Petropavlovsk-Kamchatsky, August 2004 RECURRENT CALDERA-FORMING ERUPTIONS: KSUDACH CASE STUDY Preliminary results of the ongoing petrological and experimental

Petropavlovsk-Kamchatsky, August 2004

RECURRENT CALDERA-RECURRENT CALDERA-FORMING ERUPTIONS: FORMING ERUPTIONS: KSUDACH CASE STUDYKSUDACH CASE STUDY

Preliminary results of the ongoing petrological Preliminary results of the ongoing petrological and experimental studyand experimental studyPavel IzbekovPavel Izbekov11, James , James

GardnerGardner22, Ivan Melekestsev, Ivan Melekestsev33, , and John Eichelbergerand John Eichelberger11

11 Alaska Volcano Observatory, Alaska Volcano Observatory, Geophysical Institute, University of Geophysical Institute, University of Alaska Fairbanks, AlaskaAlaska Fairbanks, Alaska

22 University of Texas at Austin, Texas University of Texas at Austin, Texas33 Institute of Volcanology and Institute of Volcanology and Seismology, Petropavlovsk-KamchatskySeismology, Petropavlovsk-Kamchatsky


IntroductionIntroduction

The development of The development of volcanic centers is volcanic centers is often cyclic: cone – often cyclic: cone – caldera – cone – caldera – cone – caldera ...caldera ...

Ksudach caldera complex in Kamchatka Ksudach caldera complex in Kamchatka appears to represent an exemplary case appears to represent an exemplary case where at least three such cycles have where at least three such cycles have been completed since middle been completed since middle Pleistocene.Pleistocene.

Landsat 7 ETM & SRTM DEM


We use petrological and We use petrological and experimental approaches to experimental approaches to determine the relationships of determine the relationships of magmas erupted at Ksudach and to magmas erupted at Ksudach and to infer their pre-eruptive conditions, infer their pre-eruptive conditions, which potentially may illuminate the which potentially may illuminate the mechanism of such cyclicity, as well mechanism of such cyclicity, as well as the mechanism of the recurrent as the mechanism of the recurrent caldera-forming eruptions. caldera-forming eruptions.

Introduction cont.Introduction cont.


LocationLocation

Southern part of Southern part of the Eastern the Eastern Volcanic Front of Volcanic Front of KamchatkaKamchatka

Ca. 150 km to the Ca. 150 km to the SW from PKSW from PK

Japan

Russia

Alaska


BackgroundBackground 5 caldera-forming eruptions at one 5 caldera-forming eruptions at one single volcanic center:single volcanic center:

2 larger Pleistocene calderas and2 larger Pleistocene calderas and 3 Holocene calderas (8800 BP, 3 Holocene calderas (8800 BP,

6300-6000 BP, and 1800 BP)6300-6000 BP, and 1800 BP) Cycles of a cone-building effusive Cycles of a cone-building effusive

activity and caldera-forming events: activity and caldera-forming events: (1) The first caldera has destroyed an (1) The first caldera has destroyed an older Pleistocene shield volcano.older Pleistocene shield volcano.(2) Then a new cone has been (2) Then a new cone has been constructed inside the caldera and was constructed inside the caldera and was subsequently destroyed by a collapse subsequently destroyed by a collapse of the second Pleistocene caldera.of the second Pleistocene caldera.(3) The intra-caldera cone has been (3) The intra-caldera cone has been constructed again and was destroyed constructed again and was destroyed by the first Holocene caldera-forming by the first Holocene caldera-forming eruption KS-4 (8800 BP). eruption KS-4 (8800 BP). Then three Then three nested calderas were formed nested calderas were formed sequentially with only a subtle intra-sequentially with only a subtle intra-caldera effusive activity between the caldera effusive activity between the caldera-forming eruptions.caldera-forming eruptions.(4) After KS-1 eruption (1800 BP) the (4) After KS-1 eruption (1800 BP) the cone-building activity resumed again cone-building activity resumed again and culminated in the explosive and culminated in the explosive eruption of 1907 AD.eruption of 1907 AD.

from Volynets et al. (1999)


QuestionsQuestionsWe focused first on the composition of the most evolved magmas erupted We focused first on the composition of the most evolved magmas erupted during a sequence of three caldera-forming eruptions occurred at Ksudach during a sequence of three caldera-forming eruptions occurred at Ksudach during the Holocene: the dacite of the initial fall deposit of KS-4 (8800 yr. BP, during the Holocene: the dacite of the initial fall deposit of KS-4 (8800 yr. BP, 67.4 wt % SiO2), the KS-3 rhyolite (~6000 yr. BP, 70.3 wt % SiO2), and the 67.4 wt % SiO2), the KS-3 rhyolite (~6000 yr. BP, 70.3 wt % SiO2), and the KS-1 rhyolites (1800 yr. BP, 71.5-72.1 wt % SiO2). KS-1 rhyolites (1800 yr. BP, 71.5-72.1 wt % SiO2).

