The acid crater lake of Taal Volcano, Philippines: hydrogeochemical and hydroacoustic data related to the 2010–11 volcanic unrest

Pedro A. Hernández, Gladys V. Melián, Luis Somoza, Mª Carmencita Arpa, Nemesio M. Pérez, Ericson Bariso, Hirochicka Sumino, Eleazar Padrón, Johan C. Varekamp, José Albert-Beltran and Renato Solidum
Geological Society, London, Special Publications, 437, 131-152, 2 February 2017, https://doi.org/10.1144/SP437.17

Abstract

Studies of the water chemistry of Taal crater lake and echo-sounding surveys have provided new insights into its chemical and physical dynamics. During the volcano-seismic unrest of April 2010–June 2011, the waters of Taal crater lake showed changes in chemical composition and increases in CO2 emissions associated with the seismic unrest. The chemical and isotopic data show that the lake water has contributions from both seawater and meteoric water and receives injections of deep hydrothermal water and gases during periods of intense volcano-seismic unrest. These inflationary periods may lead to faulting of the impermeable cap rock that usually seals the deeper Taal hydrothermal reservoir in response to degassing and convective movements in the underlying Taal magma chamber.

Acid crater lakes form when volcanic or hydrothermal fluids enter a volcanic lake filled with meteoric water (e.g. Pasternack & Varekamp 1997; Christenson et al. 2015; Varekamp 2015). Volcanic lakes are an important route for degassing to the atmosphere and represent some of the most extreme natural chemical environments at the Earth's surface (Sriwana et al. 2000; Pérez et al. 2011). Volcanic crater lakes are also a unique hazard because of the danger posed by the storage of large volumes of water in proximity to near-surface magma bodies (Rowe et al. 1992). Monitoring the physical and chemical parameters of crater lakes is an important tool in the surveillance of volcanic systems because they tend to reflect variations in the composition and input fluxes of volcanic fluids and gases (e.g. Rowe et al. 1992; Christenson 2000; Ohba et al. 2008; Bani et al. 2009; Rouwet & Tassi 2011; Caudron et al. 2012; Arpa et al. 2013; Rouwet et al. 2014; Varekamp 2015). Such variations may reflect the water–rock interactions that occur in and beneath the lake and variations in the flux of heat and volatiles derived from subsurface hydrothermal/magmatic systems (Varekamp et al. 2000). Large amounts of magmatic gases can be dissolved in lake water and CO2 degassing from volcanic lakes should be monitored as part of volcanic surveillance studies (Mazot & Taran 2009; Hernández et al. 2011; Pérez et al. 2011; Caudron et al. 2012; Bernard et al. 2013).

Taal Volcano in the Philippines is host to an acid crater lake and is one of the most active volcanoes on Earth. Hydrovolcanic eruptions dominate the activity of Taal, posing an extreme hazard to people living around the volcano. We conducted hydrochemical surveys in several vertical profiles in the Main Crater Lake (MCL) of Taal Volcano and present data on changes in the water chemistry of the MCL before and during the volcanic unrest that occurred in 2010–11 (Arpa et al. 2013). These new data and earlier published data (Delmelle et al. 1998) cover a period of almost two decades. Chemical and isotopic analyses of lake water samples and fumarolic gases highlight the importance of hydrothermal contributions to the crater lake. The data for dissolved gases in the lake waters provide a better understanding of the evolution of the MCL and its underlying hydrothermal system. We also report the results of a hydroacoustic study, which provided information on the characteristics of rising gas bubbles in the water column and the lake bottom bathymetry via an echo-sounding (ES) survey. Gas bubbles in the water column produce strong ES backscatter signals that allow the detection and quantification of gas fluxes from subaqueous fumaroles (Greinert et al. 2010).

The main aims of this study were: (1) to investigate potential correlations between temporal variations in the chemical and isotopic composition of the MCL waters and the volcanic unrest that occurred at Taal Volcano in 2010–11; (2) to document the correlations between gas emissions and observed variations in the chemistry of the MCL waters; and (3) to analyse the types and spatial distribution of active subaqueous degassing zones identified in the ES surveys.

Geological background

Taal Volcano Island is a basaltic to andesitic arc volcano related to subduction in the Manila trench (Miklius et al. 1991; Listanco 1993; Mukasa et al. 1994). It is located in the Macolod Corridor, a zone characterized by intense volcanism and active extensive NE–SW- and NW–SE-trending fault systems (Defant et al. 1988; Forster et al. 1990) (Fig. 1a). The volcanic island consists of multiple tuff cones and tuff rings (Listanco 1993) (Fig. 1b). Taal has erupted more than 30 times since 1572 (PHIVOLCS 2009) and the last eruption of its Main Crater in 1911 caused 1300 casualties (Worcester 1912). The latest eruption period (1965–77) consisted of hydromagmatic to phreatic eruptions (Moore et al. 1966). The 30 km wide Taal caldera contains Taal Lake, the surface of which is at c. 2 m a.m.s.l. and which is connected to the South China Sea by the c. 9 km Pansipit River. Taal Volcano Island, within Taal Lake, contains the largest Taal crater at 2 km diameter (Main Crater), which is now filled by the 1.2 km wide, 84 m deep, acidic (pH c. 2.8) MCL. The fluids in the MCL and the spring discharge are derived from a deep parental reservoir (T=300°C) containing a mixture of volcanic fluids, meteoric water and seawater (Delmelle et al. 1998). Taal Lake is only a few kilometres from Balavan Bay in the West Philippine Sea and the depths to the seawater–meteoric water interface in the MCL and Taal Lake are 80 and 160 m, respectively.

Fig. 1.
Fig. 1.

(a) Geographical location of Taal Volcano, Philippines. (b) Shaded relief map and aerial view of Taal Volcano showing the Main Crater Lake (MCL) of the central island of the volcano (UTM coordinates in km). (c) Aerial view of the volcano taken from the NE showing the Taal MCL (TMCL).

Two large geothermal manifestations are found inside the Main Crater and on the northern upper slopes, where east–west-trending fissures opened during the 1991–94 seismo-volcanic crises. These thermal areas are also fed by the underlying hydrothermal system (Fikos et al. 2012). A multidisciplinary study on the Taal hydrothermal system, which combined electromagnetic, geochemical and thermal surveys, concluded that the northern part of the volcano is undergoing strong thermal transfers and degassing, with activity taking place mainly along east–west-trending fissures (Zlotnicki et al. 2009).

Seismic tomography studies have indicated the presence of a low-velocity zone below much of SW Luzon, including the Taal caldera. This zone at c. 18 km depth is interpreted as a magma body (Besana et al. 1995). A shallow degassing magma chamber at c. 6 km depth, connected to this deeper and larger reservoir, was suggested by seismic data (Nishigami et al. 1994). A magnetotelluric survey has indicated the presence of a hydrothermal reservoir at c. 500 m below the MCL (Yamaya et al. 2013). This hydrothermal reservoir carries two-phase water at temperatures >200°C and is overlain by an impermeable cap of altered clays saturated with hydrothermal liquid waters at <200°C. The shallow hydrothermal reservoir is probably the source of fluids in the geysers, solfataras and steaming vents on the northern shore of the MCL (Yamaya et al. 2013). Long-period earthquakes at depths of only 100–200 m beneath the soil are interpreted as vapour migration from the hydrothermal reservoir to the surface (Maeda et al. 2013).

