| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 44 | Number 9 | Pages 1525-1559 | 2003
© Oxford University Press 2003
Kilauea East Rift Zone Magmatism: an Episode 54 Perspective
1 US GEOLOGICAL SURVEY, CASCADES VOLCANO OBSERVATORY, VANCOUVER, WA 98683, USA
2 US GEOLOGICAL SURVEY, HAWAIIAN VOLCANO OBSERVATORY, HAWAII NATIONAL PARK, HI 96718, USA
3 US GEOLOGICAL SURVEY, DENVER FEDERAL CENTER, DENVER, CO 80225, USA
* Corresponding author. E-mail: cthornber{at}usgs.gov
RECEIVED DECEMBER 12, 2001; ACCEPTED MARCH 11, 2003
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONOn January 29–30, 1997, prolonged steady-state effusion of lava from Pu'u'O'o was briefly disrupted by shallow extension beneath Napau Crater, 1–4 km uprift of the active Kilauea vent. A 23-h-long eruption (episode 54) ensued from fissures that were overlapping or en echelon with eruptive fissures formed during episode 1 in 1983 and those of earlier rift zone eruptions in 1963 and 1968. Combined geophysical and petrologic data for the 1994–1999 eruptive interval, including episode 54, reveal a variety of shallow magmatic conditions that persist in association with prolonged rift zone eruption. Near-vent lava samples document a significant range in composition, temperature and crystallinity of pre-eruptive magma. As supported by phenocryst–liquid relations and Kilauea mineral thermometers established herein, the rift zone extension that led to episode 54 resulted in mixture of near-cotectic magma with discrete magma bodies cooled to
1100°C. Mixing models indicate that magmas isolated beneath Napau Crater since 1963 and 1968 constituted 32–65% of the hybrid mixtures erupted during episode 54. Geophysical measurements support passive displacement of open-system magma along the active east rift conduit into closed-system rift-reservoirs along a shallow zone of extension. Geophysical and petrologic data for early episode 55 document the gradual flushing of episode 54 related magma during magmatic recharge of the edifice. KEY WORDS: Hawaii; Kilauea; glass thermometry; magma mixing; mineral thermometry
|
INTRODUCTION |
|---|
The well-monitored episode 54 disruption in steady-state lava effusion of Kilauea volcano presents an opportunity to evaluate the origins of mixed-magma components that are inherently associated with prolonged rift zone eruption. Like a skylight above an active lava tube, which provides an opportunity to study the physical and chemical processes associated with lava transport from the vent, the eruptive fissures of episode 54 are windows into the magmatic plumbing system between the summit reservoir and Pu'u'O'o.
For the past two decades, amidst nearly continuous rift zone eruption and recharge of new magma into the summit region, progressive outward movement of the south flank of Kilauea has accommodated an increasingly efficient magma conduit through the shallow upper–middle east rift zone. Details of magmatic conditions that may persist within and along this 16-km-long rift conduit between the summit and the vent are poorly understood, but may be broadly inferred from what is known of systematic differences in lava chemistry between summit and rift zone eruptions of Kilauea and Mauna Loa volcanoes. Kilauea summit eruptions are primarily restricted to olivine-phyric tholeiites with MgO contents of >6·8 wt %, as regulated by olivine fractionation and accumulation within the summit reservoir (Wright & Fiske, 1971). Rhodes and coworkers [see Rhodes (1995) and references therein] emphasized the preponderance of lava compositions at the low-MgO end of an olivine control trend (
7·0 wt % MgO) for both summit and rift eruptions of Kilauea and Mauna Loa. They proposed that such compositions reflect the persistence of shallow open-system reservoirs where magma is chemically ‘buffered’ at near-cotectic compositions as a result of continuous replenishment of primitive melt. Both Rhodes (1995) and Fodor et al. (1993) interpreted Hawaiian phenocryst assemblages as indicators of open-system fractionation within actively replenished, shallow magma reservoirs.
In contrast to the limited range of olivine-saturated magma that characterizes summit eruptions, a broader array of lava chemistry is erupted along rift zones. Historic Kilauea rift eruptions with relatively evolved compositions are restricted to short-lived events following repose intervals of decades (Wright, 1971; Wright & Fiske, 1971; Wright & Helz, 1996). Variations of plagioclase and pyroxene phenocryst compositions and textures in these small volumes of evolved lava indicate pre-eruptive mixing between rift-stored, fractionated magma and rift-transported, olivine-saturated magma (Helz & Wright, 1992; Fodor & Moore 1994; Clague et al., 1995; Wright & Helz, 1996; Yang et al., 1999).
Evidence that pockets of magma have persisted along Kilauea's east rift zone during the present eruption is provided by petrologic and geophysical studies. As demonstrated by Wolfe et al. (1988) and Garcia et al. (1989), olivine–plagioclase–clinopyroxene-saturated tholeiite erupted through fissures extending 7·5 km from the west edge of Napau Crater during episodes 1–3 (Fig. 1) were hybrid mixtures resulting from magmatic dike propagation through pre-existing reservoirs of cooler magma isolated within the rift. Pockets of rift-resident magma may also persist as open-system reservoirs maintained in connection with fresh magma supplied from the summit region, as is suggested by the persistence of shallow deformation sources beneath Makaophui Crater and uprift of the Pu'u'O'o vent (Okamura et al., 1988; Hoffman et al., 1990; Owen et al., 2000).
|
Approximately 97% of the
2 km3 of lava produced during this 20-year-old rift zone eruption is weakly olivine-phryic, with minor chromian spinel, and has MgO contents ranging from 6·8 to 9·6 wt % (Thornber, 2003). The majority of this volume has been issued during prolonged intervals of near steady-state eruption from July 1986 to January 1997 (episodes 48–53) and since mid-1997 (episode 55). During these episodes, olivine–liquid relations of near-vent samples indicate near-equilibrium crystallization from melts equivalent to their bulk composition, without significant loss or accumulation of olivine during transit along the length of the rift-conduit (Garcia et al., 1996; Thornber, 2001). Cyclic variations of lava composition and temperature observed throughout prolonged eruptive intervals portray the repetition of a limited range of olivine-saturated magmatic conditions, the Mg end-members of which are maintained by open-system recharge of the edifice beneath the summit region (Thornber, 2003). Steady-state magmatic and eruptive conditions at Kilauea were briefly disrupted on January 29–30, 1997, when fissures opened in Napau Crater, uprift of the active vent (episode 54). For the first time since the onset of this eruption in 1983, lower-temperature and fractionated lava compositions were erupted that are consistent with derivation from a mixture of rift-stored magma and magma in the active eruption conduit (Thornber et al., 1997; Garcia et al., 2000; Thornber, 2001). In this paper, we present and assess the detailed petrography and geochemistry of minerals, glasses and bulk samples of lava erupted before, during and after episode 54. Our petrologic interpretations of Kilauea rift zone magmatism are constrained by extensive geological and geophysical data. A brief summary of eruptive behavior and geophysical evidence for magma movement along the summit-to-vent conduit, before, during, and after episode 54 is presented below. This information provides important geologic context for the petrologic modeling of rift zone magmatic processes that follows.
|
ERUPTION NARRATIVE—BEFORE, DURING, AND AFTER EPISODE 54 |
|---|
The Pu'u'O'o eruption began on January 3, 1983, when fissures opened in Napau Crater, 12 km downrift from the summit. The onset of eruptive activity was preceded by summit deflation as a dike propagated down the east rift zone (Okamura et al., 1988). The eruption soon localized at the eventual site of Pu'u'O'o, which, once established, was the main vent from 1983 to 1986 and again from 1992 to the present, after 6 years of eruption from Kupaianaha. A more detailed narrative account of post-1983 eruptive activity has been presented by Heliker & Mattox (2003, and references therein). From mid-1994 until 1997, episode 53 of the eruption persisted in a steady-state mode with crater pond activity at Pu'u'O'o accompanied by a single continuously active vent on the uprift side of the cone. The flank vent fed a lava tube system that extended 11 km to the ocean (Fig. 1).
During the early evening hours of January 29, 1997, episode 53 ended dramatically when the Pu'u'O'o lava pond and west flank vent suddenly drained and the crater floor and west wall of the cone collapsed into the void. As confirmed by onsite observations and GOES satellite imagery (Harris et al., 1997), episode 54 began at 0240 h (HST) on January 30, when a new eruptive fissure opened on the floor of Napau Crater. During the next 22 h, 300 000 m3 of lava (dense rock equivalent) were erupted from six fissures along a northeasterly line in the Napau Crater area, 2–4 km uprift of Pu'u'O'o. These eruptive fissures were overlapping and en echelon with fissures that marked the onset of this eruption in 1983 (Fig. 2) and with earlier fissure eruptions in 1963 and 1968 (Moore & Koyanagi, 1969; Jackson et al., 1975). Fissures A–D erupted along a line of 1·3 km length, striking N60°E from the center of Napau Crater. Fissure E, 130 m in length, opened en echelon and 150 m SE of fissure D, within a pre-existing graben. An unexpected change in the northeastward progression of erupted fissures A–E occurred when fissure F opened in the west wall of Napau Crater, on-strike with earlier fissures but SW of the initial fissure, A. Most of the lava from fissure F poured out of the crater wall and ponded at its base. This last gasp of episode 54 concluded at 0033 h on January 31.
|
On the morning of February 1, observers at Pu'u'O'o saw that the west side of the cone had collapsed and the crater floor had dropped 150 m to a depth of 210 m below the low point on the crater rim, 65 m below the pre-1983 surface. Seismic records indicate that drain-back and floor collapse at Pu'u'O'o occurred between 1845 and 2015 h on January 29. At 2015 h, a low rumbling roar was heard from the Pu'u'O'o vent area, signaling the collapse of the west side of the Pu'u'O'o cone, after magma had drained and the crater floor had collapsed into the void.
The end of episode 54 was followed by a 24 day eruptive pause, the longest hiatus in the eruption since 1986. On February 24, 1997, the first lava of episode 55 appeared deep within the Pu'u'O'o crater. Initial episode 55 activity was limited to an active lava pond fed from the uprift edge of the crater. The pond rose steadily to the elevation of flank vents that began erupting on March 28. After 3 months of shield building around sporadically active Pu'u'O'o flank vents, a predominant vent on the SW flank of the cone began to steadily feed a new easterly flow (Cashman et al., 1999). This unstable interval of early episode 55 was characterized by interplay of Pu'u'O'o pond and flank vent activity. The crater overflowed several times in 1997, producing the first lava flows originating from the Pu'u'O'o crater since 1986.
