NHESSNatural Hazards and Earth System SciencesNHESSNat. Hazards Earth Syst. Sci.1684-9981Copernicus PublicationsGöttingen, Germany10.5194/nhess-16-2799-2016Typhoon Haiyan's sedimentary record in coastal environments of the
Philippines and its palaeotempestological implicationsBrillDominikbrilld@uni-koeln.dehttps://orcid.org/0000-0001-8637-4641MaySimon Matthiashttps://orcid.org/0000-0001-6762-7500EngelMaxhttps://orcid.org/0000-0002-2271-4229ReyesMichellePintAnnaOpitzStephanDierickManuelGonzaloLia AnneEsserSaschaBrücknerHelmutInstitute of Geography, Universität zu Köln, Cologne, GermanyMarine Science Institute, University of the Philippines, Quezon City, PhilippinesDepartment of Physics and Astronomy, Ghent University, Ghent, BelgiumNationwide Operational Assessment of Hazards (Project NOAH), Department of Science and
Technology, Quezon City, PhilippinesDominik Brill (brilld@uni-koeln.de)21December201616122799282221June20164July201621October201614November2016This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://nhess.copernicus.org/articles/16/2799/2016/nhess-16-2799-2016.htmlThe full text article is available as a PDF file from https://nhess.copernicus.org/articles/16/2799/2016/nhess-16-2799-2016.pdf
On 8 November 2013, category 5 Supertyphoon Haiyan made landfall on the
Philippines. During a post-typhoon survey in February 2014, Haiyan-related
sand deposition and morphological changes were documented at four severely
affected sites with different exposure to the typhoon track and different
geological and geomorphological settings. Onshore sand sheets reaching
100–250 m inland are restricted to coastal areas with significant
inundation due to amplification of surge levels in embayments or due to
accompanying long-wave phenomena at the most exposed coastlines of Leyte and
Samar. However, localized washover fans with a storm-typical laminated
stratigraphy occurred even along coasts with limited inundation due to waves
overtopping or breaching coastal barriers. On a recent reef platform off
Negros in the Visayan Sea, storm waves entrained coral rubble from the reef
slope and formed an intertidal coral ridge several hundreds of metres long
when breaking at the reef edge. As these sediments and landforms were
generated by one of the strongest storms ever recorded, they not only provide
a recent reference for typhoon signatures that can be used for
palaeotempestological and palaeotsunami studies in the region but might also
increase the general spectrum of possible cyclone deposits. Although a rather
atypical example for storm deposition due to the influence of infra-gravity
waves, it nevertheless provides a valuable reference for an extreme case that
should be considered when discriminating between storm and tsunami deposits
in general. Even for sites with low topography and high inundation levels
during Supertyphoon Haiyan, the landward extent of the documented sand sheets
seems significantly smaller than typical sand sheets of large tsunamis. This
criterion may potentially be used to distinguish both types of events.
Introduction
On 8 November 2013, Typhoon Haiyan (local name: Yolanda) made
landfall on the Philippines reaching category 5 on the Saffir–Simpson
hurricane scale. By crossing the archipelago Haiyan caused more than
6000 casualties, affected more than 16 million people, and damaged more than
1 million houses (NDRRMC, 2014; Lagmay et al., 2015). Its destructive power
resulted from exceptional surface winds reaching sustained velocities of up
to 315 kmh-1 (1 min averaged data), gusting even up to
380 kmh-1, in combination with massive storm surge flooding with
water levels up to 9 m above tide level (IRIDeS, 2014). Based on
recorded wind speeds and core pressure, Haiyan was not only an exceptional
event for the Philippines but one of the most powerful tropical cyclones ever
recorded. Against the background of the ongoing controversial discussion on
the influence of climate change on cyclone frequencies and magnitudes
(Knutson et al., 2010; Pun et al., 2013), Typhoon Haiyan could be both an
exceptional low-frequency event and/or a precursor of a new normality.
The need for robust data that provide information about long-term typhoon
risk in affected areas is highlighted by the lack of awareness and
preparedness of many inhabitants regarding Haiyan (Engel et al., 2014; Lagmay
et al., 2015). Similar to other recent coastal disasters, such as the 2004
Indian Ocean Tsunami (Brückner and Brill, 2009) or Cyclone Nargis in 2008
(Fritz et al., 2009), studies of occurrence patterns and effects of past
flooding events with exceptional magnitude have been lacking in the Central
Philippines. Although previous catastrophic typhoons with similar tracks have
been historically documented, e.g. in 1897 or 1984 (PAGASA, 2014; Soria et
al., 2016), their disastrous effects have not been taken into account
properly. This was at least partly due to limited comprehension of the term
“storm surge”, which people mainly associated with the moderate flooding of
typical typhoons in the area (Engel et al., 2014; Mas et al., 2015). This
discrepancy between real risk and perceived risk results from an
underrepresentation of local high category typhoons in instrumental and
historical records. This discrepancy is great, because cyclones usually
follow an inverse power law (Corral et al., 2010).
Geological imprints of cyclones may be used to enhance existing historical
and instrumental records, as they potentially cover periods of several
millennia and, thus, document even large events in statistically significant
numbers (Hippensteel et al., 2013; May et al., 2013, 2015a). However, using
geological archives to extend tropical cyclone histories requires the
identification of tropical cyclone traces, reliable dating of event layers,
as well as a careful consideration of potential changes in palaeogeography
and/or sea level. In particular the discrimination of similar depositional
events such as tsunamis is challenging (Shanmugam, 2012). In this regard,
modern event deposits offer the possibility to establish (locally valid)
event-specific criteria that indicate prehistoric events in the geological
record (Nanayama et al., 2000; Kortekaas and Dawson, 2007). Although all
potential discrimination criteria have been observed for both tsunami and
storm deposits on a global perspective (Shanmugam, 2012), local comparisons
assign several decimetre thick laminated deposits with inland extents of a
few tens to hundreds of metres that often show foreset bedding and tend to thin and fine landwards to typical
storm signatures, while tsunami deposits tend to be thinner, are composed of
a few layers with massive or normally graded structure, and may extend inland
for several kilometres (Tuttle, 2004; Morton et al., 2007; Switzer and Jones,
2008; Goto et al., 2011). Likewise, modern analogues may help to evaluate
reliable age (Bishop et al., 2005; Brill et al., 2012) and magnitude (Brill
et al., 2014) reconstructions of palaeoevents. Since sedimentary signatures
of cyclones are influenced by local factors (Schwartz, 1982; Matias et
al., 2008), Haiyan's geological footprint will primarily be useful to
establish criteria for inferring past typhoons from geological records found
in the same areas. However, as one of the strongest storms ever recorded,
Haiyan might increase the spectrum of possible cyclone features and, by this,
add to the discussion on discriminating between tsunamis and extreme cyclones
in general.
Here, we report on sandy sediment data from four different Haiyan-affected
coastal areas, collected during a post-typhoon field survey on the
Philippines in February 2014, i.e. 3 months after Haiyan made landfall. The
investigated areas comprise different geologies with carbonate and
volcaniclastic coasts, as well as different geomorphological settings
including steep cliff platforms, low coastal plains, and coral reefs. The
major aims of this study are to (i) document Haiyan's impact on the
sedimentology and geomorphology of heavily affected coastal areas by
recording onshore and intertidal sedimentation, coastal erosion, and
geomorphological changes. Based on these data, sedimentary and
geomorphological typhoon signatures typical for the study area shall be
established. In addition, (ii) the spatial variability of these typhoon
signatures due to site-specific characteristics such as the local topography,
bathymetry, geology, and hydrodynamics as well as the exposure to the typhoon
track shall be investigated. Finally, (iii) the potential of these modern
typhoon deposits will be evaluated in respect of possible implications for
the identification of prehistoric tropical cyclones in the geological record.
Study area
The Philippine archipelago is located in the western Pacific Ocean between 5
and 20∘ northern latitude. Field work was carried out in four
different areas (Fig. 1): (i) at the east coast of Samar between Llorente and
General MacArthur (Fig. 1b); (ii) at the northeastern coast of Leyte between
Tacloban and Dulag (Fig. 1b); (iii) on several islands northeast of Northern
Negros (Sagay); and (iv) on Bantayan (Fig. 1c). All four study areas were
significantly affected by Typhoon Haiyan but are characterized by a
different exposure to the typhoon track and by particular geomorphological
and geological settings.
Overview of the study area. (a) Philippine archipelago with
main tectonic structures (Rangin et al., 1989) and track of Typhoon Haiyan
(NDRRMC 2014); (b) position of research areas on Samar and
Leyte as well as (c) northern Negros and Bantayan
(based on ESRI base maps). S: Samar; L: Leyte; C: Cebu; N: Negros; Pa: Panay;
P: Palawan. (d) Historical cyclone tracks crossing the Philippines
since the beginning of weather recording (NOAA Historical Hurricane Track
Pool) illustrate a decreasing cyclone frequency from north to south.
(e) Mean significant wave heights (Navy METOC, 2014) indicate
highest wave energy during winter monsoon along the exposed east coasts and
usually calm conditions in the interior parts of the archipelago.
Climate and typhoon activityGeneral climatic and hydrodynamic conditions
The Philippine islands are characterized by a subtropical climate influenced
by monsoon winds and the ENSO system. While the west of the archipelago
experiences seasonal variations in rainfall and temperature with a dry
period from November to April and a wet season during summer monsoon,
seasonal variability gradually decreases towards the east reaching
all-season wet conditions along the eastern coast (PAGASA, 2011). Hence, the
study areas on Leyte and Samar show no pronounced annual rainfall
variations. Northern Negros and Bantayan are influenced by weak monsoon
seasonality. The highest swell waves occur during the winter monsoon and
along the Pacific coast, while wave action during the summer monsoon and
within the Visayan Sea is generally moderate (Fig. 1e, Navy METOC, 2014).
Tidal variations are in the range of 0.8–1.8 m. In addition, the entire
Philippine territory is located in the corridor of east-west moving
typhoons. On average, 21 tropical cyclones hit the Philippines annually,
whereas the occurrence probability increases markedly towards the north
(Fig. 1d, PAGASA, 2014).
Haiyan's path over the Philippines
Starting as a tropical depression on 3 November 2013 over the northwestern
Pacific, Haiyan continuously gained intensity, turning into a tropical storm
on 4 November and a typhoon on 5 November. On 6 November Haiyan reached
category 5 on the Saffir–Simpson hurricane scale and finally made landfall
on Samar on 8 November (IRIDeS, 2014) (Fig. 1a). After its first landfall
near Guiuan (Eastern Samar) at 4:40 PHT (Philippine time, UTC+8), Haiyan
crossed the archipelago in a western direction without reducing its strength
below category 5 (Fig. 1a). On its path over the Philippines, Haiyan made
landfall on northern Leyte at 7:00, on northern Cebu at 10:00, on Bantayan at
10:40, and later on Panay and Palawan (NDRRMC, 2014).
Depending on the exposure to the typhoon track, wind speeds, storm surge
levels, and wave heights varied significantly between the study sites.
