V43E-1450 1340h
(U-Th)/He Dating Of Volcanic Phenocrysts With High-(U-Th) Inclusions, Bandelier Tuff, New Mexico
(U-Th)/He dating is typically applied to U-Th-rich crystals of apatite, zircon and titanite for thermochronologic and geochronologic studies. In some circumstances, mineral phases devoid of U and Th but containing U-Th-rich inclusions may also be dated. The host grains may efficiently retain He released from the inclusions, simplifying the alpha ejection or diffusive loss corrections. To test the application of (U-Th)/He dating in such circumstances, we analyzed volcanic fayalite and hedenbergite crystals containing abundant U-Th-rich inclusions found from the \sim1.6 Ma Otowi (Lower) Member of the Bandelier Tuff in New Mexico. U and Th, the major parent nuclides of radiogenoic $^{4}$He, are concentrated in the britholite ((Ca, REE$^{3+}$, Na)$_{10}$(Si, P)$_{6}$O$_{24}$(OH, F) $_{2}$) and chevkinite ((Ce, La, Ca, Th) $_{4}$(Fe$^{2+}$, Mg) $_{2}$(Ti, Fe$^{3+}$)$_{3}$Si$_{4}$O$_{22}$) inclusions and are essentially absent in the host grains of fayalite and hedenbergite. A subset of phenocrysts was subjected to controlled air-abrasion to remove the alpha-ejection affected regions. We analyzed 36 fayalite (21 non-abraded + 15 abraded) and 13 hedenbergite (7 non-abraded + 6 abraded) phenocrysts, and applied alpha recoil correction assuming uniform intracrystalline distribution of inclusions and parent nuclides in the phenocrysts. The F$_{T}$ factors (alpha retentivity) of the non-abraded samples are in the range of 0.90-0.96, whereas those of the abraded samples are essentially unity. The fayalite and hedenbergite crystals yield weighted mean ages of 1.62 \pm 0.11 Ma (1 standard deviation) and 1.68 \pm 0.11 Ma, respectively. The overall weighted mean age of 1.63 \pm 0.10 Ma is identical with the previously published $^{40}$Ar/$^{39}$Ar ages of 1.61 \pm 0.01 Ma. The calculated (U-Th)/He ages may decrease by up to 0.05 myr when the secular disequilibrium of U and Th decay series is considered. Nevertheless, the two isotopic systems yield indistinguishable ages within their uncertainties. These results suggest that other inclusion-bearing volcanic phenocrysts and their host units may be dated by the (U-Th)/He method. As shown in this study, olivine and pyroxene are suitable for this approach. However, phases such as quartz, sanidine or muscovite may not retain a significant portion of trapped $^{4}$He even at low temperatures over short time scales, undermining possible application of He dating to these minerals.
V43E-1451 1340h
Geochemical investigation of magma evolution in Campi Flegrei volcanic field (Italy) based on melt inclusions
Melt inclusions are a powerful tool for understanding the evolutionary history (crystallization, volatile evolution, mixing, etc.) of magmas. They are also useful to interpret the conditions of primary melt generation and extraction. In particular the volatile content in magmas (e.g. H$_{2}$O, CO$_{2}$) is of critical importance in determining the eruptive style and magma evolution, because degassing is usually one of the major phenomena before and during an eruption. In populated and active volcanic areas such as Campi Flegrei (Italy), understanding the role of volatile content in magmas is a necessary requirement to assess volcanic risks. The goal of our project is to define the geochemical parameters that control the internal dynamics of magmatic systems of this active volcanic area. The eruptive products of Campi Flegrei were selected for study based on age, eruptive characteristic, mineralogy, chemical compositions and structural position of eruptive centers, in order to verify the existence of a possible change in plumbing systems between different eruptions. Here we report preliminary data on latitic-to-trachytic products from Fondo Riccio eruption. Compositions of the inclusions and host phenocryst were obtained by electron microprobe techniques. The compositions of melt inclusions in pyroxenes vary from trachybasalt to trachite, in those Mg-rich diopside are less evolved ( ranging from trachybasalt to shoshonite) than in Fe-rich diopside (from latite to trachyte). The compositions of inclusions in Fo 85-87 olivine are similar to those in Mg-rich pyroxenes. Homogenization temperatures were obtained by microthermometric experiments and range from 950 to $1000\deg$C for feldspars, from 1030 to $1080\deg$C for pyroxenes and from 1070 to $1120\deg$C for olivines. Volatile contents of melt inclusions, based on electron microprobe analyses, are $<$ 2 wt %. The chemical composition of minerals and melt inclusions, as well thermometric data, are consistent with a continuos crystallization process of an evolving magma body.