Through successive advancement of flows and tubes, lava reached the coast in mid-July and, by August 1997, a relatively stable tube system of 11 km length was serving as a lava pipeline from the vent to the ocean [see Harris & Thornber (1999) for detailed maps]. As was the case during episode 53, the mid-1997 to 1999 interval of episode 55 was punctuated by brief and intermittent eruptive pauses.
|
GEOPHYSICAL SIGNATURE OF SHALLOW MAGMA MOVEMENT—BEFORE, DURING AND AFTER EPISODE 54 |
|---|
Throughout the Pu'u'O'o–Kupaianaha eruption, magmatic continuity between a summit reservoir and rift vents is demonstrated by correlations between summit inflation and deflation and surges and lulls in eruptive activity (Okamura et al., 1988; Wolfe et al., 1988; Garcia et al., 1996; Thornber et al., 1996; Denlinger, 1997; Heliker et al., 1998; Thornber, 2001; Cervelli et al., 2002). During near-steady-state eruption of episode 53, an efficient pressure balance between the summit reservoir and the Pu'u'O'o vent is indicated by correlation of precursory summit deformation and short-term fluctuations in eruptive vigor at the vent (e.g. Thornber et al., 1995, 1996). In contrast to this summit-controlled eruptive behavior, the timing of events associated with episode 54 indicates that sudden shallow rift zone extension beneath Napau Crater preceded magma withdrawal from the summit reservoir. An annotated chronology of eruptive activity, seismicity and deformation associated with episode 54 is presented in Fig. 3.
|
, formation time of individual samples used in this investigation, numbered here with letters corresponding to each of the fissures sampled. (b) RSAM at MPR station located on the south rim of Makaopuhi Crater. Black bars indicate number of discrete earthquakes per 10 min intervals located in the vicinity of Makaopuhi, Napau and Pu'u'O'o (right axis scale). Earthquakes counted range from magnitude 0·9 to 3·0 at a depth range of 0–4·5 km. (c) RSAM at NPT station located within Kilauea's summit caldera. Black bars indicate the number of discrete earthquakes located in the summit and upper-east-rift region; magnitude and depth range as in (b). (d) Summit tilt (microradians) measured electronically near the west rim of the summit caldera. Deflation of summit is indicated by a decrease in microradians. Distinct rift zone seismic activity commenced 8 h before the episode 54 fissure eruption, when a swarm of shallow earthquakes (1–4 km depth) occurred near Makaopuhi Crater at 1841 h (HST) on January 29, 1997. At 1845 h, a dramatic increase in shallow tremor was recorded at STC and MPR stations (a, b), located between Napau Crater and Pu'u'O'o and at Makaopuhi Crater, respectively (see Global positioning system (GPS) deformation data presented by Owen et al. (2000) constrain the timing and geometry of summit and upper east rift zone deformation associated with magma displacement during episode 54. Their model suggests that 2 m of rift extension occurred at depths <2·5 km in the Napau Crater area during the January 1997 event. This value corresponds to the 1·8 m of cumulative surface extension estimated from our field measurements of fresh surface cracks paralleling episode 54 fissures in Napau Crater. Continuous GPS data show that the rift opening began 8 h before fissure eruption, concurrent with shallow earthquakes in this zone and before the onset of summit deflation (Fig. 3). Owen et al. (2000) demonstrated that rift-extension rates of
4 cm/h at the onset of the event decreased with time before and during fissure eruptions at Napau Crater, and they recognized a second point-source of deflation associated with this event at shallow depths beneath Makaopuhi Crater. Their geodetic model of magma withdrawal from beneath Makaopuhi crater is consistent with seismic records indicating shallow tremor and earthquake swarms located near Makaopuhi before and during the episode 54 fissure eruptions (see Fig. 3 caption). This geophysical suggestion of magma displacement from an open-system rift-reservoir sustained beneath Makaopuhi is supported by leveling data obtained in 1983 at the onset of this eruption, which identified the same point-source for inflation related to magma intrusion (Okamura et al., 1988). Steady summit inflation associated with magma recharge began when episode 54 ended and continued until steady-state magmatic and eruptive conditions were re-established in July 1997 (Owen et al., 2000; Thornber, 2001). The timing of the onset of episode 55 Pu'u O'o activity was accurately forecast at the Hawaiian Volcano Observatory by assuming continued magma recharge at a rate equivalent to an average effusion rate of episode 53.
|
PETROLOGIC SAMPLING |
|---|
The best possible assessment of pre-eruptive magmatic conditions through time is achieved by frequent sampling and analysis of vent tephra and spatter, including Pele's tears and hair, of tube-contained lava flows and of surface lava flows at the vent. Details of collection techniques used to obtain samples from 1994 to 1999 have been presented by Thornber et al. (2002) along with detailed time, location and sampling information. During periods of steady-state eruption of late episode 53 (September 1994–January 1997) and episode 55 (July 1997–January 1999), 306 rapidly quenched lava samples from near-vent locations were obtained at near-weekly intervals. Twenty-two lava samples were obtained during episode 54, the locations of which are shown in Fig. 2. The majority of these samples were quenched from a molten state at the time of collection; others were collected while still warm. The formation time of individual samples from each of the eruptive fissures (A–F) is indicated in Fig. 3a. During early episode 55 (April–July 1997), 22 samples of fountain spatter and spatter cone overflows or flow breaches from sporadic vents on the west and SW flanks of the Pu'u'O'o cone were obtained.
|
ANALYTICAL METHODS |
|---|
Microbeam quantitative analyses and imagery used in this study were performed at the US Geological Survey (USGS) Denver Microbeam Facility using a JEOL 8900 electron microprobe and a JEOL 5900 scanning electron microscope. Qualitative compositional information of electron backscatter imagery was used to augment petrographic assessments of mineral proportions and zonation characteristics. Thornber et al. (2002) has provided details of microprobe operating conditions, analytical accuracy and precision. Major element glass compositions used for all plots and calculations are the average of 10 analyses per sample. Estimated errors in MgO and total FeO (FeOt = Fe2O3 + FeO) concentrations used in glass thermometry calculations [T°C = 20·1 x wt % MgO + 1014°C (Helz & Thornber, 1987)] and in calculations involving olivine–liquid and pyroxene–liquid MgO/FeO distribution coefficients [KD = (MgO/FeO)xl/(MgO/FeO)Liq (Roeder, 1974; Grove & Bryan, 1983)] have been presented by Thornber (2001). As substantiated by wet chemical analysis (see Thornber, 2001), ferrous iron concentrations in whole rocks and glasses are calculated from the total iron oxide concentration assuming an average FeO/(FeO + Fe2O3) value of 0·9.
Whole-rock major element analyses were performed at the USGS analytical facility in Denver using X-ray fluorescence (XRF) techniques as described by Taggart et al. (1987). All major element data reported are normalized to 100% after correction for <1 wt % of volatile loss-on-ignition. Repeated XRF analyses of four USGS Kilauea standards (wt % MgO from 5·4 to 14·3) provided analytical control in each of 13 sample batches analyzed from 1994 to 1999, details of which have been presented by Thornber et al. (2002). Standard reproducibilities are less than 1
= 0·01 wt % for K2O and MnO, 0·02 for TiO2 and P2O5, 0·03 for CaO and Na2O, 0·05 for Al2O3, FeO (total-iron) and MgO, and 0·13 for SiO2.
Whole-rock trace element analyses of samples were obtained by instrumental neutron activation analysis (INAA) at USGS Denver; the techniques and analytical precision have been described by Baedecker & McKown (1987). Glass trace element analysis of small (1 mm–1 cm) episode 53 Pele's tear samples were conducted by laser ablation–inductively coupled plasma mass spectrometry (LA–ICP-MS) at USGS, Denver. Repeated analyses of a secondary standard basaltic glass, ENDV, indicates precision for most elements of 1–1·6% (Ridley & Lichte, 1998).
|
WHOLE-ROCK MAJOR AND TRACE ELEMENT GEOCHEMISTRY |
|---|
Bulk lava and glass major element chemistry for individual samples used in this study have been provided by Thornber et al. (2002) and can be accessed on the internet at http://geopubs.wr.usgs.gov/open-file/of02-017. The average and variance of major and trace element compositions of bulk lava and Pele's tear glasses collected during steady-state eruption intervals of episodes 53 and 55 are presented in Table 1. Major and trace element data for episode 54 and early episode 55 samples are provided in Tables 2 and 3, respectively.
|
|
|
Whole-rock MgO variation of lava erupted throughout the 1994–1999 interval (including episode 54, Fig. 4) defines an apparent trend of progressive low-pressure cooling and multiphase crystallization of Kilauean magma (e.g. see discussion by Thompson & Tilley, 1969; Helz & Thornber, 1987; Clague et al., 1995; Yang et al., 1996). Bulk MgO values of episode 54, fissure A–E lava cluster near an average of 6·4 wt % (1
= 0·9, for 13 samples) and the average MgO content of fissure F lava is 5·8 wt % (1 = 0·06 for nine samples). In episode 54 lava, moderately incompatible elements that are incorporated by plagioclase and clinopyroxene (e.g. Ca and Sc) below the multiphase cotectic at 7 wt % MgO are depleted, whereas highly incompatible elements (e.g. large ion lithophiles, rare earths and Ti, Zr, Hf, etc.) are variably enriched relative to olivine-controlled compositions. Episode 53 and 55 bulk lava compositions are within a range of 7·6–9·6 wt % MgO and the incompatible element variation in this range is consistent with olivine-only fractionation. Pu'u'O'o tear glasses from episodes 53 and 55 are limited to a small range of near-cotectic compositions with an average wt % MgO of 6·9 (1 = 0·16 for 90 samples) and 7·4 (1 = 0·13, for 38 samples), respectively (Fig. 4). Major and trace element variation of tear glasses samples are closely matched by the 7% MgO normalized composition (as determined by equilibrium olivine subtraction) and define the low-Mg end of the olivine control trends of episode 53 and 55 whole-rock data.
|
As pointed out by Garcia et al. (2000), differences between incompatible element ratios of episode 54 and steady-state episode 53 and 55 lava preclude a simple petrogenetic relation of comagmatic fractionation. A hybrid character of episode 54 lava is independently suggested by the geochemical data of this investigation. A comparison of relative incompatible-element variations (e.g. La/Yb vs Ce) for episode 54 lava with that of prior and subsequent steady-state eruption products (Fig. 5) demonstrates that the recent Napau Crater lava has a magmatic component that is chemically distinct from magma within the active rift conduit. Determining the nature of the mix is an objective of this investigation to assess magmatic conditions that persist within the rift zone.
|
Twenty-two vent samples collected during early episode 55 show a time progression from lower to higher Mg contents that span the range between those of steady-state bulk-liquids and tear glasses (Table 3, Fig. 4). There is petrologic evidence (presented below) indicating that this apparent olivine-controlled fractionation trend also results from mixing of magma within the shallow rift zone.