Additional variations are determined by differences in local bathymetry,
fetch, and shape of the coastline (see Sect. 2.2). In general, coastal
flooding rapidly reached peak levels that lasted for approximately 2 h and
was characterized by inflowing waves with periods of several seconds (Mas et
al., 2015; Morgerman, 2014). The resulting flooding levels at the affected
coastlines could mostly be reconstructed by means of combined storm surge and
phase-averaged wave models (Bricker et al., 2014; Mori et al., 2014; Cuadra
et al., 2014) (details concerning the specifications of each model presented
in this paper are provided in the respective references). However, around
Tacloban, significant amplification of storm surge levels to values of up to
8 m were measured (Mas et al., 2015), and three distinct flooding
pulses were observed by eyewitnesses (May et al., 2015b). Furthermore, along
the coast of Eastern Samar surprisingly high values of run-up and flow depth
were documented during post-typhoon surveys (Tajima et al., 2014; May et
al., 2015b). While the exceptional water levels in the semi-enclosed basin of
the San Pedro Bay can be explained by the funnel-shaped topography and
seiches using combined storm surge and phase-averaged wave models (Mori et
al., 2014; Bricker et al., 2014; Soria et al., 2016), the inundation pattern
along the open coast of Eastern Samar can be attributed to the impact of
infra-gravity waves due to an interaction of wind waves with the coral reef,
if phase-resolving wave models are applied (Roeber and Bricker, 2015; Kennedy
et al., 2016).
The exposed coast of Eastern Samar is characterized by a large fetch and a
steep offshore bathymetry. Hence, it experienced maximum wind speeds with the
highest model-predicted storm waves of up to 20 m, but only a limited
wind-driven surge (Bricker et al., 2014). Field evidence documents run-up of
approximately 12 m above mean sea level (msl) and up to 800 m
inundation distance (PAGASA, 2014; Tajima et al., 2014; May et al., 2015b).
Also the northeastern coast of Leyte was still exposed to the full strength
of the storm winds, since the landmasses to the east are too narrow to
significantly reduce Haiyan's intensity. However, due to the shallow water
and the resonance effects in the enclosed embayment (Mori et al., 2014),
storm-water levels were dominated by a surge set-up, while model-predicted
wave heights were < 5 m (Bricker et al., 2014). Field evidence
shows that the storm surge reached run-up levels of nearly
8 m a.s.l. and inundation of several hundred metres inland (Mas et
al., 2015; Tajima et al., 2014). However, smaller surge levels and
moderate wave heights were modelled for Northern Negros and Bantayan (Bricker
et al., 2014), due to a shallow bathymetry, low tide at the time of landfall,
and the sheltering landmasses of Samar, Leyte, and Cebu to the east. Likewise,
eyewitness accounts document maximum onshore flooding levels of only
∼ 2 m a.s.l. (Cuadra et al., 2014).
Geology and geomorphology
The Philippine islands are formed by a complex geological structure of
north–south running volcanic arcs, tectonic basins, and fragments of
continental crust reflecting subduction and collision processes between the
Eurasian and the Philippine plates (Rangin et al., 1989). As the main
tectonic structures, the Philippine and Manila trenches confine offshore
subduction zones to the east and west of the archipelago, while the
Philippine Fault crosses the archipelago in an axial position from north to
south (Fig. 1a, Rangin et al., 1989). The associated volcanoes and ophiolites
form islands dominated by steep, cliff-lined coasts that are bounded by
fringing reefs and are occasionally intersected by flat alluvial lowlands and
pocket beaches. The clastic beaches are characterized by either white sand of
coral reef origin or darker and denser minerals derived from volcanic
provinces (Bird, 2010).
The rocky carbonate coast of Eastern Samar is characterized by steep
headlands with cliffs formed of Pleistocene coral limestone and occasional
pocket beaches bordered by fringing reefs (HER, Fig. 1b). Since the fetch
over the Pacific is not restricted, Eastern Samar is exposed to high swell
waves especially during the winter monsoon (Fig. 1e) (Bird, 2010). In contrast, the northeastern coast of Leyte is dominated by alluvial plains,
sandy beaches, and beach-ridge plains (TOL, Fig. 1b, Dimalanta et al., 2006),
while rocky promontories and fringing coral reefs are scarce. Water depths in
the shallow, funnel-shaped San Pedro Bay between Leyte and Samar do not
exceed 30 m (Bird, 2010). In contrast to this, the northeastern coast
of Negros is characterized by wide coastal plains bordered by mangroves and
extensive coral reefs (Rangin et al., 1989). Within the shallow water of the
Visayan Sea to the north (water depths mainly < 50 m) numerous
coral islands occur (CAR and MOL, Fig. 1c). The island of Bantayan represents
the northeastern limit of the coral islands. The raised limestone formations
making up the island are completely surrounded by fringing reefs that partly
border steep rocky coastlines but are also associated with sandy beaches
(BAN A and B in Fig. 1c, Bird, 2010).
Methods
The topography of all studied locations was documented along transects by
means of a Topcon HiPer Pro differential global positioning system (DGPS). To
reconstruct flooding characteristics of Haiyan's storm surge, indicators for
onshore flow depth and run-up height were documented by measuring elevations
of debris lines, grass, or floated debris in trees and bushes, as well as
impact marks in the bark of palm trees relative to the sea level. All
altitudes are given in metres above msl, and – in case of flood marks –
additionally in metres above ground surface. Onshore flow directions were
deduced from oriented grass and trunks using a compass. In addition,
comparison of rectified pre- and post-Haiyan satellite images was used to
estimate inundation areas. Sediments and depositional landforms generated by
Typhoon Haiyan were documented both onshore and in the intertidal zone. This
included sandy deposits as well as coral-rubble ridges. Shore-perpendicular
trenches were used to describe and document the stratigraphy of sandy
deposits in the field. For detailed sedimentary and faunal analyses, sediment
from Haiyan's deposits and reference environments was taken in the form of
bulk samples from each stratigraphical unit at different distances from the
shoreline, as well as push cores from at least one storm deposit with
representative sedimentary structure at each investigated location.
To deduce transport processes, mode of deposition, and source environments of
the storm deposits, we analysed fine-grained samples in terms of their
granulometry, geochemistry, mineralogy, and faunal composition at the
Institute of Geography, University of Cologne. Grain-size analyses were
performed with a Laser Particle Analyser (Beckman Coulter LS 13320) for
material < 2 mm after pre-treatment with H2O2 and
Na4P2O7 to remove organic carbon and to avoid aggregation. In case
of samples with grain size > 2 mm, the granulometry was determined
on dried sediment using a Camsizer (Retsch Technology). It should be noted
that both approaches do not consider differences in particle shape and
density, which may influence the settling velocity of the grains
significantly (Woodruff et al., 2008). This is particularly important for the
interpretation of granulometric variations in storm deposits with a
significant percentage of shells. Statistical parameters (mean, sorting,
skewness) of grain-size distributions were calculated with the GRADISTAT
software (Blott and Pye, 2001) using the formulas of Folk and Ward (1957).
Further information about flow directions and mode of deposition was obtained
by visually interpreting sedimentary structures in µCT scans with
50 µm voxel resolution (using myVGL 2.1 software) performed on
two selected push cores (BAN 4 and TOL 8) at the University of Ghent,
Belgium.
The chemical characterization of deposits includes determination of organic
carbon by means of loss on ignition (LOI) measured after oven-drying at
105 ∘C for 12 h and ignition in a muffle furnace at
450 ∘C for 4 h. Carbonate contents were measured
gas-volumetrically using the Scheibler method. The bulk-mineralogical
composition was determined by X-ray diffractometry (XRD) on powder compounds
performed on a Siemens D5000 using a step interval of 0.05∘ and a
dwell time of 4 s. A fixed 1∘ divergence and anti-scatter
slit was used at diffraction angles from 5 to 75∘ 2theta. The Cu
K-alpha radiation source was operated at 40 keV and 40 mA.
The data were analysed with DiffracPlus Eva software package (Bruker AXS,
Berlin, Germany).
Magnetic susceptibility (MS) was measured with a Geotec MSCL core sampler. To
apply microfauna composition as an indicator for sediment source and mode of
transportation (e.g. Pilarczyk et al., 2016; Quintela et al., 2016), samples
were sieved to isolate fractions > 100 and < 100 µm. At
least 100 foraminifers were identified to species level, if possible, and
counted under a binocular microscope. States of reworking were assessed
semi-quantitatively on the basis of test taphonomy and classified as no, low,
medium, or strong reworking. To select appropriate parameters for
interpreting transport processes and sediment source areas, we carried out
principal component analysis (PCA) using PAST software (Hammer et al., 2001)
after removing foraminifer species with insignificant abundance (i.e.
< five
individuals in all samples) and correlating parameters based on Spearman's
rank correlation coefficient (≥ 0.95).
Results – sedimentary and geomorphological structuresEastern Samar – Hernani (HER)
At Barangay Batang, municipality of Hernani (Fig. 1b), the Pleistocene
reef retreats to a position more than 200 m from the shoreline,
resulting in a gently inclined sandy coast (Fig. 2). For Haiyan, local
residents report complete inundation of the lower coastal plain with water
levels of several metres above msl, while Agaton and Basyang – the two tropical
storms/depressions hitting the Philippines on 19 January and 1 February,
respectively (Fig. S3), which is after Haiyan and before this field survey –
caused no significant flooding (Table S1 in the Supplement).
Starting at the shoreline behind a 550 m wide intertidal reef lagoon and
destroyed mangrove stands, the landward transect crosses a sandy beach ridge
with a crest height of 2.2 m above msl at 50 m from the
shoreline. Landwards, the tree-covered back-barrier depression is crossed by
a shore-parallel channel before shallow reef outcrops occur at 120 m
from the shoreline. After a second depression cultivated with rice fields
(160–220 m), the transect ends at the inactive cliff of the elevated
Pleistocene reef platform at 230 m inland and 2.8 m above msl. While
swash lines document maximum inundation that exceeded the inactive cliff,
floated grass and litter in trees indicate water levels of at least
2.5 m above msl (equal to a flow depth of 1.5 m above
surface) at 100 m, up to 5.0 m above msl (3.7 m flow depth)
at 170 m, and heights of 2.6 m above msl (1.4 m flow depth)
at 220 m (Fig. 2).
Hernani (HER) study site with documented Haiyan flood marks and
positions of sampled typhoon deposits. (a) Local Typhoon Haiyan
inundation limit (based on Google Earth/Digital Globe 11 November 2013).
(b) Onshore sediments were investigated along a coast-perpendicular
transect (A–A′) crossing the flooded area. (c) Destroyed
mangroves in front of the beach (view from the shore-parallel main road;
photography: February 2014). (d) Topographical cross section
(A–A′) with sampling sites. With water levels of at least 5 m
above msl, flooding during Haiyan overtopped the sandy coastal barrier,
destroyed the road on top of the coastal barrier, and transported sand nearly
250 m inland to the foot of the former cliff.
