Hydrothermal exploration with the Autonomous Benthic Explorer

Abstract

We describe a three-phase use of the Woods Hole Oceanographic Institution's Autonomous Benthic Explorer (ABE), to locate, map and photograph previously undiscovered fields of high temperature submarine hydrothermal vents. Our approach represents both a complement to and a significant advance beyond the prior state of the art. Previously, hydrothermal exploration relied upon deep-tow instruments equipped with sensors that could locate sites of active “black smoker” venting to within a few kilometers. Follow-on CTD tow-yos could then resolve the sites of seafloor venting to length scales of less than a kilometer but rarely to better than a few hundreds of meters. In our new approach ABE: (i) uses sensors to locate the center of a dispersing non-buoyant hydrothermal plume 100–400 m above the seabed; (ii) makes high-resolution maps of the seafloor beneath the plume center whilst simultaneously detecting interception of any rising, buoyant hydrothermal plumes; and (iii) dives to the seafloor to take photographs in and around any new vent site to characterize its geologic setting and reveal the nature of any chemosynthetic ecosystems it may host. By conducting all of the above under long-baseline navigation, precise sites of venting can be determined to within 5 m. Our approach can be used both to address important scientific issues in their own right and to ensure much more efficient use of other deep-submergence assets such as human occupied vehicles (HOVs) and remotely operated vehicles (ROVs) during follow-on studies.

Introduction

Thirty years after the first discovery of hot springs on the Galapagos Spreading Centre (Corliss et al., 1978) more than 90% of the 55–60,000 km of global ridge crest has yet to be investigated, systematically, for the presence or absence of hydrothermal activity (Baker and German, 2004). While it has been established that hydrothermal venting can occur in every major ocean basin and at all ridge spreading rates, further exploration is required because major uncertainties remain concerning the systematic relationships between hydrothermal activity and tectonic forcing functions such as spreading rate, ridge segmentation and magma supply (e.g. German et al., 2004) and the impact of hydrothermal activity on ocean chemistry and biological cycles (e.g. Van Dover et al., 2002; German and Von Damm, 2004). For example, while hydrothermal activity has been proposed to correlate with spreading rate in a simple manner, basin-scale chemical fluxes are expected to vary in response to changes in partitioning between diffuse hydrothermal flow and high-temperature “black-smoker” venting. Thus, while total hydrothermal heat flux may be predominantly associated with the world's fastest spreading ridges (e.g. Baker et al., 1996), high-temperature vents—which dominate the hydrothermal chemical flux—may be proportionally more prevalent at slow and ultra-slow ridges (e.g. Baker et al., 2004; German and Lin, 2004; German et al., submitted for publication). Further, it is only at the world's slower-spreading ridges that conditions arise under which tectonically controlled, ultramafic-hosted, high-temperature vent sites can occur (e.g. German and Parson, 1998), with their associated and highly distinctive organic and inorganic vent-fluid chemistries (Holm and Charlou, 2001; Douville et al., 2002; Charlou et al., 2002). Biologically, the endemic vent faunas that have been found along the global ridge crest to date (from that small proportion of the global ridge crest where hydrothermal vents have been investigated by submersible) can be classified into six distinct biogeographic provinces (Van Dover et al., 2002). This is intriguing because the global ridge crest is geologically continuous along almost its entire length (Fig. 1), and its path closely mimics that of the modern thermohaline circulation, which has a timescale for deep-ocean mixing of just a few 1000 years. How, then, can different sections of ridge crest remain genetically differentiated?

Over the past decade, the international community has become increasingly adept at combining geophysical studies of the seafloor with oceanographic investigations of the overlying water column to prospect for hydrothermal activity along previously unexplored sections of ridge crest (e.g. Baker et al., 1995; Cave and German, 1998). Such studies have allowed us to demonstrate the presence of high-temperature hydrothermal activity from physical and chemical anomalies in deep waters overlying a section of ridge crest and to locate its source to within a few kilometers. Furthermore, by combining these data with co-registered geophysical data (multi-beam bathymetry, sidescan sonar) we have also been able to make preliminary interpretations concerning the neovolcanic or tectonic settings in which that venting occurs (e.g. German et al., 1996, German et al., 1998a, German et al., 2000; German and Parson, 1998; Hey et al., 2004; Martinez et al., 2006; Connelly et al., 2007). A key limitation to the approach, however, is that the precise location and detailed investigation/characterization of individual vent sites could not be achieved during the same cruise. Even with additional CTD tow-yo operations, sites of venting could not typically be located to better than some hundreds of meters (e.g. German et al., 1998b). Instead, dedicated follow-on expeditions equipped with specialist deep-submergence vehicles have been required to complete even the most rudimentary characterization of the hydrothermal fields that give rise to the detected plumes. Consequently, for numerous ridge segments worldwide, hydrothermal activity has long been known to be present but individual vent sites and their associated chemosynthetic ecosystems have remained completely uninvestigated over timescales of up to a decade (e.g. German et al., 1998a, German et al., 2000; Bach et al., 2002; Edmonds et al., 2003; Connelly et al., 2007).

What has become increasingly recognized, therefore, is the need for an autonomous underwater vehicle (AUV) that can be deployed from a ship of opportunity to carry out preliminary characterizations of vent sites, and the unique fauna that they host, even within the course of a single cruise. Here we describe a recently developed approach designed to achieve these goals using Woods Hole Oceanographic Institution's deep-diving AUV Autonomous Benthic Explorer (ABE). We have conducted such operations successfully on four research cruises in recent years (2004–2007) aboard US, UK, German, and Chinese research ships operating in the Pacific, Atlantic, and Indian Oceans. In this paper, we present the principle behind the approach and then illustrate its potential using specific examples from our first expedition, aboard the R.V. Kilo Moana, which included discovery of the Kilo Moana vent site, East Lau Spreading Centre (ELSC).