Projecto Laboratorio Geofisico-Geoquimico Volcán Masaya
Geochemical, geophysical, and petrological studies at Masaya volcano (1997-2000)

Glyn Williams-Jones(1), Pierre Delmelle(2), Peter Baxter(3), Alex Beaulieu(4), Mike Burton(3), Julio Garcia-Alvarez(5), Hélène Gaonac'h(3), Lisa Horrocks(1), Clive Oppenheimer(4), Hazel Rymer(1), David Rothery(1), Katie St-Amand(6), John Stix(6), Wilfried Strauch(5), Benjamin van Wyk de Vries(7)

(1) The Open University, Milton Keynes, UK
(2) Université Catholique de Louvain, Louvain-la-Neuve, Belgium
(3) University of Cambridge, Cambridge, UK
(4) Université du Québec à Montréal, Montreal, Canada
(5) INETER, Managua, Nicaragua
(6) McGill University, Montreal, Canada
(7) Universite Blaise Pascal, Clermont-Ferrand, France

The persistent volcanism and easy access make Masaya volcano an excellent natural laboratory to study important issues such as the potential for basaltic magmas to produce large explosive eruptions, the origin of the long-lived non-eruptive degassing, and the impact of the acid gases on the surrounding environment. The principal goals of this study were to infer the extent and structure of the shallow and deeper magmatic plumbing system, interpret the origin, nature, and time scales of magma movements, characterise the types and amounts of volcanic gases being emitted into the atmosphere, and evaluate the dispersion and deposition of the gas plume downwind from Masaya. The approach adopted to answer these questions included innovative applications of well-established and new geophysical and geochemical techniques, such as microgravity, ground-based remote sensing of gases, ground- and satellite-derived thermal data and melt inclusion petrology.

Deformation and magma movement

Subsurface mass and density changes within the volcano can be quantified and located using a combination of high-resolution ground deformation and microgravity monitoring techniques. Global Positioning System (GPS) measurements allow the vertical and horizontal co-ordinates of a point to be obtained at a precision of 1-3 cm. By making repeat measurements at a network of stations covering the volcano, vertical and horizontal ground deformation through time can be detected. Temporal fluctuations in gravity can also detected at each station and are used to estimate subsurface changes in density or mass. Microgravity measurements also may be affected by the Earth tide (for which we correct) or by changes in ground elevation during inflation/deflation of the volcano (which we can eliminate using GPS for height determination).

Figure 2: Microgravity changes at representative crater rim stations within (Santiago) Masaya caldera measured between February 1993 and April 2000.

Vertical and horizontal movements measured by GPS at Masaya did not exceed 3-4 cm between 1994 and 2000. However, microgravity data reveal substantial variations (Figure 2). Stations located near Santiago crater show a decrease of 90 mGal between February 1993 and April 1994, followed by a gradual increase of up to 56 mGal between April 1994 and March 1997. Recent microgravity surveys between March 1997 and June 1999 show a consistent decrease of up to 120 mGal, of the same order as that observed between 1993 and 1994. The most recent gravity survey suggests, however, that microgravity changes may have levelled off at values within error (± 20 mGal) of those measured in June 1999.

We have also detected short-term gravity variations at Santiago crater. Measurements were made every 15 minutes for 7 hours in 1997, during which gravity increases of 20-45 mGal were observed. In 1998, a more detailed survey was conducted at the same location, for periods of ~13 hours. Atmospheric pressure and seismicity were monitored in conjunction with the microgravity measurements. We observed gravity changes with amplitudes of 40, 20, and 35 mGal on 25 February, 6 March, and 13 March 1998. These fluctuations do not relate to atmospheric pressure and seismicity. However, a tentative correlation appears to exist between the maximum gravity changes and the maximum daily Earth tide amplitude, particularly on 15 February and 13 March 1998. The Earth tide may affect the level of magma in the shallow volcanic conduit, or less likely the density of the magma. However, the important gradients (>100 mGal/yr) between 1996 and 2000 suggest that although there may be some "noise" due to the effect of the Earth tide, they nevertheless represent significant changes in the shallow system immediately beneath Santiago crater.