Erosion was the dominant process seaward of the beach ridge, which is marked
by an erosive scarp. Behind the barrier, sedimentation of sand and coral
rubble was documented (HER 3–10), reaching its landward limit at the foot of
the inactive cliff (Fig. 2). A sharp contact separates the light brown sandy
layer covering the dark brown, densely rooted pre-Haiyan soil (Fig. 4b, c).
The thickness of the deposit decreases rapidly landwards, from 20 cm
directly behind the beach ridge (HER 9) to only 2 cm at 110 m
(HER 7) and a few mm at 210 m (HER 3). Coarse coral clasts at the
surface (up to ∼ 20 cm) occur as far as 100 m inland. While
mean grain size does not show any fining trend in the deposit, modal grain
size decreases from 1.3 cm to 220 µm along the same
section (Figs. 3, 4a).
Landward trends of grain size and thickness in typhoon deposits.
Onshore deposits at sites TOL, HER, and BAN B tend to thin landward. While
the deposits at HER and BAN are also characterized by fining trends of the
mean grain size, a monotonic landward fining is not observed at TOL. Clear
trends in sorting are not detected at all sites. Note the different scales.
Onshore deposits of Typhoon Haiyan at site HER.
(a) Landward transect illustrating the succession of typhoon
deposits (see Fig. 2 for location). The basal unit 1 (U1) continuously thins
landwards, while unit 2 (U2) is restricted to the proximal part of the
coastal plain (HER 9). (b) The 2 cm thick unit 1 at HER 7
(photography: February 2014). (c) The 20 cm thick unit 2 at HER 9
(photography: February 2014). (d) Sedimentary characteristics of
HER 10. A 12 cm thick typhoon layer can clearly be separated from the
palaeosol. The storm layer reveals repeated normally and inversely graded
laminae, in which the amount of strongly reworked foraminifera is highest
within the coarse basal sections.
Tolosa (TOL) study site with documented Haiyan flood marks and
positions of sampled typhoon deposits. (a) Local inundation limit
reached several 100 m inland (based on Google Earth/Digital Globe,
11 November 2013, i.e. immediately after Haiyan). (b) Onshore
sediments were investigated along a transect (A–A′) crossing the inundated
area (based on Google Earth/Digital Globe, 23 February2 2012; insert: Google
Earth, 11 November 2013) (c) Westward view from the beach ridge over
the back-barrier marsh. (d) Flood debris in bushes served as an
indicator for flow depth. (e) With flow levels of at least
4.5 m above msl, Haiyan flooded the beach ridge entirely and
transported sand about 150 m inland.
Typhoon Haiyan onshore deposits at site TOL. (a) Transect
illustrating the succession of typhoon deposits in landward direction. A
graded to massive basal unit (U1) is covered by a laminated unit (U2) in the
proximal part of the coastal plain (TOL 3–8). (b) The 10 cm thick
deposit at TOL 7 consists of massive sand at the base (unit 1) and laminated
sand on top (unit 2). (c) Sedimentary characteristics of TOL 5. The
16 cm thick laminated typhoon layer can clearly be separated from the
palaeosol due to lower LOI and magnetic susceptibility (MS).
µCT scans of sediment cores TOL 8 and BAN 4 (as for
locations see Figs. 5d and 10f respectively). The 3-D data support the
documentation and interpretation of sedimentary features and different units
within the typhoon deposits. The basal unit of TOL 8 is slightly normally
graded and characterized by scour structures behind bent grass stems
(unit 1). The upper part is horizontally laminated (unit 2). In BAN 4 the
deposit of Typhoon Haiyan clearly contrasts the unstructured pre-Haiyan
sediments composed of two palaeosols separated by an older sand sheet. It is
divided into a planar bedded unit 1 at the base and steeply landward-inclined laminae in unit 2.
PCA results for sandy deposits of Typhoon Haiyan. (a) XRD
studies allow for a separation of deposits from carbonate coasts (MOL, HER,
BAN) and the siliciclastic coast (TOL) on the basis of bulk mineralogy. Ca:
calcite; A: aragonite; Qz: quartz; Pl: plagioclase; Am: amphibole.
(b) Grain-size data reflect both site-specific characteristics (TOL,
BAN, HER) and differences in formation (unit 1, unit 2, palaeosol)
(b1). A discrimination of sediment formation (unit 1, unit 2,
palaeosol) due to particular clusters is even more pronounced on the
intra-site level (b2–b4).
Carbin Reef (CAR) study site with typhoon-generated coral-rubble
ridge. The ridge at the western edge of the reef platform shows no clear
trend of crest elevation, but significant differences in width and shape
between six coast-perpendicular transects (T1–T6). In the northern part of
the ridge, flat and up to 30 m wide lobes composed of algae-covered, greyish
coral fragments form the basal unit of the ridge (U1). Unit 1 is topped by
steeper lobes with a significant percentage of freshly broken coral fragments
that occur along the whole ridge section but reach widths of only
10 m (U2). The presence of two morphostratigraphical units might be
due to either changing wind directions during formation entirely by Haiyan or
the impact of successive storms where Haiyan only deposited the uppermost
unit.
Molocaboc study site (MOL) with depositional and geomorphological
effects of Typhoon Haiyan. (a) Location of studied beach section on
Molocaboc. (b) Onshore sediments of Haiyan were investigated in
three trenches (MOL 1–3) along a landward transect (T1). (c) Beach
profile after the Haiyan impact (view towards SE; date: 20 February 2014).
(d) Topographical cross section: the pre-Haiyan beach was
substantially eroded and Haiyan generated a landward thinning sand sheet
reaching up to 40 m inland in the back-barrier depression.
Bantayan study sites (BAN A and BAN B) with documented flood marks
and geomorphological impact of Typhoon Haiyan. (a–b) At BAN A,
Haiyan flooded the proximal part of the back-barrier plain and formed a
series of small, overlapping washover fans directly behind the beach ridge
(based on Google Earth/Digital Globe, 11 November 2013; inset in b
27 April 2001). (c) The washover fans reveal flat lobes at the base
(U1) and steep lobes on top (U2). (d–e) At BAN B, flooding levels
of more than 3 m above msl generated several distinct washover fans
(based on Google Earth/Digital Globe, 11 November 2013).
Typhoon Haiyan onshore deposits at site BAN A.
(a) A shore-perpendicular cross section (see Fig. 11 for location)
reveals two sedimentary units within the washover fans: a flat and planar
laminated unit at the base (U1) and steep, landward-inclined lobes on top
(U2). Below the pre-Haiyan soil, a second palaeosol is covered by a thin sand
layer (PE), possibly a former extreme wave event. (b) Sedimentary
characteristics of core BAN 4. The 26 cm thick Haiyan deposits are clearly
separated from older sediments by an initial soil formation. While the basal
unit 1 is laminated, unit 2 is characterized by gradual coarsening, changing
foraminifer composition (more Calcarina spp., less
Amphistegina spp.) and a decreasing percentage of fresh foraminifer
tests towards the top.
Sedimentary characteristics of Typhoon Haiyan deposits at site
BAN B. (a) Transect 2 (see Fig. 11 for location) reveals a landward
thinning structure directly behind a storm-induced breach in the coastal
barrier. A massive sand sheet at the base (U1) is overlain by a laminated
section (U2) in the proximal part of the washover fan. (b) BAN 1 is
composed of two distinct units: a massive unit at the base (unit 1) and a
laminated unit on top (unit 2).
Two different sedimentary facies units can be distinguished: unit 1 was not
encountered in HER 9 but comprises the entire typhoon deposit in HER 3–8
and 10 (Fig. 4a, b). It is characterized by normally graded to massive
(HER 3–7) or slightly laminated (HER 8, 10), poorly sorted (3.4–7.3),
unimodal fine-medium sand (mean of 85–230 µm). While sorting and
mean grain size remain more or less constant, unit 1 is thinning landward
from 8 cm at 90 m from the shoreline (HER 8) to only
3 mm at 210 m (HER 3). At the same time, the modal grain size
decreases from 570 to 220 µm (Fig. 3). Unit 2 is characterized by
a bimodal grain-size distribution that is significantly coarser than unit 1
(modes at 1.2 mm and 2.7 cm). It is only present at HER 9
(Fig. 4c), where it forms a 15 cm thick layer with steeply landward-inclined
layering and numerous angular coral fragments (constituting 77 % of total
mass). Both units are dominated by carbonates (nearly 100 % Mg-calcite
and aragonite). The foraminifer assemblage (determined for HER 10) is
dominated by Calcarina sp. (41 %), Amphistegina sp.
(21 %), and Baculogypsina sphaerulata (12 %), which show
significant reworking (< 8 % fresh tests, > 50 % strongly
reworked). Vertical variations of species composition and taphonomy within
HER 10 follow no clear trend. They rather roughly correlate with changes in
grain size, showing a higher percentage of abraded and broken tests for
coarser sediment sections (Fig. 4d).
Northeast Leyte (Tolosa)
Flood levels in Tacloban and in the coastal barangays to the south (Fig. 1b)
reached > 6 m above msl and locally inundation extended up to
1 km inland (Tajima et al., 2014). The typhoon was accompanied by strong
wind damage with uprooted trees, heavy beach erosion, and onshore
sedimentation. Due to the position to the typhoon centre – winds and surge
are strongest to the right of the typhoon track – water levels, beach
erosion, and sedimentation continuously decreased towards the south.
Although – according to surge levels – thickest onshore deposits are to be
expected in Tacloban and directly south of it, we report on the storm impact
on the nearshore area of Tolosa's beach-ridge plain (TOL, Fig. 1b), since the
northern areas are more densely populated and unaltered typhoon sediments
were hard to find 3 months after Haiyan. Floated grass found in trees and
bushes is evidence that Haiyan overtopped the 2.2 m high beach ridge (above
msl) and inundated the marshy back-barrier depression with water levels of at
least 4.7 m above msl (3.5 m flow depth) at a distance of
70–90 m from the shoreline (Fig. 5). At the landward end of the
transect, 300 m from the shoreline, water levels of at least
3.2 m above msl (1.8 m flow depth) have been recorded and satellite
images for the Tolosa area document a landward flooding extent of up to
800 m during Haiyan (Fig. 5a). However, since sandy deposition
stopped even more seaward, any further documentation was beyond the scope of
this survey. The direction of onshore flooding is indicated by grass bent
down in a NW–NNW direction, a westward collapsed north–south running wall
structure, and the NW orientation of floated palm trunks (Fig. 5b, c).
An erosive scarp at the seaward slope of the beach ridge indicates coastal
erosion by Typhoon Haiyan that had already partially recovered in February
2014. Pre- and post-typhoon satellite images show a shoreline retreat of
25–30 m (inset in Fig. 5b). Between 30 and 130 m inland, a sheet of
dark grey sand was deposited (TOL 3–14, Fig. 6). While 5–10 cm of
sediment was accumulated on top of the beach ridge (TOL 3–4), the maximum
thickness of 10–20 cm is reached directly leeward of the barrier
(TOL 5–8). Further landward, the thickness rapidly declines, reaching
2–5 cm at 70–90 m from the shoreline (TOL 9–11) and only a
few millimetres landwards of TOL 11 (TOL 12–14) (Fig. 6).
