Magma-hydrothermal system interaction inferred from volcanic gas measurements obtained during 2003–2008 at Meakandake volcano,Hokkaido, Japan

Phreatic eruptions occurred at the Meakandake volcano in 1988, 1996, 1998, 2006, and 2008. We conducted geochemical surveillance that included measurements of temperature, SO₂ emission rates, and volcanic gas composition from 2003 to 2008 at the Nakamachineshiri (NM), Northwest (NW), and Akanuma (AK) fumarolic areas, and the 96–1 vent, where historical eruptions had occurred. The elemental compositions of the gases discharged from the different areas are similar compared with the large variations observed in volcanic gases discharged from subduction zones. All the gases showed high apparent equilibrium temperatures, suggesting that all these gases originated from a common magmatic gas. The gases discharged from each area also exhibited different characteristics, which are probably the results of differences in the conditions of meteoric water mixing, quenching of chemical reactions, and vapor-liquid separation. The highest apparent equilibrium temperatures (about 500°C) were observed in the case of NW fumarolic gases, despite the low outlet temperature of about 100°C at these fumaroles. Since the NW fumaroles were formed as a result of the 2006 phreatic eruption, the high-temperature gas supply to the NW fumarole suggests that the phreatic eruption was caused by the ascent of high-temperature magmatic gases. The temperatures, compositions, and emission rates of the NM and 96–1 gases did not show any appreciable change after the 2006 eruption, indicating that each fumarolic system had a separate magmatic-hydrothermal system. The temperatures, compositions, and emission rates of the NM fumarolic gases were apparently constant, and these fumaroles are inferred to be formed by the evaporation of a hydrothermal system with a constant temperature of about 300°C. The 96–1 gas compositions showed large changes during continuous temperature decrease from 390° to 190°C occurred from 2003 to 2008, but the sulfur gas emission rates were almost constant at about four tons/day. At the 96–1 vent, the SO₂/H₂S ratio decreased, while the H₂/H₂O ratio remained almost constant; this was probably caused by the rock-buffer controlled chemical reaction during the temperature decrease.     

Surface alteration of basalt due to cation-migration

Basaltic lava from Kilauea, Hawaii may have a red-brown surface, indicative of Fe-(hydr)oxides. This surface is not found where exposed to weathering, but at the interface between lava lobes, or in the interior of lava channels. We use several analytical techniques to determine how these Fe-(hydr)oxide surfaces may have developed. WDS-elemental distribution line profiles from the lava surface towards the lava′s interior detect an Fe-rich film of less than 5 μm thickness. Heat treatment of quenched, fresh lava samples of the same chemical composition between 600–1,090°C helps to replicate temperatures under which such an Fe-rich film might have formed. These experiments suggest that Fe-enrichment occurs above 1,020°C, whereas at lower temperatures Ca is enriched relative to Fe. One sample was treated below the glass transition temperature, at 600°C for 164 h. A depth profile with secondary neutral mass spectrometry shows an enrichment of Mg at the outer 50 nm of the glass surface. The formation of films requires cation migration, which is driven by an oxygen chemical potential between air and the reduced basalt (Fe²⁺/Fe³⁺ ratio of 13.3). The change of surface alteration from Mg to Ca film at lower temperatures, to predominantly Fe at high temperatures, is determined by a change of cation availability, largely controlled by crystallization that already occurs below 850°C, and volume crystallization that occurs above 925°C.     

Restoration of new volcanic gas analyses from basalts of the afar region: Further evidence of CO2 -degassing trends

The restored compositions for approximately 70 new analyses reported recently for Erta'Alelava lake (LeGuern et al., 1979) are in good agreement with restored compositions (Gerlach, 1980a) based on previously published data. The results confirm earlier indications that gas collections taken at different times from the lava lake are related principally by variations in CO₂ content. Restored compositions for gas samples collected in the final stages of a November 1978 Ardoukoba eruption along the Asal Rift spreading axis resemble the Erta'Ale gases except for a much lower CO₂ content. The Ardoukoba gases fall close to a CO₂-decreasing control line for gases with initial compositions similar to the 1971–1973 Erta'Ale gases. These results suggest that gases released from basaltic lava along zones of crustal spreading follow compositional trends dominated by changes in CO₂ content.     

