GEOS 470R/570R Volcanology L11, 23 February 2015 Handing out Today’s PowerPoint slides Returning graded midterm “Our plans miscarry because they have no aim. When a man does not know what harbor he is making for, no wind is the right wind.” --Seneca MT Scores (100 possible + 10 EC) 470R Undergraduate Students 570R Graduate Students 95 + 10 = 105 86 + 8 = 94 79 + 3 = 82 74 + 6 = 80 73 + 5 = 78 73 + 6 = 79 73 + 4 = 77 69 + 3 = 72 62 + 8 = 70 63 + 6 = 69 96 + 8 = 104 95 + 8 = 103 93 + 10 = 103 65 + 7 = 102 92 + 10 = 102 87 + 8 = 95 87 + 3 = 90 73 + 8 = 81 73 + 8 = 81 75 + 4 = 79 31 + 4 = 35 Statistics 2015 n = 10 mean = 74.7 + 5.9 = 80.6 85 ~ A/B ; 70 ~ B/C Statistics 2013 n=7 mean = 71 + 7.3 = 78.3 85 ~ A/B ; 70 ~ B/C Statistics 2015 n = 11 mean = 81.5 + 7.1 = 88.6 Statistics 2013 n=4 mean = 89.5 + 8.5 = 98 Readings from textbook For L11 from Lockwood and Hazlett (2010) Volcanoes—Global Perspectives [None] For L12 from Lockwood and Hazlett (2010) Volcanoes—Global Perspectives Chapter 12 Assigned reading For today L11 None For L17, 25 March 2015 Hildreth, W., and Lanphere, M. A., 1994, Potassium-argon geochronology of a basaltandesite-dacite arc system: The Mount Adams volcanic field, Cascade Range of southern Washington: Geological Society of America Bulletin, v. 106, p. 1413-1429. Last time: Calderas History and nomenclature Calderas Cauldrons Volcano-tectonic depressions Subsidence mechanisms Origin of megabreccias and structural versus topographic margins Resurgence Caldera cycles and repose times Diameters and area-volume relationships Examples of calderas and their associated ignimbrites Long Valley, CA Valles-Toledo, NM Yellowstone, WY-ID-MT San Juan volcanic field, including Creede caldera, CO Southwest Nevada volcanic field, with numerous calderas Sierra La Primavera, Mexico Pantelleria, Strait of Sicily, Italy Crater Lake, OR Mount Pinatubo, Philippines Caldera vs. cauldron Original definitions Caldera Large volcanic depression ~circular in plan that results from collapse Collapse probably a function of several variables but favored by high eruption rate, resulting in collapse of roof of magma chamber A morphological (surface) term Cauldron Volcanic subsidence structure without regard to shape, degree of erosion, connection to surface volcanism Current usage Virtually synonymous Caldera has taken precedence Caldera formation preceded or accompanied by voluminous eruptions of pumice and block collapses coherently along ring fractures Found in continental areas, associated with the largest, silicic eruptions e.g., Long Valley, Yellowstone Wohletz and Heiken, 1992, Fig. 1.10, adapted from Hildreth, 1981, Fig. 15 Caldera formation preceded or accompanied by voluminous eruptions of pumice and block collapses chaotically or piecemeal Found in central vent arc volcanoes, especially andesitic e.g., Krakatau Wohletz and Heiken, 1992, Fig. 1.10, adapted from Hildreth, 1981, Fig. 15 Caldera formation independent of pyroclastic eruptions Magma lost by lava flows and lateral removal (drainage of chamber) Not by pyroclastic explosion from a vent in the center of the caldera or from vents along the ring fracture zone Examples Craters in shield volcanoes of Hawaii Mt. Katmai during eruption of Valley of Ten Thousand Smokes from Novarupta vent Volcano-tectonic depressions Collapse occurs along regional (tectonic) faults Resulting depression commonly rectilinear Caution: Some “volcano-tectonic depressions” are “conventional” calderas that have been dismembered by postmineral faults Original caldera geometry obscured by younger deformation, such as multiple sets of normal faults (e.g., Caetano caldera, Nevada, John et al., 2008; Colgan et al., 2008) Stages of resurgent calderas Three stages (Lipman, 1984) A. Precollapse volcanism B. Ash-flow eruption and concurrent collapse C. Clustered intermediate stratocones (or domes) Broad uplift Intracaldera tuff ponds during subsidence (asymmetric); megabreccias; order