PHYSICAL VOLCANOLOGY

The volcanic units encountered in Hole 1137A include crystal-lithic volcanic siltstone and sandstone (basement Unit 5), lithic volcanic conglomerate (basement Unit 6), crystal-vitric tuff (basement Unit 9) (Table T6), and seven lava flows (Units 1-4, 7, 8, and 10) (Table T7; Fig. F4) (see "Lithostratigraphy"). In addition, lithologic Unit I contains disseminated volcanic detritus of mixed origin in foraminifer bearing diatom ooze (Unit 1).

Unit I: Foraminifer-Bearing Diatom Ooze (interval 183-1137A-1R-1, 0 cm, through 1R-CC)

The pelagic sediment in Unit I (Core 183-1137A-1R) contains 2% disseminated silt- to sand-sized crystal-lithic volcanic fragments, glass shards, and pumice lapilli. The shards include palagonitized basaltic glass and unaltered felsic glass. The abundance of volcanic detritus decreases downward through the unit, but sediments in Core 183-1137A-1R were highly disturbed by drilling, so the glass shards may have originated from one or more discrete ash layers. Shards show an array of morphologies; commonly the felsic shards show cuspate fragmental bubble-wall shapes or are formed from disrupted tube pumice, whereas basaltic shards are more blocky and equant. The felsic glass is translucent and isotropic; mafic glass ranges from pale golden brown or dark brown to dark red. Some basaltic shards have a green rim, a few microns wide, of a more crystalline phase, probably iron-rich smectite (nontronite) after palagonite. No microcrystallites or vesicles were observed; however, shards are small (<0.3 mm). Incorporated fine to coarse sand-sized lithic, mostly basaltic, clasts are angular to subrounded.

Basement Units

Depth in core for each of the basement units is summarized in Table T7. However, because ODP curation procedures assume that the top of recovered core comes from the top of the cored interval, the curated depths and thicknesses are often poor recorders of the actual downhole positions (see "Introduction" in the "Explanatory Notes" chapter). The measurements of the depths of important boundaries and flow thicknesses from the logging data (Table T8) are discussed in the interpretive part of this section (see "Recovery of Basement Units"). The unit thicknesses and recovery are detailed in Table T9.

We divided the basaltic basement into units that generally consist of a single lava flow. The unit and subunit boundaries were chosen based on changes in color, structure, grain size, and mineral occurrence and abundance. Because of the interpretation inherent in choosing these boundaries, the description of the specific criteria used to separate each unit and subunit are discussed in the interpretive part of this section (see "Basement Units").

Unit 1 (interval 183-1137A- 24R-1, 71 cm, through 25R-3, 100 cm)

The first basement unit consists of massive plagioclase-phyric basaltic lava. However, the first 5-cm-long piece of lava is a well-cemented breccia containing 0.5-2 cm of angular fragments of vesicular lava. The recovered lava below this piece is massive with predominantly 2%-5% vesicularity. A few, thin, undulating subhorizontal vesicular regions are located at Section 183-1137A-24R-1, 80 cm, through 25R-1, 40 cm. A wider and less distinct vesicular zone is at interval 183-1137A-24R-2, 6-10 cm. Vesicularity increases systematically in Section 183-1137A-25R-2, reaching 15% at 97.5-99.5 cm.

Subunits 2A and 2B (interval 183-1137A-25R-3, 100 cm, through 27R-1, 0 cm)

Unit 2 consists of an aphyric basalt with a brecciated upper part and a massive interior. The four loose pieces of well-cemented angular breccia at the top of Unit 2 differ in character from the bulk of the brecciated top of Unit 2. Immediately below these pieces is a vesicular lobe that extends down to Section 183-1137A-25R-4, 35 cm. The vesicles are elongate in the margin and more spherical toward the center and make up ~30% of the lobe (Fig. F12). This lobe is wider than the core (6 cm) at the top and only ~4 cm wide near the base. Under this lobe, the breccia is made up of 3- to 20-cm-diameter clasts that contain 10%-40% rounded to subrounded vesicles (Fig. F13). The clasts are variously oxidized, ranging from dusky yellowish brown (10YR2/2) to reddish brown (2.5YR4/3). The larger clasts are subrounded in shape, but their margins are commonly broken into 3- to 10-mm angular fragments that can be fit back to the larger clast. A few clasts envelop other clasts. Clasts >1 cm make up 80-90 vol% of the breccia, and fines <1 mm in size are confined to the lowermost 10 cm of the breccia and make up only about 2%-3% of the volume. However, large (>>1 cm) voids between clasts make up 8% of the breccia. These voids are notably free of sediments and are filled by a variety of secondary minerals (see "Alteration and Weathering"). The breccia gradationally becomes more welded with depth and reaches a coherent lava at approximately interval 183-1137A-25R-5, 27 cm.

The upper part of the coherent lava remains quite vesicular (5%-30%) but the vesicles appear in 5- to 30-cm-scale domains with variable vesicularity, vesicle size, and vesicle shape. The core is broken into similar sized pieces, and most pieces contain a single style of vesicularity. This obscures the transitions between these vesicle domains. The lava becomes relatively vesicle poor (0.01%-3%) beginning at Section 183-1137A-25R-6, 79 cm. A 5-cm, flat megavesicle marks this point (Fig. F14). Only one undulating subhorizontal vesicular region is present at interval 183-1137A-25R-7, 90.5-91.5 cm. Wispy blebs of vesicular, glassy mesostasis are common through most of the more vesicle poor part of Unit 2. These blebs are generally 0.1-0.3 mm wide and 1-3 mm long, though in some sections they are equant spots and in other areas are 10-20 mm long. There is no significant increase in vesicularity at the base of the recovered portion of Unit 2.

