Soil Dynamics and Earthquake Engineering 53 (2013) 240–253

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

Maps of soil subsidence for Tokyo bay shore areas liquefied in the March 11th, 2011 off the Pacific Coast of Tohoku Earthquake$

Kazuo Konagai a,n, Takashi Kiyota a, Shota Suyama a, Toru Asakura b, Kenichi Shibuya c, Chikako Eto c a Institute of Industrial Science, University of Tokyo, Tokyo 153-8505, Japan b CTI Engineering International Co. Ltd., Tokyo 136-0071, Japan c Aero Asahi Corp., Saitama 350-1165, Japan article info abstract

Article history: The March 11th, 2011 Off the Pacific Coast of Tohoku Earthquake, also known as the Great East Japan Received 2 October 2012 Earthquake, has shown that a long stretch of landfills along northeastern shorelines of the Tokyo Bay had Received in revised form very high susceptibility to liquefaction, causing concerns about re-liquefactions of the area in the 21 June 2013 scenario earthquake expected in the capital's metropolitan area. An attempt was made to detect soil Accepted 25 June 2013 subsidence from raster images converted from airborne LiDAR (Light Detection and Ranging) data before Available online 13 August 2013 and after the earthquake. To eliminate deep-seated tectonic displacements and systematic errors of Keywords: LiDAR surveys, the template matching technique is used for clusters of pile-supported buildings and Liquefaction bridge piers chosen as templates in source images of the target areas. The obtained subsidence maps Great East Japan earthquake of 2011 describe the spatial distribution of soil subsidence in great detail. Soil subsidence & 2013 The Authors. Published by Elsevier Ltd. All rights reserved. Tokyo bay area

1. Introduction inundations inside levees. According to the census data, the population of Chiba Prefecture has fallen by 7724 residents since Tokyo Bay Area, which was formerly used for seaweed collec- the beginning of 2011 for the first time since record keeping began tion, fishing, and resorts, is now a part of the great Tokyo– in 1920 [4]. Cases of modern liquefaction illustrate that soil Yokohama–Chiba metropolitan area. By the mid-1970s, production subsidence caused by sand liquefaction can cause long-lasting of heavy metals and chemicals in the eastern bay area was the problems. In the July 16th, 1990 Luzon Earthquake for example, highest among Japan's industrial regions [1]. Intensive flow of the city center of Dagupan along the meandering river trace of the people and goods has made the Tokyo Bay Area a world renowned Pantal was one of the most seriously liquefied areas, where commercial and consumption hub. All the more because the area drainage systems were clogged up by the accumulated sand has been the center of the economy and urban lives, the impact of causing temporary flooding. Some RC buildings along Pelez Bou- the sand liquefaction over a long stretch of landfills along the coast levard have sunk in the liquefied sand with their surrounding soils of the Tokyo Bay in the March 11th, 2011 Off the Pacific Coast of and remained underwater for several months [5]. In the 1964 Tohoku Earthquake was serious, leaving many houses and power Niigata Earthquake, similar problems were reported. About a poles tilting and lifelines cut off [2]. After almost all sands were 640 m long stretch of Akashi Avenue has subsided due to liquefac- cleared up for rehabilitation, clear differences of level between tion and remained about 60 cm underwater. The area along Tsusen ground floors of pile-supported RC buildings and surrounding River was flooded by tsunami, and remained underwater for about sidewalks were noted. The liquefied areas along the coast of the a month [6]. These areas have been frequently inundated in heavy Tokyo Bay reportedly reached 42 km2 [3], and there yet remain rains since then. One of the more notorious of these floods was the serious concerns about sewage treatment and possible record rainfall on August 4th, 1988, which later led the Niigata Prefectural Government into the construction of a new drainage pump station at Yamanoshita lockage [7]. ☆This is an open-access article distributed under the terms of the Creative In addition, what should not be forgotten is that the liquefied Commons Attribution-NonCommercial-No Derivative Works License, which per- areas are considered to be at high risk of re-liquefaction. Waka- mits non-commercial use, distribution, and reproduction in any medium, provided matsu has confirmed that sand deposits, which were once lique- the original author and source are credited. n fied in past earthquakes, did liquefy again in the March 11th Corresponding author. Tel.: +81354526142, +81453394034; fax: +81354526144. Earthquake in at least 145 municipalities in both Kanto and E-mail addresses: [email protected], [email protected] (K. Konagai). Tohoku Regions, including Urayasu City [8]. The Earthquake

0267-7261/$ - see front matter & 2013 The Authors. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.soildyn.2013.06.012 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 241

