Deep submarine pyroclastic eruptions: theory and predicted landforms and deposits

Submarine pyroclastic eruptions at depths greater than a few hundred meters are generally considered to be rare or absent because the pressure of the overlying water column is sufficient to suppress juvenile gas exsolution so that magmatic disruption and pyroclastic activity do not occur. Consideration of detailed models of the ascent and eruption of magma in a range of sea floor environments shows, however, that significant pyroclastic activity can occur even at depths in excess of 3000 m. In order to document and illustrate the full range of submarine eruption styles, we model several possible scenarios for the ascent and eruption of magma feeding submarine eruptions: (1) no gas exsolution; (2) gas exsolution but no magma disruption; (3) gas exsolution, magma disruption, and hawaiian-style fountaining; (4) volatile content builds up in the magma reservoir leading to hawaiian eruptions resulting from foam collapse; (5) magma volatile content insufficient to cause fragmentation normally but low rise speed results in strombolian activity; and (6) volatile content builds up in the top of a dike leading to vulcanian eruptions. We also examine the role of bulk-interaction steam explosivity and contact-surface steam explosivity as processes contributing to volcaniclastic formation in these environments. We concur with most earlier workers that for magma compositions typical of spreading centers and their vicinities, the most likely circumstance is the quiet effusion of magma with minor gas exsolution, and the production of somewhat vesicular pillow lavas or sheet flows, depending on effusion rate. The amounts by which magma would overshoot the vent in these types of eruptions would be insufficient to cause any magma disruption. The most likely mechanism of production of pyroclastic deposits in this environment is strombolian activity, due to the localized concentration of volatiles in magma that has a low rise rate; magmatic gas collects by bubble coalescence, and ascends in large isolated bubbles which disrupt the magma surface in the vent, producing localized blocks, bombs, and pyroclastic deposits. Another possible mode of occurrence of pyroclastic deposits results from vulcanian eruptions; these deposits, being characterized by the dominance of angular blocks of country rocks deposited in the vicinity of a crater, should be easily distinguishable from strombolian and hawaiian eruptions. However, we stress that a special case of the hawaiian eruption style is likely to occur in the submarine environment if magmatic gas buildup occurs in a magma reservoir by the upward drift of gas bubbles. In this case, a layer of foam will build up at the top of the reservoir in a sufficient concentration to exceed the volatile content necessary for disruption and hawaiian-style activity; the deposits and landforms are predicted to be somewhat different from those of a typical primary magmatic volatile-induced hawaiian eruption. Specifically, typical pyroclast sizes might be smaller; fountain heights may exceed those expected for the purely magmatic hawaiian case; cooling of descending pyroclasts would be more efficient, leading to different types of proximal deposits; and runout distances for density flows would be greater, potentially leading to submarine pyroclastic deposits surrounding vents out to distances of tens of meters to a kilometer. In addition, flows emerging after the evacuation of the foam layer would tend to be very depleted in volatiles, and thus extremely poor in vesicles relative to typical flows associated with hawaiian-style eruptions in the primary magmatic gas case. We examine several cases of reported submarine volcaniclastic deposits found at depths as great as 3000 m and conclude that submarine hawaiian and strombolian eruptions are much more common than previously suspected at mid-ocean ridges. Furthermore, the latter stages of development of volcanic edifices (seamounts) formed in submarine environments are excellent candidates for a wide range of submarine pyroclastic activity due not just to the effects of decreasing water depth, but also to: (1) the presence of a summit magma reservoir, which favors the buildup of magmatic foams (enhancing hawaiian-style activity) and episodic dike emplacement (which favors strombolian-style eruptions); and (2) the common occurrence of alkalic basalts, the CO₂ contents of which favor submarine explosive eruptions at depths greater than tholeiitic basalts. These models and predictions can be tested with future sampling and analysis programs and we provide a checklist of key observations to help distinguish among the eruption styles.     