1. Do the products of the Holocene 1. Do the products of the Holocene caldera-forming eruptions at caldera-forming eruptions at Ksudach represent snapshots of a Ksudach represent snapshots of a single magma reservoir?single magma reservoir?

2. What are the pre-eruptive 2. What are the pre-eruptive conditions for the last Holocene conditions for the last Holocene caldera-forming eruption (1800 BP)?caldera-forming eruption (1800 BP)?


SamplesSamples

Holocene intra-caldera deposits

KS-1

KS-2

KS-3

SiO2

Generalized stratigraphic cross-section of Holocene deposits from Volynets et al. (1999). Our samples for this study are indicated by red symbols.


MethodsMethods

Mineral and glass compositions were Mineral and glass compositions were determined by electron microprobe (3-determined by electron microprobe (3-10 micron, 10 nA, 15 kV beam).10 micron, 10 nA, 15 kV beam).

Pre-eruptive conditions for KS-1 were Pre-eruptive conditions for KS-1 were reproduced experimentally in Rene and reproduced experimentally in Rene and TZM pressure vessels (NNO, water-TZM pressure vessels (NNO, water-saturated, 100 MPa).saturated, 100 MPa).

Whole rock compositions were Whole rock compositions were determined by XRF and ICP-MSdetermined by XRF and ICP-MS


Total alkalies versus silica for products of Holocene eruptions at Ksudach

3

3.5

4

4.5

5

5.5

6

6.5

7

54 59 64 69 74SiO2

Ksht-2

Ksht-1

Shtubel' cone

KS-1

Ksbt

post-KS-2 domes

KS-2

KS-3

KS-4

KO

+ N

aO

22

WR chemistryWR chemistry


10

15

20

25

30

35

40

45

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

KS-1 KS-2 KS-3 KS-4 KSht1-mafic

Nor

mal

ized

to c

hond

rite

R E E s igna tures o f the H o locene ca ldera-fo rm ing e ruptions a t K sudach as compared to tha t o f K s ht1 basa lt

WR chemistry cont.WR chemistry cont.


Fe-Ti oxides geothermometry for Ksudach

-13

-12.8

-12.6

-12.4

-12.2

-12

-11.8

-11.6

-11.4

-11.2

-11

850 870 890 910 930 950

Temperature, C

Log(

f)

O 2

KS-1 grey

KS-1 white

KS-2

KS-3

KS-4

Mineral compositionsMineral compositionsMagnetite-ilmenite Magnetite-ilmenite thermometry indicates thermometry indicates that the dacite of KS-4 that the dacite of KS-4 was last equilibrated at was last equilibrated at 914-924914-924C and fOC and fO22 of of NNO+0.4, KS-3 rhyolite NNO+0.4, KS-3 rhyolite at 894-927at 894-927C and fOC and fO22 of of NNO+0.1, KS-1 rhyolite NNO+0.1, KS-1 rhyolite at 870-907at 870-907C and C and NNO+0.6. It appears that NNO+0.6. It appears that there is a weak overall there is a weak overall cooling trend from older cooling trend from older to younger magmas of to younger magmas of similar silicic similar silicic composition. composition.


01020304050607080

Freq

uenc

yKS-1

1800 BP

0

10

20

30

40

50

Freq

uenc

y

KS-26000 BP

Freq

uenc

y

0

10

20

30

40

50

60 KS-36300 BP

05

10152025303540

25 35 45 55 65 75 85 95An, mol.%

Freq

uenc

y

KS-48800 BP

Plagioclase compositions of the Holocene caldera-forming eruptions at Ksudach

2

2

2

2

Rims Mineral compositionsMineral compositionsPlagioclase is a dominant phenocryst Plagioclase is a dominant phenocryst phase in the studied KS-1, KS-3, and KS-4 phase in the studied KS-1, KS-3, and KS-4 products. Most show oscillatory-zoning.products. Most show oscillatory-zoning.