Equilibrium calculations in volcanic gaseous systems

Equilibria calculations of high-temperature volcanic gases from lava lakes are carried out on the basis of best volcanic gas samples. The equilibrium gas composition at temperatures from 800° to 1400°K and pressures up to 25 kilobars (in ideal gas system) was calculated using the free energy minimization model as well as the Newton-Raphson methods. It is shown that the juvenile «magmatic gas » of basaltic magma consists of three components: H₂O, SO₂, CO₂; the water vapor being about 60%. The increase of temperature under constant pressure results in the increase of the SO₂ concentration and in the simultaneous decrease of H₂S. Under the same conditions the ratios CO/CO₂ and H₂/H₂O are found to increase. Methane cannot be a component of «magmatic gas» corresponding to the elemental composition of basaltic lava gases. The calculated values of \(P_{O_2 } \) are in good agreement with the experimental data obtained from direct measurements of \(P_{O_2 } \) in lava lakes and experiments with basaltic melts.     

Mechanism of energy transfer in the lava lake of Niragongo (Zaire), 1959–1977

Heat and mass transfer rates were studied at the Niragongo lava lake during two expeditions directed by H. Tazieff in 1959 and 1972. The results of this study are as follows:Heat is transferred to the surface of the lake by the movement of lava; gas discharge is a result and not the cause of convection. The chemical composition of the gases and magma has changed very little between 1959 and 1972, whereas the mass and energy outputs differ by an order of magnitude. In 1977 a catastrophic explosion seems to have been caused by tectonic factors, stopping the slow convection of magma under the volcano and hence reducing surface manifestations in the form of the lava lake and escaping fumarolic and magmatic gases. The gas discharge was, in tons day⁻¹, 5000 for H₂O, 11,000 for CO₂, 1000 for SO₂ in 1959, and in 1972 7700 for H₂O, 180,000 for CO₂ and 23,000 for SO₂. These values correspond to an energy transfer of 0.9 × 10⁹ W in 1959 and 16 × 10⁹ W in 1972.     

Crystallization of tholeiitic basalt in Alae Lava Lake,Hawaii

The eruption of Kilauea Volcano August 21–23, 1963, left 600,000 cubic meters of basaltic lava in a lava lake as much as 15 meters deep in Alae pit crater. Field studies of the lake began August 27 and include repeated core drilling, measurements of temperature in the crust and melt, and precise level surveys of the lake surface. The last interstitial melt in the lake solidified late in September 1964; by mid August 1965 the maximum temperature was 690°C at a depth of 11.5 meters. Pumice air-quenched from about 1140°C contains only 5 percent crystals — clinopyroxene, cuhedral olivine (Fo ₈₀), and a trace of plagioclase, (An ₇₀). Drill cores taken from the zone of crystallization in the lake show that olivine continued crystallizing to about 1070°C; below that it reacts with the melt, becoming corroded and mantled by pyroxene and plagioclase. Below 1070°C, pyroxene and plagioclase crystallized at a constant ratio. Ilmenite first appeared at about 1070°C and was joined by magnetite at about 1050°C; both increased rapidly in abundance to 1000°C. Apatite first appeared as minute needles in interstitial glass at 1000°C. Both the abundance and index of refraction of glass quenched from melt decreased nearly linearly with falling temperature. At 1070°C the quenched lava contains about 65 percent dark-brown glass with an index of 1.61; at 980°C it contains about 8 percent colorless glass with an index of 1.49. Below 980°C, the percentage of glass remained constant. Progressive crystallization forced exsolution of gases from the melt fraction; these formed vesicles and angular pores, causing expansion of the crystallizing lava and lifting the surface of the central part of the lake an average of 19.5 cm. The solidified basalt underwent pneumatolitic alteration, including deposition of cristobalite at 800°C, reddish alteration of olivine at 700°C, tarnishing of ilmenite at 550°C, deposition of anhydrite at 250°C, and deposition of native sulfur at 100°C. Ferric-ferrous ratios suggest that oxidation with maximum intensity between 550°C and 610°C moved downward in the crust as it cooled; this was followed by reduction at a temperature of about 100°C. The crystallized basalt is a homogeneous fine-grained rock containing on the average 48.3 percent by volume intergranular pyroxene (augite > pigeonite), 34.2 percent plagioclase laths (An_{60 70}), 7.9 percent interstitial glass, 6.9 percent opaques (ilmenite > magnetite), 2.7 percent olivine (Fo_{70 80}), and a trace of apatite. Chemical analyses of 18 samples, ranging from initially quenched pumice to lava cored more than a year after the eruption from the center and from near the base of the lake, show little variation from silica-saturated tholeiitic basalt containing 50.4 percent SiO₂, 2.4 percent Na₂O, and 0.54 percent K₂O. Apparently there was no significant crystal settling and no appreciable vapor-phase transport of these components during the year of crystallization. However, seven samples of interstitial liquid that had been filter-pressed into gash fractures and drill holes from partly crystalline mush near the base of the crust show large differences from the bulk composition of the solidified crust—lower MgO, CaO, and Al₂O₃; and higher total iron, TiO₂, Na₂O, K₂O, P₂O₅, and F, and, in most samples, SiO₂. The minor elements Ba, Ga, Li, Y, and Yb and possibly Cu tend to be enriched in the filter-pressed liquids, and Cr and possibly Ni tend to be depleted.     