Subunits 3A and 3B (interval 183-1137A-27R-1, 0 cm, through 29R-2, 64 cm)

Unit 3 is a coherent plagioclase-phyric basalt with a more vesicular part overlying a more massive section, and the vesicularity as a function of depth is shown in Fig. F15. The upper part of Unit 3 was recovered as discontinuous pieces that contain large (2-3 mm average diameter) elongated and irregular subangular vesicles. By Section 183-1137A-27R-1, 78 cm, the vesicles are consistently very large (some >5 cm in length) and elongate in a subhorizontal direction. Several pieces show brecciation in the form of vertical compression of these large vesicles (Fig. F16). Wispy blebs of mesostasis and near-spherical megavesicles replace the elongated vesicles at Section 183-1137A-27R-3, 77 cm, where vesicularity drops from >20% to 10%-15%. The mesostasis blebs and microvesicle trains persist with variable prominence until Section 183-1137A-28R-3, 50 cm. In some areas these wisps define a foliation/fabric inclined at a high angle across the core. Recovery improved in Section 183-1137A-27R-4 where vesicularity dropped to 10%. Subhorizontal vesicular regions are quite common between Sections 183-1137A-27R-4, 50 cm, and 28R-2, 4.3 cm. Vesicularity begins to increase again near the base of Unit 3 and reaches 25% in interval 183-1137A-29R-2, 48-53 cm. Although vesicle size initially increases toward the base, it becomes smaller again in the lowermost centimeters.

An enigmatic piece of lava is found in Section 183-1137A-29R-2 (Pieces 3 and 4) (Fig. F17). Because the clast has the same unoxidized color and aphyric petrology as lava of Unit 3, whereas the upper portion of Unit 4 is highly oxidized and is moderately plagioclase phyric, we include this piece of lava with Unit 3. This piece of lava is separated from the rest of Unit 3 by a 2- to 5-cm void filled with a variety of sediments and reworked lava fragments. The interior lava surface of the cavity has an intricately convoluted margin including 0.2-mm-wide, 2-mm-long protuberances. The base of the piece has a fluidal shape with a 1- to 2-cm wavelength, small stretched elongate vesicles parallel to the lower margin, and a fine grained matrix.

Unit 4 (interval 183-1137A-29R-2, 64 cm, through 33R-1, 122 cm)

Unit 4 is dominated by moderately plagioclase-phyric coherent basalt of variable vesicularity. Figure F18 plots vesicularity as a function of depth within Unit 4. However, the uppermost 8 cm of Unit 4 consists of very fine-grained, oxidized, laminated, volcanic silt that has been subsequently brecciated into 1- to 20-mm subangular clasts. The clasts in the uppermost 2 cm are flattened vertically. The top of the lava flow is a smooth pahoehoe surface at Section 183-1137A-29R-2, 72 cm, with a 2-mm-wide margin of black fine-grained rock with 10 vol% very small (<<1 mm) vesicles (Fig. F19). The remainder of the upper part of the flow is pervasively oxidized to a dusky red (10R3/4). Farther down, the lava is gray to dark gray (N4-N5), but in vesicular areas it is oxidized to a dark reddish gray or reddish black (10R3/1-10R2.5/1).

Unit 4 has 5 distinct vesicular zones (Fig. F18). The uppermost of these extends from the vesicular top down to Section 183-1137A-31R-4, 52 cm. Subhorizontal, subplanar vesicular domains appear in the denser lava at Section 183-1137A-31R-4, 105 cm. The second and third vesicular zones are close together, starting at Section 183-1137A-31R-4, 124 cm, and returning to dense lava with subhorizontal, vesicular domains at Section 183-1137A-32R-2, 0-17 cm. The fourth vesicular zone is between Sections 183-1137A-32R-4, 27 cm, and 32R-7, 17 cm. Below this the lava is again dense, and a subvertical cylindrical vesicular domain with finer-grained groundmass extends from interval 183-1137A-33R-1, 20-25 cm. A final gradual increase in vesicularity begins at Section 183-1137A-33R-1, 82 cm, reaching 20% at interval 183-1137A-33R-1, 113-120 cm. Vesicle shapes are rounded throughout Unit 4, except in Section 183-1137A-32R-2. In general, vesicle size increases in the zones of increased vesicularity. However, in the bottom 2.5 cm of Unit 4, the vesicles are very small (<<1 mm) despite making up 10 vol% of the lava.

An exquisite basal contact of the lava over sediments is preserved (Fig. F20). The base of the flow shows 1- to 5-mm amplitude, 0.5- to 2-cm wavelength undulations.

Unit 5 (interval 183-1137A-33R-1, 122 cm, through 33R-4, 121 cm)

Unit 5 is composed of interbedded dark greenish gray crystal-lithic volcanic siltstone and light gray crystal-lithic volcanic sandstone. The top of the unit (interval 183-1137A-33R-1, 122-124 cm) is black and has been baked during emplacement of the overlying lava flow. Total carbon (TC) analysis of this interval was low (0.9% TC) (see "Organic and Inorganic Geochemistry"), indicating that soil had not developed on the silty sand prior to baking. Adjacent to the sediments, the basalt has a glassy chilled margin that is pervasively altered to iron-oxide (hematite) and oxyhydroxide (goethite)-stained clay minerals, probably iron-rich smectite nontronite and saponite (Fig. F21). The black color of the sediment is attributed to baking and dehydration of the clay mineral component (>10% of total matrix) in the siltstone (Fig. F21).

Siltstone and sandstone beds range from a few millimeters to tens of centimeters in thickness. Many beds are normally graded. Sandstone intervals commonly have sharp irregular bases (e.g., interval 183-1137A-33R-3, 50-53 cm) (Fig. F22) and gradational tops, but some have sharp irregular tops. Some siltstone and sandstone beds have planar lamination and low-angle cross-stratification (Figs. F8A, F8B) (see "Lithostratigraphy"). Rare burrows penetrate the sediments. Thin, bladed flakes (<5%), probably organic material, are scattered through the siltstone and sandstone, and in places form more concentrated intervals. No marine fossils were observed. Small-scale faulting offsets bedding in some sections.

The siltstone and sandstone contain angular to subangular crystal-lithic volcanic grains, mostly basaltic and felsic volcanic fragments (Fig. F22). Some (<2%) reworked laminated lithic clasts (siltstone) are incorporated. Abundant crystals in the sandstone are dominated by simply twinned feldspar (sanidine) with subordinate plagioclase (10%-15% total feldspar). There is minor quartz (~5%), and there are isolated crystals of garnet and opaque minerals (<5%). Glass shards were not observed in these sediments; however, there is a significant component (5%-10%) of secondary grayish green smectite nontronite or saponite.

