ArticlePDF Available

Geochemical characteristics of paleotsunami deposits from the Shizuoka plain on the Pacific coast of middle Japan

Authors:

Abstract and Figures

The Pacific coast of Japan has repeatedly suffered earthquakes with magnitudes greater than Mw 8 and accompanying tsunami on coastal areas. Estimating the inundation area of paleotsunami using event deposits in sediment layers can inform and reduce possible damages from future earthquakes and tsunami. In addition to geological and sedimentological analyses, the geochemical signature of sediments can be an effective proxy for identifying event deposits in many cases. However, few geochemical analyses have been applied to study the Pacific coast of middle Japan. Therefore, the geochemical characteristics of paleotsunami deposits (∼1000, 3500, and 4000 cal BP) were assessed for core samples from the Shizuoka plain (Oya lowland) on the Pacific coast of middle Japan. In a ternary diagram for (Na2O+CaO)-Al2O3-(Cr+Ni), our data from the paleotsunami deposits from the Shizuoka plain plotted between those of the beach core samples (shoreface and dune deposits) and mud layers in the back marsh deposits. Additionally, vertical and horizontal distributions of titanium normalized values (Na/Ti, Sr/Ti, Ba/Ti, and Cr/Ti atomic ratios) and Si/Al atomic ratios in the cores from the Shizuoka plain provided important clues for discrimination of the paleotsunami deposits from other layers, such as flood deposits from river overflow and mud layers that settled in calm environments. Based on the cluster analysis, the data from the paleotsunami deposits (∼3500 cal BP) were discriminated from those of other layers in our case.
Content may be subject to copyright.
325
Geochemical Journal, Vol. 55, pp. 325 to 340, 2021 doi:10.2343/geochemj.2.0641
*Corresponding author (e-mail: watanabe.takahiro46@jaea.go.jp)
Copyright © 2021 by The Geochemical Society of Japan.
(Mw) of past earthquakes have been previously estimated
using geological records (Minoura et al., 1994; Nanayama
et al., 2003; Goto et al., 2014; Chagué-Goff et al., 2015;
Sawai et al., 2009, 2012, 2015).
An event deposit is preliminarily found using grain
size changes in a geologic stratum along with multiple
other proxies for environmental changes (Watanabe et al.,
2020). Additionally, geochemical analysis is widely used
to estimate the source of the sediments (Minoura and
Nakaya, 1991; Chagué-Goff et al., 2017). Following re-
cent major subduction zone earthquakes, the analysis of
tsunami deposits has been increasing to help clarify such
geochemical features (Chagué-Goff et al., 2017; Watanabe
et al., 2020). The geochemical signatures of tsunami de-
posits from the southwestern coasts of Spain, New Zea-
land, the Hawaiian Islands, Mexico and northeastern Ja-
pan have also been studied (Minoura et al., 1994; Chagué-
Goff, 2010; Roy et al., 2012; Kuwatani et al., 2014;
Yamasaki et al., 2015; Goff et al., 2020). Recently,
geochemical characteristics of tsunami deposits have been
well documented on the Pacific coast in the Tohoku
Geochemical characteristics of paleotsunami deposits from the Shizuoka plain
on the Pacific coast of middle Japan
TAKAHIRO WATANABE,1* NORIYOSHI TSUCHIYA,2 AKIHISA KITAMURA,3
SHIN-ICHI YAMASAKI2 and FUMIKO W. N ARA4,5
1Toki Geochronology Research Laboratory, Tono Geoscience Center, Japan Atomic Energy Agency,
959-31 Jorinji, Izumi, Toki, Gifu 509-5102, Japan
2Graduate School of Environmental Studies, Tohoku University, 6-6-20 Aramaki-Aza-Aoba, Aoba, Sendai 980-8579, Japan
3Institute of Geosciences, Shizuoka University, 836 Ohya, Suruga-ku, Shizuoka 422-8529, Japan
4Institute of Nature and Environmental Technology, Kanazawa University, Kakuma, Kanazawa 920-1192, Japan
5Graduate School of Environmental Studies, Nagoya University, Furo-cho, Nagoya 464-8602, Japan
(Received October 16, 2020; Accepted June 17, 2021)
The Pacific coast of Japan has repeatedly suffered earthquakes with magnitudes greater than Mw 8 and accompanying
tsunami on coastal areas. Estimating the inundation area of paleotsunami using event deposits in sediment layers can
inform and reduce possible damages from future earthquakes and tsunami. In addition to geological and sedimentological
analyses, the geochemical signature of sediments can be an effective proxy for identifying event deposits in many cases.
However, few geochemical analyses have been applied to study the Pacific coast of middle Japan. Therefore, the geochemical
characteristics of paleotsunami deposits (~1000, 3500, and 4000 cal BP) were assessed for core samples from the Shizuoka
plain (Oya lowland) on the Pacific coast of middle Japan. In a ternary diagram for (Na2O+CaO)-Al2O3-(Cr+Ni), our data
from the paleotsunami deposits from the Shizuoka plain plotted between those of the beach core samples (shoreface and
dune deposits) and mud layers in the back marsh deposits. Additionally, vertical and horizontal distributions of titanium
normalized values (Na/Ti, Sr/Ti, Ba/Ti, and Cr/Ti atomic ratios) and Si/Al atomic ratios in the cores from the Shizuoka
plain provided important clues for discrimination of the paleotsunami deposits from other layers, such as flood deposits
from river overflow and mud layers that settled in calm environments. Based on the cluster analysis, the data from the
paleotsunami deposits (~3500 cal BP) were discriminated from those of other layers in our case.
Keywords: tsunami deposits, geochemical signatures, titanium normalized values, cluster analysis, Shizuoka plain
INTRODUCTION
Interplate earthquakes repeatedly occurred in subduc-
tion zones and they had caused damages, in many cases
from tsunami affecting a wider coastal area (Kanaori et
al., 1993; Kitamura et al., 2013; Goto et al., 2014; Sawai
et al., 2015; Chagué-Goff et al., 2017). For example, the
islands of Japan are located on the margin of continental
plates and suffer from repeated large earthquakes and tsu-
namis (Satake and Atwater, 2007; Satake, 2014). The
paleotsunami deposits within sedimentary records pro-
vide evidence of past disasters in the area (e.g., Goff et
al., 2020). Therefore, the identification of paleotsunami
deposits is important for the risk assessment of major
subduction zone earthquakes (Satake and Atwater, 2007;
Satake, 2014). Additionally, the recurrence intervals, tsu-
nami inundation areas, and maximum moment magnitudes
326 T. Watanabe et al.
(northeast Japan), to determine the source of the deposits
(Chagué-Goff et al., 2014; Shinozaki et al., 2015;
Watanabe et al., 2020).
Following the 2011 Tohoku tsunami, Kuwatani et al.
(2014) and Watanabe et al. (2020) analyzed modern and
paleotsunami deposits from northeastern Japan using
geochemical and statistical methods. They have suggested
that inorganic geochemical approaches using alkali and
alkaline-earth elements (such as Na and Sr) measured by
X-ray fluorescence spectrometry (XRF) are an effective
method for detecting paleotsunami deposits in the Tohoku
area. Generally, geochemical characteristics of tsunami
deposits are site-specific (Chagué-Goff et al., 2017), e.g.,
in the cases of different hinterlands, sediment sources,
sedimentation dates, topographic and lithological condi-
tions (Chagué-Goff et al., 2012a, 2012b, 2014;
Szczuci´nski et al., 2012; Shinozaki et al., 2015, 2016).
In previous studies on Japan, geochemical studies of
paleotsunami deposits were limited to the Pacific coast
of Tohoku (e.g., Goff et al., 2020; Watanabe et al., 2020,
2021) and Miyazaki plain in Kyusyu island (Niwa et al.,
2020). Therefore, we carry out a geochemical analysis of
the paleotsunami deposit from the Shizuoka plain, to pro-
pose simple methods for identifying marine incursion
generated from a tsunami disaster in the Pacific coast of
middle Japan.
In this study, we performed inorganic geochemical
analyses of <2 mm fraction for the three paleotsunami
deposits (~1000, 3500, and 4000 cal BP), flood deposits,
mud layers (peaty clays; parallel lamina layers), shoreface
and dune deposits in cores from the Oya lowland, in the
Shizuoka plain on the Pacific coast of middle Japan (Figs.
Fig. 1. (A) Map showing the location of the study area in Shizuoka on the Pacific coast of middle Japan. (B) Expanded map
around Shizuoka, (C) the Oya lowland in southeast Shizuoka, and (D) the sampling sites of the boring cores (Sites 1–6) in the Oya
lowland. The core (Site 7) and modern beach sands (BS-1) were also taken from the beach on the Suruga bay coast (Kitamura et
al., 2013). Dashed lines denote the plate boundary between the Eurasian and Philippine Sea plates (Nankai-Suruga trough).
Original map data collected by the Geospatial Information Authority of Japan (GIA) and Hydrographic and Oceanographic
Department of Japan Coast Guard (JCG) were used in (A)–(C), and the plate boundary, place names, latitude/longitude, scales
were added. https://cyberjapandata.gsi.go.jp
EUR: Eurasia Plate; NA: North American Plate; PAC: Pacific Plate; PHS: Philippine Sea Plate
Geochemical characteristics of paleotsunami deposits from the Shizuoka 327
1A–D). The paleotsunami deposits were assayed for ma-
jor and trace elements (Na-Pb) by energy dispersive X-
ray fluorescence spectrometry (EDXRF; Supplementary
Table S1). Statistical analyses were also performed to
characterize and discriminate the paleotsunami deposits.
Principal component analysis (PCA) and cluster analysis
have been widely used on geochemical and clay mineral
data from core samples to estimate the chemical charac-
teristics, sources, material cycles, and chemical weather-
ing trends (Fagel and Mackey, 2008; Kuwatani et al.,
2014; Chagué-Goff et al., 2018). This study presents the
first geochemical dataset of the paleotsunami deposits on
the Pacific coast of middle Japan. Most coastal areas in
Japan have been disturbed and filled in for reinforcement
project during the early modern period, and consequently,
it is not easy to obtain paleotsunami deposits from these
areas, whose geochemical aspects have thus far not been
sufficiently studied.
PALEOTSUNAMI DEPOSITS ON THE SOUTHWEST JAPAN
Along south of the southwest Japanese islands, the
Nankai and Suruga troughs between the Philippine Sea
plate and Eurasian plate has formed along Suruga Bay
(Shizuoka) up to the area offshore from Shikoku and
Kyushu islands (Fig. 1A; Cummins et al., 2002;
Yamamoto et al., 2013). Paleotsunami deposits from late
Holocene were found on the coastal area of the Kii pe-
ninsula, Shikoku island, and Kyushu island (Fig. 1;
Nankai and Tonankai area; southwest Japan; Okamura and
Matsuoka, 2012; Fujiwara et al., 2020; Niwa et al., 2020),
as well as in Shizuoka, on the Pacific coast of middle
Japan (Tokai area; Fujiwara et al., 2013; Kitamura et al.,
2013, 2016; Kitamura and Kobayashi, 2014). The risk of
earthquakes and tsunamis in Tokai-Tonankai-Nankai has
been recognized by geologists and geophysicists for a long
time (Yoshioka and Hashimoto, 1989; Rydelek and Sacks,
2003; Hori et al., 2004; Baba et al., 2006). Additionally,
the importance of tsunami geology for disaster preven-
tion in the case of subduction zone earthquakes has been
also recognized, especially, following the 2011 Tohoku
earthquake (Mw 9) in northeast Japan (e.g., Goto et al.,
2014).
