Submarine Volcanism

Plate tectonics theory indicates that more than 70% of the Earth's volcanic eruptions occur underwater. However, over the past 100 years, less than 10% of known eruptions were submarine, primarily due to limited detection capability. Consequently, submarine eruptions are poorly understood, especially their influence on global ocean bio-geochemistry, deep ocean circulation, and long term climate. My research addresses two questions: How do we detect submarine eruptions, and test theoretical models of eruption dynamics? How do their heat and nutrient emissions affect the ocean-atmosphere system and the biosphere?

I have developed two techniques to detect previously unknown submarine eruptions as they occur: identifying deep ocean eruptive plumes using Argo floats (global autonomous temperature/salinity measurements) and finding eruption products (pumice rafts, hydrothermal fluids, aerial plumes) in satellite imagery via machine learning (ML) algorithms. Using the Argo technique, I made the first in-situ measurements of how the plume from a known submarine eruption affected water properties and dynamics [Mittal & Delbridge 2019]. With my ML algorithm, I detected multiple new pumice rafts. My analysis of these results has significantly advanced our understanding of raft behavior and longevity [Zheng, Mittal, et al., 2020].