V53E-01 INVITED
Volcanic halogen emissions: sources and consequences
We present an overview of our recent work concerning the sources of halogens in volcanic emissions; how inert hydrogen halides in volcanic gases become activated into environmentally reactive species, and some of the potential consequences of these species. Previous studies have suggested that the Cl in volcanic emissions is the result of efficient recycling of Cl taken down with the subducting slab. We will present new results from Nicaragua for volcanic Br and I emissions, as well as Cl. Our results are consistent with a subduction source for I. In general, the volcanic samples are more enriched in Cl than many other Earth system reservoirs. We will discuss some of the questions that this raises, and review the scale of global volcanic halogen emissions. Once emitted from the magma, volcanic halogen species may cause localised ozone depletion both in the troposphere (following activation by high-temperature mixing with atmospheric oxygen) and the stratosphere (due to their interactions with other volcanic plume components as shown in the recent studies of the Hekla plume in 2000).
V53E-02 INVITED
Volcanic Eruptions and Climate: Sulfates are More Important Than Halogens in Producing Climate Change
Large volcanic eruptions inject sulfur gases into the stratosphere, which convert to sulfate aerosols with an e- folding residence time of about one year. The radiative and chemical effects of this aerosol cloud produce responses in the climate system. Using examples from major eruptions of the past and results from experiments with numerical models of the climate system, this talk illustrates the major impacts. Volcanic eruptions produce global cooling, and are an important natural cause of interdecadal and interannual climate change. One of the most interesting volcanic effects is the "winter warming" of Northern Hemisphere continents following major tropical eruptions. During the winter in the Northern Hemisphere following every large tropical eruption of the past century, surface air temperatures over North America, Europe, and East Asia were warmer than normal, while they were colder over Greenland and the Middle East. This pattern and the coincident atmospheric circulation correspond to the positive phase of the Arctic Oscillation. High latitude eruptions in the Northern Hemisphere, while also producing global cooling, do not have the same impact on atmospheric dynamics. They weaken the Indian and African summer monsoon, and the effects can be seen in past records of flow in the Nile and Niger Rivers. In fact we can use records of the Nile River flow to provide an improved date for the Eldgjá eruption in Iceland, which we now date at 939 A.D. While halogens may have short term effects on ozone in plumes with high halogen concentrations, there is no evidence that they are important in the global climate response to volcanic eruptions.
V53E-03 INVITED
Halogen emissions during explosive activity on Mt. Etna
The gas composition emitted during an explosive eruption is controlled by several factors including gas transport; gas solubility; diffusion; and fragmentation rocesses. Measurements of gas composition during explosive activity are therefore a rich information source, providing valuable constraints on many volcanic processes. Halogens are particularly useful for tracking the dynamics of explosive activity in basaltic systems, primarily because they exsolve at shallower depths than CO2, H2O and S and are therefore strongly coupled to shallow volcanic processes. In this paper we present an overview of our observations of gas compositions during explosive eruptions on Etna since 2000, with particular emphasis on halogen emissions. Our findings suggest that diffusion-limited halogen degassing is a major factor in controlling the composition of emitted gases, and that this is strongly controlled by the degree of magma fragmentation.
V53E-04 INVITED
Fluid and Melt Inclusion Evidence for Halogen, Sulfur and Metal Contents in Magmatic Hydrothermal Systems
Information on halogen, sulfur and metal contents of active volcanic systems is readily available from analyses of volcanic gases and condensates. Similarly, fluid and melt inclusions in phenocrysts and vein material provide a sample of the melt and fluid that existed at depth in the magmatic system and may be used to constrain the geochemistry of the system at depth. Detailed studies of fluid and melt inclusions over the past several decades, combined with gas analyses in active systems, has led to a better understanding of the geochemical evolution and the processes that link the deeper geologic environment with the near surface. White Island, New Zealand, is an active andesitic-dacitic volcano associated with subduction of the Pacific plate beneath the Indian-Australian plate. Current annual fluxes to the atmosphere from this volcano are: 0.078 Mt S; 0.49 Mt CO2; 0.03 Mt Cl; 110 t Cu; >36 kg. Melt inclusions from White Island show several hundred ppm S, ~1000 ppm Cl and several hundred ppm Cu. The Cl/H2O ratio of melt inclusions is ~0.15, indicating that the magmatic gas phase can efficiently remove Cu from the melt during volatile separation. Over the 10,000 yr lifetime of the White Island volcano, it is estimated that 1 Mt of Cu has been discharged into the atmosphere. The tectonic and geochemical environment at White Island is analogous to that in which porphyry copper deposits form. Based on Cu contents of melt inclusions, it is estimated that extraction of Cu from about 5 km3 of melt at White Island could generate a porphyry copper deposit containing 250 Mt of 0.75 percent Cu ore. Fluid inclusions in fossil magmatic-hydrothermal systems such as the porphyry copper deposits are characterized by elevated S, Cl and metal contents. Early, magmatic fluids often contain several 10's of weight percent Cl, as well as up to several thousand ppm metals and up to 1 weight percent S. These fluids represent the "residue" that remains in the deeper parts of the magmatic-hydrothermal system after phase separation and loss of volatiles and metals to the atmosphere. Thus, even though large quantities of halogens, sulfur and metals are lost to the near surface environment, the largest proportion of these components remains in the deeper geosphere.