Investigations of trenches TOL 5 and 7 reveal the complete absence of
carbonate components and microfauna in the storm layer. Instead, the
composition is dominated by a mixture of feldspar (30–70 %), amphibole
(20–40 %), and a minor percentage of quartz (5–30 %). MS and LOI help to discriminate typhoon deposits from the
underlying soil but have constant values within the storm layer (Fig. 6c).
However, in the stratigraphy of TOL 7 (Fig. 6b) and in the µCT scan
of TOL 8 (Fig. 7), a normally graded unit 1 with scour marks at the base is
clearly discriminated from the successive horizontally laminated unit 2
within the Haiyan deposits.
Unit 1 forms a slightly normally graded to massive layer of unimodal fine to
medium sand at the base of trenches TOL 7–14. It covers the pre-Haiyan soil
and, at several places, bent stems of grass. In trenches TOL 9–14, unit 1
constitutes the entire event deposit and is covered by a thin mud cap
(Fig. 6a). While it is well sorted (1.5–1.7) throughout the entire transect,
its thickness decreases in a landward direction from 3–5 cm behind
the beach ridge (TOL 7–10) to < 1 cm at TOL 11–14. A landward fining
trend from 240 to 130 µm has been noted as well (Fig. 3). Unit 2
comprises a well-laminated layer of unimodal medium sand (Figs. 6, 7) and
makes up the upper part of the storm layer in the proximal part of the
transect (TOL 3–8). Sorting is similar to unit 1 (1.5–1.6), but the mean
grain size (245–328 µm) is slightly coarser. The lamination is
due to alternating concentrations of pyroxene (dark) and quartz (light) (XRD
results, Fig. 8a). In TOL 5, the laminae are slightly inclined landwards
(10–15∘) and show repeated coarsening and fining sequences that
superimpose a vertical coarsening upwards trend from ∼ 260 to
320 µm (Fig. 6c). Unit 2 forms a washover fan at the landward
slope of the beach ridge. While no clear trend in the thickness of the
deposit was found – values range between 5 cm at the top of the
beach ridge (TOL 4) and the landward edge of unit 2 (TOL 8) and
∼ 20 cm directly behind the ridge (TOL 5) – it slightly thins
landward (Fig. 3).
Northern Negros
All visited coral islands north of Negros (Carbin, Molocaboc, Suyac; Fig. 1c)
were affected by moderate flooding reaching a few tens of metres inland and
water levels less than 2–3 m above msl. Wind and wave directions
changed from NNE before to WSW after the passage of Typhoon Haiyan's centre
(Table S1 in the Supplement). Substantial beach erosion associated with
onshore transport of sediment was observed directly after Haiyan. While
coral-rubble ridges were formed in the intertidal zone of Suyac and Carbin
(CAR), onshore deposition was restricted to thin sand patches (few
centimetre) and small coral boulders or parts of sea walls (main axis
< 2 m) in the proximal coastal zones of Molocaboc (MOL) and Suyac.
Carbin Reef (CAR)
On Carbin Reef, a coral-rubble ridge along the western edge of the reef
platform is visible at low tide (Fig. 9), but entirely submerged at high
water. According to local fishermen (personal communication, 2014), the ridge
did not exist in its present form prior to Haiyan (Table S1), which means
that it had either been absent or significantly lower (Reyes et al., 2015).
The average crest height of the more than 300 m long section of the ridge
measured during the field survey is 0.20 m below msl (boulders reach
heights of 0.45 m above msl) without a significant trend in crest
elevation. The basal width is 10–20 m, and lobe structures at its
landward side cover corals in living position that must have been alive prior
to the typhoon. Morphology and sedimentary structure – the ridge is mainly
composed of centimetre to decimetre large coral fragments – allow the
discrimination of two units: lobes of greyish, algae-covered coral rubble
extending up to 20 m landward (unit 1) and light, freshly broken coral
branches that form steep lobes or patchy coverings extending not more than
10 m from the reef edge on top (unit 2). While both units are present
along the entire 300 m long section, they significantly broaden from
10 m to 20 m width along the northern section (Fig. 9).
Molocaboc Island (MOL)
At Molocaboc (Fig. 10), the coastal zone behind the 550 m wide intertidal
platform of the fringing reef is formed by a beach ridge with a crest height
of 2.4 m above msl, followed by a shallow back-barrier depression
densely covered by acacia shrubs at 2.0 m above msl. Behind the beach
ridge, fresh sandy deposits that rapidly thin landwards from 10 cm at
20 m shoreline distance to only 1 cm at a distance of 45 m
have been recorded. The unimodal, medium to coarse
(mean = 600–680 µm), moderately sorted (1.7–1.8) sand shows
no apparent sedimentary structures and a calcareous composition
(> 90 % Mg-calcite and aragonite); the foraminifer assemblage is
dominated by Calcarina sp. (58 %) and strongly reworked tests
(< 18 % fresh). The storm deposit contrasts with the underlying soil
which is likewise unimodal, but slightly finer (mean of
390–410 µm) and poorly sorted (3.4–4.7), and shows elevated
contents of organic matter (4–7 %) and reduced carbonate concentrations
(70–80 %). A reference sample from the shallow subtidal
(Rsubt) at 0.5 m below msl is unimodal, slightly finer
(mean = 310 µm), moderately sorted (2.4), and dominated by
different foraminifers (Quinqueloculina spp. [18 %],
Peneroplis pertusas [17 %], Elphidium sp. [14 %],
and
Ammonia beccarii [12 %]) with moderate reworking (74 %
fresh).
Bantayan
Eyewitnesses report limited storm surge elevations with moderate waves and
peak flooding arriving at low tide for the entire east coast of Bantayan
(Table S1). However, while onshore deposition is restricted to sand sheets of
a few centimetres in coastal areas directly behind active beach ridges, beach
erosion of several metres did occur. More pronounced flooding happened only
locally at the mouths of estuaries and resulted in the formation of small
washover fans (BAN A and B, Fig. 1c).
Bantayan A (BAN A)
At BAN A, the NE-exposed beach section (Fig. 11a, b) is separated from an
estuary river mouth by outcrops of the Pleistocene coral reef limestone with
1–2 m high cliffs. Cross sections reveal a succession of shore-parallel
beach ridges with elevations of 1.5–1.8 m above msl that are
interrupted by ∼ 1 m deep swales (Fig. 12a). While recent flooding
levels were indicated by debris lines at 1.4 m above msl along the
seaward and landward slopes of the second beach ridge, as well as on top of
the 2.9 m high (above msl) reef platform to the NW, the maximum wave height
during Haiyan was estimated to have reached 3.0–3.4 m above msl on
the basis of eyewitness accounts (Table S1). Furthermore, residents report
flooding of approximately 50 m inland and strong beach erosion during
Haiyan. However, for this section of the coast (E1–E5, Fig. 1c) similar
effects were observed after tropical storm Basyang in February 2014 were less intense, but struck at high tide (Fig. S3, Table S1).
Storm erosion created a steep shoreface with an erosive scarp and uprooted
palm trunks at the seaward slope of the first beach ridge. Residents report
lateral erosion of more than 5 m due to the combined influence of
Haiyan and Basyang along large sections of the beach (Table S1). However, the
formation of lobate washover fans in the back-barrier depression is already
visible on satellite images from November 2013 (Fig. 11b). The washover fans
are restricted to a 10 m wide section behind the beach but form prominent
landforms with two distinct stratigraphical units (Figs. 11c, 12a).
The basal unit 1 is related to flat washover lobes characterized by planar
lamination. It overlies a weakly developed soil with a sharp contact and
extends slightly further inland compared to the subsequent unit 2, which is
characterized by landward-inclined beds with a steep terminal front. The
internal structure is illustrated by the sedimentology (Fig. 12b) and
µCT scans (Fig. 7) of sediment core BAN 4. Below the pre-Haiyan
soil, a thin sheet of medium sand (PE) was found along the erosive cliff in
the ridge (Fig. 12a) as well as in BAN 4 (Figs. 7, 12b). The sand layer
covers an older palaeosol, composed of brown, slightly loamy sand. The basal
Haiyan deposit is composed of a laminated section of unimodal, moderately
sorted (1.8) medium-coarse sand (mean of 600–770 µm), rich in
strongly reworked foraminifer tests (35–60 %) dominated by
Calcarina sp. and Amphistegina sp. (unit 1). The uppermost
19 cm (unit 2) are composed of coarser (mean of
730–1300 µm) and slightly less sorted (2.0) unimodal sand with
similar species composition but a higher percentage of fresh or only slightly
reworked foraminifers (60–80 %) between 19 and 10 cm below the
surface (Fig. 12b). A reference sample collected at the present beach at
0.5 m above msl (BAN R1) is slightly finer than unit 1 (mean of
400 µm) and better sorted (1.6).
Bantayan B (BAN B)
Site BAN B is located in direct vicinity of an estuary mouth in northern
Bantayan, lined by a 500 m wide intertidal reef platform (Fig. 11d).
While the coast south of the estuary is characterized by a 2.8 m high
cliff (above msl) in the Pleistocene coral reef limestone, a ca. 100 m wide
sand spit backed by mangroves forms the section to the north (Fig. 11e). In a
shore-perpendicular direction, a beach ridge at 2.4 m above msl is
followed by a flat mud plain with mangroves and reef outcrops at
0.9–1.1 m above msl (Figs. 13a and S1 in the Supplement).
Eyewitnesses report significant flooding by Typhoon Haiyan, while Basyang –
different from the southern part of Bantayan – caused no marked inundation
(E6–E10 in Table S1). Flood marks in the form of floated debris on top of
the cliff or trapped in bushes and mangroves document inundation of at least
200 m inland. Minimum water levels decrease landwards from 3.1 m
above msl directly at the coastline to 2.1–2.3 m above msl at
60 m from the shoreline and only 1.4 m above msl at
160 m (Fig. 11e).
Haiyan-induced onshore sand deposition is less than 1–2 cm in the
back-barrier mangroves. Thicker deposits occur in the form of up to 30 cm
thick and 50 m wide washover fans at two sections behind breaches in the
barrier (Fig. 11e). Coast-perpendicular transects (T2 and T3, Fig. 11e)
reveal landward thinning trends from 28 cm at 10 m behind the
barrier to only 5 cm at a distance of 40 m (T2, Fig. 3) and
from 15 cm at 10 m behind the ridge to 1 cm at a
distance of 40 m respectively (T3 in Fig. S1a). Similar to site
BAN A, the pre-Haiyan soil had formed in a thin sand sheet, which has a
composition similar to the modern storm deposit and covers a second palaeosol
at the base of the trenches (PE in Fig. 13a). It might be the deposit of a
former storm.
In both washover fans, two successive sedimentary units were distinguished
and investigated in detail for cores BAN 1–3 (transect 2; Figs. 13b, S2).