Groundwater residence times and flow paths in fractured rock determined using environmental tracers in the Mission Tunnel; Santa Barbara County,California, USA

Multiple geochemical tracers [ion chemistry, stable isotopes of water, chlorofluorocarbons (CFC), tritium] and a 25-year-long record of discharge were used to understand residence times and flow paths of groundwater seeps in the fractured rock aquifer surrounding the Mission Tunnel, Santa Barbara, California. Tritium data from individual seeps indicate that seep waters are a mixture of >45-year-old (recharged prior to the nuclear bomb tests) and young groundwater. CFC data support this interpretation, however, a two-end member mixing model cannot completely explain the age tracer data. Microbial degradation and partial re-equilibration complicate the CFC signal. Spectral analysis of precipitation and groundwater seepage records shows that seepage lags precipitation by 3 months. This delay is related to the advancement of the wetting front and increasing the number of active flow paths. Additionally, the amount of seepage produced by precipitation is less during extended periods of drought than during normal or wet periods, suggesting antecedent conditions strongly affect flow through this fractured rock aquifer.     

Eastern Laurentia in Rodinia: constraints from whole-rock Pb and U/Pb geochronology

Whole-rock Pb isotopic signatures and U/Pb geochronology refute a Rodinian correlation of northeastern Laurentia and proto-Andean Amazonia. According to this previously proposed model, the Labrador–Scotland–Greenland Promontory (LSGP) of northeastern Laurentia collided with the proto-Andean margin of Amazonia, at the Arica Embayment, during the Grenville/Sunsás Orogeny (ca. 1.0 Ga). Links between the two margins were based upon the correlation of the LSGP with Arequipa-Antofalla Basement (AAB), a Proterozoic block along the proto-Andean margin of Amazonia adjacent to the Arica Embayment. Specifically, similarities in 1.8–1.0 Ga basement rocks in both regions suggested that the AAB was originally a piece of the LSGP. Furthermore, similarities in unique, post-collisional, but pre-rift, glacial sedimentary sequences also supported a link between the AAB and LSGP.Tests of these apparent similarities fail to support correlation of the AAB and the LSGP and, thus, eliminate a direct link between northeastern Laurentia and southwestern Amazonia in Rodinia. However, Pb isotopic compositions and U/Pb geochronology provide the basis for two new correlations, namely, (1) the ca. 1.3–1.0 Ga basement in the central and southern Appalachians may be an allochthonous block that was transferred to Laurentia from Amazonia at ca. 1.0 Ga, and (2) an allochthonous AAB may be a piece of the Kalahari Craton that was transferred to Amazonia at ca. 1.0 Ga. Based on these new correlations and a previously proposed Grenvillian connection between southern Laurentia (Llano) and Kalahari, we propose that Amazonia may have collided with a contiguous southeastern Laurentia/Kalahari margin at ca. 1.0 Ga.     

Ion microprobe measurements of O/O ratios of phosphate minerals in the Martian meteorites ALH84001 and Los Angeles

Oxygen isotope ratios of merrillite and chlorapatite in the Martian meteorites ALH84001 and Los Angeles have been measured by ion microprobe in multicollector mode. δ¹⁸O values of phosphate minerals measured in situ range from ∼3 to 6‰, and are similar to Martian meteorite whole-rock values, as well as the δ¹⁸O of igneous phosphate on Earth. These results suggest that the primary, abiotic, igneous phosphate reservoir on Mars is similar in oxygen isotopic composition to the basaltic phosphate reservoir on Earth. This is an important first step in the characterization of Martian phosphate reservoirs for the use of δ¹⁸O of phosphate minerals as a biomarker for life on Mars. Cumulative textural, major-element, and isotopic evidence presented here suggest a primary, igneous origin for the phosphates in Los Angeles and ALH84001; textural and chemical evidence suggests that phosphates in ALH84001 were subsequently shock-melted in a later event.     