V53E-05 INVITED
Halogens and melt viscosity: an overview.
The influence of halogens on melt viscosity ranges from well-investigated, for F, through moderately-well investigated, for Cl, to reconnaissance studies for Br and I. F is the strongest fluxing agent in lavas next to water. It has been investigated and utilised in this capacity for centuries. Under hypabyssal conditions (dykes, small intrusions) fluorine, together with water can dominate melt viscosity. Its relatiively low volatility means that under volcanic conditions of degassed magma, the role of fluorine in controlling lava rheology can be dominant over all other volatiles. Cl is far less effective in influencing melt viscosity than fluorine. Nevertheless, the substantial levels of Cl in aklaine magmas suggest that it should be included in future models. The effect of Cl on melt viscosity has long be argued to generate either slight increase or slight decrease in viscosity. Recent studies indicate that this ambivalent behavior of Cl mnay be a systematic function of melt composition. The influence of Br and I on melt viscosity is at the reconnaissance stage. Nevertheless, the results so far indicate that the Br and I additions achieved so far yield melts with viscosities similar to their halogen-free equivalents. The viscosity data set, will be reviewed and discussed in the context of the relative structural roles of the halogens, their inferred influence on the viscous flow mechanisms and the implications for flow and crystallisation kinetics of magmas and lavas. Together with recent data on the relative diffusivities of F, Cl, Br and I, implications for the degassing kinetics of such systems will also be discussed.
V53E-06
Abundance, degassing and chemistry of halogens at Kilauea Volcano
The source concentration of halogens in the mantle, as well as other volatiles, is heterogeneous and related to tectonic setting. Intra-plate magmas, which give rise to ocean island basalts and are derived from a mantle plume source, are typically enriched in halogens relative to mid-oceanic ridge basalts but depleted relative to arc magmas. Fourier Transform Infra Red (FTIR) spectroscopy measurements of volcanic gases were supplemented with electron microprobe analyses of glasses in erupted products from Kilauea Volcano, Hawaii, to provide the first detailed picture of the abundance, degassing and plume chemistry of halogens at this volcano. This study reveals that typical fluxes of both HCl and HF are 50-100 t/d; Cl/S ranges from 0.01 to 0.07 and F/S from 0.01 to 0.05. SiF4 and HBr are below detection. The abundance of halogens in melt inclusions and matrix glasses in the lavas allow formulation of their degassing path prior to eruption. Kilauea volcanic gases have lower Cl/S than volcanoes in arc settings, but are similar to those measured at other volcanoes in plume settings. F/S, on the other hand, is within range of that measured at Soufriere Hills, Masaya and Etna, indicating that there is a higher degree of homogeneity in the mantle distribution of F. The relative proportion of Cl, F and S in the volcanic plume varies considerably over time and space. This variability may be due to a number of factors: the depth of vapour-melt separation in the conduit; the degree of degassing of shallow, stagnant magma (S-rich gases exsolve first); interaction between vapour and a hydrothermal system; the relative rates of diffusion between the species; or processes occurring in the plume such as adsorption of or reaction between acidic gases and silicate particles.