Research Institute of the University of Tokyo predicted that there correcting positioning errors [13]. The laser scanner emits fast would be a 70% probability that the capital’s metropolitan area pulses from a focused infrared laser, which are beamed toward the would experience a magnitude-7 quake within 4 years and a 98% ground surface with an oscillating mirror for fast scanning in a probability within the next 30 years [9]. sinusoidal pattern. The kinematic GPS measures the spatial posi- With all mentioned above, it is very important for relevant tion of the platform aircraft, while the IMU records the pitch, roll, organizations to have quantitative information on the soil sub- and heading of the aircraft. sidence to cope with post-earthquake problems. Raster images The obtained high-resolution digital elevation maps (Digital converted from airborne Light Detection and Ranging (LiDAR) Surface Models: DSMs) before the earthquake (in December data, namely digital surface models (DSMs hereafter), were 2006–January, 2007 for the entire target areas) and after the obtained on April 20th, September 6th and November 27th, 2011 earthquake (on April 20th, 2011 for Urayasu, September 6th for the for Urayasu, Funabashi–to–Chiba and Ichikawa areas, respectively, area from Funabashi to Chiba, and November 27th for Ichikawa) and an attempt was made to compare the images with those before the earthquake [10], [11], [12]. This paper describes the Table 1 overall features of soil subsidence in all target areas. Spatial resolutions of DSMs at different times and areas.

Area Date of LiDAR survey Spatial resolution 2 2. Detection of soil subsidence from LiDAR images (points/m )

Urayasu December 2006 to January 2007 0.792 A Light Detection and Ranging (LiDAR) system is capable of Urayasu April 20th, 2011 4.089 rapid and accurate collection of topographic and elevation data. It Ichikawa-to- December 2006 to January 2007 0.792 consists of (1) a laser scanner, (2) a kinematic airborne Global Chiba Ichikawa-to- September 6th, 2011 and Nov. 27th, 1.786 Positioning System (GPS), (3) an interfaced Inertial Measurement Chiba 2011 Unit (IMU), and (4) a fixed, ground-based reference GPS station for

Ichikawa area, Nov. 27, 2011

Funabashi to Chiba area, Sept. 6, 2011

Urayasu area, April 20, 2011

DSMs before the earthquake were Tokyo Bay not available below this line.

Ichikawa area Funabashi to Nov. 27, 2011 Chiba area, Sept. 6, 2011

DSMs before the earthquake were not available below this line.

Fig. 1. DSMs obtained for the Tokyo bay shore area. (a) Air-borne LiDAR surveyed areas (b) Urayasu area, April 20, 2011 and (c) DSMs of Ichikawa area (November 27, 2011) and Funabashi to Chiba area (September 6, 2011) 242 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 are raster graphic images of pixels having information on their radius of 3 m is drawn around it, and elevations of pixels within elevations (Fig. 1). These images have different spatial pixel this circle are averaged for the representative value of the point's densities as shown in Table 1 for different areas and times, elevation (Fig. 2). depending on the safe flight altitudes allowed for the aircraft to As for the Urayasu area, the least-square regression plane fly near the Haneda and Narita International Airports. The increase shows no significant inclination with the standard error of in flight altitudes can also yield an increase of errors because the estimate, 0.017 m, and the cloud of template points is about laser-light beam can spread transversely as it propagates. The 0.2 m lower in average than that before the earthquake. Regres- arrangement of satellites in the sky also affects the accuracy of GPS sion planes for Ichikawa and Funabashi–to–Chiba areas clearly positioning. When all the visible satellites are clustered close slope down in northwest and southwest with the standard errors together within a comparatively restricted part of the sky, trian- of estimate, 0.0259 m and 0.0456 m, respectively (Fig. 3(a) gulation calculations are susceptible to errors, resulting in a poorer and (b)). These standard errors for Ichikawa and Funabashi–to– position solution [13]. This effect, caused by the geometry of the Chiba areas are a little larger than that for Urayasu probably satellite arrangement, happens from time to time on any spot of because of the higher flight altitudes for the LiDAR surveys, and the earth surface, and therefore corrections are necessary at each this fact limits discussions that require much higher accuracy. time we survey. In addition to the above-mentioned system- Comparing DSMs at different times in a straight-forward correlated errors, the effect of deep-seated tectonic deformations manner only allows us to detect displacements in the Eulerian caused by this earthquake was remarkable over the entire stretch description, in which the description of motion is made in terms of of the Pacific coastal areas of the eastern Japan [14]. the spatial coordinates which does not follow the motion of a However liquefaction-induced shallow subsoil subsidence is particular target. Therefore not only vertical components but also considered to be simply estimated with reference to elevations of lateral components of Lagrangian displacements are to be elimi- top ends of pile-supported buildings and bridge piers such that nated especially when they are due to tectonic movements. One any horizontal and vertical biases can be eliminated. Therefore, to practical method to deal with this problem is to detect edges of find the best matching depth for DSMs before and after the pile-supported buildings where elevation changes sharply, and to earthquake, a template matching technique is used for three air- keep tracking the motions of the detected edges [16]. However, as borne LiDAR surveyed areas with clusters of pile-supported described previously, the DSMs have a spatial resolution of 1.7 to structures as templates [15]. 4 pixels m2, which is a little too sparse for sharp edge detection. For each air-borne LiDAR surveyed area, a flat regression plane After the 2004 Mid-Niigata Prefecture Earthquake, Konagai et al. is fit to a three-dimensional plot of the template points in such a [17], [18] estimated the Lagrangian components of tectonic dis- way that the sum of squared errors will be minimized. Total 191 placement induced within a mountainous terrain of Yamakoshi points on flat roofs of 57 buildings and 35 piers for expressway Village, by assuming that tectonic displacement varies gently in viaducts were chosen as template points for all three areas. Ideally, space, and therefore three adjacent pixels of DSM would have the all chosen structures are desirable to be resting on end-bearing same Lagrangian displacements. The same method is applied here piles. Among the 92 chosen structures, it was confirmed that 53 for detecting lateral Lagrangian displacements of roofs with structures were on end-bearing piles. They included RC public sloping surfaces towards walls. As illustrated in Fig. 4, several facilities, schools in Urayasu (see Appendix A) and piers of viaducts cross-sections of a roof are taken first, and after excluding those for the Bay-Shore Route of Tokyo Metropolitan Expressway Co. with outshooting objects, they are averaged for the representative Ltd., which expressway continues as the bay-shore section of the roof shape with two sloping surfaces towards walls on both sides. Higashi-Kanto Express way of the East Nippon Expressway Com- If two points on the two sloping surfaces of a roof undergo a rigid- pany. Though detail dimensions of piles for the other 39 structures body-translation movement fΔy ΔzgT , their Lagrangian compo- were not known at the time of template matching, it was an nents fΔy ΔzgT can be obtained by solving the following simulta- imminent need to supplement areas with these structures where neous equations. template points were rather sparsely distributed. Therefore, sup- ()"# δ plemental points were taken on medium to high-rise RC tower z1 a1 1 Δy ¼ ð1Þ condominiums (4 to several tens stories in height), which were δz2 a2 1 Δz considered to be almost certainly on end-bearing piles, and thus fi T appropriate for further enhancing the statistical signi cance of where fgδz1 δz2 are Eulerian displacements at these two points. each regression plane. At each template point, a circle with a To obtain two orthogonal components of lateral displacement, one

