V13C-1485
The Source of Volcanic Ash in Late Classic Maya Pottery at El Pilar, Belize
The presence of volcanic ash used as temper in Late Classic Maya pottery (AD 600-900) at El Pilar has been long known although the volcano(s) contributing ash have not been identified. We use geochemical fingerprinting, comparing compositions of glass shards in potsherds with volcanic sources to identify the source(s). El Pilar is located in the Maya carbonate lowlands distant from volcanic sources. It is unlikely Maya transported ash from distant sites: ash volumes are too large, the terrain too rugged, and no draft animals were available. Ash layer mining is unlikely because mine sites have not been found despite intensive surveys. Nearest volcanic sources to El Pilar, Belize and Guatemala, are roughly 450 km to the south and east. The ash found in potsherds has a cuspate morphology. This suggests ash was collected during, or shortly after, an ash airfall event following eruption. Analyses of n=333 ash shards from 20 ceramic (pottery) sherds was conducted by electron microprobe for major elements, and LA-ICPMS for trace elements and Pb isotopes. These analyses can be compared to volcanic materials from candidate volcanoes in the region. The 1982 El Chichon eruption caused airfall deposition (< 1 mm isopach) at El Pilar which lead Espindola et. al, 2000 to suggest that earlier eruptions at El Chichon could have caused ash fall at El Pilar during the Late Classic. 333 individual glass shards found within about 20 distinct potsherds have a mean silica content of 78.3±1.1 wt. % (one-sigma). The 1982 El Chichon eruption products have a mean silica content of 60.2±7.2% (one-sigma, n=48); the circa 1475 AD eruptive products of El Chichon have a mean silica content of 53.4±3.4 wt % (one-sigma, n=8). This suggests that El Chichon was not a source of the ash at El Pilar. In order confirm or refute the El Chichon source hypothesis, comparison of trace element ratios between archaeological samples and El Chichon has been made. The atomic ratios of La/Yb, Nb/Ta, Zr/Hf, Sr/Ba and Th/U of n=215 glass shards in the potsherds are 12.2±7.1, 10.9±3.4, 31.2±11.5, 0.09±0.05 and 2.5±0.9, respectively. These ratios for 1982 El Chichon are 15.4±2.1, 26.3, 36.1±5.3, 1.4±0.06 and 3.16, respectively. Data for the 1475 AD El Chichon eruption (Macias et al, 2003) can also be compared; the ratios from are 13.2±2.2, 7.3±1.8, 30.4±9.6, 1.51±0.4 and 2.88±0.23, respectively. The mean 208Pb/206Pb ratio of n=5 potsherds is 2.0523±0.002 compared to 2.0514±0.00074 for n=7 samples from El Chichon. The two most recent eruptions from El Chichon overlap with the potsherd glass data except for Sr/Ba, which might be modified by Sr-Ca exchange during firing. In order to test the effects of pot firing on glass compositional changes, experiments were conducted in which high silica volcanic glass was fired with clay according to heating schedules used by Maya potters. Two important changes are that Na is rapidly lost preferentially to K and that the Si/Ca ratio decreases due to Ca diffusion from matrix into glass during firing. One expects that ratios of the refractory trace elements such as La/Yb and Zr/Hf are less susceptible to modification. Further experiments of trace element mobility during firing are underway.