Unit 6 (interval 183-1137A-33R-4 through 37R-1, 0 cm)

The conglomerate is dominated by well-rounded volcanic clasts in the granule, pebble, and cobble size range within a poorly to moderately well-sorted matrix of crystal-lithic volcanic clasts of fine- to coarse-sand size (see "Lithostratigraphy"). Matrix minerals include altered clinopyroxene and, possibly, olivine and titanomagnetite. The matrix is variably altered and cemented with green clay minerals (nontronite?), calcite, and silica. Some veins of calcite are present. Clasts have a range of volcanic lithologies dominated by highly plagioclase-phyric basalt (>80%), with lesser alkali-feldspar-plagioclase-clinopyroxene-phyric massive trachyte and alkali-feldspar-plagioclase-phyric flow-banded trachyte (10%-15%). There are scattered pebbles (<5%) of other lithologies, including granitoid and gneiss (see "Igneous Petrology"). Clasts are commonly oxidized along their margins with oxidation rims as wide as 1 cm (Fig. F9A) (see "Lithostratigraphy").

A point count of clasts in the lithic volcanic conglomerate allows quantification of the characteristics of Unit 6 (Fig. F23). The proportion of matrix in each section is generally <25%, but discrete intervals are granular and preserve fewer large clasts (Fig. 23A). In Section 183-1137A-35R-1, a discrete well-sorted medium sand interval is partially preserved in Pieces 17 and 18 (98-118 cm), which is considered a separate sedimentary event and not included in the granulometry (Fig. F23). A poorly sorted, coarse sand to granular interval is preserved in Section 183-1137A-35R-1 from 131 to 136 cm, and a moderately well-sorted medium sand between 58 and 65 cm in Section 183-1137A-35R-2. These sand intervals can be broadly related to planar fine-grained intervals observed in FMS log data (see "Downhole Measurements").

The average clast size in the lithic volcanic conglomerate ranges from 1.5 cm × 2.0 cm to 3.0 cm × 5.0 cm and is relatively constant in the pebble size range throughout the conglomerate. There are slightly larger average clast sizes in the upper half of Unit 6 (Fig. F23B). The maximum clast size ranges from 3.0 cm × 5.5 cm to 9.0 cm × 16.0 cm. The distribution of maximum clast size reflects the presence of two broad intervals in the conglomerate with reverse graded cobbles at the base and more normal graded cobble to pebble distribution above (Fig. F23C). One of these intervals lies below the >20-cm-thick medium sand interval in Section 183-1137A-35R-1 (98-118 cm) and the other lies above it. Sorting is relatively poor with a range in clast size from granule to cobble throughout most of the conglomerate. The calculated standard deviation for the clast size data provides a good indication of the degree of sorting (Fig. F23D). The standard deviations range from 1.4 to 3.2 cm, which is significant compared to the average clast size for the conglomerate (2.0 cm × 3.2 cm).

The conglomerate is an almost uniformly closed framework, or clast supported (Fig. F9B) (see "Lithostratigraphy") with very few open-framework, or matrix-supported, intervals. Open-framework intervals are present in the more matrix-rich parts of Sections 183-1137A-34R-1, 34R-2, 35R-1, and 36R-1 (Fig. F23A). Imbrication of cobbles is apparent in some parts of the FMS log (Fig. F94) (see "Downhole Measurements") and is present in Section 183-1137A-34R-3, from 57 to 85 cm (Fig. F9C) (see "Lithostratigraphy"). The clasts are moderately elongate, commonly subspheroidal to oblate, and subrounded to rounded (Fig. F9C) (see "Lithostratigraphy").

Subunits 7A and 7B (interval 183-1137A-31R-1, 0 cm, through 39R-2, 64 cm)

Unit 7 is a plagioclase-phyric basalt with a brecciated section on top and a coherent portion below. The breccia extends down to Section 183-1137A-38R-1, 23 cm. No contact is preserved between the Unit 6 conglomerate and Unit 7 lava flow, but the uppermost 10 cm of the recovered breccia has smaller and more rounded clasts than the remainder of the breccia. The breccia is made up of three different types of clasts and has a sediment matrix (Figs. F24, F25). Overall, 50-60 vol% of the breccia is in the form of clasts >1 cm, 40%-30% is 1- to 10-mm fragments, and ~10% is <1-mm fine-grained material. This fine-grained portion includes large sediment filled voids that make up 2% of the breccia. The three types of clasts are (1) larger clasts with 1- to 3-cm vesicles, (2) larger dense clasts with small, irregular elongate vesicles, and (3) smaller fragments of both of these larger clasts. The larger clasts are up to 30 cm in size and subrounded in shape. The largest (i.e., most intact) vesicular clasts have rounded vesicles elongated parallel to the clast margins. Interestingly, the dense clasts are common only in the lowermost part of the breccia. The dense clasts also contain a few small fragments of oxidized lava within their interiors. The smaller fragments are angular and are concentrated in subplanar zones surrounding the dense clasts. Other small angular fragments show a jigsaw-fit pattern with the margins of larger vesicular clasts.

Starting in Section 183-1137A-38R-1, 23 cm, the lava is welded into a coherent rock. Patches of small irregular angular vesicles are in interval 183-1137A-38R-1, 69-103 cm. The remainder of the coherent lava is quite massive with only 0.5%-3% vesicles. Wispy mesostasis blebs are evident in Section 183-1137A-38R-3 through 39R-2, 43 cm (Fig. F26). Vesicularity is higher (5%-15%) in interval 183-1137A-39R-2, 55-62 cm, immediately above a glassy chill margin.

Subunits 8A and 8B (Core 183-1137A-39R-2, 64 cm, through 41R-1, 103 cm)

Unit 8 is predominantly a coherent plagioclase-phyric basalt with a vesicular portion overlying a massive interior. Included at the top of Unit 8 are four pieces of breccia whose origin is not immediately obvious (Fig. F27). However, the breccia pieces have nearly identical clast and vesicle size, number density, and morphology as the breccia at the top of Unit 7.

