DRAFT 2, 8/31/02
SUMMARY OF DEEP UNDERWATER RESEARCH 1998-2002
JAPAN-USA COOPERATIVE PROJECT ON EVOLUTION OF HAWAIIAN VOLCANOES
Hawaiian volcanoes are classic examples of intraplate mantle plume volcanism. These volcanoes host frequent large eruptions, generate major earthquakes, and are prone to flank failure causing enormous landslides and gigantic tsunami. Geologic hazards associated with the volcanoes are not just restricted to those living in the Hawaiian Islands but also extend to the populated coastal regions of the Pacific. Hawaiian volcanoes are the best known volcanoes in the world, but most previous studies have focused on the easily accessible subaerial parts of these volcanoes and largely ignored their submarine flanks. Thus, knowledge of these immense volcanoes has been based largely on the tips of these volcanic 'icebergs.’
The vast bulk of Hawaiian volcanoes are under the ocean (to depths
of 5700 m), and exploration of these areas has been difficult, costly, and very
limited. Many
previous studies have focused on subaerial parts of Hawaiian volcanoes, but the
deep-water flanks of the edifices (maximum depths, 5,700 m) have remained
poorly known until recently. In
1998, a collaborative Japan-USA program was initiated to explore the evolution
of Hawaiian volcanoes including their growth and degradation, making use of the
deep-sea research capabilities of the Japan Marine Science and Technology
Center (JAMSTEC). During a four
week cruise in 1998, the ROV Kaiko (Remotely
Operated Vehicle), supported by its mother ship RV Kairei, made 10 dives at depths to 5,200
m for direct sampling and video observations, supplemented by dredging, piston
cores, and SeaBeam bathymetric surveys.
During a seven week cruise in August-September 1999, the Shinkai 6500 submersible made 29 dives from the RV Yokosuka for sampling and direct observations as deep as 5,560 m,
and further SeaBeam surveys were obtained. A third cruise in 2001, utilizing the Kairei and Kaiko , made
17 ROV dives and collected about 300 outcrop samples, obtained 10 new piston
cores, made several dredge hauls, and generated additional SeaBeam
coverage. During this year’s final
cruise in the present project, the Shinkai
6500 made 30 more dives, supplemented by additional geophysical surveys. Participating scientists have been
largely from JAMSTEC, the University of Hawaii, several universities in Japan,
the Monterey Bay Aquarium Research Institute, the U.S. Geological Survey,
and the Geological Survey of Japan.
Knowledge of the submarine
region around Hawaiian volcanoes has changed dramatically in the last few years
as a result of these JAMSTEC cruises. The expeditions have generated improved
sea-floor bathymetry and side-scan sonar images, utilized ROV and submersible
vehicles to collect a large suite of precisely located samples for petrologic
study, acquired detailed photo and video images for many critical areas, and
generated gravity, magnetic, and seismic reflection-profiles. Such materials, supplemented by
petrologic, geochronologic, and other laboratory studies, provide the basis for
investigating a wide variety of geophysical phenomena. These include the sources for and
extent of plume magmatism, processes associated with active and ancient
landslides on oceanic island volcanoes, seismic structure and tectonic
processes on active volcanoes, nature of rift zone and other submarine
volcanism, growth of oceanic island volcanoes, and hydrothermal processes. In concert with the field and
laboratory studies, recent theoretical investigations have explored the causes
and consequences of Hawaiian plume magmatism and the landslides associated with
the resulting unstable volcanoes.
Five major topics for the field surveys and laboratory research have been: (1) magmatic evolution of
the large island volcanoes of the Hawaiian Ridge in relation to the source
mantle plume, (2) morphology and structure of submarine rift zones of these
volcanoes, (3) landslide failure of their unstable flanks, (4) off-ridge
ocean-floor volcanism that samples periphery of the mantle plume, (5)
piston-core stratigraphy of the distal sedimentary record of growth of the
Hawaiian ridge, and (6) general geophysical surveys—bathymetry, side-scan
images, and gravity, magnetic, and seismic profiles.
(1) Plume volcanism: The Hawaiian
Islands are products of a mantle plume, which is thought to originate at a
boundary layer deep within the mantle, perhaps at the 660 km discontinuity or
at the core- mantle boundary. Thus, the volcanism associated with plumes provides
a window into the deeper mantle and the possibility of access to the
geochemical record of plate recycling and mantle. The Hawaiian plume is the Earth's hottest, most productive
and most thoroughly studied mantle plume.