Fig. 2. Template points on a pile-supported RC building. K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 243

more roof with its ridge in a different orientation is to be found Regression plane elevation (m) = -1.93+0.0000319x-0.000044y nearby. It is also noted that even a pile-supported building may with x= EW distance (m), not be an appropriate target for lateral template matching, y =NS distance (m) and because piles are laterally flexible enough to be easily deformed Standard error of estimate = 0.0259m by the movements of the surrounding side soils. Due to the above- mentioned constraints, only 5, 4 and 3 buildings were found as – – 0.1 appropriate templates in Urayasu, Ichikawa and Funabashi to Change in elevation (m) Chiba areas, respectively; each building must have at least two roof ridges in different orientations. However they are clustered 0.0 only on stable lands, mostly on the inland side of the liquefied coastal landfills, and fitting flat regression planes to three- -0.1 dimensional plots of these clustered buildings would cause serious extrapolation errors for the greater part of the target areas. -0.2 Therefore the extracted lateral components of Lagrangian displa- cements of the buildings were simply averaged over each target area, and the corresponding template DSM after the earthquake -0.3 was aligned to the DSM before the earthquake to eliminate the extracted lateral anomalies. These anomalies as well as their Ichikawa standard deviations are not as reliable as they should be because the numbers of samples are too low to expect stochastic conver- gence. However both the anomalies and their standard deviations for areas with spatial resolutions of 4.089–1.786 points/m2 are at least much smaller than the corresponding distances between neighboring pixels, and thus can be smaller than the errors expected from the edge detection method.

Regression plane elevation (m) = 0.342 + 0.0000123x+0.0000107y with x= EW distance (m), 3. Obtained map of soil subsidence y=NS distance (m) and Standarderror of estimate = 0.0456m 3.1. Urayasu area

Fig. 5 shows the obtained soil subsidence map of Urayasu. Since 0.4

Change in elevation (m) the LiDAR survey was conducted for Urayasu on April 20th after

0.3 almost all sand ejecta were cleared up from streets, the amount of ground elevation loss is considered to be largely due to the 0.2 removal of sand ejecta. The aerial photograph of the same area on the upper right of the figure was taken in 1948 by the US Army 0.1 [19], clearly showing that it was post-World War II when the majority of land reclamation in Urayasu was undertaken, and that 0.0 Funabashi today, the greater part of the city spreads over the reclaimed land of sand dredged from the Tokyo Bay. -0.1 On this map, subsidence can be seen over the entire stretch of the reclaimed land. In particular, remarkable subsidence is seen where the aerial photo of 1948 shows darker black indicating the presence of deeper water. On the other hand, on the long-existing natural land of Urayasu before the time of landfills, no or slight Chiba subsidence less than 0.1 m is seen. Zooming in on one of residential areas, Benten, a blue brush of subsided houses is found running across several town blocks as indicated with arrows in Fig. 6(a). In these residential town blocks, Fig. 3. Regression planes fit to a three-dimensional plot of the clouds of template where strip footings and mat foundations are the most common, points for (a) Ichikawa and (b) Funabashi to Chiba areas, respectively, on the Japanese National Grid System, Zone IX (see Appendix A). houses have sunk in the soil by about 0.5 m in average (Fig. 7(a)).

Fig. 4. Template building with sloping roofs. 244 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253

Coast in 1948 Urayasu in 1948, photo by US Army

Coast in 2011

Fig. 6(a)

-4000m

N

5000m 10000m

Fig. 5. Soil subsidence map of Urayasu.

