Richard S. Fiske

Deep magma body beneath the summit and rift zones of Kilauea Volcano, Hawaii

A magnitude 7.2 earthquake in 1975 caused the south flank of Kilauea Volcano, Hawaii, to move seaward in response to slippage along a deep fault. Since then, a large part of the volcanos edifice has been adjusting to this perturbation. The summit of Kilauea extended at a rate of 0.26 meter per year until 1983, the south flank uplifted more than 0.5 meter, and the axes of both the volcanos rift zones extended and subsided; the summit continues to subside. These ground-surface motions have been remarkably steady and much more widespread than those caused by either recurrent inflation and deflation of the summit magma chamber or the episodic propagation of dikes into the rift zones. Kilaueas magmatic system is, therefore, probably deeper and more extensive than previously thought; the summit and both rift zones may be underlain by a thick, near vertical dike-like magma system at a depth of 3 to 9 kilometers.

Volcanic Substructure Inferred from Dredge Samples and Ocean-Bottom Photographs, Hawaii

Ocean-bottom photographs from 18 stations and dredge hauls from 35 stations adjacent to the Island of Hawaii indicate that basaltic pillow lava and pillow fragments are the dominant rock type on the crest and flanks of the submarine rift zone ridges, whereas glassy basalt sand and scoria are the dominant type on the submarine flanks of the volcanoes directly downslope from land. These relations indicate that three major rock units comprise different levels of the volcanoes depending on the site of eruption: (1) pillow lavas and pillow fragments are dominant below sea level and are erupted from deep-water vents; (2) hyaloclastite rocks (vitric explosion debris, littoral cone ash, and flow-foot breccias) mantle the pillowed base of the volcano, and are erupted from shallow-water vents, subaerial vents in water-soaked ground, or are produced where subaerial lava flows cross the shoreline; and (3) thin subaerial lava flows make up the visible, subaerial shield volcano, are built atop the clastic layer, and are erupted from subaerial vents. This three-fold structure is similar to the table mountains of Iceland that are built by eruption beneath glacial ice. Large-scale slumping in the clastic layer may modify the submarine slopes of the volcanoes as well as produce faulting and downslope movement of parts of the overlying shield volcano. The slope change produced where the gentler shield meets the steeper pillowed pile can be recognized beneath sea level in the older volcanoes, where it has been submerged by regional subsidence.

Fallout of pyroclastic debris from submarine volcanic eruptions.

Volcanic fallout deposits on land, being widespread and accessible for study, have received much attention and have revealed a great deal about subaerial eruption mechanisms. In contrast, virtually nothing is known about equivalent deposits produced by submarine volcanoes, despite the probable abundance of such material in todays oceans and in accreted volcanic arc terrains. Many submarine deposits may form by the fallout of debris to the sea floor downcurrent from the umbrella region of submarine eruption columns. Experiments on water-saturated pumice and pieces of rock (lithics) show that particles settling to the sea floor at terminal velocities of 10 to 50 centimeters per second will display conspicuous bimodality of particle diameters: pieces of pumice may be five to ten times as large as codeposited lithic fragments. Similar material, erupted into the air and deposited on land, displays less well-developed bimodality; pumice diameters are generally two to three times as large as associated lithics. Submarine fallout deposits are therefore distinctive and may be used to indicate a subaqueous origin for some of the great thicknesses of nonfossiliferous volcanic debris contained in ancient volcanic terrains worldwide whose environment of deposition has been uncertain.

Submarine silicic caldera at the front of the Izu-Bonin arc, Japan: Voluminous seafloor eruptions of rhyolite pumice

Myojin Knoll caldera, a submarine rhyolitic center 400 km south of Tokyo, is one of nine silicic calderas along the northern 600 km of the Izu-Bonin(-Ogasawara) arc and the first anywhere to receive detailed, submersible-based study. The caldera, slightly smaller than the Crater Lake structure in Oregon, is 6 × 7 km in diameter; its inner walls are 500–900 m high, and it has a remarkably flat floor at 1400 m below sea level (mbsl). The caldera collapse volume is ∼18 km 3 , suggesting that more than 40 km 3 of pumiceous tephra may have been erupted at the time the caldera formed. Precaldera seafloor eruptions built a broad volcanic edifice consisting of overlapping composite volcanoes made of rhyolitic lavas, shallow intrusions, and a variety of volcaniclastic deposits—including thick accumulations of rhyolitic pumice erupted at 900– 500 mbsl. The caldera-forming eruption produced a 150–200 m deposit of nonwelded, fines-depleted pumice that resembles a colossal layer of popcorn at the top of the caldera wall. Freshly erupted pumice behaved as “sinkers” or “floaters,” depending on the environment in which it cooled. The pumice clasts deposited proximally and exposed in the caldera wall were likely quenched in eruption columns that remained below sea level. This pumice ingested seawater and sank as gases filling its vesicles cooled, particularly as steam in its vesicles condensed to liquid water. Some eruption columns may have broken through the sea surface and entered the air, especially during vigorous phases of the caldera-forming eruption. These pumices had the opportunity to ingest air as they cooled, becoming floaters as they fell back to the sea; these could have been carried distally on the sea surface by the combined effects of ocean currents and wind. The age of the caldera is unknown, but it may be as young as several thousand years. Its magmatic system at depth retains sufficient heat to sustain an actively growing intracaldera Kuroko- type polymetallic sulfide deposit, rich in gold and silver and topped by chimneys emitting fluids as hot as 278 °C. Sufficient time has elapsed, however, for a 250-m-high postcaldera dome to grow on the caldera floor and for the caldera rim to be deeply scalloped by slumping.

