Cenozoic Tectonics, Magmatism, and Stratigraphy of the Snake River
Plain–Yellowstone Region and Adjacent Areas themed issue
00969 1st pages / page 1 of 27
Rhyolites—Hard to produce, but easy to recycle and sequester:
Integrating microgeochemical observations and numerical models
I.N. Bindeman1 and A.G. Simakin2
1
Department of Geological Sciences, 1272 University of Oregon, Eugene, Oregon 97403, USA
Institute of the Physics of the Earth, Bolshaya Gruzinskaya 10, Moscow, Russia
2
ABSTRACT
Recent discoveries of isotopically diverse
minerals, i.e., zircons, quartz, and feldspars,
in large-volume ignimbrites and smaller
lavas from the Snake River Plain (SRP;
Idaho, USA), Iceland, Kamchatka Peninsula,
and other environments suggest that this
phenomenon characterizes many silicic units
studied by in situ methods. This observation
leads to the need for new models of silicic
magma petrogenesis that involve double or
triple recycling of zircon-saturated rocks.
Initial partial melts are produced in small
quantities in which zircons and other minerals undergo solution reprecipitation and
inherit isotopic signatures of the immediate
environment of the host magma batch. Next,
isotopically diverse polythermal magma
batches with inherited crystals merge
together into larger volume magma bodies,
where they mix and then erupt. Concave-up
and polymodal crystal size distributions of
zircons and quartz observed in large-volume
ignimbrites may be explained by two or three
episodes of solution and reprecipitation.
Hafnium isotope diversity in zircons demonstrates variable mixing of crustal melts and
mantle-derived silicic differentiates. The low
δ18O values of magmas with δ18O-diverse
zircons indicate that magma generation happens by remelting of variably hydrothermally
altered, and thus diverse in δ18O, protoliths
from which the host magma batch, minute or
voluminous, inherited low-δ18O values. This
also indicates that the processes that generate
zircon diversity happen at shallow depths of
a few kilometers, where meteoric water can
circulate at large water/rock ratios to imprint
low δ18O values on the protolith. We further
review newly emerging isotopic evidence of
diverse zircons and their appearance at the
end of the magmatic evolution of many longlived large-volume silicic centers in the SRP
and elsewhere, evidence indicating that the
genesis of rhyolites by recycling their sometimes hydrothermally altered subsolidus
predecessors may be a common evolutionary trend for many rhyolites worldwide,
especially in hotspot and rift environments
with high magma and heat fluxes. Next, we
use thermomechanical finite element modeling of rhyolite genesis and to explain (1) the
formation of magma batches in stress fields
by dike capture or deflection as a function of
underpressurization and overpressurization,
respectively; (2) the merging of neighboring magma batches together via four related
mechanisms: melting through the screen
rock and melt zone expansion, brittle failure of a separating screen of rocks, buoyant
merging of magmas, and explosive merging
by an overpressurized interstitial fluid phase
(heated meteoric water); and (3) mixing
time scales and their efficacies on extended
horizontal scales, as expressed by marker
method particle tracking. The envisioned
advective thermomechanical mechanisms of
magma segregation in the upper crust may
characterize periods of increased basaltic
output from the mantle, leading to increased
silicic melt production, but may also serve
as analogues for magma chambers made of
dispersed magma batches. Although not the
focus of this work, dispersed magma batches
may be stable in the long term, but their
coalescence creates ephemeral, short-lived
eruptable magma bodies that erupt nearly
completely.
INTRODUCTION
The origin of silicic magmas is a traditional
topic of igneous petrology that has fundamental significance for the origin of the first continental crust (Hutton, 1794). It was recognized
early that fractional crystallization of basaltic
magmas could only generate small, difficult to
extract, interstitial silicic liquids (Bowen, 1928;
Tuttle and Bowen, 1958), recently advocated for
the Snake River Plain and Yellowstone (western USA) plume by McCurry et al. (2008) and
Whitaker et al. (2008). An alternative mechanism, partial melting of the lower basaltic crust
(Leeman et al., 1992; Dufek and Bergantz,
2005) has problems related to heat availability,
conduction, and the large amounts of basaltic
magma required, often producing excessive
amounts of associated cumulates (e.g., Beard
and Lofgren, 1991; Petford and Gallagher,
2001; Ducea, 2001). Models involving filter
pressing, or reactivations of melts from cumulate piles and static or dynamic mushes, have
advantages (e.g., conservation of latent heat and
space), but also significant limitations on the
physical mechanisms of melt extraction and the
slow rates of these processes (Wickham, 1987;
Philpotts et al., 1996; Bachmann and Bergantz,
2004; Huber et al., 2009; Streck and Grunder,
2008; Bindeman et al., 2008b).
Many factors that militate against the production of silicic magmas, including slow extraction, the room problem, and heat, are eliminated
by a hypothesis of simple recycling of originally
silicic or intermediate protoliths, which may
have been formed by any or all of these slow and
incremental processes. If rhyolites are thought
to be produced by such recycling, estimates of
the total amount of silicic magma generated in
a given magmatically active area, the amount of
basalt required, and the amount of heat required
become significantly smaller.
A large-volume silicic eruption provides a
snapshot of an instant in the temporal evolution of a magma body; repeated eruptions over
millions of years give insights into petrogenesis
at depth (Lipman, 2007). However, the debate
over whether plutons and ignimbrites provide
comparable records of magma genesis is ongoing, especially with reference to arc and collisional environments, where batholiths may
be assembled over millions of years (Coleman
et al., 2004; Glazner et al., 2004; de Silva and
Gosnold, 2007) and magma bodies are kept
near solidus.
Geosphere; October 2014; v. 10; no. 5; p. 1–27; doi:10.1130/GES00969.1; 19 figures; 2 tables; 7 supplemental files.
Received 25 July 2013 ♦ Revision received 3 June 2014 ♦ Accepted 3 July 2014 ♦ Published online XX Month 2014
For permission to copy, contact [email protected]
© 2014 Geological Society of America
1
00969 1st pages / page 2 of 27
Bindeman and Simakin
The study of the plume-related Snake River
Plain (SRP; Idaho, USA) silicic volcanism
(Fig. 1) has an advantage over other areas of
voluminous silicic volcanism because the North
American plate is moving at a speed of 2–3
cm/yr over the Yellowstone hot magmatic zone
(e.g., the hotspot and/or propagating rift; Pierce
and Morgan, 2009), creating a conveyor-belt
record of silicic magma genesis as opposed to
more stationary examples in convergent margins
and subduction arcs, where volcanism occurs in
a localized area and younger magmatic episodes
obscure their predecessors. At least six major
nested caldera clusters have developed over the
past ~15 m.y. (Fig. 1); current activity is centered at Yellowstone. Each of the volcanic centers is the site of several major caldera-forming
eruptions and overlapping calderas, although
individual caldera boundaries are often obscured
in the older 15–11 Ma centers.
Recent work (Figs. 1 and 2) has demonstrated
that Picabo, Heise, and Yellowstone, the three
most recent caldera clusters, share common
features of magmatic evolution for the past 10.5
m.y. Eruption of the first major tuff occurred
after an ~2 m.y. hiatus from the last major eruption in the previous caldera center. The initial
Huckleberry Ridge Tuff of Yellowstone erupted
at 2.1 Ma, after the 4.4 Ma Kilgore tuff, which
finished activity at the Heise center to the southwest, and the oldest Heise tuff, the 6.6 Ma Blacktail Creek tuff, erupted ~2 m.y. after the eruption
of the 8.3 Ma Pocatello tuff, the youngest tuff
of the Picabo center (Fig. 1) (e.g., Christiansen et al., 2007; Morgan and McIntosh, 2005;
Drew et al., 2013). These relationships become
Syn and pre CRB silicic centers
Figure 1. Map of western North America showing essential magmatic and tectonic features discussed in the text
(after Drew et al., 2013; Pierce and Morgan, 2009; Coble and Mahood 2012). The yellow dashed circle is the
inferred infringement location of the Columbia River Basalt (CRB) plume head that resulted in basaltic volcanism
such as the Steens and Columbia River Basalt Groups and scattered 16.7–15 Ma silicic magmatism (Coble and
Mahood, 2012). Subsequent silicic volcanism, from ca. 13 Ma to present, along the Snake River Plain (SRP) was
more organized and occurred in northeast-migrating caldera clusters resulting from the movement of the North
American plate over the hotspot plume at a rate of 2 cm/yr (e.g., Pierce and Morgan, 2009); there were also contributions from rifting and coeval Basin and Range extension. There was a 1–2 m.y. gap before initiation of volcanic
activity after the establishment of the new magma chambers with mantle input from the underlying mantle plume.
Basaltification of the crust in the wake of the plume made crust infertile for further melting, resulting in a decrease
of silicic volcanism; however, basaltic volcanism continued throughout the SRP after direct passage of the plume.
Occurrences of recycled rhyolites with low δ18Omelt values and diverse zircons and other phenocrysts are found
throughout the SRP and in some early post-CRB centers (underlined). The 0.704 (dotted black) and 0.706 87Sr/ 86Sr
(dashed black) lines define the transition from accreted Mesozoic and Paleozoic terrain to the Precambrian terrains of North America (Fleck and Criss, 1985).
2
Geosphere, October 2014
00969 1st pages / page 3 of 27
Rhyolites—hard to produce but easy to recycle
YELLOWSTONE
(2.1–0.07 Ma)
Quartz
–3‰
Plagioclase
Clinopyroxene
Zircon
Obsidian
Calculated melt δ18O
Individual Zircon ranges (Fig. 3)
A
–1‰
HEISE
(6.6–3.9 Ma)
B
PICABO
(10.5–6.6 Ma)
Idavada
TAV
2.5 Mile
HSpr
TAF
LRS
SP
Chokecherry
Jim Sage
C
W Pocatello
CPT V
CPT III
CPT XIII CPT XV
CPT XI
–0.5‰
SP2
BRUNEAU-JARBIDGE
(12.7–8.1Ma)
CPT IX
CPT VII
–
–
D
CPT XII
Figure 2. Temporal evolution of melt and mineral δ18O values in four caldera complexes. (A) Yellowstone. (B) Heise. (C) Picabo.
(D) Bruneau-Jarbidge. The dashed line at 6‰ denotes the lowest possible normal δ18O rhyolite limit formed by crystal fractionation of
normal δ18O, 5.7‰ ± 0.2‰ basalt. An overall decreasing trend in δ18O in the three youngest caldera clusters is explained by remelting of
low δ18O hydrothermally altered materials after burial in calderas and rifts. Note the pervasive presence of low δ18O values in BruneauJarbidge magmatism; our new data are plotted in D (larger symbols for minerals, see Table 1 for analyses). Other data (shown with individual unit abbreviations) are compiled from Bindeman and Valley (2001), Bindeman et al. (2008b), Watts et al. (2011, 2012), Drew et al.
(2013), Wotzlaw et al. (2014), Boroughs et al. (2005), Bonnichsen et al. (2008) and Table 1; the ranges of the δ18O values of individual zircons
(blue vertical bars) are from these studies and from Cathey et al. (2011) for Bruneau-Jarbidge (see Fig. 3). The origin of zircon and other
phenocryst diversity is explained by segregation and mixing of isotopically diverse batches and is numerically modeled in this work.
Geosphere, October 2014
3
00969 1st pages / page 4 of 27
Bindeman and Simakin
more obscure in the older and less-well-exposed
central SRP centers (Fig. 1), including the Twin
Falls–Bruneau-Jarbidge centers, which also
exhibited lingering silicic magmatism as young
as 8–6 Ma (see Ellis et al., 2013, and references
therein), which were also likely influenced by
coeval rifting (Konstantinou et al., 2013), and
are now buried by Quaternary SRP basalts.
Ongoing investigation of the earliest ca. 15 Ma
volcanism before the Bruneau-Jarbidge center
(Colon et al., 2014), however, demonstrates
a gap of ~1–2 m.y. between scattered post–
Columbia River Basalt volcanism and more
focused central SRP volcanism. The ~2 m.y.
gap is sufficiently long to generate the first
voluminous silicic magma body at the beginning of each caldera cluster cycle, comparable to the Huckleberry Ridge Tuff batholith, ~5000–10,000 km3 in volume, assuming
an erupted versus residual magma ratio of
1:3. To produce these rhyolites, we therefore
favor slow incremental assembly of magma
by fractional crystallization of SRP basalt and
melting of the lower crust in a hot zone (via
sill emplacement; Annen and Sparks, 2002),
with insignificant contribution from the upper,
highly radiogenic crust, as is indicated by the
modestly radiogenic values of rhyolites (Leeman et al., 1992; McCurry and Rodgers, 2009),
or more rarely with an exclusively upper crustal
source (e.g., the Arbon Valley Tuff Member;
Drew et al., 2013). Usual or slightly elevated
hotspot basaltic intrusion rates of 0.001–0.01
km3/yr (Pierce and Morgan, 2009) are sufficient to explain the required volumes of silicic
magma to build the first magma bodies that are
normal δ18O and have 60%–88% mantle contribution to their radiogenic isotopes (Leeman
et al., 1992; McCurry and Rodgers, 2009; Nash
et al., 2006). The rest of this paper deals with
the recycling of these first magma bodies as a
chief mechanism of rhyolite petrogenesis.
GEOLOGICAL EVIDENCE FOR A
RECYCLED ORIGIN FOR MOST
SRP RHYOLITES AND MANY
RHYOLITES WORLDWIDE
The advent of microanalytical isotopic techniques, i.e., single crystal laser fluorination,
isotopic dilution-thermal ionization mass spectrometry, single zircon analyses with high precision, and ion microprobe analysis with worse
precision but superb spatial resolution, allows a
novel look at igneous petrogenesis. The rapid
and accelerating rate of discoveries of isotopically diverse zircons, quartz, feldspars, and
other minerals in large-volume ignimbrites and
smaller volume lavas worldwide suggests that
this phenomenon characterizes many silicic
4
units studied by in situ methods. We here summarize recent results of these investigations with
emphasis on the SRP.
Zircon and Quartz δ18O Diversity and
Isotope Diversity in Other Minerals
Early evidence of zircon and quartz δ18O
zoning and diversity in a few Yellowstone lavas
(Bindeman and Valley, 2001) has been observed
in more than half of units studied in the Columbia River Basalt–SRP large igneous province
(Figs. 1 and 3), and similar evidence is being
obtained for other areas (Fig. 4). This diversity
variably characterizes zircon δ18O, Hf isotopes,
and sometimes older U-Pb ages from nearly
half of studied SRP tuffs and lavas erupted
since ca. 15 Ma: Yellowstone (Bindeman et al.,
2008b; Watts et al., 2012; Vazquez et al., 2009),
Heise (Watts et al., 2011), Picabo (Fig. 3; Drew
et al., 2013), Bruneau-Jarbidge and Twin Falls
(Cathey et al., 2011), and the early post–Columbia River Basalt volcanism (Colon et al., 2014).
The diversity of the refractory mineral zircon,
the mineral with the slowest oxygen diffusion
rates, is not surprising, but diversity in δ18O
quartz crystals is also present (Fig. 3).
In the SRP and in Iceland (Figs. 3 and4) there
is common association of diverse zircon and
quartz crystals with low δ18O magmas, indicating their inheritance during remelting of buried
and hydrothermally altered volcanic rocks
and their subvolcanic units (e.g., Carley et al.,
2011). However, there are also diverse zircons
in several normal δ18O to moderately high δ18O
rhyolites of the SRP, Iceland, and Kamchatka
Peninsula (Fig. 4). This suggests that nearly all
studied low δ18O magmas worldwide, and some
nominally normal δ18O magmas, display δ18O
diverse zircons and sometimes other minerals.
The isotopic fingerprinting of the upper crust
by low δ18O meteoric waters values thus serves
as a tracer of the processes of remelting and
recycling of hydrothermally altered rhyolites.
If hydrothermal alteration happens at the seafloor using 0‰ seawater, a spread from low to
high δ18O values will result (Grimes et al., 2013;
Wanless et al., 2010; Bindeman et al., 2012),
potentially leading to high and low δ18O magma
batches. Finally, advocated processes may occur
in nature even if the precursor rocks have not
been hydrothermally altered with respect to
δ18O, particularly in the lower and middle crust.
High δ18O, Normal δ18O, and Abundant
Low δ18O Magmas in the SRP
Throughout the past decade there has been
rapid discovery of low δ18O magmas in the SRP
and worldwide (e.g., Bindeman et al., 2007,
Geosphere, October 2014
2012; Boroughs et al., 2005; Ellis et al., 2010;
Drew et al., 2013), and these rocks have gained
increasing attention as probes into upper crustal
petrogenesis. Nearly all nested caldera clusters
in the SRP have produced thousands of cubic
kilometers of rhyolites depleted in δ18O (the
volume of these magmas in the central SRP is
10,000–30,000 km3; Fig. 1; Bonnichsen et al.,
2008; Ellis et al., 2010, 2013), and individual
units are as voluminous as 1800 km3 (Morgan
and McIntosh, 2005). We advocate that the origin of these magmas requires a two- to threestep process: (1) hydrothermal alteration of the
protolith by low δ18O meteoric waters at large
water/rock ratios; (2) remelting and convective
mixing during which low and variable in δ18O
oxygen is homogenized in a newly formed
magma batch; and (3) formation of new crystals representing batch-averaged δ18O composition, and repetition of 1 and 2. Washington
State University researchers favor the idea that
step 1 significantly predates step 2 (e.g., alteration preceding central SRP volcanism occurred
during formation of the Eocene Idaho batholith; Boroughs et al., 2005; Ellis et al., 2013),
while the University of Oregon group favors
auto development of magmas from normal to
low δ18O values immediately before low δ18O
eruptions via successive caldera collapse and
rift burial (Bindeman et al., 2007; Watts et al.,
2011; Drew et al., 2013). Shallow crustal fingerprinting by low δ18O meteoric water may also
be accomplished via tectonically induced waterrock interaction during syngenetic rifting, core
complex formation, and alteration along normal
faults (Konstantinou et al., 2013; Drew et al.,
2013; Colon et al., 2014). Conditions for remelting are optimized when the rocks are brought
down closer to the heat source during crustal
burial processes of caldera collapses, rifting, or
both. Rifting and burial due to extension may
have created additional hydrogeological conditions for hydrothermal fluid circulation particularly in Iceland, where burial and alteration of
rocks were followed by remelting (Martin and
Sigmarsson, 2010; Bindeman et al., 2012; Figs.
3 and 4). It seems highly unlikely that low δ18O
magmas are produced in any other way than
by remelting of hydrothermally altered low
δ18O protoliths (for a review of old models, see
Taylor, 1974, 1986).