On the Pacific coast of middle Japan, at least six ma-
jor earthquakes (Mw > 8) had occurred since the 7th cen-
tury. This area faces the Nankai and Suruga troughs, and
the last major earthquake occurred in 1944 Common Era
(CE) with tsunami height under ~2 m, around the coast
of the Suruga Bay (Tonankai Earthquake). In 1854 CE
(Ansei Tokai earthquake) and 1707 CE (Hoei earthquake),
the tsunami height was estimated to be up to ~6 m. The
Keicho (1605 CE) and Meio (1498 CE) earthquakes also
occurred in this area. Additionally, old accounts of tsu-
nami based on historical documents were recorded in
1096, 887, and 684 CE (the Eicho, Ninna, Hakuho earth-
quakes, respectively; Kitamura et al., 2013).
Paleotsunami deposits during the last ~4000 years have
been well surveyed on the Shizuoka plain using cores and
outcrops from archaeological sites (Fujiwara et al., 2013,
2020; Kitamura et al., 2013, 2020). In addition,
paleotsunami deposits have been detected in the
Hamamatsu plain and Iwata area, west of the Shizuoka
plain (Fujiwara et al., 2013, 2020; Kitamura et al., 2013).
Recent tsunami deposits (15th and 19th century) have
been also found at the Izu peninsula, east of the Shizuoka
plain (Fig. 1; Kitamura et al., 2013; Kitamura, 2016). The
identification of paleotsunami deposits is achieved by
using multiple criteria, such as the dates of event layers,
lateral extent of the layer, changes in grain size, round-
ness of minerals, sedimentological features, microfossils,
and geochemical signatures (e.g., Goff et al., 2004; Moore
et al., 2006; Nelson et al., 2008; Kortekaas and Dawson,
2007; Nakamura et al., 2012; Sawai et al., 2012).
SAMPLES AND ANALYTICAL METHODS
Samples from the Oya lowland on Shizuoka plain
The Shizuoka plain is an alluvial fan of the Abe River
(~5‰ slope from the lower reach to river mouth), with
beach ridges found along the coast of Suruga Bay
(Kitamura et al., 2013). The Oya lowland is located on
the southeast Shizuoka plain between the Udo Hills and
Abe River fan (Fig. 1D), and the altitude of the lowland
is ~7–8 m above mean sea level at present (Kitamura et
al., 2013). Six borehole cores from the Oya lowland (Sites
1, 2, 3, 4, 6, and 7; Kitamura et al., 2013) and modern
beach sand on Suruga Bay coast, from the Shizuoka plain
in middle Japan, were used in this study (Fig. 1, Supple-
mentary Tables S2 and S3). The length of each boring
core was ~8 m, and landfill layers of ~1-2 m were ob-
served at Sites 1–6 (Kitamura et al., 2013). The cores
from Sites 1–6 contained mainly back marsh deposits (in-
cluding paleotsunami deposits, flood deposits, peaty clays,
and muddy parallel lamina layers), river channel depos-
its with gravels, and the landfill layer. The core from Site
4, close to the Oya River, contained thick river channel
deposits (Kitamura et al., 2013). Mud contents in the cores
have been reported by Kitamura et al. (2013) (Table 1;
Tables S2 and S3).
A paleotsunami deposit layer with a radiocarbon age
of ~1000 cal BP and archaeological remains (T0 tsunami
deposit layer; 4-T0) were found in Site 4 core (Kitamura
et al., 2013). The cores from the Suruga Bay side (Sites 1
and 3) also contained layers with paleotsunami deposits
(~3500 cal BP, T1 tsunami deposit layer; 1-T1 and 3-T1).
Additionally, the cores from the inland side (Site 2) con-
tained two paleotsunami deposit layers (~3500 cal BP for
the 2-T1 and ~4000 cal BP for the 2-T2) and flood de-
328 T. Watanabe et al.
posit layers (Kitamura et al., 2013). Kitamura et al. (2013)
used grain roundness, grain-size, sedimentary structures,
and diatom assemblages to estimate the sources of the
deposits. These fluvial flooding deposits were found based
on the grain shape of the sand (angular), which resem-
bled those seen in the river mouth bar (Kitamura et al.,
2013). Site 6 core, which was taken further inland at dis-
tance of approximately 1.6 km from the modern shore-
line, contained the T1 tsunami deposit layer (6-T1). The
ages of the T1 and T2 paleotsunami deposits have been
estimated based on radiocarbon dating of the upper and
lower adjacent layers in each core (Kitamura et al., 2013).
The key bed layers with volcanic clasts from the Kikai-
Akahoya tephra (K-Ah; ~7,300 cal BP) were also found
in each core (Sites 1–6; Kitamura et al., 2013). The K-
Ah tephra was widely dispersed over the Japanese Islands
(from Okinawa to south Tohoku) in the middle Holocene
(Machida and Arai, 2003). Site 7 core was from a beach
(~9 m in length), and modern beach sands (BS-1) were
also taken from the surface close to the modern shoreline
(Fig. 1D; Kitamura et al., 2013). The core contained
mainly gravel and medium-coarse sand (shoreface and
dune deposits; Kitamura et al., 2013). From Site 7 core,
five samples were taken for geochemical analysis, for
comparison to the paleotsunami deposits from the other
cores (Tables S2 and S3).
Pretreatments
Fifty-two samples from the cores (Sites 1, 2, 3, 4, 6
and 7; including six paleotsunami deposit samples with
three different ages) and one sample of the modern
backshore deposit, from the Oya lowland in the Shizuoka
plain, were air-dried and the samples with a grain size of
<2 mm were obtained by dry-sieving (Fig. 1, Supplemen-
tary Tables S2 and S3). The cores were sub-sampled at
intervals of ~1 cm interval, and the outer rims of the sam-
ples were removed to avoid contamination. The separate
samples were freeze-dried. Then, each dry sample was
powdered by a vibration mill (TI-100, CMT, Japan). The
powdered samples were pellet-pressed with polyvinyl
chloride rings, with a diameter of 2.5 cm, under a weight
of 20 tons by a manual hydraulic press (GS25011, Specac,
UK); the pellets were used for quantification of the ma-
jor and trace elements using the EDXRF equipped with
three-dimensional polarization optics.
Analysis
Twenty-four elements were measured in the cores and
surface samples by an EDXRF (Epsilon 5, PANalytical,
B.V., now Malvern Panalytical, B.V. the Netherlands)
equipped with a 600 W X-ray tube (Sc/W dual target) at
the Tohoku University. The EDXRF was also equipped
with selectable secondary targets (Al, CaF2, Fe, Ge, Zr,
Mo, and Al2O3; Table S1; Matsunami et al., 2010;
Yamasaki et al., 2011) for the suitable excitation energy
of each element, and had a three-dimensional design to
reduce the X-ray tube’s spectrum. In a previous study
using Cr analyses, these contents had been underestimated
Paleotsunami deposits* Flood deposits Mud layers** shoreface and dune deposits***
T1 (n = 4) T0, T1, T2 (n = 6) (n = 11) (n = 31) (n = 5)
Na/Ti (atomic ratio)
Max. 16.5 16.5 14.2 10.8 19.8
Min. 14.6 12.3 9.3 7.3 16.0
Average 15.9 14.7 11.6 9.2 17.6
SD ± 1
σ
0.9 2.0 1.5 0.9 1.4
Si/Al (atomic ratio)
3.53.64.84.8.xaM 10.5
2.89.38.43.63.7.niM
3.96.45.54.79.7egarevA
SD ± 1
σ
0.5 0.8 0.5 0.5 1.0
Mud content (mass%)****
7.8.xaM 14.8 57.8 99.8 6.2
6.67.4.niM 19.3 40.1 1.5
3.7egarevA 10.7 34.8 81.7 4.7
SD ± 1
σ
1.8 5.8 14.0 17.4 2.0
Table 1. Average Na/Ti and Si/Al atomic ratios of the paleotsunami deposits and other layers from the boring cores
(Sites 1, 2, 3, 4, 6, and 7) at the Oya lowland, Shizuoka plain on the Pacific coast of middle Japan
*T0: ca. 1000 cal BP, T1: ca. 3500 cal BP, T2: ca. 4000 cal BP (Kitamura et al., 2013).
**Peaty clay and parallel lamina layers.
***Site 7 core.
****Kitamura et al. (2013).
Geochemical characteristics of paleotsunami deposits from the Shizuoka 329
by acid digestion process in many cases, especially for
soil samples derived from serpentinite, including chromite
(Yamasaki et al., 2016). Because of the undissolved
residues following acid digestion for pretreatment, Cr and
Zr analyses have shown relatively lower values than those
from other methods such as neutron activation analysis
(NAA; Yamasaki et al., 2016). Although the acid diges-
tion residue could be analyzed, the amount of the residues
is too small to measure in most cases (usually less than
10 mg; Yamasaki et al., 2016). Therefore, in this study,
we used the EDXRF for geochemical analysis of <2 mm
fraction from the tsunami deposits and other samples.
The concentrations were calibrated using the stand-
ard reference materials JA-1, JA-2, JA-3, JB-1a, JB-2,
JB-3, JF-2, JGb-1, JGb-2, JR-1, JR-2, JR-3, JLk-1, JSd-
1, JSd-2, JSd-3, JSl-1, JSl-2, JSO-2 (The National Insti-
tute of Advanced Industrial Science and Technology, Ja-
pan; https://gbank.gsj.jp/geostandards/welcome.html),
NIST2704, NIST2709, NIST2710, NIST2711 (National
Institute of Standards and Technology, USA; NIST, 1993a,
1993b, 2000, 2002), SO-1, SO-2, SO-3, SO-4 (National
Research Council, Canada; Roubault et al., 1970),
JSAC0401, JSAC0402, JSAC0403, and JSAC0411 (The
Japan Society for Analytical Chemistry). Quantification
procedures by the EDXRF using pressed powder pellets
have been well studied by Matsunami et al. (2010),
Yamasaki et al. (2011, 2015, 2016), Kuwatani et al.
(2014), Watanabe et al. (2020, 2021), and Nara et al.
(2021). In previous study for tsunami deposits from north
Sendai plain, total contents of major elements were ~90–
100 mass% in most case (Watanabe et al., 2020). How-
ever, relative low total contents were observed in this
study (76–95 mass%, ~83 mass% on average; Table S2),
possibly caused by organic materials, dehydration of clay
minerals, and/or complicated chemical forms in untreated
natural samples (such as chloride and carbonate) from the
Oya lowland on Shizuoka plain. For tsunami deposits in
other sites, relatively high LOI (loss of ignition) values
close to ~10–20% have been also observed in lowland
areas including organic-rich materials (Goff et al., 2010;
Chagué-Goff et al., 2017). Therefore, we also evaluated
the geochemical characteristics in the tsunami deposits
using element ratios, such as titanium normalized values
in this study.
