Volcaniclastic Dispersal During Submarine Lava Effusion: The 2012 Eruption of Havre Volcano, Kermadec Arc, New Zealand

Abstract

Understanding clast dispersal from subaqueous volcanism is hampered by uncertainty in the source and extent of seafloor deposits. Extensive sampling in situ of seafloor deposits from the 2012 submarine eruption of Havre volcano provides an ideal opportunity to assess subaqueous dispersal. The 2012 Havre eruption produced 14 lavas/domes, a pumice raft, and three seafloor clastic deposits. At Havre the source of clastic deposits can be confidently identified, and deposit thickness, grain size, and distribution are also well-constrained. We examine a seafloor deposit termed subunit 3 (S3) generated in the 2012 Havre eruption to investigate dispersal of fine lapilli and ash, and the eruption conditions that generated this deposit. Subunit 3 is the third from bottom of four subunits that make up the Ash with Lapilli unit. Subunit 3 is composed of ash with highly elongate shapes, unique within the 2012 Havre deposits. It thickens and coarsens toward Lava G, also generated in the 2012 eruption, located on the southwest wall of Havre caldera. Lava G is the only lava produced during the 2012 Havre eruption that has a glassy carapace with elongated vesicles and a fibrous texture. We infer the source of unit S3 is Lava G, due to the spatial pattern of deposit thinning and fining away with distance from this lava, and the morphological and microtextural similarity of ash with the Lava G carapace rock. Grain size and transport distance of ash from S3 are used to test a simple 1D model addressing both clast dispersal by a buoyant thermal plume above an explosive eruption, and by penetrative convection during effusive lava emplacement. Comparison of calculated maximum dispersal distances with grain size and transport distance show that a jet forming eruption generating a turbulent plume is required to generate S3. We suggest that S3 was generated by hybrid explosive-effusive activity during the effusion of Lava G. Using model results we calculate maximum clast dispersal distances across a range of grain sizes for both dispersal mechanisms. The calculated maximum clast dispersal distance has wide implications globally for interpretation of ash deposits from subaqueous eruptions.