V13C-1486
Precursory magma activities leading to Aira caldera-forming eruptions in southern Kyushu, Japan
A series of volcanic products from Aira caldera (20 × 20 km) recorded a unique history of magma reservoir underneath the caldera from 100 ka until present. The largest volume was represented by Aira pyroclastic eruptions which produced more than 400km3 of rhyolitic tephra at 29 ka consisting mainly of Ito pyroclastic flow and regional Osumi pumice fall and A-T ash fall. A trasition of magma composition was recorded in Iwato tephra formation at 60 ka, which consists of 3 fall layers, 5 flow layers and minor surge layers. Some of these layers contain scoria, banded pumice and white pumice. We made petrological descriptions and whole-rock XRF analyses for the two key Aira pyroclastic materials and products of other pre-caldera-forming events. i.e., Fukuyama pumice fall deposit (Fk; 90-86 ka), Shikine andesitic lava (Sk; ca 70 ka), Iwato tephra formation (Iwt; ca 60 ka), Shimizu ryolitic lava (Sm; 39-33 ka), Otsuka pumice fall deposit (Ot; ca 33 ka), Fukaminato tephra formation (Fm; ca 31 ka), Kenashino tephra formation (Kn; ca 30 ka). Three magma types were recognized from our study, type-A (67-70 wt. % SiO2; Fk), type-B (56-59 wt. % SiO2; Sk) and type-C (73-78wt. % SiO2; Sm, Ot, Fm and Kn). Volcanic samples from all types contain Pl, Opx, Cpx and Fe-Ti oxides phenocrysts. Type A contains abundant Hb but no Qtz phenocrysts, whereas types B and C contain abundant Qtz but no Hb phenocrsyts. Iwt samples represented a wide compositional variation (58-78 wt. % SiO2). Iwt scoria belongs to type-B, whereas Iwt pumice belongs to type C. The compositions of scoria (58-59 wt. % SiO2; type B), banded pumice (65-68 wt. % SiO2) and white pumice (71-78 wt. % SiO2; type C) make a straight line in Harker diagrams, suggesting a magma mixing event. Type-C magma emerged at 60 ka and had been active until 30 ka, then continued on to major 29 ka Aira pyroclastic eruptions. This suggests that large silicic magma reservoir which eventually led to the gigantic caldera-forming eruption possibly existed for a period of 30 ka. Mafic type-B magma was only active before and at the Iwt tephra event.
V13C-1487
Structural Relationship Between Piton des Neiges and Piton de la Fournaise Volcanoes: New K-Ar Data and Geomorphological Study of the Takamaka Region (East Reunion Island, Indian Ocean)
Reunion Island (Indian ocean) is a volcanic complex resulting from hotspot activity composed of three coalescent eruptive systems. The first subaerial volcano (la Montagne massif), which only outcrops in the NW part has been dated between 2.2 and 1.8 Ma (McDougall, 1971). After a major flank collapse of this volcano (Gillot et al., 1994), Piton des Neiges (PNv) edificated from 1.2 Ma to 30 ka (McDougall., 1971; Gillot et al.,1982). then, Piton de la Fournaise volcano (PFv), one of the most active on Earth, started its activity about 530 kyr ago (Gillot et al., 1989; 1990) and was affected by 3 eastward flank collapses (Gillot et al., 1994). Its present complex morphology is characterized by large scale erosional depressions, (Cirques) cut in the volcanic structures, such as Cilaos, Mafate or Salazie in PNv, Grand Bassin between the two volcanoes, and Grand Pays in PFv. Due to the tropical conditions, deeply incised valleys are present throughout the island. The eastern part of the island (Takamaka area), where we show that products of both PNv and PFv overlap, is one of the most rainy place in the world. It is deeply incised and has been highly eroded during the coeval building stages of PFv and PNv since at least 530 kyr. In order to constrain the relationship between the PNv and PFv volcanoes and to characterise the morphological evolution of this area, we realized a new geochronological study of the different massifs based on the accurate K- Ar technique devoted to the dating of very young rocks (Cassignol technique; Gillot et Cornette, 1986). A preserved structure between the two volcano, Morne de l'Etang, is dated between 1.36 +/- 0.02 Ma, which is older than the primary known activity of Piton des Neiges (about 1.2 Ma; McDougall, 1971), to 0.97 +/- 0.02 Ma. It may either correspond to a remnant and older part of PNv or it belongs to the Proto Fournaise ‘les Alizés' volcano', which existence is still debated. Our analysis also emphasizes the fact that PFv could have built Ilet Patience and Massif du Cratere (600-400 ka) which both are juxtaposed in relief inversion to the Morne de l'Etang. As a result, PFv seems to have a wider areal extention in the eastern part of Reunion island than previously thought, and the limits between the two volcanoes correspond clearly to the Marsouins River. Geochronological datas coupled to the geomophological analysis show that this region is characterised by many relief inversions, younger and younger from East to West e.g. Morne du Bras des Lianes (about 1 Ma), Coteau Mazerin (340-220 ka), and Plaine de Bebour (140-30 ka). This sector seems to have always been submitted to intense erosion rates which have controlled the later emplacement of lavas filling the successive Cirques cut in the structure. We show here that erosion is a major process in the morphostructural evolution of Reunion Island and most particularly in the Takamaka region, where eroded volume have the same order of magnitude (10 to 100 km3) than those involved in flank collapses identified on PFv. Finally, our new data allow the elaboration of first detailed model investigating the relationships and morphological evolution between the two PNv and PFv on the eastern sector of the island.
