Upload
trinhtruc
View
222
Download
2
Embed Size (px)
Citation preview
Deforma(on Constraints
• Galápagos volcanoes undergo 3-‐phases of evolu(onary change, corresponding to juvenile, mature, and dying stages as the Nazca Plate carries them away from the hotspot. • In the juvenile stage, a diverse suite of picrite-‐to-‐basalt is erupted, because each batch of ascending magma undergoes a different thermal experience. • In the mature stage, the magmas pass through a system that is in a thermochemical steady-‐state, due to crystal-‐liquid reac(on and residence in a shallow, well-‐mixed sill. A monotonous suite of low-‐Mg basalt is produced. Most of the volume of the plumbing system is crystalline mush. • As the volcano dies, the mush column cools and solidifies, which can create fairly large volumes of crystal-‐poor rhyolite.
An Evolu(onary Model of Galápagos Magma Chambers Dennis Geist
Petrologic Constraints
Petrographic Constraints
Trace Element Constraints
A Dying System: Alcedo
3 Stage Evolu-on: 1. Juvenile Transient 2. Mature Steady State 3. Dying Transient Fernandina
Mg# 50 ± 3 Mature Steady
Roca Redonda Mg# 64 ± 10 Juvenile Transient
Ecuador Mg# 51 ± 5 Juvenile Transient
Cerro Azul Mg# 58 ± 6 Juvenile Transient
Wolf Mg# 49 ± 4 Mature Steady
Darwin Mg# 47 ± 3 End of Mature
Alcedo Mg# 38± 11 Dying Transient
Sierra Negra Mg# 42 ± 4 End of Mature
Some volcanoes at the leading edge of the Galápagos hotspot erupt picrites as well as basalts (Cerro Azul, Roca Redonda, and Volcan Ecuador).
Fernandina, Sierra Negra, and Wolf volcanoes erupt lavas with a very small composi(onal range and generally lower Mg#.
Some of the “downstream volcanoes” erupt lavas with low and variable Mg# (extending to rhyolite). Alcedo is the best example. Terry Naumann has recently discovered young icelandites on Darwin volcano.
Geode(c measurements indicate a systema(c difference between the 3 types of volcanoes. The steady systems have very shallow, flat-‐topped magma bodies. This example, from Sierra Negra, is best modeled as a 1.9 km deep sill (note the lack of mo(on of GV01, which could only be because of a shallow chamber). Fernandina and Wolf show similar deforma(on behavior.
Cerro Azul, a juvenile system, has a notably deeper magma body (Amelung et al., 2000; Naumann et al., 1999). Ecuador displays no evidence of a long-‐lived shallow magma body.
Subaerial Fernandina
Submarine Fernandina I didn’t
crystallize from this black part.
Me neither
Actually, none of us did
We subaerial plag
“phenocrysts” are exo(c too.
There are 2 remarkable observa(ons to be made about the phenocrysts in western Galápagos basalt. First, it can be demonstrated that most did not crystallize from their matrix, i.e. these are “antecrysts”, crystallized from older magmas but not yet consolidated into rock.
Second, the crystals are density segregated, with plagioclase only at the top of the system, and abundant olivine and pyroxene lower in the system.
My interpreta(on is that these crystals come from disaggregated mush. Light plagioclase mush accumula(ng at the top of the system and denser mafic mush below.
This is a GPS (me series showing the infla(on of Sierra Negra’s caldera floor since the 2005 erup(on. The overall trend is decelera(on, but there are notable signs of waxing and waning over ~ 6 month cycles.
The important point, however, is that the shallow magma chamber has been con(nuously refilled and growing since the erup(on. There are not episodic replenishments: the magma has been trickling in steadily.
tres = 3000 to 18000 years
Eruption rate = 0.0002 km3/y
Chamber volume = 0.6 to 3.6 km3
20 to 200 m thick sill
tres =C2in
C2resRin − Rres
dR /dt
Changes in conserva(ve trace element ra(os with (me can be used to es(mate the volume of the well-‐mixed part of the chamber, through the calcula(on of residence (mes. Applica(on of even extreme es(mates of the governing parameters means that the volume of the well-‐mixed magma chamber is small. For example, it cannot be a sphere with the same radius as the caldera.
We have developed a geothermometer based on the composi(ons of lavas and a thermodynamic model. Note that the steady-‐state mature systems have been thermally buffered: lavas have erupted at nearly the same (and rela(vely cool) temperature for thousands of years at the mature volcanoes. Thus, both composi(on and temperature are buffered in the mature stage. Juvenile volcanoes erupt hoier and more variable magma. Dying volcanoes erupt cool evolved magma.
.
Rb Th U Ta K La Ce Nd Sr Sm Hf Zr Ti Eu Gd Tb Y Yb Lu0.1
1
10
20
Roc
k/O
livin
e Th
olei
ite
Tholeiite
IcelanditeDaciteRhyolite
Model
Conclusions The Conceptual Model
Alcedo has erupted lavas ranging from basalt to rhyolite. The rhyolites are due to frac(onal crystalliza(on, as this trace element modeling effort demonstrates.