The dynamics of vulcanian, subplinian and plinian eruptions; Die Dynamik vulkanischer subplinianischer und plinianischer Eruption
Zusammenfassung der Projektergebnisse
Both, Doppler radar data and numerical modeling show that pulsed forcing significantly affects the dynamics of developing eruption clouds at and near the vent, which leads to considerably different eruption dynamics and cloud height when compared to the empirical model prediction at equal — but constant — mass eruption rate. The relative strength of consecutive pulses controls whether a secondary pulse disturbs the developing buoyant rise of the previous pulse or supports it. Even weak volcanic clouds rise to heights of 8–10 km, which means that dome growing volcanoes with Vulcanian activity may carry high amounts of fine grained ash to flight levels during periods of minor activity. The “eruption source parameters” used by the VAACs for the modeling of ash transport and dispersion in the atmosphere may therefore be significantly wrong, in that they underestimate the impact of weak volcanic clouds. The short-lived fluctuations in mass flux through the volcanic vent are local phenomena and can only be observed in the jet region of the eruption cloud in the first few hundred meters of rise. As a consequence weather Doppler radars are not well suited for the observation of such pulses. This study is the first to quantify the effect of a fluctuating MER on eruption cloud rise. Based on our first rough exploration of the parameter space controlling explosive eruptions more detailed parameter studies are needed to understand the interplay of a temporally varying MER, event duration, forcing function, particle velocity and mass loading of pulses and their influence on plume height and especially the injection height of particles into the atmosphere, without which ash dispersion models are useless. The numerical models used in this study are based on the same physical process (multiphase flow) but have very different numerical approximations. Comparing both models is a first step towards cross-validation. Here we find that both models are suitable for the simulation of large scale volcanic eruptions (Sub-Plinian and Plinian), but it remains questionable whether they can be applied to small scale Vulcanian explosions, like the ones observed by Doppler radar in this study. For future studies it would be desirable to have a combination of both models, i.e. the multi-phase flow approach of PDAC combined with microphysics (phase changes of water), which is included in ATHAM. But the most important improvement would be the incorporation of particles larger than a few millimeter, which is currently not possible in either of the models. With such a model, forward modeling could be used to interpret or even invert Doppler radar data of Vulcanian (and maybe Strombolian) eruptions. When we are able to constrain the dynamics at the vent from measurements of velocity and echo power at a few hundred meters above the vent, we may be able to constrain the physical processes within the shallow conduit. Monitoring at volcanoes is mostly done using seismometers measuring the ground movement. However, it has been shown that the seismicity is not always representative for the surface activity but rather for the overall state of unrest (or activity status) of a volcano. We show that the information of a pulsed MER may be hidden within the seismic and infrasonic signals, but it is not possible to retrieve pulses from either of the signals due to the coda of each pulse overlapping with the onset of the next pulse. Volcanoes may also be monitored using remote sensing techniques (on satellites or weather radars) that capture the surface activity and possibly ash distribution, but satellite measurements are very scarce due to the infrequent passages and neither satellites nor weather radars can resolve the eruption cloud dynamics near the vent (in time and space). For monitoring purposes a reliable automatic event detection and classification algorithm is needed for Doppler radar data. The picking and discrimination of the pulses into ballistic and non-ballistic was done manually from the ‘shape’ of the pulse in a velocigram. This could be replaced in the future by the application of the 2D-cross-correlation used for pulse detection in the Santiaguito study in real time. An even more sophisticated idea is the use of image matching routines that would, in addition to the detection of an event, give information on the duration and maximum velocities in the event or even in each pulse. This information could be put into 2D eruption cloud models so that a theoretical ash injection height could be retrieved for each eruption in near-real-time.
Projektbezogene Publikationen (Auswahl)
- 2012. A detailed view into the eruption clouds of Santiaguito volcano, Guatemala, using Doppler radar. J Geophys Res, 117: B04201
Scharff, L., Ziemen, F., Hort, M., Gerst, A. and Johnson, J. B.
(Siehe online unter https://doi.org/10.1029/2011JB008542) - 2014. The dynamics of the dome at Santiaguito volcano, Guatemala. Geophys J Int, 197:926–942
Scharff, L., Hort, M. and Gerst, A.
(Siehe online unter https://doi.org/10.1093/gji/ggu069) - 2015. Pulsed Vulcanian explosions: A characterization of eruption dynamics using Doppler radar. Geology 43 (11), 995-998, 2015
Scharff, L., Hort, M. and Varley, N. R.
(Siehe online unter https://doi.org/10.1130/G36705.1)