V13C-1488
Correlation of Late Quaternary tephra-fall layers as preserved in 25 sedge- Sphagnum peat cores recovered from the Kenai Peninsula, Alaska
Tephra-fall layers were detected in 25 sedge- Sphagnum peat cores recovered along a 70 x 115 km northeast to southwest transect of Alaska's Kenai Peninsula, which lies directly parallel to and generally downwind of the Cook Inlet volcanoes (Hayes, Spurr/Crater Peak, Redoubt, Iliamna, and Augustine) and other active Aleutian Arc volcanoes. Magnetic susceptibility (MS), field characteristics and stratigraphic context, petrography, geochemical trends (electron microprobe analyses), particle size, shard characteristics, and radiocarbon ages are being used to identify, characterize, and correlate tephra deposits among cores. Of the 221 undifferentiated tephra-fall layers preserved in our cores, approximately 80 percent were identified visually and 20 percent were identified by MS peaks and petrographic verification. Eighty AMS radiocarbon ages, ranging from recent to about 13,000 yr B.P., were obtained from Sphagnum sampled directly beneath visual tephra layers. Stratigraphic records from individual peat cores vary from site to site, in some cases significantly, even over very short distances (10 km), suggesting that no single peat record contains a complete volcanic history for the region. Thus, a composite tephrostratigraphy incorporating a wide geographic range is necessary in order to understand the magnitude and frequency of ash fall in the Cook Inlet region, home to over 60 percent of Alaska's population. Correlative deposits among cores must be established to show a composite tephrostratigraphy. A composite tephrostratigraphy for the Kenai Peninsula is constructed by correlating tephra layers from our 25 peat cores using multiparameter characterization. An index of correlation probability (ICP) is assigned for each correlated pair by presenting data in a matrix and assigning numerical values to candidate pairs by independently weighting tephra characteristics as parameters for correlation. This technique (as compared to geochemical similarity alone) enables a more confident correlation especially when geochemical differences between tephra deposits are small (as is the case of tephra erupted from the same vent or eruptive source area), or where individual tephra deposits are chemically heterogeneous. Tephra deposits are then assigned to source volcanoes, primarily from the western Cook Inlet, based on more general geochemical and petrographic data.
