Completo_DEF.qxd 1501 ANNALS OF GEOPHYSICS, VOL. 47, N. 4, August 2004 Key words volcanic gases – volcanic hazard – geo- chemical monitoring – geophysical monitoring – Turrialba Volcano – Costa Rica 1. Introduction Turrialba (10.03N, 83.77E) is an active strato- volcano belonging to the Holocene «Cordillera Central» volcanic belt and is located at about 35 and 15 km from the towns of San José and Carta- go, respectively (fig. 1). This region is part of the central Mountain Range Conservation Area and includes seven volcanic edifices. It was colonized by the Indians of the Oriental Huetar Kingdom and later by the Spaniards. Originally this volcano was named «Torre Alba» (White Tower) by the Spaniards, due to the presence of a large plume that looked like an enormous white tower. Since its last eruption (16 September 1864), Turrialba Volcano has been characterized by a discontinu- ous fumarolic activity. A small plume, visible from the immediate surroundings of the volcano, has occasionally been reported. In the last decade, the presence of continuous weak fumarolic dis- charges has been observed. In the present study, the results of a geochem- ical monitoring carried out from 1998 to 2002 Mailing address: Dr. Franco Tassi, Departimento di Scien- ze della Terra, Università degli Studi di Firenze, Via G. La Pi- ra, 4, 50121, Firenze, Italy; e-mail: francot@steno.geo.unifi.it Fluid geochemistry and seismic activity in the period 1998-2002 at Turrialba Volcano (Costa Rica) Franco Tassi (1), Orlando Vaselli (1), Vilma Barboza (2), Erick Fernandez (2) and Eliecer Duarte (2) (1) Dipartimento di Scienze della Terra, Università degli Studi di Firenze, Italy (2) Volcanological and Seismological Observatory, Universidad Nacional, Heredia, Costa Rica Abstract Turrialba Volcano, located in Central-Southern Costa Rica, has been characterized, since the last period of eruptive activity in 1884-1886, by a weak and discontinuous fumarolic activity in the western area of its summit. During the 1998-2002 period, fumaroles discharging from central and West craters were collected for chemical analyses of ma- jor and trace gas compounds, 13C/12C in CO2 and 18O/16O and D/H (in one fumarolic condensate), isotopic ratios. Geo- physical measurements (seismic activity and ground deformation), monitored in the same period, were compared to geochemical data to define the status of the volcanic system. Chemical and isotopic characteristics of fumaroles of Turrialba Volcano seem to be related to interaction processes between a magmatic source and a shallower hydrother- mal aquifer. Since February 1997, seismicity at Turrialba Volcano gradually increased, while since August 2001 new fumaroles start to discharge from a new fracture system located in the area between central and West craters. Since September 2001, strong compositional changes of gas discharges have been recorded at central crater. These occur- rences are possibly due to variations in the permeability of the conduit system feeding the fumaroles. Heat pulse episodes from a magmatic source have possibly caused the increase of vapour pressure at depth and, consequently, favoured the uprising of the magmatic fluids toward the surface. The observed evolution of chemical and physical parameters suggests that to forecast a possible renewal of the volcanic activity in the near future a full program of both geochemical and geophysical surveillance must be provided at Turrialba Volcano. 1502 Franco Tassi, Orlando Vaselli, Vilma Barboza, Erick Fernandez and Eliecer Duarte on the fumaroles of the summit craters are de- scribed and discussed. Data from a geophysical monitoring of the seismic activity, carried out by the Volcanological and Seismological Ob- servatory of the National University of Costa Rica, are also included. Thus, the main goals of this work are: i) to investigate the origin of the fumarolic discharges; ii) to evaluate the evolu- tion in time of both chemical parameters and seismic activity and iii) to assess the status of activity of the volcanic system. 2. Geological and volcanological background Costa Rica consists of six main geologic provinces (Castillo, 1984): 1) Cretaceous-Middle Tertiary ophiolitic suite; 2) Tertiary basins; 3) Ter- tiary volcanic ranges; 4) Active Quaternary vol- canic ranges; 5) Intra-arc basins; 6) Caribbean coastal plain (fig. 1). The active Quaternary vol- canic ranges comprise the Cordillera de Gua- nacaste to NW and the Cordillera central to SE (fig. 1). The volcanic activity in Costa Rica in the last 25 000 years has been characterized mainly by phreatic and phreato-magmatic eruptions, with rare plinian events (Alvarado et al., 1992) and is related to the subduction of the Cocos plate un- derneath the Caribbean plate (e.g., Johnston and Thorkelson, 1997 and references therein), with a subduction angle of about 30° (Protti