The southwestern extension of the brush appeared as sand blows Slightly raised spots are found scattering particularly along lined up in the school ground of Miakegawa Primary School (Fig. 7(b)). the western coastal area of Makuhari as shown in Fig. 11(a). The Later, on April 24th, there appeared a puddle as shown in Fig. 7(c) due greater part of this area is on pre-development stage with some to a heavy rain of April 23rd, 2011. buildings sparsely distributed. Some raised spots are due to land- development works, and some are damp yards of ejected sand. However, when these elevated spots are compared with the 3.2. Ichikawa–to–Chiba areas micro-topography of the area (Fig. 11(b)), they are often found in shallow dips or hollows among some mounds, indicating that Fig. 8 shows the obtained soil subsidence map of Ichikawa- surrounding mounds have slumped down under their own to-Chiba areas. An about a 1 km to 2 km wide stripe of soil weight causing their neighboring dips and hollows to be subsidence extends along the Tokyo Bay. On this stripe, spots of pushed up. serious subsidence are found clustered in Makuhari area of Chiba City (Fig. 9). These spots included Makuhari–Kaihin Park and the areas in front of JR Kaihin–Makuhari Station, where manholes 4. Verification of the method and discussions were found sticking out (Fig. 10(a)) and buildings next to the pile supported JR Kaihin–Makuhari Station fell forward (Fig. 10(b)). The small standard error of estimate for each regression plane It is noticed that these spots are roughly lined up along the does not necessarily guarantee the same level of accuracy for the foreshore front of an old tidal flat (broken line in Fig. 9(a)), which extracted soil subsidence. Therefore, differences in level between was once shown on the map compiled by the Japanese Army in ground floors of pile-supported RC buildings and surrounding 1877 [20]. Sand had been accumulating in this tidal flat since then, sunken sidewalks were measured at 13 and 16 locations in and it is seen in the aerial photo of 1972 (Fig. 9(a), [19])asa1km Urayasu and Makuhari respectively (Fig. 12), and compared with wide light-gray brush; the color is indicating that the brush is a those from the obtained subsidence maps. shallow intertidal zone. The first landfilling project for the eastern A laser level was put on the ground floor of each building to area of Makuhari was completed in 1964 as shown in Fig. 9(a), and project a laser line to help provide accurate readings of the followed by the 2nd three-years landfilling project for the west measuring rod at a distance on the sunken ground. Meanwhile, Makuhari new city complex in 1973. Fig. 9(b) [19] shows the aerial three pixels were taken from the subsidence map within a 1 m- photo of the new-city complex in 2006 with the original coast line radius circle around each measured location, and the three values of a few kilometers inland. estimated soil subsidence were averaged for each location. Fig. 13 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 245

Distance (m) 0 200 400

Fig. 7(a)

Fig. 6(b) 40000m

Change in elevation (m) -0.7-0.5 -0.25 0.0 0.25 0.5 0.7 5000 m

Distance (m) 0 100 200

Fig. 7(c) Fig. 7(b)

Change in elevation (m) -0.7 -0.5 -0.25 0.0 0.25 0.5 0.7

Fig. 6. Stripe of soil subsidence across Benten residential area. (a) Location of stripe (b) Southwest extension of the soil subsidence stripe appeared in a school-ground and (c) Benten area in 1979: The area was developed in late 1970s. Geospatial Information Authority of Japan (2011b).

confirms that the overall trend of the actual soil subsidence is in In this measurement, references were taken on the ground good agreement with the trend that appeared in the subsidence floors of the pile supported buildings, and thus the measured map with the mean value errors and the standard errors of estimate values of subsidence may be affected by the slightly fluctuating being 0.00467 m and 0.04206 m for Urayasu and 0.00203 m levels of the ground floors. Geospatial Information Authority of and 0.0282 m for Makuhari, respectively. Japan (GSI) measured elevations of several points along the main 246 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253

Fig. 7. Extension of the subsidence stripe appeared in a school ground of Miakegawa Primary School. (a) Inclined and sunken houses Photo by Kiyota T., at 35.641895, 139.89886 (b) Sand blows lined up in school ground Photo by Kiyota T., on March 19th at 35.638933, 139.895619 and (c) Puddle appeared in school ground Photo by Konagai K., on April 24th at 35.639117, 139.895471.

