Exotic Fluids in the Exosphere: When Hades Meets Apollo

Johan C. Varekamp, Wesleyan University, Middletown, Conn., and Gary L. Rowe, U.S. Geological Survey, Water Resources Division, Columbus, Ohio

The interaction of magmatic volatiles with lakes and shallow groundwater systems results in some of the most exotic fluids found at the Earth's surface. Geologic environments where such fluids occur and the chemical and physical processes responsible for their formation were the subject of a multidisciplinary Chapman Conference titled "Crater Lakes, Terrestrial Degassing and Hyper-acid Fluids in the Environment," held from September 4 to 9, 1996, at Crater Lake, Ore.

Sixty-eight scientists with backgrounds in geochemistry, hydrology, limnology, microbiology, economic geology, and volcanology attended the conference. Their presentations covered the physical and chemical characteristics of volcanic lakes, the geochemistry and microbiology of hyper-acidic (pH<<1) fluids found in crater lakes, geothermal and acid-mine drainage environments, the rates and mechanisms of magmatic degassing in near-surface environments, and acid fluid-rock reactions and their role in alteration and mineralization associated with epithermal ore deposits.

Presentations at this Chapman Conference illustrated the wide compositional spectrum of volcanic lakes, from extremely diluted lakes like Crater Lake, Oregon, to acidic brine lakes that receive inputs of heat and volatiles primarily from underlying magma bodies, and thus are subject to full-scale eruptions, such as Ruapehu in New Zealand (see Eos, May 14, 1996, p. 189). Volcanic lakes can be dominated by meteoric waters ("Apollan lakes") or by volcanic and hydrothermal inputs ("Hadean lakes"). Energy budget analysis of volcanic lakes was presented as a way of creating a quantitative classification system to replace such mythological analogs.

The meeting marked the 10th anniversary of the Lake Nyos disaster, in which over 1700 people were killed when a cloud of cold CO₂ gas was ejected from a stratified lake in Cameroon. Few disagree that the CO₂ in the Nyos lake water has a deep magmatic origin, but what actually triggered the 1986 catastrophic gas burst remains poorly understood. The CO₂ may be introduced into the lake by advection for example, by discharging CO₂-rich waters at depth or alternatively, by diffusing CO₂ through bottom sediments. Depth / CO₂ profiles presented at the meeting indicate that Lake Nyos is again stratified and that CO₂ is accumulating in the lake's bottom waters. Calculated CO₂ accumulation rates at Lake Nyos indicate that CO₂ oversaturation could occur there in less than 30 years.

A gas-lift system to extract CO₂ from the depths of such lakes has been successfully demonstrated, but the logistics of large-scale applications are quite complex. Cold, magmatic CO₂ releases also occur in volcanic and geothermal areas not covered by lakes, such as Dieng, Indonesia, and Mammoth Mountain, California. Results of recent research at Mammoth Mountain, California, indicate that diffuse degassing of large amounts of CO₂ followed the intrusion of a shallow dike in 1989. Tree rings document the apparent ages and the stable carbon isotopic composition of past CO₂ releases at Mammoth Mountain.

The placid surfaces of Nyos-type volcanic lakes contrast starkly with turbulent, sulfur-rich, acid-brine crater lakes (Figure 1). Long-term monitoring of such lakes has shown that chemical and physical properties of these lakes change in response to changing volcanic activity. Exciting efforts are now underway to develop reliable real-time surveillance techniques, including monitoring of acoustic noise levels and in situ determinations of polythionate (S_xO₆₌) concentrations in lake water. Seepage and discharge of toxic and hyper-acidic fluids from high-level crater lakes into flank watersheds may result in strongly contaminated surface waters, an environmental hazard that deserves closer scrutiny. Circulation of such hyper-acidic fluids in volcanic flank aquifers also may affect the long-term stability of the volcanic edifice.

Fig. 1. The turquoise crater lakes of Keli Mutu, Flores, Indonesia. The lake with the floating sulfur slick (at left) has a pH of about 0.3.

Studying hyper-acidic fluids with several weight percent total dissolved solids poses problems. How can we measure and interpret pH when hydronium ion concentrations exceed 1 molal, and how can we model chemical speciation, mineral solubilities, and water-rock reactions involving such fluids? A method was presented at the meeting for calibrating and accurately measuring negative pH fluids that employs the specific-ion interaction approach developed by Pitzer. Thermodynamic models that use the Pitzer approach can be used to evaluate the speciation and saturation state of such brines, whether it is from acidic mine drainage or warm crater-lake brines. Microbial populations in hyper-acidic mine waters, which may play a role in the weathering of sulfide minerals and precipitation of secondary phases, can be characterized by ribosomal DNA sequencing. Results of such work at Iron Mountain, near Redding, Calif., which carries some of the most acidic waters found to date, indicate the presence of a variety of microorganisms that are only slightly related to common genera found in less acidic waters.