V43E-1452 1340h
New Thoughts About Newberry Volcano, Central Oregon USA
New geologic mapping, geochemistry, argon dating, and paleomagnetic data suggest some significant new ideas about the history and evolution of Newberry Volcano. Previous mapping (MacLeod et al., 1995) made use of K-Ar dating done in the 1970's. A new $^{40}$Ar/$^{39Ar}$ age of 83$\pm$5 ka for the rhyolite of Paulina Peak, a thick lava flow cut by the southern caldera-bounding fault and previously thought to be hundreds of thousands of years old, indicates that the caldera formed since $\sim$83 ka. The ash-flow tuff (Qat) that forms the western rim of the caldera contains rhyodacitic pumices with unusually high Na2O (6.1 wt.%), identical in composition to a subset of pumices found in the underlying and more widespread basaltic lapilli tuff (Qbt), suggesting that these two ash-flow tuffs represent parts of the same eruptive event. Their paleomagnetic directions are very similar. Lapilli identical in composition and appearance to those in the Qbt are found in gravel underlying basalt that has been dated at about 80 ka. Taken together, the data suggest that the caldera formed about 80 ka when Qbt and Qat erupted. An earlier caldera likely formed about 300 ka, the approximate age of each of several mapped rhyolitic to dacitic ash-flow tuffs on the eastern side of the volcano, one of which was previously thought to be $>$500 ka in age. Preliminary paleomagnetic data suggest that these ash-flow tuffs have similar remnant directions. Previous identification (Kuehn, 2002) of some 40-50 Pleistocene tephras indicates either that Newberry is far more explosive than previously thought or that many of these tephras (of which no more than 6 occur in any single hand-dug pit) may represent different lobes of tephra generated by compositionally-zoned caldera-forming eruptions. The widespread Olema ash bed (a centimeter thick in the San Francisco Bay Area, $\sim$700 km SSW of Newberry Volcano) may represent one lobe of the $\sim$80-ka caldera-forming event. In the course of geologic mapping, 1- to 2-meter-sized boulders of a variety of lava flow types have been found on the tops of cinder cones at elevations as low as 5782 feet. Interpreted as glacial erratics, they suggest that ice extended far down the flanks of the volcano, rather than forming only very small cirques and limited moraines at the caldera rim (highest elevation 7984 feet), as previously thought. Large areas of mixed pumice and gravel at higher elevations may represent till that grades into outwash gravel at lower elevations. Periodic outburst floods from lakes formed in or beneath the ice cap (possibly linked to eruptive activity?) would explain the large channels that are so abundant on the eastern and, to a lesser extent, on the western (and younger) side of the volcano. As many as half a dozen postglacial ($<$$\sim$12 ka?), but pre-Mazama ($\sim$8 ka), basaltic lava flows occupy several of the eastern and southeastern dry channels. The dry channels on the NE side appear to drain into the canyon of Dry River, a large, dry box canyon several hundred feet deep, which may have been carved by multiple large floods originating primarily from Newberry Volcano.
V43E-1453 1340h
Stratigraphy and Paleomagnetism of the Pine Creek and Castle Creek Eruptive Episodes, Mount St. Helens, Washington
We present new paleomagnetic data, and we compile and re-evaluate published data for the Pine Creek and immediately following Castle Creek eruptive episodes at Mount St. Helens. For the first time, paleomagnetic sites are correlated with detailed volcanic stratigraphy for these late Holocene eruptive episodes. New and existing radiocarbon data and correlation with western North America geomagnetic secular variation data provide age control. Our work leads to a revised understanding of the eruptive sequence and suggests modifications to the traditional interpretation of Pine Creek and Castle Creek eruptive episodes. Previously, the beginning of the Castle Creek episode was defined as the first appearance of andesite in the Pine Creek- Castle Creek sequence, and a hiatus of a few hundred years was inferred to have been present between the two episodes (for example, see Mullineaux 1996, USGS PP 1563). However, a number of early Castle Creek andesitic pyroclastic deposits and lava flows have magnetic directions unlike later Castle Creek units. Instead, they are similar to Pine Creek dacitic pyroclastic flows. Correlation between Castle Creek age lava flows and pyroclastic deposits and Castle Creek tephras (B set of Mullineaux) shows that the hiatus defined by a gap in the paleomagnetic record occurs between tephras Bo and Bi instead of at the presently accepted boundary between the Pine Creek and Castle Creek eruptive episodes. Thus, Castle Creek tephras Bh and Bo and associated units have magnetic directions similar to Pine Creek, and tephras Bi and Bu and associated units have Castle Creek magnetic directions. Magnetic directions of both eruptive episodes are entirely consistent with the late Holocene western North America geomagnetic secular variation as presently understood. Pine Creek dacites erupted about 2,700-2,600 years B.P. Castle Creek andesites (and tephras Bh and Bo) probably erupted about 2,600-2,300 years B.P., and Castle Creek dacites and basalts (and tephras Bi and Bu) erupted between about 2,100-1,900 years B.P.
