In
San Francisco
[1999][1998][1997]
December 13－18，1999
Yokose H., (1999), Petrography of partially
fused basic crustal xenoliths from Kimbo volcano, northwest Kyushu, Japan: implications
for disequilibrium textures of magma mixing process. AGU 1999 Fall Meeting (
San Francisco).
To clarify the interaction between magma
and the crust concretely, crustal xenoliths and magmatic inclusions entrained
in Ishigamiyama and Araoyama andesites from Kimbo volcano northwest Kyushu,
Japan are described and are discussed. Various kind of crustal xenoliths, such
as hornblendite, hornblende gabbro, amphibolite, biotite gneiss, pyroxene
granulite, siliceous gneiss and biotite granite, are found and underwent
ultra-metamorphism and melting.
Textures
of ultra-metamorphism in the xenoliths are summarized as follows. Hydrous
minerals break down to cryptocrystalline aggregate, aggregate of fine grained
minerals (px+pl+mt), symplectic intergrowth of pl and cpx or idiomorphic large
clinopyroxene with magnetite and plagioclase inclusions. These breakdown
products resemble a opacitic hornblende phenocryst and glomero porphiritic
texture in volcanic rocks. Pyroxenes are found as breakdown products of hydrous
minerals, and as crystallization products during cooling process in the melt
that was produced by dehydration melting. Both types of pyroxene include
abundant magnetite. It is difficult to identify a phenocryst crystallized in
magma chamber with a xenocryst derived from crustal melting. In general,
plagioclase exhibits sieve texture in their core when it is accompanied with
hydrous minerals or its pseudomorph. The sieve texture indicates that these
crystals underwent incipient melting. Some large plagioclase crystals with
clear rim are found in direct contact with glass. Resorption and overgrowth
texture in plagioclase is commonly found as a phenocrystic plagioclase in mixed
magma. The dehydration melting textures present sparsely in xenoliths. The
degree of dehydration melting depends on mineral species around hydrous minerals.
The melt phases quenched immediately after eruption resemble to groundmass
texture of volcanic rocks. Highly metamorphosed hornblende gabbro xenoliths
change continuously into magmatic inclusion in their texture and composition.
This implies that magmatic inclusion can produce by dehydration melting of
hornblende gabbro.
The
disequilibrium textures in crustal xenoliths which underwent ultra-metamorphism
have been believed as a line of evidence for magma mixing. We may misunderstand
disequilibrium melting process in the crust with magma mixing process.
Dehydration melting of heterogeneous basic crust seems to be the most important
processes for volcanic rocks in island arcs and continental margins.
Yasuhara M. and Yokose H., (1999),
Geochemical evolution of Aso caldera, central Kyushu Japan. AGU 1999 Fall
Meeting ( San Francisco).
Twenty-six
horizons of volcanic edifices related to Aso caldera formation, which is divided
geologically into 32 horizons entirely, have been analyzed to exam the
geochemical evolution of Aso caldera. Aso caldera has been considered to form
by 4 large-scale pyroclastic flow eruptions (pfl): from Aso-1 (30km3,
260ka) through Aso-2 (25km3, 140ka) and Aso-3 (40km3, 120ka)
to Aso-4 (80km3, 90ka). The large scale pfl deposits intercalate
small scale air fall deposits (pfa; Aso-2/1, Aso-3/2, Aso-4/3）and
lava flow. Most of all essential
fragments from the pyroclastics have pl, cpx, opx and mt as a phenocrystic
mineral. Pumices from Aso-4 eruption only have hornblende as phenocryst additionally.