The compositional data were processed by centered
logratio transformation with the zero replacement to avoid
constant-sum constraint based on Aitchison (1982),
Kucera and Malmgren (1998), Arai and Ohta (2006), and
Ohta and Arai (2006). Subsequently, correlation matrix
and PCA of the processed geochemical data were calcu-
lated using R version 3.5.2 (http://cran.r-project.org/).
Cluster analysis was also performed using Euclidean dis-
tance and Ward’s method with the R software (R Core
Team, 2018; Maechler et al., 2018), for the normalized
geochemical data in the cores (Sites 1, 2, 3, 4, 6, and 7)
and modern beach sands (BS-1) from the Oya lowland in
the southeastern Shizuoka plain, on the Pacific coast of
middle Japan.
RESULTS
Site 1 core
Concentrations of major and trace elements in Site 1
core are shown in Tables S2 and S3 and Supplementary
Fig. S1. The SiO2 contents varied from 54.7 to 63.6
mass%, and a clear peak of the contents up to 63.6 mass%
was observed on the layer with the paleotsunami depos-
its (1-T1; 349 cm in core depth; ~3500 cal BP; Tables S2
and S3 and Fig. S1). In the layer with the 1-T1, the de-
pletion of Al2O3, TiO2, K2O, Rb, Y, Zr, La, Ce, and other
heavy metals (Cu, Zn, Pb) contents was also observed.
The Na2O contents exhibited a slight increase in the 1-
T1 up to 2.2 mass%. In contrast, MgO, CaO, Sr, and Ba
had no obvious variation around 1-T1. In the layer with
the flood deposits (1-S6; 400 cm in core depth), a de-
crease of the Na2O contents was observed, and other ele-
ments exhibit no clear variation within the layer. In the
lower mud layers (406–559 cm in core depth, above the
Kikai-Akahoya tephra layer), relatively high contents of
Fe2O3 and S were found up to 5.2, and 1.7 mass%, re-
spectively.
Site 2 core
The SiO2 contents also varied from 53.1 to 67.2 mass%
in Site 2 core (Tables S2 and S3 and Supplementary Fig.
S2). Although an increase of SiO2 contents was observed
from 378 to 355 cm depth (from 53.1 to 66.2 mass%) in
the T1 (2-T1; 355 cm in core depth; ~3500 cal BP), vari-
ation in the SiO2 contents was also found in other mud
layers and flood deposits in the core (Fig. S2A). In the
layers with the 2-T1 and 2-T2 (2-T2; 455 cm in core depth;
~4000 cal BP), the depletion of Al2O3 down to 8.5 mass%
was also observed (Fig. S2B), and another ten elements
also decreased in 2-T1 and 2-T2, the same as in Site 1
core (Al2O3, TiO2, K2O, Cu, Rb, Y, Zr, La, Ce, Pb) in the
2-T2. The Na2O contents was almost constant through-
out the core, but it showed a slight increase in 2-T1 and
2-T2. The MgO, CaO, Sr and Ba indicated no clear fluc-
tuation around the 2-T1 and 2-T2. In the layer with flood
deposits (2-S12, S10, S9, S8, and S7), a slight increase
of SiO2 contents and the depletion of Al2O3 and TiO2 were
also observed (Fig. S2). In the mud layers between 2-T1
and 2-T2 (356-435 cm in core depth), relatively high
Fe2O3, MnO, S, and As contents were observed. In addi-
tion, the contents of these elements also exhibited high
values in the lower mud layers (458–660 cm in core
depth). The Cr contents increased from bottom of the core
to the 2-T2, up to ~130 mg/kg (Fig. S2).
330 T. Watanabe et al.
Site 3, 4, and 6 cores
For the T0 paleotsunami deposit in Site 4 (4-T0, 375
cm core depth; ~1000 cal BP), the SiO2 content was rela-
tively low (56.8 mass%) in comparison to those of the T1
and T2 from other cores (Sites 1, 2, 3, and 6). Relatively
high Fe2O3 and Cr contents were observed (4.1 mass%
and 199 mg/kg, respectively) in the 4-T0 (Table S3). The
chemical compositions of the T1 in Sites 3 and 6 (3-T1 in
the layer of 397 cm depth of Site 3; 6-T1 in the layer of
440 cm depth of Site 6) were similar to those of the T1
from Sites 1 and 2 (Tables S2 and S3; ~3500 cal BP). The
contents of SiO2, Al2O3 and Na2O in the 3-T1 and 6-T1
were 63.5–64.3, 6.5–6.6, and 2.1–2.2 mass%, respectively.
For the four flood deposits in Site 6 (6-S13-S16), SiO2
varied from 56.8 to 62.1 mass%, and TiO2 was relatively
higher (0.41–0.50 mass%) than those of the 6-T1 (0.35
mass%, Table S2). The Na2O contents in the flood de-
posits (2.0–2.3 mass%) were similar to those of the 6-T1
(2.2 mass%).
Site 7 core and BS-1
The SiO2, Al2O3, and Na2O contents ranged from 62.6–
68.7, 5.5–6.6, and 2.2–2.5 mass%, respectively, in Site 7
core (Table S2). The TiO2, Sr, Zr, Ba, La, and Ce con-
tents were almost constant throughout the core (Table S2
and S3). The MgO and CaO contents fluctuated from 1.9
to 3.1 and 0.4 to 0.7 mass%, respectively. In addition, the
Cr and Ni contents varied from 160 to 670 and from 50 to
95 mg/kg, respectively (Table S3). For the BS-1, the SiO2,
Al2O3, and Na2O contents were 67.2, 4.9, and 2.5 mass%,
respectively. These content and other chemical composi-
tions in the BS-1 (Tables S2 and S3) are similar to the
those in Site 7 core from the beach, except with respect
to the Cr contents (110 mg/kg).
DISCUSSION
Geochemical ternary diagrams
For the ternary diagram (Al2O3-(Na2O+CaO)-K2O;
Nesbitt and Young, 1984; Figs. 2A and 2B), the data from
the modern beach sands (BS-1), shoreface, and dune de-
posits (7-S1-S5, Site 7) plotted close to middle between
the Al2O3 and Na2O+CaO sides. In contrast, the data for
the mud layers (Sites 1, 2, and 6) plotted closer to the
Al2O3 side. For the paleotsunami deposits (Sites 1, 2, 3,
4, and 6), the data mostly plotted in the intermediate area
between those from Site 7 core and mud layers (Figs. 2A).
In addition, most of the data from the flood deposits over-
lapped with those of the mud layers in the cores, except
for 6-S13 and 6-S16 (Fig. 2A). These two values of the
flood deposits plotted between the data of the T1 (~3500
cal BP) and T2 paleotsunami deposits (~4000 cal BP; Fig.
2B). Therefore, the ternary plot of Al2O3-(Na2O+CaO)-
K2O for the core samples from the Oya lowland in
Fig. 2. (A) Ternary diagram of Al2O3-(Na2O+CaO)-K2O for the deposits from the cores (Sites 1, 2, 3, 4, 6, and 7) and modern
beach sands (BS-1) in the Oya lowland, Shizuoka plain on the Pacific coast of middle Japan. (B) Expanded figure of the ternary
diagram (A) at close to the upper and lower boundary of the data from the paleotsunami deposits in the Oya lowland. (C) Ternary
diagram of (Na2O+CaO)-Al2O3-(Cr+Ni) for the deposits from the cores and modern backshore in the Oya lowland. The Na2O,
CaO, Al2O3, Cr, and Ni contents were normalized by average values for each element across all samples.
Mud layers: peaty clays and muddy parallel lamina layers
CIA: Chemical Alteration Index
Geochemical characteristics of paleotsunami deposits from the Shizuoka 331
Shizuoka separated the T1 paleotsunami deposits with
ages of ~3500 cal BP from the flood deposits and mud
layers (Fig. 2A).
Geochemical data have also been obtained from mod-
ern riverine and marine surface sediments from the Pa-
cific coast of the middle Japan including the Suruga Bay
(Imai et al., 2010; https://gbank.gsj.jp/geochemmap/
index.htm). In the Abe, Fuji, Ohi Rivers, and other small
rivers in the Shizuoka (Fig. 1), the average values of Al2O3
and Na2O contents were 13.1 ± 3.1 and 1.9 ± 0.4 mass%
(n = 48), respectively (Imai et al., 2010). For the marine
sediments (Suruga Bay; Fig. 1), both were 9.4 ± 2.2 and
3.1 ± 0.7 mass% (n = 217), respectively. The Na2O con-
tents in the marine sediments in Suruga Bay were clearly
higher than those from the rivers in Shizuoka. Conversely,
the CaO contents were widely dispersed from 0.2 to 9.2
mass% in the rivers, as well as from 0.3 to 38 mass% in
the marine sediments (Imai et al., 2010), because Ca dis-
tributions in the deposits could mainly depend on the
amount of shell fragments.
In a previous study on the 2011 Tohoku tsunami in
the north Sendai plain, the Na contents in the fresh tsu-
nami deposits had decreased and soluble elements had
leached from the deposits approximately 3 months after
the tsunami’s inundation (Chagué-Goff et al., 2012c).
Additionally, the leachable elements had migrated down-
ward into the soil layers and had been lost from the
depositional area in part of the north Sendai plain, Tohoku
(Chagué-Goff et al., 2012c). For the Tohoku samples, the
data plotted widely from the top (Al2O3) to the left-hand
side (Na2O+CaO) on the ternary diagram (Watanabe et
al., 2020). In addition, Watanabe et al. (2020) suggested
that some minerals containing Na (such as plagioclase
and micas) may possibly be preserved in the tsunami de-
posits, although a certain fraction of the soluble elements
had leached from the paleotsunami deposits.
In the paleotsunami deposits, the Na potentially origi-
nated from sea salts, beach, and marine sediment. In ad-
dition, the majority of Ca in the deposits is derived from
shell fragments, carbonate microfossils, and detrital car-
bonates. In contrast, the Al and K distributions are af-
fected by clay minerals contents such as phyllosilicates
in the deposits (Cuven et al., 2013; Font et al., 2013;
Watanabe et al., 2020). In this study on the Shizuoka plain,
the tsunami deposits could lose soluble elements that
originated from seawater after deposition. The Na and
other soluble elements (such as Cl and S) could have
leached from the original tsunami deposits; thus, the
chemical composition would have changed to that of the
rock materials. In addition, the Na contents in Site 7 in-
dicated a slight decrease from 60 cm to 930 cm depth in
the core (from 2.5 to 2.2 mass%; Supplementary Table
S2). The relatively low contents of Na and Ca in the sam-
ples in this study for the Shizuoka (Table S2) might also
indicate a mixture of sea/beach sand and clay minerals
from terrestrial soils, added during the tsunami inunda-
tion and/or after deposition. Additionally, the values of
the event deposits following tsunami inundation were
affected by leaching from rain and/or ground water flow.
