V040-0021
Ocean dispersal of the 2019 Tonga pumice raft and insights on submarine eruption mechanisms (SW Pacific)
Wednesday, 16 December 2020
Poster
Martin Jutzeler, University of Tasmania, School of Natural Sciences and Centre for Ore Deposit and Earth Sciences (CODES), Hobart, TAS, Australia, Erik van Sebille, Universiteit Utrecht, Marine & Atmospheric Research, Utrecht, Netherlands, Robert Marsh, University of Southampton, Ocean and Earth Science, Southampton, United Kingdom, Tushar Mittal, University of California Berkeley, Earth and Planetary Science, Berkeley, CA, United States, Rebecca Carey, University of Tasmania, Centre of Ore Deposits and Earth Sciences, Hobart, TAS, Australia, Kristen E. Fauria, Vanderbilt University, Department of Earth and Environmental Sciences, Nashville, TN, United States, Michael Manga, University of California Berkeley, Berkeley, CA, United States and Jocelyn McPhie, Univ Tasmania, Hobart, Australia
Abstract:
On the 7 August 2019, a 195 km
2 pumice raft was produced from a vent at 200 m below sea level from an unnamed submarine volcano in the Tonga Islands (Southwest Pacific Ocean). Several yachts crossed the raft as early as 2 days post‐eruption and reported raft thicknesses of between 15-30 cm thick. The pumice clasts in the raft were reported as ranging from <1 to 60 cm in size and photos show that they were partly abraded. Further examination of the clasts shows that they are grey-brown in color, andesitic (SiO
2 61.36%; Na
2O+K
2O 5.39%), and highly vesicular. The clasts have partial quench rims where vesicles are much smaller than in the clast’s core.
Despite the remoteness of the raft, excellent satellite images allowed for daily tracking of the pumice raft. The raft drifted chiefly westward from the eruption site, and reached Fiji at >750 km from source 43 days later. We conducted short‐term (1 month) forecast and long‐term (2 year) hindcast simulations in Parcels based on data from CMEMS to understand the relative influence of surface ocean currents, wind and wave action on pumice dispersal. Comparison with an extensive dataset of near real-time satellite images show that 1% windage and no Stokes drift was the best fit to model the dispersal of the newly formed raft. The impact of Stokes drift on raft dispersal is likely to increase through time, with thinning and breaking apart of the raft.
The coupling of real‐time satellite observations with oceanographic Lagrangian simulations allowed bi-weekly release of hazard maps to the Tonga and Fiji authorities for dissemination to the yachting, shipping and fishing communities; such strategy can be deployed globally.