et al., 1995). Turrialba is a composite stratovolcano (3,325 meters a.s.l.) and belongs to the Cordillera cen- tral where two more active volcanoes with his- torical eruptions are present: Irazu and Poàs. This volcanic range also includes Platanar, Con- go and Barva volcanoes. Historic eruptive activ- ity of Turrialba dates back to 1723, 1847, 1853, 1855, 1861 and 1864-1866 (Castillo, 1984). They were probably small phreatic eruptive events, but very little is currently known about these volcanic activities. The last eruption (1864-1866) was characterized by the emission of large volumes of ashes that, during the period of major activity (February 1866), extended up to Nicaragua (OVSICORI-UNA, 1998). The West crater on the North-western side of the volcano summit (fig. 2) was formed during this eruptive event. In 1920 and in the period 1950-1960 local newspapers reported that large columns of vapour were observed from the city of Cartago, showing that episodes of strong fumarolic emis- sion have occurred several times in the last cen- tury. Turrialba summit presently hosts three NE- SW oriented craters: East, central and West Fig. 1. Schematic geological map of Costa Rica. The Middle America Trench, the Caribbean Plate, the Cocos Plate, the Cordillera de Guanacaste vol- canic arc, the Cordillera central volcanic arc and the location of the Turrialba Volcano are reported. Fig. 2. Map of the summit craters of Turrialba Vol- cano with the location of principal, central and Que- mada fumaroles. 1503 Fluid geochemistry and seismic activity in the period 1998-2002 at Turrialba Volcano (Costa Rica) craters (fig. 2). Sulphur deposits of the central crater had been exploited till the end of the 80’s. The age of the craters decreases from East to West and the central and West craters currently have fumarolic emissions. 3. Sampling and analytical procedures of gas discharges Gas discharge sampling points are shown in fig. 2. Fumaroles from both the central and West craters are characterised by a relatively low flux. They seep out, mainly as diffuse emis- sions, from areas of few tens of square meters and marked by an intense rock alteration. Gas sampling of the West crater fumarole started in February 1998, whereas that of the central crater in February 1999, due to the fact that pre- viously the flux of this gas discharge was too low and diffuse. The Quemada fumarolic field, formed during summer 2001 but sampled only in April 2002, discharges from a large fracture (about 30 m long and 5 m wide) located be- tween the West and central craters (fig. 2) where burning effects on the nearby vegetation were produced by the hot gas emission. Gas sampling was performed with a titani- um tube inserted into the gas outlets. Gases were convoyed, through pyrex glass dewars, in- to pre-weighted and pre-evacuated 50 ml thori- on-tapped vials to which 20 ml of a 0.15 M Cd(OH)2 and 4 M NaOH solution was previ- ously added. Here, acid gases (CO2, SO2, H2S, HCl, and HF) dissolve into the alkaline solu- tion, while the unabsorbed gases (N2, O2, CO, He, H2, CH4, Ar, Ne and light hydrocarbons) concentrate in the head-space. Sulphur from H2S immediately precipitate as CdS (Monte- grossi et al., 2001). Major and trace inert gas compounds were analysed with Shimadzu 15A and Shimadzu 14A gas-chromatographs equipped with TCD (Thermal Conductivity De- tector) and FID (Flame Ionisation detector), re- spectively. CO was determined, as CH4, after its conversion at 400°C (Shimadzu MTN-1 metha- nizer), using H2 as carrier gas to allow the CO- CH4 Fischer-Trops reaction (in presence of Shi- malite-Ni acting as catalyser). The complete separation of Ar-O2 and He-Ne-H2 peaks was performed at low temperature (about 0°C, ob- tained with a Shimadzu CRG-15 cryogenic equipment) by using a 9 m length molecular sieve column. CO2, trapped in the caustic solu- tion of the sampling vials, was determined by acid titration with 0.5 M HCl, while SO2, H2S, S0 and HCl abundances were performed by ion- chromatography (Dionex DX100 ion-chro- matograph), as described in Montegrossi et al. (2001). HF contents in the condensates (col- lected at the «bulbe» temperature by the gentle dropping from glass tubes wrapped with paper tissue) were determined by potentiometry with a specific electrode. Analytical precision was < 1% for major components and < 5% for mi- nor and trace compounds. 13C/12C ratios of CO2 (expressed as 0/00 PDB) were determined with a Finningan Delta S mass spectrometer after the standard procedures of ex- traction and purification of the gas mixtures (Evans et al., 1998). The reproducibility of δ-val- ues for C is ± 0.10/00. δ18O and δD values (ex- pressed as 0/00 SMOW) were determined with a Finningan Delta S mass spectrometer after equil- ibrating the waters with CO2 and after the reac- tion of 10 ml of water with 0.3 g of pure Zn at 500°C, respectively (Vaselli et al., 1997, 1998). 