Synchrotron X ray absorption spectroscopy can provide information on trace metal partitioning between acidic waters, soils, and alteration minerals, and one can then evaluate metal sorption versus coprecipitation in soils and alteration assemblages exposed to metal-rich acidic fluids. Hyper-acidic fluids are also commonly found in volcano-hosted geothermal systems, and knowledge of their origin and corrosive properties is greatly important to the geothermal industry. Meeting participants discussed theoretical models that are now able to predict the origin of such HCl-rich, acidic fluids quite well. Some acidic fluids create their own sanctuaries, as in Lake Armyansk in the Crimean desert, where industrial waste acid has been discharged for several decades. Evaporation in the desert environment results in hyper-acidic lake water that reacts with the clay-rich carbonate bedrock to form a self-repairing, naturally impermeable seal on the lake bottom that is primarily composed of ferrihydrite, gypsum, and jarosite. This example may provide a natural analog for storage and treatment of other acidic waste fluids.

Correlation spectrometer measurements are the principal tools for estimating volcanic sulfur fluxes, but such data can also be derived from the physical and chemical changes in crater lakes over time. Such measurements reveal that many volcanoes emit sulfur in quantities that greatly exceed the amount dissolved in erupted magma. The sources of this "excess sulfur" include the remobilization of hydrothermal sulfur and/or degassing of unerupted magma. Sulfur-rich vapor can be extracted from large volumes of magma by vapor diffusion, bubble transport, or, more likely, through magma convection with vapor extraction at shallow levels. The scrubbing action of high-level hydrothermal fluids may absorb significant quantities of magmatic SO₂ during the earliest stages of magma ascent and degassing, which has implications for eruption forecasting. Estimates of sulfur fluxes from explosive eruptions should also take into account sulfur adsorbed onto fine-grained ash, which may be, in some cases, half the total amount of sulfur erupted.

The chemical and isotopic characteristics of hyper-acidic fluids and alteration products of acid-fluid/rock interactions provide insight into the sources and processes responsible for generating hyper-acidic fluids and associated epithermal ore deposits. Analysis of d³⁷Cl in crater-lake waters may allow us to determine if the Cl degassed from arc volcanos is derived by reflux from surface reservoirs or from the mantle. Other tracers like ¹²⁹I and ¹¹B were discussed as tools for discriminating between volatiles derived from the mantle and those from recycled subducted material. Stable isotope ratios of sulfur and oxygen in dissolved sulfate in volcanic and acid-mine waters can be used to infer levels of magmatic activity in crater-lake systems and sulfide oxidation mechanisms in acidic mine waters.

Several decades ago a conceptual problem of ore genesis was understanding how such massive amounts of metals could be transported in hydrothermal fluids within geologically reasonable timeframes. This question is now gradually being replaced by the notion that many volcano-hosted ore deposits may form over geologically short time spans (for example, 10³ - 10⁵ years). Hyper-acidic fluids with high concentrations of ore-forming metals- for example, several mg/L of Zn and Pb are common in many crater-lake brines, and those may serve as field laboratories for studying the genesis of ore deposits. How are these metals precipitated from the brines? In some Japanese crater lakes, the neutralization of hyper-acidic waters by water-rock reaction and mixing with neutral-pH surface waters has led large quantities of gold and arsenic to be deposited in lake bottom sediments.

After the meeting, some participants traveled to northern California to observe acidic fluids produced by geothermal activity at Lassen Volcanic Park, while others examined hyper-acidic mine drainage at Iron Mountain. The U. S. Environmental Protection agency spends millions of dollars annually to characterize and remediate sites affected by hyper-acidic fluids. Discussion at this conference strengthened the conclusion that further integration of studies of hyper-acidic fluids in both natural and disturbed settings is a worthy goal of continued research.

Acknowledgments: We thank the U. S. Department of Energy, the U. S. Geological Survey, and the National Science Foundation for sponsoring this Chapman Conference. Support was also provided by the National Park Service. Special thanks are extended to the Crater Lake National Park staff, to Charlie Bacon for showing the geology of the park, and to Bob Collier for his coordination work.

During the writing of this report we were saddened to learn that Rick Hutchinson, a contributor to this meeting and National Park Service researcher at Yellowstone National Park, died there in a snow avalanche in early March 1997.

Back to Science for Everyone