V13C-1489
An analysis of Late Quaternary eruption frequency as recorded by tephra-fall records from 25 sedge- Sphagnum peat cores recovered from the Kenai Peninsula, Alaska
Tephra fall (volcanic ash) is considered the principal hazard from Aleutian Arc volcanoes in terms of volume, distribution, and environmental impact. Over sixty percent of Alaska's human population resides in the Cook Inlet region, where ash fall from nearby volcanoes, including Hayes, Spurr/Crater Peak, Redoubt, Iliamna, and Augustine pose the greatest volcanic risk. Alaska's Kenai Peninsula is located generally downwind of Cook Inlet volcanoes (approx. 85 to 400 km) and other active Aleutian Arc volcanoes. Previous studies have shown that Holocene-age tephra fall is well preserved in post-glacial sediments from this region. Such studies have used individual stratigraphic records to estimate tephra-fall frequency on a regional scale, although it is unclear whether those data reflect actual eruption frequency or are biased by paleo-wind direction, basin features, geomorphology, etc. These studies have shown highly discrepant ash-fall frequency records, which may result from the restricted number of study sites (typically one or two), the lack of spatial coverage, and/or the limited preservation potential of a given location. In order to evaluate actual eruption frequency as reflected by tephra fall, our study incorporates a wider geographic range and a greater number of sample locations than previous studies. We recovered and examined 25 sedge- Sphagnum peat cores from a northeast to southwest transect of the Kenai Peninsula, directly parallel to the Cook Inlet volcanoes and covering an area of 8,050 km2 (70 km wide by 115 km long). Magnetic susceptibility (MS), petrographic and electron microprobe analyses, and radiocarbon ages have been utilized to identify, characterize, and correlate tephra deposits among cores. A total of 221 undifferentiated tephra-fall layers are preserved in our cores, of which approximately 80 percent were identified visually and 20 percent were identified by MS peaks and petrographic verification. Eighty AMS radiocarbon ages, ranging from recent to approximately 13,000 yr B.P., were obtained from Sphagnum sampled directly beneath visible tephra layers. We found that stratigraphic records from individual peat cores varied from site to site, in some cases significantly, suggesting that no single peat record contains a complete volcanic history for the region. Thus, a composite tephrostratigraphy from across the Kenai Peninsula is both necessary and advantageous to better understanding volcanic ash fall (as a proxy for eruption history) in Southcentral Alaska. Geochemical and petrographic correlations of our 221 tephra layers are in progress and will aid in identifying unique tephra deposits across this region, and linking them to their source volcanoes. These data will be used to verify and improve upon existing tephra-fall frequency records and their link to Cook Inlet explosive volcanism, and will improve our understanding of ash fall hazards in Southcentral Alaska.
V13C-1490
The Widespread Distribution of Komatiitic Tuffs in the 3.3 Ga Weltevreden Formation, Barberton Greenstone Belt, South Africa
The 3.5-3.2 Ga Barberton greenstone belt is a heavily deformed, 10-15 km thick succession of volcanic and sedimentary rocks representing one of the best preserved Paleoarchean supracrustal sequences known. It consists of the basal volcanic-dominated Onverwacht Group and the overlying sedimentary-dominated Fig Tree and Moodies Groups. Major volcanic rocks in the BGB include komatiites, tholeiitic basalts, and dacites. Although flow rocks and fragmental deposits have been identified representing all extrusive magma types, the abundance of komatiitic volcaniclastic units is remarkable considering the mechanical difficulties in explosively erupting low viscosity ultramafic lava. In the Onverwacht Group, most komatiitic tuffs contain 85-95 wt% SiO2, due to early silicification, and very low concentrations of most other elements, making original compositions somewhat uncertain. However, in the northernmost part of the BGB, north of the Inyoka Fault, the ~ 3.3 Ga Weltevreden Formation is composed largely of komatiitic flow rocks, tuffs, layered ultramafic complexes, and subordinate black and banded cherts. Previous studies have established the extrusive nature of the komatiites, but there are also many thick interlayered slaty units, previously interpreted as sheared flow rocks, which show cross-bedding, soft-sediment deformation, and other features indicating an alternate derivation. These units range from 2 to 80 m thick and may represent 10% or more of the overall stratigraphy of the Weltevreden Formation. They are characterized by low-temperature serpentinization that has commonly preserved original elemental abundances, enabling a more precise determination of primary komatiitic liquid composition. These rocks are magnesium rich, with MgO ranging from 23 to 36 wt%, and high Ni (~1500 ppm) and Cr (~2600 ppm) contents typical of komatiites. Several possible mechanisms could have produced these rocks, including (1) erosion and transport of pre-existing komatiitic flow rock, (2) volcanic base surges, (3) current reworking of fall-deposited pyroclastic material, and (4) remobilization of hyaloclastitic debris. The abundance of fine-grained sediments and of flat- and cross-laminated beds, the paucity of cr-spinels, and komatiitic immobile element ratios suggest that most of these high-Mg beds formed by minor reworking of komatiitic pyroclastic ash in a subaqueous environment.