The remelting process happens in shallow
depths available to meteoric water percolation,
thus requiring a shallow heat source to generate
thousands of cubic kilometers of magma in the
upper crust. It further reaffirms, however, that
even if remelting occurs effectively in the shallow crust, magmatic recycling in the lower and
middle crust should be even more prevalent,
given higher temperatures and melt-segregation
00969 1st pages / page 5 of 27
Rhyolites—hard to produce but easy to recycle
Age, Ma
10.4
9.1
8.1
7.9 7.1 6.6
7
8
6
7
5
6
4
5
3
2
1
A
0
δ18O(Zircon), ‰, SMOW
δ18O(Zircon), ‰, SMOW
Picabo 10.4–6.6Ma
HRT
2.1 A B C 1.8
Age, Ma
1.3
0.6–0.35
0.62
Yellowstone 6.6–4.4Ma
LCT
AB
0.25
0.07
HRT cycle
MFT
LCT
UBM
Qz
CPM
4
3
2
1
0
Modern Mantle
0
–1
–2
C
–10
εHf(0)
–3
–20
7
Age, Ma
4.8 4.4
–40
Archean Crust
B
6
5
LostRiverSinks
Blacktail
Kilgore
post-Kilgore
4
3
2
1
0
–1
–2
–14
3
εHf
–12
–10
–8
4
δ18 O
Picabo
δ18O(Zircon), ‰, SMOW
Heise 6.6–4.4Ma
–30
–50
6.6
5
6
D
Figure 3. Individual zircon δ18O and εHf diversity studied by in situ methods across the Snake River Plain (SRP) eruptive centers. (A) Picabo
zircon δ18O. (B) Picabo εHf . (C) Yellowstone zircon δ18O. (D) Heise zircon δ18O. SMOW—standard mean ocean water. The range of normal
δ18O zircons in equilibrium with mantle-derived magmas (5.3‰ ± 0.3‰; Valley, 2003) is shown by horizontal gray fields. Pink horizontal
bars show equilibrium with melt zircon values (taken as Δ18Omelt-zircn = 1.8‰) for each sample. Black vertical bars denote average ±2σ error
on individual zircon δ18O analysis. Single crystal quartz analyses are by laser fluorination (Bindeman and Valley, 2001; error bars are
comparable to symbol size), and are only shown for units that demonstrate statistical diversity in δ18OQuartz values. Most zircons and some
quartz are in disequilibrium with melt and are on both high and low sides of equilibrium δ18O and εHf values; there are no zircons or quartz
values lower than 0‰ to –3‰, although hydrothermal alteration by local meteoric waters is capable of generating –14‰ to –18‰ values.
The tremendous diversity of εHf data for Picabo and Heise (Blacktail Creek tuff) is shown in B and the inset in D; this range in values can
result from pure melting of the low εHf Archean crust (orange field) or from silicic differentiation of basalts from the modern mantle. The
U-Pb ages of these zircons are close to the eruption age with the exception of one Paleoproterozoic xenocryst shown by larger symbol in B.
See text and Figure 5 for explanation. HRT, MFT and LCT are caldera forming tuffs of Yellowstone: Huckleberry Ridge, Mesa Falls and
Lava Creek. UBM and CPM are post-LCT Upper Basin Member and Central Plateau Member rhyolitic lavas.
Geosphere, October 2014
5
00969 1st pages / page 6 of 27
Bindeman and Simakin
Isotopically diverse zircons from major ignimbrites around the world
A
SW Nevada
6
B
C
Iceland
Kamchatka
7
6
5
6
5
Karymshina, 1.78Ma
Hekla, A.D. 1158
Laufafel,
Pleistocene
4
3
Pleistocene
Askja, A.D. 1875
2
δ18O(Zircon), ‰, SMOW
δ18O(Zircon), ‰, SMOW
δ18O(Zircon), ‰, SMOW
Qz:
5
Pauzhetka, 0.45Ma
4
4
1
±2σ
3
3
0
E
SW Colorado VF
Qz: 8.13‰, Mt: 2.26‰
T‚ Δ(Qz-Mt)=760°C
6
5
cores
rims
±2σ
Toba, Indonesia
Youngest Toba, Early, sample A102
Fish Canyon, Early, BFC125
δ18O(Zircon), ‰, SMOW
δ18O(Zircon), ‰, SMOW
D
±2σ
±2σ
cores
rims
7
Qz: 9.21‰, Mt: 3.22‰
T‚ Δ(Qz-Mt)=752°C
Zircon in oxygen isotopic
equlibrium with melt
Older zircon cores
±2σ
6
cores
rims
Figure 4. Zircon δ18O diversity in large- and small-volume ignimbrites from different tectonic settings around the world. SMOW—standard
mean ocean water. Zircons from units in panels A–C were dated, and most returned U-Pb ages similar to Ar-Ar eruption ages, with the
exception of zircons shown by larger symbols, which are statistically older by as much as 1 m.y. Core and rim analyses in A, D, and E are
shown; B and C are for zircon core only. Pink horizontal bars show computed zircon δ18O values in equilibrium with host melt and major
phenocrysts. Zircon diversity and corresponding Δ18O (melt-zircon) disequilibria range from equilibrium (Fish Canyon Tuff) and modest
(Toba) to severe disequilibria (all other); in most cases, zircons are heavier than is required by equilibrium with host melt (pink bar). In
the case of Hekla, Kamchatka, and Toba, cores are both heavier and lighter than equilibrium. Quoted precision shown as vertical bars
on each graph reflects uncertainty on standards run together with the unknowns and thus denote average ±2σ errors (95% confidence) on
individual zircon analyses. Mixing of both high- and low-δ18O magma batches are required to explain zircon diversity, see text and Figure 5
for explanation. Gray fields denote normal δ18O zircon values in equilibrium with mantle-derived silicic differentiate melt (5.3 ± 0.3‰;
Valley, 2003). Data sources: (A) Bindeman et al. (2006), also note quartz δ18O diversity; (B) Bindeman et al. (2012) and Gurenko et al.
(2013); (C), (D), and (E) new data from this study, Table 1, and the Supplemental Files. Data were collected in: (A) Wisconsin; (B) Nancy;
(C) UCLA; (D) and (E) Alberta, on Cameca 1270-1280HR large radius ion microprobes. Note that Toba rims are in equilibrium with the
melt. but cores spread a ~1‰ range, analytically resolvable due to superb precision by the University of Alberta Cameca 1280HR instrument (Richard Stern, analyst). Sample courtesy: BFC125–O. Bachmann, A102 (C. Chesner, see Bindeman, 2003, for crystal size distribution of quartz and zircon in this sample).
6
Geosphere, October 2014
00969 1st pages / page 7 of 27
Rhyolites—hard to produce but easy to recycle
In situ radiogenic isotope analyses provide
further insights. The Pb and Sr isotope diversity of individual feldspar crystals within single
units (Watts et al., 2012) further fingerprint the
isotopic sources of Yellowstone lavas. Diverse
Sr and Pb isotopes in individual feldspars are a
well-known phenomenon worldwide (Davidson
et al., 2001), suggesting that many crystals are
xenocrysts and antecrysts.
The Hf isotopes within individual zircons
record crust/mantle melt proportions of each
magma batch from which the zircons crystallized (Figs. 3 and 5). In two extreme cases when
a silicic differentiate of the SRP crystallizes
high (mantle like) εHf zircon mixes with a pure
partial melt of silicic crust (ultralow εHf) zircons,
the resulting final magma may have extremely
diverse εHf zircons (see Fig. 3B). While the
evidence for diverse Hf isotopes in zircons is
growing, preliminary data suggest that both
homogeneity and heterogeneity of Hf isotopes
in zircons is common in SRP magmas (Picabo:
Drew et al., 2013; Yellowstone: Stelten et al.,
2013). Nonetheless, sporadic ultralow-Hf zircons (Fig. 3; down to –47 εHf units; Drew et al.,
2013) suggest addition of melts of Archean crust
during the late stages of magma segregation; in
most cases ultralow εHf zircons have reset U-Pb
ages, suggesting that remelts of the Archean
crust appeared and then crystallized new zircons
before mixing with other magmas.
Isotopically Diverse Crystalline Cargo:
Inheriting Isotopic Variations from
Protolith
Gaining insights into the range of isotopic
values and their spatial distribution within a
protolith undergoing melting are important for
our future modeling of the efficiency of melting
and mixing and the mechanisms of incorporation of these parameters into zircon and other
minerals (Fig. 4).
FAST, EASY, <<1 m.y.
Zircon T-t Destinies:
Temperature
UPPER CRUST:
Sweet spot, partial eutectic
melting of silicic lithologies,
batch assembly, mixing, eruption,
δ18O
zircons. Eutectic melting will
cross diverse in δ18O and ε(Hf) areas,
further rejuvenate U-Pb ages
a
Zrc out
b solidus
c
time
Temperature
Hydrothermal preconditioning of
intruded silicic rocks or batholith tops
generates diverse and low in δ18O areas
solidus
time
UPPER TO MIDDLE CRUST:
Silicic sills accumulation or
batholith formation,
future multiple recycling of
this batholith, further sequestration
of silicic magmas
Mixing generates average ε(Hf),
rejuvenated U-Pb age, normal δ18O
silicic rocks; Intrusion to the upper to
middle crust
b
c
a
Temperature
Insights from Zircon Hf Isotopes and
Whole-Rock Sr-Nd-Hf-Pb Isotopes
Multi-stage silicic magma recycling and sequestration to the upper crust
SLOW, HARD, >1–2 m.y.
efficiency (Annen et al., 2006; Dufek and Bergantz, 2005). Furthermore, this process must be
relatively rapid, faster than mineral-diffusive
and recrystallization time scales or currently
the most precise geochronologic method,
ID-TIMS U-Pb dating (102-104 yr; Wotzlaw
et al., 2014; see also the following and Bindeman, 2008). Both of these statements contradict
the currently prevailing view that silicic magma
are generated slowly (because it takes effort;
see above), and in the lower crust (where it is
hotter), or via filter-pressing extraction from the
mushy batholiths.
a
Zrc out
b solidus
c
time
LOWER TO MIDDLE CRUST:
segregate silicic melts with blended
ε(Hf) and δ18O values
Figure 5. Model of multi-stage vertical delegated sequestration of silicic, zircon-saturated
magmas leads to multiple recycling of zircons, blending together their oxygen and Hf isotopic values and rejuvenating U-Pb ages. Slow initial segregation of small volume melts
from mafic protoliths and differentiates of basalts is contrasted by the rapid process of
remelting of already segregated silicic materials containing diverse zircons. Different potential zircon temperature-time destinies are shown (e.g., A, B, C; see text).
Ranges of δ18O and εHf Variations in Protolith
Ranges. Given the –14‰ to –19‰ δ18O
value of meteoric water, it is theoretically possible to generate a volume of rock (or a vein)
with comparable δ18O values if the temperature
of exchange were >500 °C, meteoric water was
isotopically unshifted at these depths and temperature, the water/rock ratio was high, and/or
the isotope exchange was 100%. Investigation
of hydrothermally altered intracaldera tuffs and
ore deposits around the world suggests that such
conditions are rarely realized in nature. Isotopic
ranges in and around shallow plutons, subvol-
Geosphere, October 2014
canic intrusions (e.g., Taylor, 1974, 1986), and
intracaldera tuffs (McConnell et al., 1997; John
et al., 2011) demonstrate higher and narrower
δ18O ranges (–2‰ to +7‰). Thus, the lack of
observed –18‰ to –14‰ values in zircons and
quartz (Fig. 3) is partly a reflection of incomplete water-rock isotope exchange of their
protolith, shifted waters, and a lower temperature of hydrothermal alteration. Another argument is purely hydrologic; in order to expect
a high water/rock ratio, high porosity and
permeability (open cracks) are required, and
this is best achieved above the brittle-ductile
7
00969 1st pages / page 8 of 27
Bindeman and Simakin
transition (e.g., Norton et al., 1984). For silicic
volcanic rocks these temperatures are likely to
be below ~400–450 °C (Simpson, 1985). Given
Δ18Orock-water fractionation (e.g., Cole and Charraborthy, 2001), rocks are several permil higher in
δ18O than equilibrium water at ~400 °C, further
reducing the lowest possible δ18O rock value in
hydrothermally altered rock.
Spatial δ18O distribution. Does the protolith
typically contain a frequently spaced, centimeter- to meter-scale network of isotopic variations such as ultralow δ18O veins in a normal
δ18O matrix, or ultralow εHf Archean crustal
slivers in high εHf matrix? The isotopic variations that we observe in zircons and other minerals will then have a highly localized (centimeter to meter scale) origin. Partial melting of such
areas will generate lenses of low-δ18O melt as is
observed inside granodiorite blocks of the climactic Mt. Mazama eruption (e.g. Bacon et al.,
1989) or in remelted Icelandic crustal xenoliths
(Gurenko et al., 2013). If such high-amplitude
variations span across wide regions undergoing
melting (e.g., 8‰ δ18O variation over >40 km
for the Kilgore tuff; Watts et al., 2011), this will
simplify not only the efficiency of subsequent
mixing, but may also help with mass and heat
balance problems of extraction of small-volume, zircon-saturated melt batches with diverse
zircons. This extraction may be facilitated by
dikes and sills generated by external tectonic
forces (e.g., Norton et al., 1984). Natural evidence, however, suggests that high-amplitude
variations with limited isotopic magnitudes of
several permil are also accompanied by larger
scale (10–103 m) variations across greater isotopic magnitudes, but the detailed structure of
this variation still needs to be studied by isotope
mapping. Mapping of bullseye alteration patterns over plutons (e.g., Taylor, 1974, 1986),
remote sensing of alteration products, and isotopic investigation of alteration in intracaldera tuffs
(McConnell et al., 1997; Livo et al., 2007; John
et al., 2011) provide concentric patterns of δ18O
variation (10–103 m scale). Moreover, numerical modeling of meteoric fluid flow through
porous media in one dimension (Bowman et al.,
1994) and two dimensions (Norton and Taylor,
1979) suggests predominance of continuous
isotopic fronts with diffusive-hydraulic dispersion around fast pathways, isotopically shifted
waters, and overlapping, time-integrated fronts
leading to laterally smoothed δ18O distribution
on a scale of meters to kilometers (Hayba and
Ingebritsen, 1997).
Isotopic Reequilibration During Remelting:
Lessons from Skaergaard
Figures 3 and 4 demonstrate the large isotopic diversity of zircons; however, among thou-
8
sands of zircons studied by in situ methods for
U-Pb age, only a few older xenocrysts survive.
This suggests that any partial melts of older
crust were heated above zircon saturation (e.g.,
Fig. 5). In all of the subsequent melting history,
as shown in Figure 6, all processes lead to averaging out of isotopic differences, causing solution reprecipitation of zircons and other minerals. Prolonged residence leads to oxygen isotope
equilibrium between coexisting minerals, and
further homogenization of hydrothermally
induced heterogeneity.
An important insight into the partial remelting of hydrothermally altered rocks with and
without zircons is provided by the Skaergaard
intrusion in East Greenland (Taylor and Forester, 1979; Norton and Taylor, 1979; Bindeman
et al., 2008a). Following crystallization and
hydrothermal alteration of the Skaergaard intrusion, the mafic Basistoppen sill had intruded
100 m above the hydrothermally altered siliceous Sandwich Horizon. The melting of Sandwich Horizon at near eutectic compositions
generated partial melt inside of the matrix of
rock of moderately diverse δ18O (Bindeman
et al., 2008a; Wotzlaw et al., 2012). These lowdegree (<20%) silicic partial melts are difficult
to extract; however, minerals in them become
isotopically equilibrated and zircons partially
reset their U-Pb ages and acquire oxygen and
hafnium isotopes reflecting each minute magma
batch. In the remelted Sandwich horizon of
Skaergaard, isotope diversity of zircons occurs
on the 10–102 m scale (Bindeman et al., 2008a),
a scale we use for modeling. The Skaergaard
example is evidence that there are no extreme
δ18O values in the partial melts of variably
hydrothermally altered rocks, as isotopic differences induced by heterogeneous isotopic
alteration (Taylor and Forester, 1979) are further
averaged out by melting process in line with the
reasoning above (Figs. 5 and 6B; stage III). The
Sandwich horizon recrystallized 20 k.y. later
(Wotzlaw et al., 2012), without extraction of
this partial melt, which only underwent limited
intercumulus fluid flow (Tegner et al., 2009).
If the Skaergaard intrusion had another sill
intrusion soon after Basistoppen, and/or if the
Basistoppen sill intruded closer to the Sandwich
horizon, it could have resulted in greater degree
of partial melting approaching the extraction
window of ~50% melt (Marsh, 1981; Wickham, 1987; Dufek and Bachmann, 2010; stage
V in Fig. 6B) and would have segregated silicic
melt. Furthermore, if the sill intruded directly
under the Sandwich Horizon, partial melting
of this laterally confined eutectic area would
have been more efficient, and large fraction
partial melt extraction would have been thermomechanically driven by the viscous convection
Geosphere, October 2014
in the lower layer. This process is not static and
involves melting through the separating rocks
by hot magma pulses (e.g., Simakin and Bindeman, 2012), mixing and diking between magma
batches, and segregation by buoyancy; these
processes are thermomechanically modeled in
the following. This set of processes is called
bulk cannibalization (Bindeman, 2008; Watts
et al., 2011) and is similar to the reactive bulk
assimilation model of Beard et al. (2005) and
Bacon and Lowenstern (2005).
Batch Assembly, Preeruptive Mixing of
Melts, and Diverse Crystalline Cargo
Leading to Diverse Crystals
The advanced stages of segregation of
10–100 m magma batches lead to further averaging out of isotope diversity of magma, but
bring together isotopically-diverse crystal cargo
with diverse zircons. Because diffusion of oxygen and Hf is the slowest in zircons, the zircons
provide the best memory effect of the described
multistage segregation, while other minerals
undergo greater reequilibration. Most important, the majority of zircons are out of isotopic
equilibrium with the final melt they are erupted
in (Figs. 3 and 4), which suggests that a preeruptive mixing episode was efficient and isotopically homogenized the melt.
Textural Evidence of Recycling in Crystals
Abundant bright cathodoluminescence rims
in quartz and zircon provide evidence of recycling of as much as half of the crystal mass due
to remelting processes, and the time between
melting and mixing on the batholithic scale
is measured in centuries (Gualda et al., 2010;
Pamukcu and Gualda, 2010). The observation
by Poldervaart (1956) that zircons in granites
are have log-normal size distributions is supported by investigations of zircons and other
phenocryst phases in single pumice clasts from
major tuff units (Fig. 7; Bindeman, 2003). In
Simakin and Bindeman (2008) we analytically
computed that the abundant concave-down crystal size distribution of zircons and quartz may be
explained by only 1–3 remelting episodes that
recycle the outermost 10%–30% of quartz or
zircon radii (Fig. 7). Delayed nucleation leads
to concave-up distribution and often a lack of
the smallest zircon and quartz crystal sizes.
Time Scales as Recorded by Crystals
Isotopic diversity in zircons occurs within
error of U-Pb ion microprobe age dating (Watts
et al., 2011; Drew et al., 2013) or even IDTIMS age dating with the precision of several
1000 lnα18Omineral-H2O
–5
0
5
10
15
20
25
50
150
Mafic
protoliths
450
550
small W/R
Temperature (°C)
350
Starting Rock
650
750
Albite
Quartz
Kaolinite
Chlorite
–20
–15
–10
–5
0
5
–14
–2
+2
10
Water
Rock
ers
wat
ted
shif
–1‰
+7‰
350
IV
batch
melting
assembly
760
Solidus, minute melts
850
merging,
mixing,
heating
Eruption
V
Temperature, °C, increasing with time
200
III
melting
650
700
750
800
850
C 900
0
1.5 kbar
3wt% H 2O
25
% crystals
50
“bulk” melting Qz+San+Plag
75
water out
Eutectic Range
Qz out
Plag out
Zrc out 100 ppm
200 ppm
San in, liquidus
Extraction window,
Inheritance of
crystals
δ 18 O, ‰, rock
Figure 6. Modeling of hydrothermal alteration, remelting, melt extraction, and segregation. (A) Isotope effects of hydrothermal alteration of rocks
and minerals as expressed by equilibrium isotope fractionation. Complete equilibrium hydrothermal alteration at high water/rock (W/A) ratios
above the brittle-ductile transition will generate a range of δ18Orock values higher than water, and shifts water to higher δ18O values. Incomplete equilibrium and/or disequilibrium exchange will generate local diversity in δ18O on a small scale shown by zigzag pattern (dotted black line). (B) Isotopic
diversity in rocks during progressive hydrothermal alteration and subsequent melting. I–II: Hydrothermal alteration and generation of a range in
δ18Orock values at large water/rock ratio and temperatures <~400 °C (see A and text for explanation). III: Heating the rocks above solidus results in the
appearance of low-degree partial melts in areas with eutectic compositions (see Fig. 5), resulting in diverse crystalline cargos and disequilibrium
zircon compositions (similar to what is observed in Skaergaard; see text). IV: Continuing melting leads to growth of magma batches. Later zircon
paths are defined by the particular temperature-composition path of each magma batch (see Fig. 5). V: Subsequent creation of voluminous and
eruptable magma body. (C) Results of melting experiments of two silicic (beige and blue) and one mafic (green) assemblages using rhyolite-MELTS
(Gualda et al., 2010). T—temperature. Note extended near-isothermal eutectic melting range (725–750 °C) of silicic compositions and disappearance
of minerals and zircons at different levels of zircon (Zrc) concentration. At a likely melt extraction window above ~50% of partial melting only zircon,
sanidine (San) and sometimes quartz (Qz) will be inherited as xenocrysts or antecrysts (Plag—plagioclase). Melting of mafic lithology is also shown
for comparison with a narrow range of eutectic melting resulting in low degrees of partial silicic melts that are difficult to extract. At low degree
of partial melting (such as observed in Skaergaard; see text), interstitial minute melts will be retained and zircons will be partially reequilibrated.
