Groups Of Melt Inclusions Research Articles

Constraining the volatile evolution of mafic melts at Mt. Somma–Vesuvius, Italy, based on the composition of reheated melt inclusions and their olivine hosts

Abstract. Mount Somma–Vesuvius is a stratovolcano that represents a geological hazard to the population of the city of Naples and surrounding towns in southern Italy. Historically, volcanic eruptions at Mt. Somma–Vesuvius (SV) include high-magnitude Plinian eruptions, such as the infamous 79 CE eruption that occurred after 295 years of quiescence and killed thousands of people in Pompeii and surrounding towns and villages. The last eruption at SV was in 1944 and showed a Volcanic Explosivity Index (VEI) of 3 (0.01 km3 of volcanic material erupted). Following the 1944 eruption, SV has been dormant for the past nearly 79 years, with only minor fumarolic and seismic activity. During its long history, centuries of dormancy at SV have ended with Plinian eruptions (VEI 6) that signal the beginning of a new cycle of eruptive activity. Thus, the current dormancy stage demands a need to better understand the mechanism involved in high-magnitude eruptions in order to better predict future eruption magnitude and style. Despite centuries of research on the SV volcanic system, many questions remain, including the evolution of magmatic volatiles from deep primitive magmas to shallower more evolved magmas. Developing a better understanding of the physical and chemical processes associated with volatile evolution at SV can provide insights into magma dynamics and the mechanisms that trigger highly explosive eruptions at SV. In this study, we present new data for the pre-eruptive volatile contents of magmas associated with four Plinian and two inter-Plinian eruptions at SV based on analyses of reheated melt inclusions (MIs) hosted in olivine. We correct the volatile contents of bubble-bearing MIs by taking into account the volatile contents of bubbles in the MIs. We recognize two groups of MIs: one group hosted in high-Fo olivine (Fo85–90) and relatively rich in volatiles and the other group hosted in low-Fo olivine (Fo70–69) and relatively depleted in volatiles. The correlation between volatile contents and compositions of host olivines suggests that magma fractionation took place under volatile-saturated conditions and that more differentiated magmas reside at shallower levels relative to less evolved/quasi-primitive magmas. Using the CO2 contents of corrected MIs hosted in Fo90 olivine from SV, we estimate that 347 to 686 t d−1 of magmatic CO2 exsolved from SV magmas during the last 3 centuries (38–75 Mt in total) of volcanic activity. Although this study is limited to only few SV magmas, we suggest that further study applying similar methods could shed light on the apparent lack of correlation between the volatile contents of MIs and the style and age of eruptions. Further, such studies could provide additional constraints on the origin of CO2 and the interaction between the carbonate platform and ascending magmas below SV.

Geochemical and lead isotope compositions of olivine-hosted melt inclusions from the Yaeyama Graben in the southern Okinawa Trough: Implications for slab subduction and magmatic processes

Chemical compositions were measured in olivine-hosted melt inclusions in basaltic volcanic rocks from the Yaeyama Graben in the southern Okinawa Trough to investigate the magmatic processes of erupted lavas in an incipient epicontinental back-arc basin in the western Pacific. Melt inclusions have widely variable element and Pb isotope compositions, while erupted host whole-rock compositions are homogeneous, suggesting that melt inclusions preserve more information about magmatic processes than their host whole rocks. Two groups of melt inclusions can be distinguished by their trace elements and Pb isotopes. Group 1 melt inclusions have geochemical compositions similar to those of their host whole rocks, which have low 207Pb/206Pb ratios (<0.850), and may be attributed to the contributions of recycled continental crust and subducted sediments to magma sources during Philippine Sea Plate subduction. Group 2 melt inclusions have higher 207Pb/206Pb ratios (>0.850) and Li, Rb, and U contents and lower Cu, Zn, and Pb contents than their host whole rocks. High Li, Rb, and U contents may be induced by specific melts generated by possible breakdown reactions of some accessory minerals, such as phengite and cordierite, while low Cu, Zn, and Pb contents are attributed to sulfide precipitation, metal flushing and/or metallic element fractionation during magma evolution. The Pb isotopes of the group 2 melt inclusions are similar to those of the Kuanhsi-Chutung basalts, which are correlated with some intraplate volcanism in the Fujian-Taiwan region. We infer that the group 2 melts were preexisting small volumes of magma that were emplaced in the middle-lower continental crust (~11 km), experienced extensive magma evolution and were entrained by upwelling magma. Our melt inclusion data suggest that magma sources are compositionally heterogeneous and that the mixing of magmas from different sources may be widespread beneath the southern Okinawa Trough because of Philippine Sea Plate subduction and preexisting intraplate magmatism.