V13C-1491
Implications of Temporal-Compositional Variations in the Cerro Pajas Flow and Cone, Floreana Volcano, Galapagos Islands
Floreana Island is a unique volcano in the Galapagos Islands due to its abundance of mantle xenoliths and the compositional imprint of mantle metasomatism. Floreana lavas are also the most alkaline in the archipelago and represent an enriched end-member due to their high 87Sr/86Sr and 206Pb/204Pb isotope ratios and high concentrations of incompatible trace elements. The surface exposure age of the lavas ranges from 1.52 Ma to 26 ka. Detailed mapping of a single eruptive unit, the Cerro Pajas sequence, shows 18 different eruptive units consisting of lava lobes and tephra. The age of the Pajas units is 26 ka. A suite of nineteen samples was collected from each of the flow lobes and tephra in order to assess variations over the course of the eruptive event. By studying the temporal-compositional trends of a heterogeneous eruptive sequence on Floreana, the melting mechanisms at Floreana can be assessed. Large variations in MgO concentration are seen over the course of the eruption, implying possible differences in the extent of differentiation. The beginning and end of the eruption show the highest concentrations of MgO (~11.5 wt. %). Most of the eruption, however, produced lavas with MgO concentrations of about 7.5 wt. %. The change in MgO cannot simply be attributed to different proportions of olivine phenocrysts, but must reflect changes in liquid composition. In general, concentrations of incompatible elements and ratios such as Nb/Zr increased over the course of the eruption by an amount greater than can be accounted for by fractional crystallization. This may be the result of a decrease in the degree of partial melting over the course of the eruption, and reveals that compositional heterogeneity is generated during a single melt extraction event and preserved during ascent through the lithosphere.
V13C-1492
Twilight of a Volcanic Field: 11 Million Years of Basaltic Volcanism in the Southwestern Nevada Volcanic Field, USA
Following the end of major caldera-forming silicic volcanism in the Southwestern Nevada Volcanic Field (SNVF), at least 10 episodes of alkalic basaltic volcanism have occurred over the last ~11 Ma. An understanding of the past behavior of the volcanic field provides insight for forecasting future eruptive behavior for use in hazard assessment for the high-level radioactive waste repository at Yucca Mountain. A program of geophysics, drilling, Ar-Ar dating and geochemistry conducted since 2004 by Los Alamos National Laboratory and the U.S. Geological Survey, combined with previous and ongoing petrogenetic and physical volcanology studies, sheds more light on the early and middle evolution of the volcanic field, much of which has been buried in alluvial basins. Volumes of erupted basalt have drastically declined over the history of the field, from as much as 50 km3 in the Miocene to about 0.5 km3 in the Pleistocene. The volume decrease is accompanied by a drastic decrease in extension rate, suggesting a close link between magmatism and tectonism. Neodymium and strontium isotopic analyses indicate that enriched lithospheric mantle has been the source of basalt throughout the history of the field. Decreasing eruption volumes are accompanied by an approximate doubling of Ce/Yb ratios, indicating that the volume decrease reflects a decrease in degree of partial melting of the lithospheric source. Eruption style has also changed with time, reflecting an increase in magma volatile content, consistent with decreased amounts of partial melting of a volatile-bearing source. These observations are consistent with a model in which the lithospheric mantle source was hottest during the period of major silicic volcanism and the presence of an active subduction system. After the breakdown of subduction, continued thermal input into the lithosphere ceased, and the lithosphere began to conductively cool. Melt accumulation in non-convecting, static lithosphere is probably related to the presence of mantle heterogeneities enriched in hydrous minerals that are partially melted. During regional extension, these zones are relatively weak and preferentially deform, forming melt bands of increased porosity that concentrate melt and lead to dike generation. Decreasing regional extension results in less melt accumulation and decreasing eruption volumes. Without a new source of heat and limited lithospheric extension, it is likely that the next million years of volcanic activity in the field will likely be characterized by eruptions of the type that have occurred during the past million years of activity: infrequent eruptions of small-volume (<0.1 km3), volatile-rich alkali basalt magmas within the most tectonically active southern and western margins of the volcanic field.
