Tropical cyclones, or typhoons as they are known in the Western Pacifi c and in the Indian Ocean, are one of the most dangerous natural hazards. Their passage to the coast is accompanied by hurricane-force winds, extreme storms with the formation of tidal-surge floods and river floods. Extra dangerous are so-called super-typhoons passing the coast at an intervalof tens or hundreds of years(Huang and Xu, 2010). On the Philippine Isl and s, in the recent passage of supertyphoon Haiyan in November of 2013, more than four thous and people died; during the tropical cyclone passage in the bay of Bengal in November of 1970, about 300 000 people died or went missing. The probability of super-typhoons passage and the magnitude of the resulting extreme catastrophic events currently are not defi ned for most coasts of the Pacifi c and Indian Oceans. Long-term historicalobservation data similar to those performed for the coast of South China(Liu et al., 2001)are lacking here.

One of the approaches of classifying coasts with regard to typhoon hazard is the analysis of past events—the frequency of typhoon passages and the magnitude of the relating disasters. For this, historical records are often used(Liu et al., 2001)such as the results of meteorologicalobservations in conjunction with historical data(Golitsin and Vasilieva, 2001; Zhou et al., 2011) and the sedimentary record in lagoon and lake sediments. For the east coast of North America methods have recently been developed, based on tracking extreme storms traces when sedimentary material with a composition different from that of the previously accumulated deposits enters coastal salt marshes, lagoons, lakes, and swamps from the sea(Travis, 2000; Donnelly, 2005; Lambert et al., 2008). This approach can be defi ned on the basis of complex lithological, geochemical, and micropaleontological characteristics. A special term “paleotempestology” was even proposed to designate this fi eld of paleoenvironment sciences(Liu et al., 2009). However, these methods are not commonly used in the regions of the Pacific and Indian Oceans, as here the extreme supply of sediment to the coastal lowl and soccurs during the passage of tsunamis.

In this study, we used another way to reconstruct the history of the typhoons passage—by the presence(in the offshore sediments)of layers accumulated during extreme floods that accompany typhoons. The possibility of applying such an approach is based on submillimeter scanning of sediment cores using X-ray fl uorescence analysis with synchrotron radiation(XRF SR; Darin et al., 2013). The low limits of detection allow the analysis of the chemical composition of sediments, providing a much larger range of elements and more sensitivity as compared with conventional core-scanners.

Amur Bay, Sea of Japan(Fig. 1), was chosen for the research because its northern part is characterized by very high sedimentation rates(3–5 mm/a) and anoxic conditions(Tishchenko et al., 2011), hindering the development of benthic fauna and therefore reducing sediment bioturbation. The frequency of typhoons passing is not high here; they are registered not every year and no more than 1–2 times per year. Floods are the main factor that creates natural disasters during the passage of typhoons in the selected study area of South Primorye(Plotnikov and Tunegolovec, 2001).

Fig. 1 The location sites of bottom-sediment sampling in the Amur Bay by core in 2008 (08-3) and by core and box core in 2012 (12-3, 12-4, 12-5) Grey shading: areas with a lack of oxygen in the bottom waters from measurements in August 2008(Tishchenko et al.，2011); dark: less than 40 μmol/L，light: 40–80 μmol/L; black solid and dashed lines: the boundaries of the maximum mixing of sea and river water(isohaline 20×10-3)for the summer low and high river runoff，respectively(Anikiev et al.，2000); gray lines: isobaths in meters. Inset: An overview map of the study area(bottom) and a boundary of the Razdolnaya(Suifen)River basin(top).

Figure options

Depositional features in the northern part of the bay are infl uenced by the Razdolnaya(Suifen)River. Annual river runoff, according to the observations for the period since 1939, varies from 14 to 139 m ³ /s, and the total solid runoff from 2 to 32 kg/s. Such a wide range of annual variations is caused by monsoon climate(Mikhailik et al., 2011). The zone of the mixing marine and river waters is displaced into the central part of the bay during the summer floods, as demonstrated by the location of the 20×10^-3 isohaline(Fig. 1). During low-water periods and especially in winter fresh river waters do not penetrate outside the estuary. Due to the prevailing system of water circulation in the bay, water masses and suspended sediment carried by the Razdolnaya River during summer floods mainly move along the western shore of the Amur Bay, as confi rmed by satellite observations(Tishchenko et al., 2011).