250
brittle,
large W/R
Silicic
protoliths
+7
δ18 O, ‰
30
brittle
II
ductile
hydrothermal
alteration
Bulk Cannibalization
I
Partial melting,
B
T (°C)
A
00969 1st pages / page 9 of 27
Rhyolites—hard to produce but easy to recycle
Geosphere, October 2014
Retention of minute melts
thousand years (Wotzlaw et al., 2014), but
mineral-diffusive time scales suggest even
shorter time scales of mixing. Because oxygen
diffusion in quartz is >104 times more rapid
than in zircons (Watson and Cherniak, 1997;
Cole and Chakraborty, 2001), the preservation
of isotopically zoned crystals in many units
suggests short times (e.g., hundreds of years)
between segregation, mixing, and eruption. In
many other units quartz is isotopically annealed
due to residence times >~100 yr (e.g., Bindeman and Valley, 2001; Bindeman et al., 2006).
Furthermore, Cathey and Nash (2004) and
Ellis and Wolff (2012) presented evidence for
compositionally (Fe-Mg) polymodal pyroxene
crystals in the Cougar Point Tuffs (Table 1),
which are largely unzoned and record very high
(900–950 °C) temperatures. Fe-Mg interdiffusion in pyroxenes is even more rapid than
oxygen diffusion in quartz, and pyroxene crystals are typically small; the occurrence of such
crystals can help further constrain time scales
of extraction from the hot, once recycled, low
δ18O crystal batches or mushes that underwent
very rapid preeruptive mixing.
Tasks for Numerical Modeling
The microanalytical isotopic evidence mentioned here directs the numerical efforts. We aim
to quantify our above observations on (1) efficacy
of melting and formation of magma batches,
(2) efficacy of merging together of initially separated batches, (3) efficacy of mixing by convection and other processes involved in merging
the magma bodies with diverse crystal cargos,
and (4) the time scales of these processes, as
well as the conditions when magma mixing is
interrupted and magma bodies erupt.
METHODS
For this work we used proprietary written
and relatively simple two-dimensional finite
element code to model of convection of incompressible fluid at low Reynolds numbers (plug
flow approximation), valid for small magma
sills with vertical dimensions of as much as
~30 m. Additional calculations were performed
on larger grids covering 100–1000 m to analyze
the mechanics of large-scale batch assembly,
from small magma batches through slow viscous deformation of the frame rocks; numerical
details are described in the following. We also
performed several more roof melting + mixing
experiments (started in Simakin and Bindeman,
2012; technical details provided therein). Pressure was eliminated from the Navier-Stokes
equations with the penalty formulation (King
et al., 1990). Material transport is modeled by
9
00969 1st pages / page 10 of 27
Bindeman and Simakin
Lagrange markers with 10× finer resolution in
each spatial dimension. The typical grid was
180 ×180. The code was tested with the analytical solution for the convection in the liquid
with sharp viscosity contrasts (Trybitsyn at al.,
2006). Further technical details and material
properties of magmas and country rocks are
listed in Table 2.
RESULTS OF NUMERICAL
EXPERIMENTS
Formation of Larger Magma Bodies
Melting and generation of small magma
batches for hotspot and rift environments were
numerically modeled in Simakin and Bindeman (2012) and continued in this work (Fig. 8;
see Supplemental Files 11, 22, 33). We model
multistage generation of eruptable high-silica
rhyolitic magma bodies (Fig. 5). The first is
generation of silicic melts in the hot zone near
the Moho or lower crust (e.g., Annen et al.,
2006) by differentiation of basalt and melting
the lower mafic crust; these dry and near-liquidus silicic melts rise up rapidly, carrying heat
and undergoing further adiabatic decompression and overheating in pressure-temperatureX compositional space. They are then intruded
into the upper crustal silicic lithologies defining
rheological or density traps, i.e., hydrothermally
altered portions of the silicic batholith or under
low δ18O silicic caldera fill (Figs. 5 and 6). Formation of these small, ephemeral magma bodies
proceeds rapidly, in which roof-rock melting of
isotopically distinct protolith is accompanied
1
Supplemental File 1 is a generic melting by hot
rhyolite intruded into cold, solid rhyolite, both with
realistic phase diagrams (see Simakin and Bindeman, EPSL, 2012). If you are viewing the PDF of
this paper or reading it offline, please visit http://
dx.doi.org/10.1130/GES00969.S1 or the full-text
article on www.gsapubs.org to view Supplemental
File 1.
2
Supplemental File 2 is a simulation of the bridge
failure by “melting it through” without considering
gravity or stress fields leading to subsequent mixing;
The lower magma is a hot rhyolite with T=870°C, the
upper is cold, near solidus rhyolite with T=730°C,
the green bridge is solid, higher melting T. If you are
viewing the PDF of this paper or reading it offline,
please visit http://dx.doi.org/10.1130/GES00969.S2
or the full-text article on www.gsapubs.org to view
Supplemental File 2.
3
Supplemental File 3 is a melting run emphasizing two refractory blocks in the roof rock, which
have higher melting temperature (e.g. slivers of precaldera ancient rocks). They release potential energy
upon fall, promote mixing, and carry down frozen-in
rhyolitic material. If you are viewing the PDF of this
paper or reading it offline, please visit http://dx.doi
.org/10.1130/GES00969.S3 or the full-text article on
www.gsapubs.org to view Supplemental File 3.
10
by mixing via convective melting; the roof collapses, leading to inheritance of partial melts and
isotopically diverse crystalline cargo, in which
δ18O and εHf in zircons survive the longest (e.g.,
Fig. 2). The melting processes of silicic, near
eutectic, often glassy rocks are incredibly rapid
and energy efficient (Fig. 8; Simakin and Bindeman, 2012), and serve as an energy sink, with
minimal heat dissipation to the environment.
We thus consider several first-order thermomechanical mechanisms of growth and differentiation of larger volume magma bodies in the
upper crust, assembled from these melts, and
those that were produced in situ. These shallow magma bodies should be at least as large
and as hot as the volume and temperature of the
observed erupted ignimbrites, i.e., 102–103 km3,
but likely more than twice that given that some
residual magma will remain in the crust.
Expansion of the Melting Zone
With addition of heat and mass from below,
the upper crustal magma body may simply grow
by expansion of the melting zone if magma
addition is faster than cooling. Continuous supply of new hot silicic differentiates of basalt to
the base leads to rapid upward expansion of
the melt zone and sinking of separating screen
rocks. High versus low basaltic power outputs
(Lipman et al., 1972; Hildreth, 1981; Lipman,
2007; de Silva, 1989) determine particular
temperature-time trends and the subsequent
lifetimes of each of these magma bodies, pertinent to the isotopic reequilibration of inherited
crystalline cargo (Figs. 5 and 6).
Merging by Melting Through
Separating Screen Rocks
We envision that areas of melting and shapes
of magma bodies are initially complicated (e.g.,
Figs. 5 and 6) and may be originally determined
by country-rock composition, areas of higher
glass content (so less latent heat is needed), and
hydrated areas of hydrothermal preconditioning.
In addition, hydrothermal redistribution of silica
and alkalies will create local silica oversaturation conditions, favoring lower temperature
eutectic melting in selected areas. Furthermore,
screens of higher melting temperature rocks
may separate low-temperature eutectic rhyolite
lithologies. Figure 8 and Supplemental File 2
(see footnote 2) demonstrate such a scenario of
amalgamation by two melt zones via the melting
through of separate screen rocks with different
silica contents. Stopping of these screen areas
will lead to the release of gravitational energy,
enhancing convective mixing. The thermomechanics of convection play a more important
role than silica oversaturation (as also observed
in Simakin and Bindeman, 2012).
Geosphere, October 2014
Physics of Batch Assembly in
Stress and Density Fields
We consider simple numerical models with
two vertically separated and offset magma sills
(Figs. 9 and 10 and Supplemental Files 44, 55, and
66) and explore conditions of when and how these
sills can merge, and exchange and mix their crystalline cargoes. In the following we employ scaling analyses, numerical experiments, and consider relevant physical aspects of large magma
body assembly from batches in a stress field.
Formation of Magma Sills and Batches
Each magma batch forming in situ by melting, or those arriving from below into the zone
of accumulation, can mechanically interact with
neighboring batches if they are close enough
to each other, or with a large dominant magma
reservoir. This interaction occurs through the
thermomechanical influence of the stress field
around each magma sill that controls the generation and direction of dike propagation (Fig. 9A).
While there are multiple studies of the interacting vertical dikes (e.g., Takada, 1994a, 1994b),
there are limited analytical or numerical simulations of the mechanical interaction of a magma
reservoir and a rising dike. Field observations
(e.g., Horsman et al., 2009) from some shallow
silicic intrusions of ~1 km in diameter document
that when small magma batches are emplaced
successively, magma bodies internally inflate
then spread laterally, forming a series of silicic
sills emplaced on top of each other, creating
voluminous silicic stocks.
Subsequent analysis of the interaction
between magma batches and a rising dike often
employs consideration of the principle stress
(σ1) directions around a magma reservoir,
defining approximate trajectories for diking
via hydraulic fracturing of rocks (Jellinek and
4
Supplemental File 4 shows batch merging in
gravitational and stress fields in viscoplastic crust, as
described in Fig 9 and 10 of text of the paper. If you
are viewing the PDF of this paper or reading it offline,
please visit http://dx.doi.org/10.1130/GES00969.S4
or the full-text article on www.gsapubs.org to view
Supplemental File 4.
5
Supplemental File 5 shows batch merging in
gravitational and stress fields in viscoplastic crust, as
described in Fig 9 and 10 of text of the paper. If you
are viewing the PDF of this paper or reading it offline,
please visit http://dx.doi.org/10.1130/GES00969.S5
or the full-text article on www.gsapubs.org to view
Supplemental File 5.
6
Supplemental File 6 shows batch merging in
gravitational and stress fields in viscoplastic crust, as
described in Fig 9 and 10 of text of the paper. If you
are viewing the PDF of this paper or reading it offline,
please visit http://dx.doi.org/10.1130/GES00969.S6
or the full-text article on www.gsapubs.org to view
Supplemental File 6.
00969 1st pages / page 11 of 27
Rhyolites—hard to produce but easy to recycle
HRT-C
8
gro
wt
h
C
9
7
HRT-A
6
Kilgore Tuff
5
4
3
new growth
B
Relative Abundance per cm 3, ln (n/LV), cm-4
10
d is
sol
u ti
on
A
MFT
dissolution
LCT-A
2
HS14
1
2017B
HS10
0
50
100
150
200
250
Zircon Long Axis, μm
Figure 7. Textural and crystal size distribution evidence for recycling of quartz and zircon. (A) Cathodoluminescence images of recycled
zircons with variable δ18O values but similar ages from the Heise complex Kilgore tuff suggesting rapid preeruptive assembly of diverse
magma batches. (B) Normalized (to the largest crystal) crystal size distribution (CSD) of zircon and quartz in Huckleberry Ridge Tuff C
(HRT-C) of the Yellowstone Group requiring two solution-reprecipitation episodes, each recycling between 0.36 and 0.55 fractions of crystal
mass, or 10%–20% of crystal radius (Simakin and Bindeman, 2008). Recycling was likely caused by changing temperature conditions during batch assembly, given the strong zircon and quartz diversity in this sample (Fig. 3; HRT-C).
DePaolo, 2003; Simakin and Ghassemi, 2010;
Karlstrom et al., 2012). This analysis leads to
the observation that for the overpressurized
deep spherical magma reservoir (equally inelastic, viscoelastic, or viscous media), σ1 directions
are such that the greatest compressive stress is
radial to the magma body and σ3 is tangential
(Fig. 9). For the underpressurized magma reservoir, principal stress directions are transposed
and σ1 becomes tangential. Our numerical
model shows (Fig. 9B) that an overpressurized
magma body attracts a dike, causing it to propagate radially and coalesce with the magma body,
thus causing the magma body to grow.
When an underpressurized magma body
deflects a dike, it may produce a separate sill
(Fig. 9B) capable of growing into a separate,
independent magma body with a continuing
supply of magma. Pressure lower than lithostatic may develop in a magma body with a low
supply flux during the solidification of magma
due to the volume effect of crystallization and
escape of exsolved fluid phases, and due to
extensional tectonics. Numerical simulations
of viscoelastic elliptic chambers (see Simakin and Ghassemi, 2010) demonstrated that at
magma pressures lower than lithostatic pressures, expressed through the bottom-parallel
σ1 stress directions, the rock envelope presses
into the chamber and undergoes radial extension. These arguments support accumulation of
newly arriving batches of magma into a single
Geosphere, October 2014
growing magma body in sites of active magma
generation and high intrusion rate. At reduced
magma supply rates, crystallization and therefore underpressurization of magma will promote formation of separate magma batches,
but underpressurization conditions may even
exist at some high supply rates (Gregg et al.,
2013). An additional factor promoting separate
magma batch formation may be preexisting
oblique faults that serve as structural traps for
rising magmas and relieve stress from overpressured magma bodies (Fig. 9B; Simakin and
Ghassemi, 2010). This is particularly relevant
for Basin and Range faults and extension along
the SRP (Fig. 1), or any other silicic magma
generation in rift zones.
11
00969 1st pages / page 12 of 27
Bindeman and Simakin
TABLE 1. NEW LASER FLUORINATION ANALYSES OF QUARTZ, CLINOPYROXENE, AND PLAGIOCLASE OF COUGAR POINT
TUFF UNITS IN THE BRUNEAU JARBIDGE ERUPTIVE CENTER AND THE KILGORE TUFF OF THE HEISE CENTER
δ18O, ‰ SMOW
Elevation
Plag, or glass
Cpx
Magma*
Latitude
Longitude
(ft)
Description
Sample ID
Qz
Bruneau Jarbidge Eruptive Center, Central SRP
2005-ID-14
CPT-III, 12.7 Ma
4.02
3.52
41°57.602′
115°39.810′
5355
Rowland, Pack trail
2005-ID-13
CPT-V, 12.07 Ma
4.53
1.68
3.8
5381
Rowland, Pack trail
41°57.602′
115°39.810′
2005-ID-3
CPT-VII, 12.8 Ma
41°55.995′
115°25.402′
5771
Deer Creek Rd
0.71
0.21
2005-ID-12
CPT-IX, 11.6Ma
2.52
0.62
2.02
5486
Rowland, Pack trail
41°57.610′
115°39.786′
2005-ID-5
CPT-IX, 11.6Ma
2.52, 2.64
2.08
41°55.227′
115°25.716′
5145
Deer Creek Rd
2005-ID-7
CPT-XI, 11.3 Ma
3.26
2.76
41°56.138′
115°25.403′
6347
Deer Creek Rd
CPT-XII, 11.2 Ma
–
–0.91, –1.15 –1.92
–0.9
Deer Creek Rd
SB173†
2005-ID-8
CPT-XIII, 10.79 Ma
3.67
3.22
3.27
6432
Deer Creek Rd
41°56.517′
115°25.363′
2005-ID-9
CPT-XIII, 10.79 Ma
3.62
3.16
6593
Deer Creek Rd
41°56.576′
115°25.336′
2005-ID-2
CPT-XVj, 10.5 Ma
3.61
42°02.7859′
115°36.3833′
5161
Murphy Hot Springs
3.08
2.58
Twin Falls and Heise Centers
INB-1
Shoshone Falls rhyolite, Twin Falls
–
0.31
0.3
42°35.8333′
114°24.11′
3078
Twin Falls
2017B
Kilgore Tuff, Heise, 4.4 Ma
3.62
2.75 (glass)
1.28
2.9
NW of caldera
44°07.8′
112°32.4′
Note: All data were analyzed in the University of Oregon Stable isotope lab; see Fig 2 for plotted data.
*Samples were collected using descriptions from Bonnichsen (1981) and Cathey and Nash (2004). The effort was made to extract and analyze alteration-resistant minerals
quartz and clinopyroxene to check and complement published bulk feldspar data (Boroughs et al., 2005; Bonnichsen et al., 2008).
†
sample kindly provided by S. Boroughs; * calcuated from phenocrysts
Merging of Vertically Stacked Magma Batches
Sills that were closely spatially emplaced are
separated by bridges (or screens) of country
rocks, or crystal mush if emplacement happens
into a cooling and crystallizing batholith (Figs.
10–13). If certain mechanical conditions of
interaction are met, these bridges can fail, causing sills to merge to form bigger magma bodies
(e.g., Schofield et al., 2010, 2012a, 2012b).
We consider two basic controls on the
mechanical interaction between sills: overpressure (Fig. 10), and the density contrast between
magma and country rocks (Figs. 11–13). In
the modeling approach used here, when slow
motion of country rocks is simulated, dimensions of computational domain increase to tens
of kilometers at the same grid size. In our simulations the viscosity contrast between deforming rocks and magma was set to 104.5, similar
to the maximum viscosity contrast 105 used by
Bittner and Schmeling (1995) for the modeling
of the Rayleigh-Taylor instability of the melting
front in the underplated silicic crust; this value
is less than the real ratio between hot country rocks (1017–1018 Pa·s) and silicic magma
(105–106 Pa·s). However, further increase in
the viscosity ratio (above 104–105) has no effect
on the numerical solution for slow deformation of rocks. It allows us to portray large-scale
slow features (this does not require accounting
for the inertia term in the Navier-Stokes equation) instead of modeling the fast processes
with the 10 cm scale resolution used for melting processes. Consideration of a far greater
viscosity ratio or equally high Rayleigh number
(Ra) requires a supercomputer approach (Longo
et al., 2012) for implementation of the inertia
terms and integration of the small and large time
and spatial scales, but does not change our outcome for slowly deforming rocks.
12
We also simplify our analysis also by neglecting complex and fast thermally activated processes inside sills, such as convection, melting,
and crystallization, thus concentrating only on
the viscous (slow) deformation of the rocks.
Viscoelastic materials (Maxwell viscoelasticity) behave elastically at fast deformation and
flow as viscous material at slow deformation
(Gerya, 2009; Kaus and Podladchikov, 2006).
Scaling analysis defines the Debora number,
De = η/Gt0, as a measure of viscoelastic behavior, in which η/G is the characteristic time of
viscous relaxation of the deviatoric stresses, G
is the shear modulus, and t0 is the time scale of
TABLE 2. SYMBOLS AND PHYSICAL PARAMETERS USED IN THIS PAPER
Δρ
density contrast between magma and rocks →s -→m, kg/m3
l0
length scale equals size of the element in numerical grid, m
h
h0
η
Ra
P0
λP
t0
τ
G
σ
σ
Sy
Sy
β
viscosity, Pa·s
viscosity scale, Pa·s
nondimensional viscosity η/η0
Rayleigh number, Ra = 1 at t0 = η0 /Δρgl0
pressure scale Δρgl0 /Ra
overpressure, MPa
time scale, Ra η0 /Δρgl0
t/t0
shear modulus, GPa
stress, MPa
scaled stress σ/P0
yielding stress, MPa
scaled Sy /P0
Poisson’s ratio
νu undrained
Skempton’s poroelastic coefficient
thermal (T—temperature) expansion coefficient of fluid
porosity
pressure conductivity, m2/s
thermal conductivity, m2/s
fluid pressure, MPa
circular shell radius
thickness: hsill = sill, hb = bridge between sills
diapir rise velocity, m/day
magma chamber volume
sill semiwidth (km)
geometric factors, nondimensional parameters involving
characteristic spatial dimensions.
