Multi-sensor acoustic evidence of submarine volcanism across diverse tecto-volcanic regimes

Submarine volcanism and associated submarine venting can be observed using multibeam bathymetry, seafloor backscatter, water-column backscatter, and sub-bottom acoustic profiling. While numerous volcanically-driven gas accumulation and venting phenomena can be observed through these techniques (such as water-column acoustic flares, subsurface acoustic blanking, and seafloor acoustic hardgrounds), detecting active submarine volcanism remains largely enigmatic due to the rarity of observations at appropriate spatial or temporal scales and underutilisation of coincident data. This thesis uncovers and characterises active volcanic processes in three distinct submarine volcanic terrains by capitalising on rich tapestries of acoustic and ancillary data to develop multi-frequency, multi-sensor, and multi-scale acoustic analyses. The first contribution is a multi-frequency water-column backscatter investigation into mechanisms for sediment-hosted seeps and volcanic vents offshore the volcanically active Heard Island and McDonald Islands on the Central Kerguelen Plateau, an oceanic large igneous province in the southern Indian Ocean (Chapter 2). Active submarine fluid expulsion (acoustic flares) was observed with a multi-frequency split‚ÄövÑv™beam echosounder and characterised using a decibel differencing technique, combined with video, sub-bottom profile, water column ˜í¬•3He, and sediment ˜í¬•13C and ˜í¬•34S data. The second contribution uses multi-frequency seafloor backscatter to link active submarine fluid expulsion with seafloor characteristics in the Calypso hydrothermal vent field (HVF), offshore the volcanically active Whakaari / White Island in the Bay of Plenty, Aotearoa / New Zealand (Chapter 3). This work explores whether using seafloor backscatter data from a low (30 kHz) or high (200 kHz) frequency multibeam echosounder, or using a combined frequency model, increases predictive accuracy of fluid expulsion locations. The final contribution is a multi-scale and multi-platform geomorphic time-series of the submarine Havre caldera volcano, on the Kermadec Arc, Southwest Pacific (Chapter 4), covering pre-eruption (2002), post-eruption (2002) and fine-scale (2015) bathymetry. Multiple bathymetric datasets were used to parameterise geomorphological features and volcanic products, while seafloor samples and video data refined geomorphic boundaries and located vent and seep fields. The vents and seeps at Calypso HVF represent a long embedded combined volcanic and thermogenic seep system. Model results, based on multi-frequency interrogation, revealed active fluid expulsion is not typically associated with extensive, bathymetric hardgrounds observed at the Calypso HVF, as previously suggested. Unconsolidated sediment around the perimeter of and between hardgrounds hosted more active fluid expulsion, showing hardgrounds are the result of vents self-sealing over time. At Heard and McDonald Islands, volcanic and/or thermogenic gases are suggested to be the dominant source for venting proximal to the volcanically active islands. The McDonald Island vents are driven by shallow diffuse hydrothermal venting, likely CO2 dominated. Excess ˜í¬•3He indicates mantle‚ÄövÑv™derived input, suggesting proximal hydrothermal activity. The Heard Island flares indicate an alternate source: shallow biogenic CH4 and deeper‚ÄövÑv™sourced thermogenic CH4 related to geothermal heat from onshore volcanism. Gas chimneys in sub-bottom acoustic data, no ˜í¬•3He signal, ˜í¬•13C and ˜í¬•34S fractionation factors in sediment, indicate a sediment-hosted geothermal system. The lack of typical features from a well embedded seep system here suggest the Heard Island seep system is either young, slow-growing, or ephemeral. At Havre, additional growth on the primary dome emplacement between 2012 and 2015 was revealed for the first time, while voluminous loss from the northern caldera wall was revealed to be subsidence related. Video data revealed extensive active hydrothermal fluid expulsion locations across all ROV surveyed areas, revealing the rate at which recolonization and the return of vent systems can occur at a submarine caldera. This thesis demonstrates that without a suite of acoustic and complementary data, small-scale seafloor fluid expulsion or ephemeral volcanic processes are likely missed, in contrast to the obvious, voluminous eruptions of oceanic or submarine volcanoes. It is well established that monitoring and understanding submarine volcanic processes requires repeat, comparable surveys. As observed in the first contribution, remote environments such as the volcanic Heard and McDonald Islands often suffer from being sampled only once or rarely. The acoustic flares observed provide the first evidence of submarine gas escape on the Central Kerguelen Plateau, expanding our understanding of submarine volcanic expulsion and carbon cycling in the southern Indian Ocean. In the second contribution, being able to remotely predict active and inactive regions of seafloor fluid expulsion may lead to detection of seeps in legacy datasets while also highlighting seafloor characteristics that can aid in locating nascent, senescent, or extinct seeps during underway surveys. The final contribution, a rare opportunity to conduct change detection in the deep ocean, reveals submarine volcanic forms and geomorphic processes across a spectrum of scales, highlighting the importance of repeat, high-resolution bathymetric surveys. This unprecedented dataset highlights however how lacking a full suite of acoustic data can result in active ephemeral volcanic processes being overlooked. Integrated acoustic analysis of active seafloor volcanism is vital for revealing the spectrum of volcanic and volcanic-adjacent systems and processes.