Unit 1 is massive to slightly normally graded, composed of unimodal,
moderately sorted (1.5–2.6) medium sand (mean = 250–320 µm)
and shows a sharp boundary to the pre-Haiyan soil. It constitutes the entire
typhoon layer in BAN 3 but is topped by a markedly thicker unit 2 in BAN 1
and 2. Unit 2 is composed of several planar to slightly inclined, normally or
inversely graded beds of well-sorted (1.5–1.9) medium sand
(mean = 300–480) with shell and coral fragments as well as pieces of
litter. Both units contain > 90 % carbonates (aragonite and
Mg-calcite). The foraminifer
assemblage is dominated by Calcarina sp. (32 %),
Amphistegina sp. (19 %), and Ammonia beccarii
(19 %). Most of the tests are strongly reworked (< 17 % fresh),
though test preservation is much poorer (< 1 % fresh) in the
palaeosol, which is mainly composed of Amphistegina sp. (22 %),
Ammonia beccarii (21 %), Elphidium craticulatum
(12 %), and Quinqueloculina spp. (12 %). While unit 1 thins
landwards in T3, neither a clear thinning tendency nor a fining trend is
detectable in T2. Unit 2 is restricted to the proximal part of the washover
fans (BAN 1–2: Figs. 13 and S2; BAN A–B: Fig. S1a) and rapidly thins
landwards. For comparison with modern environments, reference samples were
collected at 0.6 m above msl (BAN R2) and at msl (BAN R3). Both
samples show a similar granulometry of unimodal, moderately sorted (1.9–2.0)
medium sand (mean of 280–550 µm).
Inter-site comparison of sediment characteristicsGeochemistry, mineralogy, and foraminifers
The comparison of XRD data from all four sites reveals two general types of
mineralogical compositions. While sediments from the carbonate environments
(MOL, HER, BAN) are dominated by calcite, Mg-calcite, and aragonite (at least
80 %) with minor percentages of feldspar or quartz, the samples from the
siliciclastic coast (TOL) are mainly composed of feldspar and amphibole with
a minor percentage of quartz (Fig. 8a).
A PCA on foraminifer data from the carbonate coasts indicates three
principle components (PCs) which explain 60 % of the total variability. PC1 (31 %) is
characterized by positive loadings for the genera Rosalina sp.,
Quinqueloculina spp., Peneroplis sp., Milionella
sp., Globigerinoides sp., and Coscinospira sp., while
Calcarina and Amphistegina are negatively correlated. PC2
(18 %) shows positive scores for Amphistegina sp. and strong
reworking, and negative ones for Heterostegina sp. PC3 (11 %) is
positively correlated with Challengerella sp. and negatively with
Spirillina sp. and Schlumbergerella sp. Plotting of PC1
against PC2 (Fig. S4) reveals differences between sediment from the foreshore
at Molocaboc (MOL Rsubt), on the one hand, and beach reference
samples and storm deposits from all locations on the other hand. Differences
between beach and storm deposits from different sites are less pronounced.
Granulometry
PCA combining granulometric data from all sites reveals three PCs that
explain 82 % of the total variability. Plotting of PC1 (positive loadings
for mean, sorting, gravel, and mode; negative ones for skewness and medium
sand) vs. PC2 (positive loadings for mean, skewness, and mode; negative ones
for sorting and mud) reveal both site-dependent properties as well as
distinct clusters for samples from unit 1, unit 2 and the palaeosols
(Fig. 8b1).
In addition, PCAs on grain-size parameters were performed for each site
separately. For TOL, PC1 (62.8 %) and PC2 (33.5 %) explain 96 %
of the total variability. Plotting of PC1 (positive loadings for skewness,
mode and medium sand, negative ones for sorting and mud) vs. PC2 (positive
scores for sorting, mode, and medium sand; negative ones for skewness) reveals
clusters for unit 1, unit 2, and the pre-Haiyan soil (Fig. 8b3). Samples from
HER reveal three PCs explaining 84 % of the total variability. Plotting
of PC1 (positive loadings for sorting, mud, and fine sand; negative ones for
skewness and coarse sand) vs. PC2 (positive scores for fine and medium sand;
negative ones for mode and sorting) reveals clustering of unit 1 and the
palaeosol, while unit 2 forms two separate sample groups (HER 10 and all
other trenches) (Fig. 8b2). At BAN, three PCs explain 86 % of the total
variation. Plotting of PC1 (positive loadings for mean, skewness, and medium
sand; negative ones for sorting and mud) vs. PC2 (positive scores for mean,
sorting, mode, and gravel; negative ones for medium sand) reveals clusters for
unit 1, unit 2, and the palaeosol, whereas reference samples from the upper
and lower beach plot into the cluster of unit 2 (Fig. 8b4), pointing to this
area as the main sediment source.
DiscussionSedimentary footprint of Typhoon Haiyan on the Philippines
Based on eyewitness accounts and the interpretation of satellite images, most
of the documented storm deposits can unambiguously be related to Typhoon
Haiyan. Even at BAN A, where eyewitnesses report strong coastal erosion by
tropical storm Basyang for the period between Haiyan and the field survey,
satellite images from 11 November clearly document the formation of the
washover fans by Typhoon Haiyan. Only at TOL, where both satellite images and
eye witnesses cannot unambiguously relate all of the documented onshore
deposits with Haiyan, the proximal parts of the washover sediments (unit 2 at
TOL) could be associated with post-Haiyan storm
waves of Basyang. In addition to the fine-grained storm deposits reported in
this study and summarized in Fig. 14, Haiyan moved block- and boulder-sized
reef-rock clasts at the coast of Eastern Samar, which have been discussed
elsewhere (May et al., 2015b).
Compilation of sedimentary characteristics and inferred formation
processes for the three different types of Haiyan deposits: coral ridges,
sand sheets, and washover fans. The cited references report on storm or
tsunami signatures with similar characteristics.
Sandy onshore deposits
Based on sedimentary and morphological criteria, the here presented sandy
onshore deposits of extreme wave events are classified into sand sheets and
washover fans. The classification reflects different flooding regimes and,
therefore, is assumed to represent hydrodynamic processes that lead to
different sediment characteristics (Fig. 14).
Separated from the underlying soil by a layer of bent grass,
the base of the typhoon deposits at TOL, HER, BAN B, and MOL (the local
units 1) is formed by slightly normally graded to massive layers of sand
(some of the layers were too thin to prove potential grading without
laboratory analyses). All these sand sheets extend at least 100 m inland,
are relatively thin (< 10 cm), and exhibit clear landward thinning
and slight fining trends (Fig. 3). Their appearance is similar to storm
deposits formed under inundation regimes related to extensive flooding of
back-barrier marshes described by Donnelly et al. (2006) or Wang and Horwitz
(2007). Indeed, complete inundation of coastal barriers and back-barrier
areas at HER, TOL, BAN B, and MOL is documented by flood marks. While at MOL
and BAN B flooding is facilitated by the absence of a pronounced beach ridge
or by a nearby river mouth, flow depths of nearly 4 m above surface
and the complete submergence of barriers with crests more than 3 m
above msl at TOL and HER are either due to the high storm surge levels (TOL)
or due to a combination of storm surge, high storm waves, and related
infra-gravity waves such as surf beat (HER) (Bricker et al., 2014; Roeber and
Bricker, 2015; Kennedy et al., 2016). The associated overland flow generally
followed shore-perpendicular directions, if not re-directed by shore-parallel
wall structures (TOL, Fig. 5). As indicated by scour marks (µCT scan
of TOL 8, Fig. 7) and a mostly normally graded structure of these sand
sheets, sediment dynamics related to this inundation regime (i.e. inundation
overwash; see Donnelly et al., 2006) are assumed to be characterized by
turbulent flow conditions and deposition from suspension. Even if clear
suspension grading as described by Jaffe et al. (2011) could not be detected,
the deposits are very similar to suspension-settled sediments described by
Williams (2009) for Hurricane Rita on the US coast. At least at TOL and HER
the deposition of these sand sheets was influenced by few tsunami-like
flooding pulses due to infra-gravity waves or seiches that amplified peak
inundation (Mori et al., 2014; Roeber and Bricker, 2015). Comparison of
granulometry as well as foraminifer taphonomy and species composition with
reference samples point to the beach as the dominant sediment source (BAN
R1-3) rather than foreshore (MOL Rsubt) or other
environments (Figs. 8b, S4). Nevertheless, obvious differences in the
granulometry and faunal composition of the sand sheets and modern beach sand
(Figs. 8b, S4) may indicate also minor contributions of sediments from other
source areas (the foreshore, deeper water, or landward areas), as reported by
Pilarczyk et al. (2016) for deposits of Typhoon Haiyan from Tanauan (Leyte)
and Basey (Samar). Alternatively, at least the differences in foraminifer
taphonomy may reflect alteration of the sediments due to wearing and
fracturing of foraminifer tests during transport in high energy flows
(Quintela et al., 2016).
Unit 2 at TOL, HER, and BAN B, as well as the entire storm deposits at BAN A
(the local units 1 and 2), is formed by lobate landforms of several
decimetres thickness directly behind the barrier. As the washover fans (a) are restricted to the proximal part of back-barrier
depressions (within < 50 m from the shoreline), (b) form lobes
with a gently inclined upper surface (1–5∘) and a steep landward
front, (c) show multiple grading within horizontally to inclined laminated
sections, and (d) reveal landward thinning and fining trends; these features
resemble typical storm-induced washover fans, e.g. described by Sedgwick and
Davis (2003), Phantuwongraj et al. (2013), or Williams (2015). They formed
due to wave-induced sediment transport over the coastal barriers. This took
place after the first flooding pulses inundated the coastal barriers, most
likely during the peak of the storm surge when the largest storm waves
occurred (Williams, 2009). Where coastal barriers were already inundated due
to the wind-induced storm surge (at TOL and BAN B) or due to the impact of
infra-gravity waves (at HER; May et al., 2015b), deposition is assumed to be
caused by waves breaking at the inundated barrier. At site BAN A, however,
where a basal sand sheet (unit 1 at TOL, BAN B, and HER) is absent, flood
marks indicate maximum water levels lower than the coastal barrier and
deposition was related to confined overwash (i.e. run-up overwash; see
Donnelly et al., 2006). The internal stratification and washover morphology
allow for a discrimination of a basal section with horizontal bedding
associated with flat lobes (unit 1) and a section with steeply inclined
layers with a steep avalanching front on top (unit 2). Horizontal lamination
is interpreted as the result of initial barrier overtopping into dry
back-barrier depressions associated with high flow velocities, whereas steep
lobes with inclined bedding are associated with delta-front sedimentation
into already flooded back-barrier depressions (Sedgwick and Davis, 2003;
Switzer and Jones, 2008), e.g. due to high levels of the preceding storm
surge or intensive rainfall.