The constant gas emissions from the convecting magma in the deep magma chamber through the hydrothermal reservoir may prevent pressurization of the magma system and may explain the long periods of dormancy at Taal (Kumagai et al. 2014). About 33 eruptions have occurred on Volcano Island in Taal Lake since the first recorded eruption in 1572 (Torres et al. 1995). The 1965–77 eruption created a new crater lake south of the MCL (Moore et al. 1966). A key feature of these eruptions is that the emission centres were not limited to the Main Crater, but occurred on the flanks of Volcano Island (e.g. Alanis et al. 2013). Possible intrusion events may have occurred later in 1992–94 when seismic swarms, ground fissuring and increased hydrothermal activity were detected (PHIVOLCS 1992, 1993, 1994). Prominent deformation of the MCL zone was observed in February–November 2000 and June 2004–March 2005 (Bartel et al. 2003; Galgana et al. 2014). The February–November 2000 period of inflation was characterized by c. 120 mm of uplift of the centre of Volcano Island relative to the northern caldera rim. Based on point-source model inversion, the source of inflationary deformation for the 2000 period was determined to be at 5.2 km depth beneath the NE part of Volcano Island (Bartel et al. 2003). By contrast, the 1999 deflationary source was situated south of the inflationary sources, near the south shore of the MCL, at a shallower depth of 4.2 km (Bartel et al. 2003). The 2005 inflation episode shows that the Mogi source was situated NE of the MCL at a depth of 5.1 km close to the 2000 event, indicating a nearly stationary magma source. The inflationary and deflationary patterns at Taal Volcano occur at intervals of about five to nine months, with uplift rates of as much as 220 mm a−1; these activities are much slower changes than those associated with low-viscosity magmatic systems such as Kilauea, Hawaii, where many eruptive sequences show rapid deformation pulses that span several hours to a few days (Galgana et al. 2014). Magmatic source change is still the most plausible explanation for uplift during the 2000–01 period, instead of hydrothermal inflation, because no apparent increase in surficial hydrothermal activity was observed (Bartel et al. 2003). Shallow hydrothermal sources may have affected the deformation signals of stations located near the NE quadrant of the MCL in early 2005, when strong thermal transfers and degassing were observed (Zlotnicki et al. 2009). Upward migration of shallow degassing magma from the deeper magma chamber has been related to surface deformation during the 2010 period of volcanic unrest (April 2010–June 2011), together with changes in CO2/SO2, CO2/H2O and SO2/H2S from fumaroles near the MCL (Arpa et al. 2013). Diffuse CO2 emissions through both the MCL – with the highest value of 4670 t d−1 above a baseline of about 1000 t d−1 – and soils in September 2010 during a period of volcano-seismic unrest in April 2010–June 2011 were also documented (Arpa et al. 2013). The increased CO2 emission rate and changes in the fumarole gas compositions were mostly likely caused by magma rising from a deep to a shallow reservoir, giving rise to increased degassing.

Methods

Two field campaigns were undertaken at the MCL as part of an Instituto Tecnológico y de Energías Renovables (ITER)–Philippine Institute of Volcanology and Seismology (PHIVOLCS) collaborative project in March and June 2011. Sampling of the MCL water was carried out along a 70 m deep vertical profile in both surveys (open circle in Fig. 2). The water temperature, pH and electrical conductivity were measured in situ using an Oakton Model PC-10 multiparameter probe (accuracies ±0.5, ±0.2 and ±1%, respectively); the pH meter was calibrated in the field with standard solutions of pH 2.00 and 4.00. Water samples for the laboratory determination of the major and minor ions (Ca2+, Mg2+, K+, Na+, NH4+, Cl−, SO42−, F− and Br−) and for hydrogen (δH) and oxygen (δO) stable isotopes were collected and stored for later analysis at the ITER laboratory, Spain.

Fig. 2.
Fig. 2.

Transects from echo-sounder surveys at Taal MCL in March 2011. Open circle indicates the location of the vertical profile for the March and June 2011 surveys. The locations of the echo-sounder profiles 1 and 2 of Figures 11 and 12 are also shown. Thick lines at the surface of Taal MCL indicate the transects of the echo-sounding profiles.

The anions (Cl−, SO42−, Br− and F−) were determined with a Dionex DX-500 ion chromatograph, whereas the cations (Ca2+, Mg2+, K+, Na+ and NH4+) were determined with a Metrohm 861 Advanced Compact ion chromatograph in acidified and filtered (0.45 µm) samples. The accuracy of the analyses was estimated at c. 2–3.5 and 5% for the major and minor species, respectively.

Water samples for the determination of dissolved gases were stored with no headspace in 60-ml glass bottles sealed in the field with silicon/rubber septa using a crimping tool. Before sealing the glass bottles, CuCl2 was added to prevent any post-sampling biological activity. Chemical analyses were performed on the gas phase extracted after the attainment of equilibrium (at constant temperature) between the water sample and the gaseous phase after the introduction of a carrier gas (Ne) into the bottle (Sugisaki & Taki 1987; Capasso & Inguaggiato 1998; Capasso et al. 2000). The concentrations of He (reproducibility of measurement ±300 ppb) and CO2, N2, O2 and Ar dissolved in the lake waters were determined with a Model Pfeiffer Omnistar 422 quadrupole mass spectrometer (accuracy 0.2%). The quadrupole mass spectrometer was calibrated in the laboratory with standard gases, which were used to determine the characteristic fragmentation patterns, sensitivity factors and mass calibrations for the gases dissolved in water.

The hydrogen and oxygen isotopic ratios of the waters were obtained by mass spectrometry (Thermo-Finnigan MAT 253) at the ITER laboratory. The results are expressed in delta units δ (‰), defined as: Embedded Image where δ is the isotopic deviation in ‰, s is the sample, p is the international reference standard and R is the isotopic ratio (2H/1H, 18O/16O). The accuracies for δD and δ18O were ±0.5 and ±0.05‰ v. Vienna Standard Mean Ocean Water (V-SMOW), respectively. The 3He/4He ratios of the dissolved gases from the MCL were analysed at the laboratory facilities of the Geochemical Research Center of The University of Tokyo, Japan. Lead glass sampling bottles with volumes of about 50 cm3 and high-vacuum stopcocks at both ends were completely filled with water. The gas was extracted by ultrasonic vibration and quantitatively collected into a small volume (10 cm3) following the method described by Padrón et al. (2013). The determination of He isotope ratios as well as the 4He and 20Ne concentrations was carried out according to the method of Sumino et al. (2001). Air standards were measured frequently during the analyses to determine the precision of the mass spectrometer. The correction factor for He isotope ratios was determined by measuring the inter-laboratory He standard HESJ with the recommended 3He/4He ratio of 20.63±0.10 Ra (Matsuda et al. 2002). Uncertainties given with the reported 3He/4He ratios are 1σ of each measurement series, including the error in the measured raw 3He/4He ratio. Errors in concentrations are estimated to be 10% based on repeated air standard analyses.

The ES surveys were carried out using a Lowrance HDS-5 instrument equipped with a dual-frequency (50 and 200 kHz) transducer. Average boat speeds ranged between 0.5 and 3 knots. The digital ES data were processed with Sonar Viewer 2.1.2 software. The extracted bathymetric data were converted from the Lowrance-type Mercator projection to the Universal Tranverse Mercator projection using ArcGis 10.2.1. The uppermost scattering surface marked by the automatic bottom detection, which was assumed to be the top of the degassing flares, was interpolated by kriging with Grapher software; three-dimensional surface analyses were performed with Fledermaus IVS software. We mainly used a 50 kHz sampling frequency because data at 200 kHz miss the bubbles detected at 50 kHz (Caudron et al. 2012). The ES survey was conducted along and crossing the lines shown in Figure 2.

Chemistry of Taal MCL

Physicochemical parameters and chemical composition of crater lake waters

The acid MCL waters had high concentrations of sulphate, chloride and the major cations (Table 1) during the March and June 2011 sampling campaigns. The average pH of the MCL water was 2.8, with a relative standard deviation of 3.7% during both surveys; the surface temperature was 30°C in March 2011 and 33°C in June 2011. The total dissolved solids (TDS) were high at TDS=19 808–30 400 ppm. All the MCL water samples showed high concentrations of Na+ (4889–5910 ppm) and moderately high concentrations of K+ (60–840 ppm) and Ca2+ (260–630 ppm). The waters were rich in Cl− (8330–11 100 ppm) and SO42− (1630–7570 ppm), although these results were comparable with the data reported in Delmelle et al. (1998). The pH and temperature profiles did not show strong variations with depth, except for the temperature measured at the lake surface, which was subject to environmental influences. The temperatures measured during this study in these vertical profiles indicated the absence of any significant thermal stratification. Of all the species measured, only Cl− showed some variation with depth (Fig. 3) for the June profile.

Fig. 3.
Fig. 3.

Vertical profiles of temperature, pH, electric conductivity, and major cations and anions.

View this table:
Table 1.