In another ternary diagram for (Na2O+CaO)-Al2O3-
(Cr+Ni) (Fig. 2C), our data from the beach (BS-1 and 7-
S1-S5) from the Oya lowland in Shizuoka plotted between
the top (Na2O+CaO) and the right side (Cr+Ni). In par-
ticular, relatively high concentrations of Cr in 7-S5 (673
mg/kg; 60 cm depth) and 7-S2 (610 mg/kg; 652 cm depth)
from Site 7 core (beach) were observed, and these two
data plotted on the rightmost side (Fig. 2C). Addition-
ally, Ni-rich layers were found in the deeper portions of
Site 7 core (up to 95 mg/kg; Table S3). The paleotsunami
deposits from the Oya lowland (the T0, T1, and T2) plot-
ted between the beach core samples and mud layers in
the cores. The data of the flood deposits overlapped with
those of the mud layers in the cores from the Oya low-
land (Fig. 2C). The Cr concentration of the 4-T0
paleotsunami deposits (~1000 cal BP) was relatively high
(~200 mg/kg) compared to those of other paleotsunami
deposits from the Oya lowland, and the chemical compo-
sition was similar to that of the beach core sample (7-S1;
Fig. 2C, Table S3). The characteristics of Cr distribution
from the samples in the Oya lowland might be caused by
source differences in each sample. A possible source of
the Cr-rich samples is the hinterland of the Abe River on
Shizuoka (Fig. 1; Sugiyama and Shimokawa, 1990; Ohta
et al., 2005).
Characterization by titanium normalized values
Titanium normalized values of each element and Si/
Al atomic ratios in the cores from the Oya lowland,
Shizuoka plain, are shown in Figs. 3–5 and Supplemen-
tary Figs. S3 and S4. For Site 1 core (inland side; Figs. 1
and 3), the vertical distribution in the Na/Ti atomic ratios
exhibited good agreement with those in the Si/Al through-
out the core from the Oya lowland. In the mud layers in
Site 1 core, the Na/Ti and Si/Al atomic ratios were con-
stant, and their values rapidly increased (up to 12.3 and
7.8 atomic ratios, respectively) in the T1 paleotsunami
deposit layer at a depth of 455 cm (~3500 cal BP). Al-
though mud contents decreased in the layers with the T1
and flood deposits (Figs. 3D), peaks of Na/Ti and Si/Al
were observed only in the T1 (Figs. 3A and 3B) for Site 1
core. In the Mg/Ti and Ca/Ti atomic ratios, these values
showed a slight increase in the T1, however, the ratios
had fluctuation throughout the core and increases were
also found in the bottom mud layers of the core (Figs. 3C
and 3E, ~500–600 cm depth).
In Site 2 core (Suruga Bay side; Figs. 1D and 4), in-
creases in the Na/Ti and Si/Al atomic ratios were also
observed in the T1 paleotsunami deposit layers (355 cm
332 T. Watanabe et al.
Fig. 3. Depth profiles of the Ti normalized values for major and trace elements in the paleotsunami deposits, flood deposits, and
mud layers of Site 1 core sample from the Oya lowland, Shizuoka plain on the Pacific coast of middle Japan. The T1 paleotsunami
deposits (~1000 cal BP; 349 cm depth; 1-T1) was found in the layers with a depth of approximately 340–350 cm depth. The layer
with the Kikai-Akahoya tephra (~7300 cal BP; Machida and Arai, 2003) was also found in Site 1 core samples (Kitamura et al.,
2013). The lithology of the core is also shown to the right of the profiles (Kitamura et al., 2013). The yellow bars denote the layers
with the paleotsunami deposits. The green bars denote the layers with flood deposits.
Geochemical characteristics of paleotsunami deposits from the Shizuoka 333
Fig. 4. Depth profiles of the Ti normalized values for major and trace elements in the paleotsunami deposits, flood deposits, and
mud layers of Site 2 core sample from the Oya lowland, Shizuoka plain on the Pacific coast of middle Japan. The T1 (~3500 cal
BP; 355 cm depth; 2-T1) and T2 paleotsunami deposits (~4000 cal BP; 455 cm depth; 2-T2) have been found in the core (Kitamura
et al., 2013). The yellow bars denote the layers with the paleotsunami deposits. The green bars denote the layers with the flood
deposits.
334 T. Watanabe et al.
depth), up to 14.6 and 7.3 atomic ratios, respectively.
These fluctuations in the Na/Ti and Si/Al atomic ratios
corresponded to the stratigraphy of the cores. The flood
deposits in the upper part of Site 2 had intermediate val-
ues of Na/Ti and Si/Al in between the paleotsunami de-
posits and mud layers in the core. For the T2 paleotsunami
deposits at ~4000 cal BP (455 cm depth), although the
peak of the Na/Ti ratio was relatively small compared to
that of the T1 (Fig. 4B), it could be distinguished from
that of the mud layers. In fact, Na/Ti and Si/Al atomic
ratios were clearly increased from 8.5 to 12.3 and from
4.1 to 6.3 atomic ratios, respectively, in the T2 (458 to
455 cm depth in Site 2). The fluctuations of Sr/Ti and Ba/
Ti atomic ratios throughout the core were similar to those
of Na/Ti. Same as with the results in Site 1 core (Fig. 3),
the titanium normalized values (Na/Ti, Sr/Ti, and Ba/Ti)
and Si/Al could also be useful for determining the signa-
tures of paleotsunami deposits in the cores from the Oya
lowland in Shizuoka plain. As shown in Fig. 5A, data for
Na/Ti and Si/Al plotted linearly (r = 0.97, p < 0.01). Ad-
ditionally, the atomic ratios (Si/Al and Na/Al) have nega-
tive correlations with the mud contents in the core
(Kitamura et al., 2013; Fig. 5). The Na/Al atomic ratios
were up to ~0.85 in the layers with low mud contents
(Fig. 5). Therefore, the sand layers in the cores and beach
sands could include plagioclase (rich with albite;
NaAlSi3O8). Previous studies for pleotsunami deposits
(Cuven et al., 2010, 2013; Chagué-Goff et al., 2017;
Moreira et al., 2017) have also revealed strong relation-
ships between grain size and geochemical data, notably
Si (mainly affected by quartz sand and coarse silt con-
tents) and calcium carbonates. Heavy minerals such as
zircon (ZrSiO4) potentially derived via high energy tsu-
nami inflow have been also observed within coarser
grained deposits in other areas (Moreira et al., 2017).
Moreover, Cl, Br, and I have preserved in a part of finer
deposits after marine incursions (Chagué-Goff et al.,
2017; Moreira et al., 2017). Therefore, source-depended
geochemical characteristics of paleotsunami deposits
should be cleared on wider coastal areas, and multi prox-
ies including geochemical and other signatures should be
used to estimate the past marine incursions.
Peaks of S/Ti, Mn/Ti, and Fe/Ti were observed in these
mud layers in Site 2 (356–435 cm and 458–660 cm depth).
These results indicate that sulfide minerals, such as py-
rite, could accumulated in the mud layers below the re-
dox boundaries. These elements might migrate easily with
redox conditions after burial (Nara et al., 2010, 2014).
The Cr/Ti atomic ratios exhibited a peak in the T2
paleotsunami deposits (455 cm depth). In addition, the
Mg/Ti also generated a peak in the layer with T2 up to
11.6 atomic ratio. Relatively high contents of Cr in the
marine sediments have been reported in Suruga Bay, off
the west end of Shizuoka, and around Lake Hamana (Fig.
Fig. 5. (A) Relationship between Na/Ti and Si/Al, (B) Sr/Ti and Si/Al, (C) Si/Al and mud contents, and (D) Na/Al and mud
contents from the cores and modern beach sands from the Oya lowland, Shizuoka plain on the Pacific coast of middle Japan.
Geochemical characteristics of paleotsunami deposits from the Shizuoka 335
1; Ohta et al., 2005; Imai et al., 2010); these high values
have also been found in the upper reaches of the Abe
River, possibly caused by the distribution of ultramafic
rocks around the area (Paleogene Setogawa belt, close to
the west end of the Fossa Magna and the Itoigawa-
Shizuoka Tectonic line in the area; Ohta et al., 2005).
For our study in Shizuoka, the average values of Na/
Ti and Si/Al atomic ratios from the cores (Sites 1, 2, 3, 4,
6, and 7) are listed in Table 1. The average Na/Ti ratio of
the T1 tsunami deposit layers was 15.9 ± 0.9, which was
significantly higher than that of the mud layers (9.2 ±
0.9, Table 1). The beach core samples from Site 7 had a
relatively high Na/Ti and Si/Al atomic ratios (17.6 ± 1.4
and 9.3 ± 1.0, respectively), and the values of the T1
paleotsunami deposits were coincident with those from
Site 7 within the 1 sigma error (Table 1). In addition, the
T1 tsunami deposit layers were distinguished from the
flood deposits by their Na/Ti atomic ratios (an average
Na/Ti ratio of 11.6 ± 1.5).
The horizontal distributions of the Na/Ti and Si/Al
atomic ratios in the paleotsunami deposits in Shizuoka
(Sites 1–4 and 6–7) are shown in Fig. S4. In Site 7 core
from the beach on Suruga Bay coast, the Na/Ti and Si/Al
atomic ratios maintained relatively high values through-
out the core (16.0–19.8 and 8.2–10.5 atomic ratios). In
addition, these ratios of the T1 paleotsunami deposits from
Fig. 6. Results of principal component analysis (PCA) for the geochemical data from the cores and modern beach sands in the
Oya lowland, Shizuoka plain. The compositional data were processed by centered logratio transformation with the zero replace-
ment to avoid constant-sum constraint based on Aitchison (1982), Kucera and Malmgren (1998), Arai and Ohta (2006), and Ohta
and Arai (2006). Filled circles denote the PCA results from the paleotsunami deposits from the cores. Open triangles and dia-
monds are the data from the beach core samples and modern beach sands, respectively. Open squares represent the data from the
flood deposits. Lastly, filled squares are the data from mud layers in the cores. The arrowed lines accompanied by element
symbols indicate the eigenvector values in both biplots ((A) PC2 vs. PC1; (B) PC2 vs. PC3).
Sites 1, 2, 3, and 6 (~3500 cal BP) were shown to be rela-
tively high up to 16.5 and 8.4, respectively, from the
Suruga Bay side (Site 3) to the inland area (Site 6). Also,
these values in T1 converged with approximately 6% on
relative standard deviation (RSD) for the both ratios (Fig.
S3 and Table 1). These horizontal distributions (Fig. S4)
revealed that the geochemical data were useful signatures
for the characterization and discrimination of the T1
paleotsunami deposits from other layers on the Shizuoka
plain in our case.
