Constraining Volcano-Hydrologic Interaction at Masaya Volcano, Nicaragua Through Continuous Temperature Monitoring and Modeling

Flank fumaroles are a direct result of the interaction between groundwater and magma, and therefore have good potential for monitoring volcanoes. At Masaya volcano in Nicaragua, Tough2 models of fluid transport through porous media show that three fumarole zones observed on the flank of Masaya volcano can be explained by convection within the saturated zone when a uniform heat source is applied at depth. 5 days before the formation of a small lava lake in the active crater, temperatures increased by up to 5°C in the further fumarole zone, 3-4 km away. However, magnetic modeling shows that an anomaly of up to 6000 nT corresponds extremely well with topographic offset of a fault and fracture system on the flank of the volcano that is an important control on local fluid flux. Comparisons between Tough2 models of these structures, self-potential and CO2 data show that elevated fluid flux occurs in the hanging wall of relatively impermeable faults, and flow is inhibited in the footwall. Local geological structures like faults are therefore a dominant factor in the location and response of these fumaroles. Although the fundamental source of fumaroles is groundwater-magma interaction, the response of the groundwater system to volcanic activity is on the order of months to years and is hard to resolve with current monitoring techniques. Therefore the source for variations in fumarole degassing is not simple, and a good conceptual model of the system is vital. Continuous monitoring of CO2 for 3 years has shown that there is a distinct structure to the temperature signal of fumarole gases prior to and during volcanic activity. 10-15 cycles were recorded during 4 10-day periods, two of which were associated with surficial volcanic activity. Rainfall also showed a significant, albeit imperfect, correlation with these signals. The frequency spectrum proved to be an extremely useful tool in identifying the beginning and end of the anomalous episodes. Using Tough2 to create numerical models of the system we have determined that flow of fluids from the water table is too slow to explain these responses. A pressure pulse originating from the active crater is however a possible mechanism. We are now creating models to see what magnitude and time- and spatial-scale changes can create the signals observed. Models allow us to see whether changes in diffuse degassing can emanate from variations below the water table, or are due to a shallow source like shallow fluid injection or a pressure pulse traveling through the system. This is extremely important when interpreting diffuse degassing variations in light of changes in volcanic activity.