A real-time observation network of ocean-bottom-seismometers deployed at the Sagami trough subduction zone,central Japan

We installed a real-time operating regional observation network of Ocean-Bottom-Seismometers, connected to an electro-optical fiber communication cable, at the Sagami trough subduction zone, just south of the Tokyo metropolitan area, central Japan. The network, called ETMC, has six seismic observation sites at approximately 20 km spacing. In addition, there are three tsunami observation sites along the ETMC network to monitor the propagation process of tsunamis around the Sagami trough region.The on-line data from the ETMC has been improving the detection capability of smaller-magnitude earthquakes even at areas close to the margin of the trough. The ETMC data analyzing system, which has a function of real-time digital filtering for each seismic channel, can read the arrival times of P- and S-waves precisely, constraining well the automatic on-line hypocenter locations. The network has been providing useful information regarding the bending and downgoing process of the Philippine sea plate at the Sagami trough subduction zone.The pressure sensors of the installed network have a detection capability of tsunami wave trains with an amplitude of less than 1 cm. For example, the sensors recorded the full time history of tsunami wave trains, with mm order resolution, originating from a tsunami earthquake with 5.7 M_W and the tsunami magnitude of 7.5 occurred near Tori Shima (Tori Is.) of the Izu-Bonin Is. arc on September 4, 1996. The maximum amplitude of the tsunami signals on the trough-floor was approximately 1 cm (P-P), in contrast with approximately 20 cm (0-P) at a coastal site on Izu-Oshima, near the trough. Also, the pressure sensors observed tsunamis due to a large tsunami earthquake (7.1 M_W) at the northern New Guinea, on July 17, 1998.     

Estimation of geometric route distance from its topological distance: application to narrow road networks in Tokyo

The structure of road networks has been investigated in accordance with the development of GIScience. By classifying road networks into wide and narrow ones, we can define the route as the path from the route’s origin (also called the root) on a wide road network to a narrow road segment which consists of the sequence of narrow road segments arranged by ascending order of the number of steps of adjacency to its root. The length of the route can be defined with the following geometric and topological terms: the route distance, measuring the length along the route and the depth, counting the number of road segments on the route. The depth plays the important role of being a substitute for the route distance in modelling road networks as a planar graph. Since road networks clearly exhibit irregular patterns and road segment lengths are non-uniform, it is considered appropriate to adopt a stochastic approach rather than a deterministic one to analyse the route distance. However, the relationship between the route distance and its depth has not been sufficiently investigated stochastically. Therefore, the research question is how can we estimate the route distance from its depth? Based on an empirical study in the Tokyo metropolitan region, it was found that (1) the statistical distribution of the route distance can be formulated as an Erlang distribution whose parameters are its depth and the inverse of the mean length of narrow road segments, and (2) this length is constant and close to 40 m. Therefore, we can estimate the route distance from only one parameter, the depth. Also, as a practical application, accessibility to the kth depth link in terms of firefighting was evaluated because the maximum length of the extension of fire hoses is approximately 200 m. It was found that (1) even if k?≤?5, the probability that the route distance to the kth depth link is equal to or longer than 200 m ranges from 0 to 0.45; and (2) if k?≥?8, the probability is approximately 1. These indicate the limitation of the deterministic approach because, on the basis of complete grid patterns (with intervals of 40 m between intersections), k?=?5 corresponds to a distance of 200 m from wide road networks and the route to the 5th depth link can be covered with fire hoses. Moreover, it was found that the connectivity of wide road networks is higher than that of narrow ones in terms of the smaller ratio of cul-de-sacs and the larger ratio of four-way intersections. These answers contribute substantially not only to constructing a science of cities that provides a simple model and specifies the most important parameter, but also to our understanding of the structure of narrow road networks within several hundred metres of wide road networks.     

Seasonal transition of hydrological processes in a slow‐moving landslide in a snowy region

A comprehensive understanding of seasonal hydrological dynamics is required to describe the influence of pore‐water pressure on the stability of landslides in snowy regions. This study reports on the results of continuous meteorological and hydrological observations over 2 years on a landslide body comprising Neogene sedimentary rocks in northern Japan, where a thick (3–5 m) seasonal snowpack covers the land surface. Monitoring of the volumetric water content in shallow unsaturated zones (<0.8 m depth) and pore‐water pressure in saturated bedrock at depths of 2.0 and 5.2 m revealed clear seasonality in hydrological responses to rainfall and meltwater supply. During snow‐free periods, both the shallow soil moisture and deep pore‐water pressure responded rapidly to intense rainwater infiltration. In contrast, during snowmelt, the deep pore pressure fluctuated in accordance with the daily cycle of meltwater input, without notable changes in shallow moisture conditions. During occasional foehn events that cause intense snow melting in midwinter, meltwater flows preferentially through the layered snowpack, converging to produce a localized water supply at the ground surface. This episodically triggers a significant rise in pore‐water pressure. The seasonal differences in hydrological responses were characterized by a set of newly proposed indices for the magnitude and quickness of increases in the pressure head near the sliding surface. Under snow‐covered conditions, the magnitude of the pressure increase tends to be suppressed, probably owing to a reduction in infiltration caused by a seasonal decrease in the permeability of surface soils, and effective pore‐water drainage through the highly conductive colluvial layer. Deep groundwater flow within bedrock remained in a steady upwelling state, enhanced by increasing moisture in shallow soils under snow cover, reflecting the convergence of subsurface water from surrounding hillslopes.     

Direct measurements of mid-depth circulation in the Shikoku basin by tracking SOFAR floats

Mid-depth circulation of the Shikoku Basin was measured by tracking four SOFAR floats drifting at the 1,500 m layer. Two floats were released on 17 April 1988 at 30°N, 135°59E and tracked for 433 days. Another two were released on 3 November 1988 at 29°52N and 133°25E, and tracked for 234 days. Two floats flowed clockwise around the Shikoku Warm Water Mass with a diameter of 400 km centered at 31°N and 136°E and a mean drift speed of 4.5 cm sec^–1. One of the floats showed about ten counterclockwise rotations with a period of about 8 days and a maximum speed of 80 cm sec^–1 in the sea area west to the Izu Ridge. In the east to Kyushu, a southward flow was observed under the northward flowing Kuroshio. The southward flow of 4 cm sec^–1 drift speed was considered to be a part of the counterclockwise circulation at deep layers along the perimeter of the Shikoku Basin. One float remained for 234 days in a limited area of 100 km by 150 km in the western part of the basin.