Chemical composition of the Taal MCL water

The MCL waters belong to the Na+(Mg2+)–Cl−(SO42−) compositional group (Fig. 4) typical of crater lakes in active volcanic systems (Delmelle & Bernard 1994; Tassi et al. 2009). All the samples showed relatively low SO42−/Cl− ratios, which can be linked to an important contribution from seawater and to the influence of the discharge of volcanic gases along the bottom of the crater that interacted with the MCL water. This hypothesis is supported by the compositional trend observed between the Na+ and SO42− contents in all the samples in the ternary Na+–SO42−–Mg2+ diagram (Fig. 5), suggesting the addition of S-rich gases (Yamaya et al. 2013) to the MCL waters.

Fig. 4.
Fig. 4.

Piper plot showing major ion chemistry for the Taal MCL waters.

Fig. 5.
Fig. 5.

Na+–SO42−–Mg2+ ternary diagram for Taal MCL water samples collected during the March and June 2011 surveys. Data from Delmelle et al. (1998) are also shown.

The dissolved constituents decreased in abundance in the order Na+>Mg2+>K+>Ca2+ and with Cl−>SO42−. The Na+ v. Cl− diagram (Fig. 6a) suggests binary mixing, with the variable dilution of a seawater end-member with small contributions of Cl− from other sources, similar to the observations of Delmelle et al. (1998) in samples taken in 1995. However, for Mg2+, Ca2+ and SO42− v. Cl− (Fig. 6b–d), most of the samples plotted above the mixing line between seawater and a meteoric end-member. Water–rock interactions may result in significant releases of Mg2+ and Ca2+ into the water mixture; enrichment in Na+ was unlikely to be observed because this was masked by the relatively high Na+ concentrations in seawater. Delmelle et al. (1998) suggested two possible mechanisms to explain the cation contents of the MCL: (1) rock dissolution by acidic volcanic waters, followed by water–rock interactions at shallow levels; and (2) contamination of the hydrothermal system by a marine component. The observed enrichments of Mg2+ relative to Ca2+ (half of the samples plot along the mixing line) may suggest that the removal of Ca2+ or the addition of Mg2+ occurred in the course of the mixing process between seawater and the dissolved rock end-members (Delmelle et al. 1998). This process may have been enhanced during the volcanic unrest of 2010–11 due to the interaction of high-temperature magmatic gases and groundwater with rocks, resulting in a significant input of Mg2+ and SO42− into the MCL waters.

Fig. 6.
Fig. 6.

Relationships between Cl− and (a) Na+, (b) Ca2+, (c) Mg2+ and (d) SO42− (all in meq l−1) for Taal MCL water samples collected during the March and June 2011 surveys. Data from Delmelle et al. (1998) are also shown.

The K+, Na+, Mg2+ and Ca2+ concentrations of the MCL waters cannot be explained in terms of either water–rock equilibrium or rock dissolution alone (Delmelle et al. 1998), which may well occur in the hotter hydrothermal zone beneath the lake because most of the samples lie on a simple rock–seawater mixing curve. The incorporation of seawater into the hydrothermal system of Taal is supported by the measured Br−/Cl− values (see Table 1), which suggest that the deep hydrothermal reservoir from which the surface discharges derive is ultimately a mixture of seawater and volcanic water, with variable contributions of meteoric water.

Stable isotopes of water: δD and δ18O

The stable isotope data from the waters provide an insight into the various water end-members as well as the evaporation processes in the lake (e.g. Varekamp & Kreulen 2000). The MCL waters are all enriched in both δD and δ18O (Table 2) relative to local meteoric values, but they plot far from a mixing trend between seawater and meteoric waters in a δD v. δ18O diagram (Fig. 7a, b). This may be the result of the mixing of seawater with volcanic hydrothermal fluids (VHFs), andesitic water and/or evaporation. Marked differences occur between the MCL waters collected in 1995 by Delmelle et al. (1998) and samples from the present study. The two 1995 MCL samples showed similar δ18O values (+1.9 and +2.1‰), somewhat heavier than those collected in 2011 (March and June average values of +1.1 and 1.4‰, respectively). Assuming that meteoric waters during the last 20 years have had a similar isotopic composition, the δD values shifted to lighter δD, from −5.0‰ in 1995 to −18‰ in March 2011. The June 2011 values are again closer to the 1995 isotope data. Several hypotheses can explain the observed isotopic shifts: (1) they are a result of a higher precipitation rate in 2011 (moving the 2011 data points towards the meteoric water end-member); (2) there were increases in the amount of VHF (andesitic water); (3) there were changes in the evaporation rate as a result of the increased water temperature; and (4) hydrogen exchange took place between H2S and H2O. The simple evaporation of meteoric waters (Fig. 7a) could explain the isotopic data, but not the chemical data. However, the evaporation of a ternary mixture of VHFs, seawater and meteoric water (Fig. 7b) may be able to explain the differences in the 1995 and 2011 data. The concentrations of Cl−, SO42− and most cations were lower in 2011 in the MCL compared with 1995 and some data plot on the mixing lines with meteoric water, suggesting some dilution of the lake waters. The lake water temperature in 2011 was similar or slightly lower than in 1995, so a different evaporation path as a result of an increased lake water temperature is unlikely. The 1995 to March 2011 trends in Figure 7 are not towards the meteoric end-member, so simple dilution is not a viable explanation, whereas the June 2011 and 1995 data plot on a meteoric water mixing line (Fig. 7a).

Fig. 7.
Fig. 7.

(a, b) Variation of δD and δ18O in waters and gas condensates (triangles) from Taal Volcano. The isotopic composition of andesitic waters is from Giggenbach (1992). (c) δD v. Cl− and (d) δ18O v. Cl− binary diagrams for the Taal MCL water samples collected during the March and June 2011 surveys. Data from Delmelle et al. (1998) are also shown.

View this table:
Table 2.

δ18O v.δD composition of Taal MCL waters and chemical and He isotopic composition of dissolved gases in Taal MCL waters

The H2S and SO2 gases have different solubilities in water (Symonds et al. 2001) and their relative abundances are mainly controlled by the temperature, pressure and redox conditions (Giggenbach 1987). At the relatively low fumarolic temperatures measured in 2011, similar H2S and SO2 proportions are expected. Most of the SO2 will dissolve in water, whereas the H2S remains mainly in the gas phase. As a result, H2S can produce an H isotopic exchange with water. This exchange could be produced by the greater availability of H2S during March 2011, in consort with the increase in volcanic activity during that period. Variations in SO42− and Cl− for the MCL waters relative to the VHF, meteoric water and seawater end-members are shown in Figure 8. The 1995 samples seem to be mixtures of seawater and VHF, with dilution for the June 2011 samples and the addition of S for the March 2011 samples. These trends support the idea that part of the δD trend is caused by isotopic exchange with H from H2S.

Fig. 8.
Fig. 8.

Binary diagram SO42− v. Cl− for the Taal MCL water samples collected during the March and June 2011 surveys. Data from Delmelle et al. (1998) are also shown. VHF, volcanic hydrothermal fluid.

The δD and δ18O isotopic composition of fumarolic condensates from 1995 and June 2011 show the influence of seawater contamination because they plot near the SMOW–local meteoric water mixing line. The Taal caldera lake samples with their clear seawater chemical signal also plot on this mixing line (TL waters in Fig. 7a, b). The fumarolic vapours thus have a strong seawater component. The isotopic differences between the 1995 and 2011 fumarole condensate samples (stars in Fig. 7a, b) may relate to different mixing proportions of seawater or enhanced evaporation processes in 2011. The latter may be related to the higher temperatures associated with the volcanic activity and possibly enhanced inputs of H2S. These processes may have masked any isotopic signature deriving from the direct contribution of magmatic waters (‘andesitic water’), as has been suggested by other researchers (Taran et al. 1989; Tassi et al. 2005).

The distribution of the MCL water samples in the Cl− v. δD and Cl− v. δ18O binary diagrams (Fig. 7c, d) diverge from the mixing line between an isotopically light meteoric end-member and seawater, suggesting the contribution of waters with different isotopic values. Figure 7c shows a shift in δ18O from the 1995 to the March 2011 water samples towards lighter values, which may represent an increase in the isotopic exchange of endogenous CO2 with water. Similar behaviour is observed for δD, showing a shift towards lighter δD values, which can be explained in terms of more abundant H2S and associated isotopic exchange. Both changes coincide with the significant increases observed on the diffuse CO2 emission from the MCL and in the changed CO2/H2O and SO2/H2S molar ratios from the fumaroles at Taal Volcano prior to and during the volcanic unrest in 2010–11 (Arpa et al. 2013). Quantitative modelling of the evaporation of a mixture with variable proportions of three end-members over time is possible (Varekamp & Kreulen 2000), but such simulations have too many degrees of freedom to provide truly new insights.