V43E-1454 1340h
Preliminary Geology of Gareloi Volcano, Western Aleutian Islands (Alaska)
Gareloi Island consists of Gareloi volcano (1573 m elevation), and is located nearly 2000 km west of Anchorage and 120 km west of Adak in the western Aleutian (Andreanof) Islands. A geologic mapping operation was combined with the installation of a seismic monitoring network in September of 2003 by the Alaska Volcano Observatory. This work provided the first direct observations of Gareloi volcano since Robert Coats' four-day visit in 1945. Gareloi volcano is a stratovolcano 10 km by 8 km in diameter at its base with two summit craters separated by a narrow saddle. The southern crater is a 300-m-wide amphitheater formed by the partial collapse of its southern crater wall, and contains several active fumaroles. The northern crater is enclosed, although the intra-crater eruptive stratigraphy is abruptly interrupted by near-vertical local unconformities on the northwest wall, suggesting the occurrence of a sector collapse sometime in the past. Gareloi volcano is principally composed of intercalated trachytic lava flows, ranging from 0.5 m to more than 10 m in thickness. Two prominent valleys composed of thick lava flow packages on the SW flank are clearly U-shaped, suggesting that the oldest sequence of lava flows is of at least late Pleistocene age. Lavas erupted during the Pleistocene and Holocene range from basaltic trachyandesite to basaltic andesite in composition and contain plagioclase and clinopyroxene, with minor olivine, and rare hornblende. An explosive eruption in 1929 formed a SSE trending fissure of thirteen aligned craters, ranging from 80 to 1600 m in diameter. These craters extend from sea level up to the amphitheater of the southern crater (1160 m). Fall deposits from the 1929 eruption are interbedded with thin, laterally discontinuous pyroclastic flow deposits that are mainly limited to the island's southeastern flanks. Despite an abrupt change in color from light beige pumice clasts at the base of the 1929 fall deposit to black scoria at the top, the unit is homogeneous trachyandesite. Following the explosive phase of the eruption, 4 blocky trachyandesite lava flows emerged from craters below 600 m asl. All 1929 eruptive products contain plagioclase and clinopyroxene with scarce olivine. An effusive eruption during the 1980's from the center of the south crater amphitheater produced an elaborate blocky lava flow that extends 800 m in elevation down the SE flank. This lava flow is basaltic trachyandesite, and contains abundant coarsely sieved plagioclase phenocrysts with minor clinopyroxene and olivine. The majority of Gareloi lavas contain anomalously high concentrations of K, Na, and Rb and low concentrations of Mg compared to reported findings from other Aleutian lavas, including those of the western portion of the arc. This suggests that Gareloi magmas may be unique with respect to their source region and possibly storage conditions compared to other Aleutian volcanoes.
V43E-1455 1340h
Crisis GIS: Preparing for the Next Volcanic Crisis in the United States
Geographic Information Systems (GIS) specialists from the Volcano Hazards Program (VHP) of the U.S. Geological Survey (USGS), including personnel at Menlo Park, California, the Cascades Volcano Observatory in Vancouver, Washington, the Alaska Volcano Observatory in Anchorage and Fairbanks, Alaska, the Hawaiian Volcano Observatory in Hawaii National Park, Hawaii, and the Smithsonian Institution Global Volcanism Program in Washington, DC, are developing a GIS response plan in the event of a volcano crisis. This plan, referred to as "Crisis GIS", outlines how VHP can ensure rapid, reliable delivery of spatial and ancillary information for data analysis and visualization at any required location during a volcanic crisis or event within the United States. An effective Crisis GIS needs the capacity to store multiple, large datasets, including: base layer data, elevation data, geologic maps, hazard assessment maps, satellite data, and aerial photography for volcanoes around the U.S. It must be readily accessible by VHP GIS specialists stationed around the Nation. Such a GIS should also support installations of monitoring instruments and telemetry equipment that relay monitoring signals, and provision of updates to public officials, the media, and the public during a crisis. GIS technology has proven to be an invaluable tool for crisis response. Recently, GIS was applied as part of the response efforts to two large-scale crises: the terrorist attacks of September 11, 2001, and the Southern California wildfires of Fall 2003. In each case, GIS was used to organize large quantities of spatial data and to produce electronic and paper maps that illustrated hazards, supported decision making, and communicated developing situations to responsible emergency-management authorities and to the populace affected (Kant, 2002, and Pratt, 2003). VHP GIS specialists are currently testing the software and hardware employed in recent major crisis response efforts and are learning to adapt the technology for volcano crisis response.
