научная статья по теме KOS PLATEAU TUFF (KPT) ON KALYMNOS ISLAND, AEGEAN VOLCANIC ARC: A GEOCHEMICAL APPROACH Геофизика

Текст научной статьи на тему «KOS PLATEAU TUFF (KPT) ON KALYMNOS ISLAND, AEGEAN VOLCANIC ARC: A GEOCHEMICAL APPROACH»

УДК 550

KOS PLATEAU TUFF (KPT) ON KALYMNOS ISLAND, AEGEAN VOLCANIC ARC: A GEOCHEMICAL APPROACH

© 2013 г. Dimitrios Zouzias1, Karen St. Seymour1, 2

laboratory of Ore Deposits and Volcanology, Department of Geology, University of Patras, Rion 26500 Hellas (GR),

e-mail: dizouzias@upatras.gr; dizouzias@gmail.com 2Department of Geography, Urban Planning and Environment, Concordia University, Montreal Canada H3G1M8,

e-mail: kstseymr@upatras.gr Поступила в редакцию 06.07.2010 г.

Thirteen pumice samples from the D and E ignimbrite units of Kalymnos Tuff have been analyzed for their biotite and feldspar phenocryst mineral chemistry and for bulk major and 20 trace, including 14 Rare Earth elements, to define and compare their petrochemistry with the Kos Plateau Tuff (KPT). For the same purpose major element analyses were obtained from Kalymnos Tuff and KPT glasses. Both KPT and Kalymnos pumice lapilli are rhyolites characterized by a well-developed 'silky' texture and roundish quartz. Phenocrysts of biotite and feldspars (sanidine, oligoclase) from both tuffs display compositional overlap. Crystals are characterized by undulatory extinction (quartz), fractures (sanidine, oligoclase) and bent cleavages (biotite) due to the explosive origin of their host. Both tuffs show well-defined petrogenetic trends and extensive compositional overlaps on major and trace element variation diagrams suggesting that they are consanguineous. However, D ignimbrite samples are more evolved than those obtained from E ignimbrite as indicated from major elements, alkali earths (Ba, Rb, Sr), immobile (Zr, Y), compatible (V) and hygromagmatophile trace element (Th) distributions. This evidence indicates a stratified magma chamber under a ~l6 Km caldera superstructure which is mostly submarine.

DOI: 10.7868/S0203030613050040

1. INTRODUCTION

It has been suggested by Allen [2001] that the Kos Plateau Tuff (KPT) has been erupted 161 Ka ago [Smith et al., 1996] from a superstructure now outlined by the remains of a ~16 Km caldera which is mostly submarine. Our studies using satellite imagery, subaqueous distribution of hydrothermal vents and tectonic data of ring and radial faults supports the evidence of such a caldera superstructure [St. Seymour et al., 2006; Zouzias, St. Seymour, 2006] very similar to that proposed by Allen [2001]. Kalymnos lies roughly 20 Km from Kos and represents the northernmost border of deposition of the KPT according to physical volcanology arguments presented by Allen [2001]. She argued that pyroclastic density currents from a major eruption, which occurred south of Kos 161 Ka ago [Smith et al., 1996], deposited tuffs on Kalymnos, Tilos, on the coasts of Asia Minor and other islands in the area. The apron of this ash has travelled, due to the predominant trade winds in the Aegean Sea, 300 km to the south and is thought to be equivalent with the W-3 submarine tephra layer [Federman, Carey, 1980; Vinci, 1985]. Pyroclastic density currents from the 161 ka Kos eruption traveled either on land [Pe-Piper et al., 2005] or crossed open sea [Allen, 2001; Allen, Cas,

2001; Bohla, Keller, 1987; Stadlbauer et al., 1986]. This paper presents the geochemical arguments for the consanguineous relationship of the KPT and Kalymnos pumices based on major, trace and Rare Earth element and mineral chemistry new data obtained from these tuffs. The geochemical arguments presented here entirely support conclusions derived from physical volcanology observations by Allen [2001, 1998], Allen and Cas [2001, 1998] and Allen et al. [1999].

2. ANALYTICAL METHODS

Mineral chemistry was analyzed at McGill University, Montreal, using a JXA JOEL 8900L instrument equipped with wavelength-dispersive spectrometers (WDS). Electron microprobe work on petrographic polished thin sections was performed at an accelerator voltage of 15 kV, a beam current of 20 nA and a beam diameter of 5 microns. On line ZAF corrections were performed utilizing CAMECA software. For glass analyses a highly defocused beam (20 ^m) and a beam current of15 nA was used to prevent devolatilization of alkalies from the glass [St. Seymour et al., 2004]. Natural and synthetic feldspar, diopside, rutile, chromite, hematite, spessartine, fluorite, vanadinite, obsidian and basaltic glass were used as standards. For the feld-

spars, the detection limits are: 0.03 wt % for Si, 0.05 wt % for Ti, 0.02 wt % for Al, 0.04 wt % for Fe, 0.02 wt % for Mg, 0.03 wt % for Ca, 0.07 wt % for Ba, 0.03 wt % for K and for Na 0.03 wt %. For ferromagnesian minerals the detection limits are: 0.03 wt % for Si, 0.06 wt % for Ti, 0.03 wt % for Al, 0.05 wt % for Fe, 0.05 wt % for Mn, 0.02 wt % for Mg, 0.04 wt % for Ca, 0.03 wt % for Na, 0.02 wt % for K, 0.17 wt % for F and 0.02 wt % for Cl. For the glasses, spessartine, vanadinite, apatite, obsidian and basaltic glass were used as standards and the detection limits are: 0.04 wt % for Si, 0.06 wt % for Ti, 0.03 wt % for Al, 0.06 wt % for Fe, 0.05 wt % for Mn, 0.03 wt % for Mg, 0.04 wt % for Ca, 0.05 wt % for Na, 0.03 wt % for K and 0.01 wt % for Cl.