Dissolved gases at Taal MCL

The concentrations of dissolved gases (Table 2) in the MCL samples collected during June 2011 are strongly elevated. The dissolved CO2 in the spring waters (average 876 cm3 STP l−1) is several orders of magnitude higher than that of water in equilibrium with the atmosphere (0.32 cm3 STP l−1). The measured He contents were about two orders of magnitude higher than that of water in equilibrium with the atmosphere (4.81×10−5 cm3 STP l−1). The relative amounts of dissolved O2, N2 and CO2 in the MCL waters are shown, together with the O2/N2 value of air-saturated water (ASW), in Figure 9a. All the MCL samples plot to the right of this ASW ratio, indicating lower O2/N2 ratios than the ASW values and suggesting interaction between the fumarolic gases and the MCL waters. The low O2/N2 ratio of the dissolved gases in thermal waters results from a lack of oxygen in the fumarolic gases and to the consumption of oxygen by reducing gases, such as the conversion of H2S to SO42− (Capasso et al. 2000). Both mechanisms explain the observed low O2/N2 ratios in the MCL waters because the fumarolic discharges at Taal Volcano have very low O2 contents (Delmelle et al. 1998; this work) and the acidity of the MCL waters can be explained by the dissolution and subsequent oxidation of H2S. This pathway is supported by the high SO42− content in the MCL waters. The gas composition is dominated by CO2 because the N2/CO2 ratios are four orders of magnitude smaller than the atmospheric ratio and the samples plot along the N2–CO2 line in Figure 9a.

Fig. 9.
Fig. 9.

(a) O2–CO2–N2 and (b) He–CO2–N2 ternary diagrams for dissolved gases in the Taal MCL waters during the March and June 2011 surveys. The He–CO2–N2 ternary diagram shows the He/N2 ratios of fumarolic discharge samples in April 2008, March 2011 and June 2011 as well as the fumarole data for 1995 from Delmelle et al. (1998). Vertical profiles of (c) He/36Ar ratios and CO2 and (d) N2/O2 ratios and He in the MCL waters.

The dissolved He, N2 and CO2 in the MCL samples also plot far from the ASW line (Fig. 9b). The He/N2 ratios of the fumarolic discharges sampled in 1995 (Delmelle et al. 1998) and in this work are also shown. The dissolved gases show very similar He, N2 and CO2 contents and plot close to the CO2 vertex. The physicochemical conditions of the MCL (low pH, relatively high temperature and high salinity) contribute to the behaviour of CO2 in water due to the different solubility and reactivity between He and CO2 (Capasso et al. 2000). The He/N2 ratios in dissolved gases in the MCL waters are different from those measured in the fumarolic discharges. All the samples show He/N2 ratios higher than the atmospheric ratio and plot between the ratios measured in the fumarolic discharges and in the frying pan discharge located along the lakeshore (see Delmelle et al. 1998), supporting the close relationship between MCL waters and fumarolic fluids discharged from the bottom of the crater.

The distribution of dissolved He and CO2, together with the He/36Ar and N2/O2 ratios along the MCL vertical profile, is shown in Figure 9c and d. He and CO2 show a decreasing trend towards the 50 m depth sample and then increase again to the bottom of the lake. This observation is consistent with the occurrence of significant CO2 sources rising up from the bottom of the lake through fumarolic discharges. This behaviour is also observed in the He/36Ar ratio, but this is the opposite to the CO2/N2 ratio at that depth. The decrease in He and CO2 at 50 m depth could be related to inputs of fresh meteoric waters from Taal Lake diluting the CO2 and He contents with a strong atmospheric component. This observation is supported by the distribution of He/36Ar ratios along the profile and the vertical variation of major components such as SO42− and Cl−. It is well documented that noble gases (e.g. Ar and He) are useful geochemical tracers because the 36Ar ASW composition is consistent globally and the crustal components (4He and 40Ar) are resolvable and unaffected by oxidation or microbial activity (Darrah et al. 2014). The noble gas isotopic composition along the profile reflects different mixtures of fresh meteoric waters containing 36Ar (and N2) derived from solubility equilibrium with the atmosphere during groundwater recharge and noble gases derived from a magmatic component. The distribution of N2/O2 along the vertical profile (Fig. 9d) shows an increase in this ratio close to the bottom of the lake. This behaviour can be related to the consumption of O2 due to the oxidation of H2S to SO4 and/or organic matter in the deeper waters.

The 3He/4He ratios of the MCL waters sampled along the vertical profile range from 7.4 to 7.6 R/Ra (where R is the 3He/4He isotopic ratio of the gas and Ra is the 3He/4He ratio of air and is equal to 1.39×10−6; Mamyrin & Tolstikhin 1984), with 4He/20Ne ratios in the range 0.4–6.8 (Table 2). The 3He/4He ratios of the fumarolic gas discharges sampled in 2008 (unpublished data) and 2011 range from 7.3 to 7.6 R/Ra, with 4He/20Ne ratios from 5 to 298. Fluids from mantle degassing show high R/Ra values, up to 20 or more in hotspot/plume gases, whereas the R/Ra ratios for crustal gases are as low as 0.01 (Craig & Lupton 1976). The He isotope ratios in geothermal gases were measured by Poreda & Craig (1989) for Taal as R/Ra=7.5. The measured 3He/4He values reveal a major contribution from a mantle-type source. Figure 10a shows the diagram of uncorrected R/Ra v. 4He/20Ne values. All samples have 3He/4He values above atmospheric values and the measured ratios fall on a mixing line between a local mantle value and an atmosphere-derived end-member. The data show that although the 4He/20Ne ratio varies significantly, the 3He/4He isotopic ratios remain relatively constant at c. 7.5 Ra, indicating that this ratio is characteristic and not affected by atmospheric contamination. This is supported by the distribution of 3He/4He ratios along the vertical profile (Fig. 10b). Noble gases such as He, Ar and Ne are not affected by biogeochemical processes, thus their distribution along the lake vertical profiles mainly depends on physical processes such as advection and diffusion (Cabassi et al. 2013). The MCL is a high-activity volcanic lake affected by the addition of significant amounts of hot hydrothermal–magmatic fluids and advection seems to be the main physical process responsible of the observed vertical distribution. The absence of significant variations on the 3He/4He isotopic ratios along the vertical profile supports this hypothesis.

Fig. 10.
Fig. 10.

(a) Helium isotope ratios (normalized to the atmosphere) v. 4He/20Ne ratios. The samples plot on a mixing line between the atmosphere and a mantle-type end-member. The mixing boundary lines are built with the following end-members: atmosphere (as dissolved air in water) R/Ra=1 and He/Ne=0.285; mantle R/Ra=6.5 and He/Ne=1000; continental crust R/Ra=0.01 and He/Ne=1000. (b) Profile of change in 3He/4He ratio with depth.

Acoustic imaging of Taal MCL

The acoustic imaging of MCL shows the presence of bubbles and provides information on the bathymetry of the lake. The 50–200 kHz echograms obtained during the March–June 2011 campaigns (Fig. 2) exhibit frequent vertical plumes of rising bubbles within the MCL (Figs 11 & 12). The bubble backscatter surface (Fig. 11a) has an undulating morphology very different from the lake bathymetry obtained several months earlier (January 2011; Bernard et al. 2013). This reflective surface (Fig. 11a) clearly represents the ‘wall of bubbles’ (labelled as 1 and 2 in Fig. 11b) and not the lake floor topography. The term wall of bubbles was used in Kelud volcanic lake to describe very high concentrations of bubbles (Caudron et al. 2012) from intense degassing by subaqueous fumaroles. The 50 kHz echograms show several walls of bubbles in the deep central basin of the MCL, which are mostly orientated east–west (Fig. 11a, b). These correspond to the zones of highest CO2 fluxes observed during the March 2011 survey (Arpa et al. 2013). These features (Fig. 3) show a rapid temperature increase in the water column near the top of the surface (c. 10 m depth) up to 33.5°C, which then increases gradually to the bottom of the lake. At the same time, the specific conductivity decreases rapidly from the surface of the lake to the top of this surface (Figs 5 & 11b). By contrast, the pH values do not show any change near the top of this acoustic surface.