V43E-1456 1340h
A.D. 1835 Eruption of Volc\'{a}n Cosig\"{u}ina, Nicaragua: New Data and Interpretations
The eruption of Volc\'{a}n Cosig\"{u}ina, Nicaragua, in January 1835, greatly impressed witnesses owing to its large area of ash fallout and great distance at which eruption noises were heard. Initial estimates of tephra volume of tens to more than 100 km$^{3}$ were lowered substantially by later reconnaissance studies to less than 10 km$^{3}$-more likely 3 to 6 km$^{3}$. But little has been done to map the distribution of eruptive deposits and to collect systematic stratigraphic observations around the volcano. We present the results of preliminary field and laboratory investigations that provide new details about proximal deposits. Our key conclusions are the following. Initial eruptions produced a tephra sequence of ash to fine lapilli, including a white dacitic pumice-fall deposit that is thickest and coarsest in a narrow sector to the northwest of the vent. The overlying main tephra-fall deposit is composed of dark-colored andesitic scoria that is thick and coarse in a broad sector extending clockwise from northwest to southeast, but appreciably thinner and finer-grained in south and southwest sectors, consistent with transport by high-altitude westerly winds. Pyroclastic surges and flows of dark-gray andesitic scoria and ash followed the scoria fall and formed fans and valley fills in northwest, northeast, and southeast sectors. Ash clouds, probably mostly related to pyroclastic flows, produced thick (locally $>$1 m) ashfall deposits, rich in accretionary lapilli, chiefly in the southwest sector, transported by prevailing low-altitude northeasterly winds. Bulk volume of 1835 eruptive products may lie at the low range of previous estimates, about 3 to 4 km$^{3}$.
V43E-1457 1340h
Preliminary geology of the Tanaga Island volcanic cluster, western Aleutians (Alaska)
During 2003, the northwestern portion of Tanaga Island (178$\deg$W) was mapped in detail for the first time during fieldwork by Alaska Volcano Observatory geologists in conjunction with the installation of a volcano monitoring seismic network. The northern half of the island is approximately 20 km across and comprises four discrete volcanic centers, from west to east: Sajaka (area = 22 km$^{2}$), Tanaga (20 km$^{2}$), East Tanaga (15 km$^{2}$), and Takawangha (54 km$^{2}$). The three western centers are steep-sided cones of Holocene age, and have grown in the scar formed by a catastrophic Pleistocene sector collapse directed to the northwest. To the east, a $>$300 m-thick sequence of volcanic and volcaniclastic rocks (Pre-Tanaga unit) underlies Takawangha, which has been active throughout the Pleistocene and Holocene. Holocene eruptive products from all four centers are predominantly lava flows, with minor explosive activity recorded in tephra sections. Additionally, Sajaka experienced a relatively young (mid-Holocene?) sector collapse of its west flank, accompanied by eruption of laterally-directed pyroclastic flows. A morphologically young cone of scoria and thin basalt flows has grown in the collapse scar. Whole-rock geochemical data on 130 samples of lava and scoria from the four centers and the Pre-Tanaga unit are basalts through low-SiO$_{2}$ andesites. All but ten lavas have molar Mg\# between 0.35 and 0.5 and the remainder are between 0.5 and 0.61; no primitive lavas were discovered on Tanaga Island. Lava flows of Holocene age from Tanaga and East Tanaga follow medium-K trends, all lavas from Takawangha are high-K, and Sajaka and Pre-Tanaga lavas fall along both trends. High-K lavas are enriched in other LILE (Rb, Ba, Pb) as well, and fall near or above the high end of published Aleutian lavas for these elements. The lavas exhibit petrographic as well as compositional diversity: mafic phases in Tanaga lavas are olivine, two pyroxenes, and amphibole, at East Tanaga lavas contain two pyroxenes, amphibole, and minor olivine, and Sajaka, Takawangha, and Pre-Tanaga lavas are predominantly olivine- and clinopyroxene-bearing. Thus the character of the sources as well as crustal storage conditions for magmas within the Tanaga Island cluster likely vary, despite close spacing in space and time.