The washover fans form either isolated structures such as at TOL, HER,
and BAN B or coalescing washover terraces (BAN A). Since breaching is
associated with barrier erosion and radial spread of water and sediment into
back-barrier depressions, granulometry, foraminifer assemblages, and taphonomy
indicate that storm sediments were mainly derived from the beach
(Figs. 8b, S4). The swash of multiple individual waves generates successions
of laminae which are probably caused by density separation in mixtures of
heavy minerals, quartz, and shell fragments during transport as traction load
(Komar and Wang, 1984). Vertical changes of granulometry and faunal
composition within individual washover deposits – such as the upward trend
towards coarser and stronger reworked deposits at BAN A (Fig. 12) – could be
related to slightly different sediment sources as a result of a successively
changing beach profile. However, the coarsening trend could just be
an artefact of the reduced settling velocity of platy shell fragments, which
are particularly abundant in this section of BAN 4 (Fig. 7), compared to more
spherical grains (Woodruff et al., 2008).
Intertidal coral-rubble ridges
Considering the interviews with local fishermen and the fact that parts of
the intertidal coral ridge at Carbin Reef in February 2014 covered reef
organisms which must have been alive shortly before, formation or at least
significant heightening can unambiguously be attributed to Typhoon Haiyan.
Coral-rubble ridges have repeatedly been reported to be a typical cyclone
signature (Scheffers et al., 2012), e.g. on Funafuti Atoll, where Maragos et
al. (1973) describe a cyclone-generated, intertidal coral ridge dominated by
sand to boulder-sized rubble derived from the foreshore. Diving surveys
before and after Haiyan revealed significant impact of the typhoon at the
seaward reef slope while organisms on the intertidal platform were nearly
untouched (Reyes et al., 2015), pointing to the entrainment of sediment from
the foreshore and its deposition at the reef edge. Since ridge formation
requires the repeated impact of breaking waves and wave swash is attenuated
by the high porosity of rubble ridges (Spiske and Halley, 2014), ridge
generation during Haiyan is mainly due to the impact of multiple storm waves
breaking on top of the inundated reef platform.
The two morpho-sedimentary ridge units at Carbin allow for two possible
explanations: a first interpretation is that the entire ridge was formed by
Typhoon Haiyan, whereas the distinct units would reflect changing wind and
wave directions. In accordance with a rotation of wind directions reported
for the passage of Haiyan, unit 2 is present only at the W-exposed section of
the ridge and may be interpreted as the result of stronger wind waves
connected with increased destruction of the living reef at Carbin's slope as
indicated by the fresh, angular coral fragments incorporated in unit 2.
However, the more pronounced algae cover on coral rubble of unit 1 and the
eyewitness accounts of ridge occurrence after Haiyan might also be explained
by the impact of several typhoons. In this case, formation of the initial
ridge (unit 1) that was not pronounced enough to be recognized by local
fishermen could have taken place during former storms. Afterwards, Haiyan
increased its height by adding fresh coral rubble on top (unit 2), which made
the ridges widely recognizable.
Spatial variability of fine-grained typhoon signatures
The spatial distribution and sedimentary structure of deposits formed by
Typhoon Haiyan was critically influenced by both local setting and the
hydrodynamic characteristics during inundation. Although onshore transport of
sand is rather ubiquitous, widespread sand sheets with a significant inland
extent > 100 m in the study area are restricted to locations with
exceptional surge and/or inundation levels. This includes the exposed
coastlines of Eastern Samar (HER) and the funnel-shaped San Pedro Bay (TOL),
where a remarkable amplification of surge levels to values
> 8 m above msl occurred during Haiyan (Mori et al., 2014). However,
pressure and wind driven surge alone do not explain the high inundation
levels for Eastern Samar, since numerical storm-surge models combined with
phase-averaged wave models infer rather low water levels of
∼ 2 m (Bricker et al., 2014). While surge levels in the San
Pedro Bay were additionally heightened by wave reflection in the enclosed
embayment (Mori et al., 2014), phase-resolved wave models imply that the
surprisingly high flooding levels along Eastern Samar result from
infra-gravity waves, caused by non-linear wave interactions with the fringing
reef (Roeber and Bricker, 2015; May et al., 2015b; Kennedy et al., 2016).
In contrast, washover features were even formed in areas with limited
flooding levels and restricted landward inundation such as Bantayan and
Northern Negros. Since washover deposits require high waves capable to
overtop coastal barriers, their local occurrence seems to be predetermined by
bathymetry (river estuaries) and coastal morphology (pre-existing gaps or
depressions in sandy barriers) as observed on Bantayan. Hence, they are
limited to small sections of the coast even in heavily affected areas. For
the localized occurrence of coral ridges on the reefs north of Negros,
obligatory requirements seem to be the presence of intertidal reef platforms
with coral rubble in the foreshore zone, exposure towards the main direction
of the storm waves, and waves high enough to entrain the foreshore sediment
and to lift it onto the reef platform.
Local geology, geomorphology, and sediment source are also the main factors
determining the composition of the investigated typhoon deposits (Fig. 8). On
the one hand, variations of total sediment composition (mineralogy and
granulometry) between different sites are more significant than variations
between different units at individual sites (Fig. 8a, b). The granulometry of
storm-transported sediment varies with the sediment availability at the beach
and the foreshore zone so that dominant grain size varies between HER, TOL
and BAN (Fig. 8b). Geochemistry and mineralogy basically reflect differences
between siliciclastic and carbonate coasts (Fig. 8a). Varying site-specific
compositions have also been reported for foraminifer assemblages of Haiyan
deposits by Pilarczyk et al. (2016). Likewise, clearly laminated washover
deposits (alternation of dark and light laminae) are linked to the presence
of heavy minerals at siliciclastic coastlines (TOL), while bedding structures
are less prominent in carbonate environments (BAN and HER). On an intra-site
level, on the other hand, mineralogy and geochemistry (due to insignificant
differences) and microfauna (due to a limited data set) only allow for a
separation between storm deposits and underlying palaeosols, while
sedimentary structures and grain-size data enable further discrimination of
distinct subunits (Fig. 8b). Although the number of reference samples from
recent environments is very limited in this study, the granulometric
differences between internal sublayers of the same storm deposit seem to be
related to varying sediment sources (beach vs. foreshore) or to different
hydrodynamic transport conditions as it has been described by Switzer and
Jones (2008). This is the case for the units 1 and 2 at sites TOL, HER, and
BAN (Fig. 8b). The large scatter within the unit 2 deposits at HER might be
explained by deposition of the HER 10 deposits by backwash, since the core
was taken close to a fluvial channel.
Implications for palaeotempestology
Since the sedimentary characteristics of Typhoon Haiyan's deposits show
significant site-specific variations, it is not possible to infer one
particular storm signature type. Nevertheless, storm deposits can clearly be
distinguished from most other depositional processes on the basis of
granulometry, internal structures, mineralogy, and faunal composition. Only
tsunamis might be capable to produce similar features, due to comparable
hydrodynamic characteristics that potentially allow for barrier overwash,
landward transport of sand for hundreds of metres, and the generation of
waves strong enough to entrain and lift subtidal coral rubble onto intertidal
reef platforms (Shanmugam, 2012). Although numerous features have been
established to discriminate between tsunami and storm deposits, including
among others thickness, lateral extent, granulometry, source areas, and
sedimentary structures of the event deposits (e.g. Morton et al., 2007;
Switzer and Jones, 2008), most of these sedimentary indicators seem to be of
local value only and cannot serve as universal discrimination criteria
(Shanmugam, 2012). Unfortunately, only very few tsunami signatures that might
facilitate discrimination by providing typical site-specific features (e.g.
Kortekaas and Dawson, 2007) have been described for the Philippines (Imamura
et al., 1995), making it hard to evaluate which of the sedimentary
characteristics documented for Typhoon Haiyan in this study are unique for
the impact of local typhoons.
So far, coral ridges have been reported to be exclusively generated by
cyclones (Scheffers et al., 2012), which seems straightforward. Ridge
formation requires repeated breaking waves as observed during cyclones (Nott,
2006), while the small number of inundation pulses characteristic for
tsunamis tends to produce randomly scattered boulder fields (Richmond et
al., 2011; Weiss, 2012). However, preservation of coral ridges is often
limited (Baines and McLean, 1976), and age determination remains challenging
due to potential reworking of the components (Scheffers et al., 2014). In
contrast, suspension-settled, normally graded sand sheets with large inland
extents as described at TOL or HER are typical signatures of tsunamis as well
(e.g. Jankaew et al., 2008), and also washover fans with internal lamination
as present at BAN or TOL have been reported for both storms (Switzer et
al., 2012) and tsunamis (Atwater et al., 2013). Likewise, most features
described for the sand sheets and washover fans in the study area have
already been observed for tsunami deposits: the composition and granulometry
of the storm-induced sand sheets and washover fans presented here are mainly
controlled by local geology rather than transport processes typical for
cyclones. With most sediment derived from the littoral zone, the main
sediment origin of the Haiyan deposits is similar to that of typical tsunami
deposits (e.g. Brill et al., 2014), and apart from that not analysed with
sufficient detail to determine secondary source areas that might enable
better differentiation from tsunami deposits. Finally, the documented
landward thinning and fining trends in the Haiyan deposits are similar to
those observed in many tsunami deposits (e.g. Goto et al., 2008).
Nevertheless, by adding local data to the knowledge on the manifold
expressions of storm deposits, the sediments accumulated by Haiyan offer some
valuable considerations regarding the interpretation of palaeoevent deposits
in geological records during future studies both in the Philippines (e.g. for
interpreting the young palaeoevents documented on Bantayan) and in general.
On the one hand, with not more than ∼ 250 m the inland extent
of the Haiyan-laid sand sheets presented here seems to be limited compared to
sandy deposits of many recent tsunamis with comparable inundation levels in
settings with flat topography that extend landwards for several kilometres
(e.g. Jankaew et al., 2008; Goto et al., 2011). This is true even at site
TOL that experienced onshore flooding levels > 5 m above msl and
where the flat topography did not hinder lateral inundation and sediment
transport. Although the landward limit of coastal inundation may be indicated
by mud deposits rather than by sand layers (e.g. Williams, 2010; Abe et
al., 2012; Goto et al., 2014), and our comparison is not based on tsunami
deposits from the same location (e.g. Kortekaas and Dawson, 2007), our
findings seem to corroborate previous conclusions that the landward extent of
tsunami sand sheets in low-lying coasts tends to be larger in comparison with
sandy storm deposits (see Morton et al., 2007). On the other hand, the
combined occurrence of (i) a thin (i.e. a few cm), massive, to slightly
normally graded basal unit formed by suspension settling due to surge-related
extensive coastal flooding (unit 1 at TOL and HER), and (ii) a laminated,
swash-induced, spatially limited washover unit on top (unit 2 at TOL, BAN B,
and HER; units 1 and 2 at BAN A) that in case of HER can both be
unambiguously related to Haiyan and therefore a single storm event, might be
rather indicative for cyclone-generated deposition. However, at TOL we cannot
exclude that part of this succession (i.e. the washover unit on top) is the
result of deposition by Basyang, since its time of formation could not be
proven by eyewitnesses or satellite images. In this case, the association of
the two units might also be explained by several storm events in a quick
sequence, whereas the strong beach erosion during a major event such as
Supertyphoon Haiyan would only increase the susceptibility of the sedimentary
coastline towards storm overwash during follow-up events.
