Olgeir Sigmarsson

A detailed Th, Sr and O isotope study of Hekla: differentiation processes in an Icelandic Volcano

Abstract238U−230Th disequilibria and Sr and O isotope ratios have been measured in a suite of samples from most of the known prehistoric and historic eruptions of Hekla volcano, Iceland. They cover the compositional range from basaltic andesite to rhyolite. Recent basalts erupted in the vicinity of the volcano and a few Pleistocene basalts have also been studied. Geochemical data indicate that the best tracers of magmatic processes in Hekla are the (230Th/232Th) and Th/U ratios. Whereas most geochemical parameters, including Sr, Nd and O isotopes, could be compatible with crystal fractionation, (230Th/232Th) and Th/U ratios differ in the basalts and basaltic andesites (1.05 and 3.2, respectively) and in the silicic rocks, dacites and rhyolites (0.98 and 3.4–3.7, respectively). This observation precludes fractional crystallization as the main differentiation process in Hekla. On the basis of these results, the following model is proposed: basaltic magmas rise in the Icelandic crust and cause partial melting of metabasic rocks, leading to the formation of a dacitic melt. The basaltic magma itself evolves by crystal fractionation and produces a basaltic andesite magma. The latter can mix with the dacitic liquid to form andesites. At higher levels in the magma chamber, the dacitic melt sometimes undergoes further differentiation by crystal fractionation and produces subordinate volumes of rhyolites. Together all these processes lead to a zoned magma chamber. However, complete zoning is achieved only when the repose time between eruptions is long enough to allow the production of significant volumes of dacitic magma by crustal melting. This situation corresponds to the large plinian eruptions. Between these eruptions, the so-called intra-cyclic activity is characterized by the eruption of andesites and basaltic andesites, with little crustal melting. The magmatic system beneath Hekla most probably was established during the Holocene. The shape and the size of the magma chamber may be inferred from the relationships between the composition of the lavas and the location of the eruption sites. In a cross-section perpendicular to Heklas ridge, a bell-shaped reservoir 5 km wide and 7 km deep appears the most likely; its top could be at depth of 8 km according to geophysical data.

Origin of silicic magma in Iceland revealed by Th isotopes

Th, Sr, Nd, and O isotopes have been determined in a suite of volcanic rocks from Hekla and in a few samples from Askja and Krafla volcanic centers in Iceland. Although {sup 87}Sr/{sup 86}Sr and {sup 143}Nd/{sup 144}Nd ratios are nearly the same for all compositions at Hekla, the ({sup 230}Th/{sup 232}Th) ratios differ and thus clearly show that the silicic rocks cannot be derived from fractional crystallization of a more primitive magma. Similar results are obtained for the Krafla and Askja volcanic centers, where the {delta}{sup 18}O values are much lower in the silicic magma than in the mafic magma. These data suggest that large volumes of silicic rocks in central volcanoes of the neovolcanic zones in Iceland are produced by partial melting of the underlying crust.

Systematics of U-series nuclides in primitive lavas from the 1730–36 eruption on Lanzarote, Canary Islands, and implications for the role of garnet pyroxenites during oceanic basalt formations