Highly explosive basaltic eruptions driven by CO2 exsolution

The most explosive basaltic scoria cone eruption yet documented (>20 km high plumes) occurred at Sunset Crater (Arizona) ca. 1085 AD by undetermined eruptive mechanisms. We present melt inclusion analysis, including bubble contents by Raman spectroscopy, yielding high total CO2 (approaching 6000 ppm) and S (~2000 ppm) with moderate H2O (~1.25 wt%). Two groups of melt inclusions are evident, classified by bubble vol%. Modeling of post-entrapment modification indicates that the group with larger bubbles formed as a result of heterogeneous entrapment of melt and exsolved CO2 and provides evidence for an exsolved CO2 phase at magma storage depths of ~15 km. We argue that this exsolved CO2 phase played a critical role in driving this explosive eruption, possibly analogous to H2O exsolution driving silicic caldera-forming eruptions. Because of their distinct gas compositions relative to silicic magmas (high S and CO2), even modest volume explosive basaltic eruptions could impact the atmosphere.

Ultra-depleted melts in olivine-hosted melt inclusions from the Ontong Java Plateau

The Ontong Java Plateau (OJP) represents the largest large igneous province (LIP) preserved in the geologic record. The most voluminous volcanic types of the OJP—the Kroenke and Kwaimbaita groups, which dominate the accessible portions of the plateau—have relatively flat primitive mantle normalized rare earth element (REE) patterns. With the exception of relatively small volumes of late-stage melts—referred to as the Singgalo group—that are characterized by slightly enriched REE patterns relative to a chondritic pattern, the volcanic groups that dominated the eruptive history of the OJP exhibit remarkably homogeneous, flat REE patterns. Here we isolate, for the first time, olivine-hosted melt inclusions from the OJP. We show that the melt inclusions have two clear populations defined by distinct trace element characteristics. The first population has relatively flat trace element patterns that are similar to those observed in whole rock lavas from the most voluminous volcanic groups (Kroenke and Kwaimbaita) recorded in the OJP. In contrast, a second group of melt inclusions, referred to as UDM (ultra-depleted melt) inclusions, exhibit (light-REE) LREE-depleted patterns relative to a chondritic pattern; these trace element patterns are far more depleted than any previously reported lava from OJP. The UDM have unique trace element signatures that preclude an origin by re-melting the depleted mantle source left over after melt extraction during construction of the OJP. We interpret the new UDM compositions to be the result of melting a previously unrecognized (in lavas) ultra-depleted component hosted in the OJP mantle source.

An assessment of the reliability of melt inclusions as recorders of the pre-eruptive volatile content of magmas