The lava immediately below this breccia is recovered as 2- to 5-cm-long pieces, with each piece consisting of coherent lava with no evidence for clast margins (Fig. F27). The vesicles in this coherent lava are large, rounded, elongated vesicles of a type not seen anywhere in Unit 7, but very similar to the vesicles in the upper part of Unit 3. Some of the largest vesicles show evidence for vertical compression identical to that seen in Unit 3. Vesicle sizes gradually increase with depth in Unit 8, with the average size increasing from 2.5 to 16 mm. Unit 8 continued to be recovered only as isolated 2- to 5-cm pieces until Section 183-1137A-40R-2, where vesicularity decreased below 15%. The underlying denser lava has a vesicularity of 0.1%-5%. Subhorizontal vesicle-rich regions with a distinct groundmass first appear at Section 183-1137A-40R-1, 31 cm, and are not found below Section 183-1137A-40R-3, 65 cm. A vertically elongate, 2- to 4-cm-diameter vesicle-rich zone with similar groundmass is present in interval 183-1137A-40R-3, 97-106 cm. Wispy mesostasis blebs are most common from Section 183-1137A-40R-4, 79 cm, through 40R-5, 111 cm, and near the base of the flow. Vesicularity at the base of the flow begins to increase from Section 183-1137A-41R-1, 22 cm, reaching a peak of 12% in the lowermost piece recovered. The lowermost 10 cm of recovered Unit 8 contains large, horizontally elongated vesicles of the same character as those at the top of the coherent lava. No glassy basal chill zone was recovered. However, several chips of black rock were recovered between the bottom of Unit 8 and the first recognizable piece of the crystal-vitric tuff of Unit 9.

Unit 9 (interval 183-1137A-41R-1, 103 cm, through 44R-4, 49 cm)

Unit 9 is a green, internally massive 14.8-m-thick interval of crystal-vitric tuff. The upper part of the tuff (interval 183-1137A-41R-1, 103-112 cm) is very fine grained and black with scattered broken feldspar crystals (<5%) and was probably baked during emplacement of the overlying lava flow (Fig. F10) (see "Lithostratigraphy"). Total carbon analysis on this interval was very low (0.15% TC), indicating that soil had not developed on the tuff before baking. The basalt has a glassy chilled base adjacent to the sediments. The black color of the uppermost tuff is probably caused by dehydration of the clay mineral component during baking.

The tuff consists of 30%-40% angular, euhedral, broken, or disrupted simply twinned sanidine crystals, up to 5 mm in diameter. The balance of the mineral component includes <5% amphibole, <5% plagioclase feldspar, 1%-3% quartz, and <3% opaques (see "Igneous Petrology"). The tuff contains 1%-2% subangular to rounded, granule- to pebble-sized lithic clasts. Clast lithologies include plagioclase-phyric basalt, aphyric basalt, trachyte, garnet gneiss, and granitoid (see "Igneous Petrology").

Cuspate and tricuspate, bubble-wall glass shards (40%-50%), now partially altered to green clay minerals (kaolinite, nontronite) and possibly micas (celadonite), form the matrix between crystals (Fig. F28). The tuff is more altered and oxidized toward the top, and correspondingly the glass shards are harder to resolve in thin section. Lower in Unit 9, many glass shards retain original morphologies and flattening is not apparent, but in more altered zones, the relict shards are slightly flattened and aligned. There is no evidence of welding in the tuff.

Subunit 10A and 10B (interval 183-1137A-44R-4, 49 cm, through 49R-3, 37 cm)

No clear contact between Units 9 and 10 was recovered. However, from Section 183-1137A-44R-4, 49 cm, through 45R-2, 100 cm, there is a breccia consisting of basaltic clasts in a matrix of pale green, feldspar-rich, silty sand. The silty sand differs from the overlying Unit 9 crystal vitric tuff in its grain size and mineralogy, particularly the greater abundance of biotite in the silty sand. Downhole FMS log data indicate that there is an ~1-m-thick (logged interval 360-361 mbsf) (see "Downhole Measurements") bedded interval at the base of the crystal vitric tuff (Unit 9).

Within the breccia, basalt clasts are 0.5 to 35 cm in diameter. Smaller clasts are angular to subangular, and larger clasts are subrounded to rounded with 10% to 40% vesicles. One clast has a denser margin and more vesicular interior. Vesicle shapes are variable, and one clast has a margin with vesicles elongated parallel to the margin and more spherical vesicles toward the center. More intact clasts have irregular fluidal margins, but fragmented and jigsaw-fit margins are common (Figs. F29, F30). Welded margins are rare in the upper part of the breccia but common lower down. Green silty sand fills most of the space between the clasts and fragmented clast margins. Domains of sediment make up ~4% of the breccia. Disaggregated, 3- to 20-mm, angular, fragments with some curviplanar faces appear suspended in the sediments in interval 183-1137A-45R-2, 10-25 cm (Fig. F31), but there may be some clast-clast contact between these lava fragments in three dimensions. A zone of red (10R 3/2-10R 3/4) oxidized, jigsaw-fit, dense lava fragments, 0.5-2 cm in size, is present in interval 183-1137A-45R-2, 20-67 cm (Fig. F32).

Several thin sections show the contact relationships between the basalt clasts and the sediment (Sample 183-1137A-45R-1, 80-84 cm; Sample 183-1137A-45R-2, 7-10 cm; Sample 183-1137A-45R-2, 19-21 cm) (see "Igneous Petrology"). In Sample 183-1137A-45R-1, 80-84 cm, sediment grain size ranges from silt to fine sand, and sorting is generally poor. Some discrete domains of coarser or finer grained material appear to have flowed around clasts (Fig. F33). There is no evidence of grading. The sandy silt is either internally massive or it has fine silt distributed around and between clasts with coarser sand concentrated in stringers or domains of aligned particles farther from clasts (Fig. F33). Basalt clasts are glassy and aphanitic. They have angular polygonal shapes with some curviplanar surfaces, and there is possible evidence of breakage across a few crystals (Fig. F34) and through vesicles. However, breakage seems to preferentially take place through glassy parts of the basalt.