δX
δXm
δR
marker horizontal displacement, m
δX averaged over magma volume
pair of markers separation
ν
B
αT,fl
φ0
C
kT
pfl
R
h
U
V
W1/2
Geosphere, October 2014
100–1000
0.25 (convection), 30–200 m
(slow deformation)
104–1018
108–1013
1–105
40–120
typical value 1 yr
5–50
0.1–150
0.15–0.5
0.6–1.0
9 10–4 (T = 800 °C)
0.01–0.3
5 10–3–1.3 10–5
10–6
<1; see text
0–30 m
initial δR0 = 2.5 cm
00969 1st pages / page 13 of 27
Rhyolites—hard to produce but easy to recycle
700°C
A
B
C
D
850°C
around the magma sill arises at the very beginning of deformation, and changes slowly with
the expansion of the sill, often leading to dike
initiation. When the level of stress, a product of
the second invariant of strain rate and viscosity,
exceeds the yield stress of country rocks (Fig.
10), the screen rock between magma batches
fails. Shear strain localization is facilitated by
means of dynamic power law (DPL) rheology
(proposed by Sleep, 2002, and developed by
Simakin and Ghassemi, 2005, and Simakin,
2014, and references therein) to produce sharp,
fracture-like shear zones (Fig. 10). The DPL
model treats viscosity as a dynamic parameter
accounting for the effect of the texture evolution
and described by a system of ordinary differential equations on the material Lagrange particles. As an asymptotic limit at a fixed strain rate
(ė),the DPL model yields an ordinary power-law
viscosity law [η = a(T)/ė m] where T is temperature, a is a some temperature-dependent parameter, and m is an exponent coefficient used in
description of power laws (e.g., Becker, 2006).
Calculations were performed in nondimensional form. Since density is uniform at the
overpressure approximation (i.e., we consider
magma overpressure over the lithostatic), the
gravity term can be dropped by setting the Rayleigh Number, Ra, to 0. Stresses and viscosity
are scaled with some pressure (P0) and time
scales of deformation (t0):
Sy = Sy /P0 , η = η /P0 /t0 .
Figure 8. Simulation of the bridge failure by melting through leading to subsequent mixing in the entire magma body. The simulation does not account for gravity or stress fields.
The lower magma is a hot rhyolite at 870 °C. The upper two layers are fully crystalline
rhyolites near solidus at 730 °C. The green bridge is a solid, silicic rock with excess quartz
(45%);the lower and upper layers have 35% normative quartz. Box size is 30 m. (A) 0 days.
(B) 19 days. (C) 22 days. (D) 27 days.
the process. When De < 1, mechanical evolution
of the viscoelastic solid can be approximated by
viscous flow. The viscous flow approximation of
the deforming crust is likely the correct case, as
is suggested by geodetic data on surface deformation above an overpressurized magma chamber, data that include the volume of the rocks
in the thermal aureole (Newman et al., 2006).
Newman et al. (2006) suggested that the threshold viscosity for liquid behavior of the thermal
aureole is 3 × 1015 Pa·s (G = 5 GPa), and De ≤
0.26 for several months (global positioning system and interferometric synthetic aperture radar
observations). The viscous approximation used
in the following is therefore most likely valid for
slow deformation of the rocks around magma
sills below the brittle-ductile transition depth,
typically at 400–450 °C.
Effects of Overpressure
We investigated the effects of overpressure
on a pair of vertically stacked sills (Fig. 10),
with the deeper sill undergoing gradual infill
at a constant volumetric rate. For this purpose
we use a vertical conduit of rectangular shape
(Fig. 10) and apply a parabolic vertical magma
ascent velocity (with zero values at the walls) at
the inlet. This sill inflates with time, inducing
viscous stress. At a given filling rate, a quasisteady-state stress field in the country rocks
Geosphere, October 2014
(1)
In these calculations nondimensional viscosities
of the frame rocks are kept at ηr = 105 and magma
ηm = 1 with yield stresses at Sy = 50–200. In the
following text we will use symbol Sy for the
scaled yielding stress. Then, for example, at P0
= 1 MPa, the physical value of Sy will be 50–200
MPa. Time scales can be found by setting rock
viscosity t0 = ηr /ηr /P0 . With the relatively low
viscosity of the country rocks ηr = 3.0 1018 Pa·s,
close to the value estimated for the thermal
aureole of the magma chamber underlying the
Long Valley Caldera (Newman et al., 2006), and
pressure fixed at P0 = 1 MPa we get a time scale
for deformation t0 = 1018.5–6-5 = 107.5 s, or ~1 yr.
In our experiments the normalized sill expansion rate dV/V0 /dτ was ~0.01 or 0.01/t0 =
0.01 yr –1 for the above parameters. By increasing P0 we proportionally increase the rock yield
strength Sy and the physical expansion rate of
the magma. In other words, these nondimensional results can be interpreted equally with
the choice of a high filling rate and high Sy, or
a low filling rate and small Sy , as these are linearly interrelated. The linear scale, l0, remains
arbitrary. Possible limits of l0 are constrained by
the absolute intrusion rate (km3/yr). Results of
13
00969 1st pages / page 14 of 27
Bindeman and Simakin
σb = ΔP= αη
A
B
Figure 9. Influence of overpressure and underpressure in both a spherical or sill-like
single magma chamber will either cause dike attraction or deflection. (A) Generalized picture showing principle stress 1 (S1) around an underpressurized and pressurized circular magma chamber determining trajectories of dikes. (B, C) Numerical modeling of dike
attraction and deflection in a more complex stress field involving a reverse Basin and Range
fault and free upper boundary (from Simakin and Ghassemi, 2010). B is an overpressurized
magma body, C is an underpressurized magma body. Chamber underpressure may lead
to dike deflection, but also formation of an independent thin sill below the original magma
body. If overpressure in the dike is stronger than in the chamber, the dike will be absorbed.
Here we assume that stress field around a dike’s tip decays faster than stress field around
chamber, with many orders of magnitude larger curvature. The dashed orange line depicts
saucer-type sill growth along the calculated σ1 direction. The continuous black line depicts
the edge of a saucer-type sill resulting from analogue modeling (Galland, 2012). The dotted
line corresponds to the shape of the reconstructed sill in the Karoo Basin (South Africa;
from Polteau et al., 2008). The black lines were scaled with conservation of proportions.
Blue line is a fault or frictionless contact surface.
the numerical experiments with varying Sy and a
constant filling rate (0.01) show that bridge failure occurs at Sy < 150 (150 MPa at P0 = 1 MPa).
This confirms that the order of magnitude scaling estimate of the viscous stress in the bridge
is correct, since 150 MPa is close to the typical
shear strength of the rocks (Barton, 2013).
Overpressure defines the amplitude of the
induced stress, and for the expanded two-dimensional (2-D) circular shell around a magma body
it was shown to be (e.g., Jellinek and DePaolo,
2003; Simakin and Ghassemi, 2010):
14
(
)
ΔP= 1− (R1/R2 )2 η
dV/dt
,
V0
(2)
where R1 and R2 are internal and external (R1 <
R2) radii of the 2-D circular cavity filled with
liquid. The external radius in these approximations is defined as an elliptic envelope of
deformation, which coincides with the brittleductile transition (Newman et al., 2006). In the
case of two cavities next to each other, we can
expect the level of stress in the separating screen
rock to be:
Geosphere, October 2014
dV/dτ
,
V0
(3)
where σb is the stress component normal to
the boundary (radial for circle inclusion) of the
expanded sill, and α is some geometric factor
<1, which depends on the particular geometry of
the system; α can be determined analytically or
numerically (Simakin and Ghassemi, 2010). For
a circular cavity in the shell α = 1 – (R1 /R2)2 or
(R1 /R2)2 = 0.5, which corresponds to equal volumes (areas in 2-D) of magma and rocks in the
thermal aureole, α = 0.5. For an elliptic chamber
and viscoelastic rocks the numerically estimated
value of this coefficient is 0.32 (recalculated
from Simakin and Ghassemi, 2010). By setting
α = 1/3, P0 = 1MPa, and dV/dτ/V0 = 0.01, we get
an estimate of 333 MPa for the viscous stress in
the bridge.
These numerical experiments (Fig. 10) demonstrate how the process of thin bridge failure
between closely spaced magma lenses proceeds
in 2-D. In particular, they show that the bridge is
initially faulted at one end at the very beginning
of deformation and then rotates upward onto
and into the upper sill as a handle or trapdoor,
eventually giving way for lower magma to flow
into the upper sill. The thicker bridge between
magma lenses yields two fractures, leading to
the extrusion of a part of the bridge in a pistonlike fashion into the upper magma body (see Fig.
10); we call this mechanism an internal caldera
collapse. Our analysis also demonstrates that
widely spaced magma batches can coexist for a
long time without interaction, and homogenize
their crystalline cargo and whole-rock chemical
and isotopic composition on time scales of cooling. For example, Wilson and Charlier (2009)
demonstrated coeval, probably lateral, reservoirs separated by ~20 km at the Taupo volcanic
zone (New Zealand). The numerical results here
were obtained exclusively due to an overpressure response in the viscous media without considering density terms, which we consider next.
Influence of Magma Buoyancy
When sills are intruded into sufficiently soft
country rocks or into a crystal mush, the density
difference between the magma and surrounding
media leads to slow deformations due to buoyancy. In the set of stacked magma lenses (sills),
there are two possible mechanisms for buoyant
rise: (1) the large-scale diapiric rise of one or
two magma bodies, thinning and disrupting the
screen rocks between, causing the merging of
these into a single magma body; and (2) development of Rayleigh-Taylor instabilities of the
plastic screen rocks separating the sills. Our
numerical experiments (Figs. 11–13) show that
diapir-like motion always prevails in the set of
the stacked sills, as opposed to Rayleigh-Taylor
00969 1st pages / page 15 of 27
Rhyolites—hard to produce but easy to recycle
A
B
Figure 10. (A) Numerical modeling of merging
magma bodies together at different nondimensional country-rock yield strength (Sy, in MPa;
see text) but identical magma and countryrock density. Viscosity of country rocks is 1018
Pa·s. The overpressure is created due to slow
infilling of the lower reservoir by the feeder
dike. (B) Flow structure in dike. See Supplemental Files 4, 5, and 6 (see footnotes 4, 5, and
6) for temporal evolution movies. Note the
various regimes of screen-rock failure. Three
models with Sy = 50 MPa differ in the shape of
the connecting dike.
instability of the separating screen rock. Diapirlike motions are driven by horizontal density
gradients. For diapiric batch rise, the maximum
flow rate during approximately one nondimensional unit of time varies slowly, and the vertical position of the screen midpoint rises linearly
above the ascending flow of the lower sill (Fig.
12B). For sufficiently weak screen rocks, viscous stresses overstep the shear strength (yield
stress in the frame of the viscoplastic rheology),
resulting in sharp, fracture-like shear zones
splitting the screen rock and surrounding rocks,
thus facilitating further deformation.
Numerical models of the buoyant rise can be
further rationalized with simple scaling analysis. With the nondimensional Navier Stokes
equation, we use time scale t0, length scale l0
(size of the element grid), and viscosity scale
η0. Pressure (stress) scale is derived as P0 =
η0 /t0. With choice of the time scale equal to
t0 = η0 /Δρgl0 Rayleigh number (multiplier in
the scaled gravity term) becomes unity and P0 =
Δρgl0. If Ra is set larger than unity (for practical considerations), then t0 = t0Ra. Thus gravity
flow develops identically with the fixed initial
geometry for the various density contrasts and
contrasts in viscosities and linear sizes, but on
different time scales. The characteristic diapir
velocity for a sill semi-width W can be compared with corresponding Stokes rise velocity
Us of the sphere (3-D) or circle (2-D):
Us = kS R 2Δρg/ηr ,
(4)
where for a solid sphere kS = 2/9, for a solid
circle kS = 1/4, and for a gas 3-D bubble kS = 1/3;
all values are for slow flow at small Re numbers.
Our numerical experiment is not identical to any
of these geometrically idealized models. While
tentatively setting kS = 1/3, for the half-width of
the sill, W1/2, we rearrange the equation for Us in
nondimensional form as:
Us =
2
Ra l0 (W1/2 /l0 )
.
3 t0 (ηr /η0 )
(5)
Numerical nondimensional maximum flow
velocity at the beginning of the experiment
(ηr /η0 = 105, Ra = 120, Us = 0.18) can be approximated with W1/2 /l0 = 21.2 while circle inscribing
Geosphere, October 2014
both sills has nondimensional radius 32. Because
low-density material fills only part of this large
circle, correspondence seems to be satisfactory.
Dimensional velocity of the diapiric flow (Ud)
strongly depends on the physical parameters. In
a crystal mush environment with at ηr = 1013
Pa·s, l0 = 40 m, Δρ = 300 kg/m3, and geometry
as in Figure 11, we obtain Ud = 6 m/day with
maximal rise rate of as much as 4 m/min during
plastic failure. Sills intruded into stiffer rocks
with ηr = 1018 Pa·s, a value typical for a thermal aureole of an intrusion, will undergo diapiric rise at a velocity of 2 cm/yr; i.e., a diapiric
component can only contribute little to the commonly observed deformation rates of as much
as several meters per year associated with sill
intrusion. The main factor of uplift in this case
is magma overpressure.
Cooling and Deformation Rates
Another important time scale involved is that
of the cooling and solidification of magma bodies. Unlike salt diapirs, driven by a density instability without temperature contrast with country
rocks, magma sills cool and solidify; therefore,
15
00969 1st pages / page 16 of 27
Bindeman and Simakin
A
important on the cooling time scales (3–300 yr)
inferred for the relatively small, 10–100-m-thick
magma sills considered in this work. For Ra = 2
at β = 0.067 it yields small viscosities in the
range of 1012–1015 Pa·s that are probably typical
for crystal mushes (Lavalee et al., 2007) or for
some nonwelded pyroclastic rocks and hyaloclastites. For rock viscosity 1018 Pa·s, a value
typical for rocks near the brittle-ductile transition and in thermal aureoles of intrusions (e.g.,
Newman et al., 2006), and to satisfy Equations
6 and 7 for Ra = 2 and Δρ = 300 kg/m3, the sill
sizes are W1/2 = 2165 m, and W1/2 = 3702 m, if
the Ra = 10. Therefore, a significant increase in
size is required to compensate for an increase
in viscosity, and so only less viscous bridges
are capable of deformation due to buoyancy for
geologically realistic thicknesses and cooling
time scales.
Initial smooth deformations are accelerated
and modified after plastic failure of the surrounding rocks. For the geometry presented in
Figure 10 we find that the threshold nondimensional yield strength, Sy providing for plastic
failure of the bridge, rises linearly with Ra and
is 18 at Ra = 80. The dimensional value of Sy can
be recalculated via natural scale for pressure as:
B
Sy = Sy Δρl0 g/Ra.
Figure 11. Initial stage of the slow viscous deformations in the set of two stacked sills
(Ra = 80, initial log[ηr /ηo] = 5, Sy /P0 = 16, R = 500 m, Δρ = 300 kg/m3; see text and Table 2 for
abbreviations and symbols). Numbers on vertical and horizontal axes are nondimensional
linear coordinates in units of l0. Color coding strip denotes log(ηr /ηo). Physical values of
parameters are on right (see Fig. 10). (A) Flow field at the time, t = 0.232. (B) Viscosity distribution at t = 1.19. Yielding stress is reduced to scaled value of 16 in the elliptic zones around
sills partially seen as low viscosity areas; beyond these zones Sy is 200.
in order to be effective, the time of the screenrock instability and the ability to fail between
magma sills by diapiric flow (tb = hb /Us, where
hb is screen or bridge thickness) should be less
than the characteristic conductive heat dissipation time t1: t1 = hsill2/kT, where hsill is half-thickness of the sill, and kT is thermal conductivity. A
ratio of these two time scales gives:
tb = hb /Us =
hb η
2
W 1/2
Δρg
t1 /tb = β
and
1 h ΔρgR
t1 /tb =
.
3 hb R kT ηr
2
sill
ΔρgW1/32
kT ηr
1, β =
2
1 hsill
.
3 hbW1/ 2
(7)
3
(6)
Note that the last multiplier in the above expression is a quasi-Rayleigh number, which defines
the effectiveness of slow motion versus heat
loss. Equation 6 demonstrates that in order for
deformation to remain self-similar, a factor of
16
10 increase in the sill half-width W1/2 requires
a viscosity increase by a factor of 1000. The
condition of the buoyant rise time scale over
the conductive cooling time scale can be thus
expressed as this nondimensional quasi-Rayleigh number multiplied by the geometric factor. The geometric factor allows comparison of
nondimensionalized parameters for different
geometric configurations.
The geometric factor ratio β is proportional
to the product of the inverse aspect ratio
(W1/2 /hsill = 5) and the ratio of the sill half-width
to the bridge width (hsill /hb = 1). The resulting
value of β here is 0.067.
We can compute the required viscosities that
are required for the gravity instability to be
Geosphere, October 2014
(8)
With Δρ in the range 300–500 kg/m3 and l0 in
the range 40–100 m, we estimate Sy as 10–120
kPa. Such low values of Sy are compatible only
with shear strength of the nonwelded pyroclastics at the effective mean stress approaching
zero (σn, eff = 10 kPa, Sy = 30–350 kPa, Cecconi
et al., 2010) or crystal mushes (Sy = 0.14 kPa
for the solid fraction of isometric crystals, 50%
and less; Hoover et al., 2001; or 0.3–4.3 kPa;
Philpotts and Carroll, 1996).
Our analysis demonstrates that buoyancy
does not play a significant role in the merging of
small (10–102 m) magma batches, unless they
are already located in a weak mush. The brute
force buoyancy approach in this case requires
that magma batches are already of magma
reservoir size, >~15km, in order to cause diapiric rise, but such sizes may cause caldera collapse (e.g., Ramberg, 1970; Geyer et al., 2006;
de Silva et al., 2006; Gregg et al., 2012).
Overpressure Versus Buoyancy—What
Controls the Merging Processes and Stability
of Multibatch Magmatic Systems?
Figures 10 through 12 demonstrate that overpressure due to magma production and supply
(which causes sill formation on the first place)
is primarily responsible for the merging of
magma sills. This is achieved through mechanical failure and dike focusing across separating
00969 1st pages / page 17 of 27
Rhyolites—hard to produce but easy to recycle
ity radius, in agreement with theoretical and
numerical results (e.g., Jellinek and DePaolo,
2003; Simakin and Ghassemi, 2007). The
stresses are thus reduced ~9 times when the
distance between batches is increased from 1
to 4 times their radius. This has implications
for the long-term stability of widely dispersed
magma batches with distinct crystal cargoes.
The magma batches may exist without merging,
crystallizing new zircons with rejuvenated U-Pb
ages and batch-averaged εHf and δ18O (Fig. 4).
A
Overpressure in the Interstitial Fluid Phase
Interstitial hydrothermal fluid in country rocks
may play an important role in low-pressure
upper crustal conditions, in porous rocks, and
especially in buried tuffs. Fluid pressure may rise
rapidly in the thermal aureoles around magma
lenses upon their emplacement or redistribution,
leading to mechanical failure of the screen rocks
in the fast elastic mode. Heating of the pore
fluid increases its pressure almost proportionally to the temperature (e.g., according to gas
law) at a constant volume of pores, or faster if
there is a phase change. Real rock expands and
fluid migrates in response to the internal pressurization, and the net effect is determined by
the rock poroelastic properties and permeability.