Geochemical discrimination by statistical analysis
The nonparametric correlation coefficients (r values;
Spearman’s rank correlation calculated using the R pro-
gram) among the processed geochemical data (Si-Pb; 24
elements) in Sites 1–7 cores and the BS-1 from the Oya
lowland on Shizuoka plain are shown in Supplementary
Table S4. The r values indicated that Si showed positive
correlations, which were statistically different from zero
at a 1% significance level (p < 0.01, Spearman’s rank
correlation tests), for the Na, Mg, Ca, Cr, Sr, Zr, and Ba
(Table S4). Notably, relatively high r value (r = 0.94) was
observed between Na and Si. Additionally, Na contents
had a negative correlation with mud contents (r = –0.50,
p < 0.05). These results also revealed that minerals con-
taining Na (such as albite) could be preserved in the sand
336 T. Watanabe et al.
rich layers. Additionally, positive correlations were ob-
served between Al and Ti (Table S4), which may have
originated from detrital materials.
For the statistical analysis of processed geochemical
data, PCA results are shown in Figs. 3, 4, 6, and Supple-
mentary Table S5. Cumulative proportions of the 1st, 2nd,
and 3rd principal components (PC1-3) were approxi-
mately 73.7% of the variation (Table S5). In this study,
PC1 accounted for 37.8% of the variation, and the PC1
values possibly represented redox conditions in the cores,
because of the positive PC1 eigenvector in S, Mn, Fe,
and As (Fig. 6A, PC1-2 biplot). As mentioned previously,
concentrated sulfide could occur at the redox boundary
below the tsunami deposits (Figs. S1 Q-T and S2 Q-T).
Therefore, the positive PC1 values indicated the accu-
mulation of these elements under anoxic conditions.
The negative PC2 values likely represented the
paleotsunami deposits because of the negative eigenvector
in Si, Na, Mg, Sr, and Ca (Fig. 6A, PC1-2 biplot). Cr also
exhibited negative values in the PC2 eigenvectors because
the Cr content was relatively high in the paleotsunami
deposits (T0 and T2) and the beach samples (Site 7 and
BS1). All our data from the paleotsunami deposits (T0,
T1, and T2) plotted in the area with negative PC2 values,
and the data from the sand layers plotted in the area with
negative values (Fig. 6). Additionally, all data from beach
core samples also plotted in the area with negative PC2
values. In addition, the PC2 eigenvector of Al, Ti, and
heavy metals (Cu, Zn, Pb) exhibited a positive direction.
Thus, the PC2 values might be correlated with grain size
(sandy or muddy deposits). Based on these PCA charac-
teristics, the PC2 values might help to indicate the
paleotsunami deposits from the others. The depth pro-
files of the PC1 and PC2 scores also represented these
characteristics (PC1, redox condition; PC2, grain size and
tsunami inundation; Figs. 3 N-P and 4 N-P). In addition,
the depth profiles of the PC3 score resemble in those of
heavy metal contents (Ni, Cu, Zn, and Pb; Figs. S1U-X
and S2U-X) in the cores from Sites 1 and 2.
The result of a cluster analysis using the geochemical
data in the cores (Sites 1, 2, 3, 4, 6, and 7) and BS-1 is
shown in Fig. 7. The dendrogram showed the two major
cluster; the first one included the mud layers, except for
1-S6, 6-S14, and 6-S15, and second one consists of the
beach samples (Site 7 and BS1), paleotsunami deposits
(T0, T1, and T2), and most of the flood deposits. Based
on the statistical analysis, the data from the T1
paleotsunami deposits (~3500 cal BP) were clearly dis-
criminated from other flood deposits, beach core sam-
ples, and mud layers in this study. The T0 (~1000 cal BP)
was separated from the flood deposits and other mud lay-
ers in the cores (Fig. 7). Conversely, the T2 paleotsunami
deposits (~4000 cal BP) could not be distinguished from
the flood deposits (6-S16 and 6-S13) in this case. In this
study, we have concluded that the titanium normalized
values and statistical analyses of geochemical data are
useful methods to confirm the T1 paleotsunami deposits
in continuous cores from the Oya lowland on Shizuoka
plain, Pacific coast of middle Japan.
Fig. 7. Cluster dendrogram using Euclidean distance and
Ward’s method for the geochemical data in the cores (Sites 1,
2, 3, 4, 6, and 7) and modern beach sands (BS-1) from the Oya
lowland, Shizuoka plain on the Pacific coast of middle Japan.
Geochemical characteristics of paleotsunami deposits from the Shizuoka 337
CONCLUSIONS
In this study, paleotsunami deposits (~1000, 3500, and
4000 cal BP) from the Oya lowland were analyzed using
EDXRF to survey possible geochemical proxies of ma-
rine incursion in the Shizuoka plain on the Pacific coast
of middle Japan. The geochemical ternary diagrams indi-
cated that the data can help for considering the character-
istics of the paleotsunami deposits from the Shizuoka
plain. The titanium normalized values (Na/Ti, Sr/Ti, and
Ba/Ti) and Si/Al atomic ratios in the paleotsunami de-
posits are higher than those in other layers from the cores.
In particular, the Na/Ti and Si/Al atomic ratios of the T1
paleotsunami deposits (~3500 cal BP) in the Oya low-
land were clearly high up to an atomic ratio of 16.5 and
8.4, respectively, compared with those of flood deposits
and mud layers in the cores. Additionally, we observed
increases of Cr/Ti on the T0 and T2 paleotsunami depos-
its (~1000 and ~4000 cal BP, respectively). The PCA and
cluster analyses of the geochemical data were used for
the characterization and discrimination of the T1
paleotsunami deposits from other layers in our case. Our
method had been previously adopted in another area and
applied to tsunami deposits from the Pacific coast of
northeast Japan (Tohoku area; Watanabe et al., 2020).
Therefore, the geochemical proxies of paleotsunami de-
posits in our study may be able to contribute to the iden-
tification of the event deposits on the both coastal areas.
Acknowledgments—We thank the members of the Shizuoka
University and Tohoku University for the preservation, divi-
sion and preparation of the boring cores and other geological
samples. Y. Asahara and two anonymous reviewers are acknowl-
edged for their helpful comments and suggestions for improv-
ing the manuscript. This work was partly supported by a Grant-
in-Aid from JST/RISTEX, Japan, to N.T. (FY2011, 2012-2015)
and a Grant-in-Aid for Scientific Research to T.W. (17K06989)
from JSPS, Japan.
REFERENCES
Aitchison, J. (1982) The statistical analysis of compositional
data (with discussion). Jour. Roy. Statist. Soc. B 44, 139
177; doi:org/10.1111/j.2517-6161.1982.tb01195.x.
Arai, H. and Ohta, T. (2006) Note on zero and missing values
in compositional data. Jour. Geol. Soc. Japan 112, 439
451; doi:org/10.5575/geosoc.112.439 (in Japanese with
English abstract).
Baba, T., Cummins, P. R., Hori, T. and Kaneda, Y. (2006) High
precision slip distribution of the 1944 Tonankai earthquake
inferred from tsunami waveforms: Possible slip on a splay
fault. Tectonophysics 426, 119134; doi:10.1016/
j.tecto.2006.02.015.
Chagué-Goff, C. (2010) Chemical signatures of palaeotsunamis:
A forgotten proxy? Mar. Geol. 271, 6771; doi:org/10.1016/
j.margeo.2010.01.010.
Chagué-Goff, C., Andrew, A., Szczucinski, W., Goff, J. and
Nishimura, Y. (2012a) Geochemical signatures up to the
maximum inundation of the 2011 Tohoku-oki tsunami-Im-
plications for the 869 AD Jogan and other palaeotsunamis.
Sediment. Geol. 282, 6577; doi:org/10.1016/
j.sedgeo.2012.05.021.
Chagué-Goff, C., Goff, J., Nichol, S. L., Dudley, W., Zawadzki,
A., Bennett, J. W., Mooney, S. D., Fierro, D., Heijnis, H.,
Dominey-Howes, D. and Courtney, C. (2012b) Multi-proxy
evidence for trans-Pacific tsunamis in the Hawai’ian Islands.
Mar. Geol. 299
302, 7789; doi:org/10.1016/
j.margeo.2011.12.010.
Chagué-Goff, C., Niedzielski, P., Wong, H. K. Y., Szczucinski,
W., Sugawara, D. and Goff, J. (2012c) Environmental im-
pact assessment of the 2011 Tohoku-oki tsunami on the
Sendai Plain. Sediment. Geol. 282, 175187; doi:org/
10.1016/j.sedgeo.2012.06.002.
Chagué-Goff, C., Wong, H. K. Y., Sugawara, D., Goff, J.,
Nishimura, Y., Beer, J., Szczuci´nski, W. and Goto, K. (2014)
Impact of tsunami inundation on soil salinisation—up to
one year after the 2011 Tohoku-oki tsunami. Tsunami Events
and Lessons Learned: Environmental and Societal Signifi-
cance (Kontar, Y., Santiago-Fandiño, V. and Takahashi, T.,
eds.), 193214, Springer, Dordrecht.
Chagué-Goff, C., Goff, J., Wong, H. K. Y. and Cisternas, M.
(2015) Insights from geochemistry and diatoms to charac-
terise a tsunami’s deposit and maximum inundation limit.
Mar. Geol. 359, 2234; doi:org/10.1016/
j.margeo.2014.11.009.
Chagué-Goff, C., Szczuci´nski, W. and Shinozaki, T. (2017)
Applications of geochemistry in tsunami research: a review.
Earth Sci. Rev. 165, 203244; doi:org/10.1016/
j.earscirev.2016.12.003.
Chagué-Goff, C., Sugawara, D., Goto, K., Goff, J., Dudley, W.
and Gadd, P. (2018) Geological evidence and sediment trans-
port modelling for the 1946 and 1960 tsunamis in
Shinmachi, Hilo, Hawaii. Sediment. Geol. 364, 319333;
doi:org/10.1016/j.sedgeo.2017.09.010.
Cummins, P. R., Baba, T., Kodaira, S. and Kaneda, Y. (2002)
The 1946 Nankai earthquake and segmentation of the
Nankai Trough. Phys. Earth Planet. Inter. 132, 7587;
doi:org/10.1016/S0031-9201(02)00045-6.
Cuven, S., Francus, P. and Lamoureux, S. F. (2010) Estimation
of grain size variability with micro X-ray fluorescence in
laminated lacustrine sediments, Cape Bounty, Canadian
High Arctic. J. Paleolimnol. 44, 803817; doi:org/10.1007/
s10933-010-9453-1.
Cuven, S., Paris, R., Falvard, S., Miot-Noirault, E., Benbakkar,
M., Schneider, J. and Billy, I. (2013) High-resolution analy-
sis of a tsunami deposit: Case-study from the 1755 Lisbon
tsunami in southwestern Spain. Mar. Geol. 337, 98111;
doi:org/10.1016/j.margeo.2013.02.002.
Fagel, N. and Mackay, A. W. (2008) Weathering in the Lake
Baikal watershed during the Kazantsevo (Eemian) intergla-
cial: Evidence from the lacustrine clay record. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 259, 244257; doi:10.1016/
j.palaeo.2007.10.011.
Font, E., Veiga-Pires, C., Pozo, M., Nave, S., Costas, S., Munõz,
F. R., Abad, M., Simões, N., Duarte, S. and Rodríguez-Vidal,
338 T. Watanabe et al.
J. (2013) Benchmarks and sediment source(s) of the 1755
Lisbon tsunami deposit at Boca do Rio Estuary. Mar. Geol.