Their compositions range from AnTheir compositions range from An52.352.333 in in KS-4 to AnKS-4 to An434366 in KS-1, with KS-3 in KS-1, with KS-3 plagioclase compositions being plagioclase compositions being intermediate, which is consistent with the intermediate, which is consistent with the view that these magmas have been view that these magmas have been derived from a single, slowly cooling derived from a single, slowly cooling source.source.

However, the presence of “dusty-zoned” However, the presence of “dusty-zoned” plagioclases, in which resorbed Anplagioclases, in which resorbed An46-4946-49 cores are mantled by Ancores are mantled by An72-7572-75 inclusion-rich inclusion-rich rims, as well as rare anorthite xenocrysts rims, as well as rare anorthite xenocrysts (An(An93-9693-96) indicate that the composition of ) indicate that the composition of silicic magmas was likely modified by silicic magmas was likely modified by mixing with basalt. mixing with basalt.


Experimental results: KS-1Experimental results: KS-1

500 1000 1500 2000 2500 3000Pressure, bars

700

750

800

850

900

950

Pl

Hb

CP x out

Bt

Pl+OPx +CPx+Mt+Ilm+L+V

CrystallizationMeltingBoth melting and crystallizationP =P

~NNO+0.5

H O total2

CPx+OPx+L+V

Mt+Ilm

Phase diagram

Tem

per

atur

e,

C°

P-T phase diagram for P-T phase diagram for KS-1 rhyolite based on KS-1 rhyolite based on the phase diagram for the phase diagram for Karymsky dacite, which Karymsky dacite, which has a similar bulk has a similar bulk composition. Red composition. Red symbols indicate the symbols indicate the conditions of conditions of experiments, which used experiments, which used the natural KS-1 rhyolite the natural KS-1 rhyolite pumice (sample pumice (sample 02IPE45). The 02IPE45). The equilibrium mineral equilibrium mineral phases observed in KS-1 phases observed in KS-1 experiments match those experiments match those in Karymsky experiments in Karymsky experiments with the exception of one with the exception of one run at 900ºC, where run at 900ºC, where plagioclase is stable.plagioclase is stable.


Experimental results cont.Experimental results cont.

20

25

30

35

40

45

50

75 0 80 0 85 0 90 0 95 0

An

, m

ol.

%

Temperature

Variations of experimental plagioclase composition as a function of temperature

2Shaded areas Shaded areas correspond to the correspond to the average composition average composition (±1 sigma) of natural (±1 sigma) of natural plagioclase rims and the plagioclase rims and the average temperature average temperature estimated using Fe-Ti estimated using Fe-Ti oxides (±1 sigma).oxides (±1 sigma).

100 MPa, NNO buffer100 MPa, NNO buffer


ExperimentExperimental results al results cont.cont.Shaded areas correspond Shaded areas correspond to the average to the average composition (±1 sigma) composition (±1 sigma) of matrix glass and the of matrix glass and the average temperature average temperature estimated using Fe-Ti estimated using Fe-Ti oxides (±1 sigma) for the oxides (±1 sigma) for the KS-1 natural sample.KS-1 natural sample.

100 MPa, NNO buffer100 MPa, NNO buffer

72

73

74

75

76

77

78

SiO

2

10

11

12

13

14

15

16

AlO 2

3

1

1.2

1.4

1.6

1.8

2

2.2

2.4

CaO

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

KO 2

0

0.1

0.2

0.3

0.4

0.5

0.6

MgO Fe

O

0

0.5

1

1.5

2

2.5

3

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

700 750 800 850 900 950

Temperature

TiO

2

3

3.5

4

4.5

5

5.5

700 750 800 850 900 950

Temperature

NaO 2

Variations of experimental melt composition as a function of temperature


Summary of experimentsSummary of experiments

• Compositions of natural plagioclase, Compositions of natural plagioclase, orthopyroxene, clinopyroxene, and matrix orthopyroxene, clinopyroxene, and matrix glass of KS-1 has been reproduced glass of KS-1 has been reproduced experimentally at 892±9°C, 100 MPa, and experimentally at 892±9°C, 100 MPa, and oxygen fugacity near NNO buffer.oxygen fugacity near NNO buffer.

• Water in glass inclusions averages 3.8±1 Water in glass inclusions averages 3.8±1 wt.%, which matches saturation at ~100 wt.%, which matches saturation at ~100 MPa (Andrews et al., AGU-2003, V31G-01). MPa (Andrews et al., AGU-2003, V31G-01). The pre-eruptive pressure corresponds to The pre-eruptive pressure corresponds to approximately 3-4 km depth.approximately 3-4 km depth.


