Presentation on theme: "Can eruptions be forecast? Chris Newhall, US Geological Survey Seattle."— Presentation transcript:
Can eruptions be forecast? Chris Newhall, US Geological Survey Seattle
Principal methods of volcano monitoring (All on surface, in lieu of an internal gas pressure gauge! Most, looking for repeated patterns of precursory change)
For better, process-based forecasts, we need to track what makes volcanoes erupt... GAS!
magma input from depth Gas leaks --> Lava domes Viscous magma + Slow ascent to surface --> preeruption gas leakage --> low explosive potential. H Mader/CN
silicate melt + dissolved gases (H 2 O, CO 2, S) (+ crystals) magma input from depth low viscosity magma --> gas highly mobile and can escape easily ‘Effusive’ eruptions. After H Mader
Lava pond on Mauna Ulu (Hawaii) 1969 - bubble is ~ 5 metres across.
high viscosity magma --> gas not very mobile and cannot escape easily IF ALSO rapid ascent, --> high gas content persists to near surface --> large potential expansion silicate melt + dissolved gases (H 2 O, CO 2, S) (+ crystals?) magma input from depth Explosive eruptions. After H Mader
E.g., Pinatubo, Mount St. Helens Popocatepetl (1994- ), Sakurajima, Mayon, Kilauea; Iwo-jima Closed, Tight No convection in the conduit(s); incoming gas is stored Open, Leaky Convection in the conduit, releasing most gas Two modes of conduit, eruptive behavior Strong eruption precursors; ample warnings Weak and/or short eruption precursors; warnings difficult Frequent, sm eruptions (or no eruptions at all) Infrequent, large explosive eruptions
Explosions/day Earthquakes/day, < 1 km deep Earthquakes/day, 1-15 km deep (1987) Open vent: Sakurajima Volcano, Japan K. Kamo et al.
Take home message #1: Forecast success rates? ---------------------------------------------------- Volcano restless, eruption possible? >95% Eruption next few days? >50% Hours?<50% Explosive potential? >95% Actual explosive magnitude? Tough, maybe 50%
Take home message #2: Biggest uncertainty is the degree to which magma is still trapping gas as it nears the surface, i.e., how tight or leaky the volcano. Gas content, pressure in magma cannot be measured directly. Gas monitoring is relatively rare; more often, we rely on proxies like seismicity and ground deformation.
Take home message #3: Big eruptions generally easier to forecast, with lead times of days to months. The smallest, most frequent eruptions, are tough to forecast with useful lead times. In all cases, eruption detection and tracking still critically needed!