Petrographic overview
Episode 53 and episode 55 steady-state Pele's tears, vent spatter and near-vent tube samples typically consist of vesicular glass and 1–4 vol. % of 0·6–0·7 mm, subhedral to euhedral olivine phenocrysts. Microphenocrysts of chromian spinel are commonly included within olivine crystals, which also frequently contain glass inclusions. A detailed account of petrographic and chemical variations of chromian spinel in episode 53–55 steady-state lava has been provided by Roeder et al. (2003). Acicular microlites of plagioclase (usually <50 µm long) and associated clinopyroxene are occasionally observed in upper elevation tube flow and vent spatter samples and are rarely observed in Pele's tears.
Despite the chemical differences between lava from fissures A–E and the more evolved fissure F lava, overall petrographic characteristics of episode 54 samples are indistinguishable from one another and clearly distinct from olivine-phyric episode 53 and 55 lava. As determined by the bulk-rock density methods of Houghton & Wilson (1989), episode 54 lava samples are highly vesicular, ranging from 30 to 80 vol. % vesicles with no clear correlation between vesicularity and composition or fountain height. Matrices are predominantly aphyric with proportions of glass ranging from 1 to 80 vol. %, varying significantly within and between samples owing to differences in extent of microphyric to cryptocrystalline groundmass crystallization during eruption and quenching. The most striking petrographic features observed in electron backscatter images of these samples are the diverse textural and zonation patterns among phenocrysts (0·3–1 mm and rarely up to 3 mm) and microphenocrysts (0·1–0·3 mm) (Figs 6 and 7). Phenocrystic and microphenocrystic plagioclase, clinopyroxene, and olivine (± chromian spinel) are present in all samples, either as isolated grains or in glomerophyric aggregates of phenocrysts or microphenocrysts. The phenocryst population ranges up to 5 vol. % in these samples and includes an assortment of rounded, euhedral and angular fragments of clinopyroxene and plagioclase that are occasionally intergrown as aggregates or clots and may be unzoned, normally zoned, reversely zoned or display more complex zonation patterns. This population also includes normally zoned and resorbed olivine and a resorbed and a reversely zoned orthopyroxene crystal, present in one fissure F sample. In addition, glass inclusions are preserved in some olivine, plagioclase and clinopyroxene phenocrysts. Examples of episode 54 phenocrysts, microphenocrysts and glomerophryic aggregates are shown in Fig. 6 and discussed in more detail in the following sections pertaining to individual mineral groups. Additional examples of resorbed, normally zoned and overgrown crystals of olivine in episode 54 lava have been presented by Thornber (2001).
|
|
|
Typical glomerophyric aggregates of microphenocrysts are
0·5–1 mm in diameter, ranging up to 3 mm. These microphenocrystic clots usually consist of plagioclase and clinopyroxene and, in some cases, they include intercrystalline glass that is isolated and compositionally distinct from matrix glasses. Microphenocrystic olivine–plagioclase clusters are present in some samples and olivine microphenocrysts are occasionally observed in association with plagioclase–clinopyroxene intergrowths. In many samples, aggregated microphenocrysts are texturally distinct amidst glassy or microphyric groundmass (Fig. 6h). Such aggregates occasionally appear disrupted by flow within the host lava as implied by alignment of surrounding matrix crystals. In other cases, such aggregates are obscured amid abundant isolated microphenocrysts of similar size. Both isolated and aggregated microphenocrysts may be normally zoned, reversely zoned and unzoned. Lava issued sporadically from Pu'u'O'o flank vents during early stages of episode 55 contains variably reacted phenocrysts of clinopyroxene and olivine with complex zoning features, distinct reaction rims, and plagioclase inclusions. Figure 7 shows an early episode 55 xenolithic aggregate composed of resorbed and complexly zoned olivine and a sector zoned clinopyroxene surrounded by a microphyric matrix that is itself resorbed and overgrown by relatively Mg-rich clinopyroxene. Additional examples of variably reacted olivine phenocrysts in early episode 55 lava have been presented by Thornber (2001, fig. 4c–g).
|
MINERAL AND GLASS CHEMISTRY |
|---|
Olivine, plagioclase and pyroxene phenocrysts and microphenocrysts in individual lava samples convey a record of crystal growth and reaction in a range of compositionally distinct magmas. Olivine–liquid relations of near-vent, steady-state lava demonstrate the prevalence of Fo80–82 saturated magma at
1200°C, within the active conduit (Thornber, 2001). Episode 54 and early episode 55 samples exhibit crystal–liquid reactions indicative of pre-eruptive magma mixing between conduit magma and cooler, phenocryst-laden magma. The heterogeneity among glass compositions in episode 54 samples also provides strong evidence for pre-eruptive mixing of discrete magmatic components. Compositional and textural features suggestive of magma mixing for each of the major crystalline phases and discrete glasses within episode 54 samples are presented in turn.
Olivine
Deviations from equilibrium distribution of FeO and MgO between eruption glasses and olivine in 1994–1999 eruption products were used by Thornber (2001) to define three types of olivine phenocrysts as: Type A (near-equilibrium Fo/host-glass), Type B (high Fo/host-glass) and Type C (low Fo/host-glass). In the episode 54 sample suite, both the most differentiated lava of fissure F and intermediate lava of fissures A–E contain a range of Type B and Type A olivine crystals (Table 4, Fig. 8a). The Type B olivine is compatible with glasses or bulk liquids erupted during steady-state eruptive intervals (Thornber, 2001) and episode 54 Type A compositions reflect crystallization from their host-liquids. Close inspection of microcrystals within the matrix of these samples reveals tiny remnants of resorbed Type B olivine phenocrysts persisting among euhedral and skeletal Type A microphenocrysts. Type B olivine is more apparent and larger (up to 1 mm) in fissure F samples (Fig. 6a) and less prominent and smaller (up to 0·3 mm) in fissures A–E (Fig. 6b). The Type B crystals exhibit variably resorbed outer margins, normal core-to-rim zonation and exterior overgrowth of either Type A skeletal olivine or clinopyroxene.
|
|
Two Type B olivine grains among a sparse population of resorbed olivine crystals observed in fissure A–E samples were analyzed. At Fo84·1, a remnant Type B olivine core in fissure B sample 1850 (Table 4) is a compositional match for Type B olivine in steady-state lava (Thornber, 2001). This crystal displays continuous normal zoning into a skeletal overgrowth of Fo79·8 and is locally associated with free-floating microlites of chromian spinel, a primary crystalline product of a relatively primitive magma. An Fo80·6 Type B olivine in fissure E, sample 1860 (Table 4, Fig. 6b) is an angular phenocrystic fragment (0·25 mm) with slightly rounded edges and a thin (<20 µm), Fo74 rim. This olivine core composition overlaps the range of near-equilibrium olivine in Pele's tears sampled during steady-state eruption at Pu'u'O'o (Thornber, 2001). Type A olivine microphenocrysts in fissure A–E lava range in composition from Fo79·8 to Fo74·6 and occur as normally zoned or unzoned crystals that are either isolated in the matrix or intergrown with plagioclase at near-cotectic temperatures (
1150°C).
Six Type B olivine grains analyzed in fissure F samples have core/rim compositional ranges from Fo79·2/Fo75·1 to Fo84·4/Fo77·7 and coexist with Type A microphenocrysts ranging from Fo76 to Fo72 (Table 4). Microphenocrysts of lower-Fo Type B olivine in fissure F overlap the composition of Type A (near-equilibrium) microphenocrysts in fissures A–E. Figure 6a shows a normally zoned and well-rounded olivine phenocryst from fissure F (Fo84·4 core, Fo77·7 rim) with glass inclusions (see discussion of glasses below). Small rounded Type A microphenocrysts (0·05 mm) of Fo78 composition surrounding this crystal are intergrown with plagioclase and overgrown by Fo73 skeletal olivine. The fine-scale cuspate texture at the crystal–liquid interface of the larger olivine phenocryst is evidence for partial resorption by the surrounding low-temperature liquid. Most other Type B olivine crystals in episode 54 lava have similar skeletal olivine overgrowth of angular or rounded cores with cuspate margins (see Thornber, 2001, fig. 4b). Comparable examples of cuspate olivine texture, observed in hybrid magmas erupted in the Kilauea 1959 summit eruption and in the 1955 lower east rift eruption, are interpreted as evidence for mixing of thermally and chemically disparate magmas (Thornber & Huebner, 1985; Helz, 1987; Helz & Wright, 1992).
Early episode 55 olivine phenocrysts have been discussed in detail by Thornber (2001). These include an array of normally zoned Type B compositions and Type C phenocrysts with reversed to complex zoning patterns that are unique to this post-episode 54 interval. Type C olivine within the lithic fragment in sample 1894 (Fig. 7) has an Fo77·6 core that is normally zoned to a rim of Fo73·9. An overgrowth of Type A Fo78·9 olivine surrounds the entire microphyric fragment. A second Type C olivine microphenocryst (Fo73·3), which is intergrown with plagioclase, is attached to this fragment (lower left, Fig. 7) and is reversely zoned toward the Fo78·9 rim composition. The low-Fo compositions of cores and inner rims of these complexly zoned crystals are similar to near-equilibrium olivine in episode 54 lava. Higher-Fo cores and outer rims record a history of post episode 54 magma mixing and subsequent incorporation in early episode 55 conduit magma.
Plagioclase
Variations in core composition of plagioclase phenocrysts (An63·1–76·0) and microphenocrysts (An60·7–67·3) reflect crystallization from a range of thermally and compositionally distinct magmas (Tables 4 and 5; Fig. 8b). Both phenocrysts (0·3–3 mm) and microphenocrysts (0·1–0·3 mm) occur as isolated crystals and in aggregates with clinopyroxene. The variability in the extent of crystal reaction features, including edge resorption, internal melting and core-to-rim zonation, imply differences in compatibility among different crystals and different liquids and in post-mixing residence time before eruption.