Subaqueous Pyroclastic Flows in the Ohanapecosh Formation, Washington

Subaqueous pyroclastic flows form almost one-half of the 10,000-foot Ohanapecosh Formation (Eocene and Oligocene?) in the eastern part of Mount Rainier National Park, Washington. Most of these flows probably originated by the sloughing of debris from the flanks of active underwater volcanoes during and after pyroclastic eruptions. Some of the pyroclastic flows caused directly by underwater eruptions may not have been completely quenched and could have traveled as steam-inflated slurries of pyroclastic debris and water. Most of them, however, were probably cold or only slightly warm as they flowed into deeper water. The Ohanapecosh subaqueous pyroclastic flows are extensive, nonwelded deposits of lapilli-tuff or fine tuff-breccia ranging in thickness from 10 to more than 200 feet. They are interbedded with thinner and generally finer turbidity-current and ash-fall deposits formed by smaller and more water-rich slumps of pyroclastic debris from the underwater volcanoes and by ash falls that rained into the water. The three main types of flows are thought to be related to three different kinds of volcanic activity. The most common flows—those containing a variety of lithic fragments and variable amounts of pumice—were probably produced by underwater phreatic eruptions. The flows rich in pumice and glass shards were probably caused by underwater eruptions of rapidly vesiculating magma which, on land, would have produced hot ash flows or ash falls. The least common flows—those containing only one or two kinds of lithic fragments—were probably derived from fairly homogeneous bodies, such as domes, spines, and lava flows that were erupted into water and fragmented by steam-blast explosions. Remnants of the underwater Ohanapecosh volcanoes consist chiefly of coarse tuff-breccia piled around filled volcanic vents. Unbrecciated lava flows are subordinate and were probably deposited on land when the volcanoes succeeded in growing into islands.

Caldera-forming processes and the origin of submarine volcanogenic massive sulfide deposits

Certain volcanogenic massive sulfide (VMS) ore deposits form in submarine calderas. This association is well known, but the link between caldera formation and the origins of the deposits remains poorly understood. Here we show that the size and location of a VMS deposit within a submarine caldera may be determined by how and when the caldera formed. These spatial-temporal conditions control development of the hydrothermal system associated with the VMS deposit. We propose that caldera opening along outward-dipping faults promotes magma degassing, seawater influx, and high-temperature leaching, resulting in a metal-rich hydrothermal fluid. These outward-dipping faults are considered to provide critical pathways for ore-forming fluids responsible for some caldera-hosted VMS deposits and may also be fundamentally important for the formation of many other caldera-hosted ore deposit types.

Recognition and Significance of Pumice in Marine Pyroclastic Rocks

Pumice is abundant in many ancient sequences of marine pyroclastic rocks and is regarded as important evidence that contemporaneous, or nearly contemporaneous, volcanic activity was the source of at least some of the fragmental debris. The pumice in many such sequences of rocks, however, is easily overlooked, chiefly because most marine pyroclastic rocks have been altered or metamorphosed to varying degrees, masking or obliterating the delicate cellular structures of the original pumiceous material. With care, however, and with the knowledge that pumice-rich rocks commonly occur toward the top of thick, vertically graded beds of lapilli tuff and tuff breccia, the elusive pumiceous fraction of most sequences of rocks can generally be recognized. Hand-lens examination of wetted specimens in the field will usually reveal the wispy and ragged outlines of some of the pumice that is present. More detail can be seen in thick sections and in conventional thin sections of pumiceous rocks. In general, the pumice in more altered and metamorphosed rocks can be seen by careful examination of hand specimens or thick sections; the pumice in relatively unaltered rocks can best be seen in thin section. Examples of pumice-rich rocks from the Precambrian of Arizona, the Cretaceous of Puerto Rico, and the Tertiary of Japan are described and illustrated.