Abstract Inferences on mantle melting and related parameters are often derived from U-series results on oceanic basalts. However, most of these basalts come from magma chambers in which magma mixing amongst other processes is likely to have reduced the true compositional spread created during partial melting. Here are presented U–Th–Ra disequilibrium results on historical lavas of near-primary composition from Lanzarote, Canary Islands, where magma chambers have not been detected. During the 1730–36 fissure eruption, the magma composition varied with time from basanites through alkali basalts to tholeiites. Approximately 3–5 km3 of primitive and often ultramafic xenolith-bearing lavas were produced. Abundances of Ba, Th and U generally decreased towards the end of the eruption, and show more than four-fold variations (e.g. Th = 5.75–1.55 ppm). Ba/Th and Ba/U vary by more than 75%, whereas Th/U vary by less than 10%. Excesses of 230Th over 238U range from 18% to 33% and generally increase with alkalinity. (230Th/232Th) decreased significantly from 1.02 to 0.94 with time and is positively correlated with the alkalinity of the lavas. Isotope ratios of Sr and (226Ra/230Th) also correlate with SiO2. (226Ra/230Th) increased with time from 1.10 to 1.56. The correlations between (230Th/232Th), 87Sr/86Sr and SiO2 suggest that the lava compositions represent nearly unmodified primary melts from a lithologically and compositionally heterogeneous mantle source. Decreasing (230Th/238U) and increasing (226Ra/230Th) in basanites to tholeiites can be explained by partial melting of a mixed garnet pyroxenite–lherzolite source, during which the proportion of pyroxenite and extent of melting diminished. Alternatively, the large variations of excess 226Ra and Ba/Th may reflect fluid addition to such a heterogeneous mantle source. In order to conserve the systematic variations of (226Ra/230Th) with other geochemical variables, transfer time through the lithosphere is likely to have been short relative to the half-life of 226Ra. A magma velocity through the lithosphere on the order of 10−5 m/s (0.5 km/year) is compatible with the data. Both the Th–U disequilibria and the Th-isotope ratios of the Lanzarote lavas appear to be related to the mantle source lithology and composition, where garnet pyroxenite melting produced the early alkaline basalts and increasing melting of garnet lherzolite produced the tholeiites. Furthermore, correlations between (230Th/238U) and (232Th/238U) for different volcanic provinces worldwide, strongly suggest that Th–U disequilibria in oceanic basalts are principally caused by partial melting of garnet pyroxenites.

Mantle and crustal contribution in the genesis of Recent basalts from off-rift zones in Iceland: Constraints from Th, Sr and O isotopes

Along the two volcanic off-rift zones in Iceland, the Sn˦fellsnes volcanic zone (SNVZ) and the South Iceland volcanic zone (SIVZ), geochemical parameters vary regularly along the strike towards the centre of the island. Recent basalts from the SNVZ change from alkali basalts to tholeiites where the volcanic zone reaches the active rift axis, and their87Sr/86Sr andTh/U ratios decrease in the same direction. These variations are interpreted as the result of mixing between mantle melts from two distinct reservoirs below Sn˦fellsnes. The mantle melt would be more depleted in incompatible elements, but witha higher3He/4He ratio (R/Ra≈ 20) beneath the centre of Iceland than at the tip of the Sn˦fellsnes volcanic zone (R/Ra≈ 7.5). From southwest to northeast along the SIVZ, the basalts change from alkali basalts to FeTi basalts and quartz-normative tholeiites. TheTh/U ratio of the Recent basalts increases and both (230Th/232Th) andδ18O values decrease in the same direction. This reflects an important crustal contamination of the FeTi-rich basalts and the quartz tholeiites. The two types of basalts could be produced through assimilation and fractional crystallization in which primary alkali basaltic and olivine tholeiitic melts ‘erode’ and assimilate the base of the crust. The increasingly tholeiitic character of the basalts towards the centre of Iceland, which reflects a higher degree of partial melting, is qualitatively consistent with increasing geothermal gradient and negative gravity anomaly. The highest Sr isotope ratio in Recent basalts from Iceland is observed inOr˦fajokull volcano, which has a3He/4He ratio (R/Ra≈ 7.8) close to the MORB value, and this might represent a mantle source similar to that of Mauna Loa in Hawaii.

Origin of 226Ra–230Th disequilibria in arc lavas from southern Chile and implications for magma transfer time

