Voyage Plans and Summaries

[back to voyage document index]

Franklin Voyage Summary No. FR04/00

Title

Project SHAARC: Investigation of Submarine, Hydrothermally Active Arc Volcanoes in the Tabar-Lihir-Tanga-Feni Island and Solomon Island Chains

Itinerary

Depart Rabaul 0930 hrs, Friday 5 May, 2000
Arrive Darwin 0830 hrs, Wednesday 24 May, 2000

Principal Investigators

Dr. Brent McInnes (Chief Scientist)
CSIRO Exploration & Mining
PO Box 136, North Ryde, NSW 1670
Work: 02-9490-8597 Fax: 02-9490-8921?
e-mail: brent.mcinnes@dem.csiro.au

Professor Richard Arculus
Dept. of Geological Sciences
Australian National University
Canberra ACT 0200
Work: 02-6249-3778 Fax: 02-6249-5544?
e-mail: richard.arculus@anu.edu.au

Dr. Gary Massoth
Institute of Geological and Nuclear Sciences, Ltd.
Gracefield Research Centre, 30 Gracefield Rd
PO Box 30-368, Lower Hutt, NEW ZEALAND
Work: 64-4-570-4878 Fax: 64-4-570-4657?
e-mail: g.massoth@gns.cri.nz

Scientific Objectives

The Franklin 04/00 research team conducted a multi-disciplinary investigation (volcanology, petrology, geochemistry and economic geology) of submarine volcanic-hydrothermal systems in island arc regions of Papua New Guinea and the Solomon Islands. The team carried out 3 primary scientific activities:

(i) Volcanology/Petrology (Team Leader: Arculus)
(ii) Hydrothermal Deposits/Geochemistry (Team Leader: McInnes)
(iii) Marine Chemistry/Hydrothermal Fluids (Team Leader: Massoth)

Although volcanism and associated hydrothermal activity on mid-ocean ridges (MOR) and some back-arc basins have been the focus of many detailed studies over the past decade, only a few submarine arcs and fore-arc regions have received much attention. FR04/00 investigations of the New Ireland-Solomon Islands convergent margin will generate a better understanding of the fundamental geological processes of submarine arc volcanism, including hydrothermal activity and the formation of economic concentrations of ore minerals. These data will be compared to and contrasted with hydrothermal activity occurring at divergent plate boundary volcanism.

Cruise Objectives

The FR04/00 research cruise targeted several submarine island arc volcanoes in the Tabar-Lihir-Tanga-Feni and the Solomon Island chains for water-column sampling and dredging operations. Operation time was equitably distributed to each of the 3 key science activities.

Specific sampling targets included:

1. active hydrothermal vent sites on submarine cinder cones off the southern coast of Lihir Island;

2. mineral deposits and associated rocks shed off caldera-collapse structures on eastern Lihir Island;

3. several fore-arc submarine volcanoes adjacent to the Kilinailau trench and;

4. four fore-arc and triple-junction related submarine volcanoes located in the New Georgia Group of the Solomon Islands.

We relied on the accurate SO-94 and SO-133 multibeam bathymetry maps for the Stage I portion of the cruise published by the Geological Survey of Canada, and bathymetric maps of the Solomon fore-arc from atlases published by the University of Hawaii and MMAJ/SOPAC.

Cruise Track (Figures 1 and 2)

Figure 1. Cruise track for Stage I Operations — FR04/00

Figure 2. Cruise track for Stage III Operations — FR04/00

Preliminary Results

ACTIVITY 1. Volcanology/Petrology

Lihir-Tanga-Feni Islands (Richard Arculus)

Dredging of the most prominent shallow submarine ridges at the two southern islands of Tanga and Feni recovered mostly mudstones, including some that are bioturbated and carbonate-rich, together with a small percentage of phlogopite-clinopyroxene-bearing basaltic pebbles, equivalent to known lava types on the islands. This type of result would appear to confirm previous interpretations of the regional geological setting of these islands. For example, it seems likely that the volcanic constructs are a relatively small veneer on top of tectonically uplifted blocks of sedimentary strata forming the shallower New Ireland forearc crust.