of magnitude thicker than outflow facies Resurgence and postcaldera deposition Resurgence also asymmetrical Caldera floor nearly obliterated Lake sedimentation Hydrothermal activity Schmincke, 2004, Fig. 9.49, after Lipman, 2000, Fig. 2; from Lipman, 1984 Geometries of caldera subsidence Plate (piston) Single, large-volume eruption Piecemeal Multicyclic? Trap-door Asymmetrical pluton? Downsag Small volume/deep pluton? Funnel Small/deep pluton? Schmincke, 2004, Fig. 9.48, after Lipman, 2000, Fig. 6; from Lipman, 1997 Topographic wall vs. structural margin of calderas Moat Resurgent Dome Lipman, 1976, Fig. 1 Resurgent doming Produces structural doming May have subsequent keystone graben collapse Related to rapid re-establishment of isostatic equilibrium Occurs shortly after collapse of caldera Conventional view: Buoyant upwarp of cauldron block by reinflation (e.g., Smith and Bailey, 1968) Alternate view: Central uplift by injection of shallow sills or laccoliths into thick intracaldera fill (McConnell et al., 1995) Relationship to magmatic compositions Observed for rhyolitic to rhyodacitic systems Not observed for andesitic and basaltic systems Not observed for strongly peralkaline systems (rheologically similar to andesites) Keystone graben Moat Lipman, 2000, Fig. 2c; from Lipman, 1984 Proportion of chamber erupted Evacuated portion of magma chamber erupted Large continental calderas perhaps ~10% Smaller arc calderas perhaps more (10-25%?) Remainder of magma in chamber buoyantly readjusts following eruption Leads to resurgent doming (structural) and resurgent magmatism (e.g., eruption of lava domes) May be followed by recharge from lower crust Valles and Toledo calderas (Otowi and Tshirege Members of Bandelier Tuff), Jemez Mountains, NM Open-system behavior Lifespan of magmatic system Volcanic field: >3.6 m.y. (perhaps 16.5 m.y.) One chamber: >1.79 m.y. Volume of cataclysmic eruptions 300 km3, 1.23 Ma, Valles caldera 300 km3, 1.61 Ma, Toledo caldera Repose time for cataclysmic eruptions 0.38 m.y. Greater Long Valley region Six spatially discrete, successive foci of silicic magmatism 1) 3.5-2.5 Ma: Pre-caldera dacite field NW of future caldera at Long Valley 2) 2.2-0.79 Ma: >60 pre-caldera high-silica rhyolite vents centered at Glass Mtn, NE of Long Valley (total ~100 km3) 3) 0.76-0.10 Ma: Central Long Valley system 760 ka: Caldera-forming eruption of compositionally zoned rhyolitic Bishop Tuff (~600 km3) 760-650 ka: Crystal-poor Early Rhyolite on resurgent dome (total ~100 km3) 527-101 ka: Three clusters of varied Moat Rhyolites (<10 km3) 4) 160-8 ka: Trachydacite-rhyodacite and peripheral mafic vents at Mammoth Mountain, W of Long Valley 5) <50 ka: Mono-Inyo craters, NW of Long Valley Hildreth, 2004, Fig. 6 6) <14 ka: Mono Lake, N of Mono Craters and NW of Long Valley Each silicic focus driven by locally concentrated basaltic intrusion in deep crust, a response to Extensional unloading and decompression melting of mantle Hildreth, 2004 What is the state of the Long Valley magma reservoir? Earlier view, with large contemporary magma chamber Near death, based on evidence that 100-fold decline in eruption rate after 650 ka Lack of crystal-poor rhyolite since 300 ka Limited volumes of moat rhyolite (most of it crystal-rich) Absence of post-caldera mafic volcanism inside the structural caldera or to the N and S Paleo-hydrothermal system in resurgent dome died out soon after 300 ka Hydrothermal underflow at Hot Creek, active since 40 ka, is sourced from outside the ring fault zone to the west, possibly driven by intrusion of hybrid dacites Low thermal gradients inside the caldera P. Kresan collection More recent view, with chamber largely crystallized System has been dying for last 300 ka A nearly crystalline pluton Bachmann and Bergantz, 2008, Fig. 4 Hildreth, 2004 Long Valley caldera (Bishop Tuff), CA Open-system behavior Six spatially discrete, successive foci of silicic magmatism