V33D-2240
Depths of Magma Storage Beneath Fogo Volcano, Cape Verde Islands
Fogo is one of the most active oceanic volcanoes of the world in present time and the only island of the Cape Verde archipelago with historic volcanic activity. We have carried out a barometric study of basanitic to tephriphonolitic volcanic rocks of the 1995 eruption of Fogo in order to reconstruct the depths of magma reservoirs and magma pathways prior to eruption. The pyroclastic rocks and lavas studied span the whole temporal and compositional range of this eruption. Clinopyroxene-melt thermobarometry of 75 clinopyroxenes in 9 samples yields a well-defined pressure range of 500-630 MPa (average 550 MPa). This pressure range is interpreted to reflect a major fractionation level at ca. 17-22 km depth, within the uppermost mantle, where melt and phenocrysts last equilibrated. Microthermometry of CO2-rich fluid inclusions in clinopyroxene and olivine phenocrysts indicates a broader pressure range with two apparent frequency maxima. The higher pressure range, between 430 and 510 MPa (average 480 MPa), partly overlaps with the clinopyroxene-melt barometry data. The lower pressure range, between 250 and 430 MPa (average 375 MPa), is within the lower crust to Moho and may reflect short-term stagnation of magma at 9-16 km depth prior to eruption. Our data suggest that the 1995 magmas ascended from mantle depth to the surface with residence times at shallow levels probably being restricted to less than a day. This conclusion is in accordance with the absence of plagioclase phenocrysts and microphenocrysts in the 1995 and most other recent lavas.
V33D-2241
Equilibrium Assessment for the Application of Clinopyroxene-Melt Thermobarometry
The composition of igneous clinopyroxenes depends on pressure (P), temperature (T), and composition of the melt from which they crystallized. This is the basis for clinopyroxene-melt thermobarometry [1], which has been widely used in the last years to determine the pressure of crystal fractionation and the depth of magma reservoirs. The assessment of chemical equilibrium between crystal and host melt, however, is not straightforward and has received only little attention. Because most natural clinopyroxenes show strong chemical zonations, the calculated P and T critically depend on the location of the spots analyzed. We investigated this effect by analyzing natural euhedral clinopyroxenes along traverses parallel as well as perpendicular to crystal faces, with the aim of refining criteria for equilibrium assessment and data filtering. As expected, calculated P and T variably depend on spot distance to the crystal surface even within the outermost 20 μm. Surprisingly, in some clinopyroxenes, these P-T variations are surpassed by those between different crystal faces, even though there is no apparent sector zonation. In order to select the data points that are closest to the composition in equilibrium with the host melt, we propose a data filter using the clinopyroxene components predicted by the surface saturation model of Putirka [2]. The filter is based on the sum of squared residuals (SSR) between measured and predicted clinopyroxene compositions. This sum shows a minimum in plots of SSR vs. P and SSR vs. molar Fe/Mg in clinopyroxene if sufficient points are analyzed, which can be used as a cut-off criterion for data filtering. This procedure allows for the removal of outliers thereby significantly reducing the scatter in calculated P and T of natural samples. [1] Putirka et al. (1996), Contrib. Mineral. Petrol. 123, 92-108. [2] Putirka (1999), Contrib. Mineral. Petrol. 135, 151-163.
V33D-2242
Water under-saturated phase equilibria of basaltic andesites from Westdahl volcano, Alaska
The two most abundant gases released from magmatic systems are typically H2O and CO2, however, most phase equilibria studies examining crystallization applied to natural magmatic systems over the past 200 years have relied on H2O-saturated conditions. We will present the results of new phase equilibria experiments run using natural basaltic andesite starting materials from the 1991-1992 eruption of Westdahl volcano, Alaska, examining both H2O-saturated and undersaturated conditions, using a fixed ratio of XH2O ~0.7 and XCO2 ~0.3 in the total volatile budget. The experiments were conducted at total pressures (PTotal) of 0-200 MPa and 900-1050 °C, and fO2 set to the Ni-NiO buffer. Experiments were loaded into gold and Au75Pd25 capsules, and run in a TZM alloy pressure vessel for 48 hours before rapid quenching while still at pressure. After quenching, samples were polished and examined by microprobe and reflective microscopy. Identified mineral phases include plagioclase, clinopyroxene, Fe-Ti oxides, and minor orthopyroxene in both water-saturated and under- saturated experiments. A ~25 to 50 °C shift in temperature, at similar pressures is observed in the plagioclase and pyroxene stability curves when CO2 is added. Solubility models predict relatively low amounts of CO2 dissolved in the melt at similar conditions. Thus, our experiments indicate a significant effect of CO2 on the crystallization of mafic magmas at crustal pressures in volcanic arcs.