The variety of plagioclase compositions and textural relations in samples from fissures A–E record a diverse range of crystallization and reaction histories. Figure 6c shows a melted and reacted plagioclase phenocryst in fissure E sample 1860 (Table 4). This phenocryst remnant is optically continuous but exhibits complex chemical zonation from inner to outer core of An76 to An80·5 and from inner to outer rim of An78·7 to An66·1. Glass inclusions are pervasive throughout this grain and provide evidence of internal melting. Phenocrystic and microphenocrystic plagioclase with similar complex zonation patterns have been reported by Garcia et al. (2000) in fissure D samples and are also plotted in Figure 8b. These are interpreted here as metastable remnants of relatively low-An phenocrysts that were sequentially heated and cooled during a short interval of pre-eruptive mixing of hotter and cooler magmas. The high-An, reversely zoned core compositions have persisted as the refractory residua of partial melting during incorporation of an originally lower-temperature plagioclase in a relatively high-temperature magma. The normally zoned rim compositions of this crystal reflect subsequent reaction and overgrowth in lower-temperature hybrid magma. In contrast to these reacted phenocrysts, an isolated An65·6 phenocryst in fissure B lava sample 1850 (Table 5) has only slightly rounded outer margins and no significant core-to-rim zonation. Another An65·5 phenocryst with unreacted morphology and uniform composition is intergrown with clinopyroxene in a fissure E phenocryst aggregate in sample 1859 (Table 5).
Glomerophyric plagioclase microphenocrysts in lava from fissures A–E have cores ranging from An61·9 to An67·3. In fissure A sample 1849, a glomerophyric plagioclase has distinct normal core-to-rim zonation from An67·3 to An61·9, amidst clinopyroxene overgrowth in an aggregate that also includes intercrystalline glass.
The compositional range of analyzed plagioclase phenocryst and microphenocryst core compositions in fissure F lava is similar to that in fissures A–E (Fig. 8b), with the exception of the high-An refractory compositions of resorbed phenocrysts discussed above. Plagioclase intergrown with large clinopyroxene in fissure F samples 1862 and 1867 (Fig. 6d and e) have An63·6 and An64·8 cores, respectively (Table 5), and do not show appreciable evidence of reaction. Core composition among analyzed plagioclase microphenocrysts in fissure F ranges from An64·8 to An65·7 for free-floating crystals in sample 1863 (Table 5, Fig. 6h) and from An60·7 to An66·9 in glomerophyric aggregates in samples 1861 and 1862 (Table 5).
No large plagioclase phenocrysts were found in early episode 55 lava. However, the presence of plagioclase microlites that are included in Type C olivine crystals indicates a pre-eruptive history involving short-term mechanical mixing of hotter (olivine-only) and cooler (plagioclase-bearing) magmas. Although present in early episode 55 lava, this phenomenon is not observed in steady-state eruption samples. A case for inclusion of episode 54 related plagioclase microphenocrysts in early episode 55 lava is apparent in sample 1894 (Fig. 7). Plagioclase in the lithic aggregate in this sample ranges in composition from An59·1 to An67·3 from the inner to outer portion of the aphyric fragment, with rim compositions overlapping those of microphyric plagioclase included by olivine, and those of free-floating plagioclase microphenocrysts and microlites in other early 55 glasses (Fig. 8b).
Pyroxene
Phenocrysts and microphenocrysts of augite occur as isolated crystals and in aggregates with plagioclase, and cores have a wide compositional range with reaction features that include normal, reverse and complex zoning (Fig. 8c), along with textural evidence for resorption and partial melting. Details of clinopyroxene petrology, along with those of a particularly significant orthopyroxene phenocryst, place further constraints on the nature and origin of magmas that were mixed to form hybrid lava of episode 54. Also, the presence of reacted pyroxene phenocrysts in early episode 55 lava provides additional evidence for the persistence of hybrid magma during the post-episode 54 interval of non-steady-state eruption at Pu'u'O'o.
Clinopyroxene microphenocryst cores in fissure A–E samples range in Mg# from 79·9 to 76·3, overlapping the 77·7–73·5 Mg# range of matrix microlites in these samples (Fig. 8c and Table 5). Two large clinopyroxene phenocrysts observed in fissure E samples 1859 and 1860 have distinctly different character. The isolated phenocryst in 1860 (Fig. 6f) has a core of Mg#68·2 and is extensively resorbed and reversely zoned, with a distinct rim of Mg#78·7 (Table 4). This phenocryst also has irregular-shaped glass inclusions with cuspate texture along host–inclusion boundaries. These inclusions are interpreted as the products of intracrystalline melting under conditions analogous to that of the melted plagioclase phenocryst observed in this same thin section. In contrast, the 1859 phenocryst is subhedral and intergrown with plagioclase, has a core composition of Mg#79·9, and is normally zoned to an Mg#78·0 rim (Table 5).
Analyzed clinopyroxene microphenocrysts from fissure F samples have cores of Mg#80·2–75·3. A free-floating microphenocryst in sample 1863 is normally zoned from Mg#80·2 to Mg#76·3 (Table 5). Two fissure F clinopyroxene phenocrysts in samples 1862 and 1867 are intergrown with plagioclase (Fig. 6d and e, respectively). A slight reversed zonation of the 1867 phenocryst is indicated by core to rim variation of Mg#79·0–79·4 (Table 5) and the euhedral outline of this crystal is only slightly rounded, indicating a near-equilibrium relation with the host melt. The 1862 phenocryst is sector zoned, rounded and surrounded by a thin homogeneous rim of clinopyroxene. Microprobe traverses from core to rim of hour-glass sectors that appear dark and light in electron backscatter image reveal opposite zonation patterns. The core-to-rim dark and light sector zoning ranges from Mg#79·1 to Mg#76·2 and from Mg#75·4 to Mg#77·5, respectively (Table 5). Both sectors are rimmed by clinopyroxene of Mg#77, an apparent product of reaction between the phenocryst and the host melt, indicating that development of sector zoning occurred before inclusion and reaction with fissure F hybrid magma. As discussed in the following section, an average Mg#77·3 for light and dark sector cores in this crystal provides a reconstructed composition that is compatible with the associated An63 plagioclase phenocryst. This clinopyroxene also contains orthopyroxene (Mg#78·7, Table 4) at the intersection of the light sector core and the included plagioclase. The relatively Mg-rich orthopyroxene is lamellar-like and may have formed because of exsolution associated with sector zoning in the clinopyroxene host. It is conceivable that sector zoning occurred during heating and partial resorption and that Mg-rich orthopyroxene was produced in the buffered melt reaction between a hotter olivine-melt and clinopyroxene + plagioclase (Yang et al., 1999). Similar orthopyroxene lamella development in complexly zoned augite was described in hybrid lava of the 1960 east rift zone eruption by Wright & Helz (1996, fig. 4) and is suggestive of a complex thermal history before inclusion in host lavas. In contrast, Mg-rich orthopyroxene in Kilauea east rift lava reported by Clague et al. (1995) is rounded and often encloses rounded olivine. These are interpreted based upon experimental phase relations to have crystallized from primitive magmas at elevated pressures between 2 and 0·45 GPa (Eggins, 1992).
A single free-floating orthopyroxene phenocryst was identified in fissure F lava (sample 1863, Fig. 6g). This slightly rounded phenocryst is normally zoned from Mg#73·5 to Mg#78·0 (Table 4) and is partially mantled by host-compatible clinopyroxene microphenocrysts with Mg#77·4 (Table 5). Orthopyroxene phenocrysts of similar composition, although rare at Kilauea, were reported by Anderson & Wright (1972) as compatible phases within examples of the most chemically fractionated Kilauea lava, including that produced during early stages of the 1955 east rift zone eruption [also described by Wright & Helz (1996)]. Orthopyroxene of similar composition and appearance to that in sample 1863 occurs in hybrid 1960 eruption products. These were described as metastable remnants of a more fractionated magma by Fodor & Moore (1994) and Wright & Helz (1996). As shown in experiments using Kilauean basalt by Thompson & Tilley (1969), this relatively iron-rich orthopyroxene composition is restricted to low-pressure (<0·45 GPa) crystallization from differentiated melts and appears on the low-pressure liquidus at temperatures below 1140°C from a liquid at 5 wt % MgO.
No large phenocrysts of pyroxene are observed in early episode 55 lava. Consistent with complexities of olivine and plagioclase variations in the KE55-1894 lithic fragment, the small clinopyroxene phenocryst (0·3 mm) in this cluster (Fig. 7) displays moderate sector zoning with slightly reversed zonation from inner to outer core (Mg#81·9 to Mg#83·2) and is normally zoned to an outer rim of Mg#77·0. The part of the grain in contact with the host melt is uniformly mantled by an Mg#77·5 clinopyroxene overgrowth that is intermediate in composition between the phenocryst and the Mg#78·0–80·0 clinopyroxene microlites analyzed in the matrix of this sample (Fig. 8c). As suggested by Lofgren (1980), sector zoning in clinopyroxene may develop because of rapid growth during undercooling. The Mg# within the sector-zoned core of this crystal is anomalously high compared with the range of episode 54 clinopyroxene phenocrysts, and may be explained by such metastable growth.
Glass
Glass heterogeneities in episode 54 samples attest to thermal and chemical disparities associated with magma mixing before eruption (Table 6). Among the samples analyzed, most have abundant areas of ‘clean’ matrix glass (i.e. devoid of quench crystals). Exceptions are the 1858 (fissure D), 1852 (fissure B) and 1860 (fissure E) samples, which have relatively evolved glass compositions owing to extensive development of microlitic to spherulitic groundmass during inefficient quenching and non-equilibrium crystal growth in these particular globs of fountain spatter. Such poorly quenched glasses are not useful for application of glass thermometry to assessment of pre-eruptive magmatic temperatures, but they provide analog fractionated compositions for lower-temperature liquids used to calibrate multiphase mineral thermometers.
|
Average MgO contents of ‘clean’ matrix glasses in fissure A, B and E samples range from 6·02 to 6·16 wt %, corresponding to pre-eruption temperatures of 1135–1138°C, as inferred using the glass thermometer of Helz & Thornber (1987). The average glass composition in fissure F samples ranges from 4·9 to 5·5 wt % MgO and defines a lower magmatic temperature range of 1112–1124°C (Table 6). Eruption temperatures for these temporally and spatially discrete fissures are distinctly lower than the 1146–1168°C range of steady-state eruption products. An average 10°C difference in glass temperatures between episode 54 sample groups is 7°C less than the difference of magmatic temperatures that might be inferred from their bulk-rock wt % MgO averages (see Thornber, 2001, appendix 1), thus implying a difference in mixing proportions of hotter and cooler magmas erupted from fissures A–E and fissure F.