street (Orange place marks along Prefectural Routes 242 and 270 lessons offer strategically important hints for quick investiga- in Fig. 12, [21]) with the reference taken at N35.6608519273881, tions that can be necessary in the expected scenario earth- E 139.901017021075 in the old city of Urayasu where no clear quakes in the capital's metropolitan area [9] and Tokai/Tonankai evidence of liquefaction was found. regions. Fig. 14 compares GSI measured values of soil subsidence and those from the obtained subsidence map. The values of the actual soil subsidence are generally in good agreement with those 5. Conclusions appeared in the subsidence map with the mean value error and the standard error of estimate being 0.0139 m and 0.04391 m, The March 11th, 2011 East-Japan Earthquake has caused sand respectively. Eventually, Figs. 13(a) and 14 showed similar stochas- liquefaction over a long stretch of landfills along the coast of the tic natures of error with each other, supporting the premise that Tokyo Bay. The liquefied areas along the coast of the Tokyo Bay pile-supported buildings make up a point cloud for aligning DSM reportedly reached 42 km2, and there yet remain serious long- images to extract soil subsidence. In other words, the soil sub- lasting concerns about sewage treatment and possible inunda- sidence map can be prepared only when pile-supported buildings tions inside levees, etc. An attempt was made to measure and/or their alternatives are available, and the quality of the liquefied soil subsidence with reference to elevations of top ends obtained map largely depends on how they are distributed within of pile-supported buildings and bridge piers. Two sets of Digital the target area. In so far as Bay–Shore area goes, the overall Surface Models (DSMs) before and after the earthquake were standard error of estimate will be comparable to about 0.05 m. compared with pile-supported RC buildings and bridge piers as Nonetheless, the proposed method is considered important templates for aligning the two sets, and examples of the analyzed in that it records in time a precise instance of liquefied soil images were shown for Urayasu and Ichikawa–to–Chiba areas. On subsidence, which often exceeds several tens of centimeters as the obtained map of Urayasu, subsidence can be seen over the was shown above. The authors, thus, had made a plan to cover entire stretch of the reclaimed land. However, the severity of quickly more areas liquefied in the March 11th Earthquake. liquefaction within this zone was not uniform. In particular, However, it turned out that all aircrafts for LiDAR surveys had remarkable subsidence is seen where the aerial photo of 1948 been booked up in the large post-earthquake-time procurement shows darker black color indicating the presence of deep water at boom. Almost all available small aircrafts were used in not only that time. The obtained map of the Ichikawa–to–Chiba area government-initiated surveys but also in investigation/recovery shows an about 1 km to 2 km wide stripe of sunken area works of electric power transmission lines, particularly in the extending all along the Tokyo Bay shore. The zone of more situation generated by the serious accident which happened at pronounced liquefaction appeared in Makuhari to somewhat line the Fukushima–Daiichi Nuclear Power Plant [22]. It had also up with the foreshore front of an old tidal flat once shown on the made the situation worse that 67 small aircrafts and helicopters map compiled by the Japanese Army in 1877. These facts may including those of aerial survey companies were washed away suggest that soil subsidence has a strong correlation with the by the tsunami at the Sendai International Airport [23].Allthese thickness of the liquefied layer. K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 247

Chiba City

20000m

Makuhari (Fig. 9)

Narashino City Western Makuhari (Fig. 11)

15000m

Funabashi City

Bay-shore Route of Tokyo Metropolitan Expressway

Edogawa Diversion Channel

10000m

N

Ichikawa City Urayasu City (Fig. 5)

Change in elevation 35000m 40000m

Fig. 8. Soil subsidence map of Ichikawa-to-Chiba area. 248 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253

Intertidal zone

Foreshore front of an old tidal flat in 1877 [19]

Photos: Geospatial Information Authority of Japan

JR Kaihin-Makuhari Station

Makuhari Kaihin Park

Fig. 10(b) N

Fig. 10(a)

0 1km

Distance

-0.7 -0.5 -0.25 0.0 0.25 0. 5 0. 7

Change in elevation (m)

Fig. 9. Soil subsidence map of Makuhari. (a) 1972, (b) 2006 and (c) Subsidence in 2011.

This side fell forward JR Kaihin Makuhari St.

Fig. 10. (a) Manhole sticking out of sidewalk Location: 35.646337,140.040413 and (b) a two-story building falling forward Location 35.648773,140.04176. (Photos by Kazuo Konagai on July 10th, 2011). K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 249

Change in Elevation elevation (m) (m) 0.7 10.0 0.5 7.5 0.25 5.0 0.0

-0.25 2.5

-0.5 0.0

-0.7 -2.0 0 1km

Fig. 11. Soil subsidence and micro-topography along shore line of the western Makuhari. (a) Soil subsidence and (b) Micro-topography (Contour lines at every 0.2 m interval).

Pile-supported buildings 1 Points for leveling 3 Change in 2 4 N elevation 5 (m) 6 0.7 78 8 0.5 1-3 99 4-6 7 10 0.25 11 10-11 12-13 0.0 JR Keiyo Line 12 13 -0.25 -40000 m -0.5 -0.7 Urayasu City

Distance (m)

5000 m 10000 m

4 Change in 3 elevation (m) 2 N 0.7 10 0.5 5 7 0.25 6 8 0.0 1 9 -0.25 -0.5 11 -0.7 13 12 15 14 Pile-supported 16 buildings Distance (m) -4000 m

20000 m

Fig. 12. Locations where differences in level between ground floors of pile-supported RC buildings and surrounding sunken sidewalks were measured (a) Urayasu City and (b) Makuhari area in Chiba City. 250 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253

Mean value errors = -0.00467m Mean value errors = -0.00203m Standard error of estimate = 0.04206m Standard error of estimate = 0.0282m

Building Number Building Number

Subsidence (m) Subsidence (m)

Fig. 13. Differences in level between ground floors of pile-supported RC buildings and surrounding sunken sidewalks (Yellow place marks in Fig. 11). (a) Urayasu and (b) Makuhari.