Magnetic studies on Kilauea Iki lava lake,Hawaii

Ground surveys made during August, 1961, show large vertical magnetic intensity anomalies associated with the partly lava filled crater of Kilauea Iki. A vertical magnetic variation of 11,600 gammas occurs along a north-south profile across the crater, the maximum being on the north rim of the crater and the minimum on the south edge of the encrusted lava lake below the south rim. An east-west profile shows less vertical magnetic variation, with lake-surface measurements 1500 to 2500 gammas lower than measurements on the east rim of the crater. Computed anomalies using two-dimensional potential field graticules are in good agreement with the observed anomalies and support the following conclusions: 1) Average measured values of remanent magnetization of 10^?2 cgs units and susceptibilities of 10^?3 cgs units give reasonable magnitudes to the computed anomalies. 2) The remanent magnetization is parallel to the earth’s present magnetic field. 3) The maximum vertical magnetic field value in the north-south profile is the result of reinforcement of the positive terrain effect of the north rim of the crater and the positive edge effect of the north side of the lava lake. 4) The minimum value in the same profile is the result of reinforcement of the negative terrain effect at the base of the south rim of the crater and the negative edge effect of the south side of the lava lake. 5) Variation in the east-west magnetic profile is less because the terrain and edge effects of the horizontal components of the earth’s magnetic field and remanent magnetization approach zero. Changes in vertical magnetic field values as the lake solidifies will be maximum at the north edge of the lava lake, but more consistent changes of the opposite sign will occur on the south side of the lava lake. Total change will be approximately + 2300 gammas between the August 1961 measurement at station S6 and the value at that point when the entire lava lake has cooled below 400° C. The maximum rate of change at station S6 will occur when the 500° C isotherm is 35 to 65 meters below the surface and will be about 28 gammas per meter of lowering of the 500°C surface. Because of the steep magnetic anomalies associated with the lava lake and crater rims, the permanent magnetization presently forming in the cooling lake crust will have inclinations as much as 12° less than the average 37.5° inclination in the Kilauea area.     

Die quantitative Erfassung von Methan im Seewasser

A preparation method for the quantitative analysis of methane in lake water samples has been developed. The method is based on the equilibrium reactions of gases in a closed two-phase (headspace-water) system, in which the headspace gas is analyzed gaschromatographically. The method shows a standard deviation of ±5%, including sampling and sample preparation and can also be used for the quantitative determination of other dissolved gases.      

Chemistry of the crater lake during the 1971–1972 Soufrière eruption

During the 1971–1972 eruption of Soufrière volcano on St. Vincent Island, a lava mass was extruded subaqueously in the crater lake. An investigation of the chemistry of the lake indicates that over 50,000 tons of dissolved solids were taken into solution during the eruption, in addition to 9000 tons of iron precipitated as ferric oxide in syngenetic metalliferous sediments on the crater floor. Leaching of hot disintegrating lava and volcanic glass is the principal source of cations dissolved in the lake (Na, Ca, Mg, Si and K), whereas chlorine and sulfur were introduced during injection of acid volcanic gases from the submerged lava mass. Concentrations of the common cations in the lake are not affected by mineral solubility, except in the case of Fe³⁺, but rather by the rate of leaching, evaporation, and water-rock reactions. Variations in Cl/Na, total Cl and acidity have aided in identification of distinct fumarolic phases during the eruption, which may correlate with observed increase in frequency of minor volcanic tremors in the crater. Accumulation of ferric oxide in sediments on the crater floor is thought to be due to leaching of ferrous iron at high temperature from the lava mass, followed by oxidation and precipitation of hematite in the cooler lake.     

