Stanley N. Williams

Fluxes of mantle and subducted carbon along convergent plate boundaries

The potential impact of increases in atmospheric CO 2 is a topic of considerable controversy. Even though volcanic emission of CO 2 may be very small as compared to anthropogenic emissions, evaluation of natural degassing of CO 2 is important for any model of the geochemical C cycle and evolution of the Earths atmosphere. We report here the mantle C flux in subduction zones based on He and C isotopes and CO 2 / 3 He ratios of high-temperature volcanic gases and medium- and low-temperature fumaroles in circum-Pacific volcanic regions. The calculated volcanic C flux of 3.1 x 10 12 moVa from subduction zones is larger than the flux of 1.5 x 10 12 mol/a from mid-ocean ridges, while contributions from the mantle in subduction zone is only 0.30 x 10 12 mol/a, equivalent to about 20 % of the C flux in mid-ocean ridges. Since the estimated mantle C flux in hot spot regions is insignificant, 0.029 x 10 12 mol/a, we propose that the global mantle C flux is 1.8 x 10 12 mol/a in total. The flux, if accumulated over 4.5 billion year of geological time, amounts to 8.3 x 10 21 mol which agrees well with 9 x 10 21 mol of the present inventory of C at the Earths surface. This may support a continuous degassing model of C or the idea that subducted C is recycled into the lower mantle.

Catastrophic Volcanic Collapse: Relation to Hydrothermal Processes

Catastrophic volcanic collapse, without precursory magmatic activity, is characteristic of many volcanic disasters. The extent and locations of hydrothermal discharges at Nevado del Ruiz volcano, Colombia, suggest that at many volcanoes collapse may result from the interactions between hydrothermal fluids and the volcanic edifice. Rock dissolution and hydrothermal mineral alteration, combined with physical triggers such as earth-quakes, can produce volcanic collapse. Hot spring water compositions, residence times, and flow paths through faults were used to model potential collapse at Ruiz. Caldera dimensions, deposits, and alteration mineral volumes are consistent with parameters observed at other volcanoes.

Volcanic flux of nitrogen from the Earth

Abstract The global flux of nitrogen from subduction zones is estimated by the elemental and isotopic compositions of nitrogen, argon and helium observed in volcanic gases and hydrothermal fluids in island arcs and in back-arc basin basalt (BABB) glasses. The 3He/4He ratios of island arc samples vary from 4.7 Ratm to 7.5 Ratm, indicating a typical subduction signature. The 40Ar/36Ar ratios are consistent with atmospheric values except for a few samples. The δ15N values range from +0.1‰ to +4.6‰, which is generally higher than those of BABB glasses. Taking into account data distribution in the δ15N–N2/36Ar diagram, we distinguish three nitrogen components (mantle-derived, sedimentary and atmospheric nitrogen) for the island arc samples. Contribution of mantle-derived nitrogen is 9–30% in the samples, which is consistent with that of mantle-derived carbon. It is possible to calculate nitrogen flux based on the 3He flux in the literature and N2/3He ratios corrected for elemental fractionation. The nitrogen flux of 6.4×108 mol/year from island arc is comparable with 5.6×108 mol/year from back-arc basin, but smaller than 2.2×109 mol/year from mid-ocean ridges. In detail, island arcs show a large flux of subducted sedimentary nitrogen, while back-arc basins have a relatively small but measurable subduction component. The nitrogen flux of 4.1×106 mol/year from hot spot region is significantly small, which is consistent with the characteristic of global carbon flux from the Earth. Total volcanic flux of nitrogen amounts to 2.8×109 mol/year by taking mid-ocean ridge, hot spot and subduction values. The global nitrogen flux, if it has been constant for the 4.55 billion years of geological time, leads to an accumulation of 1.3×1019 mol in total, which is one order of magnitude smaller than 1.8×1020 mol of the present inventory of nitrogen at the Earths surface.

Global carbon dioxide emission to the atmosphere by volcanoes

Global emission of carbon dioxide by subaerial volcanoes is calculated, using CO2SO2 from volcanic gas analyses and SO2 flux, to be 34 ± 24 × 10l2gCO2yr from passive degassing and 31 ± 22 × 1012gCO2yr from eruptions. Volcanic CO2 presently represents only 0.22% of anthropogenic emissions but may have contributed to significant “greenhouse” effects at times in Earth history. Models of climate response to CO2 increases may be tested against geological data.