The overall CaO vs MgO variation of matrix glasses in episode 54 samples reflects liquid depletion in clinopyroxene and plagioclase from the near-cotectic compositional range of steady-state Pele's tear glasses (Fig. 9a). The ratio of K/Ti vs K in matrix glasses overlaps that of bulk lava in fissure A–E, fissure F and steady-state eruption samples (Fig. 9b). This comparison affirms the conclusion that episode 54 bulk lava compositions are not derived from fractionation of recent rift conduit magma but rather represent two chemically distinct hybrid melt groups.
|
The coexistence of relatively evolved, lower-temperature glomerophyric glasses (1107–1123°C, Table 7 and Fig. 9) with less fractionated matrix glasses, representing the hybrid host-liquid in fissure A–E samples, provides direct evidence for late-stage mixing between discrete magmas. The K/Ti concentrations in these glomerophyric inclusions (Fig. 9b) reflect a range of mixed source characteristics. As further elaborated in our discussion of phenocryst–liquid relations, microphenocryst core compositions in these aggregates are compatible with their included glasses.
|
Intracrystalline glass inclusions in episode 54 olivine phenocrysts provide conclusive evidence for crystal growth in magmas of chemically disparate affinities. The included glass in Type B olivine (Fo84·4) of fissure F sample 1864 (Fig. 6a, Table 7) has the distinct signature of olivine-incompatible elements in olivine inclusions, matrix glasses and bulk lava of steady-state eruption intervals (Fig. 9). This and other high-Fo olivine in hybrid episode 54 lava are thus considered to have originated from relatively primitive magma associated with the long-term eruption. In contrast, the inclusions in the core and rim of lower-Fo olivine microphenocrysts of fissure A (samples 1849 and 1850) indicate crystallization from magma with higher relative concentrations of incompatible elements.
Inclusions in the fissure E sample 1860 resorbed plagioclase (Fig. 6c) have a K/Ti vs K signature that is comparable with episode 54 glomerophyric aggregates, matrix glasses and bulk lavas. However, compositions up to 7·0 wt % MgO reveal anomalously high temperatures, suggesting that this relict crystal was partially melted while in contact with a high-temperature magma. The 7·0 wt % MgO corresponds to the olivine–plagioclase cotectic composition that would be stable during plagioclase resorption by an olivine-saturated liquid. The CaO depletion of these glasses relative to normally fractionated liquids (Fig. 9) is affected by CaO retention in a refractory high-An solid residua product (An80) of plagioclase melting. The survival of this discrete high-temperature glass (1154°C) in host melts of much lower temperature attests to a short time interval between magma mixing and eruption of hybrid magmas.
|
MINERAL THERMOMETRY DEDUCED FROM PHENOCRYST–LIQUID RELATIONS |
|---|
Equilibrium crystal–liquid distribution coefficients of ferrous iron and magnesium for olivine–basalt (KDol/liq= 0·300, Roeder, 1974), clinopyroxene–basalt (KDcpx/liq= 0·230, Grove & Bryan, 1983) and orthopyroxene–basalt (KDopx/liq = 0·284, Beatie, 1993) are used in conjunction with Kilauea MgO glass thermometry (Helz & Thornber, 1987) to deduce crystallization temperatures of phenocrysts in episode 54 lava. The variation of liquid Mg number [Mg#liq = 100 x mol % MgOliq/(MgOliq + FeOliq)] with temperatures calculated for glasses and bulk liquids of the 1994–1999 eruptive interval is shown in Fig. 10. In accordance with known low-pressure phase relations for Kilauean basalt (Thompson & Tilley, 1969; Helz & Thornber, 1987), these variations of Mg#liq with cooling present a fractionation path for Kilauea magma primarily affected by olivine and clinopyroxene crystallization throughout the major crystallization interval from
1200° to 1100°C. At temperatures above 1150°C, calculated temperatures of bulk liquids erupted during episodes 53 and 55 (198 samples, Thornber, 2001) correspond to a linear olivine-controlled Mg#liq variation, which is defined as
 | (1) |
|
Below
1150°C, the Mg#liq descent with temperature is defined by glass compositions controlled by clinopyroxene and minor olivine fractionation as
 | (2) |
The lower-temperature glass data include 168 individual analyses of matrix glasses in episode 54 and early episode 55 samples and are extrapolated through the range of hybrid bulk-rock compositions for each of these samples and the average of all episode 53 steady-state tear glasses (900 analyses of 90 samples). The tear glass composition represents a steady-state liquid maintained at 1153°C (1 = 3) near the multiphase cotectic at the low end of the olivine control trend and in equilibrium with Fo80·6 olivine.
Mineral thermometers for Kilauea olivine, clinopyroxene and orthopyroxene are based upon equilib rium Mg#xl/liq KD values for each of these phases and Mg#liq/T(°C)liq relations for recent Kilauea liquids [equations (1) and (2)]. These mineral thermometers for use in determining low-pressure crystallization temperatures in Kilauea magma are defined as follows:
high-Fo olivine thermometer (>Fo80):
 | (3) |
 | (4) |
 | (5) |
 | (6) |
The correlation of percent anorthite component in plagioclase (% Anplg) with MgO liquid thermometry shown in Fig. 11 is extrapolated from mineral core analyses for 25 pairs of coexisting plagioclase–clinopyroxene crystals in this sample suite (Table 5) and using temperature values determined with the clinopyroxene thermometer [equation (5)]. The resulting linear correlation between plagioclase composition and temperature (r2 = 0·83) defines a loosely constrained empirical thermometer for plagioclase in Kilauea basalt (with low H2O content) as
|
plagioclase thermometer:
 | (7) |
Temperatures calculated using these mineral thermometers [equations (3)– (7)] are inferred to represent that of a liquid that is close to equilibrium with phenocrysts and microphenocrysts. The calculated crystal–liquid equilibrium temperatures for phenocrysts, using Kilauea mineral thermometers, are presented with corresponding mineral analyses in Tables 4 and 5. Independent validation of temperatures estimated using these Mg–Fe equilibration schemes is provided by comparison of mineral- and glass-thermometry values for adjacent crystal–glass phases in two glomerophyric aggregates (samples 1849 and 1854) and for glass-inclusion/host values of the 1860 clinopyroxene (Fig. 6h and f). In each of these cases, temperatures of the included glass are closely approximated by temperatures calculated from adjacent minerals (Table 7) and are distinctly lower than that of surrounding matrix glass.
|
PHENOCRYST DEFINED MAGMATIC COMPONENTS IN EPISODE 54 HYBRID LAVA |
|---|
The mineral chemistry of episode 54 phenocryst assemblages and the nature of phenocryst–matrix reactions provide clear indications of the hybrid origin of these rocks. With the exception of orthopyroxene in fissure F, the diversity of phenocryst chemistry is comparable for both fissure F and fissure A–E sample groups. Using Kilauea mineral thermometers to determine the range of source magma temperatures for the episode 54 phenocryst assemblages, we can ascribe a range of variably fractionated magma compositions as likely end-members for mixing. The range of temperatures calculated for episode 54 phenocrysts, microphenocrysts and glass inclusions is similar in both fissure A–E and fissure F datasets (Fig. 12) and indicates crystallization from a similar range of variably cooled and fractionated magma. The assemblage of relatively high-, intermediate- and low-temperature phenocrysts and glass inclusions in hybrid episode 54 lava suggests a magma mixing process involving at least three shallow magmatic components.
|
The pyroxene–plagioclase-saturated hybrid component
Phenocrysts that define the most fractionated end-member involved in episode 54 mixing are those with calculated core temperatures clustering near 1120°C (Fig. 12). Among this group of crystals, orthopyroxene (Mg#73·5), clinopyroxene (Mg#78·2–68·2) and plagioclase (
An64·0–60·5) phenocrysts have reversely zoned or unzoned rims (Fig. 8). A distinct population of isolated and glomerophyric plagioclase and clinopyroxene, along with associated interstitial glass inclusions, also range in temperature from 1130°C down to 1105°C. This temperature range corresponds to an initial magma with 5·75 wt % MgO and crystallizing fractions of cooler liquid, down to 4·5 wt % MgO. The presence of these low-temperature phenocrysts and microphenocrysts in lava from fissures A–E and fissure F indicates mixture of similar rift-stored magma components in both episode 54 hybrid magmas.
A temperature of 1116°C for the sector-zoned clinopyroxene phenocryst that is intergrown with plagioclase in fissure F sample 1862 (Fig. 6d) is obtained using an average Mg#77·3 of light and dark core sectors. This value corresponds well to the 1118°C temperature calculated for the coexisting An63·3 plagioclase, indicating that reconstituting an average core composition may be a viable procedure for obtaining initial crystallization temperature for sector-zoned clinopyroxene. A higher temperature of 1138°C calculated for the small orthopyoxene lath in this aggregate (Mg#82) corresponds to the 1140°C four-phase saturation temperature at which olivine reacts with melt to form orthopyroxene (Kinzler & Grove, 1992). This is consistent with orthopyroxene formation as a by-product of sector zoning of the clinopyroxene host during exposure to a higher-temperature, olivine-saturated melt. The preponderance of hybrid melt compositions corresponding to this temperature, both during episode 54 and from the same locations during episode 1, is suggestive of chemical buffering resulting from the mixing of hotter, olivine-saturated melt and a low-temperature crystal–liquid assemblage as hypothesized by Yang et al. (1999).
The near-cotectic hybrid component
The group of isolated and aggregated microphenocrysts of plagioclase (An65–67) and clinopyroxene (Mg#80·7–83·2) with crystal-core temperatures ranging from 1130°C to 1140°C, along with lower-Mg olivine crystals (Fo78·4–80) of 1140–1150°C melt affinity, all display varying degrees of normal zonation (± resorption) (Figs 8 and 13). This assemblage is derived from shallow magma, ranging in composition from 6·8 to 5·8 wt % MgO near the onset of multiphase precipitation of olivine + plagioclase + clinopyroxene. The episode 54 olivine crystals in this group overlap the range of equilibrium olivine in episode 53 tears, erupted at temperatures corresponding to the low-MgO end of olivine-only fractionation. At an average of 6·9 (1 = 0·2), episode 53 tear glasses define steady-state magmatic conditions that persist at near-cotectic conditions during actively replenished shallow storage. The presence of similar olivine and high-temperature clinopyroxene and plagioclase microphenocrysts attests to the possibility that this steady-state cotectic component is an ingredient of the hybrid mixtures.
|
The olivine-saturated hybrid component
Olivine microphenocrysts with Fo
84 core compositions display normal zonation and resorption textures demonstrating that these crystals are incompatible in cooler, episode 54 hybrid melts. These crystals are analogous to Type B olivine that crystallizes at 1200°C in equilibrium with summit-derived steady-state lava, before rift transport (Thornber, 2001). As substantiated by glass-inclusion chemistry (discussed above), such episode 54 olivine crystals are interpreted as cognate to this eruption and are testimony to the mixture of a summit-derived magma component in the hybrid mixes. In comparison with an average Type B olivine phenocryst size of 0·6–0·7 cm observed in steady-state eruption products (Thornber, 2001), the largest of these episode 54 remnant crystals is 0·3 mm. Assuming approximate dissolution rates of olivine in undercooled basalt to be of the order of 5–10 µm/h (Thornber & Huebner, 1985), the observed difference of less than 50% original size among surviving olivine crystals is consistent with mixing and resorption during the 38 h interval between the onset of rift zone extension and the end of the Napau eruption.