Fig. 11.
Fig. 11.

(a) Three-dimensional image of the top of the high scattering surface identified on the 50 kHz echo-sounding survey. The highest values (red to yellow colours) correspond to zones with walls of bubbles as observed on the profile in (b) (labelled as 1 and 2). Therefore this surface does not reflect the bathymetry, but shows the acoustic anomalies related to the intense degassing that masks the floor of the lake. (b) Echogram (50 kHz) portraying an intense degassing area (wall of bubbles). Colour scale in (b): red to yellow colours reflect high concentrations (high scattering) and blue colours reflect low concentrations (low scattering). Location of profile is shown in Figures 2 and 11a. (c) Three-dimensional image showing acoustic flares sourced from circular depressions. Location shown in (a). (d) Three-dimensional image showing the train of bubbles (‘flares’) sourced along the southern scarp of the MCL. Location shown in (a).

Fig. 12.
Fig. 12.

(a) Photograph of the MCL taken from the northwestern border during the campaign in June 2011 and showing floating rafts of yellow sulphur on the surface of the lake. Locations of the echogram profiles in (b) and (c) are also shown. (b, c) Echograms (50 kHz) showing a well-defined vertical train of bubbles (‘flares’). Location of profiles is shown in Figures 2 and 12a.

The sides of the MCL show flares as the main manifestation of hydroacoustic bubbles, with similar shapes to those detected earlier (Bernard et al. 2013, Fig. 11c, d). Flares are the term for acoustic plumes with vertically elongated shapes in echograms (Greinert et al. 2006). The echograms show that the majority of the rising gas bubbles had largely dissolved at c. 3–5 m below the lake surface (Figs 11c & 12b). Most flares stem from soft pools (Fig. 11c), which are small circular depressions up to 80 m in diameter that occur on the lake floor of the MCL (Figs 11c & 12b). The reflectivity contrast between the soft sediments and the underlying volcanic rocks of the MCL can be clearly seen on the ES echograms (Fig. 11a, b). The topography of the volcanic rocks appears strongly deformed, with extensional faults bounding the circular depressions (Fig. 12b, c). These circular depressions are imaged as negative acoustic anomalies on the echograms, but do not reflect the real lake bathymetry (Figs 11c & 12c).

The more weakly degassing part of the MCL shows flat echo trains that reflect ellipsoidal bubble shapes and allow us to calculate the bubble speed within the water column (Ostrovsky et al. 2008). The rise velocity can be directly measured from the time difference at two distinct depths along the same bubble line (e.g. Fig. 12c). The different slopes of these bubble lines indicate differences in the rise velocities. The progressive bending of the bubble lines indicates decreasing rise velocities due to bubble shrinkage as a result of CO2 dissolution (Greinert et al. 2006).

Thus some flat echoes observed within the plumes show that the speed of ascending bubbles decreases from 20 cm s−1 at 35–30 m water depths to 9.5 cm s−1 at 11–10 m water depths (Fig. 12c). These velocities fall into the theoretical range for ascending CO2 bubbles, which varies between 20 and 30 cm s−1 for bubbles between 5 and 20 mm in diameter, respectively (McGinnis et al. 2006). Some large fields with flat echoes show descending trajectories and yield velocities of 24–35 cm s−1 at 4–12 m water depths (Fig. 12b) and 48 cm s−1 at 7–15 m water depths (Fig. 12c). These descending trajectories of flat echoes are related to falling particles in the water column. These falling particles fill up the faulted bottom of the MCL and smooth out the bottom surface of the lake. Although particle material was not found floating on the surface of the lake during the June 2011 ES survey, oblique photographs from drones obtained in 2014 show white floating plumes at the south end of the MCL (identified as red arrows in Fig. 12a). These surface plumes are above the location of sub-lacustrine fumaroles imaged acoustically in June 2011 (Figs 11 & 12). The nature of this particle material from the subaqueous fumaroles will be addressed in the following section.

Discussion

Temporal trends in lake chemistry

The overall data from this study and earlier work indicates variations in the composition of the MCL waters. We explore here a model of periodically rising fluids from the high-temperature hydrothermal reservoir underlying the MCL as identified by Yamaya et al. (2013) (Fig. 13a). The cap rock of the underlying hydrothermal reservoir is composed of altered rocks and impermeable clays (Fig. 13a). During non-inflationary periods, the aquifer surrounding the MCL may inject meteoric water and seawater into the low-temperature (<200°C) hydrothermal reservoir and may thus generate low-temperature mixed-source seeps into the MCL (Fig. 13a). During inflationary periods (Fig. 13b), waters and gases from the deeper (c. 500 m depth) hydrothermal reservoir (>200°C) may rise in response to increased degassing from the c. 5 km deep magma reservoir. The inflation may also trigger faults/fissures in the overlying impermeable cap rock, similar to the behaviour of shale diapirs above doming areas (Fig. 13b). The documented increase in seismicity in 2011 could have been related to fracturing of the cap rock during the formation of the radial and crest fissures. This would create pathways for two-phase fluids from the underlying hydrothermal reservoir (Fig. 13b; PHIVOLCS Annual Report 2009). Long-period earthquakes were only recorded at depths of 100–200 m beneath the surface and are interpreted as being related to the migration of vapour from the hydrothermal reservoir to the surface through these conduits (Maeda et al. 2013).

Fig. 13.
Fig. 13.

(a) Model of the large hydrothermal reservoir beneath Taal Volcano (modified from Yamaya et al. 2013). (b) Proposed evolution of this model as a response to the episodic intrusion of the hydrothermal reservoir beneath Taal MCL during inflation periods. We infer that radial fissures are developed on the overlying cap rock as a consequence of the overpressure of the hydrothermal reservoirs. This system of radial fissures is used as pathways for rising high-temperature fluids. The changes in the lake chemistry as a consequence of the temporal intrusions of the hydrothermal reservoir into the MCL are also labelled. See text for further explanation. (c, d) Comparison between the two 50 kHz echo-sounding surveys carried out in January 2011 by Bernard et al. (2013) (available at http://www.ulb.ac.be/sciences/cvl/taal/Taal.html; and in June 2011 (this work). The surface obtained in June 2011 is interpreted to reflect acoustic anomalies (wall of bubbles and flares) caused by intense degassing that mask the floor of the lake as response to the overpressured hydrothermal reservoir.

This model may explain the increase in Mg2+/Cl− ratios in March 2011, with inputs of fluids from the deep hydrothermal reservoir. The temporal evolution of Mg2+/Cl− ratios in the MCL water (Fig. 14) is based on the maximum measured contents for each chemical component and is plotted with the recorded seismicity between 2008 and 2014 and the diffuse CO2 emission from the MCL surface (data from Arpa et al. 2013). The three parameters (maximum degassing rate, seismic activity and dissolved element ratios) show a positive correlation. The Mg2+/Cl− values increased sharply in March 2011 before the April 2011 peak of volcanic unrest and then later (June 2011) decreased to the levels of previous years. The increased magmatic activity of 2011–12 may also have provoked an increase in temperature in the subaqueous fumaroles and an increase in the volcanic gas flux. The frequency of sampling was insufficient, but the Mg2+/Cl− values during periods of volcanic unrest are clearly higher than during volcanically quiet times. Therefore the Mg2+/Cl− ratio is considered to be a useful indicator ratio for the assessment of volcanic activity in active crater lakes such as the Taal MCL (e.g. Giggenbach 1974). Similarly, increases in the Mg2+/Cl− ratio have been observed before and during periods of volcanic unrest at Ruapehu Crater Lake in New Zealand (Christenson & Wood 1993), Yugama Lake in Japan (Ohba et al. 2008) and Copahue in Argentina (Varekamp et al. 2009; Agusto & Varekamp 2015).

Fig. 14.
Fig. 14.

Mg2+/Cl− ratios in Taal MCL water samples collected during the March and June 2011 surveys. Data from Delmelle et al. (1998) are also shown. The temporal evolution of the diffuse CO2 emission rate through the water surface of the MCL and the daily number of high-frequency seismic events are also displayed as grey bars (data from Arpa et al. 2013).