Another typical feature of the Amur Bay is the lack of oxygen or its low concentrations in the bottom waters in its central part in summer. The low-oxygen period starts in May simultaneously with the heating of the surface waters, reaching minimum saturation in July and August when water stratifi cation in the bay is dramatically increasing because of warming and freshening of the surface waters(Tishchenko et al., 2011).

In 2012 we sampled three stations in the northern part of the Amur Bay(Fig. 1)using a gravity corer with diameter 60 mm and a 15 cm×15 cm×20 cm box-corer. Core 08-3 was obtained in 2008, and station 12-4(sediment column 75 cm in length)was taken in the same place. We cut longitudinal slabs of 170 mm×15 mm×7 mm size along each core, and then they were freeze dried, and penetrated by resin. For measurements, we used these resin-impregnated slabs, preserving the fine structure of the starting material. They are suitable for prolonged storage, for study under the optical microscope, and for X-ray and other current methods of microanalysis.

An experimental setup was designed and implemented allowing us to perform scanning XRF microanalysis, using synchrotron radiation of the VEPP-3 as the excitation source in the Institute of Nuclear Physics SB RAS(Novosibirsk). Analysis was performed for 25 elements(from Cl to Nb by K series, Pb, Th, U—by L series), with detection limits down to(5×10 ^-5)% and a spatial resolution of about 0.1–1 mm. The spectra were processed by a nonlinear least squares optimization method. As basic fi tting functions, we used sets of Gaussian multiplets in which the position and mutual intensity of the individual peaks were selected from fundamental tables of the energies and intensities of fl uorescence lines for different elements(Darin et al., 2013).

Concentrations of minor elements were measured along the profi les of the core samples at excitation energies of 18 and 24 keV and with steps of 0.1, 0.2, and 0.5 mm. The duration of a measurement at each point ranged from 5 to 25 s, depending on the levelof concentration of the analyzed element and the tasks to be performed. Simultaneously, we determined concentrations of Ca, K, Ti, Mn, Fe, V, Cr, Ni, Cu, Zn, Ga, Pb, Rb, Sr, Y, Zr, Br, As, Se, Nb, Mo, Th, and U with detection limits of up to 0.5×10 ^-6(Table 1) and X-ray density as well.

Table 1 Set of analyzed elements and limits of detection (LD)

Table options

Grain-size composition of sediments, mercury, organic carbon, and some major and microelements were determined by st and ard methods. The age of the sediments was evaluated by distribution of nonequilibrium ²¹⁰ Pb, which is applicable for dating the time intervalof 100–150 years(Goldberg, 1963; Appleby and Oldfield, 1978) and by the peaks of ¹³⁷ Cs content. ²¹⁰ Pb and ¹³⁷ Cs radioactivity measurements in dried and powdered samples of sediment were carried out using st and ard methods at the Department of Radiochemistry, Moscow State University. In addition, semiconductor low-background gamma-ray spectrometry based on the coaxial Ge detector with low background cryostat EGPC-192-P21, with a spectrometer equipped by processor FP-6300B, EURISYS MESURES, was used in the Institute of Geology and Mineralogy SB RAS.