V13C-1493
Differentiation of Historical Hekla Magmas
59 tephra and lava samples from 16 historical Hekla eruptions span the compositional range from basaltic andesite to dacite. The eruptive order of each of these samples is well constrained. Our analyses confirm previous work which showed that basalts are limited to lateral fissure systems flanking Hekla and have not erupted from the volcano in historical time. Most eruptions begin with a plinian to sub-plinian explosive phase which transitions in a matter of hours to an effusive phase for the remainder of the eruption. Historical tephra glasses have SiO2 ranging from 55.1-73.9 wt.%. Whole rock tephra have a slightly lower range of SiO2 (55.2- 70.0%), owing to the relatively high crystal content of the most evolved tephra. The historical lavas have a generally lower range in SiO2 (54.3-65.1 wt.%). This expanded and highly precise data base confirms Thorarinsson's (1967) observation that the SiO2 content of the first material erupted is proportional to the length of repose between eruptions. The most evolved material is always erupted during the first part of the explosive phase. The lavas of the effusive phases are dominated by basaltic andesite compositions, but they too show a time dependent zoning which invariably attains a base level of ~54 wt.% SiO2 by the end of the eruptions. Taken together, the data show remarkably coherent major and trace element trends that suggest the main series of Hekla lavas up to ~65 wt.% SiO2 is controlled by fractionation of olivine, pyroxene, Fe-Ti oxide, and apatite. Plagioclase is present as phenocrysts in all compositions but is not removed (fractionated) from the liquid until ~63 wt.% SiO2 Hekla rocks are crystal poor with <5% phenocrysts, and the crystallinity of the rocks relates directly to silica content. The mineral assemblage is consistent throughout the Hekla suite and consists of plagioclase, olivine, clinopyroxene, Fe-Ti oxide, apatite, and a few rocks contain orthopyroxene. Phenocryst compositions relate in a straightforward way to the rock compositions, lack significant zoning, and have little variation in each sample. Some lavas show thin reaction rims of pyroxene mantling olivine crystals. Our interpretation is that the phenocrysts grew from the host liquids after the SiO2 gradient was established in the magma reservoir. The data collected thus far are consistent with a persistent basaltic andesite magma chamber beneath Hekla volcano in which the top of the magma column evolves through fractional crystallization proportional to the length of time between eruption events. Melt inclusions in olivine from the 2000 basaltic andesite tephra and 1104 rhyolite tephra have H2O contents ranging from 2.5-2.7 wt.% and 4.6-6.0 wt.%, respectively, but CO2 contents for all melt inclusions are below detection. Calculated vapor saturation pressures at magmatic temperatures range from ~0.6 kbars for the basaltic andesite to 1.4-2.2 kbars for the rhyolite. These values are significantly less than the >3 kbar estimate for the present-day Hekla magma chamber indicating either: 1) crystallization under vapor-undersaturated conditions in the magma reservoir; or 2) formation of at least some crystals within the conduit system above the reservoir. The ubiquitous zoning in Hekla magmas suggests a stable thermal state and recharge rate through time extending at least since the last large rhyolite eruption.