The
secular variation of Na2O shows systematical increasing (3.2 to 4.6wt.\%) and that
of TiO2, Fe2O3, MgO, P2O5 and Cu show systematical decreasing with time: TiO2 (0.7
to 0.6wt.\%), Fe2O3 (6.3 to 2.8wt.\%), MgO (3.0 to 0.8wt.\%), P2O5 (0.3 to 0.1wt.\%)
and Cu (46.7 to 8.3 ppm). Systematic differences of geochemical character are
also found between pumices from pfls and pfas. SiO2 concentration in
pumice from pfl deposits increase with time (60wt.\% to 65wt.\%), but that of
pumice from pfa deposits show constant value (62 wt.\%). K2O and Rb
concentration in pumice from pfl deposits show constant values (K2O= ~ 5.8wt.\%;
Rb = ~ 157ppm). On the contrary, the concentrations of pfa deposits increase
with time (K2O: 1 to 3 wt.\%; Rb: 70 to 100 ppm ). Rb/K ratio of both pfl and
pfa do not indicate secular variation ( pfl = 39, pfa = 43 ), but Zr/Nb ratio
of pfl and pfa increase from different level with time. Ga/Al ratio of pfl and
pfa decrease from deferent level with time.
Crystallization differentiation or
contamination in single magma chamber can not explain the fluctuation of
incompatible element ratios with time. The chemical variation corresponding to eruption
modes do not support zoned magma chamber model. From Nb/Zr ratio and Ga/Al
ratio, the secular variation of geochemical data may indicate source rock heterogeneity.
Nakamura Y.,
Yokose H., Matsui, T. and Yokose R., (1999), Petrogenesis of large scale felsic
magma from Kikai, Ata and Aira caldera, southern Kyushu, Japan. AGU 1999 Fall
Meeting ( San Francisco).
Pyroclastic
flow deposits (Aira caldera, Ata caldera and Kikai caldera）and
granitic rocks (Takakumayama granitic rocks, Minami-Osumi granitic rocks and
Yakushima granitic rocks）adjacent to the calderas southern Kyushu,
Japan, have been compared geochemically in order to examine the formation of
large-scale silicic magma body. Strong regional correlations are found in
Harker diagram of the major elements except for Na2O and K2O. Especially, SiO2
contents of the rhyolitic and granitic magmas are almost identical (Aira
caldera = Takakuma Gr = 75wt.\%；Ata caldera = Minami-Osumi Gr = 68wt.\%;
Kikai caldera = Yakushima Gr = 72wt.\%). This systematic coherence on major
element composition implies that they have genetical relationships. However,
there are no correlation between volcanic rocks and granitic rocks in
concentration of incompatible element. Volcanic rocks indicate linear trend on
many elemental variation diagrams. Basaltic rocks are plotted on extension of
the trends. Especially, the differences between Outer zone granitoids and
volcanic rocks in concentration of LILE are remarkable. The concentration of K,
Ba, Rb in granitic rocks are twice as much as those of volcanic rocks. The
plotted data of LILE/LILE ratio in granitic rocks are dispersed largely, but
volcanic rocks indicate linear trend. Nb/Zr ratio of the granitic rocks
resemble pelitic rocks from Shimanto supergroup (Nb/Zr＝0.1).
Nb/Zr ratios of the volcanic rocks have different value from each caldera. In Y
vs. Zr diagram, basaltic rocks and rhyolitic rocks do not seem to be the same lineage.
It suggest that these magmas could not be produced through fractional crystallization
or mixing process.
The
scatterings of LILE/LILE ratio in Outer zone granitoids are contribution to
fluid phase contamination. On the other hand, source rock heterogeneity rather
than metasomatic fluid can interpret the difference in Nb/Zr ratio of the
granitic rocks and volcanic rocks. We can conclude that genesis of these
rhyolitic magma is different from those of granitic magma in southern Kyushu.
December 6－10，1998
Nakamura Y. and Yokose H.(1998), Origin of
rhyodacite magma from Ata caldera south Kyushu, Japan: implication for crustal
evolution. AGU 1998 Fall Meeting (San Francisco).
A
large-scale silicic volcanism and granitic batholith play an important role at
the chemical differentiation in the crust. The Ata pyroclastic flow deposit
cover southernmost Kyushu, Jpan, and underlain directly the outer zone granitoids
solidified at 14Ma and sedimentary rocks of Shimanto supergroup. The magmas
related to Ata caldera and the outer zone granitoids may share with the same
crust for their source. We discussed the chemical differentiation process in
the crust by comparing the chemical composition of the pumice of Ata caldera,
the outer zone granitoids and the sediments of Shimanto supergroup.