The direct proportionality of fluid pressure with
respect to temperature and its dependence on
other parameters at the impermeable boundary
of the heated porous half-space was provided by
McTigue (1986). After simplifications,
B
pfl =
Figure 12. Development of the magma flow structure between vertically stacked sills in
due to buoyancy inside of soft country rocks (crystal mush or pyroclastic rocks). (A) Large
viscosity contrast of 105 between magma and country rocks at nondimensional t = 0.312.
(B) Small viscosity contrast of 300 at t = 0.378. Strongest deformation occurs in the bridge,
causing its failure and rapid flow of material from the lower to the upper sill. Numbers on
vertical and horizontal axes are nondimensional linear coordinates in units of l0 (see text).
When l0 = 40 m, domains becomes 6.4 × 3.6 km. Reference vector represents nondimensional
flow rate. For Ra = 80, Δρ = 300kg/m3, country-rock viscosity of 1013 Pa·s (see text), we get
2.1 m/h rate per 1 nondimensional rate unit, corresponding to 146 m/h for non-dimensional
scale vector 68.8 shown in (A) and the above parameters.
screen rocks connecting dikes, followed by the
initiation of magma coalescence. Density difference between rock and magma may cause rapid
coalescence of magma lenses only if they are
very close, longer lasting, and larger than the
country rock screens, which have low viscosities
and attributes of partially molten rocks, crystal
mushes, or nonwelded pyroclastic rocks. Buoyant merging may be important at later stages of
magma coalescence on a caldera scale, and may
be one of the causes of some caldera-forming
eruptions (Malfait et al., 2014; Caricchi et al.,
2014), although the mechanisms for achieving the excess buoyancy postulated remain
unclear. Overpressure or external tectonic controls (Allan et al., 2012) were proposed as an
explanation for the Taupo volcanic zone (New
Zealand) eruptions.
The analysis here demonstrates that widely
separated magma batches may exist without
interaction. In 2-D, stresses around circular
cavities decay as 1/(R/Rc)2, where Rc is the cav-
Geosphere, October 2014
2GB 2 φ oα T, fl ΔT
,
9(νu − ν) 1+ c/kT
(
)
(9)
where G is the shear modulus, ν and νu are the
drained and undrained Poisson’s ratio, respectively, jo is the porosity, c and kT are the pressure and temperature conductivities, aT,fl is the
fluid expansion coefficient, and ΔT is the temperature increment. Inserting typical parameter
values from Detournay and Cheng (1993) and
calculating aT,fl for an ideal gas at T = 800 °C,
we get a fluid pressure in the range of 16–470
MPa (e.g., typical for sandstone and marble).
The fluid pressure increase is the greatest for
rocks with low permeability and high porosity,
such as impermeable pumice, poorly consolidated, clay-rich sediments, or hydrothermally
altered rocks. Intrusion of sills in such sediments at depths of <2 km results in fluidization
of the rocks (Reid, 2004; Buttinelli et al., 2011;
Schofield et al., 2010), sometimes giving rise
to clastic dikes (Svensen et al., 2006). Even at
greater depths of 6–8 km, common for magma
reservoirs, pressurization effects might reach
tens of megapascals; the details depend on the
17
00969 1st pages / page 18 of 27
Bindeman and Simakin
poroelastic properties and permeability of the
country rocks. We stress that the rocks with
large isolated pores, characteristic of vesiculated
pyroclastic materials, are candidates for failure
due to the strong pressurization and mechanical weakening at the thermal aureoles of sills,
which we model. This makes realistic scenarios
of sills merging due to buoyancy (see Figs. 11
and 12), as rocks with low viscosity and Sy can
be interpreted as requiring fluid overpressuredegraded cataclastic materials.
A
Mixing Inside and Between Reservoirs
B
Melting
Front
Convective
C
Diverse in δ18O
roofrocks, zircons
Figure 13. Mixing induced by compositional convection during roof melting. (A) Screenshot of melting process showing descending plumes (illustrated by a numerical model from
Simakin and Bindeman, 2012); colors depict composition (scale: roof rock–magma). See
Supplemental File 7 (see footnote 7) for vorticities distribution movies. B and C are conceptual model for a convective random walk, which can explain bringing together separated
chemical markers, in this case isotopically diverse zircons from roof rock. (B) Localized,
closely spaced distribution of chemical markers (e.g., δ18O, shown as veins and squiggly
areas), with spacing of tens of meters. (C) Large-scale gradient distribution of δ18O over
hundreds of meters to several kilometers. Zircon crystallized from different δ18O areas (red
and yellow) are brought together in one area by convection.
18
Geosphere, October 2014
We next consider mixing effectiveness within
merged batches. Many papers in the past 50 yr
have documented abundant natural examples
of efficient magma mixing and mingling on
short time scales of magma chamber refill, or
on mineral dissolution time scales for a variety
of compositions. Examples including mixed
and unmixed ignimbrites are also abundant
(for recent numerical treatment of this problem, see Huber et al., 2009, 2012; for a review
of diffusion time scales, see Cooper and Kent,
2014). For example, buoyant overturn caused
by bubbles demonstrates the efficiency of this
process on the order of hours or days (Ruprecht
et al., 2008), and work in the Taupo volcanic
zone demonstrates effective rhyolite mixing on
eruptive time scales (Wilson and Charlier, 2009;
Allan et al., 2013).
Mixing Within a Single Magma Batch
In these sets of numerical experiments we
studied the dynamics of the magma reservoir
with fine spatial resolution, down to several
centimeters, sufficient to resolve fine flow
structures in the compositional field and thus
evaluate goodness of mixing (Figs. 13 and 14;
see Supplemental File 77). This effort continues
from Simakin and Bindeman (2012), where
coupled Navier-Stokes and heat transfer equations were solved with realistic phase diagrams
and proper viscosity models. We observed in
these and previous melting experiments that
convective melting induces vigorous rates
of compositional convection, significantly
exceeding the intensity of thermal convection
in the solidifying magma chamber. This is
essentially due to the high efficiency of in situ
melting in generating density instabilities, as
compared with the far-field control of chamber
cooling. Although somewhat counterintuitive,
7
Supplemental File 7 is a temporal evolution of
vorticity distribution, which is responsible for mixing. If you are viewing the PDF of this paper or
reading it offline, please visit http://dx.doi.org/10
.1130/GES00969.S7 or the full-text article on www
.gsapubs.org to view Supplemental File 7.
00969 1st pages / page 19 of 27
Rhyolites—hard to produce but easy to recycle
A
m
B
m
ment δXm is plotted as a function of time for the
averaged trajectories of 25 markers, originally
located in the same square. During the first
2 days, this parameter oscillates around 1 m,
which corresponds to the lifetime of the individual plume (vortex), where particles move vertically with small periodic horizontal displacements. An abrupt rise of δXm corresponds to
the moment of pulling apart the magma parcel.
Later, the δXm increases, yielding oscillations
with periods of 2.5–2 days and average rates of
~0.5 m/day.
Dependence δXm on time (t) after 10 days is
nearly linear with a rate of 0.5 m/day (Fig. 15B;
R 2 = 0.969). However, square root dependence
on time is expected on the basis of physical 1D
Markov chains model with diffusive stochastic
jumps (Levin et al., 2008). If so, the diffusional
relationship allows the definition of the marker
diffusion coefficient:
δX = (Dt)1/ 2, D = δx 2/δt,
Figure 14. Distribution of horizontal displacements δX(x,y) of the 5 × 5 cm averaged markers
after days of convective melting in the rhyolitic magma chamber (see Fig. 12B). (A) After
5 days. (B) After 17 days. Images are 25 × 50 m and color scale signifies the magnitude of
horizontal displacement in meters. Brown color corresponds to the surrounding unmelted
(static) rocks, static condition, δX(x,y) = 0; green stripes correspond to the maximum relative horizontal displacement of markers. Although patterns appear as convective flow lines,
they here represent goodness of horizontal mixing. Large-scale flows are generated near the
left steep edge of the chamber. Although the current model does not include a chemical diffusional transport or crystal-melt separation, it is expected that efficiency of diffusive compositional mixing is a direct function of the total travel distance traveled by markers; their
prolonged exposure to larger volume of melt will be completely homogenize melt within
each 5 × 5 cm volume.
it is thus the melting process that leads to convection and homogenization inside of magma
bodies (Fig. 14; see Supplemental Files 1, 2,
and 3 [see footnotes 1, 2, and 3]). We observed
(Simakin and Bindeman, 2012) that any roof
heterogeneity in density or compositional character (e.g., higher melting temperature blocks)
will release gravitational energy and induce
mixing upon downward movement through the
magma body.
Here we pay greater attention to the motion
and mixing in the horizontal direction responsible for the homogenization in sills with large
aspect ratios (D/H), relevant for magma bodies parental to large-scale ignimbrite eruptions. Thermocompositional plumes chaotically
descending from the roof cause rotations of
the multiphase fluid, and are characterized by
a spotty vorticity distribution. We observe that
due to relatively small scale stochastic flow,
liquid particles caught by successive vortexes
rotate in different directions. Their horizon-
tal motion and horizontal displacement can be
approximately represented by random jumps
with steps proportional to a vortex diameter in
size. An example of the distribution of the absolute difference between initial and final horizontal coordinates, i.e.,
∑( x
j1/4
δX(x, y,τ) =
m
where δx is the size of the jump each δt time
units, and D is the effective diffusion coefficient.
By using δXm (t) data for time >4 days we can
derive the square root of time approximation
with diffusion coefficient, D = 4 m2/day. These
are obviously very fast rates as compared to the
chemical diffusion of SiO2 or ZrO2 in rhyolitic
magma, 10–9 cm2/day (Baker, 1991; Harrison
and Watson, 1983).
Another way to treat mixing is to apply average separation between the markers approach
(e.g., Ruprecht et al., 2008), as implemented
for magma chambers and mantle mixing problems. Average separation is defined as the average distance between initially neighboring
pairs of material points in some domain Ω with
NΩ points:
δR(t) =
∑ ( x (t) − x
Ω
)
(t = 0) − xm (t = τ) nj1/4 ,
(10)
of the magma markers inside a model sill of
7 × 24 m is presented in Figure 13, recorded in 5
and 17 days from the beginning of the numerical
experiment. Absolute displacements are averaged over a subgrid with size equal to one-half
of the velocity elements (one-quarter by area).
Trajectories of 6 markers initially placed
in the filled square are mapped after 8 days in
Figure 15A. Markers first move coherently and
then spread across the studied magma domain.
In Figure 15B, absolute horizontal displace-
Geosphere, October 2014
(11)
i
i+1
)
)
(t) + ( yi (t) − yi+1 (t) NΩ . (12)
2
2
This parameter δR as a function of time was
calculated for 16,000 markers initially located
inside the 1 × 1m square shown in Figure 15A,
and results of the numerical run are plotted in
Figure 15C. Until ~4 days, separation almost
does not change; in the time interval 4–11 days
it undergoes exponential stretch, and demonstrates square root dependence on time afterward. Ruprecht et al. (2008) treated separation
as an exponential function of time, using a
Lyapunov parameter formulation with exponent
defined as:
)
λ(t) = lim ro −>0 ln ( δR(t)/r0 (t = 0) t. (13)
19
00969 1st pages / page 20 of 27
Bindeman and Simakin
Height, m
A
Width, m
14
B
12
δ X, m
10
y = 1.6t-1.1
2
R =0.969
8
6
1/2
y=(36.8t) - 6
4
R =0.903
2
2
0
0
2
4
6
8
10
time, days
C
0.8 Lyapunov exponent:
40
σ day −1
0.6
,
0.4
0.2
0.0
δ X, m
30
0
20
4
8
12
time, days
16
square root
10
induction
0
exponential
0
4
8
12
16
Figure 15. Redistribution of markers in the course
of convection. (A) Trajectory of markers initially
located inside blue box (1x1 m). Dotted yellow line
depicts part of trajectory where markers move
together in the same plume (as in Fig 13A and
in “induction” period in C), before being pulled
apart by convection. (B) Horizontal distance δX
between two neighboring markers, averaged for
25 markers, initially located in blue box in A,
averaged as a function of time, t, plotted as filled
diamonds. Long-term divergence trends can be
approximated by the square root (solid red line)
or linear (dashed red line) functions with nearly
equal accuracy. (C) Average separation between
initially neighboring pairs of points, calculated for
1640 markers initially located inside the 1 × 1 m
blue box in A. Note the change in functional dependence on time; initially, markers travel with the
blue box for 4 days without separation (induction), then they undergo exponential separation.
Inset shows computed Lyapunov exponent for
separation; after that, square root separation is
a better approximation. See text for discussion
and implications. A–C represent particle separation without side-wall effects. In the closed-box
experiment representing small magma chambers,
mixing may reach an asymptotic limit for longer
time scales due to side-wall effects defined by the
magma pool width. It can be shown that during
perfect mixing δXm will approach a value of W/3,
where W is width of the pool.
time, days
This temporal variation of the exponent
parameter in our experiments is displayed in
the Figure 15C inset. The Lyapunov exponent
attains a maximum plateau with a level of 0.6
day –1 that characterizes the intensity of convective mixing in our case. Obviously λ decreases
with time, since convective velocities are finite
while constant λ implies growth of average separation with an exponentially increasing rate.
20
We observe in our runs that exponential mixing
with the Lyapunov exponent is only a transient
phenomenon.
The random walk model with either square
root time or linear dependence allows for extension of the results obtained in the restricted space
domain toward the geologically more relevant
larger dimensions and larger magma bodies. In
particular, retaining the same dynamic parame-
Geosphere, October 2014
ters with a simple increase in the size of magma
chamber results in a linear regression prediction of δX = 1 km in 2 yr. The diffusional model
predicts 1 km dispersion in 684 yr. However, as
dispersion (mixing) occurs in parallel with melting (Figs. 8, 13, and 14), horizontal parameters
are also affected by the overall rates of displacement of the convective body of the magma. During 2 yr of roof melting, the melting front will
00969 1st pages / page 21 of 27
Rhyolites—hard to produce but easy to recycle
displace the solid-melt boundary by 6–20 m,
given modeled melting rates, um, of 4–12 m/yr;
during 684 yr of melting, the magma body may
migrate 300–1000 m, comparable with overall
vertical chamber size, further contributing to
homogenization.
It is likely that this model of displacement
due to repeated rotations in random directions in descending plumes provides a minimum proxy for chemical mixing and horizontal homogenization in magma bodies. This is
because it neglects the contribution of (1) the
curved trajectories of the descending interacting plumes, and (2) occasional larger scale
flows (magmatic avalanches), as seen in some
runs (See Supplemental Files 1, 2, and 3 [see
footnotes 1, 2, and 3]) and as widely speculated to occur in nature (e.g., layered intrusions;
Wager and Brown, 1967). In the latter case,
horizontal flows are occurring with significant
horizontal velocities of meters per hour, much
faster than either the linear or ~t 1/2 horizontal spreading rates mentioned here. However,
compositional homogenization on a centimeter
scale is ensured by the slower and smaller scale
motions. Efficiency of mixing is also dependent
on the amplitude and spatial distribution of
tracer heterogeneity (e.g., Figs. 14, 16, and 17),
whether large scale or small scale. The handspecimen isotopic homogeneity is achieved
when isotopically distinct individual minerals
such as zircons with diverse εHf and δ18O (Figs.
4 and 14) are brought together from areas that
were originally tens of meters apart.
In summary, our numerical results demonstrate that thorough compositional mixing is
easy, and proceeds on time scales faster than
cooling, but slower than mineral diffusive time
scales. The process generates homogeneous
whole-rock composition with isotopically
diverse crystalline cargo, including zircons.
Comparison with Natural Data on Mixing
There are attempts to study convective
homogenization in a magma reservoir in experiments with real silicate melts. De Campos et al.
(2011) experimentally performed mixing of siliceous and basaltic magmas in a high-T cell. The
experimental setup with independently rotated
noncoaxial cylinders generated a chaotic (nonperiodic) flow field (Fig. 17). Mingling in the
experiments at the large viscosity contrast of a
factor of ~1000 proceeds through formation of
the continuously twisted magma layers rather
than separated drops. The textures produced are
similar to the commonly observed interlayering of dark and light bands of glasses in obsidians. Geologic observations (e.g., Slabunov,
1995) show us that partially melted xenoliths
of amphibole-bearing schist in an andesitic melt
Figure 16. Size distributions of parcels of melted roof rocks in the main convecting volume
of magma sill (see Figs. 13A and 15) in numerical melting runs after 5.5 and 17 days in 15 ×
80 m convecting magma body. The convective model does not consider chemical diffusion;
in reality the small-scale heterogeneities will be annealed by diffusion. Size distribution of
partially molten xenoliths in the analogous natural process is shown (data from Slabunov,
1995). In the inset image of numerical mixing after 17 days, dark indicates pieces of the
melted roof; size of the area is 1 × 1 m. L is the largest axis of the ellipsoidal inclusions, and
n/nmax is normalized on the maximum value of particles size distribution.
are characterized by elliptic shapes in section
and a unimodal distribution of the their short
axes with median values of 1.5–2 cm, a mean
of ~5 cm, and a maximum of 18 cm. In our
numerical runs we observe comparable size distributions of the partially molten roof material
scattered in the lower convecting rhyolite (see
Fig. 16, inset). This means that our experiments
correctly reproduce the initial stage of homogenization, continuing with the fast diffusive
dissolution of the smallest drops. Our results
contradict earlier published conclusions of the
small efficiency of the convection in a magma
reservoir and its inability to achieve real mixing
(Gutierrez and Parada, 2010). Both natural data
(Vazques et al., 2009; Girard and Stix, 2010)
and numerical models (Huber et al., 2009; Burgisser and Bergantz, 2011) demonstrate high
efficiency of mixing, extraction, and differentiation in crystallizing rhyolites, and relative crystal-melt separation promotes local scale mixing.
We thus consider that the large-scale merging
of magma batches that we modeled here creates
many possibilities for compositional convection
Geosphere, October 2014
and mixing. A three-layer melting experiment in
Figure 8 further illustrates how compositional
heterogeneities are efficiently mixed when
magma batches are merged together, in this case
by melting through the screen rock between.
DISCUSSION
Case for Short-Lived Batch Magma Bodies
in the Stress Field of the Upper Crust
Local high-resolution seismic tomography
(e.g., Fig. 18) provides information about the
depths, sizes, and approximate shapes of possible magma bodies under Yellowstone and
elsewhere (Beachly et al., 2012). Miller and
Smith (1999) and Husen et al. (2004) imaged
two separate regions of low Vp under the Sour
Creek (Yellowstone) resurgent domes at the
same depth of 7 km (Fig. 18). The isotopically
diverse crystal cargoes in eruptive products best
argues that that the supposed magma chambers
in the SRP and elsewhere are short lived and
ephemeral, like those shown in Figure 18, rather
21
00969 1st pages / page 22 of 27
Bindeman and Simakin
A
than a single convecting batholith. Provided
enough basalt and heat flux is maintained, dispersed magma bodies can be stable in this form,
crystallize, and be hydrothermally altered, or
rapidly reactivated. Thermomechanical stability analysis of widely dispersed melt pockets is
outside the scope of this paper, but instability
can be triggered by a greater supply of basalt
(thermal reactivation), or by changing stress
conditions (tectonic reactivation). Our modeling demonstrates that formation and mixing of
large, eruptable magma volumes of 102–103 km3
may proceed rapidly by either amalgamation of
these numerous bodies or continuous growth of
one or several large ones, consuming the others.
C
B
Accounting for the Amount of
Silicic Magma Produced
Figure 17. Comparison of the numerical and experimental mixing of two liquids during
rotation. (A) Initial configuration of a circular chamber with a solid, noncoaxial rotor and a
small circular liquid inclusion with identical viscosity. (B) Numerical structure after two full
rotations. (C) Experimental texture of glycerin with dye after 10 alternate rotations (from
De Campos et al., 2011). In both cases stretching and folding form multiple, thin layers
rather than separate blobs.