Figure 3: Microgravity changes at one station near Santiago crater measured over a period of about 13 hours on (a) February 25, 1998 and (b) March 13, 1998. Also shown is the Earth tide amplitude calculated for the same period of time.

Gas plume compositions and gas emission rates

One of the most characteristic phenomena at Masaya is the cycles of prolonged passive degassing. There may have been at least five such cycles of activity at Masaya since the formation of Santiago crater (1859). The most recent cycle of activity began in June 1993 and continues to date. Although direct sampling of the gases emitted from Santiago crater is not practical, their remote detection from the ground is easily performed using the network of roads surrounding Masaya caldera. The column amounts of sulphur dioxide (SO2) can be monitored with a correlation spectrometer (COSPEC) driven beneath the ash-free plume and approximately perpendicular to its direction. In these calculations, the plume velocity is assumed to correspond to the wind velocity (measured at the INETER seismic station in El Crucero). Most of our COSPEC measurements were conducted on the Pan-American highway, which runs partly along the Llano Pacaya ridge at a distance of about 15 km from Santiago (Figure 1). Since March 1996, we have conducted more than 460 measurements of the SO2 emission rate at Masaya (Table 1). The standard deviation for a single day of measurements (comprising between 6 and 15 traverses) was typically less than 30%.

Since 1996, the SO2 emission rate at Masaya has fluctuated significantly (Figure 3). It averaged about 600 ± 290 metric tonnes per day (t d-1) in March 1996, 390 ± 200 t/d in February-March 1997, 1870 ± 960 t/d in February-April 1998, 670 ± 520 t/d in September 1998, 1650 ± 560 t/d in February-March 1999, and 1310 ± 430 t/d in March-April 2000. The relatively large standard deviation associated with each period of observation results from instrument errors and plume-atmosphere interactions. For example, wet weather and low wind velocity conditions hampered COSPEC measurements in September 1998. Transient changes in source strength (i.e. changes in the amount of gas being released at depth) also may contribute to the variations. In particular, under normal operation conditions, we measured SO2 emission rates above 3000 t/d on a few days in February-April 1998 and February-March 1999 (Table 1).

Figure 4: Daily average SO2 emission rate at Masaya volcano measured by COSPEC between March 1996 and April 2000. Error bars represent ± 1s standard deviation.

Supplementing the COSPEC measurements of February-March 1998, March 1999 and March-April 2000, we determined the gas plume composition by open-path Fourier transform infrared (FTIR) spectroscopy. The power of the FTIR technique lies in the ability to measure ratios of gas concentrations and their changes through time. FTIR spectral data for the gas plume were collected from the rim of Santiago either actively using a silicon carbide lamp or passively using the Sun as a source of infrared radiation. The full complement of gases retrieved from the FTIR spectra includes SO2, CO2, hydrogen fluoride (HF), and hydrogen chloride (HCl). The FTIR measured SO2/HCl, CO2/SO2, and HCl/HF mass ratios of 2.8, 1.7, and 8.2, respectively, in March-April 1998, and 2.9, 1.6, and 8.2 in March 1999. Similar values were observed in March-April 2000. These ratios are typical of high-temperature gases released by convergent plate volcanoes. Remarkably, the FTIR-derived SO2/HCl gas ratios were constant during each period of study, while the HCl/HF gas ratios displayed more variations. Changes in the HCl/HF ratio may reflect different physico-chemical behaviour of the gases entering the atmosphere and/or uncertainties linked to the spectral data reduction procedure for HF.

By combining the COSPEC measurements with these gas ratios, we deduce emission rates for CO2, HCl, and HF of 3150 t/d, 670 t/d, and 90 t/d, respectively, for the March-April 1998 FTIR results, and 2800 t/d, 610 t/d, and 70 t/d for the March 1999 FTIR results. Although simultaneous determinations of gas emissions exist for only a few subaerial volcanoes, our data suggest that passive degassing at Masaya contribute large amounts of carbon, sulphur, and halogen gases to the atmosphere.