The presence of high-temperature, refractory plagioclase remnants (up to An80·5) further attests to pre-eruptive mixing of an 1200°C summit-magma component. At 7·0 wt % MgO and 1154°C, glass associated with melted plagioclase phenocrysts preserves a record of near-cotectic liquid production during plagioclase dissolution by higher-temperature, olivine-only, melts.
|
GEOCHEMICAL MODELS FOR THE ORIGINS OF EPISODE 54 MIXED MAGMAS |
|---|
Phenocrysts and glass inclusions in episode 54 lava imply the pre-eruptive mixture of at least three shallow magmatic components with high-, intermediate- and low-temperature affinities. Both geophysical and petrologic evidence presented herein indicates that the hotter and intermediate olivine-saturated components of mixing were derived from magma associated with prolonged steady-state eruption before this event. The most evolved mixing components in both fissure F and fissure A–E lava are indicated by low Mg# ortho- or clinopyroxene phenocrysts to be liquids of
5·0 wt % MgO. These are compositionally analogous to the most fractionated lava known to have erupted along the lower Kilauea east rift zone during the early phases of the 1955 eruption (Wright & Fiske, 1971) and which contains unreacted orthopyroxene of similar composition (Anderson & Wright, 1972). Likely candidates for the low-temperature, rift-stored components in fissures A–E and F hybrids are the fractionated equivalents of magma last erupted from overlapping Napau Crater fissures in 1983, 1968 and 1963. New major and trace element data were obtained for four such lava samples from episode 1 of the current eruption in January 1983 (Table 2). These were collected from fissures that overlap or are en echelon with episode 54 fissures (samples KE1-49, -50, -51, -52; Fig. 2). Sample KE1-49 is from the initial extrusion of episode 1 lava from a fissure in the NW corner of Napau Crater. At 6·8 wt % MgO, the major element composition of this lava is similar to the tear glasses and the 7 wt % MgO-normalized compositions of episode 53. Samples KE1-50 to -52 are from the east Napau Crater area of episode 54 fissures D and E. These episode 1 samples have major element compositions that are indistinguishable from the episode 54 fissure A–E products, averaging 6·4 wt % MgO, and were also interpreted as having mixed-magma affinities by Garcia et al. (1989). Before 1983, the most recent magma erupted at a west Napau Crater location was in 1963. This magma is represented by sample 63-3 of Moore & Koyanagi (1969), and had an MgO content of 7·4 wt %. In east Napau Crater, the last pre-1983 magma erupted is that represented by sample 68-4 of Jackson et al. (1975) with an MgO concentration of 7·7 wt %. Both reports on the earlier eruptions provide major element compositional data but trace element analyses for these samples are not available.
An early 1955 eruption sample with 5·0 wt % MgO (TLW67-34, Wright & Fiske, 1971) provides an analog end-member composition as targeted by major element fractionation calculations, performed using the multiple linear regression routine (PETMIX) of Wright & Doherty (1970). Major element composition of the analog evolved component was well matched (i.e. low squared-residual sums for 10 unweighted major-oxide components) by fractionating appropriate phenocryst compositions (Tables 4 and 5) from each these prospective parent compositions. Results indicate that the west Napua magma could be derived from either a 1983 or 1963 parent (KE1-49 or 63-3) with 36–40% fractionation of plagioclase and clinopyroxene (with minor olivine). Major element compositions of magma trapped beneath east Napau Crater are well matched by 28% of fractionation of a similar phenocryst assemblage from a KE1-50 to -52 parent and 37% fractionation of the 1968 parent (68-4).
Major element bulk-mixing models using analog closed-system and known open-system reservoir components to produce episode 54 hybrid magmas yielded good matches with low residuals (Table 8b). The mass-balanced proportions reproducing the fissure A–E average composition are: 32% rift-stored component (TLW67-34 analog), 67% cotectic component (7% MgO-normalized episode 53 steady-state average composition, E53ss7%NRM, Table 1) and <0·4% summit–conduit component (episode 55 steady-state average composition, E55ss, Table 1). The average fissure F composition is well matched by a mixture of these respective components in proportions of 65%, 35% and <1%. Binary mixing curves utilizing Ca/Ti variation with Mg (Langmuir et al., 1978) demonstrate the feasibility of mixing paths between the analog rift-stored component and magma in the active conduit (Fig. 13). Both hybrid episode 54 magmas are slightly Mg rich relative to a binary mixing line between the highly evolved, closed-system magma and a near-cotectic, open-system melt, as a result of the minor contribution of a hotter summit–conduit component to the mix.
As demonstrated in a plot of La/Yb vs K2O (Fig. 14), the incompatible element signatures of west and east Napau 1983 magmatic fractionates preclude their parental association with the low-temperature components of episode 54 fissure F or fissure A–E hybrids. These fractionation products were calculated assuming Rayleigh fractionation (e.g. equations of Allègre & Minster, 1978), using partition coefficients of Pietruszka & Garcia (1999) and assuming fractionated mineral proportions as defined by PETMIX major element calculations. Based upon a general comparison of incompatible-element ratios of unfractionated lava compositions in a spatially and temporally diverse sample suite, Garcia et al. (2000) came to a similar conclusion regarding the lack of genetic association between earlier products of this eruption and those of episode 54.
|
La/Yb values of 1961 and 1968 summit eruptions (from Pietruszka & Garcia, 1999) were used in lieu of such data on 1963 and 1968 Napau lava for assessing fractionation mixing and relations. This is considered a reasonable assumption based upon strong evidence for the steady decline in La/Yb ratios throughout this time span for summit and rift zone eruption products (Pietruszka & Garcia, 1999; Garcia et al., 2000; Thornber, 2003). As shown in Fig. 14, the episode 54 hybrid magmas lie along binary mixing lines between near open-system near-cotectic magma and the fractionated residue of spatially correlated magmas erupted in 1963 and 1968. The relative end-member proportions of these hybrid mixtures are consistent with those estimated from major element mixing calculations.
Hybrid lava produced during early stages of episode 55 reflects a continuous progression from lower to higher Mg compositions until steady-state eruptive conditions were established. This trend is quantified by major element mixing calculations for samples spaced throughout this April–July 1997 interval, which show progressive binary mixing of near-cotectic rift zone magma with a gradually dominant influx of hotter, summit-derived conduit magma (Fig. 13). Trace element modeling of the sequence of early episode 55 hybrid melts was also accomplished using mass balances between proportions of end-members defined by major element differences. Near-perfect matches for the major element binary-mixing model were obtained. This time-integrated trend of gradually increasing proportions of summit–conduit magma to near-cotectic, open-system reservoir magma reflects the flushing of cooler magma residing uprift of Pu'u'O'o as the fluid-pressure balance within the shallow summit–rift zone plumbing system became re-established.
|
INTEGRATION OF GEOPHYSICAL CONSTRAINTS AND PETROLOGIC MODELS |
|---|
Our petrologic interpretation of magmatic components involved in the production of episode 54 hybrid lava is further substantiated by geophysical evidence. Owen et al. (2000) hypothesized that the sudden shallow rift opening beneath Napau Crater resulted from extensional stress related to long-term movement of Kilauea's south flank. We suggest that this zone was prone to extensional release owing to inherent weakness associated with the presence of small and isolated shallow pockets of crystallizing 1963 and 1968 magma above the active rift zone conduit. Petrologic data implying that such rift-stored magmatic components were mixed with magma persisting within and around the conduit are supported by geodetic data indicating that disruption of the pressurized summit–rift zone conduit diverted magma from both uprift and downrift areas into an extensional gap beneath Napau Crater. Shallow extension associated with the event encompassed a volume of 23 x 106 m3 (Owen et al., 2000). This volume is matched by the cumulative volume of magma displaced from within the shallow magmatic plumbing system. Displacement of 1·5 x 106 and 1·2 x 106 m3 of magma are inferred from modeled sources beneath the summit caldera and Makaopuhi Crater, respectively (Owen et al., 2000). Our measurements of pre- and post-collapse dimensions of the Pu'u'O'o cone and crater indicate that 12·7 x 106 m3 of magma was diverted from beneath the vent area. [By comparison, the size of the Pu'u O'o reservoir from 1983 to 1986 was estimated as (10–12) x 106 m3 by Hoffman et al. (1990) and 18 x 106 m3 by Garcia et al. (1992).] Including our estimate of 300 000 m3 of lava erupted during episode 54, the sum of displaced magma volumes from the summit reservoir, Makaopuhi and Pu'u'O'o still leaves
7·3 x 106 m3 of rift expansion volume unaccounted for. We infer that the missing volume is the volume of the east rift zone conduit, which can be estimated by multiplying the time required to repressurize the conduit (the 24 day eruptive pause), by an average episode 53 flux rate of 300 000 m3/day (Thornber et al., 1997).
Of the total episode 54 extensional volume, 55% was accommodated by transfer of magma in an uprift direction from beneath Pu'u'O'o and 36% was filled by pre-existing conduit magma. Magma displaced in a downrift direction from beneath Makaopuhi Crater made up only 6% of the extensional volume. The Pu'u'O'o and Makaopuhi sources were probably persisting as open-system reservoirs within the active conduit at the time of the event. The consistent 7 wt % MgO composition of Pele's tears at Pu'u'O'o during episode 53 reflects the near-cotectic condition that is maintained by continuous replenishment with olivine-saturated magma along the conduit.