343, 114; doi:org/10.1016/j.margeo.2013.06.008.
Fujiwara, O., Ono, E., Yata, T., Umitsu, M., Sato, Y. and
Heyvaert, V. (2013) Assessing the impact of 1498 Meio
earthquake and tsunami along the Enshu-nada coast, cen-
tral Japan using coastal geology. Quat. Int. 308
309, 412;
doi:org/10.1016/j.quaint.2012.12.009.
Fujiwara, O., Aoshima, A., Irizuki, T., Ono, E., Obrochta, S.,
Sampei, Y., Sato, Y. and Takahashi, A. (2020) Tsunami de-
posits refine great earthquake rupture extent and recurrence
over the past 1300 years along the Nankai and Tokai fault
segments of the Nankai Trough, Japan. Quat. Sci. Rev. 227,
105999 (https://doi:org/10.1016/j.quascirev.2019.105999).
Goff, J., McFadgen, B. G. and Chagué-Goff, C. (2004) Sedi-
mentary differences between the 2002 Easter storm and the
15th-century Okoropunga tsunami, southeastern North Is-
land, New Zealand. Mar. Geol. 204, 235250; doi:org/
10.1016/S0025-3227(03)00352-9.
Goff, J., Pearce, S., Nichol, S., Chagué-Goff, C., Horrocks, M.
and Strotz, L. (2010) Multi-proxy records of regionally-
sourced tsunamis, New Zealand. Geomorphology 118, 369
382.
Goff, J., Witter, R., Terry, J. and Spiske, M. (2020)
Paleotsunamis in the Sino-Pacific region. Earth Sci. Rev.
210, 103352; doi:org/10.1016/j.earscirev.2020.103352.
Goto, K., Fujino, S., Sugawara, D. and Nishimura, Y. (2014)
The current situation of tsunami geology under new poli-
cies for disaster countermeasures in Japan. Episodes 37,
258264; doi:org/10.18814/epiiugs/2014/v37i4/005.
Hori, T., Kato, N., Hirahara, K., Baba, T. and Kaneda, Y. (2004)
A numerical simulation of earthquake cycles along the
Nankai Trough in southwest Japan: lateral variation in fric-
tional property due to the slab geometry controls the nu-
cleation position. Earth Planet. Sci. Lett. 228, 215226;
doi:10.1016/j.epsl.2004.09.033.
Imai, N., Terashima, S., Ohta, A., Mikoshiba, M., Okai, T.,
Tachibana, Y., Ikehara, K., Katayama, H., Noda, A., Togashi,
S., Matsuhisa, Y., Kanai, Y. and Kamioka, H. (2010)
Geochemical map of sea and land of Japan. Geological Sur-
vey of Japan, AIST, Tsukuba, 207 pp. (in Japanese).( https:/
/gbank.gsj.jp/geochemmap/index.htm).
Kanaori, Y., Kawakami, S. and Yairi, K. (1993) Space-time
correlations between inland earthquakes in central Japan
and great offshore earthquakes along the Nankai trough:
Implication for destructive earthquake prediction. Eng.
Geol. 33, 289303; doi:org/10.1016/0013-7952(93)90031-
7.
Kitamura, A. (2016) Examination of the largest-possible tsu-
namis (Level 2 tsunami) generated along the Nankai and
Suruga troughs during the past 4000 years based on studies
of tsunami deposits from the 2011 Tohoku-oki tsunami.
Prog. Earth Planet. Sci. 3, 12; doi:10.1186/s40645-016-
0092-7.
Kitamura, A. and Kobayashi, K. (2014) Geologic evidence for
prehistoric tsunamis and coseismic uplift during the AD
1854 Ansei-Tokai earthquake in Holocene sediments on the
Shimizu Plain, central Japan. The Holocene 24, 814827;
doi:10.1177/0959683614530447.
Kitamura, A., Fujiwara, O., Shinohara, K., Akaike, S., Masuda,
T., Ogura, K., Urano, Y., Kobayashi, K., Tamaki, C. and
Mori, H. (2013) Identifying possible tsunami deposits on
the Shizuoka Plain, Japan and their correlation with earth-
quake activity over the past 4000 years. The Holocene 23,
16841698; doi:10.1177/0959683613505345.
Kitamura, A., Ohashi, Y., Ishibashi, H., Miyairi, Y., Yokoyama,
Y., Ikuta, R., Ito, Y., Ikeda, M. and Shimano, T. (2016)
Holocene geohazard events on the southern Izu Peninsula,
central Japan. Quat. Int. 397, 541554; doi:org/10.1016/
j.quaint.2015.04.021.
Kitamura, A., Yamada, K., Sugawara, D., Yokoyama, Y.,
Miyairi, Y. and Hamatome team (2020) Tsunamis and sub-
marine landslides in Suruga Bay, central Japan, caused by
Nankai-Suruga Trough megathrust earthquakes during the
last 5000 years. Quat. Sci. Rev. 245, 106527; doi:org/
10.1016/j.quascirev.2020.106527.
Kortekaas, S. and Dawson, A. G. (2007) Distinguishing tsu-
nami and storm deposits: An example from Martinhal, SW
Portugal. Sediment. Geol. 200, 208221; doi:org/10.1016/
j.sedgeo.2007.01.004.
Kucera, M. and Malmgren, B. A. (1998) Logratio transforma-
tion of compositional data—a resolution of the constant sum
constraint. Mar. Micropaleontol. 34, 117120; doi:org/
10.1016/S0377-8398(97)00047-9.
Kuwatani, T., Nagata, K., Okada, M., Watanabe, T., Ogawa, Y.,
Komai, T. and Tsuchiya, N. (2014) Machine-learning tech-
niques for geochemical discrimination of 2011 Tohoku tsu-
nami deposits. Sci. Rep. 4, 7077; doi:10.1038/srep07077.
Machida, H. and Arai, F. (2003) Atlas of Tephra in and around
Japan (revised edition). The University of Tokyo Press,
Tokyo, 336 pp. (in Japanese).
Maechler, M., Rousseeuw, P., Struyf, A., Hubert, M. and Hornik,
K. (2018) Cluster: Cluster Analysis Basics and Extensions.
R package version 2.0.7-1.
Matsunami, H., Matsuda, K., Yamasaki, S., Kimura, K., Ogawa,
Y., Miura, Y., Yamaji, I. and Tsuchiya, N. (2010) Rapid si-
multaneous multi-element determination of soils and envi-
ronmental samples with polarizing energy dispersive X-ray
fluorescence (EDXRF) spectrometry using pressed powder
pellets. Soil Sci. Plant Nutr. 56, 530540; doi:10.1111/
j.1747-0765.2010.00489.x.
Minoura, K. and Nakaya, S. (1991) Traces of tsunami preserved
in inter-tidal lacustrine and marsh deposits: some examples
from northeastern Japan. J. Geol. 99, 265287; doi:org/
10.1086/629488.
Minoura, K., Nakaya, S. and Uchida, M. (1994) Tsunami de-
posits in a lacustrine sequence of the Sanriku coast, north-
east Japan. Sediment. Geol. 89, 2531; doi:org/10.1016/
0037-0738(94)90081-7.
Moore, A., Nishimura, Y., Gelfenbaum, G., Kamataki, T. and
Triyono, R. (2006) Sedimentary deposits of the 26 Decem-
ber 2004 tsunami on the northwest coast of Aceh, Indone-
sia. Earth Planets Space 58, 253258; doi:org/10.1186/
BF03353385.
Moreira, S., Costa, P. J. M., Andrade, C., Lira, C. P., Freitas,
M. C., Oliveira, M. A. and Reichart, G. J. (2017) High reso-
lution geochemical and grain-size analysis of the AD 1755
tsunami deposit: Insights into the inland extent and inunda-
Geochemical characteristics of paleotsunami deposits from the Shizuoka 339
tion phases. Mar. Geol. 390, 94105; doi:org/10.1016/
j.margeo.2017.04.007.
Nakamura, Y., Nishimura, Y. and Putra, P. S. (2012) Local vari-
ation of inundation, sedimentary characteristics, and min-
eral assemblages of the 2011 Tohoku-oki tsunami on the
Misawa coast, Aomori, Japan. Sediment. Geol. 282, 216
227; doi:org/10.1016/j.sedgeo.2012.06.003.
Nanayama, F., Satake, K., Furukawa, R., Shimokawa, K.,
Atwater, B. F., Shigeno, K. and Yamaki, S. (2003) Unusu-
ally large earthquakes inferred from tsunami deposits along
the Kuril trench. Nature 424, 660663; doi:org/10.1038/
nature01864.
Nara, F. W., Watanabe, T., Kakegawa, T., Seyama, H., Horiuchi,
K., Nakamura, T., Imai, A., Kawasaki, N. and Kawai, T.
(2010) Climate control of sulfate influx to Lake Hovsgol,
northwest Mongolia, during the last glacial-postglacial tran-
sition: Constraints from sulfur geochemistry. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 298, 278285; doi:10.1016/
j.palaeo.2010.10.001.
Nara, F. W., Watanabe, T., Kakegawa, T., Yamasaki, S., Inoue,
C. and Tsuchiya, N. (2014) Sources of tsunami deposits on
the Pacific coasts of Iwate, Miyagi, and Fukushima areas
by the 2011 off the Pacific coast of Tohoku Earthquake,
northeast Japan inferred from geochemical signatures (TOC/
TN ratio and stable sulfur isotope). J. Geog. 123, 871882;
doi:org/10.5026/jgeography.123.871 (in Japanese with Eng-
lish abstract).
Nara, F. W., Yokoyama, T., Yamasaki, S., Minami, M., Asahara,
Y., Watanabe, T., Yamada, K., Tsuchiya, N. and Yasuda, Y.
(2021) Characteristics in trace elements compositions of
tephras (B-Tm and To-a) for identification tools. Geochem.
J. 55, 117133; doi:10.2343/geochemj.2.0619.
Nelson, A. R., Sawai, Y., Jennings, A. E., Bradley, L.-A.,
Gerson, L., Sherrod, B. L., Sabean, J. and Horton, B. P.
(2008) Great-earthquake paleogeodesy and tsunamis of the
past 2000 years at Alsea Bay, central Oregon coast, USA.
Quat. Sci. Rev. 27, 747768; doi:org/10.1016/
j.quascirev.2008.01.001.
Nesbitt, H. W. and Young, G. M. (1984) Prediction of some
weathering trends of plutonic and volcanic rocks based on
thermodynamic and kinetic considerations. Geochim.
Cosmochim. Acta 48, 15231534; doi:org/10.1016/0016-
7037(84)90408-3.
NIST (1993a) Certificate of Analysis, Standard reference ma-
terial 2709 (https://www-s.nist.gov/srmors/certificates/ar-
chive/2709.pdf).
NIST (1993b) Certificate of Analysis, Standard reference ma-
terial 2710 (https://www-s.nist.gov/srmors/certificates/ar-
chive/2710.pdf).
NIST (2000) Report of Investigation, Reference material 8704
(https://www-s.nist.gov/srmors/certificates/8704.pdf).