Similar basalts and sedimentary rocks predominate among the fragmental material recovered from the small submarine cone of Edison, off the southern coast of Lihir. In contrast, most of the cobble- to pebble-size materials obtained from the summit of Tubaf, another small submarine explosive cone in this region, comprise phlogopite-amphibole-clinopyroxene-bearing basalt enclosing a variety of xenolithic lithologies:

1. harzburgite
2. gabbro
3. mudstone/carbonaceous mudstone/volcaniclastic sedimentary rocks.

While the external form of these cobbles and pebbles is rounded, the xenoliths are mostly strikingly angular and fractured. In addition, many of the gabbros have a tectonic fabric. As previously reported by McInnes and coworkers these xenoliths are likely to be representative of the deeper lithologies of the forearc lithosphere.

An ubiquitous albeit minor fraction of many of the dredges in the Lihir-Tanga-Feni region is dacitic pumice. The most likely source of this pumice is the 1994-present eruption of Rabaul in northeastern New Britain.

New Georgia Group (Richard Arculus)

Within the complex region enclosing the Woodlark Spreading Ridge-Transform system intersection with the Solomons island arc, dredging, grabbing, and coring operations concentrated on 3 major tectonic settings: 1. relatively small volcanic cones astride the intersection of the Ghizo and Simbo Ridges (a putative extinct spreading segment and leaky transform of the Woodlark Ridge, respectively); 2. the major volcanic seamounts of Kana Keoki and Coleman, comprising the easternmost Ghizo Ridge; 3. the active submarine volcano of Kavachi in the New Georgia forearc.

1. Ghizo-Simbo Ridges: Seamounts A, B, C, and I

   Lithologies recovered from these seamounts, respectively, were: A. fine-grained to glassy, slightly vesicular clinopyroxene-plagioclase-phyric basalt (crystalline phases in order of volumetric abundance); B. plagioclase-clinopyroxene-orthopyroxene-phyric, slightly vesicular, fine-grained to glassy andesite pillow lava; C. fine-grained clinopyroxene-plagioclase (-olivine?) slightly vesicular basalt; I. sparsely phyric (clinopyroxene-plagioclase), slightly vesicular andesite.

2. Kana Keoki and Coleman Seamounts

   At the summit region of Kana Keoki, light grey, highly vesicular (pumiceous) dacite is the major lithology that is locally, extensively hyrothermally altered. Some of the dacite contains fresh black to translucent glass, and in addition, unusual amphibole megacrysts (up to 5mm) were observed. Two minor lithologies recovered from the summit flanks of Kana Keoki were: i. clinopyroxene-plagioclase-olivine-phyric basalt; and ii., plagioclase-phyric basalt. The major lithology comprising the summit and flanks of Coleman is glassy- to fine-grained, clinopyroxene-plagioclase-phyric basalt.

3. Kavachi Volcano

In addition to strikingly fresh, glassy, black vesicular scoria from the active eruption, three different petrologic types were recovered from Kavachi, in order of abundance: i. black, vesicular, glassy plagioclase-clinopyroxene-olivine-phyric basalt; ii. dark grey, non-vesicular clinopyroxene-plagioclase-phyric basalt; iii. coarsely-phyric (10mm) clinopyroxene-plagioclase-bearing basalt.

Woodlark Spreading Center (John Chadwick)

The Woodlark mid-ocean ridge is a 650-km-long system of spreading centers and transform faults. The system extends from the New Britain Trench, where the Simbo Ridge transform is being subducted, to the D’Entrecasteaux Islands near the Papua New Guinea mainland where the ridge system is propagating to the west and creating metamorphic core complexes. Active for 5-6 million years, the spreading segments on the eastern part of the ridge system have been opening at about 70 mm/yr and forming 4 km-deep, 5-10 km wide rift valleys. These spreading segments are offset by (leaky) transforms that also form deep, V-shaped rift basins.

Previous sampling of the Woodlark Basin area found that typical mid-ocean ridge basalts (MORB) formed >150 km from the Solomon Islands arc, but those forming closer to the arc have back arc basin affinities, transitional between arc tholeiites and MORB. The Woodlark basin was in a backarc position prior to about 10 my during southward subduction of the Pacific Plate at the Kilinailau trench; consequently, the mantle under the present Woodlark Basin may have been enriched by subducted Pacific Plate-derived components prior to ~ 10Ma.