Only 1 vol. % of magma displaced into the zone of extension erupted as episode 54 hybrid lava. The erupted volume of 300 000 m3 represents a minimum for the total volume of magma involved in the pre-eruptive mixtures. Geochemical mixing models imply that the 270 000 m3 from fissures A–E comprised a mixture of 67% of an open-system reservoir (near-cotectic composition) and 32% closed-system evolved component. The <30 000 m3 of fissure F lava contained a mixture of 65% closed-system reservoir component and 35% near-cotectic, open-system reservoir component. These proportions translate to minimum volume estimates for the isolated closed-system magma pockets of 86 000 m3 beneath east Napau and 19 500 m3 beneath west Napau areas. Of the total volume of displaced magma, only a very small amount of near-cotectic magma is needed to accommodate this mixing scenario. This magmatic component is likely to be present as a density-stratified magma associated with open-system reservoirs or as a marginal facies located both uprift (Makaopuhi) and downrift (Pu'u'O'o) of Napau Crater.
Olivine phenocryst reaction characteristics are compatible with mixing that occurred within the 38 h duration of this event. The relative timing and volumes of magma displacement from downrift and uprift locations corresponds to the eruptive sequence and volumes of the two distinct episode 54 lava compositions. The initial west-to-east propagation of fissures A–E probably occurred in response to greater initial extension in the magma-laden zone beneath east Napau Crater. This suddenly created gap was rapidly filled by a large volume of magma displaced from the downrift conduit and the Pu'u'O'o open-system reservoir. The last, low-volume gasp of episode 54 in the west wall of Napau Crater (fissure F) may have resulted from a delayed encounter between the relatively small volume of magma displaced from a Makaopuhi reservoir and that of an isolated pocket beneath the west side of the crater.
The steady rate of summit inflation after episode 54 is evidence that the influx of primitive magma into the summit magma chamber continued without interruption, despite a 24 day eruptive pause. Re-establishment of the fluid-pressure balance in the shallow summit–rift zone conduit was marked by recovery of summit deformation along with the reappearance of lava in the Pu'u'O'o crater. Petrology and geochemistry of lava erupted over the next 5 months indicates that cooler hybrid magma in the disrupted conduit, uprift of Pu'u'O'o, was gradually flushed out of the vent as steady-state eruptive conditions were re-established.
|
SUMMARY |
|---|
Episode 54 has provided the first conclusive evidence for pre-eruptive mixing of hotter and cooler rift zone magmas since the onset of this eruption in early 1983. Petrologic and geophysical evidence supports our conclusion that distinct magmatic components along the summit–rift zone conduit were passively intruded beneath Napau Crater and mixed immediately before the fissure eruption uprift of the long-lived Pu'u'O'o vent. These components include variably fractionated magmas that we interpret as persisting in both closed-system and open-system reservoirs associated with the long-term eruptive plumbing system.
During periods of steady-state rift zone eruption, an efficient pressure balance is maintained along the magma conduit between the summit and the vent, and erupted magmas are restricted to a small range of olivine-saturated bulk compositions. This stable condition is the end product of continuous input and turbulent mixing of primitive magma into the shallow summit reservoir (Thornber, 2003). As the summit is continually replenished, so it continually replenishes the magma moving through the shallow rift zone to the vent. The interaction of hotter, summit-derived magma with open-system reservoirs along the rift zone may account for a historic preponderance of rift zone lava compositions maintained at the low end of a Kilauean olivine-control trend. Pele's tears at Pu'u'O'o reflect persistence of this near-cotectic condition beneath the vent and support the likelihood that similar magma exists in open-system reservoirs along the active rift zone conduit, particularly beneath Makaopuhi Crater, which has long been known as a point source of deformation.
During episode 54, steady-state eruption conditions were disrupted by extensional release of lithostatic pressure along the active conduit between the summit and the vent. Geodetic, seismic and observational data indicate that the extensional gap beneath Napau Crater was filled with magma from both uprift (Makaopuhi) and downrift (Pu'u'O'o) segments of the active conduit. A 23 h eruption ensued from a sequence of fissures overlapping those of 1963, 1968 and 1983 rift zone eruptions. Phenocryst–liquid relations in episode 54 lava samples indicate that hotter and cooler magmas were mixed during the pre-eruptive interval of extension and magma displacement. Major and trace element mixing models for episode 54 hybrid lava imply the coexistence and mixture of evolved, closed-system fractionates and open-system reservoirs that are actively replenished by olivine-saturated magma. The chemical lineage of the rift-stored components indicates they are of discrete origin, resulting from closed-system fractionation of magma erupted from these same fissures in 1963 and 1968. Petrologic data imply that such rift-stored magmatic components were mixed with magma persisting within and around the conduit.
Within the context of eruption characteristics from 1994 to 1999, episode 54 has presented an opportunity to evaluate shallow magmatic processes that are likely to typify those associated with prolonged rift zone eruptions of Hawaiian shield-building volcanoes.
|
ACKNOWLEDGEMENTS |
|---|
This research on Kilauea eruption products is accomplished as part of a comprehensive and long-term volcano monitoring effort by the Geological Survey's Hawaiian Volcano Observatory and with the essential support of observatory staff and numerous student volunteers. We greatly appreciate the contribution of David Okita (Volcano Helitours) to our field efforts. Dave Seims (USGS, Denver) completed high-quality major element XRF analyses with incredible efficiency. This paper has benefited from formal technical reviews by Drs R. T. Helz, L. G. Mastin and W. E. Scott. The senior author is especially grateful to Drs Aaron Pietruszka, Michael Rhodes and Marc Norman, all experts on the petrology and geochemistry of Hawaiian volcanism, for providing very insightful and constructive reviews that significantly contributed to this paper.
|
REFERENCES |
|---|
Allègre, C. J. & Minster, J. F. (1978). Quantitative models of trace element behavior in magmatic processes. Earth and Planetary Science Letters 38, 1–25.[CrossRef][Web of Science]
Anderson, A. T. & Wright, T. L. (1972). Phenocrysts and glass inclusions and their bearing on oxidation and mixing of basaltic magmas, Kilauea Volcano, Hawaii. American Mineralogist 37, 188–216.
Baedecker, P. A. & McKown, D. M. (1987). Instrumental Neutron Activation Analysis of Geological Samples. US Geological Survey Bulletin 1770, 16 pp.
Beatie, P. (1993). Olivine–melt and orthopyroxene–melt equilibria. Contributions to Mineralogy and Petrology 115, 103–111.[CrossRef][Web of Science]
Cashman, K. V., Thornber, C. R. & Kauahikaua, J. P. (1999). Cooling and crystallization of lava in open channels, and the transition of pahoehoe lava to ’a’a. Bulletin of Volcanology 61, 306–323.[CrossRef][Web of Science]
Cervelli, P., Amelung, F., Owen, S., Meertens, F., Segall, P., Miklius, A. & Lisowski, M. (2002). The September 12, 1999 Upper East Rift Zone dike intrusion at Kilauea Volcano, Hawaii. Journal of Geophysical Research 107(B7), 10.1029/2001JB000602.
Clague, D. A., Moore, J. G., Dixon, J. E. & Friesen, W. B. (1995). Petrology of submarine lavas from Kilauea's Puna Ridge, Hawaii. Journal of Petrology 36, 299–349.
Denlinger, R. P. (1997) A dynamic balance between magma supply and eruption rate at Kilauea Volcano, Hawaii. Journal of Geophysical Research 102, 18091–18100.[CrossRef]
Eggins, S. M. (1992). Petrogenesis of Hawaiian tholeiites: 1 Phase equilibria constraints. Contributions to Mineralogy and Petrology 17, 85–115.
Fodor, R. V. & Moore, R. B. (1994). Petrology of gabbroic xenoliths in 1960 Kilauea basalt: crystalline remnants of prior (1955) magmatism. Bulletin of Volcanology 56, 62–74.[CrossRef][Web of Science]
Fodor, R.V., Rudek, E. A. & Bauer, G. R. (1993). Hawaiian magma–reservoir processes as inferred from the petrology of gabbro xenoliths in basalt, Kahoolawe Island. Bulletin of Volcanology 55, 204–218.[CrossRef][Web of Science]
Garcia, M. O., Ho, R. A., Rhodes, J. M. & Wolfe, E. W. (1989). Petrologic constraints on rift-zone processes: results from episode 1 of the Pu'u'O'o eruption of Kilauea Volcano, Hawaii. Bulletin of Volcanology 52, 81–96.[CrossRef][Web of Science]
Garcia, M. O., Rhodes, J. M., Wolfe, E.W., Ulrich, G.E. & Ho, R.A. (1992). Petrology of lavas from episodes 2–47 of the Puu Oo eruption of Kilauea Volcano, Hawaii: evaluation of magmatic processes. Bulletin of Volcanology 55, 1–16.[CrossRef][Web of Science]
Garcia, M. O., Rhodes, J. M., Trusdell, F. A. & Pietruszka, A. J. (1996). Petrology of lavas from the Puu Oo eruption of Kilauea Volcano: III. The Kupaianaha Episode (1986–1992). Bulletin of Volcanology 58, 359–379.[CrossRef][Web of Science]
Garcia, M. O., Pietruszka, A. J., Rhodes, J. M. & Swanson, K. (2000). Magmatic processes during prolonged Pu'u’O’o eruption of Kilauea Volcano, Hawaii. Journal of Petrology 41, 967–990.
Grove, T. L. & Bryan, W. B. (1983). Fractionation of pyroxene-phyric MORB at low pressure; an experimental study. Contributions to Mineralogy and Petrology 84, 293–309.[CrossRef][Web of Science]
Harris, A. J. L. & Thornber, C. R. (1999). Complex effusive events at Kilauea as documented by the GOES satellite and remote video cameras. Bulletin of Volcanology 61, 382–395.[CrossRef][Web of Science]
Harris, A. J. L., Keszthelyi, L., Flynn, L. P., Mouginis-Mark, P. J., Thornber, C., Kauahikaua, J., Sherrod, D., Trusdell, F., Sawyer, M. W. & Flament, P. (1997). Chronology of the episode 54 eruption at Kilauea Volcano, Hawaii, from GOES-9 satellite data. Geophysical Research Letters 24, 3281– 3284.[CrossRef][Web of Science]
Heliker, C. & Mattox, T. N. (2003). The first two decades of the Puu ’O’o–Kupaianaha eruption: chronology and selected bibliography. In: Heliker, C. C., Swanson, D. A. & Takahashi, T. J. (eds) The Pu'u'O'o–Kupaianaha Eruption of Kilauea Volcano, Hawaii, the First Twenty Years. US Geological Survey Professional Paper 1676, 1–27.
