V44C-01 INVITED
Zircon U-Pb Ages Chronicle 3 Myr of Episodic Crystallization in the Composite Miocene Tatoosh Pluton, Mount Rainier National Park, Washington Cascades
Zircon geochronology of upper crustal plutons can constrain longevities of intermediate to silicic magmatic systems. As part of a larger study of the geochemistry and metallogeny of Tertiary Cascades magmatic arc rocks, we used the USGS-Stanford SHRIMP RG to determine 20 to 28 238U-206Pb ages for zircons from each of 6 quartz monzodiorite (qmd), quartz monzonite (qm), or granodiorite (grd) samples representative of the Tatoosh pluton, and one grd from the nearby Carbon River stock. The 7x12 km composite Tatoosh pluton, discontinuously exposed on the south flank of Mount Rainier, consists of at least 4 petrographic/compositional phases, here termed Pyramid Peak, Nisqually, Reflection Lake, and Tatoosh. These collectively intrude gently folded and weakly metamorphosed basaltic andesite flows and volcaniclastic rocks of the Eocene Ohanapecosh Formation, silicic ignimbrites and sedimentary rocks of the Oligocene Stevens Ridge Formation, and basaltic to intermediate volcanic rocks of the Miocene Fifes Peak Formation. Histograms and relative probability plots of U- Pb ages indicate 2 to 4 age populations within each sample. The weighted mean age of each of the youngest populations (all ±2σ) is interpreted as the time of final solidification: Pyramid Peak qmd (58.5% SiO2) 17.4±0.2 Ma, Nisqually grd (in Paradise Valley; 65.4% SiO2) 16.7±0.2 Ma, Nisqually grd (at Christine Falls; 66.4% SiO2) 17.3±0.2 Ma, Reflection Lake qm (along Pinnacle Peak trail; 66.6% SiO2) 17.1±0.2 Ma, Tatoosh grd (in Stevens Canyon; 67.8% SiO2) 18.2±0.2 Ma, Tatoosh grd (south of Louise Lake; 69.3% SiO2) 19.3±0.1 Ma, and Carbon River grd (68.0% SiO2) 17.4±0.3 Ma. The older Nisqually grd age is indistinguishable from a TIMS zircon age of 17.5±0.1 Ma reported by Mattinson (GSA Bulletin 88:1509-1514, 1977) for grd from a nearby locality. None of the 164 SHRIMP-RG U-Pb ages, including cores, is older than 21 Ma. The relatively small, high-level pluton likely was emplaced and solidified in pulses; zircons with ages significantly greater than the final solidification age of a given sample are considered to be antecrysts recycled from earlier crystallization episodes within the larger magmatic system. Although interpretation of the age populations is subjective, we have identified 17.4-Ma antecrysts in the 16.7-Ma Nisqually grd and antecrysts of 1-3 ages averaging 18.1, 18.8, 19.3, and 20.0 Ma in the other samples. Notably, age populations are separated by 0.5-0.7 Myr, intervals similar to lifetimes of large arc volcanoes. The total duration of pluton assembly is ~2.6 Myr, or as much as ~3.6 Myr if the oldest antecrysts are considered. The oldest, most differentiated rocks are on the east side of the pluton (Tatoosh phase) and the youngest, least differentiated on the west (Pyramid Peak phase). Two samples each of the Nisqually and Tatoosh phases yield apparent crystallization age differences of 0.7- and 1.1-Myr, respectively, for petrographically and chemically similar rocks. The Tatoosh U-Pb data appear to chronicle repeated waxing and waning of a long-lived igneous system as recorded by crystal-rich magma that periodically solidified near the system's roof at the close of well-defined crystallization episodes.