Thirteen bulk samples from the D and E ignimbrite units as defined by Allen [2001], from Kalymnos (samples KL-1 to KL-17) have been analyzed for major, trace and Rare Earth elements in the Actlabs laboratory of Ontario Canada and the analytical results are presented in Table 1. Major elements have been analyzed by ICP methods under analytical code WRA + trace 4Litho. Accuracy is at the 0.001 wt % level for TiO2 and MnO and 0.01 % for the rest of the major elements. Trace elements and REE have been analyzed by ICP/MS methods. Detection limits are 0.05 ppm for Pr, Eu, Tm and 0.04 ppm for Lu and 0.1 ppm for the rest of the REE. Detection limits for Sc, Be, Co, Ga, Ge, Nb, Sb, Cs, Hf, Ta, Tl, Bi, Th, U and W are 1 ppm or better. For Sr, Y and Rb are 2 ppm, for Ba and Zr are 3 ppm and 4 ppm respectively, for V and Pb are 5 ppm and for Zn are 30 ppm. Probe mineral analyses are presented in Tables 2 and 3 and glass analyses in Table 4.

3. SAMPLING 3.1 Geological setting and location of samples

The studied samples represent part of a record of volcanic activity in a key area that comprises Miocene high-K volcanic products deposited in easterly trending grabens [Ulusoy et al., 2004] which continue into the Aegean Sea [Kurt et al., 1999; Doutsos, Kokkalas, 2001; Bozkurt, 2003] and subduction-related calc-al-kaline volcanic rocks of the presently active Quaternary Hellenic Volcanic Arc. Subduction-related centers in the area (Fig. 1, inset) have been particularly explosive and because the depositional setting has been mostly marine the primary pyroclastic subaerial products from the eruptions are often limited [Allen et al., 1999]. Caldera structures are at the best interpreted from volcanological data [Allen et al., 1999; Dala-bakis, Vougioukalakis, 1993], from remotely sensed (satellite) images and structural data [St. Seymour et al., 2006; Zouzias, St. Seymour, 2006]. Of these, the 161000 yrs B.P. Kos Plateau Tuff eruption gave the most

voluminous pyroclastic products with a minimum estimated volume ~60 km3 D.R.E of which ~9 km3 was erupted during the emplacement of earlier KPT units A, B and C [Allen, Cas, 1998; Allen et al., 1999]. Distal submarine fall layers (W-3 ash) were deposited at a distance up to 300 km south of Kos [Federman, Carey, 1980; Vinci, 1985]. The intensity of this explosion is depicted by the occurrence of quite unusual platy-shaped pumice clasts in pyroclastic fall and density current deposits due to a shear mechanism in the magma conduit [Palladino et al., 2008].

In Kalymnos the pumice is the youngest alloctho-nous unit. Pumice samples were collected from two northwesterly trending fault—related valleys, these of Horion and Vathy grabens (Fig. 1, Table 5). Pumice exposure in the Horion graben is limited due to human settlement. The Horion valley is underlain by a Neopaleozoic basement which consists of limestone and schist—sandstone with lenses of mafic metavolca-nic rock. Triassic to Upper Jurassic limestone, dolo-mitic limestone and dolomite lie transgressively on Upper Paleozoic formations. These are overlain by massive to thick—bedded crystalline Lower Cretaceous limestone (Fig. 1).

On western Kos, the KPT overlies the Kefalos tuff ring deposits (3.4—2.7 Ma) and Triassic Lower Cretaceous limestone [Dalabakis, Vougioukalakis, 1993]. Allen [1998] and Allen et al. [1999] subdivided the KPT into six mappable, texturally distinct, strati-graphic units from oldest to youngest A, B, C, D, E and F. Units A and the upper part of F are mainly fallout deposits and the rest have resulted from pyroclastic density current deposition such as the D and E horizons which comprise mappable ignimbrite subunits [Stadlbauer et al., 1986; Allen et al., 1999; Allen, 1998]. Our samples represent the most pristine possible pumice lapilli- and bomb-size samples obtained from units D and E both on Kos and Kalymnos.

4. RESULTS 4.1 Petrography

A phenocryst assemblage of plagioclase — biotite — quartz and less abudant K-feldspar embedded in highly vesiculated clear glass characterizes the Kalymnos and KPT tuffs. Accessory minerals include monazite, zircon, apatite, ilmenite and magnetite. Rare clinopy-roxene and amphibole crystals showing signs of resorption are locally present in these tuffs and have a probable xenocrystic origin from the deeper levels of the volcanic reservoir. All phenocrysts display textures of fracturing to various degrees, attributed to the explosive origin of their host.

Table 1. Geochemistry of Kalymnos pumice: Major element contents in wt % and trace element and REE's in ppm

KL1 KL2 KL3 KL4 KL6 KL7 KL8 KL9 KL10 KL14 KL15 KL16 KL17

SiO2 % 71.91 73.3 73.06 72.55 70.86 72.41 73.58 73.26 72.25 70.97 69.75 70.94 71.04

TiO2 % 0.18 0.191 0.172 0.184 0.188 0.201 0.184 0.187 0.18 0.18 0.18 0.18 0.17

Al2O3 % 13.08 12.6 13.13 13.5 12.82 13.1 12.62 13.28 12.96 13.23 14.19 12.52 13.34

Fe2O3(T)% 1.

Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.

Показать целиком