Nine unusually large tsunami deposits from the past 4000 years at Kiritappu marsh along the southern Kuril Trench

Large earthquakes along the Kuril subduction zone in northern Japan are known to have caused damaging tsunami, although there is a little information on historical earthquakes and tsunami in this area because no documents exist before the 19th century that might refer to tsunami events. To determine the likely timing and size of future events we need information on their recurrence intervals and to do this for the prehistoric past we have investigated sediments located in the Kiritappu marsh in eastern Hokaido that we interpret as laid down by tsunami. Using reliable multiple lines of evidence from sedimentological, geomorphological, micropaleontological, and chronological results, we identify 13 tsunami sands. Two of these lie within a peat bed above a historical tephra, Ta-a (AD 1739); the upper one probably corresponds to the AD 1843 Tempo Tokachi-oki earthquake (M 8.2) tsunami, and the lower to either the AD 1952 Tokachi-oki earthquake (M 8.2) tsunami or the AD 1960 Chilean earthquake (M 9.5) tsunami. Underlying are 11 prehistoric tsunami sand beds (nine large sand beds and two smaller sand beds) deposited during the past 4000 years. Because of the wide spatial distribution of the large sand beds, and inundation distances inland of between 1200 to 3000 m, we suggest that they record unusually large tsunamis along the Kuril subduction zone. According to our analyses, these tsunami sands were derived from the coastal area and, although they do not show clear graded bedding, they commonly have gradational upper boundaries and erosional bases and include internal sedimentary structures such as plane beds, dunes, and current ripples, reflecting bedload transportation. Based on our results we calculate the recurrence interval of unusually large earthquakes (probably M 8.6) along the Kuril subduction zone as about 365–553 years and estimate the youngest large event to have occurred in the 17th century.     

A nonlinear parametric model based on a power law relationship for predicting the coastal tsunami height

When a subduction-zone earthquake occurs, the tsunami height must be predicted to cope with the damage generated by the tsunami. Therefore, tsunami height prediction methods have been studied using simulation data acquired by large-scale calculations. In this research, we consider the existence of a nonlinear power law relationship between the water pressure gauge data observed by the Dense Oceanfloor Network System for Earthquakes and Tsunamis (DONET) and the coastal tsunami height. Using this relationship, we propose a nonlinear parametric model and conduct a prediction experiment to compare the accuracy of the proposed method with those of previous methods and implement particular improvements to the extrapolation accuracy.

     

Interannual variability of North Pacific eastern subtropical mode water formation in the 1990s derived from a 4-dimensional variational ocean data assimilation experiment

The interannual variability of eastern subtropical mode water (ESTMW) formation in the North Pacific is examined using a new ocean dataset constructed by a 4-dimensional variational data assimilation experiment covering the decade of the 1990s. The volume of newly formed ESTMW varies due to interannual variability in the following three physical processes taking place in the surface layer: (1) convergence in the transport of surface saline water induced by Ekman flow in the vicinity of the formation region, (2) thermal stratification in the preconditioning phase in association with the insolation anomaly induced largely by low-level cloud coverage, and (3) wintertime surface cooling in the eastern subtropics. We find that, in addition to the surface forcing, the properties of both the ESTMW and the upper mixed-layer water are broadly controlled by the volume of the new ESTMW component, and that the variations in the upper mixed-layer water affect the properties of ESTMW formed in the following winter. Due to the combined effect of these processes, the ESTMW subducts down to subsurface layers with a wide range of σ_θ values lying between 24.8 and 25.4 and with significant interannual variation in water mass formation.     

Long-term continuous in situ potentiometrically measured redox potential in anoxic groundwater with high methane and iron contents

The in situ redox potential (Eh) in anoxic groundwater with high methane and iron contents (approximately 12.3 and 28.4 mg/L, respectively) was potentiometrically measured to identify the processes that control Eh. The measured Eh ranged from −213 to −187 mV; it had an inverse correlation with the concentration of methane and no correlation with that of iron. The saturation indices indicate that goethite and amorphous FeS were nearly at solubility equilibrium. A comparison of the measured Eh with those calculated for the particular redox pairs indicates that either Fe²⁺/FeOOH or CH₄/CO₂, but not sulfur redox pairs, controlled the measured Eh. The inverse relationship between measured Eh and methane concentration suggests possible control of the redox conditions by the CH₄/CO₂ redox pair. Furthermore, the equilibrium solubility state of goethite, which has higher crystallinity and lower solubility than Fe(OH)₃ indicates that the iron reaction was electrochemically irreversible. This further supports the contribution of the CH₄/CO₂ pair to controlling the measured Eh of groundwater.     

Dating of the Dome Fuji shallow ice core based on a record of volcanic eruptions from AD 1260 to AD 2001

We measured the concentration of non-sea-salt sulfate () in the Dome Fuji shallow ice core (Antarctica) from the surface to 40 m depth with the aim of dating the core with reference to the record of volcanic eruptions. Three huge spikes related to large-scale volcanic eruptions were detected at depths of 12.5, 29.9, and 38.8 m, correlated to the eruptions of Tambora (AD 1815), Kuwae (AD 1452) and an unknown event (AD 1259), respectively. We identified another nine spikes related to accurately dated eruption events. The shallow ice core was dated from AD 1260 to AD 2001 based on these 12 eruption events and the assumption of constant annual snow accumulation in the periods between eruption events. The results yield a maximum correction of ∼20 years compared with the dating proposed in a previous study. The annual accumulation varied within ±∼15% of the average water equivalent value over the study period (25.5 mm).     