V44C-02
Size and Longevity of Magma Chambers in the Tuolumne Batholith: A Comparison of Thermal Modeling and Cooling Thermochronology
Our field studies on the 1200 km2 Tuolumne batholith and the controversy about how this 95-85 Ma composite batholith was constructed, motivated us to evaluate the thermal evolution of volumetrically large magma bodies formed by an amalgamation of a few to many pulses. We use finite difference 1D and 2D models with full spatial heterogeneity of rock properties, fine-scale internal grid spacing that allows for the definition of intricate rock geometries, and small internal time steps for calculations over any time duration. Careful code construction for numerical stability, computational efficiency, and resource management (dynamic memory allocations and CPU parallelization) allows us to model at scales between sub-m to km's for time durations of days to many millions of years. Several types of initial and boundary conditions including thermal gradients and heat flux and the effects of latent heat of fusion are installed. We have modeled a number of chamber construction scenarios. (1) single intrusions of rectangular or elliptical geometry (i.e., sills, dikes, or blobs) emplaced at variable time; (2) A sequence of intrusions emplaced at specified but arbitrary times or according to a time rate. Shapes in the sequence are fixed or set to randomly vary within a range of dimensions and aspect ratios. (3) A sheeted dike complex can be created wherein the thermal model actually expands according to a growth (extension) rate to accommodate the emplacement of new but thin dikes. Dike width and the time between dikes are coupled based on growth rate; (4) Finally, irregularly shaped bodies from a series of maps or cross-sections can be emplaced into our thermal model and digitally rendered into rock types, which are assigned thermal properties. These mapped shapes are emplaced into the thermal model at specified times so that they represent new thermal pulses. The use of maps or cross-sections allows us to examine the thermal behavior of observed field geometries. This wide range of chamber construction scenarios naturally result in a wide range of length and timescales of magma chambers. But many likely scenarios indicate that the lobes of the Tuolumne batholith will crystallize in less than 500,000 years, whereas large parts of the main chamber stayed above the solidus for 1-3 million years. Both our field studies throughout this batholith and thermochronology (U/Pb zircon, Ar40/Ar39 of hornblende and large and small biotite populations) along three transects support this conclusion.
V44C-03
Longevity of the Silurian Vinalhaven Intrusive Complex, Maine, USA
Many silicic intrusions preserve apparent stratigraphic sections of interlayered silicic, hybrid and mafic rocks. These sections provide a potential temporal framework that can be used to constrain the longevity of plutonic systems and delineate detailed histories of pluton construction. Field relations indicate that the Silurian (419 Ma) Vinalhaven intrusive complex (VIC) has a stratigraphic record of magma chamber evolution and pluton growth by crystal accumulation on magma chamber floors that was interrupted by episodes of replenishment, rejuvenation and eruptions (Wiebe and Hawkins, this volume). The system preserves evidence for a wide range of magma chamber processes including, multiple replenishments of compositionally varied magmas, mingling and mixing of those magmas, and mafic replenishments that rejuvenated granitic crystal mush (Wiebe et al., 2004) and perhaps triggered roof collapse associated with caldera formation (Hawkins and Wiebe, 2004). The purposes of this study are to test the stratigraphic model interpreted from field relationships in the VIC and to constrain the timescale of pluton construction via high-precision ID-TIMS U-Pb zircon geochronology. We collected four samples of cg granite from different stratigraphic levels in the intrusion and one sample of fg granite from a large-volume intrusion interpreted as a silicic replenishment late in the system's history. The U-Pb ages were determined from the weighted mean of the 206Pb/238U dates of 5 to 9 concordant analyses of either whole single crystals of zircon or individual crystal terminations (tips) sliced from whole zircon crystals. Each grain or grain fragment was either air-abraded or annealed and chemically abraded prior to dissolution. Ages have uncertainties of 110 – 70 k.y. at 419 Ma and are statistically robust. The three cg granite samples from the layered granite-hybrid-gabbro unit near the base of the intrusion yield ages that are consistent with the interpreted stratigraphy and span about 300 k.y. In that time interval, more than 100 individual mafic intrusions were trapped within the resident magma chamber(s) and silicic magma repeatedly replenished the system to construct several kilometers of layered section. A fourth cg granite sample collected in the upper portion of the intrusion is indistinguishable in age from the youngest sample collected in the layered section. This indicates that about 300 k.y. after the oldest granite sample formed, the intrusion had a thick wall/roof of crystallized granite, at least locally. The sample of fg granite is about 700 k.y. younger than the granite at the margin of the intrusion at the base of the layered unit. Therefore the intrusion, as exposed today, was constructed in at least 700 k.y.