in greater Long Valley Wilson and Hildreth, 2007, Fig. 18 Abandoned silicic reservoirs have crystallized Mammoth Mtn and Mono-Inyo chains invaded caldera’s moat But outside its western ring-fracture zone Lifespan of magmatic system Volcanic field: 3.5 my One chamber: Vigorous for ~0.5 my (~790-~288 ka) Central Long Valley system near death Volume, age of cataclysmic eruptions 750 km3, 0.76 Ma Repose time for cataclysmic eruptions ??? Hildreth, 2004, Fig. 6 Yellowstone, WY-ID-MT Open-system behavior Lifespan of magmatic systems Along hotspot trace: >16.6 m.y. Volcanic field: >2.2 m.y. One chamber: >2.2? m.y. Volume of cataclysmic eruptions 1,000 km3, 0.64 Ma 280 km3, 1.29 Ma 2,500 km3, 2.06 Ma Repose times for cataclysmic eruptions 0.774, 0.646 m.y. Lowenstern and Hurwitz, 2008, Fig. 3 Pantelleria, Strait of Sicily, Italy Open-system behavior Lifespan of magmatic system One chamber: >0.093 my? Volcanic field: >0.25 my Volume of cataclysmic eruptions--small 3 km3, 55 ka ? km3, 93 ka Repose times for cataclysmic eruptions 38 ka Summary: Calderas Calderas no longer solely a morphological term Volcano-tectonic depressions Some are structurally-controlled collapse features; others are calderas modified by faulting Variety of subsidence mechanisms Collapse Syneruptive failure of caldera walls produces megabreccias Collapse of structural walls topographic wall beyond structural margin of caldera Resurgence Occurs shortly after collapse of caldera Produces structural doming Related to rapid re-establishment of isostatic equilibrium Caldera cycles and repose times Diameters and area-volume relationships Large calderas: ~10% of chamber volume is erupted Examples of calderas and their associated ignimbrites Long Valley, CA Valles-Toledo, NM Yellowstone, WY-ID-MT San Juan volcanic field, including Creede caldera, CO Southwest Nevada volcanic field, with numerous calderas Sierra La Primavera, Mexico Pantelleria, Strait of Sicily, Italy Crater Lake, OR Mount Pinatubo, Philippines Lecture 11: Dynamics of magma chambers Yellowstone hot spot controversy Follow up to caldera discussion Dynamics of magma chambers Emphasis on large silicic systems Open-system behavior Basaltic underplating and shadow zones for basaltic volcanism Construction of batholithic magma chmabers and their geometry Lifespans of magmatic systems Repose times for caldera-forming eruptions Evolution of silicic magma chambers with time Crystal mush model Successor to thermogravitational diffusion model As developed for greater Long Valley region Yellowstone, WY-ID-MT: Present site of a subcontinental hotspot track? Perkins and Nash, 2002, Fig. 1 Yellowstone hotspot: History Yellowstone is the archetype of a continental hotspot First interpreted as being related to a convective plume from the deep mantle by Morgan (1972) “Hotspot” coined by Wilson (1963) Linear chains of oceanic volcanoes that appear to record the motions of the overlying plates Long-lived centers of enhanced magma production Commonly associated with voluminous surface volcanism that has propagated linearly w.r.t. plates on which they occur Christiansen et al., 2002 Grotzinger and Jordan, 2010, Fig. 12.27 The Yellowstone controversy What is the nature of the “melting anomaly” (hotspot)? Is it the near-surface expression of convective plumes that rise through the entire mantle from a thermal boundary layer at the base of the mantle? Are there alternative explanations that relate to an upper mantle origin? Ignimbrites that occur regionally Perkins and Nash, 2002, Table 1 Ash-fall tuffs that occur regionally Perkins and Nash, 2002, Table 3 Time-space plot of volcanic rocks related to Yellowstone “hotspot track” Perkins and Nash, 2002, Fig. 2 Yellowstone, WY-ID-MT: Present site of a subcontinental hotspot track? Christiansen et al., 2002, Fig. 1 The Yellowstone controversy Evidence cited by Christiansen et al. (2002) to argue that melting anomaly is not consistent with a