The breccia becomes more welded in Section 183-1137A-45R-2, 70 cm, and the lava becomes coherent from about Section 183-1137A-45R-2, 100 cm. However, a dense margin is present in Section 183-1137A-45R-3, 47 cm, around a >3-cm-elongate irregular shaped void, which extends out of the core. Irregular vesicular clumps are in the coherent lava down to Section 183-1137A-45R-3, 120 cm. Thin, wispy subhorizontal vesicle and glassy mesostasis-rich regions first appear at Section 183-1137A-45R-4, 36.5 cm. These grade upward in size and are associated with megavesicles through Section 183-1137A-46R-1, 61 cm. The last of these subhorizontal vesicle-rich regions is closely followed by a vertically elongate pod of similar material extending down to Section 183-1137A-46R-1, 83 cm, where it exits the core. Other than these vesicle-rich domains, the lava is <2% vesicles through Core 183-1137A-46R, and the hole ends with massive lava in Section 183-1137A-46R-3, 36 cm.

Lithologic Units

Unit I: Foraminifer-Bearing Diatom Ooze

Disseminated volcanic components in this unit are a mixture of mafic and felsic volcanic ash, which were partially homogenized during drilling. The low concentration of volcanic components is consistent with these materials being deposited by settling through the water column, possibly reworked on the seafloor, and disturbed during coring.

Basement Units

Recovery of Basement Units

Downhole logging data allowed in situ determination of the depths of lithologic boundaries and thicknesses of flow units. The curated depths and depths determined from the logging data are compared in Table T8. In the following sections, all depths and thicknesses are derived from the logging data, unless otherwise stated.

Based on the depths derived from logging data (Table T8), and the amount of rock recovered from each subunit, it is possible to calculate the actual recovery of each unit and subunit (Table T9). There is consistently excellent recovery of the massive portions of lava flows (92%-100%) and significantly poorer recovery of breccias (62%-87%). The material that proved most difficult to retrieve was the rock with long, elongated vesicles in the upper portions of Units 3 and 8 (27%-49% recovery). These patterns of preferential recovery of massive lava are only apparent in light of the true unit thicknesses derived from logging data.

Unit 1

The recovered rocks from Unit 1 are from the massive interior and basal part of a single thick lava flow. The subhorizontal vesicular regions with more glassy mesostasis are interpreted to be thin, horizontal vesicle sheets (HVSs) but the 5- to 6-cm diameter of the core is inadequate to confirm a sheet-like morphology. However, these features have at least a 20:1 horizontal to vertical aspect ratio. The increase in vesicularity at the base of Section 183-1137A-25R-2 is interpreted to be a ~30-cm-thick lower vesicular crust. However, the actual chill margin at the base of the flow is missing. What happened to the top of the flow is equivocal. The logging data indicate that no coherent basalt exists above 227 mbsf. However, the zone with high gamma-ray intensity at 223-226 mbsf might be consistent with an extremely altered flowtop if the basalt is highly altered to clay minerals. With both the flow top and bottom missing, it is difficult to interpret the morphology of the flow. However, it is at least 7 m thick, and, if the high gamma region is the flow top, it might have had a total thickness of about 10 m. The internal features are consistent with, but not diagnostic of, an inflated pahoehoe flow.

Unit 1/Unit 2 Boundary

Although the basal glassy chill from Unit 1 was not recovered, the four loose, slightly brecciated pieces at the top of Unit 2 appear to be from the upper margin of the lobe at the top of Unit 2. The downward necking of this lobe suggests that it is intruding the breccia from above. Thus, it is possible that this lobe is part of Unit 1 or represents a completely different lava flow. However, the most likely scenario is that this lobe is part of the Unit 2 flow-top breccia. The oxidation of the upper part of Unit 2 is consistent with exposure in a subaerial environment, and probably there was a significant, but geologically very short, time interval between the emplacement of these two units.

Subunit 2A. The breccia at the top of Unit 2 does not fit any established category of basaltic lava flow. It distinctly differs from aa or slab pahoehoe. Instead, the breccia appears to consist of variously fragmented 15- to 30-cm-diameter pahoehoe lobes. The brecciation must have taken place during the emplacement of the flow for clasts to be able to entrain fragments of earlier clasts. The elongated, ellipsoidal vesicles in the margins of many clasts indicate that the flow was not in the extremely non-Newtonian flow regime of aa breccia formation, and the fact that the clasts are surrounded by such margins indicates that they were not broken from within by surges, as in the case of slab pahoehoe. It is likely that the earlier lobes were fragmented by a combination of the intrusion of more lobes from below and shearing of the mangled flow top. The range of oxidation of the different parts of the breccia supports the idea that the breccia formed over a significant time interval with the last clasts to form being the least oxidized.

Subunit 2A/Subunit 2B Boundary. The boundary between the breccia and the massive interior of Unit 2 is marked by a gradational increase in welding over a distance of tens of centimeters. As such, the precise location of this boundary is arbitrary. The vesicular patches within the interior of the flow (interpreted to be remelted vesicular breccia clasts) show that the breccia was being entrained into the interior of the flow.

Subunit 2B. The massive interior of Unit 2 is less vesicular than that of Unit 1, suggesting that it is more thoroughly degassed. The dearth of HVSs and the presence of finer mesostasis wisps may indicate that the segregated material had difficulty separating from the rest of the lava. This can occur if the flow did not stop moving (and mixing) until it was too crystalline and viscous for the differentiated material to move far before complete crystallization. The lack of a basal vesicular zone or breccia probably reflects the poor (or nonexistent) recovery of the contact zone between Units 2 and 3.

Unit 2/Unit 3 Boundary

Although the boundary between Units 2 and 3 was not recovered, the lowermost pieces of Unit 2 include a piece with vesicles and oxidation similar to Unit 3. This is probably a result of drilling disturbance.

Subunit 3A. The poor recovery of the upper part of Unit 3 is probably directly related to the fact that the individual vesicles in the top of Unit 3 are similar in size to the core diameter. Furthermore, the lava is brecciated by collapse of many of the largest vesicles. This collapse must have happened after solidification of the lava and might plausibly have been caused by the weight of Unit 2 riding over the top of Unit 3. Although overall vesicularity and vesicle shapes in Unit 3 are similar to those in many inflated pahoehoe flows, their size is quite unusual. The formation of such large vesicles suggests an unusually high degree of bubble coalescence and, thus, an unusually long time in transit from the vent. The elongate but round shapes indicate moderately high shear of a relatively fluid lava.