A
The estimates of the volcanic/plutonic ratio
for the erupted rhyolites vary widely in any volcanic area, from ~10% erupted, 90% remaining
(Smith, 1979), to 1:2 (Bonnichsen et al., 2008),
or even 1:1, as we favor here due to efficient
near-eutectic melting and extraction. These estimates affect the corresponding amount of heat
and basalt required to produce the observed
amounts of silicic melt. Basalt differentiation or
partial melting of the mafic crust (slow process;
Fig. 6), or a combination of the two of these
processes, is a strong function of the melting
efficiency, i.e., how much heat is wasted for
heating of the country rock prior to and during
B
δ18O
Depth, km
high
low
C
fractal
reservoir
assembly
Figure 18. Seismic tomography data (Miller and Smith, 1999; Husen et al., 2004). (A) Yellowstone. (B) Isotopically diverse and dispersed
batch magma bodies based on the results of this work.
22
Geosphere, October 2014
00969 1st pages / page 23 of 27
Rhyolites—hard to produce but easy to recycle
What Remains in the Crust
After Recycling?
The volcanic-tectonic processes that we
envision to characterize the evolution of an
SRP-type environment leave behind basaltified,
topographically reduced (Fig. 1), and bimodal
terrain in the wake of the plume. Internally, it
should lead to the complex intracrustal structures of dikes and sills, not common in typical
clean granite plutonic examples such as those
of the Sierra Nevada (California). However,
plutons represent significantly prolonged histories after cessation of volcanism (Paterson and
Vernon, 1995; Zak et al., 2007). In particular,
multi-cyclic caldera
zoned ignimbrites
Melt3 LSR
Brittle-Ductile
Brittle-D
transit
transition
PATHWAY
PLUTONS
periodic rapid remelting
batch assembly, mixing, eruption
Melt3d HSR
RECYCLING
BATCHES
Melt 3 ... N
SEQUESTRATION
Hydroth
e
precon rmal
ditionin
g
partial melting (e.g., Annen et al., 2006; Dufek
and Bergantz, 2005). Also, many such models
do not take into account temperature-dependent
thermal diffusivities and/or conductivities,
which, if incorporated, would reduce the heat
contribution needed from basalts by a factor of
~2 (Whittington et al., 2009). Hot zone models
require significant amounts of mafic cumulate to
be produced at the end (Fig. 19), often in excess
of the observed thickness of the lower crust,
requiring that these cumulates delaminated and
sank into the mantle. Thus the room problem
exists not only for silicic plutons in the crust,
but also for parental cumulates and restites in
the lower crust and upper mantle. In addition,
the material versus heat contribution of basalt is
one of the longest-discussed problems of igneous petrology (e.g., Hildreth and Moorbath,
1988; Grunder, 1995; and many others).
However, all of these mass, space, and heat
constraints are significantly relaxed if silicic
rocks are produced by remelting and recycling
of already existing silicic magma bodies, which
were formed by any of the above mechanisms.
As we advocate here, such remelting is easy
and fast (fast process; Fig. 5), and leads to the
sequestration of silicic melts into the upper
crust. For example, in the Andes, the U-Pb zircon evidence suggests that nonerupted parts
of older eruptions in the same magmatic system or sequence are commonly recycled (e.g.,
Folkes et al., 2011), and a similar process must
also characterize the multicyclic San Juan volcanic field of Colorado, USA (Lipman, 2007),
although we note that these systems appear not
to be fingerprinted by low δ18O values, indicating recycling of previously erupted and altered
material (e.g., Folkes et al., 2013). Much recent
work has been done on Taupo volcanic zone in
New Zealand, also pointing to zircon recycling
and short residence times of magma segregation
(Charlier et al., 2008; Wilson and Charlier, 2009;
Allan et al., 2013): oxygen isotopic evidence on
recycling there is still to be investigated.
Upper Crust
Melt2d
Melt2
Lower Crust
dynamic melt segregation
dy
Melt1
Melt
t1
“Hot Zone”
Basaltic Magma Flux
Flowing, delaminating Cumulates
Figure 19. Rhyolite genesis by progressive recycling and crustal evolution by upward sequestration of silicic material (Melt1.3d HSR;
d stands for differentiated) leading to low-silica rhyolites (LSR)
and high-silica rhyolites (HSR) in the upper crust. Each episode of
recycling involves melting, segregation, and differentiation of previously emplaced silicic material (fingerprinted by locally diverse isotopic values), simplifying the room and heat problem. High magma
supply rates yield effective sequestration and large volcanic/plutonic
ratios, and create transient, short-lived, batch magma chambers in
the upper crust (see Fig. 18).
large screens of intracaldera roof rocks may
sink to the bottom of a lower magma body, as
well as result in mafic magma recharge. Furthermore, plutons may undergo doming and upward
motion on the order of 105–106 m.y. time scales,
creating internal density redistribution (Ramberg, 1970; Petford, 2003; Schubert et al.,
2013). Here we emphasize an important conclusion: observations from plutonic examples
significantly postdate the stages of magmatic
evolution documented by the volcanic evidence
(or processes inferred for the volcanic magma
chamber), and thus the observational records
may not be compatible (e.g., Glazner et al.,
2004; Coleman et al., 2004) (e.g., the geophysically imaged gabbro or gabbro-dioritic sill that
remains in the crust after passage of the Yellowstone plume; Husen et al., 2004; DeNosaquo
et al., 2009). Discordance between plutonic and
volcanic records should be expected because
plutonic rocks represent significant posteruptive
Geosphere, October 2014
thermal evolution. So, links between plutonic
and volcanic records should not be discarded
just because there is chronological discordance.
Predicting Yellowstone’s Future
From Studying its Past
The causes of the production and eruption of
magmas in the SRP (Fig. 1), have direct implications for the 620 ka Yellowstone caldera, and
in particular on whether it represents the last of
a series of eruptions in a caldera cluster (e.g.,
Watts et al., 2012) and a crustal block that has
been remelted and recycled from top to bottom,
and thus exhausted its eruptive potential, or if
generation and eruption of large volumes of
silicic magma is still possible. In other words, it
is important to estimate if the current (past 600
k.y.) voluminous rhyolitic eruptions at Yellowstone represent the path of a dying cycle, and
if a new Huckleberry Ridge Tuff–size magma
23
00969 1st pages / page 24 of 27
Bindeman and Simakin
reservoir is currently being slowly established in
southwest Montana with anticipated completion
in 1–2 m.y. If so, the next eruption is to be normal δ18O, assembled slowly and with difficulty
from the lower crust and differentiation of SRP
basalt, with limited zircon diversity.
The low δ18O magmas with diverse zircons
(Figs. 2–4) that appeared at the end of each caldera cluster evolution (and throughout the central SRP) for at least the past 12 m.y., and likely
as early as 15 Ma (Colon et al., 2014) are identified as derived by recycling of buried silicic (zircon saturated) hydrothermally altered volcanic
and plutonic predecessors from the same crustal
block. Relative similarities in the radiogenic
isotope ratios and trace elemental abundances
in early normal δ18O and subsequent low δ18O
rhyolites (Hildreth et al., 1991; Hanan et al.,
2008; Christiansen and McCurry, 2008; Drew
et al., 2013) suggest that the low δ18O rhyolites
are derived by recycling and remelting of the
normal δ18O rhyolites after they became depleted
in 18O by high-temperature alteration, which did
not disturb, or minimally affected elemental and
other isotopic abundances. Given that Yellowstone has already erupted >600 km3 of recycled
low δ18O rhyolites with diverse crystal cargo for
the past 620 k.y., we consider that it is on a dying
cycle, and that the crustal block under Yellowstone has been depleted in silicic melt, which
was sequestered upward and recycled.
Big Picture View of Silicic Magma
Petrogenetic Models and Crustal
Evolution by Sequestration
The large-scale recycling model of rhyolite
petrogenesis in the upper crust that we propose
here (Fig. 19) serves as a previously underappreciated, important end-member to existing
models of formation of silicic magmas in the hot
zone in the lower crust during successive mafic
sill intrusions (Hildreth and Moorbath, 1988;
Huppert and Sparks, 1988; Annen and Sparks,
2002; Dufek and Bergantz, 2005; Annen et al.,
2006), and the resulting assembly of large silicic
magma bodies or batholiths in the middle and
upper crust (Schaltegger et al., 2009). Extraction of the partial silicic melt from lower crustal
mafic protoliths (slow and difficult process
above and in Fig. 7) is facilitated by the slow
coeval deformation (e.g., Spiegelman, 2003;
Rabinowicz and Vigneresse, 2004; Simakin and
Talbot, 2001). The batholiths may be initially
mushy or fully crystalline (e.g., Coleman et al.,
2004; Bachmann and Bergantz, 2004; Allan
et al., 2013), or may not be batholiths at all,
but an agglomeration of segregated silicic sills
and stocks (Fig. 19; Horseman et al., 2009).
Our model of upward sequestration (fast and
24
easy process) then requires that these already
segregated areas are remelted more than once,
each time accompanied by further fractionation that increases crustal differentiation indices (e.g., increasing Rb/Sr, Eu anomalies, and
SiO2, decreasing MgO). Each time, melting of
siliceous lithologies proceeds rapidly; the 50%
melt extraction window (e.g., Fig. 5) is achieved
quickly as the heat of intruding hot silicic differentiates is very efficiently consumed by the
eutectic melting, without much heat loss on
rapid melting time scales. Basaltic magma may
also intrude the upper crust along the magma
pathways (Fig. 19; e.g., Fowler and Spera,
2010), delivering heat but offsetting the progress of sequestration. Rhyolitic bodies are then
segregated in large volumes by upward sequestration, and our modeling demonstrates easy
assembly, amalgamation, and effective mixing
between disparate magma bodies.
We point to the observation that bimodal
basaltic-rhyolitic provinces predominantly
occur within extensional and hotspot environments with high basaltic power outputs, but
also in continental arc areas with delamination
features. Rising basaltic magmas are trapped by
density filters of silicic magma batches at different levels in the crust, leading to a Daly Gap
(i.e., a lack of intermediate compositions; Seligman et al., 2014). The described process (Fig.
19) furthermore leads to crustal sequestration
of silicic material in the upper crust and mafic
cumulates in the lower crust. We further speculate that because high basaltic output conditions
were predominant on the early Earth, the recycling and sequestration processes were responsible for the segregation of the upper silicic and
lower mafic crust, with significant components
of thermomechanical lower crustal convection
(e.g., Gerya, 2009). A dynamic view of silicic
magma petrogenesis by sequestration is a must.
ACKNOWLEDGMENTS
Supported by National Science Foundation EARCAREER-0844772 and EAR-1049351, and Russian
Foundation for Basic Research grant 07-05-00629.
We thank Matthew Loewen, Dana Drew, Dylan
Colon, and Taras Gerya for comments, Chris Folkes,
Graham Andrews, Ben Ellis, and two anonymous
reviewers for their extensive comments, and Shan de
Silva for helpful editorial handling.
REFERENCES CITED
Allan, A.S.R., Wilson, C.J.N., Millet, M.-A., and Wysoczanski, R.J., 2012, The invisible hand: Tectonic triggering
and modulation of a rhyolitic supereruption: Geology,
v. 40, p. 563–566, doi:10.1130/G32969.1.
Allan, A.S.R., Morgan, D.J., Wilson, C.J.N., and Millet,
M.A., 2013, From mush to eruption in centuries:
Assembly of the super-sized Oruanui magma body:
Contributions to Mineralogy and Petrology, v. 166,
p. 143–164, doi:10.1007/s00410-013-0869-2.
Annen, C., and Sparks, R.S.J., 2002, Effects of repetitive
emplacement of basaltic intrusions on thermal evolu-
Geosphere, October 2014
tion and melt generation in the crust: Earth and Planetary Science Letters, v. 203, p. 937–955, doi:10.1016
/S0012-821X(02)00929-9.
Annen, C., Blundy, J.D., and Sparks, R.S.J., 2006, The genesis of intermediate and silicic magmas in deep crustal
hot zones: Journal of Petrology, v. 47, p. 505–539, doi:
10.1093/petrology/egi084.
Bachmann, O., and Bergantz, G.W., 2004, On the origin of
crystal-poor rhyolites: Extracted from batholithic crystal mushes: Journal of Petrology, v. 45, p. 1565–1582,
doi:10.1093/petrology/egh019.
Bacon, C.R., and Lowenstern, J.B., 2005, Late Pleistocene
granodiorite source for recycled zircon and phenocrysts in rhyodacite lava at Crater Lake, Oregon: Earth
and Planetary Science Letters, v. 233, p. 277–293, doi:
10.1016/j.epsl.2005.02.012.
Bacon, C.R., Adami, L.H., and Lanphere, M.A., 1989,
Direct evidence for the origin of low-δ18O silicic magmas: Quenched samples of a magma chamber’s partially fused granitoid walls, Crater Lake, Oregon: Earth
and Planetary Science Letters, v. 96, p. 199–208, doi:
10.1016/0012-821X(89)90132-5.
Baker, D.R., 1991, Interdiffusion of hydrous dacitic and
rhyolitic melts and the efficacy of rhyolite contamination of dacitic enclaves: Contributions to Mineralogy and Petrology, v. 106, p. 462–473, doi:10.1007
/BF00321988.
Barton, N., 2013, Shear strength criteria for rock, rock joints,
rockfill and rock masses: Problems and some solutions:
Journal of Rock Mechanics and Geotechnical Engineering, v. 5, p. 249–261, doi:10.1016/j.jrmge.2013.05.008.
Beachly, M.W., Hooft, E.E., Toomey, D.R., and Waite, G.P.,
2012, Upper crustal structure of Newberry Volcano
from P-wave tomography and finite difference waveform modeling: Journal of Geophysical Research,
v. 117, B10311, doi:10.1029/2012JB009458.
Beard, J.S., and Lofgren, G.E., 1991, Dehydration melting
and water saturated melting of basaltic and andesitic
greenstones and amphibolites at 1–3 and 6–9 kbar:
Journal of Petrology, v. 32, p. 365–401, doi:10.1093
/petrology/32.2.365.
Beard, J.S., Ragland, P.C., and Crawford, M.L., 2005, Reactive bulk assimilation: A model for crust-mantle mixing in silicic magmas: Geology, v. 33, p. 681–684, doi:
10.1130/G21470.1.
Becker, T.T., 2006, On the effect of temperature and strainrate dependent viscosity on global mantle flow, net
rotation, and plate-driving forces: Geophysical Journal
International, v. 167, p. 943–957.
Bindeman, I.N., 2003. Crystal sizes in evolving silicic
magma chambers: Geology, v. 31, p. 367–370.
Bindeman, I.N., and Valley, J.W., 2001, Low-δ18O rhyolites
from Yellowstone: Magmatic evolution based on analyses of zircons and individual phenocrysts: Journal of
Petrology, v. 42, p. 1491–1517, doi:10.1093/petrology
/42.8.1491.
Bindeman, I.N., Schmitt, A.K., and Valley, J.W., 2006, U-Pb
zircon geochronology of silicic tuffs from the Timber Mt/Oasis Valley caldera complex, Nevada: Rapid
generation of large-volume magmas by shallow-level
remelting: Contributions to Mineralogy and Petrology,
v. 152, p. 649–665, doi:10.1007/s00410-006-0124-1.
Bindeman, I.N., Watts, K.E., Schmitt, A.K., Morgan, L.A.,
and Schanks, P., 2007, Voluminous low δ18O magmas in
the late Miocene Heise volcanic field, Idaho: Implications for the fate of Yellowstone hotspot calderas: Geology, v. 35, p. 1019–1022, doi:10.1130/G24141A.1.
Bindeman, I.N., Brooks, C.K., McBirney, A.R., and Taylor,
H.P., 2008a, The low-δ 18O late-stage ferrodiorite
magmas in the Skaergaard Intrusion: Result of liquid
immiscibility, thermal metamorphism, or meteoric
water incorporation into magma?: Journal of Geology,
v. 116, p. 571–586, doi:10.1086/591992.
Bindeman, I.N., Fu, B., Kita, N.T., and Valley, J.W., 2008b,
Origin and evolution of Yellowstone silicic magmatism
based on ion microprobe analysis of isotopically-zoned
zircons: Journal of Petrology, v. 49, p. 163–193, doi:10
.1093/petrology/egm075.
Bindeman, I.N., Gurenko, A., Carley, T., Miller, C., Martin,
E., and Sigmarsson, O., 2012, Silicic magma petrogenesis in Iceland by remelting of hydrothermally-altered
crust based on oxygen isotope diversity and disequilib-
00969 1st pages / page 25 of 27
Rhyolites—hard to produce but easy to recycle
ria between zircon and magma with implications for
MORB: Terra Nova, v. 24, p. 227–232, doi:10.1111/j
.1365-3121.2012.01058.x.
Bittner, D., and Schmeling, H., 1995, Numerical modelling of melting processes and induced diapirism in the
lower crust: Geophysical Journal International, v. 123,
p. 59–70, doi:10.1111/j.1365-246X.1995.tb06661.x.
Bonnichsen, B., Leeman, W.P., Honjo, N., McIntosh, W.C.,
and Godchaux, M.M., 2008, Miocene silicic volcanism
in southwestern Idaho: Geochronology, geochemistry,
and evolution of the central Snake River Plain: Bulletin
of Volcanology, v. 70, p. 315–342, doi:10.1007/s00445
-007-0141-6.
Boroughs, S.B., Wolff, J., Bonnichsen, B., Godchaux, M.,
and Larson, P., 2005, Large volume, low-δ18O rhyolites
of the central Snake River Plain, Idaho, USA: Geology,
v. 33, p. 821–824, doi:10.1130/G21723.1.
Bowen, N.L., 1928, The evolution of the igneous rocks:
Princeton, New Jersey, Princeton University Press,
334 p.
Bowman, J.R., Willett, S.D., and Cook, S.J., 1994, Oxygen
isotopic transport and exchange during fluid-flow—
One-dimensional models and applications: American
Journal of Science, v. 294, p. 1–55, doi:10.2475/ajs
.294.1.1.
Brandeis, G., and Marsh, B.D., 1990, Transient magmatic
convection prolonged by solidification: Geophysical
Research Letters, v. 17, p. 1125–1128, doi:10.1029
/GL017i008p01125.
Burgisser, A., and Bergantz, G.W., 2011, A rapid mechanism to remobilize and homogenize highly crystalline
magma bodies: Nature, v. 471, p. 212–215, doi:10
.1038/nature09799.
Buttinelli, M., De Rita, D., Cremisini, C., and Cimarelli, C.,
2011, Deep explosive focal depths during maar forming magmatic-hydrothermal eruption: Baccano Crater,
central Italy: Bulletin of Volcanology, v. 73, p. 899–
915, doi:10.1007/s00445-011-0466-z.
Caricchi, L., Annen, C., Blundy, J., Simpson, G., and Pinel,
V., 2014, Frequency and magnitude of volcanic eruptions controlled by magma injection and buoyancy:
Nature Geoscience, v. 7, p. 126–130, doi:10.1038
/ngeo2041.
Carley, T.L., Miller, C.F., Wooden, J.L., Bindeman, I.N.,
and Barth, A.P., 2011, Zircon from historic eruptions
in Iceland: Reconstructing storage and evolution of
silicic magmas: Mineralogy and Petrology, v. 102,
p. 135–161, doi:10.1007/s00710-011-0169-3.
Carlino, S., Cubellis, E., Luongo, G., and Obrizzo, F.,
2006, On the mechanics of caldera resurgence of
Ischia Island (southern Italy), in Troise, C., et al., eds.,
Mechanisms of activity and unrest at large calderas:
Geological Society of London Special Publication 269,
p. 181–193, doi:10.1144/GSL.SP.2006.269.01.12.
Cathey, H.E., and Nash, B.P., 2004, The Cougar Point Tuff:
Implications for thermochemical zonation and longevity
of high-temperature, large-volume silicic magmas of
the Miocene Yellowstone hotspot: Journal of Petrology,
v. 45, p. 27–58, doi:10.1093/petrology/egg081.