Many studies have used melt inclusions (MI) to track the pre-eruptive volatile history of magmas. Often, the volatile contents of the MI show wide variability, even for MI hosted in the same phenocryst. This variability is usually interpreted to represent trapping of a volatile-saturated melt over some range of pressures (depths) and these data are in turn used to define a magma degassing path. In this study, groups of MI that were all trapped at the same time (referred to as a melt inclusion assemblage or MIA) based on petrographic evidence, were analyzed to test the consistency of the volatile contents of MI that were all trapped simultaneously from the same melt. MIA hosted in phenocrysts from White Island (New Zealand) and from the Solchiaro eruption on the Island of Procida (Italy) were analyzed by secondary ion mass spectrometry (SIMS). In most MIA, H2O, F, and Cl abundances for all MI within the MIA are consistent (relative standard errors <27%, with the exception of two MIA), indicating that the MI all trapped a melt with the same H2O, F, and Cl concentrations and that the composition was maintained during storage in the magma as well as during and following eruption. In several MIA, S abundances are consistent (relative standard errors <33%, with the exception of five out of 28 MIA). Conversely, CO2 (White Island and Solchiaro MIA) showed wide variability in several MIA. The result is that some MIA display a wide range in CO2 content at approximately constant H2O. Similar trends have previously been interpreted to represent degassing paths, produced as volatile-saturated melts are trapped over some significant pressure (depth) range in an ascending (or convecting) magma body. However, the CO2 vs. H2O trends obtained in this study cannot represent degassing paths because the MI were all trapped at the same time (same MIA). This requires that all of the MI within the MIA trapped a melt of the same composition (including volatile content) and at the same temperature and pressure (depth). The cause of the variable concentration of CO2 within some MIA is unknown, but may reflect micrometer-scale heterogeneities within the melt during trapping, heterogeneities within individual MI, post-entrapment crystallization within the MI, or C-contamination during sample preparation. These results suggest that trends showing variable CO2 and relatively uniform H2O obtained from MI may not represent trapping of volatile-saturated melts over a range of pressure, and care must be taken when interpreting volatile contents of MI to infer magma degassing paths. Results of this study have been used to estimate the uncertainties in volatile concentrations of MI determined by SIMS analysis. The H2O, F, and Cl contents have an average estimated uncertainty of 11, 9, and 12%, respectively, which is consistent with the SIMS analytical error. In contrast, the S and CO2 contents have an average estimated uncertainty of 24 and 69%, respectively, which is considerably larger than the SIMS analytical error.

Volatile and trace element composition of melt inclusions from the Lower Bandelier Tuff: Implications for magma chamber processes and eruptive style

The Lower Bandelier Tuff, erupted from the Valles Caldera at 1.51 Ma, is composed of a Plinian tephra and associated ignimbrite. Based on ion and electron microprobe analyses of melt inclusions (MI) in quartz, sanidine, and pyroxene phenocrysts, a strong volatile gradient was present in the upper portion of the magma chamber. In the 20 km3 of magma that produced the Plinian tephra, most H2O and F contents of MI are between 4 to 5 wt % and 0.18 to 0.20 wt%, respectively, and between 3 to 4 wt% and 0.12 and 0.14 wt% in the first‐deposited basal ignimbrite. Measurements of the H2O and F contents of the 380 km3 of magma which produced the bulk of the ignimbrite are concentrated between 2 to 3 wt% and 0.05 to 0.14 wt% respectively. The invariance of H2O (and possibly Cl and B) relative to F in the magma which formed the Plinian tephra suggests that this portion of the magma chamber was saturated with respect to an H2O‐rich vapor phase. The trace element composition of MI are varied, and overlap with composition of pumice lumps. A group of MI are rich in Ti, Sr, Ba and low in B, Cl, Rb, Y, Nb, and Th, and are compositionally similar to analyses of matrix glass from some ignimbrite pumice lumps. These represent the mixing of a second rhyolite into the base of the Lower Bandelier magma chamber. MI influenced by the second magma have not been found in Plinian samples, but occur in ignimbrite samples from base to top of the deposit. When these MI are removed from the data set, the remainder show strong linear correlations between Nb and Rb, Y, Zr, and Th. These correlations can be most easily explained by fractionation of approximately ∼40% quartz and alkali feldspar (with trace amounts of chevkinite). MI from the Plinian tephra are similar to bulk Plinian pumice composition, suggesting that magmatic evolution was well‐progressed at the time that the MI were trapped. However, fractional crystallization is difficult to reconcile with the observed distribution of some other elements, including H, B, Li, F and Cl. The enrichment of H2O and F in the upper portion of the magma chamber, along with their lack of correlation with Nb, Rb, Zr, Y, and Th, suggest that these sets of elements were decoupled during magmatic evolution. The upward enrichment of H2O and F must have occurred at a faster rate than other trace elements. The observed volatile gradient may have influenced the eruption dynamics of the magma, initially causing discrete layers of magma to be removed when the gradient was large, followed by chaotic eruption of the bulk of the ignimbrite when the gradient was small. Comparison with the Bishop Tuff eruption, however, suggests that the chaotic eruption style of the LBT may be related to vent geometry and conduit evolution, promoting high discharge rates, not differences in the absolute concentration of volatiles.