A Proposed Method to Estimate In Situ Dissolved Gas Concentrations in Gas‐Saturated Groundwater

Gas‐saturated groundwater forms bubbles when brought to atmospheric pressure, preventing precise determination of its in situ dissolved gas concentrations. To overcome this problem, a modeling approach called the atmospheric sampling method is suggested here to recover the in situ dissolved gas concentrations of groundwater collected ex situ under atmospheric conditions at the Horonobe Underground Research Laboratory, Japan. The results from this method were compared with results measured at the same locations using two special techniques, the sealed sampler and pre‐evacuated vial methods, that have been developed to collect groundwater under its in situ conditions. In gas‐saturated groundwater cases, dissolved methane and inorganic carbon concentrations derived using the atmospheric sampling method were mostly within ±4 and ±10%, respectively, of values from the sealed sampler and pre‐evacuated vial methods. In gas‐unsaturated groundwater, however, the atmospheric sampling method overestimated the in situ dissolved methane concentrations, because the groundwater pressure at which bubbles appear (P_critical) was overestimated. The atmospheric sampling method is recommended for use where gas‐saturated groundwater can be collected only ex situ under atmospheric conditions.     

Comparison of potentially toxic metals leaching from weathered rocks at a closed mine site between laboratory columns and field observation

Potentially toxic metals, such as Cu, Pb and Zn, are leached from weathered rocks at many closed mine sites due to the acidic environments and mineralogical modifications. The mobilized toxic metals may cause contamination of surrounding water bodies. In this study, both laboratory column experiments and field observations at a former mine located in the north of Japan were carried out to compare the leaching behavior of Cu, Pb and Zn. The thickness of the surface weathered rock was varied (10, 20 and 30 cm) for the columns experiments while porous cups for porewater sampling were set up at different depths (GL-15, -45, -70, and -95 cm) for the field observations. Deionized water was added once a week over 75 weeks to the columns to simulate rainfall while porewater was extracted by a vacuum pump in several sampling campaigns (over 18 months). Similar low pH and leaching behavior of potentially toxic metals were observed for column experiments and field observations. A moderate increase in concentration with depth was observed for Cu and Zn. However, no increase in concentration was observed for Pb. This suggests that the leaching of Cu and Zn is enhanced by the length of the flow pathway through the weathered rock layer while Pb concentration is restricted by the precipitation of insoluble Pb sulfate. Thus, the thickness of the weathered rock layer is an important parameter that should be taken into consideration to estimate the loads of some potentially toxic metals, which is important when designing remediation schemes.     

Atmospheric observations of liquid water in cloud and of chemical species in aerosols and gases near the ground and in fallen snow at Svalbard, Arctic

Columnar observations of liquid water and of radar echo intensity in cloud were carried out, using a microwave radiometer and a vertically pointing radar respectively, in Ny-Ålesund, Svalbard. Chemical concentrations were also measured in aerosols, gases and snowfalls. Clouds with a large proportion of liquid water moved over the site after snow clouds, with a much lower liquid water content, had been present for about 16 h. The mass concentrations of most chemical species in snowfalls were lower from the first set of clouds than the second. The NO₃⁻ and SO₄²⁻ concentrations in gases and aerosols associated with the first set of clouds were higher than in the second set, but Cl⁻ concentration was less for the first set than the second.     

In situ filtration sampler for taking filtered water samples

A new type ofin situ filtration sampler is described. The all Teflon sampler is hung on a nylon rope attached to a steel wire (Fig. 3). By dropping a messenger, the sampler piston is pulled down by a released weight of 50 kg, and the sampler is filled with about 400 ml of filtered water. The sample is not contaminated by metals because there is no metal around the sampler.     

Groundwater radon anomalies associated with earthquakes

Earthquake-related changes in groundwater radon have been detected at a sensitive observation site located right on a major active fault in Northeast Japan. A time-series analysis based on Bayesian statistics was successfully applied to remove background variations from the observed radon data, enabling us to examine the earthquake-related changes in detail.

We set a simple criterion of amplitude and duration for an anomaly observed in our radon data; we define an anomaly as a radon change that kept its level beyond 2σ (a standard deviation over the whole observation period) during a period longer than one day. We have observed 20 radon anomalies that satisfied this criterion from January 1984 to December 1988. Most of these anomalies have turned out to be related to large earthquakes that occurred in East Japan and its surrounding area; we have identified 12 post-seismic and 2-pre-seismic radon anomalies out of a total of 30 earthquakes with magnitude M 6.0 and hypocentral distance D 1000 km.

The typical pattern of the post-seismic anomalies is a radon decrease which started just after an earthquake, lasting for periods ranging from a few days to more than one week. The amplitude of the post-seismic anomalies depends on both magnitude and hypocentral distance, and can, in general, be expressed by a simple magnitude-distance relationships.

A possible pre-seismic anomaly was observed about one week before the largest earthquake that occurred in this region during the observation period (March 6, 1984; M = 7.9, D = 1000 km). Another possible pre-seismic anomaly was observed about three days before two nearby large earthquakes that occurred at almost the same place in a time interval of 53 min (February 6, 1987; M = 6.4 and M = 6.7, D = 130 km).