Sediment accumulation rates were evaluated by ²¹⁰ Pb measurements in box cores 12-3, 12-4 and 12-5 as 5.2, 4.2 and 3.6 mm/a, respectively(Fig. 2). Further, all geochemical profi les on the timescale are based on the these accumulation rates. For core 08-3 it was 4.1 mm/a, as confi rmed by ²¹⁰ Pb and ¹³⁷ Cs distribution(Fig. 3). The maximum of ¹³⁷ Cs on the 19.5 cm horizon corresponds to 1963, when there was a global atmospheric fallout of ¹³⁷ Cs in connection with nuclear tests in the atmosphere(Appleby and Oldfield, 1978). The top peak on the 9.5 cm horizon may correspond to the time of the man-made disaster at Chernobyl in April 1986 or to the explosion of a nuclear submarine reactor in the Chazhma Bay in October 1985(Soifer et al., 1999). That accident occurred 54 km southeast of the sampling site in the direction of the radioactive fall-out. Other events of the last century are marked by lithological and geochemical indicators(Fig. 3). The drastic change of pelite fraction content(Pl-profi le)around AD 1900 marks the beginning of intensive agriculture in the region. At the turn of 1960, the anthropogenic pollution of mercury, lead and zinc increases by the combustion of coal and industrial development. At the same period, organic carbon content in the sediments increases due to global warming.

Fig. 2 Distribution of 210 Pb ex content in the sediment columns from box corers A-12-3bc, A-12-4bc and A-12-5bc Sampling points represent mid of analyzed intervals 10 mm in length, shown as shaded area. Standard error is marked by vertical bars. Instrumental error 2s=~20% for the range 30–100 Bk/kg and ~15% for 100–300 Bk/kg. Calculated sediment accumulation rates—see in the text.

Figure options

Fig. 3 The distribution of lead, zinc, mercury, organic carbon, grain-size fraction ＜0.01 mm (Pl), 137 Cs and 210 Pb in sediment along the core 08-3 Sediment accumulation rate calculated by 210 Pb as well as 137 Cs distribution is 4.1 mm/a on wet sediment. 137 Cs peak on depth 9.5 cm (right scale) is corresponded to 1985–1986 (left scale), and on 19.5 cm to 1963 (see also text).

Figure options

When analyzing the results of XRF SR scanning of sediment cores with increments of 0.5 mm in the northern part of the bay, a specifi c distribution of bromine different from the distribution of allother elements was identified(Fig. 4). Distinct 3–8 mm thick sediment layers with a low content of Br(LCB)can be distinguished. They are not revealed in box core 12-3 from the near estuarine zone(see Fig. 1). The general background of bromine in sediments of cores from the central part of the bay increases in the zone of maximal anoxia up to(50–60)×10 ^-6 and more, but in the LCB layer minimal Br concentrations are similar to the background contents((20–50)×10 ^-6)measured in box core 12-3. LCB layers, distinguished by the analysis at 0.5 mm increments, are also present in the graphs showing measurement with 1.06 mm increments(Fig. 5), but they are identifi ed not as individual characteristic events but only as ordinary variability.

Fig. 4 Variations of the element content in the box core 12-5bc by the results of XRF SR scanning in increments of 0.5 mm XRD: X-ray density (conventional units).

Figure options

Fig. 5 Bromine content in sediments from Station 12-4 Top: the box core 12-4bc, measured in increments of 0.5 mm (dash line is the trend of linear filtration); below: the upper part of core 12-4, measured in increments of 1.06 mm.

Figure options

Such an anomalous distribution of bromine can be explained by its specifi c geochemical behavior. Bromine is a typicaloceanic element; up to 75% of its amount in the earth’s crust is located in the ocean(Harvey, 1980), Marine organisms play an important role in the natural cycle of bromine, concentrating it from seawater in the biogenic matter(Gribble, 1998). The bromine( and iodine)content as well as the ratio of Br(I)/C org are often used for the discrimination of marine and continental sediments or for the determination of the origin of their organic matter(Malcolm and Price, 1984; Gribble, 1998; Mayer et al., 2007). Therefore it can be assumed that the LCB layers in the sediments of the middle bay only contain land-derived organic carbon with a few autochthonous material.