Downwind dispersion and deposition of the SO2 plume

The large amount of SO2 released in the atmosphere by Masaya produces volcanic air pollution manifested by poor air quality, hazy atmospheric conditions, and acid rain. We monitored the average dispersion of the SO2-bearing plume with a network of diffusion tubes. These passive samplers were exposed during about four weeks in March-April 1998 and February-March 1999. The 1999 surveys reveal that Masaya's plume affects a region of ~1250 km2 (Figure 4). The atmospheric SO2 concentrations measured within about 20 km of the volcano compare to, or exceed largely those reported in urbanized areas and downwind of industrial point sources. The U.S. National Air Quality Standard over the year is 30 ppbv of SO2. Concentrations above this threshold value occur within a 400-km2 area -including El Crucero- inhabited by several thousand people. Importantly, the diffusion tube data are not capable of revealing short-term (hours to days) concentration peak excursions which could be much higher than the observed time averages, thereby increasing the hazard. Elevated levels of atmospheric SO2 may induce breathing difficulties in people exposed to it, with those suffering from asthma being particularly sensitive even at low concentrations.

SO2 may also damage vegetation. Long-term exposures to concentrations up to 50 ppbv can alter the composition of plant communities, decrease the productivity of agricultural systems, and cause visible foliar injury. Such devastating effects are conspicuous immediately behind the western caldera wall and in the Llano Pacaya area that is subjected to intense fumigation. This area used to be planted for coffee and citrus fruit until a volcanic gas crisis in 1924 destroyed most of the plantations. Nowadays, only a few herb species survive in the part of Llano Pacaya most exposed to the volcanic gases. In addition, the volcanogenic pollution also is held responsible for damage to machinery and buildings.

Figure 5: Contour maps showing the dispersion of the SO2 plume measured in February-March 1999 downwind from Masaya volcano. Near-ground SO2 concentrations are given in ppbv. Contour interval is 25 ppbv.

In conjunction with the atmospheric SO2 measurements, the dry deposition of the volcano-derived acid emissions was monitored using lead dioxide-coated sulfation plates. The total dry deposition of SO2 measured within 44 km from Masaya volcano in February-March 1999 corresponds to a daily mean of 1.5 x 105 kg. Of this quantity, only about 1.4 % comes from the background deposition of SO2, and therefore the remainder is derived from volcanic emissions. Thus, based on the present data, about 8% of the total mass of the SO2 emitted daily by the volcano (i.e. 1800 t day-1) is deposited within 44 km of the source. It follows that more than 90% of the SO2 emitted is transported beyond the Pacific Coast. Besides SO2, we also detected the dry deposition of HCl onto the sulfation plates. We calculated a total deposition of 5.7 x 104 kg day-1 for HCl which corresponds to about 9 % of the total HCl mass emitted daily by the volcano. Although these numbers are preliminary, they suggest that the degassing activity at Masaya produces extreme acid loading (SO2 and HCl together) in the areas exposed to the gas plume. This may have dramatic effects on soil acidification processes and functioning of the ecosystems in the Masaya area.

Petrological observations

Lava bomb fragments ejected by a small explosive eruption on 17 November 1997 at Santiago crater were analysed by electron microprobe to determine the bulk chemistry and the sulphur (S) and chlorine (Cl) contents. The glass in both the matrix and phenocryst inclusions is a tholeiitic basalt with a composition similar to juvenile material erupted by Masaya in historic times. Melt inclusions in phenocrysts contain an average of 100 ± 50 ppm S and 440 ± 60 ppm Cl as compared with 30 ± 20 ppm and 370 ± 40 ppm, respectively, in the matrix glass. For an arc basaltic system, such S contents are low and are unlikely to reflect the deep magma conditions, but rather suggest partial degassing of magma before melt entrapment in phenocrysts. In addition, the generally lower S and Cl contents in the matrix glasses suggest that degassing continues to occur after melt entrapment. The S/Cl ratio varies from 0.04 to 0.57 in the melt inclusions and from less than 0.01 to 0.30 in the matrix glasses. While some of this variability arises from the low S contents (at or near the detection limit), the generally lower S/Cl value in the matrix glasses is consistent with preferential fractionation of S with respect to Cl in the vapour phase during gas exsolution from the magma.