Nonetheless, a lively debate continues on such fundamental questions as
the nature and scale of heterogeneities within the plume, magma generation
processes, and the composition and temperature of the magmas that supply
Hawaiian volcanoes. One major
result of the 1998-2002 deep underwater research has been recognition that
many, perhaps all the Hawaiian volcanoes have evolved through a succession of
compositions, rather than having distinctive individual compositions,
especially during the tholeiitic shield-building stage. Thus, the previous
distinctions between Kea-type (e.g., Kilauea), Loa type (Mauna Loa), and Koolau
type now are recognized as compositions
that can erupt sequentially during growth of individual edifices.
Another controversy continues over
whether the bulk of magmas that supply Hawaiian volcanoes are basaltic, similar
to those commonly erupted (8-12 wt % MgO), or whether they are picritic (MgO
>15 wt %). Deep dives along
distal parts of rift zones of several volcanoes (Puna Ridge of Kilauea, Hilo
Ridge of Kohala(?), Hana Ridge of Haleakala) has shown that olivine-rich lavas
(picrite) are much more abundant than subaerially, but no evidence has been
found for common eruption of high MgO liquids. Resolution of such questions is critical to understand the
nature and dynamics of the Hawaiian plume and the processes of magma generation
within it.
Determining the long-term compositional variations of lavas from the tholeiitic shield-building stage of Hawaiian volcanoes is also necessary to resolve these questions. The subaerial portions of oceanic volcanoes provide samples of only a small fraction of a volcano's overall history (5-10%), biased in favor of its most recent development, and may be nonrepresentative of the bulk of the volcano, and especially its early growth. Break-away scarps of submarine landslides and their deposits provide the opportunity to sample the deep interior of Hawaiian volcanoes, and much new data, especially on extremely alkalic early volcanism from ancestral Kilauea volcano, has been obtained from the JAMSTEC program. Especially important for further petrologic studies, these new results for the younger volcanoes provide confirmation of appropriate sampling strategies that can in the future be applied elsewhere along the Hawaii-Emperor Ridge.
(2)
Volcanic submarine rift zones: Another major objective
of the cooperative JAMSTEC program has been to explore the stratigraphic and
structural evolution of oceanic islands including their growth and degradation. Major emphasis during the recent cruises
has been on the growth of underwater volcanic rift zones and on landslide
failures on the flanks of the volcanoes.
Rift-zone studies have focused on the Puna Ridge (submarine east rift
zone of Kilauea), Hana Ridge (submarine east rift zone of Haleakala), and the
Hilo Ridge (east rift zone of Kohala?).
Dives and other data have explored the diversity of constructional
features along the rift zones, including elongate vents, steep cones, and
flat-topped cones that probably represent lava ponds behind marginal
levees. Flanks of rift zones are
draped by pillow lavas, with complex morphology probably due to satellitic
venting from fractures and lava tubes.
Low on the flanks, primary pillow lava accumulations are disrupted by
slumping and lateral gravitationally driven spreading. Gravity and magnetic profiles are being
used to evaluate the distribution of dike intrusions longitudinally and
transversely along the rift zones.
(3) Landslide failure of volcanic flanks:
Giant landslides are now widely recognized along the flanks of many
oceanic volcanoes, such as Hawaii, Marquesas, La Reunion, Galapagos, and Canary
Islands. The landslides appear to
form both early during the growth of the volcano as it is centered over the hot
spot and after it drifts off. The
abundance of landslides demonstrates that mass-wasting processes play an
important role in the construction and evolution of oceanic-island
volcanoes. Not only do such
processes modify the surfaces and slopes of the islands, they also are closely
linked with major geologic hazards, including earthquakes associated with slope
failure, large-scale submergence or emergence of coastlines, and massive
tsunamis which can destroy life and property. Due to the unpredictable and sporadic nature of such massive
landslides, the processes and timing associated with these events remain poorly
understood. Yet the significance
of landslide features in the evolution of volcanic islands, and their
extraordinary destructive potential, make it imperative that we understand
their history and behavior, and assess their impact on human development of
volcanic islands and adjacent coastal areas. The 1998-2000 project initially focused on the active the
Hilina slump offshore of Kilauea volcano and the enormous Nu’uanu and Wailau
slides north of Oahu and Molokai’I; during the more recent cruises, the
landslide studies were expanded to include the South Kona slide complex and the
Wai’anae slump.
The focus on landslide deposits and the
scars they produce has provided a window into the deep structure and
compositional evolution of Hawaiian volcanoes. This approach also offers opportunities to reconstruct the
deformational sequence of Hawaiian slides, to better constrain static and kinematic
models for landslide initiation and movement, and to provide data for the
development of models for destructive landslide-generated tsunamis.