V44C-04
Evolution of Volcano-Plutonic Centers in the Northern Colorado River Extensional Corridor, Nevada-Arizona: Protracted Cycles of Replenishment, Mush Accumulation, Fractionation, and Melt Extraction
Three major magmatic systems were simultaneously active during early development of the northern Colorado River extensional corridor. These systems, centered on Spirit Mountain batholith, Searchlight pluton, and Aztec Wash and Nelson plutons, include hypabyssal intrusions as well as coarser, deeper-seated rocks emplaced at depths of 5-13 km. Erupted products of Searchlight are clearly exposed; connections between the other systems and extensive coeval volcanic sequences, while very likely, remain unverified. Intrusion at each center began at 17-18 Ma and terminated with a dike swarm at ca. 15.5 Ma. Dikes, sills, and evidence for mingling document the frequent replenishment suggested by longevity of the centers. Quenched magmas in pillows, dikes, and chilled margins indicate that input included trachybasalt (49- 52 wt pct SiO2), trachydacite (quartz monzonite; 62-65 wt pct SiO2), and low-Si rhyolite (granite; ca. 73 wt pct SiO2); similar magmas formed a large part of the regional volcanic sequence. Some of the basalt may represent juvenile magma from enriched mantle, but Sr and Nd isotopic data indicate that all other input magmas are hybrids with both juvenile and ancient crustal components. Although local mixing is evident from field and geochemical evidence, the system-wide hybridization occurred at deeper crustal levels prior to emplacement into the upper crust. Whole-rock elemental compositions, field relations, crystal-size distributions and textures within the volcanic rocks and co-genetic intrusions indicate repeated cycles of magma emplacement and extraction of fractionated melt from cumulate mush. Cumulates are enriched in Sr and Ba and have positive Eu anomalies relative to input magmas. The melt-rich extracts have high-silica rhyolite compositions. They are exposed in plutons as small dikes and large subhorizontal sheets and roof zones comprising fine-grained, commonly vesicle-rich aplitic granite, and they erupted from the Searchlight center and probably from the others. These rocks are extremely depleted in Sr, Ba, and Eu, and middle REE, low in light REE, P, Ti, and Zr, and enriched in Rb, reflecting fractionation of feldspars and accessory minerals. Their high SiO2 (77-79 wt pct) attests to fractionation at shallow levels, consistent with emplacement depths of the upper parts of the plutons. SHRIMP U-Pb ages and compositional zoning in zircon also indicate repeated cycles of growth from fractionating melts and recycling into less evolved melts. Most samples have two or more distinguishable age populations, and many individual grains show evidence for resorption. Compositions of zones indicate that grains experienced dramatic temperature fluctuations and were transferred from highly fractionated to unfractionated melts. Taken together, these data indicate that the intrusions formed and were modified by repeated (a) felsic replenishment that eventually formed thick crystal mush; (b) mafic replenishments that helped to maintain a thermal balance; (c) extraction of fractionated melt into local conduits or ponding zones, to the roof, or to erupt at the surface; (d) rejuvenation of the mush by the preceding processes, entraining and transporting crystals and blurring previous intrusive contacts.
V44C-05
Constraining differentiation processes and timescales from mineral-scale isotopic data
The mechanisms by which magmas diversify en route to the surface and the timescales over which this differentiation occurs have received a great deal of attention over the past decade. Many magma systems appear to be characterised by crystal recycling and cannibalisation of progenitor products on short timescales. Single crystal isotopic data indicate that magmas contain crystal cargoes that have formed in different places and at different times, and have been aggregated shortly before, or during, emplacement/ eruption. Integration of isotopic zoning with textural features in crystals commonly reflects multiple recharge and mixing of different composition magmas. The fact that individual crystals from the same rock may have different isotopic zoning profiles further indicates that they do not share a common differentiation history. The c. 5000 cubic kilometer Fish Canyon Tuff, for example, exhibits extreme inter- and intra-crystal Sr isotopic diversity, despite being relatively homogeneous with respect to major and trace elements at the bulk rock scale. Such heterogeneity is more likely inherited from progenitor magma systems which have incubated in and interacted with the crust, rather than from mantle sources. The timescales over which the processes of contamination, recharge and mixing operate can be estimated from the compositional gradients in the crystals. The failure of small biotite grains in the Fish Canyon Tuff to isotopically equilibrate with the host magma requires entrainment and eruption within a timescale less than that required for diffusive equilibration – in this case a few thousand years. The common observation that crystal rims are sometimes not in isotopic equilibrium with the host glass is consistent with entrainment of crystals into the host melt shortly before eruption. The resolution of isotopic sampling is poor but timescales indicated are typically less than 1000 years. Trace element profiling constrains timescales better and can suggest pre- eruption mixing timescales of weeks to years.