Abstract Improved understanding of mantle melting processes and melt transport requires knowledge of how fast magma is generated and transferred from source region to surface. The rate of magma transfer can in favorable cases be estimated from radioactive disequilibria between nuclides of the 238U series. Young lavas from southern Chile, in which 238U–230Th disequilibria have been measured [Sigmarsson et al., Nature 346 (1990) 163–165; Sigmarsson et al., Nature 394 (1998) 566–569], were analyzed for 226Ra abundances. The disequilibrium between 226Ra and 230Th in these lavas is found to correlate with 238U–230Th disequilibria and 10Be/Be [Morris et al., Nature 344 (1990) 31–36]. These correlations strongly suggest that the excess of 226Ra over 230Th is due to the addition of a slab-derived fluid to the magma source, since Ra and U are fluid-mobile elements and the cosmogenic 10Be is most likely derived from the subducting Nazca plate beneath the Andes. The largest slab signature is observed in the lavas of Villarrica volcano, which is the most active volcano in South America. A model for subduction fluxing is discussed, in which the U series disequilibria in arc lavas will reflect the integrated dehydration process during metamorphism of the subducting plate and the metasomatized mantle, but be principally controlled by the latest hydrous mineral breakdown in the mantle wedge. Repeated precipitation and dehydration mineral reactions of the hydrated mantle could be the homogenization process of the slab input needed to explain the 10Be/Be–B/Be correlation for different arcs [Morris et al., Nature 344 (1990) 31–36]. The fact that excesses of 226Ra and 238U over 230Th are correlated indicates that linear arrays on the (230Th/232Th)–(238U/232Th) diagram are not isochrons reflecting time elapsed since a fluid addition but rather mixing lines between a fluid phase and melts. The 226Ra–230Th disequilibrium in arc lavas suggests significantly shorter timescales for magma transfer, or less than 8000 years. This disequilibrium is consistent with minimum magma transfer rate through the mantle wedge on the order of 10 m/year. Finally, the correlations of (226Ra/230Th) with (238U/232Th) and 10Be/Be in Andean magmas imply that magma chamber residence time is of the same order of magnitude beneath the stratovolcanoes studied.

Gradual caldera collapse at Bárdarbunga volcano, Iceland, regulated by lateral magma outflow

Driven to collapse Volcanic eruptions occur frequently, but only rarely are they large enough to cause the top of the mountain to collapse and form a caldera. Gudmundsson et al. used a variety of geophysical tools to monitor the caldera formation that accompanied the 2014 Bárdarbunga volcanic eruption in Iceland. The volcanic edifice became unstable as magma from beneath Bárdarbunga spilled out into the nearby Holuhraun lava field. The timing of the gradual collapse revealed that it is the eruption that drives caldera formation and not the other way around. Science, this issue p. 262 Magma flow from under the Bárdarbunga volcano drove caldera collapse during the 2014 eruption. INTRODUCTION The Bárdarbunga caldera volcano in central Iceland collapsed from August 2014 to February 2015 during the largest eruption in Europe since 1784. An ice-filled subsidence bowl, 110 square kilometers (km2) in area and up to 65 meters (m) deep developed, while magma drained laterally for 48 km along a subterranean path and erupted as a major lava flow northeast of the volcano. Our data provide unprecedented insight into the workings of a collapsing caldera. RATIONALE Collapses of caldera volcanoes are, fortunately, not very frequent, because they are often associated with very large volcanic eruptions. On the other hand, the rarity of caldera collapses limits insight into this major geological hazard. Since the formation of Katmai caldera in 1912, during the 20th century’s largest eruption, only five caldera collapses are known to have occurred before that at Bárdarbunga. We used aircraft-based altimetry, satellite photogrammetry, radar interferometry, ground-based GPS, evolution of seismicity, radio-echo soundings of ice thickness, ice flow modeling, and geobarometry to describe and analyze the evolving subsidence geometry, its underlying cause, the amount of magma erupted, the geometry of the subsurface caldera ring faults, and the moment tensor solutions of the collapse-related earthquakes. RESULTS After initial lateral withdrawal of magma for some days though a magma-filled fracture propagating through Earth’s upper crust, preexisting ring faults under the volcano were reactivated over the period 20 to 24 August, marking the onset of collapse. On 31 August, the eruption started, and it terminated when the collapse stopped, having produced 1.5 km of basaltic lava. The subsidence of the caldera declined with time in a near-exponential manner, in phase with the lava flow rate. The volume of the subsidence bowl was about 1.8 km3. Using radio-echo soundings, we find that the subglacial bedrock surface after the collapse is down-sagged, with no indications of steep fault escarpments. Using geobarometry, we determined the depth of magma reservoir to be ~12 km, and modeling of geodetic observations gives a similar result. High-precision earthquake locations and moment tensor analysis of the remarkable magnitude M5 earthquake series are consistent with steeply dipping ring faults. Statistical analysis of seismicity reveals communication over tens of kilometers between the caldera and the dike. CONCLUSION We conclude that interaction between the pressure exerted by the subsiding reservoir roof and the physical properties of the subsurface flow path explain the gradual near-exponential decline of both the collapse rate and the intensity of the 180-day-long eruption. By combining our various data sets, we show that the onset of collapse was caused by outflow of magma from underneath the caldera when 12 to 20% of the total magma intruded and erupted had flowed from the magma reservoir. However, the continued subsidence was driven by a feedback between the pressure of the piston-like block overlying the reservoir and the 48-km-long magma outflow path. Our data provide better constraints on caldera mechanisms than previously available, demonstrating what caused the onset and how both the roof overburden and the flow path properties regulate the collapse. The Bárdarbunga caldera and the lateral magma flow path to the Holuhraun eruption site. (A) Aerial view of the ice-filled Bárdarbunga caldera on 24 October 2014, view from the north. (B) The effusive eruption in Holuhraun, about 40 km to the northeast of the caldera