The measured 3He/4He ratios at the MCL are relatively uniform between the lake waters and fumaroles, suggesting that the 3He/4He ratio (Rc/Ra=7.5) is representative of the magmatic He isotope ratio. The origin of He has implications for the origin of other gases such as CO2. The similar He isotope signature measured in the CO2-rich gases (Arpa et al. 2013) suggests that the CO2 is also of magmatic origin

We also hypothesize that, during the inflationary episodes, the flow in the aquifers might be reversed, flowing from the MCL to Taal Lake, which may explain the fish kills periodically observed in Taal Lake.

Subaqueous vents linked to intrusions of the high-temperature hydrothermal reservoir beneath the MCL

The 2010–11 volcanic unrest caused two changes in the hydroacoustic features that we could document from a comparison of our data from May–June 2011 (Fig. 13d) and the ES survey carried out in January 2011 (Bernard et al. 2013, Fig. 13a). Positive acoustic anomalies were found along the central axis of the MCL in June 2011 (Fig. 13b), masking the lake bathymetry determined in January 2011, which was caused by more intense degassing from sub-lacustrine fumaroles. The acoustic anomalies align towards the east with the shore areas with geothermal activity (Fig. 13d). Negative anomalies are related to linear arrays of ‘apparent pools’ located at c. 85 m water depth (Fig. 11c), which seem to be almost 15 m deeper than the 71 m recorded during the ES survey of January 2011 (Bernard et al. 2013) (Fig. 14). These pools are probably acoustic artefacts instead of local deeper pool areas. These anomalies can be related to local emissions of high-temperature fluids that reduce the speed of sound in the water column and thus would increase the apparent lake depth determined acoustically.

Hydroacoustic evidence for injections of high-temperature fluids from the deeper hydrothermal reservoir include the presence of particles near the sub-lacustrine vents. Four stages in the intensity of subaqueous fumaroles can be distinguished based on the temperature of the vents (Takano et al. 1994). At temperatures below 119°C, the fumaroles discharge low-temperature gases with only traces of SO2, whereas at 119–150°C the fumaroles also release spherules of native S that form when rising gases traverse through pools of molten S at the bottom of the lake. Temperatures >150°C create a drastic increase in the viscosity of molten S. At still higher temperatures, phreatic or geyser-like eruptions are observed in crater lakes (e.g. Christenson et al. 2010).

The occurrence of down-going density currents with small particles and the floats of particles at the surface (profiles 2 and 3 in Fig. 12) of water overlying the subaqueous vent areas suggest that the vents produce small spheres of native S. The occurrence of S particles floating on the MCL is not a strange phenomenon because the MCL is locally referred to as the Yellow Lake, in reference to the yellow particles of S floating on its surface. Similar zones of floating spherules of S have been observed at Santa Ana lake (Colvin et al. 2013). The underlying local vents must therefore carry relatively high-temperature waters (>119°C) resulting from the injection of hotter hydrothermal waters and gases, presumably from the deeper reservoir.

Concluding remarks

Our chemical and acoustic surveys of the MCL indicate that the major chemical species dissolved in Taal MCL water and the distribution of dissolved gases in the water column both show a slightly inhomogeneous distribution. The lake water contains contributions of seawater, meteoric water and hydrothermal fluids (Delmelle et al. 1998). Dissolved CO2 is considerably higher near the lake bottom, presumably as a result of the discharge of fluids rich in volcanic gases into the lake. The emission of gases from a deep magma chamber (Kumagai et al. 2014) may be the ultimate source affecting the chemistry of the lake. Observed changes in the crater lake chemistry during the 2010–11 period of volcanic unrest can be correlated with a significant increase in seismicity and rates of emission of CO2 and SO2 (Arpa et al. 2013). We relate these changes to changes in pressure at depth, which may have triggered faulting of the impermeable cap rock, creating leaks from the underlying hydrothermal reservoir of Taal Volcano (Yamaya et al. 2013). The rate of subaqueous degassing and changes in the composition of the MCL contributes to monitoring of the magma chamber underlying the hydrothermal system. Periodic hydrogeochemical and hydroacoustic surveys of the MCL may thus help to document changes in the state of activity of these high-risk volcanic zones and lead to the development of better precursor parameters for magmatic volcanic activity.

Acknowledgments

This research was supported by the projects financed by The Spanish Agency for International Development Cooperation (AECID) (08-CAP2-1202) by the R+D+I Spanish National Plan 2008–11 (CGL2005-07509) and SUBVENT (CGL2012-39524-C02), as well as by the Cabildo Insular de Tenerife (Spain). We are also grateful to the staff of Taal Volcano Observatory as well as to Dra Dácil Nolasco for her help during the field surveys. Special thanks are given for logistical support from the residents of Taal Volcano Island, particularly our boat men, Jerry and Joem, who skilfully paddled the canoe to each measurement site for the many surveys. Inputs and comments from reviewers improved the manuscript.

