Littoral blasts: Pumice‐water heat transfer and the conditions for steam explosions when pyroclastic flows enter the ocean

Abstract

[1] Steam explosions, or littoral blasts, generated when pyroclastic flows interact with seawater may be a common, although rarely documented, phenomena. The development of steam explosions rather than passive steam production is related to the rate of thermal energy transfer from hot pyroclasts to water. We conduct a series of laboratory experiments to quantify the heat transfer and steam production rates when hot pyroclasts encounter water. Hot pumice (>200°C) rapidly ingests water while remaining at the surface, producing measurable amounts of steam during the process. Approximately 10% of the thermal energy of the pumice particles is partitioned into the production of steam, and smaller particles have greater steam production rates. The laboratory experiments are used to develop a subgrid model for steam production that can be incorporated into a multiphase numerical framework. We use this model to study the critical steam production rates required to initiate explosive events. For conditions typical of many pyroclastic flows, particles smaller than $1-5 mm are required to initiate a littoral blast. A second set of two-dimensional numerical simulations is conducted to simulate the 12-13 July Soufrière Hills dome collapse event that reached the sea. The simulations predict that the focus of the blast is likely generated several hundred meters offshore and although the landward directed base surge is primarily dry (<15% water vapor), the area immediately above the blast is steam-rich and may be a likely site for the production of accretionary lapilli.

Characteristics of submarine pumice-rich density current deposits sourced from turbulent mixing of subaerial pyroclastic flows at the shoreline: field and experimental assessment

“…The floating, soaking, and sinking patterns of 2-4 mm pumices are qualitatively the same as for coarser clasts (Manville et al 1998). The thermal interactions with water are predominantly controlled by the finer grains which facilitate rapid heat transfer during the short duration of experimental interactions (Freundt 2003;Dufek et al 2007). The behavior of experimental density currents may be characterized by two types of Froude number.…”

“…Processes of littoral explosions, turbulent mixing, and transformation into turbidity currents suggests that pyroclastic flows crossing the shoreline were particularly hot (>150-200°C) and poorly sorted or had low density contrast between the pyroclastic flow and water (McLeod et al 1999;Dufek et al 2007;Freundt 2003). Components and facies of the Znp imply that the pyroclastic flow mixed turbulently and pumice clasts quenched on contact with water, and that this interaction was accompanied by steam explosions.…”

“…Thermal interactions with water are predominantly controlled by ash that facilitates rapid heat transfer during the steam explosions (e.g., Freundt 2003;Dufek et al 2007). Such liberated ash would eventually settle onto the water surface and form vertical density currents transporting it to the seafloor if sufficient concentrations were reached (e.g., Carey 1997).…”

“…Hot pumice clasts enter water where subaerial pyroclastic density currents cross coastlines, or directly from submarine explosive eruptions (Whitham and Sparks 1986;Dufek et al 2007;McPhie 2000, 2009). During transport in subaerial pyroclastic flows, air quickly replaces magmatic gasses but pumice lapilli remain hot as the heat capacity of air is small (Stroberg et al 2010).…”

“…Turbulent mixing in water provides the means to sort pumice lapilli from co-erupted ash and dense pyroclasts in order to form pumice breccia. This turbulent mixing can occur within submarine explosive eruption columns, where subaerial pyroclastic flows enter the shoreline, during littoral steam explosions or during transport in watersupported turbidity currents (Freundt 2003;Dufek et al 2007;Allen and McPhie 2009).…”