Cathey, H.E., Nash, B.P., Seligman, A.N., Valley, J.W., Kita,
N., Allen, C.M., Campbell, I.H., Vazquez, J.A., and
Wooden, J.L., 2011. Low δ18O zircons from the Bruneau-Jarbidge eruptive center: A key to crustal anatexis
along the track of the Yellowstone hotspot: American
Geophysical Union, Fall meeting, abs. V11A–2510.
Cecconi, M., Scarapazzi, M., and Viggiani, G.M.B., 2010,
On the geology and the geotechnical properties of
pyroclastic flow deposits of the Colli Albani: Bulletin
of Engineering Geology and the Environment, v. 69,
p. 185–206, doi:10.1007/s10064-009-0250-x.
Charlier, B.L.A., Wilson, C.J.N., and Davidson, J.P., 2008,
Rapid open-system assembly of a large silicic magma
body: Time-resolved evidence from cored plagioclase crystals in the Oruanui eruption deposits, New
Zealand: Contributions to Mineralogy and Petrology,
v. 156, p. 799–813, doi:10.1007/s00410-008-0316-y.
Christiansen, E.H., and McCurry, M., 2008, Contrasting origins of Cenozoic silicic volcanic rocks from the western
Cordillera of the United States: Bulletin of Volcanology,
v. 70, p. 251–267, doi:10.1007/s00445-007-0138-1.
Christiansen, R.L., Lowenstern, J.B., Smith, R.B., Heasler,
H., Morgan, L.A., Nathansen, M., Mastin, L.G.,
Muffler, L.J.P., and Robinson, J.E., 2007, Preliminary
assessment of volcanic and hydrothermal hazards in
Yellowstone National Park and vicinity: U.S. Geological Survey Open-File Report 2007-1071, 94 p.
Coble, M.A., and Mahood, G.A., 2012, Initial impingement
of the Yellowstone plume located by widespread silicic
volcanism contemporaneous with Columbia River
flood basalts: Geology, v. 40, p. 655–658, doi:10.1130
/G32692.1.
Cole, D.R., and Chakraborty, S., 2001, Rates and mechanisms of isotope exchange: Reviews in Mineralogy and
Geochemistry, v. 43, p. 83–223, doi:10.2138/gsrmg.43
.1.83.
Coleman, D.S., Gray, W., and Glazner, A.F., 2004, Rethinking the emplacement and evolution of zoned plutons:
Geochronologic evidence for incremental assembly
of the Tuolumne Intrusive Suite, California: Geology,
v. 32, p. 433–436, doi:10.1130/G20220.1.
Colon, D., Bindeman, I.N., Ellis, B.S., Schmitt, A.K., Fisher,
C., Vervoort, J., Mark, D., and Brueseke, M., 2014,
Hydrothermal alteration and batch melting of crust
by Columbia River Basalt magmas, fingerprinted by
post-CRB rhyolites of the J-P Desert and the Jarbidge
Mountains, Idaho and Nevada, USA: Earth and Planetary Science Letters, v. 403 (in press).
Cooper, C.M., and Kent, A.R., 2014, Rapid remobilization
of magmatic crystals kept in cold storage: Nature,
v. 506, p. 480–483, doi:10.1038/nature12991.
Davaille, A., and Jaupart, C., 1993, Transient high-Rayleighnumber thermal convection with large viscosity variations: Journal of Fluid Mechanics, v. 253, p. 141–166,
doi:10.1017/S0022112093001740.
Davidson, J., Tepley, F., III, and Palacz, Z., 2001, and
Meffan-Main, S., 2001, Magma recharge, contamination and residence times revealed by in situ laser ablation isotopic analysis of feldspar in volcanic rocks:
Earth and Planetary Science Letters, v. 184, p. 427–
442, doi:10.1016/S0012-821X(00)00333-2.
De Campos, C.P., Perugini, D., Ertel-Ingrisch, W., Dingwell,
D.B., and Poli, G., 2011, Enhancement of magma mixing efficiency by chaotic dynamics: An experimental
study: Contributions to Mineralogy and Petrology,
v. 161, p. 863–881, doi:10.1007/s00410-010-0569-0.
DeNosaquo, K.R., Smith, R.B., and Lowry, A.R., 2009,
Density and lithospheric strength models of the
Yellowstone–Snake River Plain volcanic system from
gravity and heat flow data: Journal of Volcanology and
Geothermal Research, v. 188, p. 108–127, doi:10.1016
/j.jvolgeores.2009.08.006.
de Silva, S.L., 1989, Altiplano-Puna volcanic complex of the
central Andes: Geology, v. 17, p. 1102–1106, doi:10
.1130/0091-7613(1989)017<1102:APVCOT>2.3.CO;2.
de Silva, S.L., and Gosnold, W.D., 2007, Episodic construction of batholiths: Insights from the spatiotemporal
development of an ignimbrite flare-up: Journal of
Volcanology and Geothermal Research, v. 167, p. 320–
335, doi:10.1016/j.jvolgeores.2007.07.015.
de Silva, S.L., Zandt, G., Trumbull, G., Viramonte, J.G.,
Salas, G., and Jimenez, N., 2006, Large ignimbrite
eruptions and volcano-tectonic depressions in the Central Andes: A thermomechanical perspective, in Troise,
C., et al., eds., Mechanisms of activity and unrest at
large calderas: Geological Society of London Special
Publication 269, p. 47–63, doi:10.1144/GSL.SP.2006
.269.01.04.
Detournay, E., and Cheng, A.H.D., 1993, Fundamentals of
poroelasticity, in Fairhurst, C. ed., Comprehensive rock
engineering: Principles, practice and projects: Volume
II, Analysis and design method: Pergamon Press,
p. 113–171.
Drew, D.L., Bindeman, I.N., Watts, K., Schmitt, A.K., Fu,
B., and McCurry, M., 2013, Crustal-scale recycling in
caldera complexes and rift zones of the Snake River
Plain: O and Hf isotopic evidence in diverse zircons
from voluminous rhyolites of the Picabo volcanic field,
Idaho: Earth and Planetary Science Letters, v. 381,
p. 63–77, doi:10.1016/j.epsl.2013.08.007.
Ducea, M., 2001, The California arc, thick granitic batholiths, eclogitic residues, lithospheric-scale thrusting,
and magmatic flare-ups: GSA Today, v. 11, p. 4–10, doi:
10 .1130 /1052 -5173 (2001)011 <0004: TCATGB>2.0
.CO;2.
Geosphere, October 2014
Dufek, J., and Bachmann, O., 2010, Quantum magmatism:
Magmatic compositional gaps generated by melt-crystal dynamics: Geology, v. 38, p. 687–690, doi:10.1130
/G30831.1.
Dufek, J., and Bergantz, G.W., 2005, Lower crustal magma
genesis and preservation: A stochastic framework for the
evaluation of basalt-crust interaction: Journal of Petrology, v. 46, p. 2167–2195, doi:10.1093/petrology/egi049.
Ellis, B.S., and Wolff, J.A., 2012, Complex storage of rhyolite in the central Snake River Plain: Journal of Volcanology and Geothermal Research, v. 211, p. 1–11,
doi:10.1016/j.jvolgeores.2011.10.002.
Ellis, B.S., Barry, T., Branney, M.J., Wolff, J.A., Bindeman,
I.N., Wilson, R., and Bonnichsen, B., 2010, Petrologic
constraints on the development of a large-volume, high
temperature, silicic magma system: The Twin Falls
eruptive center, central Snake River Plain: Lithos,
v. 120, p. 475–489, doi:10.1016/j.lithos.2010.09.008.
Ellis, B.S., Wolff, J.A., Boroughs, S., Mark, D.F., Starkel,
W.A., and Bonnichsen, B., 2013, Rhyolitic volcanism
of the central Snake River Plain: A review: Bulletin of
Volcanology, v. 75, p. 745, doi:10.1007/s00445-013
-0745-y.
Folkes, C., de Silva, S., Schmitt, A., and Cas, R., 2011, A
reconnaissance of U-Pb zircon ages in the Cerro Galán
system, NW Argentina: Prolonged magma residence,
crystal recycling and crustal assimilation: Journal of
Volcanology and Geothermal Research, v. 206, p. 136–
147, doi:10.1016/j.jvolgeores.2011.06.001.
Folkes, C.B., de Silva, S.L., Bindeman, I.N., and Cas,
R.A.F., 2013, Tectonic and climate history influence
the geochemistry of large-volume silicic magmas: New
δ18O data from the Central Andes with comparison to N
America and Kamchatka: Journal of Volcanology and
Geothermal Research, v. 262, p. 90–103, doi:10.1016/j
.jvolgeores.2013.05.014.
Fleck, R.J., and Criss, R.E., 1985, Strontium and oxygen
isotopic variations in Mesozoic and Tertiary plutons of
central Idaho: Contributions to Mineralogy and Petrology, v. 90, p. 291–308, doi:10.1007/BF00378269.
Fowler, S.J., and Spera, F.J., 2010, A metamodel for crustal
magmatism: Phase equilibria of giant ignimbrites:
Journal of Petrology, v. 51, p. 1783–1830, doi:10.1093
/petrology/egq039.
Galland, O., 2012, Experimental modelling of ground deformation associated with shallow magma intrusions:
Earth and Planetary Science Letters, v. 317, p. 145–
156, doi:10.1016/j.epsl.2011.10.017.
Gerya, T., 2009, Introduction to numerical geodynamic modelling: Cambridge, Cambridge University Press, 358 p.
Geyer, A., Folch, A., and Marti, J., 2006, Relationship
between caldera collapse and magma chamber withdrawal: an experimental approach: Journal of Volcanology and Geothermal Research, v. 157, p. 375–
386, doi:10.1016/j.jvolgeores.2006.05.001.
Giletti, B.J., and Yund, R.A., 1984, Oxygen diffusion
in quartz: Journal of Geophysical Research, v. 89,
p. 4039–4046, doi:10.1029/JB089iB06p04039.
Girard, G., and Stix, J., 2010, Rapid extraction of discrete
magma batches from a large differentiating magma
chamber: The Central Plateau Member rhyolites,
Yellowstone Caldera, Wyoming: Contributions to Mineralogy and Petrology, v. 160, p. 441–465, doi:10.1007
/s00410-009-0487-1.
Glazner, A.F., Bartley, J.M., Coleman, D.S., Gray, W., and
Taylor, R.Z., 2004, Are plutons assembled over millions of years by amalgamation from small magma
chambers?: GSA Today, v. 14, p. 4–11, doi:10.1130
/1052-5173(2004)014<0004:APAOMO>2.0.CO;2.
Gregg, P.M., de Silva, S.L., Grosfils, E.B., and Parmigiani,
J.P., 2012, Catastrophic caldera-forming eruptions:
Thermomechanics and implications for eruptions triggering and maximum caldera dimensions on Earth:
Journal of Volcanology and Geothermal Research,
v. 241–242, p. 1–12, doi:10.1016/j.jvolgeores.2012.06
.009.
Gregg, P.M., de Silva, S.L., and Grosfils, E.B., 2013, Thermomechanics of shallow magma chamber pressurization: Implications for the assessment of ground deformation data at active volcanoes: Earth and Planetary
Science Letters, v. 384, p. 100–108, doi:10.1016/j.epsl
.2013.09.040.
25
00969 1st pages / page 26 of 27
Bindeman and Simakin
Grimes, C.B., Ushikubo, T., Kozdon, R., and Valley, J.W.,
2013, Perspectives on the origin of plagiogranite in
ophiolites from oxygen isotopes in zircon: Lithos,
v. 179, p. 48–66, doi:10.1016/j.lithos.2013.07.026.
Grunder, A.L., 1995, Material and thermal roles of basalt in
crustal magmatism—Case study from eastern Nevada:
Geology, v. 23, p. 952–956, doi:10.1130/0091-7613
(1995)023<0952:MATROB>2.3.CO;2.
Gualda, G.A.R., Pamukcu, A.S., Claiborne, L.L., and
Rivers, M., 2010, Quantitative 3D petrography using
X-ray tomography 3: Documenting accessory phases
with differential absorption tomography: Geosphere,
v. 6, p. 782–792, doi:10.1130/GES00568.1.
Gurenko, A., Bindeman, I.N., and Sigurdsson, I., 2013,
Insights into magmatic digestion of the crust and the
origin of silicic magmas in Iceland: Partially melted
leucocratic xenoliths: 2013 Goldschmidt Conference
Abstracts, p. 1234, doi:10.1180/minmag.2013.077.5.7.
Gutierrez, F., and Parada, M.A., 2010, Numerical modeling of
time-dependent fluid dynamics and differentiation of a
shallow basaltic magma chamber: Journal of Petrology,
v. 51, p. 731–762, doi:10.1093/petrology/egp101.
Hanan, B.B., Shervais, J.W., and Vetter, S.K., 2008, Yellowstone plume-continental lithosphere interaction beneath
the Snake River Plain: Geology, v. 36, p. 51–54, doi:10
.1130/G23935A.1.
Harrison, T.M., and Watson, E.B., 1983, Kinetics of zircon
dissolution and zirconium diffusion in granitic melts
of variable water content: Contributions to Mineralogy and Petrology, v. 84, p. 66–72, doi:10.1007
/BF01132331.
Hayba, D.O., and Ingebritsen, S.E., 1997, Multiphase
groundwater flow near cooling plutons: Journal of
Geophysical Research, v. 102, p. 12235–12252, doi:10
.1029/97JB00552.
Hildreth, W., 1981, Gradients in silicic magma chambers:
Implications for lithospheric magmatism: Journal of
Geophysical Research, v. 86, p. 10153–10192, doi:10
.1029/JB086iB11p10153.
Hildreth, W., and Moorbath, S., 1988, Crustal contributions
to arc magmatism in the Andes of central Chile: Contributions to Mineralogy and Petrology, v. 98, p. 455–
489, doi:10.1007/BF00372365.
Hildreth, W., Halliday, A.N., and Christiansen, R.L., 1991,
Isotopic and chemical evidence concerning the genesis and contamination of basaltic and rhyolitic magmas beneath the Yellowstone Plateau Volcanic Field:
Journal of Petrology, v. 32, p. 63–138, doi:10.1093
/petrology/32.1.63.
Hoover, S.R., Cashman, K.V., and Manga, M., 2001, The
yield strength of subliquidus basalts—Experimental results: Journal of Volcanology and Geothermal
Research, v. 107, p. 1–18, doi:10.1016/S0377-0273
(00)00317-6.
Horsman, E., Morgan, S., de Saint-Blanquat, M., Habert,
G., Nugent, A., Hunter, R.A., and Tikoff, B., 2009,
Emplacement and assembly of shallow intrusions from
multiple magma pulses, Henry Mountains, Utah: Royal
Society of Edinburgh Transactions, Earth and Environmental Science, v. 100, p. 117–132, doi:10.1017
/S1755691009016089.
Huber, C., Bachmann, O., and Manga, M., 2009, Homogenization processes in silicic magma chambers by stirring
and mushification (latent heat buffering): Earth and
Planetary Science Letters, v. 283, p. 38–47, doi:10
.1016/j.epsl.2009.03.029.
Huber, C., Bachmann, O., and Dufek, J., 2012, Crystal-poor
versus crystal-rich ignimbrites: A competition between
stirring and reactivation: Geology, v. 40, p. 115–118,
doi:10.1130/G32425.1.
Huppert, H.E., and Sparks, R.S.J., 1988, The generation of
granitic magmas by intrusion of basalt into continental
crust: Journal of Petrology, v. 29, p. 599–624, doi:10
.1093/petrology/29.3.599.
Husen, S., Smith, R.B., and Waite, G.P., 2004, Evidence for
gas and magmatic sources beneath the Yellowstone Volcanic Field from seismic tomographic imaging: Journal of Volcanology and Geothermal Research, v. 131,
p. 397–410, doi:10.1016/S0377-0273(03)00416-5.
Hutton, J., 1794, Observations on granite: Royal Society of
Edinburgh Transactions, v. 3, p. 77–85, doi:10.1017
/S0080456800020305.
26
Jellinek, A.M., and DePaolo, D.J., 2003, A model for the
origin of large silicic magma chambers: Precursors of
caldera-forming eruptions: Bulletin of Volcanology,
v. 65, p. 363–381, doi:10.1007/s00445-003-0277-y.
John, D.A., Rockwell, B.W., Henry, C.D., and Colgan,
J.P., 2011, Hydrothermal alteration of the late Eocene
Caetano ash-flow caldera, north-central Nevada: A
field and ASTER remote sensing study, in Steininger,
R., and Pennell, B., eds., Great Basin evolution and
metallogeny: Geological Society of Nevada 2010 Symposium Proceedings, v. 2, p. 1055–1083.
Karlstrom, L., Rudolph, M.L., and Manga, M., 2012, Caldera size modulated by the yield stress within a crystalrich magma reservoir: Nature Geoscience, v. 5, p. 402–
405, doi:10.1038/ngeo1453.
Kaus, B.J.P., and Podladchikov, Y.Y., 2006, Initiation of
localized shear zones in viscoelastoplastic rocks: Journal of Geophysical Research, v. 111, no. B4, B04412,
doi:10.1029/2005JB003652.
King, S.D., Raefsky, A., and Hager, B.H., 1990, ConMan:
Vectorizing a finite element code for incompressible
two-dimensional convection in the Earth mantle:
Physics of the Earth and Planetary Interiors, v. 59,
p. 195–207, doi:10.1016/0031-9201(90)90225-M.
Konstantinou, A., Valley, J., Strickland, A., Miller, E.,L.,
Fisher, C, Vervoort, J., and Wooden, J., 2013, Geochemistry and geochronology of the Jim Sage volcanic
suite, southern Idaho: Implications for Snake River
Plain magmatism and its role in the history of Basin
and Range extension: Geosphere, v. 9, p. 1681–1703,
doi:10.1130/GES00948.1.
Lavallée, Y., Hess, K.-U., Cordonnier, B., and Dingwell,
D.B., 2007, Non-Newtonian rheological law for highly
crystalline dome lavas: Geology, v. 35, p. 843–846,
doi:10.1130/G23594A.1.
Leeman, W.P., Oldow, J.S., and Hart, W.K., 1992, Lithosphere-scale thrusting in the western United States Cordillera as constrained by Sr and Nd isotopic transitions
in Neogene volcanic rocks: Geology, v. 20, p. 63–66,
doi:10.1130/0091-7613(1992)020<0063:LSTITW>2.3
.CO;2.
Levin, D.A., Peres, Y., and Wilmer, E.L., 2008, Markov
chains and mixing times: Providence, Rhode Island,
American Mathematical Society, 371 p.
Lipman, P.W., 2007, Incremental assembly and prolonged
consolidation of Cordilleran magma chambers: Geosphere, v. 3, p. 42–70, doi:10.1130/GES00061.1.
Lipman, P.W., Prostka, H.J., and Christiansen, R.L., 1972,
Cenozoic volcanism and plate-tectonic evolution of the
western United States. I. Early and middle Cenozoic:
Royal Society of London Philosophical Transactions,
ser. A, v. 271, p. 217–248, doi:10.1098/rsta.1972.0008.
Livo, K.E., Kruse F.A., Clark, R.N., Kokaly, R.F., and
Shanks, W.C., III, 2007, Hydrothermally altered rock
and hot- spring deposits at Yellowstone National
Park—characterized using airborne visible- and infrared-spectroscopy data, in Morgan, L.A., ed., Integrated
Geoscience Studies in the Greater Yellowstone Area—
Volcanic, Tectonic, and Hydrothermal Processes in the
Yellowstone Geoecosystem: US Geological Survey
Professional Paper 1717, p. 491–507.
Longo, A., Papale, P., Vassalli, M., Saccorotti, G., Montagna,
C.P., Cassioli, A., Giudice, S., and Boschi, E., 2012,
Magma convection and mixing dynamics as a source of
ultra-long-period oscillations: Bulletin of Volcanology,
v. 74, p. 873–880, doi:10.1007/s00445-011-0570-0.