Halogen acids in fumarolic gases of Kilauea volcano

The water-rich condensates of fumarolic gases, obtained from degassing lavas of the 1959–60 eruption of Kilauea volcano, contain unexpectedly high concentrations of hydrofluoric and hydrochloric acids, and thereby suggest that halogens are significant constituents of basaltic magmas. Vents on the pumice hill of Kilauea Iki yielded one sample that contained, in parts per million, 21,000 HF (1.1 N) with 2,920 HCl, and another sample, 20 HF with 70,500 HCl (2.0 N). Samples from vents elsewhere on the volcano had from one-fortieth to one-thousandth as much of the two acids. A rough correlation exists between the temperature of the fumaroles (range 110 to 820°C) and the total concentration of the halogen acids. This correlation is mainly due to progressive dilution of the magmatic halogen acids by water of probable meteoric origin in fumaroles of lower temperatures. Variations in the HF/HCl ratio (range 0.0003 to 7.2) may be explained by means of two different processes whose relative importance cannot be assessed with the data at hand (1). In their migration to the surface, the acid gases may have reacted with the lava to a variable extent owing to the widely different configurations of the several vents (steaming areas in glassy pumice, glowing cracks, and drillhole in lava lake). In the reaction, relatively more HF could have reacted at temperatures around 300°C with the glassy pumice (2). There is some indication that the HF/HCl ratio increases with time,suggesting that the crystallizing lava may have released HCl early, with HF concentrated in the later exhalations. The Br/Cl ratio ranges from 0.0036 down to 0.0014, as compared to 0.0034 of seawater.     

Erta'ale lava lake: heat and gas transfer to the atmosphere

Data on uncontaminated samples of volcanic gases can be counted on the fingers of one hand, yet estimation of total volcanic gas flow cannot be made without such data. In this paper the flux of gas from the lava lake to the atmosphere is calculated by a heat budget based on the excess heat loss caused by combustion of H₂ and CO and by the mass rate of loss of other gases on the basis of their ratios to H₂ and CO in the unoxidized gas samples. The estimated rates of loss of H₂O, CO₂, SO₂ and HCl are consistent with the rate of loss of heat if this heat is generated by crystallization and if the initial magma contains concentrations of gas appropriate for submarine basalt from oceanic ridges. The moderate activity of permanent degassing from the two active lava ponds studied gives a lower flux than that of other volcanoes.     

Gas composition of Popocatépetl Volcano between 2007 and 2008: FTIR spectroscopic measurements of an explosive event and during quiescent degassing

On December 1, 2007, the solar absorption infrared spectra of the Popocatépetl volcanic plume was recorded during an eruptive event and complementarily on November 17, 2008, the passive quiescent degassing was measured from the same site. A portable FTIR spectrometer with a scanning mirror for fast tracking of the sun provided the flexibility, quality, and simplicity needed for field deployment. Slant columns of the gases SO₂, HCl, HF, and SiF₄ were retrieved and strong differences could be observed when comparing gas ratios in both time periods. During the explosive eruption, the SO₂/HCl ratio was three times greater and the HF/HCl ratio was slightly smaller than during passive degassing.While the ratios among SO₂, HCl, HF, and SiF₄ describe the chemical composition of the volcanic gas mixture, the SiF₄/HF ratio provides information about the equilibrium temperatures of the stored gases which in this study were calculated at 150° and 185 °C for the explosive and quiescent degassing episodes, respectively. We conclude that cooling of lava domes in the crater precedes Vulcanian explosions as suggested by Schaaf et al (2005). Based on SO₂ flux (Grutter et al., 2008) and measurements and data from the November 2008 event, the average fluxes for HCl, HF, SiF₄, and F through quiescent degassing are estimated to be 204, 22.7, 9.8, and 31.7 tons/day, respectively. These values are similar to those reported by Love et al. (1998) more than 10 yrs ago.     