V43E-1458 1340h
Geochemistry of Late Quaternary Terrestrial Tephra Deposits From the Eastern Mediterranean
Tephra layers provide geologists and archaeologists with isochronous marker deposits, which, if interpreted correctly, can provide valuable chronostratigraphic information. Explosive volcanic activity in the Hellenic arc over the last 150 000 years has left a record of terrestrial, lacustrine and marine tephra deposits across the Aegean and western Anatolia which are of great value in stratigraphic correlation. Some of these eruptions have produced substantial volumes of ash (e.g. Kos, 150 ka, Santorini 3600 years BP) and for stratigraphic correlation accurate recognition of these individual tephra deposits is imperative. This project investigates the tephrochronology of the Eastern Mediterranean which is believed to be important in two contexts, (i) in placing time constraints on the evolution of civilisations in that area and (ii) in understanding the evolution of the Hellenic arc. Results from analysis of single glass shards from tephra deposits on Kos, Yiali, Rhodes and Santorini as well as other minor islands and selected sites in Turkey are presented. Major and trace elements were analysed by electron probe, ion probe and LA-ICP-MS techniques. A trace element database has been generated for both distal and proximal samples and has aided in the understanding of the magma chamber evolution of the volcanoes in question. These analyses also provide a detailed, single grain, chemical stratigraphy of many islands which will enable temporal and spatial correlations and ultimately, will aid in the discussions of the impact volcanic eruptions exerted upon Aegean civilisations. Arguments related to the timing of Santormi's eruptions are considered and data are also-presented for tephra from the Alaskan volcano, Aniakchak. Thus, adding further weight to disproving a 1645 BC date for the Minoan eruption, but confirming the time of the eruption of Aniakchak. Finally the study provides a comparison between two trace element analytical techniques, ion probe and the LA-ICP-MS. This serves to increase confidence in comparisons between previous studies which employed these techniques.
V43E-1459 1340h
On the Age of the Laschamp Geomagnetic Excursion
The Laschamp geomagnetic excursion is a critical tie-point found directly in deep sea sediment cores and revealed in polar ice as an abrupt change in the rate of cosmogenic nuclide flux. Despite the importance of this excursion to quantifying paleoclimate proxy records archived in sediment and ice, and to providing an independent calibration of the radiocarbon calendar, its timing remains poorly known. Previous K-Ar, 40Ar/39Ar, and U-Th isochron determinations from lava flows at the type locality in the Massif Central, France, vary widely, are imprecise, and suggest a mean age of about 46.2 +/- 2.5 ka (2sigma). Results of 6 new unspiked K-Ar and 13 40Ar/39Ar incremental heating experiments on subsamples from 3 sites on the Laschamp and Olby flows are concordant and give a weighted mean age of 40.4 +/- 1.1 ka (2sigma uncertainty including analytical sources only) that is 10% younger than the previous estimates. Considering that the 40K - 40Ar decay constant is not known to a precision better than 2.4%, the most probable radioisotopic age for the Laschamp excursion is 40.4 +/- 2.0 ka (2sigma, analytical plus decay constant uncertainties). This new age for the Laschamp excursion agrees precisely with that deduced from the NAPIS-75 deep sea sediment paleointensity stack when calibrated against the GISP2 ice core chronology using the O isotopes in ice and the magnetic properties of the marine cores.