References

  1. ↵
    Agusto M.R. & Varekamp J.C. 2015. Geochemistry of the acidic waters of the Copahue volcano-hydrothermal system, with applications for volcanic surveillance. In: Tassi F., Vaselli O. & Caselli A.T. (eds) Copahue Volcano. Springer Series in Volcanology, 199–230.
  2. ↵
    Alanis P.K.B., Yamaya Y., Takeuchi A., Sasai Y., Okada U. & Nagao T. 2013. A large hydrothermal reservoir beneath Taal Volcano (Philippines) revealed by magnetotelluric observations and its implications to the volcanic activity. Proceedings of the Japan Academy, Series B, 89, 383–389, https://doi.org/10.2183/pjab.89.383
    OpenUrl
  3. ↵
    Arpa M.C., Hernández P.A. et al. 2013. Geochemical evidence of magma intrusion inferred from diffuse CO2 emissions and fumarole plume chemistry: the 2010–11 volcanic unrest at Taal Volcano, Philippines. Bulletin of Volcanology, 75, 1–13.
    OpenUrl
  4. ↵
    Bani P., Oppenheimer C., Varekamp J.C., Quinou T., Lardy M. & Carn S. 2009. Remarkable geochemical changes and degassing at Voui crater lake, Ambae volcano, Vanuatu. Journal of Volcanology and Geothermal Research, 188, 347–357, https://doi.org/10.1016/j.jvolgeores.2009.09.018
    OpenUrlCrossRefWeb of Science
  5. ↵
    Bartel B.A., Hamburger M.W., Meertens C.M., Lowry A.R. & Corpuz E. 2003. Dynamics of active magmatic and hydrothermal systems at Taal Volcano, Philippines, from continuous GPS measurements. Journal of Geophysical Research, 108, 2475.
    OpenUrlCrossRef
  6. ↵
    Bernard A., Maussen K., Villacorte O., Reniva P., Campita N. & Bornas M. 2013. High CO2 enrichment in surface waters of MCL lake at Taal volcano, Philippines. In: IAVCEI 2013 Scientific Assembly, 20–24 July 2013, Kagoshima, Japan, 1013.
  7. ↵
    Besana G.M., Shibutani T., Hirano N., Ando M., Bautista B., Narag I. & Punongbayan R.S. 1995. The shear wave velocity structure of the crust and uppermost mantle beneath Tagaytay, Philippines, inferred from receiver function analysis. Geophysical Research Letters, 22, 3143–3146.
    OpenUrl
  8. ↵
    Cabassi J., Tassi F. et al. 2013. Biogeochemical processes involving dissolved CO2 and CH4 at Albano, Averno, and Monticchio meromictic volcanic lakes (Central–Southern Italy). Bulletin of Volcanology, 75, 1–19, https://doi.org/10.1007/s00445-012-0683-0
    OpenUrl
  9. ↵
    Capasso G. & Inguaggiato S. 1998. A simple method for the determination of dissolved gases in natural waters. An application to thermal waters from Vulcano island. Applied Geochemistry, 13, 631–642, https://doi.org/10.1016/S0883-2927(97)00109-1
    OpenUrl
  10. ↵
    Capasso G., Favara R. & Inguaggiato S. 2000. Interaction between fumarolic gases and thermal groundwaters at Vulcano Island (Italy): evidences from chemical composition of dissolved gases in waters. Journal of Volcanology and Geothermal Research, 102, 309–318.
    OpenUrl
  11. ↵
    Caudron C., Mazot A. & Bernard A. 2012. Carbon dioxide dynamics in Kelud volcanic lake. Journal of Geophysical Research, 117, B05102, https://doi.org/10.1029/2011JB008806
    OpenUrl
  12. ↵
    Christenson B., Nemeth C., Rouwet D., Tassi F., Meulenbroeck J. & Varekamp J.C. 2015. An introduction to volcanic lakes. In: Rouwet D., Christenson B., Tassi F. & Van de Meulenbroeck J. (eds) Volcanic Lakes. Springer, Heidelberg, 1–20.
  13. ↵
    Christenson B.W. 2000. Geochemistry of fluids associated with the 1995–96 eruption of Mt. Ruapehu, New Zealand: signatures and processes in the magmatic–hydrothermal system. Journal of Volcanology and Geothermal Research, 97, 1–30.
    OpenUrlCrossRefWeb of Science
  14. ↵
    Christenson B.W. & Wood C.P. 1993. Evolution of a vent-hosted hydrothermal system beneath Ruapehu crater lake, New Zealand. Bulletin of Volcanology, 55, 547–565.
    OpenUrlCrossRefWeb of Science
  15. ↵
    Christenson B.W., Reyes A.G., Young R., Moebis A., Sherburn S., Cole-Baker J. & Britten K. 2010. Cyclic processes and factors leading to phreatic eruption events: insights from the 25 September 2007 eruption through Ruapehu Crater Lake, New Zealand. Journal of Volcanology and Geothermal Research, 191, 15–32.
    OpenUrl
  16. ↵
    Colvin A., Rose W.I. et al. 2013. Crater lake evolution at Santa Ana Volcano (El Salvador) following the 2005 eruption. In: Rose W.I., Palma J.L., Delgado Granados H. & Varley N. (eds) Understanding Open-Vent Volcanism and Related Hazards. Geological Society of America, Special Papers, 498, 23–44.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Craig H. & Lupton J.E. 1976. Primordial neon, helium and hydrogen in oceanic basalts. Earth and Planetary Science Letters, 31, 369–385.
    OpenUrlCrossRefWeb of Science
  18. ↵
    Darrah T., Vengosh A., Jackson R.B., Warner N.R. & Poreda R.J. 2014. Noble gases identify the mechanisms of fugitive gas contamination in drinking-water wells overlying the Marcellus and Barnett shales. Proceedings of the National Academy of Sciences of the USA, https://doi.org/10.1073/pnas.1322107111
  19. ↵
    Defant M.J., De Boer J.Z. & Oles D. 1988. The western central Luzon volcanic arc, the Philippines: two arcs divided by rifting? Tectonophysics, 145, 305–317.
    OpenUrlCrossRefWeb of Science
  20. ↵
    Delmelle P. & Bernard A. 1994. Geochemistry, mineralogy and chemical modelling of the acid crater lake of Kawah Ijen Volcano, Indonesia. Geochimica et Cosmochimica Acta, 58, 2445–2460.
    OpenUrlCrossRefWeb of Science
  21. ↵
    Delmelle P., Kusakabe M., Bernard A., Fischer T., De Brouwer S. & Del Mundo E. 1998. Geochemical and isotopic evidence for seawater contamination of the hydrothermal system of Taal Volcano, Luzon, the Philippines. Bulletin of Volcanology, 59, 562–576.
    OpenUrl
  22. ↵
    Fikos I., Vargemezis G. et al. 2012. Electrical resistivity tomography study of Taal volcano hydrothermal system, Philippines. Bulletin of Volcanology, 74, 1821–1831, https://doi.org/10.1007/s00445-012-0638-5
    OpenUrlCrossRef
  23. ↵
    Forster H., Oles D., Knittel U., Defant M. & Torres R. 1990. The Macolod Corridor: a rift crossing the Philippine island arc. Tectonophysics, 183, 265–271.
    OpenUrlCrossRefWeb of Science
  24. ↵
    Galgana G.G., Newman A.V., Hamburger M.W. & Solidum R.U. 2014. Geodetic observations and modeling of time-varying deformation at Taal Volcano, Philippines. Journal of Volcanology and Geothermal Research, 271, 11–23.
    OpenUrl
  25. ↵
    Giggenbach W. 1974. The chemistry of crater lake, Mt Ruapehu (New Zealand) during and after the 1971 active period. Journal of the Royal Society of New Zealand, 17, 33–45.
    OpenUrl
  26. ↵
    Giggenbach W.F. 1987. Redox processes governing the chemistry of fumarolic gas discharges from White Island, New Zealand. Applied Geochemistry, 2, 143–161.
    OpenUrlCrossRef
  27. ↵
    Giggenbach W.F. 1992. Isotopic shifts in waters from geothermal and volcanic systems along convergent plate boundaries and their origin. Earth and Planetary Science Letters, 113, 495–510.
    OpenUrlCrossRefWeb of Science
  28. ↵
    Greinert J., Artemov Y., Egorov V., De Batist M. & McGinnis D. 2006. 1300-m-high rising bubbles from mud volcanoes at 2080m in the Black Sea: hydroacoustic characteristics and temporal variability. Earth and Planetary Science Letters, 244, 1–15, https://doi.org/10.1016/j.epsl.2006.02.011
    OpenUrlCrossRefWeb of Science
  29. ↵
    Greinert J., McGinnis D.F., Naudts L., Linke P. & De Batist M. 2010. Atmospheric methane flux from bubbling seeps: spatially extrapolated quantification from a Black Sea shelf area. Journal of Geophysical Research, 115, C01002, https://doi.org/10.1029/2009JC005381
    OpenUrl
  30. ↵
    Hernández P.A., Mori T., Padrón E., Sumino H. & Pérez N.M. 2011. Carbon dioxide emission from Katanuma volcanic lake, Japan. Earth, Planets, Space, 63, 1151–1156.
    OpenUrl
  31. ↵
    Kumagai H., Lacson R. Jr., Maeda Y., Figueroa M.S. II. & Yamashina T. 2014. Shallow S wave attenuation and actively degassing magma beneath Taal Volcano, Philippines. Geophysical Research Letters, 41, 6681–6688, https://doi.org/10.1002/2014GL061193
    OpenUrl
  32. ↵
    Listanco E. 1993. Space-time patterns in the geologic and magmatic evolution of calderas: a case study at Taal Volcano, Philippines. PhD thesis, University of Tokyo.
  33. ↵