Subunit 3A/Subunit 3B Boundary. The boundary between the vesicular and massive portions of Unit 3 was drawn where overall vesicularity dropped below 5%. This break is clear in the core and is easily discerned in the logging data. However, the boundary between the vesicular upper crust and the interior of the flow that cooled stagnantly is 1.5-2 m higher up, within what is designated Subunit 3A. This critical internal boundary within Unit 3 is located just above the first HVS. Overall vesicularity drops from ~25% to ~10% at this point and the vesicles switch from elongated shapes to near spherical megavesicles. These megavesicles have been interpreted to be related to coalescence of bubbles grown during crystallization (i.e., the same population that produces HVSs). Using the Thordarson (1995) technique (see "Interpretation" in "Physical Volcanology" in the "Explanatory Notes" chapter), the estimated emplacement duration for Unit 3 is about a year. Using the Subunit 3A/Subunit 3B boundary as the upper crust, massive interior boundary would overestimate the emplacement duration.

Subunit 3B. The HVSs and megavesicles are generally confined to the upper part of Subunit 3B (and the lowermost part of Subunit 3A), as would be expected in the massive interior of an inflated pahoehoe lava flow. The wispy mesostasis blebs defining a high angle foliation are not understood, though they have been observed in some large inflated pahoehoe flows. The distribution of vesicles at the base of Unit 3 is a classic example of a vesicular basal crust of an inflated pahoehoe flow, except for the presence of the sediment-filled cavity.

Unit 3/Unit 4 Boundary

The sediments inside the cavity in the bottom of Unit 3 clearly indicate the periodic influx of water-borne fine-grained sediments followed by settling in a quiescent environment. Some of the pulses of water were energetic enough to rip up and transport small clay flakes and the oxidized lava fragment with attached sediments. A lake or ephemeral stream provides such a depositional environment, and we speculate that a very similar environment could exist where water moves through an extremely permeable lava flow.

The presence of the volcanic silt horizon between the base of Unit 3 and the first lava in Unit 4 indicates a time interval between emplacement of the flows. The brecciation of this thin laminated volcanic silt horizon is probably a secondary alteration texture that formed after induration of the silt.

Unit 4

The morphology of the top of the Unit 4 lava indicates that it is a pahoehoe flow. The thickness and vesicularity of the chill margin are diagnostic of subaerial emplacement. The five vesicular zones are interpreted to be the upper vesicular crust, three horizontal vesicular zones (HVZs), and a lower vesicular crust. The uppermost vesicular zone is continuous to the top of the flow and is a typical example of the vesicular upper crust of an inflated pahoehoe flow. The base of the vesicular crust transitions to a dense interior including HVSs. The gradual increase and decrease in vesicularity below this is an excellent example of an HVZ, as described in "Interpretation" in "Physical Volcanology" in the "Explanatory Notes"  chapter. The near spherical bubbles, the fining of vesicle size both upward and downward from the peak in vesicularity, and the generally large size and moderate number density all identify this zone of vesicularity as an HVZ and distinguish it from HVSs and vesicular flow bases. The preferred interpretation for HVZs is that they represent renewed injection of bubble-laden lava into the molten interior of a recently stagnated inflated flow. The third vesicular zone also has a gradational change within the flow and is interpreted as a second HVZ. However, this HVZ exhibits somewhat less smooth changes, and poor recovery precludes definitive identification as an HVZ as opposed to a flow base and flow top. The return to massive lava with HVSs suggests that after this second injection of bubble laden lava, the flow again stagnated and the separation of differentiated material restarted. The fourth vesicular zone is somewhat more difficult to interpret. The size of the vesicles is smaller and the distribution is not that different from that of a contact between a vesicular base and a vesicular flow top. Furthermore, it is extremely rare to find an HVZ this deep within a flow. Most commonly, HVZs are confined to the upper half of the flow. However, the excellent recovery of Unit 4 suggests that the chill contact between two flows should be in the cores, if it existed. There is no glassy chill zone seen in either the core or the logging data but there is a minor decrease in groundmass crystal size (see the barrel sheet for Section 183-1137A-32R-4 in "Site 1137 Core Descriptions"). We suggest that this is an unusual HVZ relatively deep within a flow that received a third pulse of lava late in its cooling history. The last increase in vesicularity is a vesicular lower crust, including the basal chill zone. The smaller vesicles in the lower (once) glassy margin are the most common type found in the basal chill margins of large inflated pahoehoe flows. The observation that the sizes and shapes of vesicles in the lower chill are identical to those 27 m above in the upper chill zone support the idea that this is a single inflated pahoehoe flow that initially was a relatively slow moving (<<1 m/s), small (<<1 m thick) lobe. If this flow was indeed active during the freezing of 23-24 m of upper crust, the Thordarson (1995) technique would imply a flow that remained intermittently active for about a decade.

Unit 4/Unit 5 Boundary

Immediately below the basal chill margin of Unit 4, the top of the siltstone and sandstone of Unit 5 are black (N1) for 5 cm and grade to a grayish brown color over the following 5 cm. The dark color of the sediment near the contact probably results from the baking caused by the emplacement of Unit 4.

Unit 5

The fine grain size of the crystal-lithic volcanic siltstone and sandstone and the preserved sedimentary structures including lamination, normal grading, and low-angle cross-stratification are consistent with their deposition in a relatively quiet water environment with a limited amount of current activity. Together with the underlying conglomerate, the siltstone and sandstone form a sedimentary succession that was probably deposited in a braided river environment (see "Lithostratigraphy").

The siltstone and sandstone are dominantly composed of volcanic detritus generated by physical erosion and chemical weathering of volcanic rocks. The crystal content in the sandstone is high (40%-50%), and the abundance of feldspar, especially K-feldspar, supports the observation that more evolved lavas (trachytes and rhyolites) form a significant volume of the eruptives in this area (see "Igneous Petrology"). Lithic clasts within the sediment include basalt, trachyte, and possibly rhyolitic volcanic fragments. Although this sediment is well sorted, it is relatively immature (angular to subangular), reflecting its proximity to the source area. However, many clasts have subangular to subrounded shapes reflecting reworking in a fluvial environment.

We interpret these sediments to have accumulated in a quieter section of a braided river. The larger clasts of a channel succession are represented by the underlying volcanic conglomerate. Once this channel system has migrated laterally or flow no longer continues along the channel, a quiet water environment is established. Low-angle cross-stratification indicates that some directed flow continued to take place; however, the system had considerably lower energy than was required to emplace the conglomerate.