Hydrogen emissions from Erebus volcano, Antarctica

The continuous measurement of molecular hydrogen (H₂) emissions from passively degassing volcanoes has recently been made possible using a new generation of low-cost electrochemical sensors. We have used such sensors to measure H₂, along with SO₂, H₂O and CO₂, in the gas and aerosol plume emitted from the phonolite lava lake at Erebus volcano, Antarctica. The measurements were made at the crater rim between December 2010 and January 2011. Combined with measurements of the long-term SO₂ emission rate for Erebus, they indicate a characteristic H₂ flux of 0.03?kg s^–1 (2.8?Mg? day^–1). The observed H₂ content in the plume is consistent with previous estimates of redox conditions in the lava lake inferred from mineral compositions and the observed CO₂/CO ratio in the gas plume (～0.9 log units below the quartz–fayalite–magnetite buffer). These measurements suggest that H₂ does not combust at the surface of the lake, and that H₂ is kinetically inert in the gas/aerosol plume, retaining the signature of the high-temperature chemical equilibrium reached in the lava lake. We also observe a cyclical variation in the H₂/SO₂ ratio with a period of ～10?min. These cycles correspond to oscillatory patterns of surface motion of the lava lake that have been interpreted as signs of a pulsatory magma supply at the top of the magmatic conduit.     

Field measurements of volcanic gases. Vulcano Island (Italy), Kilauea (Hawaii,U.S.A.), Merapi (Java,Indonesia)

A field gas chromatograph, built in 1978, was used in the field to directly analyse volcanic gases before water vapor condensation. Tested in Vulcano (Italy), Kilauea (Hawaii) and Merapi (Indonesia), this field measurement technique provides the actual composition of the volcanic gas mixture. The technique avoids the depletion of sulfur gases and the dissolution of the acid gases in the condensed water during the cooling. Thus the mixture of H₂S and SO₂ in fumarolic and high temperature gases (up to 819°C) in equilibrium at the emission temperature was examined.     

Gas analyses from the Pu'u O'o eruption in 1985, Kilauea volcano,Hawaii

Volcanic gas samples were collected from July to November 1985 from a lava pond in the main eruptive conduit of Pu'u O'o from a 2-week-long fissure eruption and from a minor flank eruption of Pu'u O'o. The molecular composition of these gases is consistent with thermodynamic equilibrium at a temperature slightly less than measured lava temperatures. Comparison of these samples with previous gas samples shows that the composition of volatiles in the magma has remained constant over the 3-year course of this episodic east rift eruption of Kilauea volcano. The uniformly carbon depleted nature of these gases is consistent with previous suggestions that all east rift eruptive magmas degas during prior storage in the shallow summit reservoir of Kilauea. Minor compositional variations within these gas collections are attributed to the kinetics of the magma degassing process.     

On the distribution of noble gases in Allende: a differential oxidation study

We have developed a closed-system combustion technique and utilized it to progressively oxidize a gas-rich, highly carbonaceous acid residue and a fine-grained (<4 μm) matrix sample from the Allende C3V meteorite and analyze the released gases mass spectrometrically. For the residue complete gas mobilization occurs at temperatures below 600°C. The temperature interval over which most of the gases are released coincides with that for combustion of most of the carbon. Release is primarily due to chemical attack rather than thermal activation of the gases. There are somewhat different oxidation thresholds for the heavy gases (Ar, Kr, and Xe) and the light gases (He, Ne), indicating chemically different sites for the two groups. Relative enhancement of isotopically anomalous components near 600°C is as large as in any open-system oxidation method. Differential combustion of the matrix sample reveals three distinct outgassing peaks, the first matching the release from the carbonaceous residue (“combustibles”), the second attributed to sulfides, and the third tentatively assigned to silicates. They comprise about 53%, 7% and 40% of the total heavy gases respectively. While the “sulfides” exhibit a small fission-like component, the Xe in the “silicates” appears isotopically uniform with roughly AVCC composition. The “combustibles” of the matrix contain relatively less excess¹²⁹Xe than the residue, perhaps indicating that ～10% of the total¹²⁹Xe in the residue was acquired from “silicates” by redistribution during acid treatment. We cannot rule out the possibility that planetary gases assigned to “sulfides” or “silicates” actually reside in carbonaceous phases somehow sheltered within soluble mineral assemblages, or in acid-soluble carbonaceous phases resistant to oxidation. Integrated releases below and above 600°C during the matrix combustion exhibit virtually identical heavy gas elemental composition, implying similar fractionation during planetary gas entrapment in various materials or in the same material in various environments.     