4. Gas chemistry and isotopic composition The analytical results for chemical and iso- topic composition of the fumarolic discharges from the West crater (10 gases sampled from February 1998 to April 2002), central crater (6 gases sampled from February 1999 to April 2002) and Quemada fracture (sampled in April 2002) are shown in tables I and II. Fumarole temperatures ranged between 86 and 92°C, with no remarkable differences among the three sites of gas discharges. Gas chemical composition is characterized by the predomi- nance of H2O (up to 980 000 µmol/mol) and CO2 (up to 580 000 µmol/mol), with less N2, H2S, HCl, and H2 contents (6700, 2400, 570 and 428 µmol/mol, respectively). S0, HF, CO, He, Ne, CH4 and light hydrocarbons are pres- ent as trace compounds. Ar contents are < 11 µmol/mol, indicating a very limited air-con- tamination on the composition of gas samples. 1504 Franco Tassi, Orlando Vaselli, Vilma Barboza, Erick Fernandez and Eliecer Duarte T ab le I. C he m ic al c om po si ti on o f in or ga ni c ga s co m po un ds a nd δ 13 C , δ1 8 O a nd δ D v al ue s of t he T ur ri al ba V ol ca no f um ar ol es . G as c om po un d co n- te nt s ar e in µ m ol /m ol . n. a. : no t an al ys ed . T °C C O 2 H C l H F S O 2 H 2S S H 2O N 2 C H 4 A r O 2 N e H 2 H e C O δ 13 C δ 18 O δD W es t 02 -1 99 8 91 17 74 8 40 .5 2. 73 3 < 0. 01 19 2. 2 < 0. 00 1 98 15 51 41 1. 1 0. 00 7 1. 04 1. 0 0. 00 11 51 .9 6 0. 37 7 0. 00 23 n. a. –4 .7 6 –5 5. 1 W es t 02 -1 99 9 90 19 40 4 87 .4 0. 49 4 < 0. 01 83 .8 < 0. 00 1 97 99 85 41 0. 0 0. 04 3 2. 68 1. 1 0. 00 25 24 .6 0 0. 56 7 < 0. 00 05 – 2. 71 n. a. n. a. W es t 02 -2 00 0 91 11 26 5 20 5. 7 0. 24 4 < 0. 01 51 .6 5. 14 4 98 80 26 37 8. 7 0. 01 4 2. 39 1. 3 0. 00 39 63 .6 4 0. 34 4 < 0. 00 05 – 3. 07 n. a. n. a. W es t 02 -2 00 1 92 21 21 9 7. 8 0. 21 4 < 0. 01 44 .4 0. 26 1 97 83 26 29 9. 0 0. 01 3 1. 72 17 .4 0. 00 10 83 .8 0 0. 35 9 < 0. 00 05 n. a. n. a. n. a. W es t 09 -2 00 1 90 41 57 9 4. 1 0. 24 5 < 0. 01 20 .7 0. 50 3 95 80 75 22 3. 7 0. 01 5 0. 35 1. 9 0. 00 02 94 .3 5 0. 25 1 < 0. 00 05 n. a. n. a. n. a. W es t 11 -2 00 1 88 50 84 3 6. 4 0. 37 4 < 0. 01 34 .8 0. 43 9 94 86 65 31 3. 1 0. 01 6 0. 47 3. 3 0. 00 04 13 3. 47 0. 31 5 < 0. 00 05 n. a. n. a. n. a. W es t 01 -2 00 1 90 51 54 9 21 .3 1. 51 2 < 0. 01 14 .5 0. 64 5 94 80 13 20 9. 6 0. 01 9 0. 35 1. 8 0. 00 02 90 .7 8 0. 22 0 < 0. 00 05 n. a. n. a. n. a. W es t 06 -0 2- 20 02 90 55 92 3 91 .9 4. 83 8 < 0. 01 14 .4 0. 49 9 94 33 42 54 0. 3 0. 01 2 0. 99 14 8. 7 0. 00 06 44 .5 4 0. 53 1 < 0. 00 05 n. a. n. a. n. a. W es t 19 -0 2- 20 02 90 46 25 2 11 1. 3 5. 30 1 < 0. 01 17 .6 0. 51 1 95 29 91 45 6. 8 0. 01 8 0. 76 98 .7 0. 00 06 71 .1 2 0. 18 4 0. 00 13 – 2. 7 n. a. n. a. P ri nc ip al 0 4- 20 02 90 63 40 2 2. 8 0. 14 5 < 0. 01 44 .5 0. 30 6 93 63 57 14 2. 1 0. 01 4 0. 28 1. 0 0. 00 03 49 .6 6 0. 20 9 < 0. 00 05 – 2. 83 n. a. n. a. C en tr al 0 2- 19 99 89 20 42 7 17 .6 0. 09 8 < 0. 01 28 9. 3 < 0. 00 1 97 80 14 11 98 .0 0. 05 3 10 .0 7 37 .8 0. 00 97 3. 40 2. 13 5 < 0. 00 05 n. a. n. a. n. a. C en tr al 0 2- 20 01 90 19 40 3 6. 4 < 0 .0 01 < 0. 01 16 .3 0. 21 8 97 95 08 85 7. 1 0. 00 3 8. 05 20 1. 3 0. 00 60 0. 12 0. 03 7 < 0. 00 05 n. a. n. a. n. a. C en tr al 0 9- 20 01 89 18 12 98 31 .1 1. 01 9 < 0. 01 10 5. 0 2. 88 9 81 71 08 10 10 .5 0. 07 0 2. 12 12 .2 0. 00 12 42 8. 11 1. 25 2 < 0. 00 05 n. a. n. a. n. a. C en tr al 1 1- 20 01 86 58 73 67 25 2. 4 8. 26 8 70 03 8. 45 24 04 .6 29 .2 62 33 31 15 67 04 .0 0. 08 9 10 .5 2 61 .9 0. 00 61 13 .7 8 2. 97 1 < 0. 00 05 – 2. 72 n. a. n. a. C en tr al 0 2- 20 02 89 19 22 23 51 0. 2 5. 93 2 15 82 4. 60 34 5. 9 4. 23 3 78 74 61 30 12 .0 0. 03 2 8. 10 59 4. 2 0. 00 65 20 .3 9 0. 62 2 0. 14 46 – 2. 55 n. a. n. a. C en tr al 0 4- 20 02 89 32 40 20 24 5. 5 33 .2 43 34 64 7. 04 13 82 .8 1. 22 9 63 85 91 11 04 .2 0. 02 2 1. 94 2. 0 0. 00 19 2. 84 1. 66 6 0. 06 65 n. a. n. a. n. a. Q ue m ad a 04 -2 00 2 90 85 32 0 10 .6 0. 78 4 19 38 .0 2 25 .2 0. 25 2 91 20 58 61 3. 8 0. 09 0 2. 02 7. 0 0. 00 14 25 .3 8 0. 31 8 < 0. 00 05 n. a. n. a. n. a. 1505 Fluid geochemistry and seismic activity in the period 1998-2002 at Turrialba Volcano (Costa Rica) T ab le II . C he m ic al c om po si ti on o f or ga ni c ga s co m po un ds o f T ur ri al ba V ol ca no f um ar ol es . G as c om po un d co nt en ts a re µ m ol /m ol . C 2H 6 C 2H 4 C 3H 8 C 3H 6 i- C 4H 10 n- C 4H 10 1- C 4H 8 i- C 4H 8 t- 2- C 4H 8 c- 2- C 4H 8 i- C 5H 12 n- C 5H 12 C 5H 10 n- C 6H 12 n- C 7H 14 n- C 8H 16 C 6H 6 C 7H 8 C 8H 10 m -C 8H 10 W es t 02 -1 99 8 0. 00 05 8 < 0. 00 00 5 0. 00 02 3 0. 00 02 3 0. 00 01 4 0. 00 05 3 0. 00 05 3 0. 00 06 6 0. 00 03 2 < 0. 00 00 5 0. 00 11 4 0. 00 02 3 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 06 0 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 W es t 02 -1 99 9 < 0. 00 00 5 < 0. 