Four grabs and one dredge were performed near the Woodlark spreading center by RV Franklin, in areas that had not previously been investigated. The easternmost segment is within 150 km of the arc, and therefore may show evidence of enrichment by Pacific Plate subduction. Two sampling attempts were performed on seamounts, and three were performed on the seafloor just north of the rift. All but one grab (sediments and pumice only) retrieved glassy, aphyric basalt samples. One sample, acquired from a seamount in the rift zone on the axis of the spreading center, was highly vesicular in spite of its eruption depth (3500 m).

A single dredge on the inner western wall of the apparent caldera of isolated "Seamount J" (9° 44.43’S, 155° 46.30’E ), about 10 nm south of the Woodlark Spreading Cente (and with no counterpart to the north of the spreading centre), retrieved a few glassy, aphyric basalt chips and light brown mud.

ACTIVITY 2. Hydrothermal Deposits/Geochemistry

(Tim McConachy and Brent McInnes)

Kana Keoki

Although sulfide samples containing ore-grade gold concentrations had been collected previously at Kana Keoki during a MMAJ/SOPAC research cruise, our research program extended the knowledge of this occurrence by defining its areal extent by a series of parallel and cross-cutting precision dredge operations conducted in calm sea conditions.

On the western summit of Kana Keoki, in water depths ranging between 750-650m, we completed 5 dredges within an area of about 1 sq. km. There was an amazing variety of rock types dredges including fresh and variously hydrothermally altered cobbles of pumiceous dacite with black glass through to white, soft sericitic clay with sulfide. An area of approximately 250x250m contains the most intense clay alteration with massive sulfide and concomitant silicification. A variety of orange-brown oxyhydroxides and friable, black manganiferous wad appears to be peripheral to the central zone of clay/silica/sulfide alteration. The sulfide mineralisation has some characteristics of vein-style epithermal systems. Except for one 3-cm-size piece of massive pyrite which had a chimney-like shape, we did not note the occurrence of chimney-like features. Most sulfide-bearing dacite blocks have features indicative of seafloor oxidation (brown-orange jarositic stains). Mineralisation appears to be associated with an apparent fault-controlled N-S trending topographic low (possibly a graben). Fine-grained, grey dacite was dredged 2 km SE of the western summit, whereas fresh dacite with biological debris (corals, gastropods, cone shells and red shrimp) was obtained from a site on the western flank in an area described by the MMAJ/SOPAC cruise as a hydrothermal vent community.

Weepip Seamount

Samples obtained from Weepip Seamount (water depth >750m), northeast of Vella Lavella, included a peculiar assemblage of deep water biota such as branching corals, gastropods, brachiopods, scallops and red shrimp. Given the depth of the seamount and distance from the photic zone, we postulate that these organisms may be associated with a hydrothermal vent community. This was not, however, supported by CTD work, although low temperature venting would not necessarily give rise to a significant hydrothermal plume. Rock sampling of this area was limited and no obvious alteration was observed; however, barite crystals were observed in mudstones and mudstone impressions in scallop shells.

Kavachi

Hydrothermally altered basalt and sulfide-rich volcanic breccias were sampled at Kavachi. Magmatic-hydrothermal activity is probably the cause of staining of the water observed around the volcano, and the pumice-like volcanic precipitates seen during eruptions, and the sulfide mineralization and sulfur coatings on some of the lavas. Cream to dark grey (in stark contrast to black fresh volcanic rock) , hydrothermally altered basalt comprises up to 5-10% by volume of the material recovered in grabs, and typically is in the order of around 1% by volume. This material was particularly prominent in a grab sample to the NW of Kavachi but some hydrothermal alteration was evident at nearly all stations sampled. Fine- to medium-grained pyrite (and other as yet unidentified sulfides) infill vesicles and coat fractures, closely associated with late stage silica alteration and the most intense bleaching. Fine-grained sulfides also occur as disseminations and veinlets within mm-cm-size altered basalt fragments comprising volcanic breccia. These observations indicate that sulfide deposition has been prolonged and multi staged.