Although
their silica concentration overlap (pumice of the Ata pyroclastic flow, 62-74
wt%; outer zone granitoid, 60-79 wt%; sedimentary rocks, 63-84 wt%), the
differences between the granitoids and the pumice are observed in many
elements. On ACF diagram the pumice, the granitoids and the sedimentary rocks
are plotted on the area of I-type, S-type and I-type and S-type respectively.
Because the mixing trend between the pumice and other geological units cannot
draw on variation diagram, the outer zone granitoids and sedimentary rocks may
not affect magma chemistry of the pumice.
The characters of incompatible elements
in the pumice and in the outer zone granitoids are quite different. The
K2O/Na2O ratios of the pumice indicate 0.6 and are almost constant without
relation to silica. On the other hand, the K2O/Na2O ratios of the outer zone
granitoids indicate 1.3. The outer zone granitoids are strongly enriched in Rb,
K and Ba in comparison with the pumice. The K/Rb ratios of the outer zone
granitoids indicate the lowest values (250~60). The LILE/HFSE, Nb/Zr and
LREE/HREE ratios in the outer zone granitoids are higher than the pumice. The
REE patterns of the pumice are almost flat. Thus highly incompatible elements
are depleted in the pumice compared with the outer zone granitoids. It is
possible to explain that the magmas formed the Ata caldera are produced by the
melting of the restitic crust from which the outer zone granitoids extracted.
Yasuhara M. and Yokose H.(1998), Origin of trachytic rocks of Aso volcano on
volcanic front in central Kyushu, Japan. AGU 1998 Fall Meeting (San Francisco).
The
Aso caldera, central Kyushu, Japan, was formed by 4 cycles of volcanic
eruption, which are named Aso 1, 2, 3 and 4. These eruptions happened during
300ka to 70ka. Although they are
located on the volcanic front, many of the volcanic rocks are composed of trachyte
and trachy andesite that is rich in alkali compared with the other volcanic
rocks on the volcanic front. We analyzed the geochemical character of the
volcanic rocks from Aso caldera in order to understand why Aso volcanic rocks
are rich in alkali and how they produced.
Plausible mecanisms for enrichment of
alkali elements in magma are such as source enrichment, low degree of melting
and metasomatism. The acidic rocks
are only high in K2O among Aso volcanic rocks. The basaltic rocks (the pre Aso
volcanic rocks and Aso central cone) belong to high alumina basalt series and
are common for volcanic front. The
trends on Harker diagram with the
range of a intermediate rocks implies mixing process. These trends are
supported by the petrographical evidences observed in the andesites and
basaltic andesites from central cones. Therefore, the genesis of acidic rocks
rich in alkali elements may be the key to solve the main problem.
The
positive correlations are observed in LILE-LILE, HFSE-HFSE, LILE-HFSE diagrams.
Because the positive correlation are observed in LILE-HFSE diagram, the metazomatic processes can not be a
candidate for the alkali enrichment. The ratios of LILE/LILE have a nearly
constant values. On the other hand, the ratios of LILE/HFSE and HFSE/HFSE are
different. Crystallization
differentiation processes for the genesis of the trachytic rocks is also
unsuitable, because the ratios of the incompatible elements (Ba, Rb, K, Zr, Nb,
REE)/SiO2 decrease with the volcanic cycle. This dilution of the incompatible
elements can be interpreted as increasing of degree of melting. This fact is
consistent with the eruption volumes that is increasing from Aso 1 to Aso 4.
The geochemical characters of the caldera
forming magma are identical to a rift volcanism on continental region. The
genesis of the trachytic magma in Aso caldera may involve in the spreading of
Okinawa trough which is southern extent of Beppu-Shimabara graben, not in the
subducting of the Philippine Sea plate.
Taguchi Y. and Yokose H.(1998), Geochemical character of dacitic rocks from
Unzeb volcano in Beppu-Shimabara graben, central Kyushu, Japan. AGU 1998 Fall
Meeting (San Francisco).