00 00 5 0. 00 05 1 0. 00 07 9 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 69 1 < 0. 00 00 5 0. 00 21 9 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 39 2 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 W es t 02 -2 00 0 0. 00 17 6 0. 00 04 7 0. 00 05 5 0. 00 09 1 0. 00 00 5 0. 00 01 6 0. 00 12 6 0. 00 00 5 0. 00 01 6 0. 00 10 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 02 1 < 0. 00 00 5 < 0. 00 00 5 0. 00 13 4 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 W es t 02 -2 00 1 0. 00 09 0 0. 00 16 9 < 0. 00 00 5 0. 00 18 1 < 0. 00 00 5 < 0. 00 00 5 0. 00 45 0 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 04 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 W es t 09 -2 00 1 0. 00 18 7 0. 00 03 2 0. 00 10 1 0. 00 06 8 0. 00 03 3 0. 00 03 4 0. 00 01 8 0. 00 07 8 < 0. 00 00 5 < 0. 00 00 5 0. 00 02 4 0. 00 01 1 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 01 1 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 W es t 11 -2 00 1 0. 00 20 4 0. 00 03 4 0. 00 11 5 0. 00 06 5 0. 00 04 8 0. 00 03 9 0. 00 01 5 0. 00 09 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 02 5 0. 00 01 0 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 01 0 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 W es t 01 -2 00 1 0. 00 28 2 < 0. 00 00 5 0. 00 16 8 0. 00 08 5 0. 00 05 4 0. 00 06 7 < 0. 00 00 5 0. 00 26 6 0. 00 00 7 0. 00 00 9 0. 00 05 3 0. 00 00 9 < 0. 00 00 5 < 0. 00 00 5 0. 00 00 6 0. 00 00 7 0. 00 01 6 0. 00 00 5 0. 00 00 5 0. 00 00 9 W es t 06 -0 2- 20 02 0. 00 14 1 < 0. 00 00 5 0. 00 07 6 0. 00 03 9 0. 00 01 8 0. 00 03 0 < 0. 00 00 5 0. 00 07 0 0. 00 02 0 0. 00 00 6 0. 00 01 8 0. 00 02 0 < 0. 00 00 5 < 0. 00 00 5 0. 00 00 5 0. 00 00 5 0. 00 00 9 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 W es t 19 -0 2- 20 02 0. 00 23 3 < 0. 00 00 5 0. 00 13 7 0. 00 08 0 0. 00 04 3 0. 00 03 6 < 0. 00 00 5 0. 00 15 2 0. 00 00 8 0. 00 01 0 0. 00 02 0 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 00 5 0. 00 00 5 0. 00 01 7 0. 00 00 5 < 0. 00 00 5 0. 00 00 6 W es t 04 -2 00 2 0. 00 14 4 < 0. 00 00 5 0. 00 07 9 0. 00 04 9 0. 00 03 0 0. 00 02 6 < 0. 00 00 5 0. 00 05 5 0. 00 00 5 < 0. 00 00 5 0. 00 02 2 0. 00 01 1 0. 00 00 5 0. 00 02 5 0. 00 00 7 0. 00 00 6 0. 00 04 6 0. 00 00 6 0. 00 00 5 0. 00 00 6 C en tr al 0 2- 19 99 < 0. 00 00 5 0. 00 14 9 0. 00 00 0 < 0. 00 00 5 0. 00 20 3 0. 00 10 8 < 0. 00 00 5 0. 01 16 3 < 0. 00 00 5 0. 00 09 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 73 0 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 C en tr al 0 2- 20 01 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 C en tr al 0 9- 20 01 0. 00 93 9 0. 00 16 2 0. 00 50 3 0. 00 32 7 0. 00 15 6 0. 00 16 5 0. 00 04 9 0. 00 52 4 0. 00 04 4 < 0. 00 00 5 0. 00 13 2 0. 00 10 8 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 08 1 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 C en tr al 1 1- 20 01 0. 00 51 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 47 3 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 C en tr al 0 2- 20 02 0. 00 36 5 < 0. 00 00 5 0. 00 11 6 0. 00 09 4 0. 00 05 4 < 0. 00 00 5 < 0. 00 00 5 0. 00 34 3 < 0. 00 00 5 0. 00 09 0 0. 00 12 3 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 00 5 0. 00 01 2 0. 00 03 5 0. 00 01 2 0. 00 00 8 0. 00 00 5 C en tr al 0 4- 20 02 0. 00 09 0 0. 00 01 6 0. 00 03 6 0. 00 02 3 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 06 2 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 < 0. 00 00 5 0. 00 00 7 0. 00 00 6 0. 00 00 5 0. 00 03 9 0. 00 00 5 0. 00 00 6 0. 00 00 5 Q ue m ad a 04 -2 00 2 0. 00 41 8 < 0. 00 00 5 0. 00 08 4 0. 00 23 4 0. 00 04 2 0. 00 03 1 < 0. 00 00 5 0. 00 24 9 0. 00 03 0 0. 00 01 9 0. 00 02 8 0. 00 01 4 0. 00 02 6 0. 00 17 4 0. 00 05 1 0. 00 01 9 0. 00 20 5 0. 00 04 6 0. 00 11 6 0. 00 05 1 1506 Franco Tassi, Orlando Vaselli, Vilma Barboza, Erick Fernandez and Eliecer Duarte The wide variability of O2 contents (ranging from 1 to 594 µmol/mol) is possibly related to its partial consumption during fluid circulation, since it is involved in redox reactions with both the other gas compounds and the hosting rocks. SO2 in the West gas samples is below the de- tection limit (< 0.01 µmol/mol). On the con- trary, since November 2001, SO2 has been de- tected in the central fumarole (up to 70 000 µmol/mol). Quemada fumarole shows an inter- mediate composition between the West and the central fumaroles (tables I and II). δ 13C in CO2 range between – 3.1 and – 2.7 0/00 PDB, with no remarkable difference among the different gas discharges. δ 18O and D/H ratios, analysed in the condensate of the gas sample col- lected in 1998 at the West fumarole, are – 4.76 and – 55.1 0/00 SMOW, respectively. 