Possible scenarios: Model 1Possible scenarios: Model 1

A single, long-lived, fractionating magma chamber continually generates A single, long-lived, fractionating magma chamber continually generates silicic magmas, which periodically ascend and erupt. Eruptions may be silicic magmas, which periodically ascend and erupt. Eruptions may be triggered by basaltic replenishments, which do not however, prevent the triggered by basaltic replenishments, which do not however, prevent the magma system from a gradual cooling.magma system from a gradual cooling.

Pro’s: The volume of silicic material appears to increase in time, as well as Pro’s: The volume of silicic material appears to increase in time, as well as the concentration of SiOthe concentration of SiO22 in magmas and apparent period of quiescence in magmas and apparent period of quiescence between eruptions.between eruptions.

8800

BP

Newly generated silicic melt erupts first, followed by predominantly andesitic magma

Cretaceousbasement

Volcano-sedimentarydeposits

Compositionally stratifiedmagma body

6300

BP

The new portion of the generated by fractionation silicic magma ascends and erupts

Cretaceousbasement



6000

BP

~ 300 years later it is followed by a portion of remaining andesite magma. No silicic magma has been generated for such a short period of time.

Possible basaltic recharge

Cretaceousbasement



1800

BP

Fractionating magma body produces more silicic magma, which successfully erupts again

Compositionally stratifiedmagma bodyCretaceous

basement



Possible scenarios: Model 2Possible scenarios: Model 2

Silicic inputs originate from a partially molten basement. They segregate and ascend Silicic inputs originate from a partially molten basement. They segregate and ascend periodically, flushing an evolved, shallow-level andesitic reservoir. Basaltic magmas periodically, flushing an evolved, shallow-level andesitic reservoir. Basaltic magmas from even deeper source transport heat and matter, and may trigger volcanic from even deeper source transport heat and matter, and may trigger volcanic eruptionseruptions

Pro’s: Magmas of contrasting composition are involved in each eruption. The younger Pro’s: Magmas of contrasting composition are involved in each eruption. The younger KS-1 rhyolite has lower REE concentrations as compared to the slightly less evoloved KS-1 rhyolite has lower REE concentrations as compared to the slightly less evoloved KS-3 and KS-4 rhyodacites. Also, the oxidation state of the magmas appears to vary KS-3 and KS-4 rhyodacites. Also, the oxidation state of the magmas appears to vary with time.with time.

Cretaceousbasement


Partially molten basement -source of silicic magmas


Silicic magma intrudes and erupts mobilizing resident andesitic magma beneath the edifice.

8800

BP

Cretaceousbasement




Silicic magma ascends and erupts again

6300

BP

Resident andesitic magma erupts, possibly triggered by a fresh basaltic input

Possible basalticrecharge



Cretaceousbasement


6000

BP



Silicic magma intrudes at 3-4 km depth corresponding to ca. 100 MPa, re-equilibrates and erupts

Cretaceousbasement


1800

BP


ConclusionsConclusions The rhyolite of KS-1, the most recent caldera The rhyolite of KS-1, the most recent caldera

forming eruption at Ksudach, was last forming eruption at Ksudach, was last equilibrated at 892±9°C, 100 MPa, at water equilibrated at 892±9°C, 100 MPa, at water saturated conditions. The pre-eruptive pressure saturated conditions. The pre-eruptive pressure corresponds to approximately 3-4 km depth, which corresponds to approximately 3-4 km depth, which coincides with the regional stratigraphic boundary coincides with the regional stratigraphic boundary between Cretaceous metamorphic basement and between Cretaceous metamorphic basement and an overlying volcano-sedimentary layer.an overlying volcano-sedimentary layer.

The silicic magmas of the Holocene caldera-The silicic magmas of the Holocene caldera-forming eruptions at Ksudach most likely has forming eruptions at Ksudach most likely has originated from the same reservoir and then their originated from the same reservoir and then their temperatures may reflect its cooling from ca. temperatures may reflect its cooling from ca. 920°C at 8800 BP to ca. 890°C at 1800 BP. This 920°C at 8800 BP to ca. 890°C at 1800 BP. This reservoir, however, may not be at the depth, reservoir, however, may not be at the depth, where the erupted magmas has last equilibrated.where the erupted magmas has last equilibrated.

Documents

Petropavlovsk-Kamchatsky, August 2004 RECURRENT CALDERA-FORMING ERUPTIONS: KSUDACH CASE STUDY Preliminary results of the ongoing petrological and experimental