V53E-07
Chlorine Isotope Geochemistry as a Monitor of Fluid-Rock Interaction in Volcanic Systems
Near-surface interaction of aqueous fluid with ascending magma may cause phreatomagmatic eruptions, enhanced stoping and fracturing leading to rapid ascent of magma and, in some cases, the formation of epithermal mineral deposits. Many of the conservative tracers commonly used in hydrology, such as stable isotope and chlorine or bromine chemistry, fall apart in the higher temperature systems, where phase separation associated with boiling modifies a fluid's chemistry. Chlorine isotope geochemistry provides a novel technique for tracing fluid-rock interaction in volcanic systems. The value of fumaroles from the Central American arc cover a range of over 16 ‰ (Barnes et al., this meeting). The remarkable spread is probably caused by partitioning of HCl gas from an acidic solution. Measured equilibrium fractionation between HCl(g) and acidic chloride solutions are 1.57 ‰ (at 70°C), in excellent agreement with theoretical estimates (Schauble et al., 2003). However, when a kinetic 'flow-through' experiment is conducted, fractionations as high as 8 ‰ occur, related to a continuous distillation process, in which 35Cl-enriched HCl preferentially dissolves in the aqueous condensate along the flow path. The similarly high δ37Cl values found in a number of Central American volcanic fumaroles are explained by continual vaporization and condensation in the near-surface plumbing system. The low δ37Cl values (down to -5‰) found in other volcanoes are more difficult to explain. One possibility is that the gases represent the residue of extensive volatilization, whereby heavy Cl has left the system, lowering the δ37Cl value of the remaining fluid. In this case, the system must be essentially closed, so that no reintroduction of chloride-rich waters occurs. In contrast, all high δ37Cl value fumaroles are associated with large volcanic lakes, where a more-or-less infinite reservoir of Cl exists. The temporally constant high values indicate a high degree of interconnectivity in these systems. The very different delta values constrain the plumbing systems of each volcano.
V53E-08
The Dramatic Effects of C-S-O-H-Cl on the Melt-Fluid Partitioning of Cl and the Challenge of Accurately Modeling Cl Concentrations of Evolving Magmatic Fluids
New experimental constraints on the distribution of Cl between Vesuvius phonolite melt and H2O-, CO2-, SO2-, and Cl-bearing vapor, vapor plus saline liquid, or saline liquid have been determined at 200 MPa and 900°C. Some experiments involve melts saturated in all four volatile components. The addition of small quantities of SO2, CO2, and SO2 plus CO2 to fluids dominated by water and alkali chlorides causes dramatic reductions in DXCl (mole fraction Cl in fluid(s)/mole fraction Cl in silicate melt). Experiments with XCO2fluid of 0.15 or XSO2fluid of 0.15 are characterized by values of DXCl that are an order of magnitude lower than those of S- and C-free runs. This observation has important consequences for open systems that exsolve CO2- and/or SO2-enriched fluids "early" in the chemical differentiation of magma. Extrapolation of our experimental results to other pressures and for other melt compositions indicates that CO2- and/or SO2-bearing fluids will not sequester significant abundances of Cl from magma. Thus, CO2- and/or SO2-enriched fluids that exsolve and escape "early" from magma will not dramatically alter the magmatic Cl content. This is consistent with the oft-quoted, general degassing order of: C (first), S, Cl, and H2O (last) associated with decreasing pressure. To apply these new partitioning data to models of the exsolution and chemical evolution of magmatic fluids, they must be integrated with experimental constraints on other parameters that also strongly influence the distribution of Cl between melts and fluids (e.g., pressure, temperature, melt composition, and the Cl content of the bulk system). Expressed as DXCl, published values for Cl partitioning range from 20 (with felsic melts) to ca. 0.4 (with mafic melts), but DXCl is a complex function of these parameters. For example, DXCl increases by an order of magnitude as temperature decreases from 1000° to 800°C with vapor-saturated felsic melts at 200 MPa. Conversely, DXCl decreases by an order of magnitude in experiments involving a pressure reduction of 280 to 80 MPa with phonolite melt saturated in Cl-bearing aqueous fluid(s), but other experiments show that DXCl increases with decreasing pressure. DXCl also varies positively with the Cl concentration of the bulk system. In addition, the chemical evolution of fluid-saturated magma from basaltic to rhyolitic melt compositions also involves an order of magnitude increase in DXCl with all other parameters equal. In summary, as magmas ascend through the crust, cool, and differentiate, values of DXCl (and hence the Cl content of coexisting fluids) tend to increase due to the attendant reduction in temperature, because of the evolution of magma to increasingly felsic compositions, and due to the potential loss of CO2 and/or SO2 from magma to fluids that exsolve and escape "early". Conversely, values of DXCl (and the Cl content of coexisting fluids) may also tend to decrease because of the reduction in pressure during ascent. Moreover, the crystallization of Cl-free and Cl-poor minerals will tend to increase the Cl content of the magma and increase values of DXCl; whereas, the loss of Cl to C- and S-poor aqueous fluid(s) in open magmatic systems will reduce the Cl content of the system and tend to reduce DXCl during late-stage fluid-melt interactions.