Unit 5/Unit 6 Boundary

The boundary between the siltstone and sandstone (Unit 5) and the conglomerate (Unit 6) is not well preserved, but it appears that the finer sediments are an onlapping facies in the same succession as the conglomerate.

Unit 6

The clasts in the lithic volcanic conglomerate are reworked from a volcanic hinterland, dominated by basaltic and trachytic lithologies. In addition to the presence of felsic volcanic rock, isolated pebbles of continental lithologies, including gneiss and granitoid pebbles, require that continental rocks must have been exposed near this site (see "Igneous Petrology").

Clasts are dominantly subspheroidal to oblate in shape. The high number of clast-clast contacts in the conglomerate, high degree of rounding of large clasts, and presence of imbrication in the sediments as well as lack of marine fossils, supports a fluvial environment, probably a braided stream, as the environment of deposition for this conglomerate (see "Lithostratigraphy").

At least two large-scale reverse- to normal-graded sequences were identified in the granulometric data for the conglomerate (Fig. F23C). These may reflect the lateral migration of a channel. Initially, pebble and cobble size coarsens as the axis of the channel migrates toward the drill site and coarse clast size decreases as it migrates past or away again. At Site 1137, this channel may have migrated at least twice. The two periods of channel migration are separated by an 80-cm-thick sandy interval deposited in a quieter period (interval 183-1137A-35R-1, 98-118 cm) (see Fig. F94) (see "Downhole Measurements"). Other intercalated sands may reflect periods of quieter or off-axis deposition during emplacement of the conglomerate.

Clasts are variably oxidized, reflecting differing intensities of chemical weathering before incorporation of eroded material into the conglomerate.

Unit 6/Unit 7 Boundary

Although deposition of the volcanic conglomerate (Unit 6) took place in a high energy environment, the underlying flow-top breccia (Subunit 7A) has not been severely eroded. However, the increased rounding and sorting of the top 10 cm of Subunit 7A is consistent with some reworking of the uppermost part of the breccia.

Subunit 7A. Given that a conglomerate lies over Unit 7, it would be reasonable to expect the breccia at the top of the unit to be largely composed of reworked material. However, the breccia on the top of Unit 7 is broadly similar to the breccia on Unit 2, and the evidence for welding near the base of the breccia suggests that Unit 7 formed a brecciated top during emplacement. The welding also argues against interaction with significant volumes of water or wet sediments during emplacement. Like Unit 2, this breccia is unlike any recognized category of basaltic lava flow and probably involved the break-up of intruding pahoehoe lobes. The lower number of large open voids relative to the breccia on Unit 2 and the rounding of the uppermost recovered clasts may be evidence for some reworking of these breccia by sedimentary processes. The dense clasts in the lower part of the breccia also are unlike the clasts in the Unit 2 breccia. These clasts/lobes are composed of lava similar to the interior of Unit 7, providing additional evidence that the breccia was being intruded by material from inside the massive part of the flow.

Subunit 7A/Subunit 7B Boundary. The increase in welding and the coherence of the lava is gradational over at least 1 m, and the precise location of this boundary is arbitrary.

Subunit 7B. The two vesicular patches within Subunit 7B are interpreted to be entrained breccia clasts. The lack of any HVSs and the relatively high abundance of wispy mesostasis blebs suggest that late-stage liquids were not able to segregate into large sheets or cylinders. This indicates that the lava continued to move even as it became largely crystalline. This idea is supported by the dense intrusions in the breccia in Subunit 7A. The increase in vesicularity at the base indicates that the massive part of the lava was cooled from the base while it still contained ~15% vesicles.

Unit 7/Unit 8 Boundary: Unit 7 Basal Breccia

The origin of the breccia at the base of Unit 8 is not immediately clear, but following examination of clast and vesicle morphology, we have interpreted it as a basal breccia for the lava flow in Unit 7. This interpretation suggests that Unit 8 had a coherent top.

Subunit 8A. The vesicular top of Unit 8 is almost identical to the vesicular top of Unit 3, including its poor recovery. The only puzzle is exactly how the extremely large elongated vesicles formed. It must have involved the coalescence of smaller bubbles and significant shear of a lava that could not return to a spherical shape before freezing. However, despite the evidence of shear of a relatively viscous lava, the very rounded, elliptical shapes of the vesicles clearly indicates that the lava was not close to the regime in which aa forms. The lowermost part of Subunit 8A is not interpreted to be part of the vesicular upper crust, as explained in the next section.

Subunit 8A/Subunit 8B Boundary. The change from vesicular to dense lava in Unit 8 is gradational and rather poorly recovered. However, the first appearance of HVSs, which coincides with the drop below 15% vesicularity, is interpreted as the boundary between the vesicular upper crust and the part of the flow that crystallized in a more stagnant fashion. Because of the poor recovery, it is not possible to locate this contact with precision, but it must be at least 30 cm above the Subunit 8A/8B boundary (where overall vesicularity drops below 10%). The logging data do not clearly locate this contact because it is not possible to distinguish the vesicles in HVSs from primary flow-top vesicles. It is noteworthy that this leaves a relatively thin (~2.5 m) upper vesicular crust on Unit 8. This suggests that Unit 8 could have been emplaced in ~6 weeks (see "Physical Volcanology" in the "Explanatory Notes" chapter), which is quite fast for a 10-m-thick inflated pahoehoe flow.

Subunit 8B. The upper part of the interior of Unit 8 is quite vesicular because of the concentration of HVSs and has been placed in Subunit 8A. The remaining, more massive part of Unit 8 contains a few more HVSs in the upper portions and what is interpreted to be a vesicle cylinder that passes through the core obliquely. The basal vesicular zone is very similar to other lower crusts, except that the vesicles are larger and more elongated than is typical of most pahoehoe flows. These vesicles have exactly the same form as the vesicles interpreted to be the top of the lava flow, which makes up the bulk of Unit 8.

Unit 8/Unit 9 Boundary

A dark colored, sanidine-bearing indurated material beneath Unit 8 is interpreted to be a baked horizon at the top of the crystal-vitric tuff (Unit 9).