Strain in metamorphosed volcaniclastic rocks and its bearing on the evolution of orogenic belts

Orogenic belts commonly contain thick deposits of volcaniclastic and associated rock types that bear abundant strain markers. Detailed mapping of a deformed assemblage of volcaniclastic rocks in a well-exposed part of the Ritter Range (central Sierra Nevada, California) has yielded five types of strain markers: lithic lapilli, accretionary lapilli, tuff-breccia fragments, reduction spots, and “ash-flow ellipsoids.” The axial ratio and angle between the long axis of the markers to a reference line were measured in a large number of specimens, and the strain was determined using the shape factor grid method and data on initial fabrics. Strain in these rocks is markedly heterogeneous over short distances owing mostly to the heterogeneous nature of the stratigraphic section. The mean strain ellipsoid ( X > Y > Z ) for the area has the following values: a 62 percent increase in X , a 15 percent increase in Y , and a decrease of 46 percent in Z (shortening normal to the slaty cleavage). The mean strain magnitude (¯e s ) for the area is ¯e s =0.77, considerably below that of the mean slate (¯e s =1.44); the symmetry of the deformation is principally one of flattening, however, and has a Lode9s value (ν = 0.40) close to that of the mean slate (ν = 0.43). Results indicate that thick stratigraphic sections can be profoundly affected by internal strain: for the volcanogenic assemblage in the Ritter Range, calculations made show that tectonic deformation has thinned the part of the pendant studied by more than 50 percent, from a thickness of 9.0 to 4.3 km. It appears likely that significant portions of the central Sierra Nevada country rock may have undergone comparable thinning of section. A plot of mean strain ellipsoids from a number of orogenic belts ranging in age from Archean to Mesozoic define two mean deformation paths, one clearly a flattening deformation in which extensions of the X and Y axes of the main strain ellipsoids have a ratio of (% X /% Y ) ≈ 4, while the other path lies close to a plane strain ( Y ≈ 1). The elongate shape characteristic of most orogenic belts is probably the most important factor controlling the extension ratios of thick stratigraphic sections during deformation, and hence their mean deformation paths. The fact that mean strain ellipsoids from orogenic belts spanning a considerable period of geologic time (2.5 b.y.) show closely comparable symmetries, suggests that present-day deformation is likely to provide realistic models for interpreting paleostrain found in ancient orogenic belts.

Middle Cretaceous ash-flow tuff and caldera-collapse deposit in the Minarets Caldera, east-central Sierra Nevada, California

A 2.3-km section of ash-flow tuff and associated caldera-collapse deposit, representing the extrusive facies of part of the Sierra Nevada batholith, is totally exposed from its floor to its top in the Minarets Caldera, east-central Sierra Nevada. Rapid burial and subsequent hornblende hornfels facies metamorphism resulted in remarkable preservation of primary textures and structures, despite the development of cleavage domains in parts of the caldera fill; late Tertiary uplift and Quaternary erosion have produced a rugged terrain where every meter of section is available for study. Large-scale caldera-filling eruption of ash-flow tuff was interrupted by emplacement of a wedge-shaped mass of caldera-collapse deposit as much as 2 km thick, whose volume exceeded 70 km 3 . Individual clasts in the caldera-collapse deposit range to as much as 1.8 km across and include a wide variety of andesitic to rhyoliic lavas and related volcaniclastic rocks, remnants of a precaldera volcanic field that was probably much more extensive than the caldera itself. The caldera-fill sequence rests with angular unconformity on a rugged surface eroded into older volcanic rocks; the sequence is capped by bedded volcaniclastic rocks, including delicately laminated tuffs of probable caldera lake origin. The total aerial extent of the Minarets Caldera is not known, but the area studied, plus scattered pendants of ash-flow tuff and associated volcaniclastic rocks to the west, defines a 30- x 22-km elliptical area that may approximate its original shape. The caldera fill is invaded by a body of quartz monzonite porphyry, locally miarolitic,that was probably emplaced during an episode of caldera resurgence.

Paleomagnetic evidence for hot pyroclastic debris flow in the shallow submarine Shirahama Group (Upper Miocene-Pliocene), Japan

The cooling history of coarse andesite blocks erupted and deposited proximally in the shallow marine environment has been investigated by paleomagnetic studies and heat conductivity calculations. Six to 13 samples were cored and measured from six essential andesite blocks 1–5 m in diameter. Thermal demagnetization techniques were used to detect the directional change of remanence through demagnetization steps, each representing different blocking temperatures (Tb). Each sample has three components of remanent magnetization (C1, C2, and C3). C1 is characterized by a Tb of 250°C and an orientation coincident with the present geomagnetic field. C2 has a Tb of 250°–500°C and shows a consistent direction with reversed polarity. C3 has a Tb of 500°–600°C, and its orientation is nearly consistent from samples collected from nearby parts of the same clast and are random from samples from different clasts. On the basis of the remanence direction of each component, the differing Tb, and the petrography of the samples, C1 is interpreted as secondary viscous remanence, whereas C2 and C3 are the partial thermoremanent magnetizations obtained during and immediately after emplacement. C3 was probably acquired while the blocks were rolling or sliding downslope, and C2 formed during continued cooling of the blocks after they came to rest. Heat conductivity calculations suggest that the blocks were emplaced less than an hour after being erupted. If the eruptive vent was only 2–3 km distant, as suggested by field relations, a downslope transport rate of less than 36 km/h is implied.