Evaluation of volcanic gas analysis from surtsey volcano, iceland 1964–1967

Methods used previously to remove compositional modifications from volcanic gas analyses for Mount Etna and Erta'Ale lava lake have bean employed to estimate the gas phase composition at Nyiragongo lava lake, based on samples obtained in 1959. H₂O data were not reported in 11 of the 13 original analyses. The restoration methods have been used to estimate the H₂O contents of the samples and to correct the analyses for atmospheric contamination, loss of sulfur and for pre- and pest-collection oxidation of H₂S, S₂, and H₂. The estimated gas compositions are relatively CO₂-rich, low in total sulfur and reduced. They contain approximately 35–50% CO₂ 45–55% H₂O, 1–2% SO₂, 1–2% H₂., 2–3% CO, 1.5–2.5% H₂S, 0.5% S₂ and 0.1% COS over,he collection temperature range 102° to 960° C. The oxygen fugacities of the gases are consistently about half an order of magnitude below quartz-magnetite-fayalite. The low total sulfur content and resulting low atomic S/C of the Nyiragongo gases appear to be related to the relatively low f_O2 of the crystallizing lava. At temperatures above 800°C and pressures of 1–1.5 k bar, the Nyiragongo gas compositions resemble those observed in primary fluid inclusions believed to have formed at similar temperatures and pressures in nephelines of intrusive alkaline rocks. Cooling to 300°C, with f_O2 buffered by the rock, results in gas compositions very rich in CH₄ (50–70%) and resembling secondary fluid inclusions formed at 200–500°C in alkaline rocks. Below 600°C the gases become supersaturated in carbon as graphite. These inferences are corroborated by several reports of hydrocarbons in plutonic alkaline rocks, and by the presence of CH₄-rich waters in Lake Kivu — a lake on the flanks of Nyiragongo volcano.     

Geochemistry of volcanic and hydrothermal gases of Mutnovsky volcano,Kamchatka: evidence for mantle,slab and atmosphere contributions to fluids of a typical arc volcano

We report chemical compositions (major and trace components including light hydrocarbons), hydrogen, oxygen, helium and nitrogen isotope ratios of volcanic and geothermal fluids of Mutnovsky volcano, Kamchatka. Several aspects of the geochemistry of fluids are discussed: chemical equilibria, mixing of fluids from different sources, evaluation of the parent magmatic gas composition and contributions to magmatic vapors of fluids from different reservoirs of the Kamchatkan subduction zone. Among reactive species, hydrogen and carbon monoxide in volcanic vapors are chemically equilibrated at temperatures >300°C with the SO₂-H₂S redox-pair. A metastable equilibrium between saturated and unsaturated light hydrocarbons is attained at close to discharge temperatures. Methane is disequilibrated. Three different sources of fluids from three fumarolic fields in the Mutnovsky craters can be distinguished: (1) magmatic gas from a large convecting magma body discharging through Active Funnel, a young crater with the hottest fumaroles (up to 620°C) contributing ~80% to the total volcanic gas output; (2) volcanic fluid from a separate shallow magma body beneath the Bottom Field of the main crater (96–280°C fumaroles); and (3) hydrothermal fluid with a high relative and absolute concentrations of CH₄ from the Upper Field in the main crater (96–285°C fumaroles). The composition of the parent magmatic gas is estimated using water isotopes and correlations between He and other components in the Active Funnel gases. The He-Ar-N₂ systematics of volcanic and hydrothermal fluids of Mutnovsky are consistent with a large slab-derived sedimentary nitrogen input for the nitrogen inventory, and we calculate that only ~1% of the magmatic N₂ has a mantle origin and <<1% is derived from the arc crust.     