Provenance of basaltic tephra from Vatnajökull subglacial volcanoes, Iceland, as determined by major- and trace-element analyses:

The Holocene eruption history of subglacial volcanoes in Iceland is largely recorded by their tephra deposits. The numerous basaltic tephra offer the possibility to make the tephrochronology in the North Atlantic area more detailed and, therefore, more useful as a tool not only in volcanology but also in environmental and archaeological studies. The source of a tephra is established by mapping its distribution or inferred via compositional fingerprinting, mainly based on major-element analyses. In order to improve the provenance determinations for basaltic tephra produced at Grímsvötn, Bárdarbunga and Kverkfjöll volcanic systems in Iceland, 921 samples from soil profiles around the Vatnajökull ice-cap were analysed for major-element concentrations by electron probe microanalysis. These samples are shown to represent 747 primary tephra units. The tephra erupted within each of these volcanic system has similar chemical characteristics. The major-element results fall into three distinctive compositional groups, all of which show regular decrease of MgO with increasing K2O concentrations. The new analyses presented here considerably improve the compositional distinction between products of the three volcanic systems. Nevertheless, slight overlap of the compositional groups for each system still remains. In situ trace-element analyses by laser-ablation-inductively-coupled-plasma-mass-spectrometry were applied for better provenance identification for those tephra having similar major-element composition. Three trace-element ratios, Rb/Y, La/Yb and Sr/Th, proved particularly useful. Significantly higher La/Yb distinguishes the Grímsvötn basalts from those of Bárdarbunga and Rb/Y values differentiate the basalts of Grímsvötn and Kverkfjöll. Additionally, the products of Bárdarbunga, Grímsvötn and Kverkfjöll form distinct compositional fields on a Sr/Th versus Th plot. Taken together, the combined use of major- and trace-element analyses in delineating the provenance of basaltic tephra having similar major-element composition significantly improves the Holocene tephra record as well as the potential for correlations with tephra from outside Iceland.

Why are so many arc magmas close to 238U-230Th radioactive equilibrium?

Abstract New analyses of 238U-230Th disequilibria are reported for four active volcanoes: Merapi and Krakatoa in the Sunda arc (Indonesia), Masaya in Central America (Nicaragua), and Ambrym in the New Hebrides island arc. Despite a large range in ( 230 Th 232 Th ) ratios (from 0.65 in Merapi andesites to 2.5 in Masaya basalts), 238U and 230Th are close to radioactive equilibrium, as in many other arc magmas. In several mantle sources, Th/U ratios have clearly been modified by metasomatic processes associated with subduction. This is demonstrated in Central America by the correlation between ( 230 Th 232 Th ) and 10 Be 9 Be ratios for several active volcanoes along the arc. It is proposed that the 238U-230Th radioactive equilibrium found in many arc magmas is the result of disequilibrium melting involving an easily melted, slab-derived, metasomatic component which dominates the uranium and thorium budget of the mantle sources. The departure from equilibrium may be either due to mixing with 230Th enriched melts derived from unmetasomatized mantle sources or to a late stage uranium addition by fluids. This latter process, producing uranium enriched magmas, has a greater influence in uranium and thorium poor magmas.