An integrated model of magma degassing at Masaya

Episodes of strong magma degassing occur repeatedly at Masaya. However their triggering mechanism is poorly understood. The onset of the current gas crisis was preceded by a decrease in gravity at stations surrounding Santiago crater, which was accompanied by significant increase in gas emissions (Figures 2 and 4). These changes are probably not coincidental and reveal different aspects of the same process, or processes, occurring beneath the crater. It has been suggested that overturn of magma within a magma chamber allowed replacement of a high-density, degassed magma by a vesiculated (lower-density) gas-rich magma at shallow levels. Another gravity decrease and gas emission increase trend between March 1997 and June 1999 is likely related to the same mechanism as in 1993-1994. Significantly, the apparent levelling off of gravity change and slight decrease in SO2 gas flux suggests that the end of the current degassing crisis may be imminent.

Thus, it seems that the shallow magmatic plumbing system can be supplied periodically with buoyant, gas-rich magma from deeper levels. The SO2 emission rate and ground- and satellite-based remote sensing of the thermal radiance from Santiago vent further constrain the model. We estimate that approximately 1.2 x 109 kg of sulphur has been emitted at Masaya from mid-1993 through March 1999. If magma releases about 500 ppm sulphur (more reasonable than the 370 ppm measured in the 1997 bomb) during passive degassing, then petrological calculations suggest that about 0.3 km3 of magma has been degassed at shallow levels since the reactivation. However, the presence of such a large magma volume near the surface is not consistent with geophysical data. We believe that gas-rich material continually arrives at the shallow vesiculated layer where it degasses (with minimal cooling) and then sinks to be replaced by gas-rich magma rising from depth.

The magmatic plumbing system at Masaya may be viewed as a near-surface magma body linked to, and periodically replenished by, a long-lived crustal magma reservoir. Once buoyant, gas-rich magma from deep levels is emplaced near the surface, it tends to convect and degas vigorously. The steady nature of the convection process is inferred from the notable stability over different time scales of the SO2/HCl gas ratio measured by FTIR. This suggests that the gases are released continuously under stable conditions over a fixed depth range. The similarity in SO2 emission rates obtained by COSPEC in February-April 1998, February-April 1999, and March-April 2000 is consistent with this view, although fluctuations over short time scales have been measured.

Conclusions

This multidisciplinary collaboration is improving our understanding of activity at Masaya volcano. However, many interesting questions remain to be answered. The observed fluctuations in gravity over long time scales (several years) at Santiago crater may represent drastic changes in the vesicularity of the shallow magma, which ultimately may relate to the physical state of magma stored in a long-lived crustal reservoir. In contrast, the short time scale variations (hours) in gravity may indicate changes in the level of the near-surface magma in relation to non-magmatic processes such as the Earth tide. Clearly, quantitative assessment of the relationship linking gravity and vesicularity on one hand, and gravity and Earth tide on the other hand is necessary if variations in magmatic activity at Masaya are to be fully understood. Combined geochemical, gas emission and composition, and thermal radiance data sets also will allow us to better constrain evolution of the magmatic system, including sizes and depths of reservoirs involved and degassing processes. Finally, the use of the FTIR and COSPEC in conjunction with passive monitoring techniques at different distances downwind of the volcano summit opens an exciting field of investigations to understand the atmospheric chemistry involved in plume dispersal and to address the nature of the apparent short-time scale variations in gas emission rates.

Acknowledgements

We would like to thank all the staff of the Departamento de Geofisica (INETER) and the Parque Nacional del Volcán Masaya (MARENA) for their invaluable and generous help.