Gas-driven lava lake fluctuations at Erta ’Ale volcano (Ethiopia) revealed by MODIS measurements

The long-lived lava lake of Erta ’Ale volcano (Ethiopia) is remotely monitored by moderate resolution imaging spectroradiometers (MODIS) installed on satellites. The Normalised Thermal Index (NTI) (Wright et al. Remote Sens Environ 82:135–155 2002) is shown to be proportional to the volume of the lava lake based on visual observations. The lava lake’s variable level can be plausibly related to a stable foam, i.e. a mixture composed of densely packed non-coalescing bubbles in suspension within a liquid. This foam is trapped at the top of the magma reservoir, and its thickness changes in response to the gas flux feeding the foam being successively turned on and off. The temporal evolution of the foam thickness, and the resulting variation of the volume of the lava lake, is calculated numerically by assuming that the gas flux feeding the foam, initially constant and homogeneous since December 9, 2002, is suddenly stopped on December 13, 2002 and not restarted before May 2003. The best fit between the theoretical foam thickness and the level of the lava lake deduced from the NTI provides an estimate of both the reservoir radius, 155–170 m, and the gas flux feeding the foam, 5.5×10^?3–7.2×10^?3 m ³ s ^?1 when existing. This is in agreement with previous estimates from acoustic measurements (Bouche et al. Earth Planet Sci Lett 295:37–48 2010). The very good agreement between the theoretical foam thickness and that deduced from MODIS data shows for the first time the existence of a regime based on the behaviour of a stable foam, whose spreading towards the conduit (“wide” conduit condition), can explain the long-lived activity. Our predictive model, which links the gas flux at the vent to the foam spreading, could potentially be used on any volcano with a long-lived activity. The underlying gas flux and the horizontal surface area of the magma reservoir can then be deduced by combining modelling to continuous measurements of gas flux. The lava lake, when high, often shows regular rise and fall of its level. We have recognised a minimum of 26 very well marked cycles between January 2001 and December 13, 2002, corresponding to a typical return time of 10.8 ± 2.3 days and a gas volume of 8.3×10⁵ ± 2.0×10⁵ m ³. This corresponds to a gas volume fraction in the reservoir equal to 0.023–0.063 %. The yearly gas flux, estimated between December 13, 2002 and September 27, 2004, varies between 2.3×10^?6 and 5.9×10^?6 m ³ s ^?1 at the depth of the reservoir. The long-time series provided by infra-red sensors mounted on satellites could be used on any persistent volcano to detect potential periodic variations in the level of lava lakes or lava columns, providing that the vent has a funnel shape, as often, and is sufficiently large.     

DUCKS: Low cost thermal monitoring units for near-vent deployment

During 1999 we designed and tested a thermal monitoring system to provide a cheap, robust, modular, real-time system capable of surviving the hostile conditions encountered proximal to active volcanic vents. In November 2000 the first system was deployed at Pu'u 'O'o (Kilauea, Hawai'i) to target persistently active vents. Aside from some minor problems, such as sensor damage due to tampering, this system remained operational until January 2004. The success of the prototype system led us to use the blueprint for a second installation at Stromboli (Aeolian Islands, Italy). This was deployed, dug into a bomb-proof bunker, during May 2002 and survived the April 2003 paroxysmal eruption despite being located just 250 m from the vent. In both cases, careful waterproofing of connectors and selection of suitable protection has prevented water damage and corrosion in the harsh atmosphere encountered at the crater rim.The Pu'u 'O'o system cost ∼US$10,000 and comprises four modules: sensors, transmission and power hub, repeater station and reception site. The sensor component consists of three thermal infrared thermometers housed in Pelican™ cases fitted with Germanium–Arsenide–Selenium windows. Two 1° field of view (FOV) sensors allow specific vents to be targeted and a 60° FOV sensor provides a crater floor overview. A hard wire connection links to a Pelican™-case-housed microprocessor, modem and power module. From here data are transmitted, via a repeater site, to a dedicated PC at the Hawaiian Volcano Observatory. Here data are displayed with a delay of ∼3 s between acquisition and display. The modular design allows for great flexibility. At Stromboli, 1° and 15° FOV sensor modules can be switched depending changes in activity style and crater geometry. In addition a direct line of site to the Stromboli reception center negates the repeater site requirement, reducing the cost to US$5500 for a single sensor system.We have also constructed self-contained units with internal data loggers for US$1500/unit. These have been tested at Kilauea, Stromboli, Etna, Masaya, Santiaguito, Fuego, Pacaya, Poas, Soufriere Hills, Villarrica and Erta Ale. These instruments have proved capable of detecting thermal signals associated with: (1) gas emission; (2) gas jetting events; (3) crater floor collapse; (4) lava effusion; (5) lava flow in tubes; (6) lava lake activity; (7) lava dome activity; and (8) crater lake skin temperature.