NIST (2002) Certificate of Analysis, Standard reference mate-
rial 2711 (https://www-s.nist.gov/srmors/certificates/ar-
chive/2711.pdf).
Niwa, M., Kamataki, T., Kurosawa, H., Saito-Kokubu, Y. and
Ikuta, M. (2020) Seismic subsidence near the source region
of the 1662 Kanbun Hyuganada Sea earthquake:
Geochemical, stratigraphical, chronological, and
paleontological evidences in Miyazaki Plain, southwest Ja-
pan. Island Arc 29, e12341; doi:org/10.1111/iar.12341.
Ohta, T. and Arai, H. (2006) Problems in compositional data
analysis and their solutions. Jour. Geol. Soc. Japan 112 ,
173187; doi:org/10.5575/geosoc.112.173 (in Japanese with
English abstract).
Ohta, A., Imai, N., Terashima, S. and Tachibana, Y. (2005) Ap-
plication of multi-element statistical analysis for regional
geochemical mapping in Central Japan. Appl. Geochem. 20,
10171037; doi:org/10.1016/j.apgeochem.2004.12.005.
Okamura, M. and Matsuoka, H. (2012) Recurrences of Nanaki
Earthquake estimated from paleo tsunami sediment. Kagaku
82, 182-191 (in Japanese) (http://sc1.cc.kochi-u.ac.jp/
~mako-ok/pdf/Kagaku_2012.pdf).
R Core Team (2018) R: A language and environment for statis-
tical computing. R Foundation for Statistical Computing,
Vienna, Austria (https://www.R-project.org/).
Roubault, M., de la Roche, H. and Govindaraju, K. (1970)
Present status (1970) of the cooperative studies on the
geochemical standards of the “Centre de Recherches
Pétrographique et Géochimiques”. Sci. Terre 15, 354393.
Roy, P. D., Jonathan, M. P., Macias, M. C., Sánchez, J. L.,
Lozano, R. and Srinivasalu, S. (2012) Geological charac-
teristics of 2011 Japan tsunami sediments deposited along
the coast of southwestern Mexico. Geochemistry 72, 91
95; doi:org/10.1016/j.chemer.2012.02.003.
Rydelek, P. A. and Sacks, S. (2003) Triggering and inhibition
of great Japanese earthquakes: the effect of Nobi 1891 on
Tonankai 1944, Nankaido 1946 and Tokai. Earth Planet.
Sci. Lett. 206, 289296; doi:10.1016/S0012-
821X(02)01095-6.
Satake, K. (2014) Advances in earthquake and tsunami sciences
and disaster risk reduction since the 2004 Indian ocean tsu-
nami. Geosci. Lett. 1, 15; doi:10.1186/s40562-014-0015-7.
Satake, K. and Atwater, B. F. (2007) Long-term perspectives
on giant earthquakes and tsunamis at subduction zones.
Annu. Rev. Earth Planet. Sci. 35, 349374; doi:org/10.1146/
annurev.earth.35.031306.140302.
Sawai, Y., Kamataki, T., Shishikura, M., Nasu, H., Okamura,
Y., Satakae, K., Thomson, K. H., Matsumoto, D., Fujii, Y.,
Komatsubara, J. and Aung, T. T. (2009) Aperiodic recur-
rence of geologically recorded tsunamis during the past 5500
years in eastern Hokkaido, Japan. J. Geophys. Res. 11 4,
B01319 (https://doi:org/10.1029/2007JB005503).
Sawai, Y., Namegaya, Y., Okamura, Y., Satake, K. and
Shishikura, M. (2012) Challenges of anticipating the 2011
Tohoku earthquake and tsunami using coastal geology.
Geophys. Res. Lett. 39, L21309; doi:org/10.1029/
2012GL053692.
Sawai, Y., Namegaya, Y., Tamura, T., Nakashima, R. and
Tanigawa, K. (2015) Shorter intervals between great earth-
quakes near Sendai: Scour ponds and a sand layer attribut-
able to AD 1454 overwash. Geophys. Res. Lett. 42, 4795
4800; doi:10.1002/2015GL064167.
Shinozaki, T., Goto, K., Fujino, S., Sugawara, D. and Chiba, T.
(2015) Erosion of a paleo-tsunami record by the 2011
Tohoku-oki tsunami along the southern Sendai Plain. Mar.
Geol. 369, 127136; doi:org/10.1016/j.margeo.2015.08.009.
Shinozaki, T., Sawai, Y., Hara, J., Ikehara, M., Matsumoto, D.
and Tanigawa, K. (2016) Geochemical characteristics of
340 T. Watanabe et al.
deposits from the 2011 Tohoku-oki tsunami at Hasunuma,
Kujukuri coastal plain, Japan. Island Arc 25, 350368;
doi:org/10.1111/iar.12159.
Sugiyama, Y. and Shimokawa, K. (1990) Geology of the
Shimizu district. With geological Sheet Map at 1:50,000.
Geological Survey of Japan, 103 pp. (in Japanese with Eng-
lish abstract) (https://www.gsj.jp/data/50KGM/PDF/
GSJ_MAP_G050_08089_1990_D.pdf).
Szczuci´nski, W., Kokocinski, M., Rzeszewski, M., Chagué-
Goff, C., Cachão, M., Goto, K. and Sugawara, D. (2012)
Sediment sources and sedimentation processes of 2011
Tohoku-oki tsunami deposits on the Sendai Plain, Japan-
Insights from diatoms, nannoliths and grain size distribu-
tion. Sediment. Geol. 282, 4056; doi:org/10.1016/
j.sedgeo.2012.07.019.
Watanabe, T., Tsuchiya, N., Yamasaki, S., Sawai, Y., Hosoda,
N., Nara, F. W., Nakamura, T. and Komai, T. (2020) A
geochemical approach for identifying marine incursions:
implications for tsunami geology on the Pacific coast of
northeast Japan. Appl. Geochem. 118, 104644; doi:org/
10.1016/j.apgeochem.2020.104644.
Watanabe, T., Ishii, C., Ishizaka, C., Niwa, M., Shimada, K.,
Sawai, Y., Tsuchiya, N., Matsunaka, T., Ochiai., S. and Nara
F. W. (2021) Quantitative and semi-quantitative analyses
using a portable energy dispersive X-ray fluorescence
spectrometer: Geochemical applications in fault rocks, lake
sediments, and event deposits. J. Mineral. Petrol. Sci. 116,
140–158; doi:10.2465/jmps.201224.
Yamamoto, S., Obana, K., Takahashi, T., Nakanishi, A., Kodaira,
S. and Kaneda, Y. (2013) Imaging of the subducted Kyushu-
Palau Ridge in the Hyuga-nada region, western Nankai
Trough subduction zone. Tectonophysics 589, 90102;
doi:org/10.1016/j.tecto.2012.12.028.
Yamasaki, S., Matsunami, H., Takeda, A., Kimura, K., Yamaji,
I., Ogawa, Y. and Tsuchiya, N. (2011) Simultaneous deter-
mination of trace elements in soils and sediments by polar-
izing energy dispersive X-ray fluorescence spectrometry.
BUNSEKI KAGAKU 60, 315323; doi:org/10.2116/
bunsekikagaku.60.315 (in Japanese with English abstract).
Yamasaki, S., Takeda, A., Watanabe, T., Tagami, K., Uchida,
S., Takata, H., Maejima, Y., Kihou, N., Matsunami, H. and
Tsuchiya, N. (2015) Bromine and iodine in Japanese soils
determined with polarizing energy dispersive X-ray fluo-
rescence spectrometry. Soil Sci. Plant Nutr. 61, 751760;
doi:org/10.1080/00380768.2015.1054773.
Yamasaki, S, Takeda, A., Kimura, K. and Tsuchiya, N. (2016)
Underestimation of chromium and zirconium in soils by
hydrofluoric acid digestion and inductively coupled plasma-
mass spectrometry. Soil Sci. Plant Nutr. 62, 121126 (http:/
/dx.doi:org/10.1080/00380768.2016.1149437).
Yoshioka, S. and Hashimoto, M. (1989) The stress field induced
from the occurrence of the 1944 Tonankai and the 1946
Nankaido earthquakes, and their relation to impending earth-
quakes. Phys. Earth Planet. Inter. 56, 349370; doi:org/
10.1016/0031-9201(89)90169-6.
SUPPLEMENTARY MATERIALS
URL (http://www.terrapub.co.jp/journals/GJ/archives/
data/55/MS641.pdf)
Figures S1 to S5
Tables S1 to S5
... Ti-normalized values have been applied to characterize and estimate the source of tsunami deposits in previous studies (Cuven et al., 2013;Chagué-Goff et al., 2017;Watanabe et al., 2020Watanabe et al., , 2021aWatanabe et al., , 2021b. In the present study, an obvious increase in S/Ti atomic ratio from 3.1 to 9.4 was observed at the boundary between the SF1 and SF2a layers (Fig. 11A). ...
... In Shizuoka on the Pacific coast of middle Japan, increases in Zr/Ti atomic ratios within the tsunami deposits (~4000 cal BP; pre historical tsunami) were also observed accompanying Cr/Ti peaks ( Fig. 14H; Supplementary Tables S15; Watanabe et al., 2020). However, the Zr/Ti peaks were occurred on the flood deposits on the upper part of the core (angular sands; depth, ~200-300 cm; Watanabe et al., 2021b). Therefore, multiple proxies from geochemical and geological studies are needed for tsunami discrimination. ...
... In this area, multiple proxies, such as Na/Ti, Sr/Ti, Cr/Ti, and Zr/Ti atomic ratios, can support the discrimination of tsunami deposits. Fe/Ti, S/Ti, and As/Ti atomic ratios increased just below the tsunami deposits, likely because of sulfide accumulation with redox changes after the marine incursion (Watanabe et al., 2021b). ...
Article
Earthquakes and tsunami repeatedly occur in the coastal areas of the Japanese islands, which are located on the margin of continental plates. Paleotsunami deposits in these areas provide important clues regarding the reconstruction of past marine incursions by earthquakes and tsunami. The Miyazaki plain on the Pacific coast of southwest Japan along the Nankai–Suruga trough has been also affected by marine incursions by previous interplate earthquakes and tsunami. Tsunami deposits are usually identified by their geological, sedimentological, and geochemical characteristics. In the present work, portable energy dispersive X-ray fluorescence spectrometry (portable XRF), X-ray analytical microscopy (XGT), statistical analyses, and heavy mineral analyses were performed to characterize the paleotsunami deposits of the Kanbun Hyuganada earthquake (age, 1662 common era) in the SY03 core from the south Miyazaki plain. Our new findings can be summarized as follows: (1) Increases in the Ti-normalized values of S, Ca, Fe, and Sr, and depletions of K were observed in the deposits; (2) Accumulation of Zr and heavy minerals such as zircon, likely related with the high-energy tsunami, were found in the deposits; (3) Cluster analyses of the geochemical data acquired from the SY03 core by the portable XRF distinguished the paleotsunami deposits on the south Miyazaki plain. Furthermore, we also compiled the geochemical data of samples from the Shizuoka (central Japan), Sendai (Pacific coast of northeast Japan), and Miyazaki plains. Increases in Ti-normalized values (e.g., Sr/Ti) were found in these paleotsunami deposits. Our studies indicate that geochemical approaches are suitable for the characterization and exploration of paleotsunami deposits on the several coastal areas of Japan.