Conclusions
The deposits of Typhoon Haiyan are strongly influenced by local
factors causing a wide variety of site-specific sedimentary and morphological
characteristics. Nevertheless, in spite of their spatial variability, the
deposits exhibit several storm-related depositional patterns – including
washover fans, sand sheets, and coral-rubble ridges – that can be related to
specific hydrodynamic processes and resemble typical storm features from all
over the world (Fig. 14).
Massive to normally graded onshore sand sheets extend 100–250 m inland,
tend to fine and thin landwards, and are related to suspension settling
during initial surge pulses that cause widespread inundation of coastal
lowlands and complete submergence of the coastal barrier. They have formed in
areas of significant amplification of the storm surge. In contrast, washover
fans are composed of small (10–50 m inland) sand lobes with steep landward
fronts and a distinct internal stratification that occur localized behind
breaches or depressions in coastal barriers as the result of traction
transport by overtopping waves and repeated overwash. Coral-rubble ridges
were formed on intertidal reef platforms by storm waves entraining coral
fragments from the reef slope and breaking at the reef edge.
Since Haiyan was one of the most powerful cyclones ever recorded, these and
other findings from the Philippines particularly add to the general knowledge
of extreme wave deposits and, ultimately, may contribute to discriminating
sediments of strong cyclones and tsunamis. While coral-rubble ridges so far
seem to be a unique feature of strong storms, the sandy onshore deposits left
by Haiyan – both sand sheets and washover fans – resemble those generated
by tsunamis in terms of sedimentary structure, granulometry, and sediment
sources (Fig. 14). However, the inland extents of sand sheets documented in
this study are significantly smaller compared to those of large tsunamis with
comparable flooding levels in similar topographical settings. Although more
extensive sand layers have been reported for other tropical cyclones, and moderate tsunamis generate sand sheets in the range of the
Haiyan deposits, the inland extent of onshore sand sheets at least provides a
valuable tendency for the discrimination of cyclones and strong tsunamis. In
addition, the combined occurrence of basal sand sheets and overlying confined
and well-stratified sand units may be indicative for cyclones, since they
result from the succession of surge-related extensive coastal inundation and
subsequent wave overwash during Haiyan at HER, which is not known to be
typical for tsunamis. However, the ambiguous origin of the respective units
at TOL, where they might as well result from two successive storms, prevents
a more definite conclusion.
Data availability
All raw data this study is
based on were currently uploaded to research gate and can be accessed and
cited with 10.13140/RG.2.2.24805.81125.
The Supplement related to this article is available online at doi:10.5194/nhess-16-2799-2016-supplement.
Acknowledgements
This research was financially supported by the Faculty of Mathematics and
Natural Sciences, University of Cologne (UoC), and a UoC postdoctoral grant.
Invaluable logistic support was provided by Karen Tiopes and Verna Vargas
(Department of Tourism, Leyte Branch). We are very thankful for the great
hospitality throughout the Visayas archipelago and the first-hand insights
provided by local interviewees, which is even more admirable considering the
trauma after the disaster. Ramil Villaflor is acknowledged for guiding us
safely through the islets north of Negros. We thank P. Costa and B. Jaffe for
their constructive comments during the review
process.
Edited by: T. Glade
Reviewed by: P. Costa and B. Jaffe
References
Abe, T., Goto, K., and Sugawara, D.: Relationship between the maximum extent
of tsunami sand and the inundation limit of the 2011 Tohoku-oki tsunami on
the Sendai Plain, Japan, Sediment. Geol., 282, 142–150, 2012.
Atwater, B. F., Cisternas, M., Yulianto, E., Prendergast, A. L., Jankaew, K.,
Eipert, A. A., Starin Fernando, W. I., Tejakusuma, I., Schiappacasse, I., and
Sawai, Y.: The 1960 tsunami on beach-ridge plains near Maullín, Chile:
Landward descent, renewed breaches, aggraded fans, multiple predecessors,
Andean Geol., 40, 393–418, 2013.
Baines, G. B. K. and McLean, R. F.: Sequential studies of hurricane deposit
evolution at Funafuti Atoll, Mar. Geol., 21, 1–8, 1976.
Bird, E. C. F. (Ed.): Philippines, in: Encyclopedia of the World's Coastal
Landforms, Springer Netherlands, Dordrecht, 1151–1156, 2010.
Bishop, P., Sanderson, D., Hansom, J., and Chaimanee, N.: Age-dating of
tsunami deposits: lessons from the 26 December 2004 tsunami in Thailand,
Geogr. J., 171, 379–384, 2005.
Blott, S. J. and Pye, K.: GRADISTAT: A grain size distribution and statistics
package for the analysis of unconsolidated sediments, Earth Surf. Proc.
Land., 26, 1237–1248, 2001.
Bricker, J., Takagi, H., Mas, E., Kure, S., Adriano, B., Yi, C., and Roeber,
V.: Spatial variation of damage due to storm surge and waves during Typhoon
Haiyan in the Philippines, Journal of Japan Society of Civil Engineers, 70,
I_231–I_235, 2014.
Brill, D., Klasen, N., Brückner, H., Jankaew, K., Scheffers, A.,
Kelletat, D., and Scheffers, S.: OSL dating of tsunami deposits from Phra
Thong Island, Thailand, Quat. Geochronol., 10, 224–229, 2012.
Brill, D., Pint, A., Jankaew, K., Frenzel, P., Schwarzer, K., Vött, A.,
and Brückner, H.: Sediment transport and hydrodynamic parameters of
tsunami waves recorded in onshore geoarchives – A case study from Thailand,
J. Coastal Res., 30, 922–941, 2014.
Brückner, H. and Brill, D.: Der Tsunami von 2004 – fünf Jahre
danach, Geographische Rundschau, 61, 4–10, 2009.
Corral, A., Ossó, A., and Llebot, J. E.: Scaling of tropical-cyclone
dissipation, Nat. Phys., 6, 693–696, 2010.
Cuadra, C., Biton, N. I., Cabacaba, K. M., Santiago, J., Suarez, J. K.,
Lapidez, J. P., Lagmay, A. M. F., and Malano, V.: Development of Inundation
Map for Bantayan Island, Cebu Using Delft3D-Flow Storm Surge Simulations of
Typhoon Haiyan, NOAH Open-File Reports, 3, 37–44, 2014.
Dimalanta, C. B., Suerte, L. O., Yumul, G. P., Tamayo, R. A., and Ramos,
E. G. L.: A Cretaceous supra-subduction oceanic basin source for Central
Philippine ophiolitic basement complexes: Geological and geophysical
constraints, Geosci. J., 10, 305–320, 2006.
Donnelly, C., Kraus, N., and Larson, M.: State of knowledge on the
measurement and modeling of coastal overwash, J. Coastal Res., 22, 965–991,
2006.
Engel, M., Brill, D., May, S. M., Brückner, H., and Reyes, M.:
Philippinen: “Haiyans” Erbe, Geographische Rundschau, 66, 54–57, 2014.
Folk, R. L. and Ward, W. C.: Brazos River bar: a study in the significance of
grain size parameters, J. Sediment. Petrol., 27, 3–26, 1957.
Fritz, H. M., Blount, C. D., Thwin, S., Kyaw Thu, M., and Chan, N.: Cyclone
Nargis storm surge in Myanmar, Nat. Geosci., 2, 448–449, 2009.
Goto, K., Imamura, F., Keerthi, N., Kunthasap, P., Matsui, T., Minoura, K.,
Ruangrassamee, A., Sugawara, D., and Supharatid, S.: Distribution and
significance of the 2004 Indian Ocean tsunami deposits: initial results from
Thailand and Sri Lanka, in: Tsunamiites – Features and Implications, edited
by: Shiki, T., Tsuji, Y., Yamazaki, T., and Minoura, K., Elsevier B.V.,
105–122, 2008.
Goto, K., Chague-Goff, C., Fujino, S., Goff, J., Jaffe, B., Nishimura, Y.,
Richmond, B., Sugawara, D., Szczucinski, W., Tappin, D. R., Witter, R., and
Yulianto, E.: New insights of tsunami hazard from the 2011 Tohoku-oki event,
Mar. Geol., 290, 46–50, 2011.
Goto, K., Hashimoto, K., Sugawara, D., Yanagisawa, H., and Abe, T.: Spatial
thickness variability of the 2011 Tohoku-oki tsunami deposits along the
coastline of Sendai Bay, Mar. Geol., 358, 38–48, 2014.
Hammer, Ø., Harper, D. A. T., and Ryan, P. D.: PAST: Paleontological
statistics software package for education and data analysis, Palaeontol.
Electron., 4, 9 pp., 2001.
Hippensteel, S. P., Eastin, M. D., and Garcia, W. J.: The geological legacy
of Hurricane Irene: Implications for the fidelity of the paleo-storm record,
GSA Today, 23, 4–10, 2013.
Imamura, F., Synolakis, C. E., Gica, E., Titov, V., Listanco, E., and Lee,
H. J.: Field survey of the 1994 Mindoro Island, Philippines tsunami, Pure
Appl. Geophys., 144, 875–890, 1995.IRIDeS (International Research Institute of Disaster Science): IRIDeS
fact-finding missions to Philippines, Initial Report, Tohoku University,
Japan, available at:
http://irides.tohoku.ac.jp/media/files/IRIDeS_Report_Haiyan_20140527.pdf,
last access: 31 May 2015, 2014.
Jaffe, B. E., Buckley, M. L., Richmond, B. M., Strotz, L., Etienne, S.,
Clark, K., and Gelfenbaum, G.: Flow speed estimated by inverse modeling of
sandy sediment deposited by the 29 September 2009 tsunami near Satitoa, east
Upolu, Samoa, Earth-Sci. Rev., 107, 23–37, 2011.
Jankaew, K., Atwater, B. F., Sawai, Y., Choowong, M., Charoentitirat, T.,
Martin, M. E., and Prendergast, A.: Medieval forewarning of the 2004 Indian
Ocean tsunami in Thailand, Nature, 455, 1228–1231, 2008.Kennedy, A., Mori, N., Zhang, Y., Yasuda, T., Chen, S.-E., Tajima, Y., Pecor,
W., and Toride, K.: Observations and modelling of coastal boulder transport
and loading during super typhoon Haiyan, Coast. Eng. J., 58, 1640004,
10.1142/S0578563416400040, 2016.
Knutson, T. R., McBride, J. L., Chan, J., Emanuel, K., Holland, G., Landsea,
C., Held, I., Kossin, J. P., Srivastava, A. K., and Sugi, M.: Tropical
cyclones and climate change, Nat. Geosci., 3, 157–163, 2010.
Komar, P. and Wang, C.: Processes of selective grain transport and the
formation of placers on beaches, J. Geol., 92, 637–655, 1984.