One 5-cm-long push core through a grab sample collected from the SW flank of Kavachi showed distinct layering and colour changes associated with grain size differences of sand and silty mud. This section was sampled in detail, and subsequent laboratory analysis is expected to reveal important clues about Kavachi’s recent eruption history.

ACTIVITY 3. Marine Chemistry/Hydrothermal Fluids

Hydrocast/CTD Operations (Ed Baker)

Hydrographic and optical data during FR04/00 were acquired with the Franklin's Neil Brown CTD system #8. Coupled to the CTD fish were two optical sensors, a SeaTech 25 cm-pathlength transmissometer (S/N 247D) and a SeaTech light backscattering sensor (S/N 265). These optical sensors were used to identify possible hydrothermal plume targets in real time to guide bottle sampling and to provide a continuous profile of the distribution of suspended hydrothermal particulates where present. CTD data were processed as 1 s averages of each parameter on each up- and down-cast, or dip. Because no provision was made during the cruise for salinity determinations of bottle samples, the final salinity and density data will be uncorrected for inaccuracies of the Neil Brown conductivity sensor.

CTD operations during the cruise were of two types, vertical casts and tow-yos. Vertical casts provided stationary profiles at chosen locations. Tow-yos provided two-dimensional representations of hydrographic and optical parameters along a track line. Tow-yo data were acquired by continuously winching the CTD package up and down through a chosen portion of the water column while Franklin steamed along a defined track at ~1.5 kts, thus producing a saw-tooth track of the CTD over the ground. A total of 25 casts and 7 tow-yos were conducted during the cruise.

Fluid Chemistry (Gary Massoth and Cornel deRonde)

Seawater (plume) samples collected using the conductivity-temperature-depth-optical sensor (CTDO) profiling and sampling system were analyzed onboard for pH and processed for subsequent shore-based determinations of d ³He (John Lupton, NOAA/PMEL), CH₄ (Jun-ichiro Ishibashi, Kyushu University), total dissolvable and dissolved metals (TDMe and DMe, where Me = Fe and Mn, Gary Massoth, GNS), and suspended particulate matter (SPM) concentration, composition, and morphology (Richard Feely, NOAA/PMEL).

Seawater aliquots were drawn from CSIRO General Oceanics standard 10-L Niskin bottles refitted with silastic tubing springs for closure. The first aliquot was always for d ³He, to minimize contamination by atmospheric He drawn into bottle headspace during sub-sampling. He samples were stored in 25cm long sections of 5/8" diameter Cu refrigeration tubing cold-welded at both ends to insure sample integrity. Samples for methane were next drawn into 50 or 100 mL glass flasks, preserved with 0.025% (wt./vol.) HgCl₂, and sealed with rubber stoppers held in place using crimped Al foil. These samples were stored in the dark and at refrigerator temperature until analysis. Aliquots for pH were collected in 60-mL plastic bottles tightly sealed with convex closures to preclude air, and stored in a water bath to obtain a homogeneous temperature for analysis. Sub-samples for TDMe, DMe, and SPM were collected on-deck, directly from the Niskin samplers, into acid-cleaned HDPE reservoirs via a shielded transfer tube. Metal samples were acidified to pH 1.9 using ultra-clean HCl. DMe, SPM, and SEM samples were obtained by vacuum filtration through acid-cleaned and pre-weighed 37 mm diameter, 0.4 µm pore size polycarbonate membranes.

Determinations of pH were made within an hour of sub-sampling, after the samples had attained laboratory temperature ± 1°C. An Orion ROSS Sure-Flow combination semi-micro pH electrode and an Orion 290A pH meter were used to conduct the determinations of pH. Samples were stirred during analyses until equilibrium mV readings were obtained (about one minute). The measurement system was calibrated daily using NBS referenced buffers measured at room temperature. Adjustments for the small temperature offsets observed were not attempted.

Plume Reconnaissance

A total of 25 CTDO vertical profiles, 7 tow-yo distributions, and 282 discrete water samples were collected during SHAARC 2000. Over 280 determinations of pH were made (contact Gary Massoth for full data set). Discrete sub-samples for determination of d ³He (281), methane (272), TDMe (282), DMe (10), SPM (4) and SEM (12) were obtained in various quantities. We describe below the sub-distributions for the three study regions and our respective preliminary findings.