Fluxes and sources of volatiles discharged from kudryavy, a subduction zone volcano, Kurile Islands

Abstract The Kudryavy volcano, a 996-m-high basaltic-andesite cone on the northeastern shore of Iturup Island in the Kuriles, erupted last in 1883 and has since been in a persistent state of high-temperature, >900°C fumarolic activity. Its flux of SO2, measured by COSPEC, is 73±15 t/d, or 416 Mmol/a. In combination with the chemical composition of the parent gas supplying the high-temperature vents and the isotopic compositions of He and C, it allows the evaluation of contributions from major source components, such as the mantle, the crust, and subducted sediments and carbonate. The 3He/4He ratio of 6.7 RA corresponds to a 84% mantle origin and a flux of 2200 mol/a of mantle He. At a He concentration of 2200 mol/Mt, the mass of mantle material required to generate this flux is 1.0 Mt/a. The same mass produces a flux of 0.025 mol/a of 3He and of 50 Mmol/a of mantle CO2 at a CO2/3He ratio of 2·109. In conjunction with the C-isotopic composition of fumarolic CO2 of −7.2‰, about 12% of the CO2 are derived from the mantle, 67% from marine carbonate in subducted, altered oceanic crust, 21% are of subducted organic sedimentary origin. The flux of 280 Mmol/a of carbonate-derived CO2 requires 0.41 Mt/a of oceanic crust with a CO2 content of 3 wt%, and 0.35 Mt/a of sedimentary material to supply the organic CO2 flux of 86 Mmol/a. Nitrogen from the mantle contributes at most 2% to the total N2 flux of 5.4 Mmol/a. Assuming N to be derived from the subducted sediments, its concentration there is 460 mg/kg. The total volume of mantle and subducted material required to maintain the flux of volatiles over the 100 a period of high-temperature fumarolic activity of Kudryavy is 0.07 km3. Steady-state release of volatiles from the depth of arc magma generation to the fumaroles and continuously high heat flow from the mantle are proposed as the main process supporting the long-term high-temperature degassing at Kudryavy. In this steady-state system, the calculated volatile fluxes are balanced over time by volatiles originating from subducted sediments, hydrothermally altered oceanic crust below the Kudryavy volcano and the mantle wedge. This has significant implications for volatile cycling from the Earths crust and mantle to the atmosphere.

The chemical and isotopic composition of fumarolic gases and spring discharges from Galeras Volcano, Colombia

Abstract Galeras fumarole discharges have been collected since its reactivation, in 1988, through December 1995. The gases are dominated by H2O, CO2, S (as SO2 and H2S) and HCl. The relative proportions of these gases classify them as ‘magmatic’. Thermodynamic equilibrium temperatures of the gases range from 260 to > 600 °C. The relative abundance of inert gases, N2, Ar and He, can be used as ‘tracers’ to identify the source of the fumarole discharges. At Galeras the majority of the samples have a composition characteristic of gases originating from arc-related magmas, with relatively high N2 contents and minor He and Ar. During 1993, the year of frequent eruptions, the gas composition changed to basaltic or ‘mantle-derived’ gases, with significantly higher He contents. This is interpreted to be the result of injection of volatiles from a basaltic magma body at depth prior to and during the increased eruptive activity of 1993. The δ13C values for CO2 in fumarole discharges are typical of andesitic volcanoes and may indicate addition of MORB-derived CO2. The δ15N values for N2 may indicate significant contribution of N2 from marine sediments and only minor contribution of MORB-derived N2. The δ D and δ18O values of the discharging steam lie on a mixing trend between the isotopic composition of ‘arc-related’ magmatic water and18O-shifted meteoric water. The most magmatic discharges have δ D values of −30 to −35‰; while the most meteoric discharges have values of −70 to −75‰, similar to Galeras thermal spring waters. Galeras thermal water discharges consist of acid sulfate and bicarbonate waters.S/Cl ratios in the acid sulfate waters are similar to fumarole ratios, suggesting direct absorption of magmatic gases into shallow ground waters. This is supported by the essentially meteoric δD and δ18O values of the discharges and by elevated3He/4He ratios of thermal spring waters. The absorption of acid S- and Cl-rich gases yield acid waters which are capable of dissolving rocks. The thermal waters, however, are far from equilibrium with Galeras lavas and pyroclastic rocks, providing evidence of the immaturity of the Galeras hydrothermal system. The SO4 and Cl content, as well as the O and H isotopic composition of Galeras thermal springs vary with the activity of the volcano. The 7-year sampling program at Galeras revealed intriguing results concerning the activity of Galeras, its magmatic-hydrothermal system and the origin of the volatiles. Despite decreasing outlet temperatures since 1992, deep temperatures remain high, implying continued unrest in the Galeras magmatic system.