High-resolution seafloor mapping of the U.S. Exclusive Economic Zone (EEZ)
using the GLORIA side-scan sonar system previously revealed the presence of
more than 68 giant landslides along the flanks of the Hawaiian volcanoes. Using
the new results for slides from Kilauea and Koolau as models, we have gained
greater insight into the landslide processes and are better able to assess the
potential hazards they present to human life and property in Hawaii and around
the Pacific Rim from the associated earthquakes and seismic sea waves.
(4) Off-ridge volcanism:
A fourth focus of the Hawaii project has been to image and sample
off-ridge volcanic areas, especially the North and South Arch volcanic fields
and the newly documented Southwest Oahu volcano. These sites of alkalic volcanism provide special
opportunities to study magmatic processes at the periphery of the mantle plume
and associated with crustal warping that has produced the Hawaiian Arch as the
axis of the Hawaiian Ridge subsides under the load of growing volcanic
edifices. Of special interest will
be the noble gas compositions of these deep-water lavas, which have thereby
been shielded from low-pressure gas separation.
(5)
Distal sedimentation record: Gravity piston cores,
obtained for ocean-floor sediments at 15 distal sites at distances of 100-200
km from shorelines of the Hawaiian volcanoes, have yielded a remarkable
turbidite stratigraphy of basaltic sand layers that appear to record major
slope failure events on the Hawaiian Ridge during the last few million
years. When such layers can be
correlated by microprobe analysis of glass sand grains, and ages directly
determined from paleomagnetic correlation or inferred indirectly from
sedimentation rates, they can provide records of major events in the growth and
degradation of Hawaiian volcanoes for which proximal evidence is completely
buried or otherwise unavailable.
(6)
Geophysical surveys: These include seafloor mapping with a
multibeam sonar system (SeaBeam 2112), , and single channel seismic, gravity,
and magnetic surveys. The JAMSTEC sonar surveys are the
most comprehensive synoptic data-set for the Hawaiian Islands since the
lower-resolution GLORIA surveys by the U.S. Geological Survey in the
1980’s. The SeaBeam system
produces wide-swath contiguous contour maps and acoustic backscatter
images. The surveys generated
detailed maps of previously known features and imaged some features for the
first time. The new surveys determined block distribution, orientation, and
structure of the enormous Nu’uanu and Wailau landslides, north of the islands
of Oahu and Molokai.. Many of the
volcanic rift zones were surveyed, including the Puna, Hilo and Hana ridges in
their entirety. Slumped flanks of
several islands were mapped, including Hilina, South Kona, Laupahoehoe, and
Pololu (Hawaii Island), Hana (Maui), and Clarke’s landslide (SW of Lanai).
The gravity and magnetic profiles are
the first comprehensive potential field data set, which have been consistently
located by rigorous GPS navigation.
These data should provide important new interpretations of deep volcano
structures, especially when merged with existing on-land potential field
data. Seismic surveys were
especially valuable for determining sediment thickness and structure in ponded
basins within landslides and slumps, and on low-relief seafloor distal to the
submarine island flanks.
Shipboard
summary of: major 2002 cruise results
The 30 Shinkai dives in 2002 focused on many topics, including completing multi-year studies on the Nu’uanu landslide, Hilina slump, Puna Ridge, and Lo’ihi Seamount. Major additional foci of multiple dives during the 2002 cruise included Hana Ridge, South Kona slump-slide complex, Southwest Oahu volcanic field, and diverse features on the northeast flank of Kohala volcano including the Pololu slide, Laupahoehoe benches, and Hilo Ridge. Much will be learned from laboratory analysis of samples and images acquired during this cruise, but some preliminary observations and interpretations already seem significant. Among important new results, based on direct visual observations during dives and initial shipboard examination of video records and rock samples, were:
(1) The most distal large block of the
Nu’uanu landslide consists largely of subaerially erupted material (clasts of
oxidized pahoehoe lava), despite its location about 140 km north of its source
region on Oahu, and requiring that Ko’olau volcano have developed a large
subaerial edifice by the time of sliding (S700). The proximal bench also contains abundant subaerial clasts,
suggesting that this feature results from post-Nu’uanu depositional processes
(S713) rather than being a remnant of a pre-landslide “mid-slope bench,” as
previously inferred. In both
suites of dive samples, determining relative proportions of Ko’olau versus Loa
and Kea petrologic basalt types will be of great interest in interpreting the geometry
of the source volcanic edifice.