Malfait, W.J., Seifert, R., Petitgirard, S., Perrillat, J.P.,
Mezouar, M., Ota, T., Nakamura, E., Lerch, P., and Sanchez-Valle, C., 2014, Supervolcano eruptions driven by
melt buoyancy in large silicic magma chambers: Nature
Geoscience, v. 7, p. 122–125, doi:10.1038/ngeo2042.
Marsh, B.D., 1981, On the crystallinity, probability of occurrence and rheology of lava and magma: Contributions
to Mineralogy and Petrology, v. 78, p. 85–98, doi:10
.1007/BF00371146.
Martin, E., and Sigmarsson, O., 2010, Thirteen million years
of silicic magma production in Iceland: Petrogenetic
processes and tectonic settings: Lithos, v. 116, p. 129–
144, doi:10.1016/j.lithos.2010.01.005.
McConnell, V.S., Valley, J.W., and Eichelberger, J.C., 1997,
Oxygen isotope compositions of intracaldera rocks:
Hydrothermal history of the Long Valley Caldera,
Geosphere, October 2014
California: Journal of Volcanology and Geothermal
Research, v. 76, p. 83–109, doi:10.1016/S0377-0273
(96)00071-6.
McTigue, DF, 1986, Thermoelastic response of fluid-saturated porous rock: Journal of Geophysical Research,
v. 91, p. 9533–9542, doi:10.1029/JB091iB09p09533.
McCurry, M., and Rodgers, D.W., 2009, Mass transfer along
the Yellowstone hotspot track I: Petrologic constraints
on the volume of mantle-derived magma: Journal
of Volcanology and Geothermal Research, v. 188,
p. 86–98, doi:10.1016/j.jvolgeores.2009.04.001.
McCurry, M., Hayden, K.P., Morse, L.H., and Mertzman,
S., 2008, Genesis of post-hotspot, A-type rhyolite of
the E Snake River Plain volcanic field by extreme
fractional crystallization of olivine tholeiite: Bulletin
of Volcanology, v. 70, p. 361–383, doi:10.1007/s00445
-007-0143-4.
Miller, D.S., and Smith, R.B., 1999, P and S velocity structure of the Yellowstone volcano field from local earthquakes and controlled-source tomography: Journal of
Geophysical Research, v. 104, p. 15105–15121, doi:10
.1029/1998JB900095.
Morgan, L.A., and McIntosh, W.C., 2005, Timing and development of the Heise volcanic field, Snake River Plain,
Idaho, western USA: Geological Society of America
Bulletin, v. 117, p. 288–306, doi:10.1130/B25519.1.
Nash, B.P., Perkins, M.E., Christensen, J.N., Lee, D.C., and
Halliday, A.N., 2006, The Yellowstone hotspot in space
and time: Nd and Hf isotopes in silicic magmas: Earth
and Planetary Science Letters, v. 247, p. 143–156, doi:
10.1016/j.epsl.2006.04.030.
Newman, A.V., Dixon, T.H., and Gourmelen, N., 2006, A
four-dimensional viscoelastic deformation model
for Long Valley Caldera, California, between 1995
and 2000: Journal of Volcanology and Geothermal
Research, v. 150, p. 244–269, doi:10.1016/j.jvolgeores
.2005.07.017.
Norton, D., and Taylor, H.P., Jr., 1979, Quantitative simulation of the hydrothermal systems of crystallizing
magmas on the basis of transport theory and oxygen
isotope data, an analysis of the Skaergaard intrusion:
Journal of Petrology, v. 20, p. 421–486, doi:10.1093
/petrology/20.3.421.
Norton, D., Taylor, H.P., Jr., and Bird, D.K., 1984, The geometry and high-temperature brittle deformation of the
Skaergaard intrusion: Journal of Geophysical Research,
v. 89, p. 10178–10192, doi:10.1029/JB089iB12p10178.
Pamukcu, A.S., and Gualda, G.A.R., 2010, Quantitative 3D
petrography using X-ray tomography 2: Combining
information at various resolutions: Geosphere, v. 6,
p. 775–781, doi:10.1130/GES00565.1.
Paterson, S.R., and Vernon, R.H., 1995, Bursting the bubble of ballooning plutons—A return to nested diapirs
emplaced by multiple processes: Geological Society of
America Bulletin, v. 107, p. 1356–1380, doi:10.1130
/0016-7606(1995)107<1356:BTBOBP>2.3.CO;2.
Petford, N., 2003, Rheology of granitic magmas during
ascent and emplacement: Annual Review of Earth and
Planetary Sciences, v. 31, p. 399–427, doi:10.1146
/annurev.earth.31.100901.141352.
Petford, N., and Gallagher, K., 2001, Partial melting of mafic
(amphibolitic) lower crust by periodic influx of basaltic magma: Earth and Planetary Science Letters, v. 193,
p. 483–499, doi:10.1016/S0012-821X(01)00481-2.
Philpotts, A.R., and Carroll, M., 1996, Physical properties of partly melted tholeiitic basalt: Geology, v. 24,
p. 1029–1032, doi: 10 .1130 /0091 -7613 (1996)024
<1029:PPOPMT>2.3.CO;2.
Philpotts, A.R., Carroll, M., and Hill, J.M., 1996, Crystalmush compaction and the origin of pegmatitic segregation sheets in a thick flood-basalt flow in the Mesozoic
Hartford basin, Connecticut: Journal of Petrology,
v. 37, p. 811–836, doi:10.1093/petrology/37.4.811.
Pierce, K.L., and Morgan, L.A., 2009, Is the track of the
Yellowstone hotspot driven by a deep mantle plume?
Review of volcanism, faulting, and uplift in light of
new data: Journal of Volcanology and Geothermal
Research, v. 188, p. 1–25, doi:10.1016/j.jvolgeores
.2009.07.009.
Poldervaart, A., 1956, Zircon in rocks: 2. Igneous rocks:
American Journal of Science, v. 254, p. 521–554, doi:
10.2475/ajs.254.9.521.
00969 1st pages / page 27 of 27
Rhyolites—hard to produce but easy to recycle
Pollarda, D.D., Arvid, M., and Johnson, A.M., 1973,
Mechanics of growth of some laccolithic intrusions
in the Henry mountains, Utah, II: Bending and failure of overburden layers and sill formation: Tectonophysics v. 18, p. 311–354, doi:10.1016/0040-1951
(73)90051-6.
Polteau, S., Mazzini, A., Galland, O., Planke, S., and
Malthe-Sorenssen, A., 2008, Saucer-shaped intrusions:
Occurrences, emplacement and implications: Earth and
Planetary Science Letters, v. 266, p. 195–204, doi:10
.1016/j.epsl.2007.11.015.
Rabinowicz, M., and Vigneresse, J.-L., 2004, Melt segregation under compaction and shear channeling:
Application to granitic magma segregation in a continental crust: Journal of Geophysical Research, v. 109,
B04407, doi:10.1029/2002JB002372.
Ramberg, H., 1970, Model studies in relation to intrusion
of plutonic bodies, in Newhall G., et al., eds., Mechanisms of igneous intrusion: Geological Journal Special
Issue, v. 2, p. 261–286.
Reid, M.E., 2004, Massive collapse of volcano edifices triggered by hydrothermal pressurization: Geology, v. 32,
p. 373–376, doi:10.1130/G20300.1.
Ruprecht, P., Bergantz, G.W., and Dufek, J., 2008, Modeling
of gas-driven magmatic overturn: Tracking of phenocryst dispersal and gathering during magma mixing:
Geochemistry Geophysics Geosystems, v. 9, Q07017,
doi:10.1029/2008GC002022.
Schaltegger, U., Brack, P., Ovtcharova, M., Peytcheva, I.,
Schoene, B., Stracke, A., Marocchi, M., and Bargossi,
G.M., 2009, Zircon and titanite recording 1.5 million
years of magma accretion, crystallization and initial
cooling in a composite pluton (southern Adamello
batholith, northern Italy): Earth and Planetary Science
Letters, v. 286, p. 208–218, doi:10.1016/j.epsl.2009.06
.028.
Schofield, N., Stevenson, C., and Reston, T., 2010, Magma
fingers and host rock fluidization in the emplacement of sills: Geology, v. 38, p. 63–66, doi:10.1130
/G30142.1.
Schofield, N., Brown, D.J., Magee, C., and Stevenson, C.T.,
2012a, Sill morphology and comparison of brittle and
non-brittle emplacement mechanisms: Geological Society of London Journal, v. 169, p. 127–141, doi:10.1144
/0016-76492011-078.
Schofield, N., Heaton, L., Holford, L., Archer, S.G., Jackson, C.A.L., and Jolley, D.W., 2012b, Seismic imaging
of ‘broken bridges’: Linking seismic to outcrop-scale
investigations of intrusive magma lobes: Geological
Society of London Journal, v. 169, p. 421–426, doi:10
.1144/0016-76492011-150.
Schubert, M., Driesner, T., Gerya, T.V., and Ulmer, P., 2013,
Mafic injection as a trigger for felsic magmatism: A
numerical study: Geochemistry Geophysics Geosystems, v. 14, p. 1910–1928, doi:10.1002/ggge.20124.
Seligman, A., Bindeman I., Jicha B., Ellis, B., Ponomareva,
V., and Leonov, V. 2014. Multi-cyclic and isotopically
diverse silicic magma generation in an arcvolcano:
Gorely Eruptive Center, Kamchatka, Russia: Journal
of Petrology, doi:10.1093/petrology/egu034.
Simakin, A.G., 2014, Numerical modeling of the late stage
of subduction zone transference after an accretion
event: Terra Nova, v. 26, p. 22–28, doi:10.1111/ter
.12065.
Simakin, A.G., and Bindeman, I.N., 2008, Evolution of crystal sizes in the series of dissolution and precipitation
events in open magma systems: Journal of Volcanology
and Geothermal Research, v. 177, p. 997–1010, doi:10
.1016/j.jvolgeores.2008.07.012.
Simakin, A.G., and Bindeman, I.N., 2012, Remelting in
caldera and rift environments and the genesis of hot,
“recycled” rhyolites: Earth and Planetary Science Letters, v. 337–338, p. 224–235, doi:10.1016/j.epsl.2012
.04.011.
Simakin, A., and Ghassemi, A., 2005, Modelling deformation of partially melted rock using a poroviscoelastic
rheology with dynamic power law viscosity: Tectonophysics, v. 397, p. 195–209, doi:10.1016/j.tecto.2004
.12.004.
Simakin, A.G., and Ghassemi, A., 2007, The mechanics of
a magma chamber-fault system in trans-tension with
application to Coso: Journal of Structural Geology,
v. 29, p. 1971–1983, doi:10.1016/j.jsg.2007.08.009.
Simakin, A.G., and Ghassemi, A., 2010, The role of magma
chamber-fault interaction in caldera-forming eruptions: Bulletin of Volcanology, v. 72, p. 85–101, doi:10
.1007/s00445-009-0306-6.
Simakin, A., and Talbot, C.C., 2001, Tectonic pumping
of pervasive granitic melts: Tectonophysics, v. 332,
p. 387–402, doi:10.1016/S0040-1951(00)00297-3.
Simpson, C., 1985, Deformation of granitic rocks across the
brittle-ductile transition: Journal of Structural Geology,
v. 7, p. 503–511, doi:10.1016/0191-8141(85)90023-9.
Slabunov, A.I., 1995, Xenoliths as indicators of mass movement in magmatic chamber (an example of Archean
batholith of N Karelia, Baltic Shield): Geochemistry
International, v. 10, p. 1506–1511.
Sleep, N., 2002, Self-organization of crustal faulting and tectonics: International Geology Review, v. 44, p. 83–96,
doi:10.2747/0020-6814.44.1.83.
Smith, R.L., 1979, Ash-flow magmatism, in Chapin, C.E.,
and Elston, W.E., eds., Ash-flow tuffs: Geological
Society of America Special Paper 180, p. 5–27, doi:10
.1130/SPE180-p5.
Spiegelman, M., 2003, Linear analysis of melt band formation
by simple shear: Geochemistry, Geophysics, Geosystems, v. 4, no. 9, p. 8615, doi:10.1029/2002GC000499.
Stelten, M.E., Cooper, K.M., Vazquez, J.A., Reid, M.R.,
Barfod, G.H., Wimpenny, J., and Yin, Q., 2013, Magma
mixing and the generation of isotopically juvenile
silicic magma at Yellowstone caldera inferred from
coupling 238U-230Th ages with trace elements and Hf
and O isotopes in zircon and Pb isotopes in sanidine:
Contributions to Mineralogy and Petrology, v. 166,
p. 587–613, doi:10.1007/s00410-013-0893-2.
Streck, M.J., and Grunder, A.L., 2008, Phenocryst-poor
rhyolites of bimodal, tholeiitic provinces: The Rattlesnake Tuff and implications for mush extraction model:
Bulletin of Volcanology, v. 70, p. 385–401, doi:10
.1007/s00445-007-0144-3.
Svensen, H., Jamtveit, B., Planke, S., and Chevallier, L.,
2006, Structure and evolution of hydrothermal vent
complexes in the Karoo Basin, South Africa: Geological Society of London Journal, v. 163, p. 671–682, doi:
10.1144/1144-764905-037.
Takada, A., 1994a, Accumulation of magma in space and
time by crack interaction, in Ryan, M.P., ed., Magmatic
systems: International Geophysics, v. 57, p 241–257,
doi: 10.1016/S0074-6142(09)60099-1.
Takada, A., 1994b, Development of a subvolcanic structure
by the interaction of liquid-filled cracks: Journal of
Volcanology and Geothermal Research, v. 61, p. 207–
224, doi:10.1016/0377-0273(94)90004-3.
Taylor, H.P., 1974, The application of oxygen and hydrogen
isotope studies to problems of hydrothermal alteration
and ore deposition: Economic Geology and the Bulletin of the Society of Economic Geologists, v. 69,
p. 843–883, doi:10.2113/gsecongeo.69.6.843.
Taylor, H.P., Jr., 1986, Igneous rocks: II. Isotopic case studies of circumpacific magmatism, in Valley, J.W., et al.,
eds., Stable isotopes in high temperature geological
processes: Mineralogical Society of America Reviews
in Mineralogy 16, p. 273–316.
Taylor, H.P., Jr., and Forester, R.W., 1979, An oxygen and
hydrogen isotopic study of the Skaergaard intrusion
and its country rocks: A description of a 55 Ma fossil hydrothermal system: Journal of Petrology, v. 20,
p. 355–419, doi:10.1093/petrology/20.3.355.
Tegner, C., Thy, P., Holness, M.B., Jakobsen, J.K., and
Lesher, C.E., 2009, Differentiation and compaction in
the Skaergaard intrusion: Journal of Petrology, v. 50,
p. 813–840, doi:10.1093/petrology/egp020.
Trubitsyn, V.P., Baranov, A.A., Evseev, A.N., and Trubitsyn,
A.P., 2006, Exact analytical solutions of the Stokes
equation for testing the equations of mantle convection with a variable viscosity: Izvestiya, Physics
of the Solid Earth, v. 42, p. 537–545, doi:10.1134
/S1069351306070019.
Geosphere, October 2014
Tuttle, O.F., and Bowen, N.L., 1958, Origin of granite in the
light of experimental studies in the system NaAlSi3O8KAlSi3O8-SiO2 -H2O: Geological Society of America
Memoir 74, 153 p., doi:10.1130/MEM74-p1.
Valley, J.W., 2003, Oxygen isotopes in zircon: Reviews in
Mineralogy and Geochemistry, v. 53, p. 343–385, doi:
10.2113/0530343.
Vazquez, J.A., Kyriazis, S.F., Reid, M.R., Sehler, R.C.,
and Ramos, F.C., 2009, Thermochemical evolution of
young rhyolites at Yellowstone: Evidence for a cooling but periodically replenished postcaldera magma
reservoir: Journal of Volcanology and Geothermal
Research, v. 188, p. 186–196, doi:10.1016/j.jvolgeores
.2008.11.030.
Wager, L.R., and Brown, G.M., 1967, Layered igneous
rocks: San Francisco, California, W.H. Freeman and
Co., 588 p.
Wanless, V.D., Perfit, M.R., Ridley, W.I., and Klein, E.,
2010, Dacite petrogenesis on mid-ocean ridges: Evidence for oceanic crustal melting and assimilation:
Journal of Petrology, v. 51, p. 2377–2410, doi:10.1093
/petrology/egq056.
Watson, E.B., 1996, Dissolution, growth and survival of zircons during crustal fusion: Kinetic principles, geologic
models and implications for isotopic inheritance: Royal
Society of Edinburgh Transactions, Earth Sciences,
v. 87, p. 43–56, doi:10.1017/S0263593300006465.
Watson, E.B., and Cherniak, D.J., 1997, Oxygen diffusion
in zircon: Earth and Planetary Science Letters, v. 148,
p. 527–544, doi:10.1016/S0012-821X(97)00057-5.
Watts, K., Bindeman, I.N., and Schmitt, A.K., 2011, Largevolume rhyolite genesis in caldera complexes of the
Snake River Plain: Insights from the Kilgore Tuff of
the Heise volcanic field, Idaho, with comparison to
Yellowstone and Bruneau-Jarbidge rhyolites: Journal
of Petrology, v. 52, p. 857–890, doi:10.1093/petrology
/egr005.
Watts, K.E., Bindeman, I.N., and Schmitt, A.K., 2012, Crystal scale anatomy of a dying supervolcano: An isotope
and geochronology study of individual phenocrysts
from voluminous rhyolites of the Yellowstone caldera:
Contributions to Mineralogy and Petrology, v. 164,
p. 45–67, doi:10.1007/s00410-012-0724-x.
Whitaker, M.L., Nekvasil, H., Lindsley, D.H., and McCurry,
M., 2008, Can crystallization of olivine tholeiite give
rise to potassic rhyolites? An experimental investigation: Bulletin of Volcanology, v. 70, p. 417–434, doi:10
.1007/s00445-007-0146-1.
Whittington, A.G., Hofmeister, A.M., and Nabelek, P.I.,
2009, Temperature-dependent thermal diffusivity of
the Earth’s crust and implications for magmatism:
Nature, v. 458, no. 7236, p. 319–321, doi:10.1038
/nature07818.
Wickham, S.M., 1987, The segregation and emplacement of
granitic magmas: Geological Society of London Journal, v. 144, p. 281–297, doi:10.1144/gsjgs.144.2.0281.
Wilson, C.J.N., and Charlier, B.L.A., 2009, Rapid rates of
magma generation at contemporaneous magma systems, Taupo Volcano, New Zealand: Insights from
U-Th model-age spectra in zircons: Journal of Petrology, v. 50, p. 875–907, doi:10.1093/petrology/egp023.
Wotzlaw, J.F., Bindeman, I.N., Schaltegger, U., Brooks,
C.K., and Naslund, H.R., 2012, High-resolution
insights into episodes of crystallization, hydrothermal
alteration and remelting in the Skaergaard intrusive
complex: Earth and Planetary Science Letters, v. 355–
356, p. 199–212, doi:10.1016/j.epsl.2012.08.043.
Wotzlaw, J.F., Bindeman, I.N., Watts, K.E., Schmitt, A.K.,
Caricchi, L., and Schaltegger, U., 2014, Linking rapid
magma reservoir assembly and eruption trigger mechanisms at evolved Yellowstone-type supervolcanoes:
Geology, v. 42, doi:10.1130/G35979.1
Zak, J., Paterson, S.R., and Memeti, V., 2007, Four magmatic fabrics in the Tuolumne batholith, central Sierra
Nevada, California (USA): Implications for interpreting fabric patterns in plutons and evolution of magma
chambers in the upper crust: Geological Society of
America Bulletin, v. 119, p. 184–201, doi:10.1130
/B25773.1.
27