New Ireland Fore-arc

Eleven vertical CTDO profiles and two tow-yo distributions together with 116 discrete water samples were collected in this region. Our primary objective at this and the other study areas was to conduct a systematic regional survey of submarine hydrothermal discharge as evidenced by the presence of hydrothermal plumes. Plume prospecting, when conducted in systematic fashion, is probably the most efficient and conclusive way to survey for hydrothermal activity over large regions of the seafloor. The suite of tracers we detect, both physically and chemically, have proven on numerable occasions to be reliable indicators of hydrothermal activity. Here, we systematically assessed a 225-km section of the New Ireland fore-arc extending linearly from east of Feni Island to west of Lihir Island (see location figure). Despite sampling near sites inferred to be hydrothermally active based on previous surveys (e.g., Edison and New World seamounts), we found no clear evidence from our sensors or onboard determinations for active hydrothermal discharge anywhere within the New Ireland basin study region. These results do not preclude the presence of hydrothermally active sites, but the fluids they discharge could not be metal or CO₂ rich and thus are unlikely to be efficient transporters of heat and mass to the seafloor and ambient ocean. It is still quite possible, however, that laboratory determinations of d ³He, methane, Fe, and Mn may reveal low-intensity plume distributions here that may be of interest in a regional oceanographic sense.

Solomon Islands — New Georgia Group (fore-arc)

We obtained 9 CTDO profiles, five tow-yo distributions, and 111 discrete seawater samples from the Solomon "fore-arc" study area. Five sites were targeted for exploration: directly offshore a geothermally active zone on Vella Lavella Island, an historically active submarine volcanic site south of Simbo Island, a 42-km-long survey section linearly aligned with and including the Kana Keoki and Coleman submarine volcano summits, and Kavachi submarine volcano. With the exception of the latter site, no "real-time" indications of hydrothermal activity were detected in this region of the Solomon fore-arc. At Kavachi, however, we witnessed numerous phreato-magmatic eruptions that breached the sea surface. Our onboard plume mapping capabilities provided an unprecedented opportunity to view, from a submarine perspective, the phenomenon of volcanic arc island building. This survey, intended to identify hydrothermal plumes emanating away from the eruption site, may also have illuminated a mode of mass wasting associated with arc volcanism (turbidite plumes). Using the CTDO profiling and sampling package in tow-yo fashion, we were able to construct a hydrographic "ring fence" at a distance of about 750 m from the volcanic eruptive center, completely encircling the volcano near the 500 m isobath. Three additional vertical profiles were obtained in a triangular pattern 5-km equidistant from the volcano summit. A complex layering of particle plumes was imaged, extending from the sea surface, throughout the mid-water region, and down to near-seafloor depths. Light scattering anomalies suggest that eruption-related particulates were being advected predominantly into the northern quadrants, and to distances of at least 5 km. Preliminary examination by binocular microscope of suspended particles from the deepest plumes show that course-grained volcanic fragments are dominant. By contrast, samples collected from the surface waters are smaller in grain size although still composed primarily of volcanic fragments. Onboard results for pH (no detectable anomalies) provide indirect evidence that a CO₂ rich plume did not emanate from the eruptive site. This suggests that indeed many of the light scattering plume anomalies are probably of non-hydrothermal (i.e., turbidite flows?) origin. Whether a substantial chemical plume was associated with this shallow eruption of an arc volcano remains to be determined. Regardless of the outcome of our shore-based analyses for magmatic components in the observed plumes, these results will have interesting implications regarding the transfer of heat and mass to seawater when exposed directly to molten lava.