5. Discussion 5.1. Origin of gases The chemical composition of Turrialba gas discharges is relatively enriched in the less solu- ble acidic compounds, such as CO2 and H2S, al- though the presence of HCl, SO2 and HF (table I) indicates that, at least partly, the magmatic sig- nature still remains. The relative abundances of N2, Ar and He are plotted in fig. 3, where basaltic magmatism («mantle»), convergent plate bound- aries («andesite») and crust fields («crust») are also reported (Giggenbach, 1996). N2/Ar ratios of all the samples (up to 640) are higher than that in the air (83), suggesting a non-atmospheric ori- gin for N2, probably partly released from sub- ducted sedimentary material, as commonly ob- served for gases collected in areas of convergent active tectonics (Giggenbach, 1996), although He relative contents seem to differ sensitively from the typical volcanic arc gas compositions (Sano and Wakita, 1985; Sano et al., 1987). As shown in fig. 4, Ne/Ar ratios are between those of air and Air Saturated Water (ASW), suggest- ing a common atmospheric source for these two compounds. Concerning the organic gas frac- tion, up to 20 different hydrocarbon compounds pertaining to the C2-C8 species (the so-called «light hydrocarbons») were detected. Relative Fig. 4. Ar-N2/100-1000Ne ternary diagram for the Turrialba fumaroles. Air and Air Saturated Water (ASW) compositions were reported. Symbols as in fig. 3. Fig. 3. N2/100-Ar-10He ternary diagram for the Turrialba fumaroles. Basaltic magmatism («man- tle»), convergent plate boundaries («andesite») and crust fields («crust») are also reported (Giggenbach, 1996). ASW: Air Saturated Water open squares: Principal crater fumarole; open up triangles: central crater fumarole; open down triangles: Quemada fu- marole. 1507 Fluid geochemistry and seismic activity in the period 1998-2002 at Turrialba Volcano (Costa Rica) high contents of hydrocarbons pertaining to both aromatic and alkene groups were systematically observed in fumaroles from volcanic areas and geothermal fields with respect to those measured in gases from quiescent volcanism (Capaccioni et al., 1993). The composition of the organic gas fraction can be referred to the occurrence of two chemical processes: i) catalytic reforming reac- tions and ii) thermal cracking reactions (Capac- cioni et al., 1995). The catalytic reforming process is typical in hydrothermal environment, while cracking reactions are considered negligi- ble at T < 300°C. At Turrialba Volcano produc- tion of hydrocarbon gas species is probably con- trolled by catalytic-type oxidation and hydration reactions (Seewald, 2001), which may lead to the formation of both alkane and aromatic gas compounds. However, a notable influence of ki- netic processes, through the breakage of carbon- carbon bonds, produces a general reduction of the molecular weight and possibly leads to the generation of alkene gas compounds. Preliminary analytical results of R/Ra (where R is the 3He/4He ratio in the sample and Ra is the 3He/ 4He ratio in the air: 1.34 × 10–6) values are up to 7.5 and suggest that Turrialba gas discharges are strongly influenced by a magmatic component (Vaselli et al., 2002). Val- ues of δ13C (table I; fig. 6) suggest that CO2 de- rived from a mixing between a possible mantle source and a sedimentary end-member (Giggenbach et al., 1993; Rollison, 1993; Sano et al., 1997). As shown in fig. 7, where the «an- desitic water» area (Taran et al., 1989; Giggen- bach, 1992) is also reported, a condensate sam- ple of West fumarole (sampled in 1998) has iso- topically heavier δ 18O and D/H ratios than those of the meteoric water (Lachniet and Pat- terson, 2002), considering that the altitude of the recharge area is at about 3500 m. This sup- ports the hypothesis that Turrialba fumaroles were recharged by significant contribution of magmatic water. From Xa = (δg-δm) / (δa-δ m) (Giggenbach and Corrales, 1992), where δg rep- resent the isotopic ratio of 18O/16O or D/H, δa is the corresponding ratios of the «andesitic waters» and δm is the isotopic composition of the local meteoric recharge, a fraction of about 30% of «andesitic water» was estimated. aromatics alkanes alkanes Fig. 6. δ 18O-δD diagram for the condensate of the central crater fumarole at Turrialba Volcano. The SMOW field, the «Andesitic water» field (Taran et al., 1989) and the Local Meteoric Water Line (Lach- niet and Patterson, 2002), were reported. Symbols as in fig. 3. δ18O δ 2 H MW: δD=7.6 *δ18O+10.5 Fig. 5. Alkanes-aromatics-alkanes ternary diagram for the Turrialba fumaroles. Symbols as in fig. 3. 1508 Franco Tassi, Orlando Vaselli, Vilma Barboza, Erick Fernandez and Eliecer Duarte Fig. 7a-c. Temporal variations of 10∗H2S and SO2 (a), 20∗HF and HCl (b) and CO2 and 2000∗H2 (c) contents for central crater fumarole at Turrialba Volcano. a b c As a consequence, the chemical composition of the gas discharges at Turrialba Volcano is like- ly to be related to the interaction between hot flu- ids ascending from the deep magmatic system, and a shallow aquifer. The latter is able to lower the temperature of the uprising fluids (tempera- tures of gas discharges are below 92°C) and dis- solve the deep-originated acid species. 