Heliker, C. C., Mangan, M. T., Mattox, T. N., Kauahikaua, J. P. & Helz, R. T. (1998). The character of long-term eruptions: inferences from episodes 50–53 of the Pu'u'O'o–Kupaianaha eruption of Kilauea Volcano. Bulletin of Volcanology 59, 381–393.[CrossRef][Web of Science]
Helz, R. T. (1987). Diverse olivine types in lava of the 1959 eruption of Kilauea volcano and their bearing on eruption dynamics. In: Decker, R. W., Wright, T. L. & Stauffer, P. H. (eds) Volcanism in Hawaii, Volume 1, US Geological Survey Professional Paper 1350, 691–722.
Helz, R. T. & Thornber, C. R. (1987). Geothermometry of Kilauea Iki lava lake, Hawaii. Bulletin of Volcanology 49, 651–668.[CrossRef]
Helz, R. T. & Wright, T. L. (1992). Differentiation and magma mixing on Kilauea's East Rift Zone, a further look at the eruptions of 1955 and 1960. Part I. The late 1955 lavas. Bulletin of Volcanology 54, 361–384.[CrossRef][Web of Science]
Hoffman, J. P., Ulrich, G. E. & Garcia, M. O. (1990). Horizontal ground deformation patterns and magma storage during the Pu'u'O'o eruption of Kilauea volcano, Hawaii: episodes 22–42. Bulletin of Volcanology 52, 522–531.[CrossRef][Web of Science]
Houghton, B. F. & Wilson, C. J. N. (1989). A vesicularity index for pyroclastic deposits. Bulletin of Volcanology 51, 451–462.[CrossRef][Web of Science]
Jackson, D. B., Swanson, D. A., Koyanagi, R. Y. & Wright, T. L. (1975). The August and October 1968 East Rift Zone Eruptions of Kilauea Volcano, Hawaii. US Geological Survey Professional Paper 890, 33 pp.
Kinzler, R. J. & Grove, T. L. (1992). Primary magmas of mid-ocean ridge basalts. 1. Experiments and methods. Journal of Geophysical Research 97, 6885–6907.
Langmiur, C. H., Vocke, R. D., Jr, Hanson, G. N. & Hart, S. R. (1978) A general mixing equation with applications to Icelandic basalts. Earth and Planetary Science Letters 37, 380–392.[CrossRef][Web of Science]
Lofgren, G. E. (1980). Experimental studies on the dynamic crystallization of silicate melts. In: Hargraves, R. B. (ed.) Physics of Magmatic Processes. Princeton, NJ: Princeton University Press, pp. 487–551.
Moore, J. G. & Koyanagi, R. Y. (1969). The October 1963 Eruption of Kilauea Volcano Hawaii. US Geological Survey Professional Paper 614-C,18 pp.
Okamura, A. T., Dvorak, J. J., Koyanagi, R. Y. & Tanagawa, W. J. L. (1988). Surface deformation associated with dike propagation. In: Wolfe, E. W. (ed.) The Puu Oo Eruption of Kilauea Volcano, Hawaii: Episodes 1 through 20, January 3, 1983, through June 8, 1984. US Geological Survey Professional Paper 1463, 165–182.
Owen, S., Segall, P., Lisowski, M., Miklius, A., Murray, M., Bevis, M. & Foster, J. (2000). January 30, 1997 eruptive event on Kilauea Volcano, Hawaii, as monitored by continuous GPS. Geophysical Research Letters 27, 2757–2760.[CrossRef][Web of Science]
Pietruszka, A. J. & Garcia, M. O. (1999). A rapid fluctuation in the mantle source and melting history of Kilauea Volcano inferred from the geochemistry of its historical summit lavas (1790–1982). Journal of Petrology 40, 1321–1342.[CrossRef][Web of Science]
Rhodes, J. M. (1995). The 1852 and 1868 Mauna Loa picrite eruptions: clues to parental magma compositions and the magmatic plumbing system. In: Rhodes, J. M. & Lockwood, J. P. (eds) Mauna Loa Revealed. Geophysical Monograph, American Geophysical Union 92, 241–262.
Ridley, W. I. & Lichte, F. E. M. (1998). Trace, and ultratrace element analysis by laser ablation ICP-MS, applications of microanalytical techniques to understanding mineralizing processes. In: McKibben, M. A., Shanks, Wayne, C., III & Ridley, W. I. (eds) Reviews in Economic Geology 7, 199–215.
Roeder, P. L. (1974). Activity of iron and olivine solubility in basaltic liquids. Earth and Planetary Science Letters 23, 397–410.[CrossRef][Web of Science]
Roeder, P. L., Thornber, C., Poustovetov, A. & Grant A. (2003). Morphology and composition of spinel in Pu'u’O’o Lava (1996–1998), Kilauea Volcano, Hawaii. Journal of Volcanology and Geothermal Research 123, 245–265.[CrossRef][Web of Science]
Taggart, J. E., Jr., Lindsay, J. R., Scott, B. A., Vivit, D. V., Bartel, A. J. & Stewart, K. C. (1987). Analysis of geological materials by wavelength-dispersive X-ray fluorescence spectrometry. In: Baedecker, P. H. (ed.) Methods for Geochemical Analysis. US Geological Survey Bulletin 1170, E1–E19.
Thompson, R. N. & Tilley, C. E. (1969). Melting and crystallization relations of Kilauean basalts of Hawaii. The lavas of the 1959–60 Kilauea eruption. Earth and Planetary Science Letters 5, 469–477.[CrossRef][Web of Science]
Thornber, C. R. (2001). Olivine–liquid relations of lava erupted by Kilauea Volcano from 1994–1998: implications for shallow magmatic processes associated with the ongoing east rift zone eruption. In: Canil, D. Jamieson, H. E. & Martin, R. F. (eds) Special Roeder Volume, ‘Homogeneous and Heterogeneous Equilibria in Magmas’. Canadian Mineralogist 39, 323–350.
Thornber, C. R. (2003). Magma reservoir processes revealed by geochemistry of the ongoing Pu'u'O'o–Kupaianaha eruption. In: Heliker, C. C., Swanson, D. A. & Takahashi, T. J. (eds) The Pu'u'O'o–Kupaianaha Eruption of Kilauea Volcano, Hawaii, the First Twenty Years. US Geological Survey Professional Paper 1676, 121–136.
Thornber, C. R. & Huebner, J. S. (1985). Dissolution of olivine in basaltic liquids: experimental observations and application. American Mineralogist 70, 934–945.[Abstract]
Thornber, C. R., Mattox, T. N., Sutton, A. J., Elias, T., Okubo, P., Tomori, A. & Denlinger, R. (1995). The eruptive pulse of Kilauea Volcano as revealed by continuous temperature monitoring of an active lava tube. EOS Transactions, American Geophysical Union 76, 681.
Thornber, C. R., Heliker, C. C., Reynolds, J. R., Kauahikaua, J. P., Okubo, P., Lisowski, M., Sutton, J. & Clague, D. (1996). The eruptive surge of February 1, 1996: a highlight of Kilauea's ongoing East Rift Zone eruption. EOS Transactions, American Geophysical Union 77, 798.
Thornber, C. R., Sherrod, D., Heliker, C. C., Kauahikaua, J. P., Trusdell, F., Lisowski, M. & Okubo, P. (1997). Kilauea's on-going eruption: Napau Crater revisited after 14 years. EOS Transactions, American Geophysical Union 78, S329.
Thornber, C. R., Sherrod, D. R., Siems, D. F., Heliker, C. C., Meeker, G. P., Oscarson, R. L. & Kauahikaua, J. P. (2002). Whole-rock and glass major-element chemistry of Kilauea Volcano, Hawaii, near-vent eruptive products: September 1994 through September 2001. US Geological Survey Open File Report 02-17 (available at http://geopubs.wr.usgs.gov/open-file/of02-017).
Wolfe, E. W., Neal, C. A., Banks, N. G. & Duggan, T. J. (1988). Geologic observations and chronology of eruptive events. In: Wolfe, E. W. (ed.) The Puu Oo Eruption of Kilauea Volcano, Hawaii: Episodes 1 through 20, January 3, 1983, through June 8, 1984. US Geological Survey Professional Paper 1463, 1–99.
Wright, T. L. (1971). Chemistry of Kilauea and Mauna Loa Lava in Time and Space. US Geological Survey Professional Paper 735, 40 pp.
Wright, T. L. & Doherty, P. C. (1970). A linear programming and least squares method for solving petrologic mixing problems. Geological Society of America Bulletin 81, 1995–2008.
Wright, T. L. & Fiske, R. S. (1971). Origin of the differentiated and hybrid lavas of Kilauea Volcano, Hawaii. Journal of Petrology 12, 1–65.
Wright, T. L. & Helz, R. T. (1996). Differentiation and magma mixing on Kilauea's east rift zone: a further look at the eruptions of 1955 and 1960. Part II. The 1960 lavas. Bulletin of Volcanology 57, 602–630.[CrossRef][Web of Science]
Yang, H.-J., Kinzler, R. J. & Grove, T. L. (1996). Experiments and models of anhydrous olivine–plagioclase–augite saturated melts from 0·001–10 kbar. Contributions to Mineralogy and Petrology 124, 1–18.[CrossRef][Web of Science]
Yang, H.-J., Frey, F. A., Clague, D. A. & Garcia, M. O. (1999). Mineral chemistry of submarine lavas from Hilo Ridge, Hawaii: implications for magmatic processes within Hawaiian rift zones. Contributions to Mineralogy and Petrology 135, 355–372.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. A. Lange, H. M. Frey, and J. Hector A thermodynamic model for the plagioclase-liquid hygrometer/thermometer American Mineralogist, April 1, 2009; 94(4): 494 - 506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. Marske, M. O. Garcia, A. J. Pietruszka, J. M. Rhodes, and M. D. Norman Geochemical Variations during Kilauea's Pu'u 'O'o Eruption Reveal a Fine-scale Mixture of Mantle Heterogeneities within the Hawaiian Plume J. Petrology, July 1, 2008; 49(7): 1297 - 1318. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. D. Putirka Thermometers and Barometers for Volcanic Systems Reviews in Mineralogy and Geochemistry, January 1, 2008; 69(1): 61 - 120. [Full Text] [PDF]
|
|
|
|
|
|
M. Norman, M. O. Garcia, and A. J. Pietruszka Trace-element distribution coefficients for pyroxenes, plagioclase, and olivine in evolved tholeiites from the 1955 eruption of Kilauea Volcano, Hawai'i, and petrogenesis of differentiated rift-zone lavas American Mineralogist, May 1, 2005; 90(5-6): 888 - 899. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. D. Putirka Igneous thermometers and barometers based on plagioclase + liquid equilibria: Tests of some existing models and new calibrations American Mineralogist, February 1, 2005; 90(2-3): 336 - 346. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||