Article
Full-text available
Quantitative and semi–quantitative procedures using a portable energy dispersive X–ray fluorescence spectrometer (portable XRF) were applied to geochemical studies in fault rocks, lake sediments, and soils, including paleotsunami deposits. The results obtained are as follows: 1) Correlation coefficients (r2; measured values by portable XRF versus reported values) for working curves of major elements obtained using powdered reference standard materials were 0.70–0.99 with p < 0.01, except for Mg, and r2 values for the trace elements were 0.72–0.99 with p < 0.05, except for Cl, Ba, and U. 2) In the fault rocks from Central Japan (e.g., Shiraki–Nyu fault, Atera faults; powdered and nonpowdered rock fragment samples), measured values of K, Ca, Ti, Mn, Fe, Rb, Sr, Zr, and Th contents obtained using portable XRF are consistent with the reported values (r2 = 0.47–0.90, with p < 0.01). Clear Fe and Mn enrichments and K and Rb depletions were observed within the gouges of the Shiraki–Nyu fault from Central Japan. 3) The depth profiles of chemical components in lake sediments (powdered) from the Noto Peninsula in Central Japan measured using portable XRF methods were similar to those measured using the stationary–type XRF (r2 = 0.61 in Ti and r2 = 0.40 in Zr contents, with p < 0.05). 4) For paleotsunami deposits on the Pacific coast of northeast Japan, the chemical components and titanium–normalized values (Si/Ti, Ca/Ti, Sr/Ti, and As/Ti) in powdered samples measured using portable XRF correlated well with those in a previous study measured using stationary–type XRF (r2 = 0.84–0.98, with p < 0.01). Cluster analyses of geochemical data conducted using portable XRF were useful to characterize event deposits on the north Sendai Plain. In each kind of sample, the element concentration values (SiO2, TiO2, Fe2O3, MnO, CaO, K2O, S, V, Cr, Cu, Zn, As, Rb, Sr, Zr, Pb, and Th) obtained via quantitative analyses using portable XRF were consistent with the reported values obtained via other methods, such as stationary–type XRF and inductively coupled plasma–mass spectrometry.
Article
Full-text available
The absolute date of the Millennium Eruption (ME) of Changbaishan Volcano is widely recognized as AD 946. The Baegdusan-Tomakomai (B-Tm) tephra dispersed during the ME is a robust-age key bed. In order to identify the tephra, refractive index and major-element compositions of volcanic glass shards are conventionally used. Meanwhile, trace- element analysis using the inductively coupled plasma mass spectrometry (ICP-MS) technique for the volcanic glass shards and bulk samples of the B-Tm tephra is still limited. Here we give a detailed description of major- and trace- element compositions for the glass shards and bulk sediments of the B-Tm and Towada caldera eruptions (To-a) tephra deposits from the Lake Ogawara sediment core, Tohoku region, northern Japan to identify the tephras. The depth profiles of the major and trace elements show the significant peaks for the K2O and some trace elements (Zn, Rb, Zr, Nb, Sn, Y, La, Ce, Nd, Th, and U) at the B-Tm tephra layer in the Lake Ogawara sediment core, but no peaks of these elements at the To- a tephra layer. High concentrations of the trace elements in the B-Tm tephra layer were observed in individual glass shards as well as in the bulk sediment. These concentrations are highlighted by the elemental abundance pattern normalized by the crustal abundance. The elemental pattern in individual glass shards from other Japanese tephras showed significant differences from those of the B-Tm tephra, especially in REE compositions. The trace-element compositions of the glass shards and bulk sediments show a strong tool for distinguishing the B-Tm tephra from other Japanese tephras.
Article
Full-text available
Japanese historical documents reveal that Mw 8 class earthquakes have occurred every 100–150 years along the Suruga and Nankai troughs since the 684 Hakuho earthquake. These earthquakes have commonly caused large tsunamis with wave heights of up to 10 m in the Japanese coastal area along the Suruga and Nankai troughs. From the perspective of tsunami disaster management, these tsunamis are designated as Level 1 tsunamis and are the basis for the design of coastal protection facilities. A Mw 9.0 earthquake (the 2011 Tohoku-oki earthquake) and a mega-tsunami with wave heights of 10–40 m struck the Pacific coast of the northeastern Japanese mainland on 11 March 2011, and far exceeded pre-disaster predictions of wave height. Based on the lessons learned from the 2011 Tohoku-oki earthquake, the Japanese Government predicted the tsunami heights of the largest-possible tsunami (termed a Level 2 tsunami) that could be generated in the Suruga and Nankai troughs. The difference in wave heights between Level 1 and Level 2 tsunamis exceeds 20 m in some areas, including the southern Izu Peninsula. This study reviews the distribution of prehistorical tsunami deposits and tsunami boulders during the past 4000 years, based on previous studies in the coastal area of Shizuoka Prefecture, Japan. The results show that a tsunami deposit dated at 3400–3300 cal BP can be traced between the Shimizu, Shizuoka and Rokken-gawa lowlands, whereas no geologic evidence related to the corresponding tsunami (the Rokken-gawa–Oya tsunami) was found on the southern Izu Peninsula. Thus, the Rokken-gawa–Oya tsunami is not classified as a Level 2 tsunami.
Article
Full-text available
Faunal compositions and 14C ages of emerged sessile assemblages at four sites in the southern part of Izu Peninsula, central Japan, indicate that co-seismic uplift occurred at 1256–950 BC, AD 1000–1270, AD 1430–1660, and AD 1506–1815. The data suggest that the stress field in the southern part of Izu Peninsula changed to its current north–south compression at ca. 3100 BP, and that the recurrence intervals for uplift-inducing earthquakes have become shorter during the last 1000 years. The main faults responsible for this seismicity appear to be located offshore from the south part of Izu Peninsula. This study also examined a 1-cm-thick scoria layer deposited between 6940 and 6810 cal BP in the south part of Izu Peninsula. The layer, which had not been previously examined petrographically, consists of scoria grains derived from Izu-Oshima volcano located 40 km east of the study area, rather than from volcanoes located on Izu Peninsula itself.
Article
Study of prehistoric to medieval-age tsunami deposits along a riverbank site near the eastern Nankai Trough, central Japan, show that, not only did Tokai earthquakes occur with a higher frequency than previously thought, but that contemporaneous ruptures of the Tokai and Nankai fault segments were also more common. The site revealed a ∼1-km long coast-normal cross section of the strand plain and exposed four sandy tsunami deposits, each of which indicates inundation over 2 km inland of the coast. Radiocarbon dating of previously studied and newly discovered deposits in the region indicates a shorter recurrence time for Tokai earthquakes and clarifies their linkage with Nankai earthquakes. We attribute the younger two tsunami deposits to the 1498 and 1096 CE Tokai earthquakes. The older two deposits confirm the occurrence of the Tokai earthquakes in 887 CE and in the latest 7th century. These events are not reliably recorded in historical documents in the Tokai region but were noted in the Nankai area. The 887 CE earthquake likely represents a full-length rupture of the Tokai and Nankai segments, as was the case for the 1707 CE earthquake. Integrated with the previous studies, these new results show that nine Tokai earthquakes occurred over the last 1300 years, the oldest in the latest 7th century, and in 887, 1096, 1361, 1498, 1614, 1707, 1854 and 1944 CE. Recalculated recurrence intervals range from 90 to 265 years. Except for the 1498 Meio Tokai earthquake, the Tokai earthquakes occurred simultaneously with Nankai earthquakes.
Article
In the study of palaeotsunamis it is crucial to decipher the sedimentological record, to derive intensity of past events and to infer different inundation phases. To achieve this goal, it is important to apply high-resolution techniques that allow magnifying intra-deposit details (at a sub-centimetric scale) that otherwise would not be perceived; consequently, valuable information could be overlooked. In this work, we applied successfully high-resolution geochemical and grain-size analyses – XRF core-scanning and image analysis, respectively – to the AD 1755 tsunami deposit. This quartz sand enriched in bioclast deposit (exhibiting high Si/Al and Ca/Ti) was recognized in the coastal stratigraphic sequence of Salgados lagoon due to its contrasting composition when compared with the under and overlying mud layers with scarce bioclasts (exhibiting low Si/Al and Ca/Ti). In the absence of textural evidence, the identification of peaking concentrations of Cl, S and Br (all major constituents of sea salt) in a continuous muddy sequence allowed slightly extending farther inland the limit of inundation. In addition, grain-size analysis data attested the fining inland of the deposit. Furthermore, despite the macroscopic massive structure of the tsunami deposit, throughout the lagoon, grain-size results revealed more complexity and allowed inferring up to four depositional sequences directly associated with the AD 1755 tsunami inundation. https://authors.elsevier.com/a/1VF6Z5mVdWeyy
Article
Much progress has been made since the first published studies of tsunami deposits nearly 30 years ago. Geochemistry is now a much more widely used proxy in tsunami research, mainly due to its increasingly recognised value in the identification of historical and/or prehistorical deposits, at times even providing the conclusive proof when other proxies are missing or equivocal, but also its significance in environmental impact assessments following recent tsunamis. The rapid advance in analytical techniques has also made it a more approachable and popular method, as it is now most often faster and cheaper. Here we provide a review of the applications of geochemistry, including the techniques used, as well as a database of studies that used chemical proxies in their investigation of recent and old events, including onshore and offshore tsunami deposits. Chemical signatures are often used as markers of marine inundation, either as salinity indicators, where they can also allow the identification of the limit of tsunami inundation, or tracers of the incorporation of marine-sourced carbonates. Their applications as indicators of source material are nevertheless expanding, thereby potentially providing additional information on the hydrodynamic processes associated with tsunami inundation, although they are largely site-specific. The effects of post-depositional processes in different climatic regimes are examined, with a particular emphasis on water-leachable components and implications for post-event recovery of coastal ecosystems. We demonstrate the usefulness of chemical proxies in studies of the geological record of tsunamis extending back thousands of years, suggest new approaches and discuss limitations and existing knowledge gaps.
Article
A sparsely documented tsunami in 1454 may subdivide the recurrence interval between the 869 and 2011 tsunamis near Sendai, as judged from geomorphic, stratigraphic, and archival evidence. Pond-filled breaches cut across beach ridges on century-old topographic maps. The basal pond deposit in one of these breaches postdates 1454. Stratigraphy on Sendai Plain includes a sand sheet that contains marine and brackish diatoms. Radiocarbon ages suggest that the sheet dates to 1406–1615 (2σ), and written records for this interval in Tohoku mention a tsunami in 1454. The inferred inundation extended 1.0–2.5 km inland from an approximate medieval shoreline. Simulated tsunamis that best account for the sand sheet require a thrust earthquake of moment magnitude 8.4 or larger. If the sand sheet represents the 1454 tsunami, the two most recent intervals between great thrust earthquakes in Sendai region spanned 585 and 557 years.