Eastern Woodlark Basin, Segment 5b

Five vertical CTDO profiles and 55 discrete samples were obtained from this eastern-most segment of the Woodlark spreading center system. With limited time available to study the Woodlark spreading system, we opted to conduct a systematic survey of a single spreading segment rather than isolated casts on several of the 13 segments that comprise this 550-km-long system. We selected segment 5b (using the terminology of Martinez et al., 1999) for several reasons: 1. it is 35 km in length, and thus manageable relative to our systematic survey philosophy and time constraint; 2. it is the fastest spreading (67 mm/y) of all the Woodlark segments; and 3., it has the unusual morphology of a deep (>4400 m) valley rather than an inflated ridge crest more typical of this spreading rate. Of particular interest to our fore-arc work, the basalts of the Woodlark have a distinctive arc-like geochemical overprint, which is most-pronounced closest to the arc at segment 5b. Here, localized near the western end of the segment (Station SSHV00A24) and just below basin sill depth, we detected an apparent hydrothermal plume identified by a strong light scattering anomaly coincident with a slight decrease in pH. While experience would suggest these anomalies are not a false positive, final confirmation must await the results of shorebased chemical determinations. If confirmed, this discovery would be the first instance of significant and localized hydrothermal activity within the Woodlark spreading system. Subsequent location of the discharge site and characterization of the possibly "arc contaminated" venting fluids could provide unique insight into subduction related volcano-hydrothermal processes occurring in arc/back-arc settings

Cruise Narrative (Brent McInnes)

The cruise began in the port of Rabaul and ended at the port of Darwin. The cruise had 4 stages, with two stages dedicated to research and two stages dominated by transit.

We left Rabaul at 0930 on May 5^th and began our research operations in the Tabar-Lihir-Tanga-Feni island arc after a 0.5 day transit. Because the US authorities turned off GPS dithering on May 1, 2000, FRO4/00 operated with GPS accuracy of ± 10m throughout the length of the cruise. DGPS (± 2m) was in operation only until May 11^th at 1700hrs, as we passed outside of OPTUS satellite footprint coverage east of 156° 32.31’E. All times reported are Australian Eastern Standard Time.

Stage I (Fig. 1) involved 30 sampling stations and 2 echo-sounding traverses in the TLTF arc region spanning a time period of 5 days. The general path of activities took us from SE to NW along the Tabar-Lihir-Tanga-Feni chain, beginning at Feni island. Dredging operations were conducted along the submarine flanks of each of these volcanoes in order to acquire glass-bearing samples for petrological and geochemical study. Although the combination of unfavourable wind and current conditions limited the number of sampling locations, these sampling programs were successful. Dredge and grab sampling programs were also successfully implemented on 2 seamounts south of Lihir island (Edison and Tubaf Seamounts) for the primary purpose of collecting xenolith samples of the oceanic lithosphere. These programs recovered over 250 xenoliths of peridotite, gabbro, basalt and crustal sediments.

A research program designed to investigate the sedimentation history of the post-collapse features of the Kinami and Luise calderas involved 5 sediment core and 2 grab sampling operations. Grab samples were taken in place of sediment coring when the bottom substrate was considered too sandy to allow adequate penetration of the core barrel.

CTD/Hydrocast operations (vertical and tow-yo) were carried out over Edison, Conical and New World Seamounts, with the goal of investigating the hydochemical signature of hydrothermal plumes thought to be emanating from Edison seamount (Herzig et al., 1994) and a possible methane seep identified during SO-133. An additional 10 vertical hydrocasts were carried out during Stage I in order to define the regional background signal and to search for new anomalies. All CTD/Hydrocast operations were carried out successfully. Although significant anomalies were not detected in the TLTF region by CTD telemetry and shipboard water analysis, subtle anomalies await post-cruise completion of more sophisticated analyses.

Prior to leaving the TLTF region, FR04/00 conducted an echo-sounding survey NE of Lihir island to search for a seamount supposedly rising 1400m above the seafloor near the Kilinailau trench. This operation was aborted after an echo sounding survey covering a 6 nautical mile block centred on 2° 37.06’S, 153° 13.64’E failed to find any bathymetric feature, and it is presumed that the bathymetric map showing a seamount at this location is incorrect. Some members of the science team referred to it as "Lassiter’s seamount".

Stage II involved a 1.5 day transit to the Solomon Islands research sites via the eastern coast of Bougainville. An echo sounding traverse was conducted en route.