simple deep-mantle plume hypothesis: There is a Miocene to Recent northwestward migration of magmatism, concurrent with the northeastward “hotspot” migration Claim: Seismic imaging provides no evidence for, and several contraindications of, Vertically extensive plume-like structure beneath Yellowstone, or Broad trailing plume head beneath eastern Snake R. Plain And some evidence that it does not. Claim: No evidence that the system extends deeper than ~200 km Christiansen et al., 2002, Fig. 1 (detail) Alternative model Christiansen et al. (2002) argue that The melting anomaly is a product of feedback between upper-mantle convection and regional lithospheric tectonics Stay tuned for new developments! Dynamics of magma chambers Emphasis here On large silicic systems Open-system behavior Basaltic underplating Significance of shadow zones for basaltic volcanism Construction of batholith-sized magma chambers Relationship to why batholiths have shoulders or floors Lifespans of magmatic systems Repose times for caldera-forming eruptions Evolution of silicic magma chambers with time Crystallization stages of magma chambers Crystal mush model As developed for greater Long Valley region Successor to thermogravitational diffusion model Magma supply vs. percolation rate model VIII VI Magmatism Fundamentally basaltic IV Note that “magma supply” to the lithosphere and eruption rate V May differ by 1 to 3 orders of magnitude * = rapid transient injection of basalt at restricted crustal levels Resulting in generation and separation of rhyolite but few intermediates VII Ia IIa IIb Hildreth, 1981, Fig. 16 Ib III Underplating and shadow zone Wohletz and Heiken, 1992, Fig. 1.10, adapted from Hildreth, 1981, Fig. 15 Dynamic model of large, mature Cordilleran magmatic system Based on Bishop Tuff, Long Valley caldera, CA, and Yellowstone, WY System can be reestablished over time Different volcanic centers in same volcanic field Multiple eruptions in same caldera complex, with repose times correlated directly to volumes of eruptions Hildreth, 1981, Fig. 12 Why batholiths have shoulders or floors Field and geophysical evidence that batholiths have flat bottoms in mid-crust Luhr Hill Granite phase of Yerington batholith has a shoulder or floor at Luhr Hill Silicic melt arrived in shallow crust only in small batches, building reservoir incrementally Large silicic magma bodies never rose en masse from below Yellowstone: Lowenstern and Hurwitz, 2008, Fig. 3 Magma chamber, once formed, may migrate more locally en masse During resurgence following caldera eruption Yerington: Luhr Hill granite (PG) intrudes Bear quartz monzonite (QM) Yerington batholith: Dilles, 1987, Fig. 3 Another view of batholithic magma chamber construction Based on study of characteristics of accessory zircon from tuffs from a Yellowstone-type chamber Heise caldera cycle, eastern Idaho Kilgore Tuff, 1800 km3, 4.5 Ma Location of Heise caldera and Yellowstone Plateau; Yellowstone Pwave seismic tomography High-precision U-Pb geochronology By isotope dilution-thermal ionization mass spectrometry (ID-TIMS) Precision 0.1% = several thousand years Coupled with oxygen and bulk crystal Hf isotope analyses of the same zircon crystals By in situ secondary ion mass spectrometry (SIMS) U-Pb crystallization age 4.4876 ± 0.0023 Ma Indistinguishable from 40Ar/39Ar sanidine ages Wotzlaw et al., 2014, Fig. 1 Results Zircons in Kilgore Tuff display Significant intercrystalline and intracrystalline O isotopic heterogeneity Vast majority are 18O depleted Interpretation Zircons crystallized from isotopically distinct magma batches That were generated by remelting of subcaldera silicic rocks previously altered to low-δ18O meteoric-hydrothermal fluids Magma batches were assembled and homogenized into single big chamber because zircons have indistinguishable U-Pb ages This requires that Shallow crustal melting, assembly of isolated batches into unified magma