Desorption kinetics of radiocesium from subsurface sediments at Hanford Site, USA

The desorption of ¹³⁷Cs⁺ was investigated on sediments from the United States Hanford site. Pristine sediments and ones that were contaminated by the accidental release of alkaline ¹³⁷Cs⁺-containing high level nuclear wastes (HLW, 2 × 10⁶ to 6 × 10⁷ pCi ¹³⁷Cs⁺/g) were studied. The desorption of ¹³⁷Cs⁺ was measured in Na⁺, K⁺, Rb⁺, and NH₄⁺electrolytes of variable concentration and pH, and in presence of a strong Cs⁺-specific sorbent (self-assembled monolayer on a mesoporous support, SAMMS). ¹³⁷Cs⁺ desorption from the HLW-contaminated Hanford sediments exhibited two distinct phases: an initial instantaneous release followed by a slow kinetic process. The extent of ¹³⁷Cs⁺ desorption increased with increasing electrolyte concentration and followed a trend of Rb⁺ ≥ K⁺ > Na⁺ at circumneutral pH. This trend followed the respective selectivities of these cations for the sediment. The extent and rate of ¹³⁷Cs⁺ desorption was influenced by surface armoring, intraparticle diffusion, and the collapse of edge-interlayer sites in solutions containing K⁺, Rb⁺, or NH₄⁺. Scanning electron microscopic analysis revealed HLW-induced precipitation of secondary aluminosilicates on the edges and basal planes of micaceous minerals that were primary Cs⁺ sorbents. The removal of these precipitates by acidified ammonium oxalate extraction significantly increased the long-term desorption rate and extent. X-ray microprobe analyses of Cs⁺-sorbed micas showed that the ¹³⁷Cs⁺ distributed not only on mica edges, but also within internal channels parallel to the basal plane, implying intraparticle diffusive migration of ¹³⁷Cs⁺. Controlled desorption experiments using Cs⁺-spiked pristine sediment indicated that the ¹³⁷Cs⁺ diffusion rate was fast in Na⁺-electrolyte, but much slower in the presence of K⁺ or Rb⁺, suggesting an effect of edge-interlayer collapse. An intraparticle diffusion model coupled with a two-site cation exchange model was used to interpret the experimental results. Model simulations suggested that about 40% of total sorbed ¹³⁷Cs⁺ was exchangeable, including equilibrium and kinetic desorbable pools. At pH 3, this ratio increased to 60-80%. The remainder of the sorbed ¹³⁷Cs⁺ was fixed or desorbed at much slower rate than our experiments could detect.     

Paleohydrology of arid southeastern maui, hawaiian islands, and its implications for prehistoric human settlement

Arid slopes on the southeastern side of Maui are densely covered with archaeological remains of Hawaiian settlement from the late prehistoric to early postcontact period (ca. A.D. 1500-1860). Permanent habitation sites, agricultural features, and religious structures indicate perennial occupation and farming in a subregion called Kahikinui, yet there is presently no year-round water source. We explore the possibility that postcontact deforestation led to the loss of either (1) perennial channel flow or (2) perennial springs or seeps. To investigate the first possibility, we estimated ancient peak flows on 11 ephemeral channels in Kahikinui using field measurements and paleohydrology. Peak-flow estimates (3-230 m³/s) for a given drainage area are smaller than those for current perennial Maui streams, but are equivalent to gauged peak flows from ephemeral and intermittent streams in the driest regions of Hawai’i and Maui islands. This is consistent with the long-term absence of perennial channel flow in Kahikinui. On the other hand, others have shown that canopy fog-drip in Hawai’i can be greater than rainfall and thus a large part of groundwater recharge. Using isolated live remnants and snags, we estimate the former extent of the forest upstream from archaeological sites. We use rough estimates of the loss of fog-drip recharge caused by deforestation and apply a simple, steady-state hydrologic model to calculate potential groundwater table fall. These order-of-magnitude estimates indicate that groundwater could have fallen by a minimum of several meters, abandoning perennial seeps. This is consistent with archaeological evidence for former perennial seeps, such as stonewalls enclosing potential seeps to protect them. Although longer-term reductions in rainfall cannot be ruled out as a factor, deforestation and loss of fog-drip recharge are obvious and more immediate reasons for a recent loss of perennial water in Kahikinui, Maui.