V44C-06
Tracking the Longevity and Evolution of the Youngest Toba Tuff Using Allanite and Zircon Chronology, Thermometry, and Chemostratigraphy
Accessory phases are powerful repositories of information about the thermal and chemical evolution of silicic magma bodies. Allanite and zircon from the relatively youthful and chemically variable Youngest Toba Tuff (YTT) of Indonesia provide key opportunities to explore gestation of Earth's largest Quaternary eruption. The YTT is compositionally zoned from 63 to 77 wt.% SiO2 but high silica rhyolite is predominant. Chemical diversity within the YTT can be explained largely by crystal fractionation. We have previously determined that allanites entrained by the eruption began to crystallize as much as 150 k.y. before eruption of the YTT at ca. 73 ka, but most crystallized within 50 k.y. of eruption. Age ranges within individual allanites can be as large as 150 k.y. Allanites from felsic enclaves yield ages that are within tens of thousands of years before eruption, but are distinct in composition from allanites in their host pumice. Zircons, on the other hand, range in age to more than 500 k.y. before eruption, although most are within 300 k.y. of it. Rare xenocrysts of Paleocene (ca. 60 Ma) and late Proterozoic (ca. 700 Ma) age and probable antecrysts from the earlier Oldest Toba Tuff are also present. Individual zircons crystallized over ca. 30 to more than 500 k.y. intervals. Taken at face value, the wider age distribution of zircon compared to allanite could be explained by secular cooling of the magma reservoir, with zircon saturation occurring much earlier than allanite saturation. Zircon saturation temperatures, FeTi-oxides, and allanite compositions collectively suggest that most of the YTT crystallized and evolved over the temperature interval 700-800°C. Temperatures based on Ti-in-zircon geothermometry expand this range to 670- 830°C. The zircons thus appear to retain a more complete record of differentiation than the host lavas and may represent crystals retained from more mafic progenitors of and/or intrusions into the YTT as well as zircons remobilized from a (semi-)solid state. Zircon rims are, on average, more than 50°C lower than their cores and thus do not suggest mush remobilization by mafic intrusion. Some features of the chemical compositions of the zircons (e.g., Hf) correlate with the variation in temperature but neither the distribution of temperatures nor of chemical compositions are unique to pumice composition. Many cores are chemically distinct and even rim Y contents scatter despite expected bulk D(Y)~1. When combined with chronological information, there are no obvious secular chemical changes in zircons. The age, temperature, and chemical variations corroborate evidence from allanite for protracted crystallization within different, yet coeval, reservoir domains that differed in temperature and melt composition.
V44C-07
Do Zircon age Spectra Record Magmatic Cyclicity at Soufrière (Saint Lucia, Lesser Antilles)?