In central Kyushu Japan, Unzen volcano is
located on Beppu-Shimabara graben that has been believed as the north extension
of Okinawa trough. The chemical variation in the magma is explained due to the
interaction between OIB type basaltic magma, which are derived by rift
volcanism, and the crust. The volcanic rocks, however, are dacite characterized
by convergent plate margin magma system. We discussed chemical character of the
andesitic magma in this area in which subducted plate (Philippine Sea plate) does
not exist beneath the volcano.
We collected the volcanic rocks from Unzen
volcano and from basement volcanic rocks erupted during early Pleistocene to
Pliocene. The major element and the trace element of these samples have been
determined. Phenocryst in the volcanic rocks from Unzen volcano mainly consist
of plagioclase and amphibole. Many lavas have basic inclusion which belong to
calc-alkaline rock series. The basement volcanic rocks consist of olivine basalt and
pyroxene andesities. The compositional gaps are evident between the basaltic rocks of basement
(50%-53% SiO2) and the lavas from Unzen volcano (58%-67% SiO2). The basic
inclusions in the dacites of Unzen volcano are plotted on the compositional gap
in Harker diagrams. However, the trace element data of the basic inclusions
scatter largely on variation diagram We can not find a linear trend that
implies the hypothetical mixing process between basalts and dacites.
The volcanic rocks are divided into two types by
Nb/Zr ratio. Because the ratio can not change secondary process, the ratio is
inherited in their source. So the basalts and dacites have different source. We
defined the depletion of Nb as Nb/Nb* ratios in the spider diagram and examined
the relation with the chronological order change and SiO2 content. Nb/Nb*
ratios indicate a constant value, even if SiO2 content changed. The depletion of Nb in the dacites
decrease with time, but the change is quite a small. These indicate that island
arc signature does not need subducted slab.
Nakada S. and Yokose H. (1998), Pliocene-Pleistocene volcanism in Sendai
area, southwest Kyushu, Japan: implication for the opening of Okinawa trough.
AGU 1998 Fall Meeting (San Francisco)．
The
space and time distribution of the Cenozoic volcanic rocks in Sendai area,
southwest Kyushu, Japan has been discussed. In this area, the volcanisms are
closely related to tectonic events: the subduction of the Philippine Sea Plate
(6Ma), opening of the Kagoshima graben (3Ma) and opening of the Okinawa Trough
(2Ma).
The
Cenozoic volcanism in this area can be divided into 3 stages: early stage
(4-2Ma), middle stage (2-1 Ma) and late stage (0.5-0 Ma). The volcanic rocks in
the early stage mainly comprise calc-alkali andesites. In the beginning of this
stage, some andesites have high Sr/Y ratio. The volcanic rocks of the middle
stage comprise basalt (Sendai basalts) and rhyodacite. The volcanism of this
stage, as a whole, can be denoted a bimodal volcanism. The chemical character
of the basalts subdivided into three rock series: high alumina basalt, low
alkali tholeiite and alkali basalt. Alkali basalts have a OIB type geochemical
character and the other basalts have island arc characters. Their chemical characters
of these rocks including rhyolites are identical to the rocks dredged from the
base of the Okinawa Trough. The volcanic activities of the late stage are
limited to the east side of the Kagoshima graben. This stage is divided into
three subgroups: andesitic stratovolcanos (Imuta and Sakurajima volcano),
basaltic maars (Yonemaru and Sumiyoshi), large-scale felsic volcanism (Aira
caldera). Although the compositional gap is small, the volcanism of this stage
is also bimodal. The basalts in the late stage have small LILE/HFSE ratios and
Nb depletion on a spider diagram in comparison with those of the middle stage.
The
space and time distribution of the basaltic rocks constrains mantle convection
pattern. The chemical variations in the basalts are interpreted as increasing
of degree of melting and as replacing the depleted source for the enriched
source beneath the arc crust. The replacement of the mantle wedge may have
begun with the deeper part of the west, after that it may have spreaded in the
shallow part of the east. The volcanism in southwest Kyushu has been controlled
by not dehydration of the subducted slab but by the mantle flow from the
back-arc side to the mantle wedge.
Yokose H. and
Inoue T. (1998), Chemical evolution of rhyolitic magma in Aira Caldera, south
Kyushu, Japan: zoned magma chamber or progressive partial melting. AGU 1998
Fall Meeting (San Francisco).