Stage III (Fig. 2) involved 59 sampling stations and 11 echo-sounding traverses in the southwest Solomon Islands (Stage IIIa) and eastern Woodlark Basin (Stage IIIb) spanning a total time period of 7.5 days. Stage IIIa began with the unexpected discovery of a 300m seamount NW of Vella Lavella. Dredge, grab and vertical CTD hydrocast operations of "Weepip" seamount did not detect signs of active volcanic and/or hydrothermal processes. We left Weepip with the feeling that we might have missed a subtle hydrothermal system, but to keep to schedule we needed to proceed to Vella Lavella to investigate the possible submarine extension of the geothermally active Paraso Graben. Sediment cores taken along the main graben fault structure will be studied and analysed for metallic trace elements.

The next phase of research involved the systematic sampling of key bathymetric features of the Simbo and Ghizo Ridges in the forearc region of the New Georgia Group, Solomon Islands. Although this region has excellent multibeam bathymetry maps, the petrology, geochemistry and metallogeny of this volcanic environment is poorly understood. We retrieved volcanic rock samples by dredge/grab sampling 7 volcanic seamounts in this region, the majority of which had never been sampled before. The geochronology and geochemistry of these newly obtained samples will generate a better understanding of the geological evolution of the Simbo-Ghizo Ridge region.

The two principal volcanic features along the Ghizo Ridge are the Coleman and Kana Keoki seamounts. These giant volcanoes have basal diameters of approximately 17 km and tower 2.4 km above the surrounding seafloor. Because they are of Mount Fuji proportions, we conducted our dredge and grab sampling programs with a substantial degree of humility, choosing targets based on topographic features knowing that we would not be able to systematically sample these volcano giants. A 34 km long hydrocast/CTD operation (combined vertical and tow-yo) across the summits of both Coleman and Kana Keoki seamounts was interspersed with sampling operations.

Figure 3. A phreatomagmatic eruption at Kavachi volcano on May 14^th 2000.

On May 14th, FR04/00 visited Kavachi Seamount in the Solomon Islands and observed it for a period of approximately 20 hours. Upon arrival at 0600 Australian Eastern Standard Time we observed waves breaking on an apparent volcanic peak and the volcano in almost a continuous state of eruption (Figure 3). This was an unanticipated and unforgettable experience for the Franklin crew and FR04 scientists!

The science team put their heads together to implement an ad hoc research program to document this apparently new phase of island building eruptive activity:

1. Bathymetric mapping - We were able to accurately fix the position of the eruption column at 8° 59.65'S, 157° 58.23'E and estimate that the peak of the volcano lies 2-5 m below current water level. The eruptions ejected ash and incandescent blocks of lava up to 70 m above sea level, and sulfurous steam plumes mushroomed to 500 m. A bathymetric map was produced by several transits across the volcano at distances between 0.75 and 5 km of the eruptive centre. The bathymetric map is significantly different than the map provided in Johnson and Tuni (1987).

2. Hydrocast/CTD operations - Both vertical and ring-fence tow-yo hydrocast/CTD surveys were conducted and detected numerous chemical and particle plumes in the water column that extend at least 5 km from the eruptive centre.

3. Systematic rock and ash sampling - Grab and dredge sampling of rocks from the flanks of the volcano returned glassy, vesicular basalts and hydrothermally altered sulfide-rich volcanic breccias.

4. Photography and Filming - Approximately 500 photos and 2 hours of digital video of the eruptions were shot in daylight conditions. Apart from an 80 minute hiatus before noon, the volcano erupted every 5-7 minutes and lasted for 2-3 minutes, and we have captured approximately 20 eruptions on video.

After completing the Kavachi study, Franklin turned west to complete the Ghizo and Simbo Ridge seamount dredging programs. This consisted of dredging of 4 (of 5) seamounts on the eastern and southernmost flank of the Simbo Ridge, taking less than 24 hours.