The Soufrière Volcanic Center (Saint Lucia, Lesser Antilles) is a long-lived arc-volcanic system that evolved over the past 5 - 6 Ma. Its most recent volcanic activity between 20 and 40 ka was concentrated within the prominent Qualibou topographic depression and produced two voluminous pyroclastic deposits: Choiseul and the overlying Belfond. In addition, several dacitic lava domes exist within the Qualibou depression. Because evidence of earlier volcanic activity in long-lived magma systems is frequently obliterated by subsequent eruptive or volcano-tectonic events, high spatial resolution U-Th dating of zircon combined with (U-Th)/He dating is a powerful tool to identify magma crystallization episodes at depth and to link these to the eruptive record. U-Th model ages and disequilibrium corrected U-Pb ages for 56 individual zircons from Soufrière lavas (Morne Bonin, Belfond, Terre Blanche) and pumice (Choiseul, Belfond) were determined by secondary ionization mass spectrometry. The majority of results is on unpolished zircons where analysis pits integrate over the outermost ~10 μm of individual grains with a lateral spatial resolution of ~40 μm. Selected grains were subsequently analyzed by (U-Th)/He methods. Belfond and Terre Blanche (U-Th)/He zircon ages (~20 ka) agree with previous 14C charcoal ages, whereas Morne Bonin ages are much older (~250 ka). Overall, the U-Th zircon crystallization age spectrum reveals a remarkable range between ~20 and ~600 ka and displays multiple peaks, among which the most prominent are tentatively identified at ~40 ka, ~80 ka, ~130 ka, ~200 ka and ~500 ka. The distribution of rim ages indicates that most zircons lack overgrowth dating from just prior to the eruption, but the youngest ages for each sample overlap with the eruption ages. Soufrière zircons thus reveal magma intrusion, cooling, and crystallization cycles within the underlying plutonic system for which the volcanic stratigraphic record is sketchy.
V44C-08
Zircon from Mount St. Helens Reveals Residence Times of Tens to Hundreds of Thousands of Years at Low Magmatic Temperatures Prior to Eruption
U-series and U-Pb geochronology of zircons from four samples that span the 300,000 year eruptive history of Mount St. Helens, combined with zircon trace element geochemistry and application of the Ti-in-zircon thermometer, provide critical constraints on the time-temperature-compositional history of the sub-volcanic magmatic system. Preliminary results indicate that prior to and throughout its eruptive history, one or more relatively cool, crystal-rich reservoirs have been accumulating beneath the volcano. SHRIMP-RG U-Pb ages of the oldest sample, a dacite erupted ~300 ka, reveal that zircons grew between ~320 and 520 ka, suggesting magmatic activity may have begun 200 ka before eruption. 238U-230Th age spectra in the three youngest samples indicate multiple ages of growth for each sample. The oldest of these three young samples (eruption constrained to ~35-50 ka) contains zircons ranging from ~50 to ~200 ka, with the main concentration of ages ~100 ka. Zircons from a 35 ka dacite range from ~65 to ~230 ka in age, with a dominant episode of growth ~130 ka. Dacite from the current eruption, sampled from the dome in 2005, contains zircons from ~40 to ~170 ka in age, with distinct populations at ~130 and ~170 ka. Taken together, these ages of tens to hundreds of thousands of years prior to eruption and the distinct episodes of growth suggest repeated injection and accumulation of one or more crystal-rich reservoirs beneath the volcanic edifice, a scenario that is further supported by Ti-in-zircon geothermometry (Watson et al. 2006). Ti concentrations indicate zircons grew at temperatures from ~840 to ~640 C, with 90% of analyzed spots recording temperatures between ~770 and ~670 C (T's carry uncertainties of tens of degrees, mostly from uncertainty in a(TiO2)). These temperatures are significantly lower than the eruption temperatures of their host magmas, which range from ~950-800 C. The rounded, resorbed surface morphology of many of the grains attests to the undersaturation of the phase in these hotter magmas. Calculated dissolution rates indicate that zircons would have completely dissolved in magmas such as these in tens to hundreds of years, suggesting that residence of the zircons at higher temperatures prior to eruption must have been brief. Gravity data (Williams et al., 1987) tentatively suggest that an intrusive complex lies beneath the volcano and may extend as far west as Goat Mountain, the earliest eruptive center of the St. Helens system. The accumulation of this complex may correspond with the ages represented by the zircons, although the conditions of magma storage in these intrusions are unknown. The zircons in these intrusions could be stored in sub- solidus rock, in discrete reservoirs of crystal mush, or in a larger crystal mush that continuously mixes magma batches and recycles zircon populations. Regardless of the conditions, it is apparent that eruptive events sample this storage zone and the zircons contained therein and carry them rapidly to eruption.