Kortekaas, S. and Dawson, A. G.: Distinguishing tsunami and storm deposits:
An example from Martinhal, SW Portugal, Sediment. Geol., 200, 208–221, 2007.
Lagmay, A. M. F., Agaton, R. P., Bahala, M. A. C., Briones, J. B. L. T.,
Cabacaba, K. M. C., Caro, C. V. C., Dasallas, L. L., Gonzalo, L. A. L.,
Ladiero, C. N., Lapidez, J. P., Mungcal, M. T. F., Puno, J. V. R., Ramos,
M. M. A. C., Santiago, J., Suarez, J. K., and Tablazon, J. P.: Devastating
storm surges of Typhoon Haiyan, International Journal of Disaster Risk
Reduction, 11, 1–12, 2015.
Maragos, J. E., Baines, G. B. K., and Beveridge, P. J.: Tropical Cyclone Bebe
creates a new land formation on Funafuti Atoll, Science, 181, 1161–1164,
1973.Mas, E., Bricker, J., Kure, S., Adriano, B., Yi, C., Suppasri, A., and
Koshimura, S.: Field survey report and satellite image interpretation of the
2013 Super Typhoon Haiyan in the Philippines, Nat. Hazards Earth Syst. Sci.,
15, 805–816, 10.5194/nhess-15-805-2015, 2015.
Matias, A., Ferreira, Ó., Vila-Concejo, A., Garcia, T., and Dias, J. A.:
Classification of washover dynamics in barrier islands, Geomorphology, 97,
655–674, 2008.May, S. M., Engel, M., Brill, D., Squire, P., Scheffers, A., and Kelletat,
D.: Coastal hazards from tropical cyclones and extratropical winter storms
based on Holocene storm chronologies, in: Coastal Hazards, edited by: Finkl,
C. W., Springer, Part VII, 557–585, 10.1007/978-94-007-5234-4_20,
2013.May, S. M., Brill, D., Engel, M., Scheffers, A., Pint, A., Opitz, S.,
Wennrich, V., Squire, P., Kelletat, D., and Brückner, H.: Traces of
historical tropical cyclones and tsunamis in the Ashburton Delta (NW
Australia), Sedimentology, 62, 1546–1572, 10.1111/sed.12192, 2015a.May, S. M., Engel, M., Brill, D., Cuadra, C., Lagmay, A. M. F., Santiago, J.,
Suarez, J. K., Reyes, M., and Brückner, H.: Block and boulder transport
in Eastern Samar (Philippines) during Supertyphoon Haiyan, Earth Surf.
Dynam., 3, 543–558, 10.5194/esurf-3-543-2015, 2015b.
Morgerman, J.: Super Typhoon Haiyan in Tacloban City & Leyte, Philippines,
iCyclone Chase Report, 47 pp., 2014.
Mori, N., Kato, M., Kim, S., Mase, H., Shibutani, Y., Takemi, T., Tsuboki,
K., and Yasuda, T.: Local amplification of storm surge by Super Typhoon
Haiyan in Leyte Gulf, Geophys. Res. Lett., 41, 5106–5113, 2014.
Morton, R., Gelfenbaum, G., and Jaffe, B.: Physical criteria for
distinguishing sandy tsunami and storm deposits using modern examples,
Sediment. Geol., 200, 184–207, 2007.
Nanayama, F., Shigeno, K., Satake, K., Shimokawa, K., Koitabashi, S.,
Miyasaka, S., and Ishii, M.: Sedimentary differences between the 1993
Hokkaido-nansei-oki tsunami and the 1959 Miyakojima typhoon at Taisei,
southwestern Hokkaido, northern Japan, Sediment. Geol., 135, 255–264, 2000.Navy METOC: http://www.metoc.gov.au/products/wms_globwave.php (last
access: 14 November 2014), 2014.NDRRMC (National Disaster Risk Reduction and Management Council): SitRep No.
107 Effects of Typhoon “YOLANDA” (HAIYAN), 14 March 2014, NDRRMC, Quezon
City, Philippines, available at:
http://www.ndrrmc.gov.ph/attachments/article/1329/Effects_of_Typhoon_YOLANDA_(HAIYAN)_SitRep_No_107_14MAR2014.pdf
(last access: 19 March 2014), 2014.
Nott, J.: Tropical cyclones and the evolution of the sedimentary coast of
Northern Australia, J. Coastal Res., 22, 49–62, 2006.
PAGASA: Climate change in the Philippines, Climatology and Agrometeorology
Division, Quezon City, Philippines, 2011.
PAGASA: Storm surge occurences during the passage of Typhoon Yolanda on
November 8, 2013 in selected areas of Leyte and Samar, Report of PAGASA
Stride team, 33 pp., 2014.
Phantuwongraj, S., Choowong, M., Nanayama, F., Hisada, K.-I., Charusiri, P.,
Chutakositkanon, V., Pailoplee, S., and Chabangbon, A.: Coastal geomorphic
conditions and styles of storm surge washover deposits from Southern
Thailand, Geomorphology, 192, 43–58, 2013.
Pilarczyk, J., Horton, B., Soria, J. L., Switzer, A., Siringan, F., Fritz,
H. M., Khan, N., Ildefonso, S., Doctor, A. A., and Garcia, M.:
Micropaleontology of the 2013 Typhoon Haiyan overwash sediments from the
Leyte Gulf, Philippines, Sediment. Geol., 339, 104–114, 2016.
Pun, I. F., Lin, I. I., and Lo, M. H.: Recent increase in high tropical cyclone
heat potential area in the Western North Pacific Ocean, Geophys. Res. Lett.,
40, 4680–4684, 2013.
Quintela, M., Costa, P. J. M., Fatela, F., Drago, T., Hoska, N., Andrade, C.,
and Freitas, M. C.: The AD 1755 tsunami deposits onshore and offshore of
Algarve (south Portugal): sediment transport interpretations based on the
study of Foraminifera assemblages, Quatern. Int., 408, 123–138,
2016.
Rangin, C., Müller, C., and Porth, H.: Neogene geodynamic evolution of
the Visayan Region, Geologisches Jahrbuch, Reihe B, 70, 7–27, 1989.
Reyes, M., Engel, M., May, S. M., Brill, D., and Brückner, H.: Life and
death after Super Typhoon Haiyan, Coral Reefs, 34, p. 419, 2015.
Richmond, B. M., Watt, S., Buckley, M., Jaffe, B., Gelfenbaum, G., and
Morton, R. A.: Recent storm and tsunami coarse-clast deposit characteristics,
southeast Hawaii, Mar. Geol., 283, 79–89, 2011.
Roeber, V. and Bricker, J.: Destructive tsunami-like wave generated by surf
beat over a coral reef during Typhoon Haiyan, Nature Communications, 6, 1–9,
2015.
Scheffers, A., Engel, M., Scheffers, S., Squire, P., and Kelletat, D.: Beach
ridge systems – archives for Holocene coastal events?, Prog. Phys. Geog.,
36, 5–37, 2012.
Scheffers, A. M., Engel, M., May, S. M., Scheffers, S. R., Joannes-Boyau, R.,
Hänßler, E., Kennedy, K., Kelletat, D., Brückner, H., Vött,
A., Schellmann, G., Schäbitz, F., Radtke, U., Sommer, B.,
Willershäuser, T., and Felis, T.: Potential and limits of combining
studies of coarse and fine-grained sediments for the coastal event history of
a Caribbean carbonate environment, Geological Society, London, Special
Publications, 388, 503–531, 2014.Schwartz, R. K.: Bedform and stratification characteristics of some modern
small-scale washover sand bodies, Sedimentology, 29, 835–849, 1982.
Sedgwick, P. E. and Davis, R. A.: Stratigraphy of washover deposits in
Florida: implications for recognition in the stratigraphic record, Mar.
Geol., 200, 31–48, 2003.
Shanmugam, G.: Process-sedimentological challenges in distinguishing
paleo-tsunami deposits, Nat. Hazards, 63, 5–30, 2012.
Soria, J. L. A., Switzer, A. D., Villanoy, C. L., Fritz, H. M., Bilgera,
P. H. T., Cabrera, O. C., Siringan, F. P., Yacat-Sta. Maria, Y., Ramos,
R. D., and Fernandez, I. Q.: Repeat storm surge disasters of Typhoon Haiyan
and its 1897 predecessor in the Philippines, B. Am. Meteorol. Soc., 97,
31–48, 2016.Spiske, M. and Halley, R. B.: A coral-rubble ridge as evidence for hurricane
overwash, Anegada (British Virgin Islands), Adv. Geosci., 38, 9–20,
10.5194/adgeo-38-9-2014, 2014.
Switzer, A. D. and Jones, B. G.: Setup, deposition, and sedimentary
characteristics of two storm overwash deposits, Abrahams Bosom Beach,
southeastern Australia, J. Coastal Res., 24, 189–200, 2008.
Switzer, A., Srinivasalu, S., Thangadurai, N., and Ram Mohan, V.: Bedding
structures in the Indian tsunami deposits that provide clues to the dynamics
of tsunami inundation, Geological Society, London, Special Publications, 361,
61–77, 2012.Tajima, Y., Yasuda, T., Pacheco, B. M., Cruz, E. C., Kawasaki, K., Nobuoka,
H., Miyamoto, M., Asano, Y., Arikawa, T., Ortigas, N. M., Aquino, R., Mata,
W., Valdez, J., and Briones, F.: Initial report of JSCE-PICE joint survey on
the storm surge disaster caused by Typhoon Haiyan, Coast. Eng. J., 56,
14500065, 10.1142/S0578563414500065, 2014.
Tuttle, M. P., Ruffman, A., Anderson, T., and Jeter, H.: Distinguishing
tsunami from storm deposits in eastern North America: the 1929 Grand Banks
Tsunami versus the 1991 Halloween Storm, Seismol. Res. Lett., 75, 117–131,
2004.
Wang, P. and Horwitz, M. H.: Erosional and depositional characteristics of
regional overwash deposits caused by multiple hurricanes, Sedimentology, 54,
545–564, 2007.
Weiss, R.: The mystery of boulders moved by tsunamis and storms, Mar. Geol.,
295–298, 28–33, 2012.
Williams, H. F. L.: Stratigraphy, sedimentology and microfossil content of
Hurricane Rita storm surge deposits in Southwest Louisiana, J. Coastal Res.,
25, 1041–1051, 2009.
Williams, H. F. L.: Storm surge deposition by Hurricane Ike on the McFaddin
National Wildlife Refuge, Texas: Implications for palaeotempestology studies,
J. Foramin. Res., 40, 210–219, 2010.
Williams, H. F. L.: Contrasting styles of Hurricane Irene washover
sedimentation on three east coast barrier islands: Cape Lookout, North
Carolina; Assateague Island, Virginia; and Fire Island, New York,
Geomorphology, 231, 182–192, 2015.
Woodruff, J. D., Donnelly, J. P., Mohrig, D., and Geyer, W. R.:
Reconstructing relative flooding intensities responsible for
hurricane-induced deposits from Laguna Playa Grande, Vieques, Puerto Rico,
Geology, 36, 391–394, 2008.