V33D-2243
Phenocrysts Crystallisation Pressures and Temperatures and Melts Evolution at La Fossa Volcano (Vulcano Island, Italy)
La Fossa Volcano (Vulcano Isl., Aeolian Arch., Italy) erupted last, explosively, in 1888-1890. Its eruptive history includes at least four eruptive cycles of mixed eruptions with strombolian and hydromagmatic phases followed generally by small lava flows ranging in composition from latites to trachytes and rhyolites. Crystallisation temperatures and pressures of phenocrysts and melts chemical evolution, have been modeled via thermochemical calculations and HP-HT laboratory experiments. The crystallisation temperatures and pressures of olivine and clinopyroxene phenocrysts of latitic and trachitic lavas (Punte Nere, Grotte dei Palizzi) were obtained via the empirical olivine-clinopyroxene-liquid thermobarometer of Sugawara (2000) and the olivine geothermometer of Ariskin et al. (1993) which gave consistent values of 1120°C - 60 MPa and 1080°C - 50 MPa. For the trachytic lava of Grotte dei Palizzi and the rhyolitic blocks of 1888-90 eruption, T - P of 1030°C - 50 MPa and 1000°C - 40 MPa were obtained using the two feldspars thermochemical equilibrium model of Green and Usdansky (1986). The %(H2O)m of trachytic and rhyolitic melts, in the range of 1.7 - 2.7% and 2.4 - 2.7% respectively, was obtained after the experimental calibration at P = 100 MPa and T = 1000-1020°C of X(An)plg against the %(H2O)m. The phase relations and melts composition under the above indicated conditions were finally investigated by the MELTS code (Ghiorso and Sack, 1995), allowing us to show how the more primitive latitic melt can evolve toward the more evolved trachytic and rhyolitic compositions.
V33D-2244
Magma Chamber Depths at Kverkfjoll volcanic system, Iceland and Connections to Recent Dyke Injection at Mt. Upptyppingar, Iceland.
Rifting along mid-ocean ridges (MOR) is facilitated by injection of dykes and magma chambers within the crust in rifting events which occur at discrete locations along the axis of a rift zone. These volcano-tectonic events in Iceland's rift zones are often accompanied by outpouring of lava at the Earth's surface. We have used mineral-melt equilbria relationships to determine the pressure and temperature of magmas erupted from the Kverkfjoll volcanic system, located along the eastern margin of the Northern Volcanic Zone (NVZ). This volcanic system consists of the central volcano, Kverkfjoll volcano and a series of fissure eruptions extending northward approximately 50 km to Mt. Upptyppingar, a large mound of hyaloclastites and pillow lavas. Pressures and temperatures were calculated for melts lying along the olivine, plagioclase and augite cotectic. Input data were published analyses of basaltic glass in samples of pillow lava, as well as samples of glass in melt inclusions in phenocrysts. The results provide evidence for partial crystallization at average pressures of 426 MPa and 822 MPa, corresponding to depths of 15 km and 29 km. We conclude that the sampled lavas were erupted from magma chambers at these depths. These depths are consistent with seismic data for recent swarms of microseismicity beneath Upptyppingar (February 2007 to May 2008). The seismic activity is interpreted to reflect influx of magma into a dyke at a depth of 15 to 18 km. The seismic unrest at Upptyppingar might signal a rifting episode in the NVZ of Iceland. The latter accommodates the full 20 mm/yr spreading in this portion of the mid-Atlantic Ridge, but it has been over 20 years since the last major rifting episode in the NVZ at Krafla. Eruption of magmas from deep chambers will promote rapid degassing and may lead to explosive activity that could inject volcanic aerosols sufficiently high in the atmosphere to affect the Earth's climate.