Mean value errors = -0.00139 m Science, University of Tokyo, who has kindly offered professional Standard error of estimate = 0.04391 m advice on the use of LiDAR data.

Point Number Appendix A. End-bearing piles in Urayasu and along Tokyo Bay-shore line

Urayasu city published a list of pile-supported facilities on web (Table A1,after[24]). The list was one of the materials for the Liquefaction Countermeasure Task Force of Urayasu City. The list says that pile lengths vary from 30 m to 50 m. RC school buildings higher than 3 stories in height are all on end-bearing piles. Fig. A1 shows soil profile along JR Keiyo Railway and reference points for template matching close by this railway. Numbers put next to the reference points show corresponding pile lengths. The lengths of piles within an about street block distance (about 200 m) from this railway are Subsidence (m) projected upon the subsurface soil cross-section. Liquefaction Coun- termeasure Task Force of Urayasu City [24] has suggested that fills Fig. 14. GSI measured values of soil subsidence along Prefectural Routes 242 and 270 (Blue place marks in Fig. 11) and those from the obtained subsidence map. (F) and shallow parts of the underlying alluvial sandy soil deposit (As) could have been liquefied. Tokimatsu et al. [25] also examined liquefaction safety factors (FL) at various locations in Urayasu. In In this manner, the proposed method records in time a precise Mihama–Irifune, Takasu and Akemi–Hinode districts in particular, instance of liquefied soil subsidence, which often exceeds several the examined FL values in layers shallower than 20 m were below tens of centimeters as was shown above. In so far as Bay–Shore 1.0 supporting the view of the taskforce. Beyond 20 m, the settlement area goes, the overall standard error of estimate was comparable was unlikely to occur. Therefore, the embedment depths of piles for to about 0.05 m. The error is expected to be further minimized for the chosen reference points are considered to be substantially large multifactorial analyses of soil subsidence. enough not to allow them to sink. Piers for highway viaducts constructed since 1970s rest essen- tially on end-bearing piles in conformity to both the Design Guide of 1964 for Road Bridge Foundations [26], and the Seismic Design Acknowledgments Guide of Foundations for Road Bridges of 1972[27], the latter reflects strongly lessons from the Niigata Earthquake of 1964 in The authors are indebted to both the Foundation for the which many friction-pile supported bridges collapsed seriously in Promotion of Industrial Science and the Japan Society for the liquefied soils. The Metropolitan Expressway Company Ltd. took a Promotion of Science who have provided the authors with financial more thorough countermeasure for its Bay-shore line to confine supports for conducting the airborne LiDAR surveys over Ichikawa, loose liquefiable sands by installing consecutive underground steel- Funabashi and Mihama of the Chiba. The authors'deepest apprecia- pipe wall around and near the existing pile foundations [28].The tion also goes to Mr. Horii, Public Sewerage Division of Urayasu City, installation of these underground walls completed in 1999. Mr. Shigeru Hataya, Waterworks Department of Chiba Prefecture, Mr. Toshihiro Yokota, Water Quality Control Department of the National Institute for Land and Infrastructure Management, Minis- Appendix B. Geographic reference system try of Land, Infrastructure and Transport, who have provided the authors with valuable pieces of information about the damage to The soil subsidence maps were prepared on the Japanese underground sewage systems. Special thanks of authors go to Dr. National Grid System. The Japanese National Grid System Wataru Takeuchi, Associate Professor at the Institute of Industrial divides Japan into a set of 19 zones assigned with Greek K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 251

Table A1 Template points on end-bearing pile supported buildings in Urayasu (after [24]).

Template point Building Piles

ID North latitude East longitude ID Building name Friction/ End- Type Diameter Number Length (deg.) (deg.) bearing (mm) (m)