In sediments from box-corer 12-3, collected near the mouth of the Razdolnaya River, the bromine content is minimal and corresponds to that of the LCB layers of the middle part of the bay(Fig. 6), while the organic carbon content in the sediments of these areas is approximately the same(1.5%–2.5%)(Tkalin et al., 1996). That may indicate the presence of only allochthonous(terrigenous)organic matter with a bromine content that is low in comparison to autochthonous(marine)organic in sediments of the estuary seaside and in the LCB layers. The latter can be explained by very rapid accumulation of sediment in the middle of the bay due to mass accumulation of suspended matter from the river coming from the mouth of the Razdolnaya River under extreme flood conditions, and to active mixing of coastal waters. Such conditions in the region are typical for the period of typhoons and , to a lesser extent, of deep cyclones(Plotnikov and Tunegolovec, 2001). When comparing the bromine content in the sediments of the two cores from the central part of the bay on the timescale, the LCB layers are confi ned to certain time intervals(Fig. 6). Nine such events were revealed in sediment from two cores for the period from 1966 to 2012. The age of the events is in a good agreement with published dates on the floods in Vladivostok and in the Razdolnaya River basin(Plotnikov and Tunegolovec, 2001; Pacific typhoon season http://en.academic.ru/ dic.nsf/enwiki/1519766). Small shifts can be explained by errors in the timescale due to uneven rates of sedimentation. In some cases, when typhoons came in two consecutive years, it was not possible to exactly determine to which one the LCB layer was related. Only Typhoon “Melissa” of September 17– 21, 1994, accompanied by extreme precipitation in Vladivostok(188 mm precipitation in the fi rst four days) and by floods of up to 7 m on the rivers to the east of it(Plotnikov and Tunegolovec, 2001), is not explicitly mapped to any LCB horizon. This is probably due to the predominant infl uence of the Razdolnaya River on sedimentation in the central part of the Amur Bay and to the small role of the runoff from its east coast in that period.

Fig. 6 Synchronization between negative peaks of bromine content in box cores 12-4bc, 12-5bc, 12-3bc and extreme annual precipitation (broken line) caused by typhoons or deep cyclones (shaded bands) The timescale for sediment cores is based on 210 Pb dating (see Fig.2); Dashed lines with numbers: 1: typhoon Talas, 09. 2011; 2: cyclone, 08. 2001; 3: typhoon Melissa, 09. 1994; 4: typhoon Robin, 07. 1990; 5: typhoon Judy, 07. 1989; 6: typhoon Orchid, 09. 1980; 7: typhoon Irving, 08. 1979; 8: typhoon Gilda, 06. 1974; 9: cyclone, 09. 1968.

Figure options

Application of the method of submillimeter X-ray scanning, using synchrotron radiation as excitation source, allowed us to study the distribution of bromine in the core sediments of the Amur Bay, Sea of Japan in detail. The detected layers with a low content of Br formed by sedimentary material supplied by the Razdolnaya River under extreme floods followed by storms at sea. Under such conditions a large amount of terrigenous matter is supplied to the central part of the bay, which is usually accumulated in river estuaries or on the shallow estuary shelf. It does not contain organic materialof marine origin as documented by a very low bromine and possibly iodine content. These indicators are proposed to identify signs of paleo-typhoons in shelf sediments with a uniform lithological composition. The proportion of these elements to the organic carbon could be used for such purposes as well in the analysis of sediments with different particle size distribution. In the Amur Bay, Sea of Japan, the possibility to distinguish paleo-typhoon layers depends on a favorable combination of high sedimentation rate and the presence of anoxic conditions. The latter provides better preservation of the organic matter coming with plankton remains, which is the main hub of bromine and other halogens, as well as less bioturbation of sediments because of inhibition of the benthic fauna. At a sedimentation rate of ~4 mm/a, paleo-typhoon layers and their thickness are identifi ed at discrete measurements at 0.5 mm resolution, but at a resolution of 1.06 mm their diagnosis becomes problematic.

The authors are grateful to M. V. Ivanov, V. V. Kalinchuk and A. N. Kolesnik for their assistance during the expeditions. We appreciate the contribution of the editors and reviewers in improving the quality of submitted article.