Geothermobarometry of the 2010 Eyjafjallajökull eruption: New constraints on Icelandic magma plumbing systems

The 2010 Eyjafjallajokull eruption in Iceland produced mildly alkaline basalt that was emitted during the initial flank eruptive phase, whereas tephra predominately of benmorite composition was erupted during the second explosive phase from the summit of the volcano. These latter magmas show pervasive magma mingling between basalts and silicic magma. Glass and coexisting equilibrium mineral analyses have been used to define pressure-temperature crystallization paths for the eruption based on melt, clinopyroxene-melt and plagioclase-melt thermobarometry. Temperature calculations show that the early basaltic eruptions from the flank eruption have magmatic temperatures of around 1170°C (±25°C) and a narrow temperature range (<30°C) at any given depth. In contrast, benmoritic products crystallized at lower temperatures (1000-1060°C). Pressure estimates yield an average pressure of 5.6-6.4 kbar (±1.5 kbar) for the basaltic tephra and variable but lower pressures for the benmoritic samples ranging down to 0.6 kbar. The mafic magma mainly crystallized in the deeper crust (16-18 km), whereas mingled magma from the summit eruption crystallized at more shallow crustal levels (2-5 km) suggesting multistage magma ascent. Magmatic water concentrations were estimated with plagioclase-melt hygrometry. The maximum average water content of 1.8 wt % H2O, obtained in one of the summit samples, is in agreement with melt inclusion observations. Water concentration of this or lower levels is demonstrated to only have limited effect on the pressure-temperature calculations.

U, Ra and Ba incorporation during precipitation of hydrothermal carbonates: implications for 226Ra-Ba dating of impure travertines

Abstract We studied U, Ra and Ba incorporation in calcite in a natural CO 2 -rich hydrothermal area from the French Massif Central. Along the western border of the Limagne graben, several springs are exploited for the petrifaction of various artifacts with calcite. These sites offer the opportunity to sample the water and the calcite layers downflow from the spring, and thus to follow the evolution of their U, Ra and Ba contents as precipitation proceeds. Our results show that the apparent partition coefficients of U, Ra and Ba between water and calcite decrease during precipitation for the three elements. We found no direct relation between this variation and the main factors able to influence the partition coefficient, such as precipitation rate, which suggests that the incorporation of these trace elements could result from a composite process of adsorption and coprecipitation. Ra and Ba have a similar behaviour, with an apparent partition coefficient decreasing from 0.80 to 0.47 for Ra and 0.96 to 0.68 for Ba, resulting in a small (≤10%) variation of the Ra/Ba ratio. The apparent partition coefficient of U decreases from 0.38 to 0.20. These apparent coefficients are much higher than equilibrium values but might be applicable to natural systems with high precipitation rates. We also investigated the possibility of using the decay of the 226 Ra-excess, or the decrease of the ( 226 Ra)/Ba ratio to date older deposits. Whereas the 226 Ra initial activity at the time of deposition has not remained constant, and can not be used for dating, the ( 226 Ra)/Ba method gives better results, when appropriate corrections for detrital contamination in Ba are made. Mixing diagrams using Th as an indicator of contamination allow calculation of the ( 226 Ra)/Ba ratio of the pure carbonate component. The calculated ages of five travertine layers range from 330 to 800 years, suggesting a mean deposition rate of about 1 cm/yr. The relatively good agreement of the calculated ages and stratigraphic positions of the samples suggests that this method could be successfully applied to date impure hydrothermal carbonates in the range 0–10 ky.