reservoir, homogenization, and eruption occurred extremely rapidly Within resolution of dating: thousands to tens of thousands of years Connection of magma batches vertically distributed over several km would cause a substantial increase of buoyancy overpressure Providing an eruption trigger mechanism That is a direct consequence of the reservoir assembly process Wotzlaw et al., 2014 Model for assembly over time: t1 Middle crustal basaltic sill complex forms hot, dry differentiates (low-Si rhyolites or dacites) By fractionation and crustal melting Their intrusion triggers remelting of Variably hydrothermally altered subcaldera tuffs and intrusions Colors of magma batches Refer to equilibrium oxygen isotopic composition Note: Vertical exaggeration of ~10:1 Wotzlaw et al., 2014, Fig. 3a Model for assembly over time: t2, 3 Progressive amalgamation of magma batches by Thermal erosion/reactive bulk assimilation and/or Mechanical failure of intra-reservoir crust and Convective homogenization prior to eruption Wotzlaw et al., 2014, Fig. 3b, c Model for assembly over time: t13 Colors of magma batches Refer to equilibrium oxygen isotopic composition Crystal of same color as host magma Are in oxygen isotopic equilibrium Note increasing zircon oxygen isotopic diversity with progressive assembly prior to eruption While other phenocrysts equilibrate Wotzlaw et al., 2014, Figs. 1B, 3 Lifespans of magmatic systems Yellowstone Along hotspot trace: >16.6 my Volcanic field: >2.2 my One chamber: >2.2? my Valles Volcanic field: >3.6 m.y. (perhaps 16.5 m.y.) One chamber: >1.79 m.y. Long Valley Volcanic field 3.5 m.y. One chamber: vigorous for ~0.5 m.y. Pantelleria One chamber: >0.093 my? Volcanic field: >0.25 my Repose times: Proportional to volume of magma chamber Yellowstone (volumes 2500, 280, 1,000 km3) 0.646 m.y. Henry’s Fork (Island Park) to Yellowstone 0.774 m.y. Big Bend Ridge to Henry’s Fork (Island Park) Valles (volumes 300, 300 km3) 0.38 m.y. Toledo to Valles Pantelleria (volumes ?, 3 km3) 0.038 m.y. La Vecchia to Cinque Denti Crystallization stages of magma chambers (which can be reversed) A. Low-crystallinity phase (<45 vol% xls)—Magma Convection B. Medium-crystallinity phase (~45-60 vol% xls)—Mush Absence of convection High permeability provides favorable window for xl-melt separation C. High-crystallinity phase (>60 vol% xls)—Rigid sponge Permeability too low for efficient extraction of high-viscosity melt Bachmann and Bergantz, 2008, Fig. 6 Context for crystal mush model Voluminous eruptions of zoned rhyolite like the Bishop Tuff Generally start with high-silica magma poor in crystals Tap progressively or stepwise into less evolved magma richer in crystals Many such eruptions terminate with withdrawal of extremely crystal-rich mush (>50 % phenocrysts) General inference follows Large rhyolite reservoirs do not crystallize from top down Instead, water-enriched, highly evolved melt accumulates at roof Hildreth, 2004, Fig. 7 (detail) Water-rich melt progressively escapes from the mushy, qtz-fsp (anhydrous) cumulate plutonic reservoir below Water enrichment Inhibits crystallization of high-silica roof zone, in spite of its lower temperature Escape of water-rich melt possible because viscosity range of hydrous rhyolitic melt at 700–800°C (Scaillet et al., 1998) is far lower than formerly supposed Crystal mush model of rhyolite melt extraction from plutonic crystal mush of intermediate hybrid composition xp = crystal-poor (0 –5%), xm = intermediate crystal content; xr = crystal-rich (15– 55%) Pre-climactic Long Valley reservoir xp: Crystal-poor (0-5%) magma xm: Intermediate crystal content magma xr: Crystal-rich (15-55%) magma Root zones Depict mush columns consisting of quartz + feldspar-rich (meltdepleted) cumulates and Migmatized protoliths extending to zones of partial melting in lower crust Rejuvenation Prior to death, systems can be rejuvenated by new arrivals of mantle basalt Hildreth, 2004, Fig. 7 New , central Long Valley focus Dying Glass Mtn focus Glass Mtn Stage, while separating two xl-poor batches Evidence for crystal mush model Many zoned ignimbrites represent volumes of high-silica melt so great that A proximal source volume of cumulate mush many kilometers thick must have underlain the segregated melt That such mush reservoirs can persist in granitoid plutons without solidifying Demonstrated by great eruptions of ‘‘monotonous intermediate’’ ignimbrite containing 40–60% crystals and volumes of 500–5000 km3 Also provides explanation for uncommon but real ‘‘reversely zoned’’ ignimbrites Eruptions that tap marginal mush first Hildreth, 2004, Fig. 7 (detail) Crystal mush model and Long Valley Evolution in time from Glass Mountain stage (bottom) to Long Valley stage (top) Characteristics of many other stages xp = crystal-poor (0 –5%), xm = intermediate crystal content; xr = crystal-rich (15– 55%) Pre-climactic Long Valley reservoir Can be envisioned Note shifting mantle-driven thermal focus From beneath Glass Mountain To beneath central Long Valley Upper-crustal reservoir of Glass Mountain Merged with dying Glass Mountain focus Which ceased being mantle-sustained and crystallized after the calderaforming eruption Elliptical shape of caldera Caused by eruption during shift merging of upper crustal reservoirs? Recall “unzipping” ring fracture Hildreth, 2004, Fig. 7 New , central Long Valley focus Dying Glass Mtn focus Glass Mtn Stage, while separating two xl-poor batches The next “supervolcano”? Singer et al., 2014, Cover photo, GSA Today Laguna del Maule, Chile Unusually large and recent concentration of silicic eruptions Encircle Laguna del Maule 230 km E of epicenter of MW 8.8 Maule earthquake of 27 Feb 2010 Since 2007, the crust at Laguna del Maule has been inflating At astonishing rate of 25 cm/yr (InSAR, GPS data) Apparent inception of magma system in late Pleistocene 350 km3 of lavas and tuffs of basaltic to rhyolitic composition erupted during Pleistocene 13 km3 of rhyolite erupted in last 20 k.y. Most recent “super-eruption” Oruanui, New Zealand 26,500 yr Singer et al., 2014, Fig. 1a Geologic map A dozen xl-poor glassy rhyolitic lava flows Erupted in Holocene (<10.k.y.) Greatest concentration of post-glacial rhyolite in the Andes 36 silicic lava flows comprising 6.4 km3, with a likely equivalent volume of tuff preserved in Argentina Only comparable volume of rhyolite is >4 km3 of rhyolite lava and tephra at Mono Craters, California Singer et al., 2014, Fig. 1b, adapted from Hildreth et al., 2010 Interpreted magma chamber Interpretation of magmatic system Feeding the crystal-poor rhyolitic eruptions encircling Laguna del Maule Along bent cross section of Fig. 1b Supports observations of Rapid uplift Shallow earthquakes Active intrusion of mafic magma at depth of 5 km Normal faulting and geodetic data that record extension Singer et al., 2014, Fig. 6, adapted from Hildreth, 2004 Summary One can argue that all magmatism is “fundamentally basaltic” Magmatic system are dynamic, open systems for much of lives Batholiths have shoulders or floors Silicic chambers act as a shadow zone for basaltic magmatism Assembling quickly, leaking out top, being replenished from below Because they are built by arrival of silicic melt in small batches Large silicic magma reservoirs never rose en masse from below Once formed, magma may migrate en masse locally and homogenize Repose times are roughly proportional to Lifespans of magmatic systems Crystal mush model Next time: Silicic magmas interacting with water: Subaqueous and subglacial eruptions The size of the magma chamber (caldera), size of major eruptions Volcanic fields typically ~10 m.y.; single chambers, at most a few m.y. Means of assembly and homogenization could trigger major eruption As developed for greater Long Valley region Successor to thermogravitational diffusion model Crystallization stages of magma chambers (which can be reversed)