(2) At the southwest “corner” of the Hilina bench,
the abundance in breccias of oxidized vesicular pahoehoe clasts, that must have
been subaerially derived and seem to have come from Mauna Loa, indicate that a
clastic submarine apron from this volcano likely extends at shallow depth
beneath much of the southwestern Hilina slump (S710). Also traversed were the
upward continuation of a frontal-scarp dive S508 in 1999, to provide a
continuous section to the mid-slope bench (S708), and a long pillow-lava rib
above the mid-slope bench to search for the change from alkalic to tholeiitic
Kilauea magma (K208, S709).
(3) Deep east and west flanks of Lo’ihi
Seamount, exposed in slump scarps, contain relatively old-appearing pillow
lavas as judged by thickness of palagonitized glass and Mn coatings. Their aphyric textures suggest probable
alkalic compositions, possibly datable by K-Ar or Ar-Ar methods (S696-697).
(4) Visual dive observations, samples,
and bathymetric and sidescan
surveys show that the Hana Ridge is morphologically, structurally, and
compositionally complex. The
arcuate distal amphitheater appears to result partly from slope failure (north
prong), partly from primary construction of a younger southerly rift prong
(S684). Diverse pillow lithologies
collected along the rift flanks are available for detailed petrologic study
(S687, S691). Arcuate ridges,
benches and closed basins on the north side of Hana Ridge, that are known only
from bathymetric and sidescan data, seem most analogous to the Hilina and Laupahoehoe
slumps, and appear to record unique volcanic spreading along the relatively
distal portion of a long rift zone
(5) The benches and slump blocks of the
South Kona complex contain exceptionally diverse lithologies, in various dives
consisting dominantly of coherent pillow lava (S693), subaerially erupted
pahoehoe (S695), finely cominuted a’a clinker (K210), and heterogeneous
subaerial and probable submarine lava clasts (S689, S514-515). Glass sandstone and coarse
hyaloclastite were sparse in all 2002 dives. Rocks in the South
Kona slide complex appear largely to have resulted from large-scale
gravitational slumping and landslide failure of subaerial portions of the west
flank of Mauna Loa’s SW rift zone.
In these respects, South Kona is notably different from Hilina slump
south of Kilauea.
(6) The Southwest Oahu volcanic field consists of
both flat-topped and steep-sided cones and low-relief lava flows (S501-503).
The presence of evolved alkalic magmas distinguishes it from alkalic lava fields on the North
and South Arch.
(7) Kohala is a larger and more complex
volcano than previously recognized.
Lower slopes of the Laupahoehoe slump contain diverse hyaloclastite
breccias and sandstones that record relatively early slope failure of the north
flank of this pre-Mauna Kea edifice (S711). In contrast,
the Pololu slump contains large amounts younger-appearing weakly indurated black sand, along with
abundant clasts of subaerially erupted pahoehoe, that may provide the record of
slope failure at a later stage of volcano growth (S712). Glasses from the Laupahoehoe and Pololu
slumps may provide the first reliable compositions for Kohala’s tholeiitic
stage, unmodified by subaerial weathering. The Hilo Ridge, far to the east, contains both
tholeiitic and transitional pillow
lavas (K215, S699); samples from these dives should help evaluate its somewhat
controversial interpretation as the distal part of a long east rift zone of
Kohala volcano.
(8) Following the discovery in 1998 of
early submarine-erupted alkalic products of Kilauea at the Hilina slump, it was
anticipated that similar very early stages would be exposed in the failed
flanks of other Hawaiian volcanoes.
Examination of the Nu’uanu landslide does show that Ko’olau magma
compositions changed with time, but no exposures have been found of the early
alkalic phases of any volcano other than Kilauea and the previously known
Lo’ihi Seamount. Among other
possibilities, it may be that the thoeiitic stages of mature Hawaiian volcanoes
are so much larger than the early alkalic stage that complete edifice failure
would be required to expose the early alkalic core.
Initial results of the 1998-99 cruises were reported in a special session on “Magmatic and Tectonic Evolution of Hawaiian and Other Hot-Spot Volcanoes” (E. Takahashi, convenor), at the AGU Western Pacific Geophysics Meeting in Tokyo, June 2000; AGU Monograph 128, based on the special session at WPGM, was published late in 2001; and initial results from the 2001 cruises have already been reported at meetings in Japan and the USA. Additional research, including results from the 2002 cruise, will be reported at the 2002 Fall AGU meeting in San Francisco, at Goldschmidt and IUGG international meetings being held in Japan in 2003 and at the International Geological Congress (IGC) to be held in Florence, Italy, in 2004. The visual observations, rocks samples, and geophysical data obtained during the JAMSTEC Hawaii cruises should also provide materials for research that will continue for many years in the future