1 35.66374 139.8939 1 Urayasu City Library end-bearing N/A(existing) PHC piles 600 (New) 135 36-47 2 35.66371 139.8942 (New) 3 35.66369 139.8943 4 35.66366 139.8945 5 35.66603 139.9004 2 Hokubu Primary School end-bearing AHS piles 500 140 29 6 35.66597 139.9006 7 35.66587 139.9007 8 35.6658 139.9008 9 35.65539 139.8914 3 Minami Primary school end-bearing Steel piles 500 154 36-40 10 35.65532 139.8914 11 35.65477 139.8925 12 35.65609 139.9022 4 Higashi Primary School end-bearing Steel piles 600 162 43 13 35.65618 139.9023 14 35.65629 139.9027 15 35.65624 139.9028 16 35.65062 139.9006 5 Higashino Primary School end-bearing PHC/SC piles 600-1000 139 44 17 35.65057 139.9007 18 35.65044 139.9008 19 35.65039 139.901 20 35.64869 139.9046 6 Tomioka Primary School end-bearing Steel piles 500 180 50 21 35.64875 139.9046 22 35.64885 139.905 23 35.64876 139.9051 24 35.64867 139.9052 25 35.63946 139.8954 7 Miakegawa Primary School end-bearing Steel piles 600 182 49-50 26 35.63943 139.8955 27 35.63937 139.8956 28 35.6555 139.9148 8 Mihamakita Primary School end-bearing AC piles 600 91 40-43 29 35.65544 139.9149 30 35.65515 139.9152 31 35.6551 139.9152 32 35.65146 139.9076 9 Mihamaminami Primary end-bearing Steel piles infilled with 600 158 46-51 33 35.65141 139.9077 School concrete 34 35.65128 139.9078 35 35.6512 139.908 36 35.651 139.9082 37 35.65419 139.9191 10 Irifunekita Primary School end-bearing Steel piles infilled with 500, 600 192 45-52 38 35.654 139.9194 concrete 39 35.65382 139.9196 40 35.64565 139.9125 11 Irifuneminami Primary end-bearing SC/AC piles 600 216 52-55 41 35.64561 139.9127 School 42 35.64554 139.9127 43 35.6454 139.9129 44 35.64212 139.8854 12 Maihama Primary School end-bearing PHC piles 600 192 39-44 45 35.64203 139.8855 46 35.64195 139.8856 47 35.64169 139.8858 48 35.63601 139.9166 13 Takasu Primary School end-bearing PHC/SC piles 500, 600, 700 192 51-52 49 35.63573 139.9169 50 35.64512 139.9202 14 Meikai University end-bearing N/A N/A N/A N/A 51 35.64512 139.9206 52 35.64799 139.9139 15 Tower East end-bearing N/A N/A N/A N/A 53 35.64791 139.914 54 35.64795 139.9141 55 35.64795 139.9139 56 35.65362 139.9015 16 Urayasu City cultural Hall end-bearing SPC/Ac piles 600 400 32-42 57 35.65346 139.9016 58 35.65395 139.9017 59 35.65368 139.9018 60 35.65182 139.8998 17 Urayasu Post Office end-bearing N/A N/A N/A N/A 61 35.65159 139.9003 62 35.64494 139.8917 18 Urayasu High School end-bearing N/A N/A N/A N/A 63 35.64467 139.892 64 35.64505 139.8924 numerals from I to XIX in principle in a row-by-row pattern Appendix C. Distribution of soil subsidence map images starting from the zone at the southwest corner. Exceptions (from XIV to XIX) are for isolated islands. The surveyed Tokyo The soil subsidence maps prepared here are one of the results Bay area is included in Zone IX with its origin located at of a cooperative survey between Konagai Laboratory, Institute of 139150’E, 36100'N. Industrial Science, University of Tokyo and Aero Asahi Corporation. 252 K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253

29m 36 - 47m

40 - 43m 32 - 42m 42 - 52m 36 - 40m 43m NA 46 - 51m

44m 50m NA NA 52 - 55m

39 - 44m

49 -50m

Fig. A1. Soil profile along JR Keiyo Railway and reference points for template matching close by the railway.

Aero Asahi Corporation retains copyright control over Digital [3] Yasuda S, Harada K. Liquefaction-induced damage in the reclaimed lands along Surface Models as their works. Though the authors retain copy- Tokyo Bay. Japan Geotechnical Society 2011;7:38–41 in Japanese. right control over the soil subsidence maps, the map images are [4] Chiba Prefecture. Prefecture Government-appointed study committee on the change in population of Chiba Prefecture; 2011 (in Japanese). Available from: openly available, in view of their merits for informational, aca- 〈http://www.pref.chiba.lg.jp/toukei/press/2012/24jinkoukaigi.html〉. demic, public interest and noncommercial purposes, as Keyhole [5] Yanagisawa E, Iai S, Wakamatsu K, Yamazaki J, Yamazaki F. Damage in Dagupan Markup Language Zipped (kmz) place-mark files at the following area. Reconnaissance report on the July 16, 1990 Luzon Earthquake, the Philippines, – website: Japan Society of Civil Engineers; 1993; Series 1 p. 101 24 (in Japanese). [6] Hokuriku Regional Development Bureau. Geo-hazard in Niigata, Ministry of http://shake.iis.u-tokyo.ac.jp/home-new/ Land, Infrastructure and Transport; 2008 (in Japanese). Available from: 〈http:// www.hrr.mlit.go.jp/bosai/hokurikunobosai/jisin/jisinindex.html〉. [7] Niigata Regional Development Bureau. Yamanoshita lockage drainage pump Appendix D. Supporting information station. Niigata Prefecture; 2007. Available from: 〈http://www.pref.niigata.lg. jp/niigata_seibi/1201107670226.html〉) (in Japanese). [8] Wakamatsu K. Geotechnical aspects of the damage caused by the March 11th Supplementary data associated with this article can be found in off the Pacific Coast of Tohoku Earthquake, 11th Seminar on earthquake theonlineversionathttp://dx.doi.org/10.1016/j.soildyn.2013.06.012. disaster management. Japan Society of Civil Engineers; 2011 (in Japanese). [9] Outreach and public relations office. Seismic activity changes in capital region since 3.11. Earthquake Research Institute, University of Tokyo; 2011 Available from 〈http://outreach.eri.u-tokyo.ac.jp/eqvolc/201103_tohoku/shutoseis/capi 〉 References talregion/ (in Japanese). [10] Konagai K, Shibuya K, Eto C, Kiyota T. Map of soil subsidence in Urayasu, Chiba, caused by the March 11th 2011 East-Japan Earthquake. Bulletin of Earthquake [1] Britannica academic edition. Keiyo Industrial Zone 〈http://www.britannica. Resistant Structure Research Center, IIS, University of Tokyo. vol. 44; 2012 p. com/EBchecked/topic/314276/Keiyo-Industrial-Zone/%C5%8D-Industrial- 45–8. Zone〉;2012. [11] Konagai Laboratory. Soil subsidence map of Urayasu liquefied in the March [2] Bhattacharya S, Hyodo M, Goda K, Tazoh T, Taylor CA. Liquefaction of soil in 11th East Japan Earthquake. Konagai Laboratory, Institute of Industrial Science, the Tokyo Bay area from the 2011 Tohoku (Japan) earthquake. Soil Dynamics University of Tokyo; 2011. Available from: 〈http://shake.iis.u-tokyo.ac.jp/ and Earthquake Engineering 2011;31:1618–28. home-new/〉. K. Konagai et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 240–253 253