Abstract
The major elements, trace elements and Sr isotopic ratios of the pumices in pyroclastic deposits from Aira caldera southern Kyushu, Japan, has been determined. The chemical evolution of the magma related to the caldera formation has been discussed. Samples are collected from 13 geological units erupted during past 70ka: Fukuyama, Iwato, Otsuka, Fukaminato, Arasaki, Kenashino, Osumi, Tarumizu, Tsumaya, Ito, Moeshima, Takano, and Sakurajima volcano that is post caldera stage. Iwato pyroclastic flow deposits have also scoria and banded pumice. The phenocrystic mineral in the pumice mainly consists of Pl and Opx, and minor Qz, Mt and Cpx are also included. The pumice in the Fukuyama pyroclastic deposits have hornblende as a phenocryst. Phenocrysts/glass ratios are 1/10 or less.
Bulk data are plotted on the I-type area defined by ACF diagram and are indicated as a liner trend on Harker diagram. Most of the pumices are indicated in the boundary of LTH and HAB rock series in the total alkali-SiO2 diagram. The K2O concentration indicates the range of medium K. The K/Rb and Ba/Rb ratios indicate a constant value without depending on the horizon. These characters suggest that each magma may be derived from single igneous source. Based on the abundance of HFS elements, we can divide essential fragments into two groups: High HFSE group (Iwato scoria, Moeshima, Takano and Sakurajima) and low HFSE group (the other samples). The High HFSE group is rich in Zr, Nb, Y, Zn, REE, Ga in comparison with low HFSE group and has higher HFSE/LILE ratios. The Low HFSE group indicates an extremely homogeneous Sr isotopic ratio (0.7059$\pm$0.0002). The High HFSE group have high Sr isotopic ratios (0.7066-0.7076). These chemical relationships between LILE and HFSE cannot explain as mixing of different magma, fractional crystallization or assimilation.
Similar chemical differences have already been reported in A-type magma with granitic system. Considering the genesis of A-type magma, we can explain the chemical evolution of the magma in the Aira caldera as changing in the melting condition at the source region. We propose that the chemical variations in the magma formed Aira caldera is not produced by the secondary processes in the magma chamber, but produced by progressive partial melting in the lower crust.
December 8－12，1997
Yokose H.
(1997), Lateral Variation of Geochemical Character in andesites from Central
Part of Honshu arc, Japan: Role of the Lower Crust in Andesitic Magmatism. AGU
1997 Fall Meeting (San Francisco).
Major elements, trace
elements, and isotopic compositions of Quaternary volcanic rocks from central part
of Honshu arc, Japan, were determined. Although these rocks are dominantly
andesitic in composition, some element ratios (Zr/Nb, Ce/Pb and K/Rb for
example) and isotopic compositions (Sr and Nd) are available for inferring
characters of their source. The systematic lateral variations of the
geochemical character in this studied area are observed both across-arc and
along-arc directions. The lateral variations, however, correlate neither with
configuration of the subducted Pacific plate nor with the crustal thickness.
Some variation diagrams show that geochemical characters of the andesites
cannot be explained by mixing between N-MORB source and subducted oceanic
sediments and by assimilation of the upper crust.
The lateral variations of the andesites correspond
to arrangement of their basement rocks. The basement rocks in the studied area
can be divided into three tectonic provinces: (1) pre-Neogene basement rocks
formed as continental margin of Eurasia plate; (2) Neogene basement rocks
formed during opening of Japan Sea; (3) Izu-Bonine arc crust formed in oceanic environment.
Because each tectonic province has quite different geological history, it is
expected that their lower crust also have different geochemical character.
Therefore, it is considered that the geochemical character of the andesites may
have inherited from those of the lower crust underlying the volcanoes.
Furthermore, strong positive correlation is also recognized between Sr isotopic
ratio of Quaternary volcanic rocks and initial Sr ratios of Mesozoic to
Cenozoic plutonic rocks. This correlation suggests that both magmas can be
derived from the some source materials.
I propose that the most important process for the genesis of andesite is
anatexsis of the lower crust underlying the volcano.