    Maeda Y., Kumagai H., Lacson R., Figueroa M.S. II. & Yamashina T. 2013. Source process of long-period seismic events at Taal Volcano, Philippines: vapor transportation and condensation in a shallow hydrothermal fissure. Journal of Geophysical Research, 118, https://doi.org/10.1002/jgrb.50205
  34. ↵
    Mamyrin B.A. & Tolstikhin I.N. 1984. Helium Isotopes in Nature. Elsevier, Amsterdam.
  35. ↵
    Matsuda J., Matsumoto T. et al. 2002. The 3He/4He ratio of the new internal He Standard of Japan (HESJ). Geochemical Journal, 36, 191–195.
    OpenUrl
  36. ↵
    Mazot A. & Taran Y. 2009. CO2 flux from the volcanic lake of El Chichon (Mexico). Geofisica Internacional, 48, 73–83.
    OpenUrlWeb of Science
  37. ↵
    McGinnis D.F., Greinert J., Artemov Y., Beaubien S.E. & Wüest A. 2006. Fate of rising methane bubbles in stratified waters: how much methane reaches the atmosphere? Journal of Geophysical Research, 111, C09007, https://doi.org/10.1029/2005JC003183
    OpenUrlCrossRef
  38. ↵
    Miklius A., Flower M.F.J., Huijsmans J.P.P., Mukasa S.B. & Castillo P. 1991. Geochemistry of lavas from Taal Volcano, southwest Luzon, Philippines: evidence for multiple magma supply systems and mantle source heterogeneity. Journal of Petrology, 32, 593–627.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    Moore J.G., Nakamura K. & Alcaraz A. 1966. The 1965 eruption of Taal Volcano. Science, 151, 955–960.
    OpenUrlFREE Full Text
  40. ↵
    Mukasa S.B., Flower M.F.J. & Miklius A. 1994. The Nd-, Sr- and Pb-isotopic character of lavas from Taal, Laguna de Bay and Arayat volcanoes, southwestern Luzon, Philippines: implications for arc magma petrogenesis. Tectonophysics, 235, 205–221.
    OpenUrlCrossRefWeb of Science
  41. ↵
    Nishigami K., Shibutani T. et al. 1994. Shallow crustal structure beneath Taal Volcano, Philippines, revealed by the 1993 seismic explosion survey. Bulletin of the Disaster Prevention Research Institute, Kyoto University, 44, 123–137.
    OpenUrl
  42. ↵
    Ohba T., Hirabayashi J. & Nogami K. 2008. Temporal changes in the chemistry of lake water within Yugama Crater, Kusatsu-Shirane Volcano, Japan: implications for the evolution of the magmatic hydrothermal system. Journal of Volcanology and Geothermal Research, 178, 131–144.
    OpenUrl
  43. ↵
    Ostrovsky I., McGinnis D.F., Lapidus L. & Eckert W. 2008. Quantifying gas ebullition with echo sounder: the role of methane transport by bubbles in a medium-sized lake. Limnology and Oceanography: Methods, 6, 105–118, https://doi.org/10.4319/lom.2008.6.105
    OpenUrlCrossRef
  44. Ozima M. & Podosek F.A. 2002. Noble Gas Geochemistry. Cambridge University Press, Cambridge, UK.
  45. ↵
    Padrón E., Pérez N. et al. 2013. Diffusive helium emissions as a precursory sign of volcanic unrest. Geology, 41, 539–542.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    Pasternack G.B. & Varekamp J.C. 1997. Volcanic lake systematics. I.: physical constraints. Bulletin of Volcanology, 58, 528–538.
    OpenUrlCrossRefWeb of Science
  47. ↵
    Pérez N.M., Hernández P.A. et al. 2011. Global CO2 emission from volcanic lakes. Geology, 39, 235–238, https://doi.org/10.1130/G31586.1
    OpenUrlAbstract/FREE Full Text
  48. ↵
    Philippine Institute of Volcanology and Seismology (PHIVOLCS) 1992. Annual Report 1992. PHIVOLCS, Quezon City.
  49. ↵
    Philippine Institute of Volcanology and Seismology (PHIVOLCS) 1993. Annual Report 1993. PHIVOLCS, Quezon City.
  50. ↵
    Philippine Institute of Volcanology and Seismology (PHIVOLCS) 1994. Annual Report 1994. PHIVOLCS, Quezon City.
  51. ↵
    Philippine Institute of Volcanology and Seismology (PHIVOLCS) 2009. Catalog of Philippine Volcanoes. PHIVOLCS, Quezon City.
  52. ↵
    Poreda R. & Craig H. 1989. Helium isotope ratios in circum-Pacific volcanic arcs. Nature, 338, 473–478.
    OpenUrlCrossRef
  53. ↵
    Rouwet D. & Tassi F. 2011. Geochemical monitoring of volcanic lakes. A generalized box model for active crater lakes. Annals of Geophysics, 54, 161–173.
    OpenUrl
  54. ↵
    Rouwet D., Tassi F., Mora-Amador R., Sandri L. & Chiarini V. 2014. Past, present and future of volcanic lake monitoring. Journal of Volcanology and Geothermal Research, 272, 78–97, https://doi.org/10.1016/j.jvolgeores.2013.12.009
    OpenUrl
  55. ↵
    Rowe G.L., Brantley S.L., Fernández M., Fernández J.F., Borgia A. & Barquero J. 1992. Fluid-volcano interaction in an active stratovolcano: the volcanic lake system of Poas Volcano, Costa Rica. Journal of Volcanology and Geothermal Research, 49, 23–51.
    OpenUrlCrossRefWeb of Science
  56. ↵
    Sriwana T., Van Bergen M.J., Varekamp J.C., Sumarti S., Takano B., Van Os B.J.H. & Leng M.J. 2000. Geochemistry of the acid Kawah Putih lake, Patuha Volcano, West Java, Indonesia. Journal of Volcanology and Geothermal Research, 97, 77–104.
    OpenUrlCrossRefWeb of Science
  57. ↵
    Sugisaki R. & Taki K. 1987. Simplified analyses of He, Ne, and Ar dissolved in natural waters. Geochemical Journal, 21, 23–27.
    OpenUrl
  58. ↵
    Sumino H., Nagao K. & Notsu K. 2001. Highly sensitive and precise measurement of helium isotopes using a mass spectrometer with double collector system. Journal of the Mass Spectrometry Society of Japan, 49, 61–68.
    OpenUrlCrossRef
  59. ↵
    Symonds R., Gerlach T.M. & Reed M.H. 2001. Magmatic gas scrubbing: implications for volcano monitoring. Journal of Volcanology and Geothermal Research, 108, 303–341.
    OpenUrlCrossRefWeb of Science
  60. ↵
    Takano B., Saitoh H. & Takano E. 1994. Geochemical implications of subaqueous molten sulfur at Yugama crater lake, Kusatsu-Shirane volcano, Japan. Geochemical Journal, 28, 199–216.
    OpenUrlWeb of Science
  61. ↵
    Taran Y.A., Pokrovsky B.G. & Esikov A.D. 1989. Deuterium and oxygen-18 in fumarolic steam and amphiboles from some Kamchatka volcanoes: ‘andesitic waters’. Doklady Akademii Nauk SSSR, 304, 440–443.
    OpenUrl
  62. ↵
    Tassi F., Vaselli O. et al. 2005. The hydrothermal-volcanic system of Rincón de la Vieja volcano (Costa Rica): a combined (inorganic and organic) geochemical approach to understanding the origin of the fluid discharges and its possible application to volcanic surveillance. Journal of Volcanology and Geothermal Research, 148, 315–333.
    OpenUrlCrossRefWeb of Science
  63. ↵
    Tassi F., Vaselli O., Fernández E., Duarte E., Martínez M., Delgado A. & Bergamaschi F. 2009. Morphological and geochemical features of crater lakes in Costa Rica: an overview. Journal of Limnology, 68, 1–13.
    OpenUrl
  64. ↵
    Torres R.C., Self S. & Punongbayan R.S. 1995. Attention focuses on Taal: decade volcano of the Philippines. EOS, Transactions of the American Geophysical Union, 76, 241–247.
    OpenUrl
  65. ↵
    Varekamp J.C. 2015. The chemical composition and evolution of volcanic lakes. In: Rouwet D., Christenson B.W., Tassi F. & Van de Meulenbroeck J. (eds) Volcanic Lakes. Springer, Heidelberg, 93–124.
  66. ↵
    Varekamp J.C. & Kreulen R. 2000. The stable isotope geochemistry of volcanic lakes with examples from Indonesia. Journal of Volcanology and Geothermal Research, 97, 309–327.
    OpenUrlCrossRefWeb of Science
  67. ↵
    Varekamp J.C., Pasternack G.B. & Rowe G.L. Jr.. 2000. Volcanic lake systematics II. Chemical constraints. Journal of Volcanology and Geothermal Research, 97, 161–179.
    OpenUrlCrossRefWeb of Science
  68. ↵
    Varekamp J.C., Herman S., Ouimette A., Flynn K., Bermudez A. & Delpino D. 2009. Naturally acid waters from Copahue volcano, Argentina. Applied Geochemistry, 24, 208–220.
    OpenUrlCrossRefWeb of Science
  69. ↵
    Worcester D.C. 1912. Taal volcano and its recent destructive eruption. National Geographic, 23, 310–368.
    OpenUrl
  70. ↵
    Yamaya Y., Alanis P.K.B. et al. 2013. A large hydrothermal reservoir beneath Taal Volcano (Philippines) revealed by magnetotelluric resistivity survey: 2D resistivity modelling. Bulletin of Volcanology, 75, 729, https://doi.org/10.1007/s00445-013-0729-y
    OpenUrl
  71. ↵
    Zlotnicki J., Sasai Y. et al. 2009. Combined electromagnetic, geochemical and thermal surveys of Taal volcano (Philippines) during the period 2005–2006. Bulletin of Volcanology, 71, 29–47, https://doi.org/10.1007/s00445-008-0205-2
    OpenUrl
View Abstract
Back to top
Alerts
Citation tools
Permissions
View PDF
Email to
Print
Download PPT

Related Articles

Similar Articles

Cited By...