V33D-2245
Petrological Constraints on Magma Chamber Depths, Magma Plumbing Systems, and Crustal Thickness along the Reykjanes Ridge, North Atlantic
Plate spreading is accompanied by intrusion of dikes and eruption of lava mainly along the axis of mid-ocean ridges. The dikes are fed by magma chamber(s) located beneath the ridge. It has been suggested that the depth of magma chambers is related to the rate of spreading. In order to examine this hypothesis we determined the depths of magma chambers beneath the slow spreading Reykjanes Ridge that extends form the Charlie Gibbs fracture zone at 53° N to the southern tip of Iceland at 64° N in the North Atlantic. The method involves comparing the compositions of natural liquids with those of liquids in equilibrium with the minerals olivine, plagioclase, and clinopyroxene. We used pressures calculated by this method to estimate the depths of partial crystallization of the liquids in sub-crustal chambers or reservoirs. Chemical analyses of mid-ocean ridge basalts (MORB) glasses collected along the Reykjanes Ridge were used as liquid compositions. The glasses form by rapid cooling of magma when quenched by contact with seawater, and provide unambiguous samples of natural basalt liquids. The results indicate that the depth of magma chambers decreases from 4 to 8+/-0.75 km near the Charlie Gibbs fracture zone to 1.17+/-0.53 km at 55.67° N. Further north, the average depth of chambers increases to 9.71+/-3.17 km as the ridge approaches Iceland. The shallow depths obtained for chambers beneath the southern part of the ridge are contrast with results obtained for slow-spreading ridges elsewhere. This may reflect increased magma flux associated with the Iceland plume, and this is consistent with the crust thickening farther north towards Iceland as interpreted from the northerly increase in the maximum depths calculated for chambers along the ridge. The northward thickening of crust indicated by the maximum calculated depths of magma chambers is consistent with seismic data for this region. The influence of the Iceland plume is apparent from the chemical analyses of the glasses. The abundances of Ti, Na, K, P, and Fe increase whereas the abundances of Si, Mg, Al, and Ca decrease as Iceland is approached.
V33D-2246
Tectonic Influence on Magma Storage and Ascent During the Older Evolutionary Stages (223-105 ka) of the Lipari Island (Aeolian Archipelago, Southern Italy)
Pure CO2 fluid inclusions are observed in fifteen quartz-rich xenoliths collected in basaltic-andesitic to
andesitic volcanic products relevant to the older evolutionary stages of Lipari Island (223-105 ka). In
volcanics forming central composite volcanoes (M. Mazzacaruso, 223-127 ka; M. S.Angelo, 105 ka), fluid
inclusions are trapped during two distinct events: early Type I inclusions formed before host magma
transport, and late (i.e. secondary) Type II inclusions trapped during magma ascent. Early Type I inclusions
show homogenization temperatures corresponding to densities from 0.9 to 0.6 g/cc, while Type II inclusions
record a considerably lower density interval between 0.38 and 0.1 g/cc. At the estimated trapping
temperatures between 950 and 1090°C, obtained density values correspond to pressures of 0.58-
0.25 GPa (22-10 km) for Type I, and 0.13-0.03 GPa (5.5-1 km) for Type II inclusions, respectively.
In those
magmas erupted from fissural eruptive vents aligned along the main regional NNW-SSE and E-W faults
systems (Timpone Ospedale, Monterosa and M. Chirica; 223-127 ka) only early Type I inclusions are
observed. Density values form to two distinct intervals between 0.87 and 0.6 g/cc (0.53-0.25 GPa; 20-10 km;
M. Chirica), and between 0.68-0.18 g/cc (0.32-0.05 GPa; 12-2 km; Timpone Ospedale and Monterosa). Fluid
inclusion data together with tectonic features outline a complex magma storage and ascent evolution during
the Lipari's older evolutionary stages. Beneath the central volcanoes of M. Mazzacaruso, M. S.Angelo and
the M. Chirica, two magma reservoirs, located at lower crustal depths (~22 km; close to the Moho) and
at very shallow levels (5.5-1 km), are present. Mantle-Derived magmas are accumulated into the deep
magma chamber and may then reside in the shallower reservoir for a short period of time before being
erupted to the surface. Such a magma feeding system is similar to those outlined for the Alicudi and
Stromboli volcanoes, and for most of the Vulcano's eruptive stages. Conversely, an intermediate magma
reservoir at middle crustal levels (~12 km) is shown to play an important role in storage and
differentiation processes of mafic magmas relevant to the fissural eruptive vents of Timpone Ospedale,
Monterosa, and M. Chirica. Magmas more likely arise from the deepest magma storage level located close to
the Moho, as outlined by M. Chirica eruptive vent. The proposed scenario is that the regional fault systems
control the magma storage evolution, contributing to create a zone of preferential accumulation at Mid-
Crustal levels. Fault systems may also influence magma ascent and determine the upward magma movement
directly to the eruptive system without an effective ponding in the shallow reservoirs located at 5-1 km depth.