V43E-1460 1340h
Origin of a basalt-to-trachyandesite/dacite suite in the Jemez volcanic field (New Mexico, USA) by upper crustal contamination
Magmatic compositions in large, long-lived volcanic fields typically evolve through a variety of petrogenetic processes. Determining their relative roles may be difficult because (1) multiple processes may act simultaneously at a single volcano, (2) different processes may produce similar compositional results, and (3) individual processes may differ significantly at different volcanoes. To isolate the major processes and avoid the complex overprinting of multiple processes in the main Jemez volcanic field (JVF), we obtained whole-rock major- and trace-element compositions and mineral compositions from a small basalt-to-trachyandesite/dacite volcanic field (Cerros del Rio, CDR) adjacent to the main JVF. Lavas of the CDR field were erupted during development of the middle to upper crustal magma chamber (4-1 Ma) that resulted in formation of Toledo and Valles calderas at 1.6-1.2 Ma, but were not affected by mixing with the main chamber. The mafic end members of this suite typically range from basanite/alkali olivine basalt to hawaiite (SiO2 44-52%; Na2O+K2O 4.2-6.3%), likely derived from an ocean-island-basalt (OIB)-modified lithospheric source (McMillan, N. Mex. Geol. Soc. Guidebook 49, 107-116, 1998). Compositions range to trachyandesite and dacite (SiO2 65-67%; Na2O+K2O 6.9-7.2%) (classification of Le Bas et al., J. Petrol., 27, 745-750, 1986). Trace elements (e.g., Th, Zr) are not well correlated with major elements. Phenocrysts range from olivine and plagioclase in the least evolved lavas to high- and low-Ca pyroxene and sodic plagioclase in the intermediate to silicic members. Primitive members of this series contain gabbroic xenoliths and evolved members are typified by quartzofeldspathic xenoliths and by xenocrysts of quartz and feldspar, which we infer were derived from middle to upper crust. Using only whole-rock compositions observed in the CDR, we can model the major-element compositional range assuming removal of observed phenocryst phases coupled with addition of observed xenolithic/xenocrystic material. Modeling of trace elements, assuming world-wide average values (Rudnick and Gao, Treatise on Geochem., Chap. 3, 2004), is not compatible with a significant lower crustal component in these magmas. We conclude that the simplest model for the origin of this basalt-to-trachyandesite/dacite suite is fractional crystallization of OIB-modified basalt combined with significant assimilation of middle to upper crustal rocks. Although we cannot rule out more complicated models involving partial fusion of lower crust and mixing of magmas, we do not think they are necessary or likely.
V43E-1461 1340h
Volcanic Rocks Collected With ROV Tiburon From Rodriguez Seamount, Located at the Continental Slope of the California Borderland
Volcanic rocks were collected from Rodriguez Seamount at the outer margin off the Continental Borderland with MBARI's ROV Tiburon in October 2003 and April 2004. Six dives recovered lava and volcaniclastic samples from the deep flanks ($\sim$2120 m) to the summit at 630 m. Whole rock compositions of plagioclase-olivine-clionpyroxene bearing lava samples are predominantly alkalic basalt ($<$8% MgO) and hawaiite with minor mugearite (MgO=1.5%). Glass compositions of pillow rims and of volcaniclastic fragments in breccia and bedded sandstone are predominantly hawaiite, mugearite and minor evolved alkalic basalt. The lava samples include one rhyolite and one basaltic andesite with subduction-related chemistry; they are probably erratics. Other clearly identifiable erratics include granite, quartzite, amphibolite, and bored, erosion-sculpted sandstone, resembling typical beach deposits. Most of these erratics are pebble- to small cobble-size and occur in conglomerate and crossbedded sandstone that surround the summit at a break in slope that most likely marks the shoreline when Rodriguez was an island. The lava outcrops on the gently domed platform of the summit are dense, oxidized \`{a}\`{a}-like flows without glassy rinds. Sulfur content of glass, collected from the flanks of the volcano, ranges from 1300 ppm of a glass inclusion in an olivine crystal to $\sim$160 ppm of volcaniclastic grains, indicating extensive degassing. Petrographically and chemically these lavas are virtually identical to those erupted on Miocene seamounts offshore central California (e.g. Davidson, Guide, Pioneer, Gumdrop seamounts, Davis et al, 2002) as well as Northeast Bank on the continental shelf south of Rodriguez and seamounts farther offshore from the Continental Borderland (e.g. Little Joe, San Marcos, San Juan seamounts, Clague et al, unpublished; Davis et al., 1995). Trace element abundances and ratios (e.g. LREE, Zr/Nb, Ta/Nb) also completely overlap with those from the other sites, suggesting similar origins by small percentage of melting of somewhat enriched MORB-like sources. Ar-Ar dating for these samples is in progress but published ages (Davis et al, 2002; Bohrson and Davis, 1994) indicate sporadic eruptions during the late Miocene along the continental margin, extending from the continental shelf out to the seamount province. This volcanism postdates the change from a subduction-related regime by at least 7 million years and does not appear to be related to migration of a triple junction. Instead episodes of localized extension in response to movement along the transform fault systems allowed small degree of decompression melting of mantle rising along zones of weakness.