5.2. Temporal evolution of gas composition During the period the present study was car- ried out (February 1998-April 2002), major chemical changes in the central fumarole were detected, whilst the West fumarole did not suf- fer any substantial compositional modification (tables I and II). The newly formed fracture sys- tem (Quemada) released gases with a chemical composition approaching that of the central fu- marole. In spite of the measured changes in gas composition, no variations of the outlet temper- atures of all the fumaroles were observed. On the basis of the data gathered, the chemical evo- lution of the fumarolic discharges at Turrialba Volcano can be summarised as follows: 1) Between February 1998 and February 2001 gas discharges were characterised by H2O contents > 940 000 µmol/mol, minor contents of CO2 and H2S (up to 20 000 and 290 µmol/mol, respectively) and very low concen- trations of typical high temperature compounds (up to 205, 5.3, 0.0023 and > 0.01 µmol/mol for HCl, HF, CO and SO2, respectively). 2) In September 2001 the chemical compo- sition of central fumarole started to change, with a dramatic increase in both CO2 and H2 contents (fig. 7c). 3) In November 2001, the same fumarole showed a further increase in CO2 (up to 587 000 µmol/mol), appearance of relatively high contents of SO2 (up to 70 000 µmol/mol), and an increase in H2S, HCl and HF concentra- tions of about one order of magnitude (from 105 to 2400, 31 to 252 and 1 to 38 µmol/mol, respectively) (fig. 7a,b). 4) In February and April 2002 significant CO contents were measured (up to 0.1146 µmol/mol) (table I). The compositional evolution of central fu- marole has to be related to a change of the equi- libria between the hot deep source and the cold 1509 Fluid geochemistry and seismic activity in the period 1998-2002 at Turrialba Volcano (Costa Rica) perature rises up to 1161°C in April 2002), on which, as suggested by chemical and physical ev- idence, the magmatic influence on fluid chem- istry is particularly high. 5.4. Seismic activity and ground deformation Further information on the state of activity of Turrialba has been provided by the geophys- ical surveillance that started in 1990 with a sin- gle station located at the summit of the volcano. Since 1999 VDAP-USGS assistance to OVSICORI- UNA has provided 4 seismic telemetric stations, including a three component short-period sta- shallow liquid buffer, probably due to a «deep hot pulse» which occurred in summer 2001. The sudden increase in H2 contents recorded in September 2001 in central fumarole was possi- bly produced by a more effective dissociation of H2O related to the described increase of heat flux from depth. As a further consequence of the increase in the vaporization rate of the shal- low aquifer, major changes in the permeability of the fracture system that feeds the fumarolic discharges occurred. This allowed the acid species such as SO2, HCl and HF, to reach the surface, as confirmed by the composition of central crater fumarole after November 2001. On the contrary, the temperature continued to be buffered by the shallow system (the fu- maroles remained below 90°C, the boiling tem- perature of water at about 3000 m of altitude). 5.3. Geothermometry Equilibrium temperatures calculated by H2/Ar geothermometer (Chiodini et al., 2001) are 621 and 495°C for West and central fumaroles, respectively (table III). It is possible to speculate that such temperatures are consistent with ther- modynamic conditions regulating the volcanic system at depth. The difference among the results obtained from the chemical composition of the two fumaroles is likely to be related to air-con- tamination, which particularly affected Ar con- tents of central gas discharge due to the low gas flux. The relatively high temperature calculated for the sample collected in September 2001 at central fumarole (about 140°C above the mean value; table III), is related to the previously men- tioned sudden increase in H2 (table I). Different- ly, geothermometric calculation based on CH4/CO2 redox pair (Giggenbach and Goguel, 1989) indicates mean equilibrium temperatures of 868 and 963°C for West and central fumaroles, respectively. Since these CH4/CO2 temperatures are too high to correspond to realistic thermody- namic conditions at depth, they are possibly due to partial methane destruction by thermal degra- dation, able to modify CH4/CO2 ratios estab- lished through thermodynamic equilibria. This effect is more evident for the central gas samples collected after February 2001 (equilibrium tem- Table III. Equilibrium temperatures calculated with H2/Ar and CH4/CO2, geothermometers