Geochemical surveillance of magmatic volatiles at Popocatépetl volcano, Mexico

Surveillance of Popocatepetl volcanic plume geochemistry and SO 2 flux began in early 1994 after fumarolic and seismic activity increased significantly during 1993. Volatile traps placed around the summit were collected at near-monthly intervals until the volcano erupted on December 21, 1994. Additional trap samples were obtained in early 1996 before the volcano erupted again, emplacing a small dacite dome in the summit crater. Abundances of volatile constituents (ppm/day of Cl, S total , F, CO 2 , Hg, and As) varied, but most constituents were relatively high in early and late 1994. However, ratios of these constituents to Cl were highest in mid-1994. δ 34 S-S total in trap solutions ranged from 1.5‰ to 6.4‰; lowest values generally occurred during late 1994. δ 13 C-CO 2 of trap solutions were greatly contaminated with atmospheric CO 2 and affected by absorption kinetics. When trap data are combined with SO 2 flux measurements made through November 1996, Popocatepetl released about 3.9 Mt SO 2 , 16 Mt CO 2 , 0.75 Mt HCl, 0.075 Mt HF, 260 t As, 2.6 t Hg, and roughly 200 Mt H 2 O. Near-vent gas concentrations in the volcanic plume measured by correlation spectrometer (COSPEC) and Fourier transform infrared (FTIR) commonly exceed human recommended exposure limits and may constitute a potential health hazard. Volatile geochemistry combined with petrologic observations and melt-inclusion studies show that mafic magma injection into a preexisting silicic chamber has accompanied renewed volcanism at Popocatepetl. Minor assimilation of Cretaceous wall rocks probably occurred in mid-1994.

The relationship between fumarole gas composition and eruptive activity at Galeras Volcano, Colombia

Forecasting volcanic eruptions is critical to the mitigation of hazards for the millions of people living dangerously close to active volcanoes. Volcanic gases collected over five years from Galeras Volcano, Colombia, and analyzed for chemical and isotopic composition show the effects of long-term degassing of the magma body and a gradual decline in sulfur content of the gases. In contrast, short-term (weeks), sharp variations are the precursors to explosive eruptions. Selective absorption of magmatic SO{sub 2} and HCl due to interaction with low-temperature geothermal waters allows the gas emissions to become dominated by CO{sub 2}. Absorption appears to precede an eruption because magmatic volatiles are slowed or retained by a sealing carapace, reducing the total flux of volatiles and allowing the hydrothermal volatiles to dominate gas emissions. Temporal changes in gas compositions were correlated with eruptive activity and provide new evidence bearing on the mechanism of this type of `pneumatic` explosive eruptions. 18 refs., 5 figs.

A model of degassing at Galeras Volcano, Colombia, 1988-1993

Galeras volcano reactivated in early 1988, as evidenced by increased seismicity, gas emissions, and fumarole temperatures, and had a series of explosive eruptions in May 1989. During 1989-1990 the volcano was characterized by high SO 2 flux, reaching 5000 t (metric tons)/day. This period of strong degassing was followed by upward movement of magma (July-September 1991) and the emplacement of a lava dome (October-November 1991). Since November 1991, long- period seismicity and SO 2 flux have decreased, indicating that the conduit has partially sealed itself, probably by crystallization and solidification of the magma. On July 16, 1992, the dome was destroyed by an explosion, consistent with pressurization due to sealing and degassing in a nonporous medium. Continued pressurization beneath the volcano is indicated by eruptions on January 14, March 23, April 4, April 13, and June 7, 1993. Glass inclusion and matrix glass analyses support a hypothesis of significant magma degassing at depth in the magma chamber and in the conduit. This degassing has reduced the H 2 O, S, and Cl contents of the magma, but the F content appears to have increased due to crystallization of anhydrous mineral phases and to the low fluid-melt partition coefficient for F. Despite its degassed state, Galeras is currently a dangerous volcano because the sealing process forces gas-saturated magma to release volatile components into confined spaces, thereby causing pressurization and explosive eruptions.

SO2 fluxes from Galeras Volcano, Colombia, 1989-1995: Progressive degassing and conduit obstruction of a Decade Volcano

Abstract Galeras volcano has emitted a minimum of 1.9 Tg SO 2 during the past seven-year period of re-activation from 1989 to 1995. Emissions from the volcano are almost exclusively by passive degassing and represent at least 1.4% of the annual volcanic output of SO 2 . The volcano was characterized by high SO 2 flux during 1989–1990, frequently exceeding 2000 metric tons/day. Fluxes decreased to less than 1700 tons/day in October 1990. A further decrease to less than 1000 tons/day occurred in late 1991 after emplacement of a lava dome. Currently, SO 2 fluxes are less than 500 tons/day. The general decrease of the flux with time is interpreted as: (1) progressive degassing of a single batch of magma; and (2) obstruction of the conduit by emplacement of the dome. During 1992–1993, six explosive eruptions occurred which destroyed the dome and excavated the crater. Three of the largest eruptions were immediately followed by intense long-period seismicity (up to 600 events/day) and by comparatively large SO 2 fluxes. This high seismicity and SO 2 flux are interpreted as rapid degassing of partially solidified magma that had been unroofed by the eruptions. Although the volcano is currently in a comparatively degassed state, the six recent eruptions demonstrate that there is still residual magmatic gas which is unable to be released freely. This can lead to pressurization, resulting in short, periodic eruptions or gas emissions that rapidly dissipate the pressure buildup.