Unit 9

The presence of broken bubble-wall (cuspate and tricuspate) shards and abundant embayed, broken, and disrupted crystals is consistent with genesis of the tuff in an explosive volcanic eruption. The high sanidine crystal content (40%-50%) and well-preserved delicate glass shards indicate that the sediments have not been transported long distances from their source. Incorporated lithic clasts are not genetically related to the tuff and are interpreted to be accidental clasts that were incorporated during transport.

The coarse grain size of both the enclosed pebbles and the primary sanidine crystals precludes deposition of this material by settling from an ash cloud. Transport in a pyroclastic flow is more likely. However, no internal stratification is preserved, and the absence of a basal breccia or fine flow top, the total lack of normal grading of lithics and crystals or reverse hydraulic grading of glassy material strongly suggests that this pyroclastic material has been reworked.

The even distribution of crystals and of pebbles throughout the tuff and the massive internal texture of the deposit provides evidence for plastic mass flow redeposition of these sediments to their present location. It is not apparent whether this redeposition took place in a subaerial or submarine environment, although it is likely that water helped mobilize the tuff. However, the upper part of the crystal-vitric tuff is discolored by red/orange oxidation. This staining of secondary minerals with iron oxides (hematite) and oxyhydroxides (goethite) suggests some subaerial weathering.

Several mechanisms may have concentrated crystals in this interval, including explosive eruption of crystal-rich magma, elutriation of fine particles during eruptive transport, and elutriation of fine particles during mass-flow redeposition of the crystal-rich pyroclastic sediment.

The crystal vitric tuff has a green clay-rich matrix, and, although internally massive, it has a subtle, subhorizontal alteration fabric. This fabric is related to the flattening of partially altered glass shards caused by load compaction during diagenesis.

Unit 9/Unit 10 Boundary

A contact between the crystal-vitric tuff (Unit 9) and the underlying basaltic breccia (Subunit 10A) was not recovered. However, the top of basalt (Unit 10) is mingled with a green, fine-grained altered silty sand. FMS logs show that this interval of bedded silty sand at the base of the crystal vitric tuff is ~1 m thick (see "Results").

Subunit 10A. The breccia at the top of Unit 10 may have formed in part by the interaction of hot magma with fluidized, silty sand and, in part, by disruption of the breccia by intrusion of basalt lobes. The sediment commonly has massive internal textures or preserves textures consistent with dynamic interaction with the brecciated clasts, and the brecciated clasts show irregular to angular morphologies. The textures are consistent with the quench fragmentation of basaltic magma in water saturated sediment resulting in peperite formation.

Penetrative brecciation of the glassy basalt observed in thin section is consistent with magma-sediment interaction at the time of quenching (Fig. F34). Jigsaw-fit textures indicate in situ brecciation of the basalt by quenching and infiltration of fractures by fluidized sediment (Figs. F29, F30, F34). The lower part of the breccia shows less interaction with sediment and more interaction between breccia clasts and crosscutting lobes. Chilled margins, lobes incorporating earlier clasts, and welding of lobe-lobe contacts indicate that the lobes were hot.

An alternative to Subunit 10A being a peperite is that it is a flow-top breccia like Subunits 2A and 7A that has been reworked by sedimentary processes. The in situ brecciation could be the result of cooling in air and secondary alteration. The sediments in the breccia could have been deposited after the flow was emplaced.

Lava Emplacement Styles

Two principal types of lava flow are represented at Site 1137: inflated pahoehoe flows and flows with a flow-top breccia composed of pahoehoe lobes. Flows within each basic type show a great deal of individuality. The thickest pahoehoe flow (27 m of Unit 4) suggests a single lobe being active for a decade, longer than has been previously suggested for any single pahoehoe lobe (Self et al., 1998). The thinnest pahoehoe flow (10.6 m of Unit 8) is estimated to have been emplaced in a matter of weeks. The lack of a breccia on top of Unit 8 means that even such relatively rapidly emplaced ponded flows moved as smooth surfaced pahoehoe and did not involve wholesale disruption of their flow tops.

The brecciated flow type seen at Site 1137 is similar to some flows in the Columbia River Basalt, but this flow type has never been described in detail (see "Physical Volcanology" in the "Explanatory Notes" chapter). The interior of these flows is noticeably less vesicular than that of the inflated pahoehoe flows, suggesting more efficient degassing during the brecciation process. There are few or no horizontal vesicle sheets or other signs of large-scale concentration of late stage differentiates. Factors that should hinder this type of late-stage segregation include relatively rapid cooling of the flow interior by entrained clasts and continued motion of the flow stirring the interior.

The presence of a basal breccia on Unit 7 means that brecciation we observe at Site 1137 was taking place at the flow front and is not a remobilized pahoehoe flow (see "Interpretation" in "Physical Volcanology" in the "Explanatory Notes" chapter). It is worth noting that the only characteristic separating Unit 7 from classic aa flows is the insufficient angularity of the clasts and vesicles.

Eruption Environment

The basement at Site 1137 formed in environments ranging from subaerial to fluvial. The environment in which Unit 1 was erupted is equivocal, but the internal characteristics of the lava are consistent with known subaerial pahoehoe flows emplaced on gentle slopes (1°). Unit 2 is subaerial, based on both the morphology of the breccia and its oxidation. Unit 3 has poor recovery of key portions but is again consistent with subaerial eruption. The sediments within the base of Unit 3 suggest periodic influx of silt-laden water, most easily explained by the lateral migration of ground water and deposition near the water table (see "Lithostratigraphy"). Unit 4 appears to be subaerial, based on its highly vesicular nature, the thin chill margin, and deep oxidation. However, the overlying laminated siltstone was probably deposited in quiet, shallow water. Units 5 and 6 indicate a fluvial setting. The lithic volcanic conglomerate (Unit 6) probably formed in a braided river environment, near a region of significant relief. The welding and morphology of the breccia on Unit 7 suggests eruption in a subaerial environment with subsequent infilling by water-lain sediments. The features of Unit 8 are consistent with subaerial emplacement on a gentle slope. Mass flow redeposition of the crystal vitric tuff (Unit 9) probably involved water, but mass flow redeposition in a subaerial environment is possible. The breccia at the top of Unit 10 has characteristics that suggest that lava may have intruded into wet sediments.