At middle crustal depths, magmas may reside for long time, and low rates of fractional crystallization occur.
The occurrence of an intermediate storage level at similar depths beneath the rhyolitic Lentia domes at
Vulcano, which are aligned along the main NNW-SSE tectonic trend, supports present model.
V33D-2247
Thermobarometry of Clinopyroxenes From Yapoah Crater and the Chichinautzin Volcanic Field: Implications for Crystallization Depths and Magma Mixing
To investigate the crystallization depths of mafic arc magmas, we used clinopyroxene thermobarometry (Putirka et al., 2003) on basaltic andesite samples from one cinder cone in the Oregon Cascades (Yapoah Crater) and four cinder cones in the Chichináutzin Volcanic Field (CVF) in central Mexico. Magmatism in both of these regions is related to subduction, and melt inclusion studies show that the mafic magmas contained 3-5 wt% H2O (Cervantes & Wallace, 2003, Geology; Ruscitto et al., unpublished). Clinopyroxene phenocrysts are uncommon in mafic lava and tephra from these two areas. We determined that clinopyroxene phenocrysts in the samples we studied had crystallized at relatively low pressures (<5 kbar) and at temperatures ranging from 1060°C-1110°C. Melt inclusion studies from these regions have found that olivine crystallized at similar pressures; for the CVF olivine crystallized at 1-5 kbar and at Yapoah it crystallized at <3 kbar (Cervantes & Wallace, 2003, Geology; Ruscitto et al., unpublished). Most of the clinopyroxenes had Mg/Fe ratios that indicated they had crystallized from magmas that were more evolved than those in which they were erupted, and thus we infer that magma mixing was a common process in the clinopyroxene-bearing lavas. Some clinopyroxenes from Yapoah Crater, however, formed at very low pressures (<1 kbar) and probably were in equilibrium with their host magma. Based on our results, we conclude that hydrous arc magmas in central Oregon and the CVF crystallized clinopyroxene at relatively low pressures and that magma mixing was an important process.
V33D-2248
Pre-eruptive conditions of the ~31 ka rhyolitic magma of Tlaloc volcano, Sierra Nevada Volcanic Range, Central Mexico
Tlaloc volcano is located at the northern tip of the Sierra Nevada Volcanic Range in Central Mexico. This Pleistocene to Recent volcanic range consists from north to south of Tlaloc-Telapón-Teyotl-Iztaccíhuatl-and- Popocatépetl volcanoes. While andesitic to barely dacitic volcanism dominates the southern part of the range (i.e. Popocatépetl and Iztaccíhuatl); dacitic and rare rhyolithic volcanism (i.e. Telapón, Tlaloc) dominates the northern end. The known locus of rhyolitic magmatism took place at Tlaloc volcano with a Plinian-Subplinian eruption that occurred 31 ka ago. The eruption emplaced the so-called multilayered fallout and pumiceous pyroclastic flows (~2 km3 DRE). The deposit consists of 95% vol. of juvenile particles (pumice + crystals) and minor altered lithics 5% vol. The mineral association of the pumice fragments (74-76 % wt. SiO2) consists of quartz + plagioclase + sanidine + biotite and rare oxides set in a glassy groundmass with voids. Melt inclusions in quartz phenocrysts suggest that prior to the eruption the rhyolitic contain ~7% of H2O and <110 ppm of CO 2, suggesting pressure conditions around ~2500 bars and therefore depths ~8 km below the volcano. Such depths suggest that inception of rhyolitic magmatism at Tlaloc volcano halted at deeper conditions than andesitic to dacitic eruptions of Popocatépetl volcano (~6 km) in the southern part of the Sierra Nevada Volcanic Range and than Nevado de Toluca volcano (~6 km) some 50 km to the southwest.