[12] Konagai K. Soil subsidence map of Tokyo Bay Area liquefied in the March 11th, Transition Map; 2012. Available from: 〈http://habs.dc.affrc.go.jp/index.html〉). 2011 Great East Japan Earthquake. In: Proceedings of the International (in Japanese). symposium on engineering lessons learned from the 2011 Great East Japan [21] Geospatial Information Authority of Japan. Liquefaction in the Tokyo Bay Area – – Earthquake March 1 4 2012, Tokyo, Japan; 2012, p. 855 64. Available from: and measured subsidence in Urayasu. Provisional report no. 1. Geospatial 〈 〉 http://www.jaee.gr.jp/event/seminar2012/eqsympo/proceedings.html . Information Authority of Japan; 2011 Available from: 〈http://www.gsi.go.jp/ [13] National Oceanic and Atmospheric Administration (NOAA) Coastal Services common/000062445.pdf〉. (in Japanese). Center. Lidar 101: An introduction Lidar technology, data, and applications; [22] International Atomic Energy Agency. Fukushima nuclear accident update 2008. Available from: 〈http://www.csc.noaa.gov/digitalcoast/_/pdf/What_is_Li Log; 2011. Available from: 〈http://www.iaea.org/newscenter/news/tsunamiup dar.pdf〉. 〉 [14] Geospatial Information Authority of Japan. Crustal movement caused by the date01.html . March 11th 2011 off the Pacific Coast of Tohoku Earthquake detected through [23] Haneda Air Field. Press release, 3rd Regional Coast Guard Headquarters; continuous GPS monitoring; 2011. Available from: 〈http://www.gsi.go.jp/ March 29 2012. chibankansi/chikakukansi40005.html〉. (in Japanese). [24] Liquefaction Countermeasure Task Force of Urayasu City. Damage to buildings [15] Brunelli R. Template matching techniques in computer vision: Theory and and liquefaction countermeasures, Material no. 2-3-4 for the 2nd taskforce practice. Chichester, United Kingdom: Wiley; 2009 ISBN978-0-470-51706-2. meeting; 2011 (Available from: 〈http://www.city.urayasu.chiba.jp/secure/ [16] Lindeberg T. Edge detection and ridge detection with automatic scale selec- 26052/06_siryo2-4-3kenchikubutunohigai.pdf〉.(in Japanese). tion. International Journal of Computer Vision 1998;30(2):117–54. [25] Tokimatsu K, Tamura S, Suzuki H, Katsumata K. Damages to soils and [17] Konagai K, Fujita T, Ikeda T, Takatsu S. Tectonic deformation buildup in folded foundation, Chapter 10 in the preliminary reconnaissance report of the 2011 mountain terrains in the October 23, 2004, Mid-Niigata earthquake. Soil Tohoku-Chiho Taiheiyo-Oki Earthquake, Architectural Institute of Japan, – Dynamics and Earthquake Engineering 2009;29(2):261 7. Springer, ISBN 978-4-431-54097-7; 2012. [18] Konagai K. Earthquake-induced soil displacements and their impact on [26] Japan road association. Design guide of 1964 for road bridge foundations; rehabilitations. Proceedings of the Japan Academy, Series B 2011;69 1964. (in Japanese). (5):433–49. [19] Geospatial Information Authority of Japan. Archives of aerial photographs for [27] Japan road association. Seismic design guide of foundations for road bridges; tracing the history of land; 2011. Available from: 〈http://www.gsi.go.jp/ 1972. (in Japanese). tizu-kutyu.html〉. (in Japanese). [28] Metropolitan Expressway Company Ltd. Seismic countermeasures for piers; [20] National Agriculture and Food Research Organization. Rapid kanto map 2011. Available from: 〈http://www.shutoko.co.jp/efforts/prevention/earth originally compiled by the Japanese Army in 1877, in: Agricultural Land Use quake〉. [accessed on 28.4.13]. (in Japanese).