Stage IIIb began after we had completed our objectives as specified in the cruise plan ahead of schedule. The science team decided to continue exploration west of the Simbo Ridge and into the eastern Woodlark Basin designated by the geophysical community as Segment 5b. This region is a virtually unexplored mid-ocean ridge spreading system, and worthy of research because of: 1. A moderately fast spreading rate coupled with unusually deep bathymetry; 2. Proximity to a triple-junction (the location where 3 tectonic plates intersect a subduction zone); 3. And the fact that it lies to the back (north) of the Paleocene Pocklington arc system. With less than 36 hours of operation remaining, we assumed a reconnaissance philosophy and interspersed 5 CTD operations with 4 grab operations, all at depths exceeding 3500m. The fourth CTD operation really got the juices flowing as an apparent hydrothermal plume was detected. Unfortunately, this was not detected in the 5^th and final CTD operation, and because we were running out of time, we marked an "X" on the map and in Macarthur tradition, vowed to return. Grab sampling was considered to be a more time-efficient method of deep water sampling than dredging, and John Chadwick implemented a "two-drop, petroleum jelly" grab technique to ensure that the grab fired appropriately and that at the very least, some of the bottom material would get stuck into the jelly at the bottom of the grab even if the grab returned empty. The Chadwick method was 5 for 5 in sample collection, and 4 for 5 in basalt collection. Fortunately, modern analytical techniques do not require big samples. The final sampling station was a dredge site at Seamount "J", a substantial volcanic edifice sitting in the middle of the Woodlark Basin, which was completed approximately 1 hour ahead of the time of beginning the Darwin departure.

Stage IV involved the final transit stage to the port of Darwin. It took 1 day to transit from the final station at Seamount "J" in the Woodlark Basin across the Solomon Sea to the Louisiade Archipelago, and an additional 4 days to transit from the Louisiade Archipelago (eastern tip of PNG) across the top of Australia to the port of egress in Darwin.

Summary

During the 12.5 days of non-transit research operations, a total of 88 sample sites were visited. The success rate for obtaining data (rock, sediment and/or water samples) was 94%. The total number of sites where samples were successfully acquired exceeded the cruise plan by 205%! Although the original sampling plan was conservative in nature, this variance can be attributed to 4 reasons:

1. Efficient teamwork of the FR04 scientists

2. High level of professionalism of the Franklin crew

3. Good weather and sea conditions

4. Limited operations down-time due to equipment malfunctions.

Overall, the scientific participants were pleased with their experience using the Franklin. The quality of the meals and the amicability of the crew was outstanding. Some frustration was experienced with the computer facilities required to carry out operations, particularly the multiple operating systems and the non-user friendly interface. Crashes in the navigation software while preparing to hold station and deploy equipment was the cause of a few tense moments for both the helm and science watch, and lead to time wasted "going around" while the computer was rebooted and the positioning data refreshed. These computer problems are not a reflection of shortcomings of the ORV Support personnel who performed admirably on problems beyond their control, but rather point to the need for an investment on integrating and upgrading the current system. Ethernet access linking all cabins and operations rooms would improve the transfer of data between carry-on laptop computers and reduce the e-mail usage overload on available shipboard workstations. The ergonomics of the operations room could also be improved to streamline staff movement between computers, display terminals and map tables.

Personnel

Scientific Party

ORV Support

CEM = CSIRO Exploration & Mining
CMR = CSIRO Marine Research
GNS = Institute of Geological and Nuclear Sciences
ANU = Australian National University
NOAA = National Oceanic and Atmospheric Administration
UPNG = University of Papua New Guinea
UF = University of Florida
1 = Chief Scientist, 2 = Principal Investigator, 3 = Graduate Student

Crew

Acknowledgements

The scientific participants on FR04/2000 express their sincere appreciation to Master Neil Cheshire and the crew of the RV Franklin for their skill and professionalism. The RV Franklin is managed by CSIRO Marine Research as a National Facility and we acknowledge the efforts of Neal Denning, Don McKenzie and Ron Plaschke in assisting in our cruise planning and preparations. The Chief Scientist would like to thank Ray Binns, Chris Yeats and Joanna Parr of CSIRO Exploration and Mining for advice, technical expertise, equipment and effort in the preparation for FR04/2000. Richard Arculus acknowledges support from the Australian Research Council. Ed Baker acknowledges support from the NOAA Vents Program. Gary J. Massoth and Cornel E.J. de Ronde acknowledge New Zealand PGSF Programme Geothermal and Minerals Research (GMR).

Brent McInnes
Chief Scientist, FR04/2000

Updated: 31/01/03