for the Turrialba Volcano gases. Temperatures are in °C. CH4/CO2 H2/Ar West 02-1998 885 325 West 02-1999 702 550 West 02-2000 754 581 West 02-2001 832 598 West 09-2001 894 647 West 11-2001 914 648 West 01-2001 892 646 West 06-02-2002 963 595 West 19-02-2002 884 616 West 04-2002 963 635 Mean West 868 621 Central 02-1999 687 456 Central 02-2001 1020 367 Central 09-2001 887 638 Central 11-2001 1018 495 Central 02-2002 1005 513 Central 04-2002 1161 498 Mean Central 963 495 Quemada 04-2002 773 559 1510 Franco Tassi, Orlando Vaselli, Vilma Barboza, Erick Fernandez and Eliecer Duarte tion, to improve the geophysical monitoring of the area. In the period 1990-1996, seismic ac- tivity was relatively low (0-91 events/month). In May 1996, small magnitude LP events (M < 1.6 in the Richter scale) occurred (Fer- nández et al., 1996). Since February 1997 seis- micity gradually has increased, up to 2000 events/month, and in October 2000 a further in- crease was recorded. In January 2001, late March 2001, October 2001 and October 2002, four seismic swarms characterized by hybrid (dominant frequency 2.1-3.5 Hz) and LP events (dominant frequency < 1.8 Hz) with a maxi- mum magnitude of 2.0 were detected. Determi- nation of hypocenters located the seismic events at an estimated depth of 4-6 km, beneath a 4 km elongated area comprising the summit craters and the northern flank. In addition to seismic records, dry tilt measurements carried out at the «La Silvia» station located 2.5 km W of the volcano summit showed, between Octo- ber 1996 and January 1999, an accumulated ra- dial inflation of 28 µrad (Barboza et al., 2003). The characteristics of the geophysical parame- ters seem to indicate the occurrence of hy- drofracturaction phenomena at depth (Barboza et al., 2003), promoted by an increase in pres- sure due to fluid vaporization possibly triggered by an enhanced heat input, that have also caused a reduction of the efficiency of the aque- ous buffer system. Eventually, at the top of the volcano, the ground deformation caused the opening of new fractures with radial and con- centric direction with respect to the position of the central crater. 6. Concluding remarks The evolution of the chemical composition of the fumarolic gas discharges along with the geophysical parameters suggests that a relative- ly high-temperature system is still active at Tur- rialba Volcano and major modifications of the present status of activity are occurring. The de- scribed behaviour of the volcanic system is not necessarily related to an incipient resumption of eruptive activity at Turrialba Volcano, that would eventually be preceded by further changes in gas chemistry, ground deformation and seismic signals. Moreover, previous studies on these fumarolic discharges carried out in the earlier eighties detected the presence of SO2, H2 and CO in the gas phase (Cheminée et al., 1982). This may suggest that, before the chem- ical changes recorded after September 2001, variations in gas chemistry of this volcanic sys- tem had already occurred. By coupling these observations with the last volcanic events, mainly phreatic eruptions, at Turrialba two dif- ferent scenarios can be hypothesised: i) the in- creasing pressure may result in an enhanced fracturing system at the craters, that may allow a release of the accumulated energy without eruptive events; ii) the overpressure may trigger small-to-medium size phreatic eruptions. Alter- natively, we can speculate that the persistence of high heat-flux from depth may cause the to- tal consumption of the shallow aquifer. In this case, a further rapid evolution of the chemical characteristics of the fluids and a strong in- crease in the temperatures of the fumaroles should be detected prior to the occurrence of an eruptive event, probably characterised by a sig- nificant magmatic contribution. In order to monitor the activity of the Turri- alba volcanic system, a bi-weekly or at least monthly gas sampling collection, with an im- proved seismic and ground deformation sur- veillance, is highly recommended. Acknowledgements Many thanks are due to F. Bergamaschi for his help during the sampling campaign in Febraury 2001. This work was supported by the Italian Space Agency (resp. OV). REFERENCES ALVARADO, G.E., S. KUSSMAUL, S. CHIESA, P.Y. GILLOT, H. APPEL, G. CORNER and C. RUNDLE (1992): Resumen cronoestratigrafico de las rocas igneas de Costa Rica basado en dataciones radiometricas, J. South. Am. Earth Sci., 6 (3), 151-168. BARBOZA, W., E. FERNANDEZ, E. DUARTE, W. SAENZ, M. MARTINEZ, N. MORENO, T. MARINO, R. VAN DER LAAT, E. HERNANDEZ, E. MALAVASSI and J. VALDES (2003): Changes in the activity of Turrialba Volcano: seismicity, geochemistry and deformation, in 8th Gas 1510 1511 Fluid geochemistry and seismic activity in the period 1998-2002 at Turrialba Volcano (Costa