V43E-1462 1340h
H$_{2}$O and Hydrogen Isotopic Study of Glasses from Subglacial Volcanoes in the Neovolcanic Zone of Iceland
Worldwide interest in subglacial volcanism was rekindled by the spectacular 1996 Gjalp eruption in Iceland. In Iceland, the flat-topped volcanoes that form beneath glaciers are called table mountains or tuyas. Sections of the basaltic subglacial volcanoes in Iceland reveal a crude stratigraphic progression from pillow lavas to hyaloclastite deposits from the base upward. Locally the sections are capped by subaerial basaltic lava flows. Based on classic models, the glassy pillow basalts and overlying hyaloclastite deposits indicate eruption of basalt within melt water caverns of increasingly shallow depths. Capping subaerial flows would record the emergence of lava above the ice sheet. Large thicknesses of ice and melt water overlying the basaltic volcanoes significantly impede the release of volatiles from the ascending magma. As degassing and ascent rate are viewed as the critical variables controlling eruptive style, H$_{2}$O contents and hydrogen isotopic analyses of glasses were conducted in order to improve our understanding of degassing processes at subglacial volcanoes. These data may also help to refine current models of subglacial basaltic eruptions. Consequently, samples of glassy pillow basalts, pillow breccias, and hyaloclastites were collected at Hlodufell, Raudafell, Efstadalsfjall, and Geitafell. Major element compositions of the glasses measured by electron microprobe reveal uniform basaltic compositions with relatively primitive MgO values ranging only from 7.37-7.68 wt%. The glass compositions show both tholeiitic and alkalic affinities on an alkalis-silica classification diagram. In order to definitively track degassing at these subglacial volcanoes, pillow and hyaloclastic glasses were analyzed for their H$_{2}$O concentration and hydrogen isotopic composition. Glasses from the Raudafell section have H$_{2}$O contents ranging from 0.21-1.24 wt% and bulk $\delta$D values from -87.2 to -105.7$\permil$. Vapor saturation pressures calculated from these dissolved H$_{2}$O contents range from 5 to 151 bars assuming that associated CO$_{2}$ contents are zero. This range in pressure corresponds to an ice thickness of 57-1710 m based on an ice density of 0.9 g cm$^{-3}$. Step heating hydrogen isotopic analyses were undertaken to evaluate the nature of the H$_{2}$O in the basaltic glasses. A Raudafell sample released 43% of its water at temperature steps of 200 and 400$\deg$C indicating a meteoric origin, whereas the remaining 57% of the H$_{2}$O liberated during the fusion step is likely magmatic in origin. Accurate determination of the magmatic H$_{2}$O component would reduce the estimation of ice thickness above these subglacial volcanoes.
V43E-1463 1340h
Evidence That Three Seamounts Off Southern California Were Ancient Islands
Eleven ROV Tiburon dives in 2003 and 2004 explored Rodriguez and San Juan Seamounts, and Northeast Bank off southern California. Now submerged, these seamounts appear to have been subaerially exposed while the volcanoes were active. The summit of Rodriguez seamount, now at 630 m depth, is a smooth, gently domed platform. The shallowest points are low, rough hills standing above the platform and consist of thick, dense, degassed `a`a flows erupted and oxidized subaerially. Coarse bedded and cross-bedded sandstones and rounded cobbles interpreted to be beach deposits occur near the top of a major break-in-slope at about 700 m. Rodriguez stood at least 70 m above sea level and formed a small island 6.8 km$^{2}$ in area prior to subsiding at least 700 m. The summit of Northeast Bank, now at about 360 m depth, is another smooth, gently domed platform. Wave-sculpted lava flows, sandstone, and pebbly conglomerate beach deposits were observed between 510 and 554 m depth, below the sharp break-in slope at about 500 m depth. Although the bathymetry of the entire summit is poorly known, Northeast Bank was a large island perhaps 200 m above sea level and 90 km$^{2}$ in area. The summit of San Juan Seamount, now at 560 m depth, is a northeast-southwest trending series of rough ridges rather than a dome. However, subaerially oxidized `a`a-like lava flows were found above 700 m. San Juan's summit ridge emerged as a line of eight small islands with a total area of about 2.8 km$^{2}$, and the tallest island rose 140 m above sea level. The 550 to 700 m of subsidence recorded by Northeast Bank within the California borderland, Rodriguez Seamount on the continental slope, and San Juan Seamount on the adjacent ocean crust occurred since these seamounts formed, most likely since the late Miocene.
