Bubble dynamics during volcanic eruptions: in situ measurements and modelling of physical processes of a rising and exploding gas slug

在这项研究的最初两年，我们在瓦努阿图的Yasur火山收集了长达16天的多学科数据集，其中包括地震学和声学数据，红外视频片段，以及用于量化火山喷发动力学的多普勒雷达测量，以及用于SO2通量估计的DOAS测量。在此期间，总共观测到约26000次喷发，初步数据处理表明，喷发方式从几乎不含火山灰的爆炸性喷发到富含火山灰的羽状喷发，再回到几乎不含火山灰的喷发。在这个建议的框架内，我们将利用我们的多学科数据集探索这种喷发方式变化背后的物理学。特别是，我们将把处理后的地震和声学数据与雷达数据结合起来，以验证Yasur火山的哪个喷口是活跃的，并确定喷口出口以下的岩浆水平位于管道中的岩浆水平。利用雷达数据，我们将确定每次单次爆炸释放的动能，并由此确定气泡超压或气塞体积，而声波和地震数据将为我们提供地震和声波能量释放的证据。这将使我们对活火山的能量收支有更全面的了解。雷达数据还将允许我们定义不同的喷发类型，并使用神经网络对所有记录的事件进行分类。这些不同的处理步骤将有助于改进当前的喷发动力学模型。

During the initial two years funding of this research we collected a 16 days long multidisciplinary data set at Yasur volcano, Vanuatu, that includes seismological and acoustic data, infrared video footage, as well as Doppler radar measurements to quantify eruption dynamics, and DOAS measurements for SO2 flux estimation. During this time a total of about 26000 eruptions were observed and initial data processing indicated a transition in eruption style from explosive nearly ash free eruptions to very ash rich plumes and back to nearly ash free eruptions. In the frame of this proposal we will explore the physics behind this change in eruption style using our multidisciplinary data set. In particular we will combine the processed seismic and acoustic data with the radar data to verify which vent of Yasur volcano was active and determine at what depth below the vent exit the magma level in the conduit is located. Exploiting the radar data we will determine the kinetic energy released during each single explosion and from that determine either bubble overpressure or gas slug volume, while the acoustic and seismic data will give us evidences for the seismic and acoustic energy release. This will lead to a more complete understanding of the energy budget at an active volcano. The radar data will also allow us to define different eruption types and using a neuronal network we will then classify all recorded events. These different processing steps will help to improve current models for eruption dynamics.