Rica) Workshop, IAVCEI, March-April, Nicaragua-Costa Rica. CAPACCIONI, B., M. MARTINI, F. MANGANI, G. GIANNINI, G. NAPPI and F. PRATI (1993): Light hydrocarbons in gas- emissions from volcanic areas and geothermal fields, Geochem. J., 27, 7-17. CAPACCIONI, B., M. MARTINI and F. MANGANI (1995): Light hydrocarbons in hydrothermal and magmatic fu- maroles: hints of catalytic and thermal reactions, Bull. Volcanol., 56, 593-600. CASTILLO, R.M. (1984): Geologia de Costa Rica, Editorial de la Universidad de Costa Rica, San Josè, Costa Rica, pp. 188. CHEMINÉE, J., M. JAVOY and H. DELORME (1982): Tempera- ture and Gas Data from Turrialba, 01/83 (SEAN 08:01). CHIODINI, G., L. MARINI and M. RUSSO (2001): Geochemi- cal evidence for the existence of high-temperature hy- drothermal brines at Vesuvio Volcano, Italy, Geochim. Cosmochim. Acta, 13, 2129-2147. EVANS, W.C., L.D. WHITE and J.B. RAPP (1998): Geochem- istry of some gases in hydrothermal fluids from the south- ern Juan de Fuca ridge, J. Geophys. Res., 15, 305-313. FERNÁNDEZ, E., E. DUARTE, V. BARBOZA, R. VAN DER LAAT, E. HERNANDEZ, M. MARTINEZ and R. SÁENZ (1996): Number of Microseismic Events Continues to Increase, 12/96 (BGVN 21:12). JOHNSTON, S.T. and D.J. THORKELSON (1997): Cocos-Nazca slab window beneath central America, Earth Planet. Sci. Lett., 146, 465-474. GIGGENBACH, W.F. (1992): Isotopic shifts in waters from geothermal and volcanic systems along convergent plate boundaries and their origin, Earth Planet. Sci. Lett., 113, 495-510. GIGGENBACH, W.F. (1996): Chemical composition of vol- canic gases, in Monitoring and Mitigation of Volcanic Hazards, edited by M. SCARPA, R. TILLING (Springer, Berlin), 221-256. GIGGENBACH, W.F. and R.S. CORRALES (1992): Isotopic and chemical composition of water and steam discharges from volcanic-magmatic-hydrothermal systems of the Guanacaste geothermal province, Costa Rica, Appl. Geochem., 7, 309-332. GIGGENBACH, W.F. and R.L. GOGUEL (1989): Collection and analysis of geothermal and volcanic water and gas dis- charges, DSIR Chemistry, Rep. 2401. GIGGENBACH, W.F., Y. SANO and H. WAKITA (1993): Iso- topic composition of helium, and CO2 and CH4 con- tents in gases produced along the New Zealand part of a convergent plate boundary, Geochim. Cosmochim. Acta, 57, 3427-3455. LACHNIET, M.S. and W.P. PATTERSON (2002): Stable isotope values of Costa Rican surface waters, J. Hydrol., 260, 135-150. MONTEGROSSI, G., F. TASSI, O. VASELLI, A. BUCCIANTI and K. GAROFANO (2001): Sulphur species in volcanic gas- es, Anal. Chem., 73, 3709-3715. OVSICORI-UNA (1998): Report Turrialba activity, Smith- sonian Institution: Global Volcanism Network Bullet- tin, 23/3. PROTTI, M., G. GUENDEL and K. MCNALLY (1995): Cor- relation between the age of the subducting Cocos Plate and the geometry of the Wadaty-Benioff zone under Nicaragua and Costa Rica, in Geologic and Tectonic Development in the Caribbean Plate Boundary in Southern Central America, edited by P. MANN, Geol. Soc. Am. Spec. Pap. 295, 309-326. SANO, Y. and H. WAKITA (1985): Distribution of 3He/4He ratios and its implications for geotectonic structure of the Japanese Islands, J. Geophys. Res., 23, 109- 123. SANO, Y., H. WAKITA and W.F. GIGGENBACH (1987): Island arc tectonics of New Zealand manifested in helium iso- tope ratios, Geochim. Cosmochim. Acta, 51, 1855-1860. ROLLISON, H. (1993): Using geochemical date. Longman Scientific and Technical, pp. 352. SANO, Y., T. GAMO and S.N. WILLIAMS (1997): Secular vari- ations of helium and carbon isotopes at Galeras Vol- cano, Colombia, in Galeras Volcano Colombia: inter- disciplinary study of a Decade Volcano, edited by J. STIX, M.L. CALVACHE and S.N. WILLIAMS, J. Volcanol. Geotherm. Res., 77, 255-265. SEEWALD, J.S. (2001): Aqueous geochemistry of low mo- lecular weight hydrocarbons at elevated temperatures and pressures: constraints from mineral buffered labo- ratory experiments, Geochim. Cosmochim. Acta, 65, 1641-1664. TARAN, Y.A., B.G. POKROVSKY and A.D. ESIKOV (1989): Deuterium and oxygen-18 in fumarolic steam and am- phiboles from some Kamchatka volcanoes: «andesitic waters», Dokl. Akad. Nauk SSSR, 304, 440-443. VASELLI, O., P. CENSI, D. ABBADO, A. MINISSALE, M. PAOLIERI and F. TASSI (1997): Regolamento operativo per le analisi dei rapporti isotopici 18O/16O e 13C/12C nei carbonati, 13C/12C nella CO2 dei gas e 18O/16O nelle acque, Internal Report CNR-CSMGA, 1, 1-14. VASELLI, O., P. CENSI, A. MINISSALE, A. CASIGLIA, M. PAOLIERI and F. TASSI (1998): Regolamento operativo per le analisi dei rapporti isotopici D/H nelle acque, In- ternal Report CNR-CSMGA, 2, 1-12. VASELLI, O., F. TASSI, E. FERNANDEZ, E. DUARTE, V. BAR- BOZA, G. MAGRO and A. MINISSALE (2002): Fluid geo- chemistry evolution at Turrialba Volcano (Costa Rica). Colloque Intl., Mt. Pelèe 1902-2002, Explosive Vol- canism in Subduction Zones, 12-16 May, 2002, Ilede la Martinique (France), 90.