V43E-1464 1340h
Imaging Structure, Stratigraphy and Groundwater with Ground-Penetrating Radar on the Big Island, Hawaii
A series of exploratory ground-penetrating radar (GPR) surveys were carried out on the Big Island, Hawaii in March of 2004 to evaluate the efficacy of using GPR to address hydrological, volcanological, and tectonic issues in extrusive basaltic materials. Target sites included beach sands, nearshore lava flows, well-developed soil covers, lava tubes, and major fault zones. Surveys were carried out with a Sensors and Software T Pulse Ekko 100, which was equipped with 50, 100, and 200 MHz antennae. Both reflection profiles and CMP expanding spreads were collected at most sites to provide both structural detail and in situ velocity estimation. In general, the volcanic rocks exhibited propagation velocities of ca 0.09-0.10 m/ns, a value which we interpret to reflect the large air-filled porosity of the media. Penetration in the nearshore area was expectedly small (less than 1 m), which we attribute to seawater infiltration. However, surveys in the volcanics away from the coast routinely probed to depths of 10 m or greater, even at 100 MHz. While internal layering and lava tubes could be identified from individual profiles, the complexity of returns suggests that 3D imaging is required before detailed stratigraphy can be usefully interpreted. A pilot 3D survey over a lava tube complex supports this conclusion, although it was prematurely terminated by bad weather. Although analysis of the CMP data does not show a clear systematic variation in radar velocity with age of flow, the dataset is too limited to support any firm conclusions on this point. Unusually distinct, subhorizontal reflectors on several profiles seem to mark groundwater. In one case, the water seems to lie within a lava tube with an air-filled roof zone. Surveys over part of the controversial Hilana fault zone clearly image the fault as a steeply dipping feature in the subsurface, albeit only to depths of a few meters. The results suggest, however, that deeper extensions of the faults could be mapped by more aggressive surveying techniques (e.g. lower frequencies or high levels of signal stacking). In summary, GPR has proven to be a very effective technique in mapping stratigraphy, structure, and hydrology in the volcanic rocks of Hawaii. Future work is planned to evaluate more fully the merits of 3D imaging, the systematics of propagation velocity with composition and age of extrusion, the internal morphology of successive flows, and whether GPR can discriminate aa from pahoehoe flows at depth.
V43E-1465 1340h
Geochemical And Petrological Investigations Of The Representative Cretaceous Bentonite Beds, Wyoming : Tectonic Implications
Representative bentonite samples were collected from the exposed Cretaceous siliciclastics in Bighorn Basin, north-central Wyoming. Bentonite beds constitute a significant stratigraphic importance with respect to local and regional correlation tool and are associated with the Thermopolis Shale (Lower Cretaceous), Mowry Shale (Lower Cretaceous), and Frontier Formation (Upper Cretaceous). These beds range in thickness from few inches to ten feet and are interbedded with thick cross-bedded sandstone, pebbly sandstone, polymictic conglomerate, siliceous-rippled shale, and lignitic shale. Preliminary petrological and geochemical investigations were carried out to establish a distinctive geochemical signature for each bed. Emphasis was given on the overall distribution of the immobile traces, high refractory elements, and ultrastable heavy mineral components among the selective bentonite beds. The outcome of petrological, bulk, and trace element studies involving multiple bentonite beds indicated a subtle difference in terms of abundance of trace-element and detrital components among the studied samples and can be attributed to the source region characteristics, distinctive diagenetic pathways, and depositional setting. Furthermore, geochemical analyses involving multi-element plots suggest to an evolving source terrain located in close proximity to the bentonite depositional basin.
V43E-1466 1340h
ROBOVOLC: The Geochemical Package
ROBOVOLC is a special designed robot equipped with different scientific packages to collect solid, liquid and gas samples and make several measurements. This project was developed to reduce the risks to researchers working in volcanic areas (Muscato G. et al., 2003). In particular, the geochemical package allows to collect some different typology of gas samples (dray gases, NaOH flask, acid condense) and perform some direct measurements (gas temperature, mass and energy release, dry gas/steam ratio). The gases from fumaroles are collected by a 1.2m heated probe in order to keep the robot away from the fumaroles. The gas temperature is measured by several thermocouples located in different part of the probe (on the top to measure the fumarolic temperature) and along the gas canalization in order to check the temperature decreasing. The first step in the sampling procedure consists of the thermal stabilization of the equipment. In this phase, the fumarolic gases flow inside the system for a variable period of time depending on the temperature and flux of the fumaroles. Dray gases and acid condense (Capasso et al., 1992) are collected after the fumarolic gas is cooled by a pyrex glass steam condenser. This stage allows to measure the volumes of both liquid and gas phase and, therefore, to calculate the gas/steam ratio. NaOH flask (Giggenbach, 1975) can be collected excluding the condensed by some electro valves. The gas flux inside the flask can be regulated by a motorized system and a video camera. All the gas canalization is made in pyrex glass and PTFE.