10.1016/j.chemgeo.2006. 01.013 Pardo N, Cronin SJ, Palmer AS, Ne´meth K (2011) Reconstructing the largest explosive eruptions of Mt. Ruapehu, New Zealand: lithostratigraphic tools to understand subplinian–plinian eruptions at andesitic volcanoes. Bull Volcanol. doi:10.1007/s00445011-0555-z Portnyagin M, Hoernle K, Plechov P, Mironov N, Khubunaya S (2007) Constraints on mantle melting and composition and nature of slab components in volcanic arcs from volatiles (H2O, S, Cl, F) and trace elements in melt inclusions from the Kamchatka Arc. Earth Planet Sci Lett 255:53–69 Prata AJ, Grant IF (2001) Retrieval of microphysical and morphological properties of volcanic ash plumes from satellite data: application to Mt Ruapehu, New Zealand. Q J R Meteorol Soc 127:2153–2179 Price RC, Gamble JA, Smith IEM, Maas R, Waight TE, Stewart RB, Woodhead J (2012) The anatomy of an andesitic volcano: a time-stratigraphic study of andesite petrogenesis and crustal evolution at Ruapehu Volcano, New Zealand. J Petrol 53: 2139–2189 Putirka KD (2008) Thermometers and barometers for volcanic systems. Rev Mineral Geochem 69:61–120 Reed JJ (1945) Activity at Ruapehu, March-April 1945. NZ J Sci Tech B26:17–23 Roedder E (ed) (1984). Fluid Inclusions. Mineralogical Society of America, Rev Mineral Geochem 14 Saucedo R, Macı´as JL, Gavilanes JC, Arce JL, Komorowski JC, Gardner JE, Valdez-Moreno G (2010) Eyewitness, stratigraphy, chemistry, and eruptive dynamics of the 1913 Plinian eruption of Volca´n de Colima, Me´xico. J Volcanol Geotherm Res 191: 149–166 Saunders K, Blundy JD, Dohman R, Cashman KV (2012) Linking petrology and seismology at an active volcano. Science 336: 1023–1027 Sherburn S, Bryan CJ, Hurst AW, Latter JH, Scott BJ (1999) Seismicity of Ruapehu volcano, New Zealand 1971–1996: a review. J Volcanol Geotherm Res 88:255–278 Spilliaert N, Allard P, Me´trich N, Sobolev AV (2006) Melt inclusion record of the conditions of ascent, degassing, and extrusion of volatile-rich alkali basalt during the powerful 2002 flank eruption of Mount Etna (Italy). J Geophys Res 111:B04203 Wallace P (2005) Volatiles in subduction zone magmas: concentrations and fluxes based on melt inclusion and volcanic gas data. J Volcanol Geoth Res 140:217–240