CRATER LAKE EVOLUTION DURING VOLCANIC UNREST: CASE STUDY OF THE 2005 PHREATIC ERUPTION AT SANTA ANA VOLCANO, EL SALVADOR

Santa Ana volcano in El Salvador hosts a small acid crater lake and fumaroles in its summit crater. From 2000 to 2004, volcanic activity was characterized by relative quiescence interrupted by episodes of hydrothermal venting. Aqueous chemistry analyses and temperature measurements revealed an acid (pH~1) sulfate-chloride (SO4/Cl~1.5) lake that was consistently several degrees warmer than ambient temperature. Energy budget modeling elucidated the stark contrast between the relatively cool crater lake with a small heat input (< 15 MW) and the adjacent high temperature fumaroles (875 degrees C). In June 2004, unrest began and progressed into a volcanic crisis in late August 2005 when incandescence was observed in the crater fumaroles. The volcanic crisis was marked by anomalously high sulfur dioxide fluxes and uncharacteristic seismicity (VT, LP, “banded” tremor). The month-long crisis ultimately culminated in a phreatic eruption (VEI 3) on 1 October 2005. The 2005 eruption significantly changed the crater geometry. Satellite images and visual observations show that the lake changed color and position, drowning the adjacent high temperature fumaroles. Degassing directly into (and through) the lake at the newly subaqueous fumarole caused the lake to acidify (pH>0.5) and warm (<65 degrees C, <830 MW), resulting in periodic evaporations during 2006 and 2007.

Eruption precursors could be interpreted to signify an arrival of new magma within a few km of the surface prior to the October 2005 eruption, but lack of a conclusive magmatic component among eruptive products and lack of geochemical evidence of water-magma interaction suggest that new magma was not involved, or alternatively, that an intrusion of new magma stalled at depth below the hydrothermal system. Continued lake instability post-2005, i.e. lake evaporations, suggests that magma may still reside at shallow levels in the edifice. The reality of a magmatic intrusion becomes more likely the longer high lake temperatures and high heat flow continue to be recorded in the lake.