THE CHEMICAL AND HYDROLOGIC STRUCTURE OF POAS VOLCANO, COSTA-RICA

Citation
Gl. Rowe et al., THE CHEMICAL AND HYDROLOGIC STRUCTURE OF POAS VOLCANO, COSTA-RICA, Journal of volcanology and geothermal research, 64(3-4), 1995, pp. 233-267
Citations number
65
Categorie Soggetti
Geology
ISSN journal
03770273
Volume
64
Issue
3-4
Year of publication
1995
Pages
233 - 267
Database
ISI
SICI code
0377-0273(1995)64:3-4<233:TCAHSO>2.0.ZU;2-H
Abstract
Comparison of the chemical characteristics of spring and river water d raining the flanks of Poas Volcano, Costa Rica indicates that acid chl oride sulfate springs of the northwestern flank of the volcano are der ived by leakage and mixing of acid brines formed in the summit hydroth ermal system with dilute flank groundwater. Acid chloride sulfate wate rs of the Rio Agrio drainage basin on the northwestern flank are the o nly waters on Poas that are affected by leakage of acid brines from th e summit hydrothermal system. Acid sulfate waters found on the northwe stern flank are produced by the interaction of surface and shallow gro undwater with dry and wet acid deposition of SO2 and H2SO4 aerosols, r espectively. The acid deposition is caused by a plume of acid gases th at is released by a shallow magma body located beneath the active crat er of Poas. No evidence for a deep reservoir of neutral pH sodium chlo ride brine is found at Poas. The lack of discharge of sodium chloride waters at Poas is attributed to two factors: (1) the presence of a rel atively volatile-rich magma body degassing at shallow depths (< 1 km) into a high level summit groundwater system; and (2) the hydrologic st ructure of the volcano in which high rates of recharge combine with ra pid lateral flow of shallow groundwater to prevent deep-seated sodium chloride fluids from ascending to the surface. The shallow depth of th e volatile-rich magma results in the degassing of large quantities of SO2 and HCl. These gases are readily hydrolyzed and quickly mix with m eteoric water to form a reservoir of acid chloride-sulfate brine in th e summit hydrothermal system. High recharge rates and steep hydraulic gradients associated with elevated topographic features of the summit region promote lateral flow of acid brines generated in the summit hyd rothermal system. However, the same high recharge rates and steep hydr aulic gradients prevent lateral flow of deep-seated fluids, thereby ma sking the presence of any sodium chloride brines that may exist in dee per parts of the volcanic edifice. Structural, stratigraphic, and topo graphic features of Poas Volcano are critical in restricting flow of a cid brines to the northwestern flank of the volcano. A permeable lava- lahar sequence that outcrops in the Rio Agrio drainage basin forms a h ydraulic conduit between the crater lake and acid chloride sulfate spr ings. Spring water residence times are estimated from tritium data and indicate that flow of acid brines from the active crater to the Rio A grio source springs is relatively rapid (3 to 17 years). Hydraulic con ductivity values of the lava-lahar sequence calculated from residence time estimates range from 10(-5) to 10(-7) m/s. These values are consi stent with hydraulic conductivity values determined by aquifer tests o f fractured and porous lava/pyroclastic sequences at the base of the n orthwestern flank of the volcano. Fluxes of dissolved rock-forming ele ments in Rio Agrio indicate that approximately 4300 and 1650 m3 of roc k are removed annually from the northwest flank aquifer and the active crater hydrothermal system, respectively. Over the lifetime of the hy drothermal system (100's to 1000's of years), significant increases in aquifer porosity and permeability should occur, in marked contrast to the reduction in permeability that often accompanies hydrothermal alt eration in less acidic systems. Average fluxes of fluoride, chloride a nd sulfur calculated from discharge and compositional data collected i n the Rio Agrio drainage basin over the period 1988-1990 are approxima tely 2, 38 and 30 metric tons/day. These fluxes should be representati ve of minimum volatile release rates at Poas in the last 10 to 20 year s.