Field Reconnaissance: 19 November 2004

Landslides: South Yamakoshi and East Ojiya

Ellen Rathje, Randy Jibson, Keith Kelson, James Bay, and Robert Pack

GEER Beyond Reconnaissance Team

On 19 November, the reconnaissance team investigated the landslides located south of Yamakoshi and two landslides on the edge of the Shinano River valley in east Ojiya. The team drove to the center of Yamakoshi and again parked at the Yamakoshi Village Office. Team members Randy Jibson, Keith Kelson and Ellen Rathje hiked from the village office to points south (Figure 1) along Highway 291. Team members Jim Bay and Bob Pack performed a LIDAR survey of a landslide / soil slump near the Yamakoshi Village Office and then traveled approximately 6 km west to the two landslides in east Ojiya along Highway 291. A LIDAR survey was performed on one of these landslides. Team members were accompanied by Prof. I. Towhata of the University of Tokyo and several students from the University of Tokyo and Tokyo Denki University.

Fig. 1. Satellite image of reconnaissance route along Highway 291 to investigate landslides (satellite image courtesy of DigitalGlobe, imagery acquired 24 October 2004).

Fig. 2. The area south of Yamakoshi consists of steep, narrow canyons in folded Pliocene and Pleistocene weakly cemented sandstone and mudstone. Fairly shallow slides of weathered regolith and colluvium were abundant.

Fig. 3. In many areas, large parts of the canyon walls were denuded by the strong shaking. The relatively thin layer of weathered soil was stripped from the slopes, leaving remnants of trees and shrubs in place.

Fig. 4. Many of the slopes above roads were covered with shotcrete and had snow fences, both of which slid along with the surficial slope material that failed during the strong shaking.

Fig. 5. Streams in canyon bottoms were dammed in several places by landslides from the canyon walls. In the relatively narrow canyon, slides of only a few thousand cubic meters could locally dam the streams.

Fig. 6. Slope material that was nearly saturated by antecedent rainfall mobilized into debris flows. Ongoing rain after the earthquake continued to mobilize debris flows in the disturbed material remaining on the slope.

Fig. 7. Above stream blockages, saturated sediment accumulated from upstream landslides.

Fig. 8. Failures of road fills were numerous and ranged from minor slumping of outboard slopes to general failure of the entire road surface.

Fig. 9. In some places, considerable lengths of road were disrupted by fill failures.

Fig. 10. In addition to shallow landslides, some deep earth flows were triggered by the earthquake. Such landslides have been historically less common than shallow, disrupted landslides, but the exceptionally heavy seasonal antecedent rainfall probably increased the susceptibility to these much deeper landslides.

Fig. 11. Significant damage was observed to residential homes in a small town (labeled in Fig. 1) along the reconnaissance route (37.3091N, 138.899E).

EARTH SLUMP AT YAMAKOSHI VILLAGE SURVEYED ON NOVEMBER 19, 2004

A LIDAR survey was conducted of an earthquake-generated earth slump in Yamakoshi Village on November 19, 2004. The location was chosen as it illustrates a slumping seen in many areas proximal to the epicenter. The survey was conducted from two locations on a hillside opposite the slump. The view perspective is approximately as shown in the first photo below. The survey took approximately three hours and was conducted in steady rain. The data indicate that the slump head scarp is a maximum of 8 meters high, the total relief is 42 meters and the average slope is 16.5°. The slump is approximately 170 meters wide and 140 meters from head to toe.

Fig. 12. Overview photo of area in Yamakoshi Village that suffered an earth slump.

Fig. 13. Close-up of slump head scarp and slump block. The back-tilted house was likely originally at the elevation of the blue house above.

Fig. 14. Small pond created by the damming of a creek by the earth slump.

Fig. 15. Image acquired by the imager onboard the LIDAR system. Superimposed on the image is the location of the profile plotted below.

Fig. 16. Profile of slump showing a total relief of approximately 42 meters and an average slope of approximately 16.5°.

Fig. 17. Plan view of earth slump showing the limits as determined by the LIDAR survey. The black areas are areas that could not be seen from the two tripod setups. Survey data of the front faces of buildings show up as lines when projected in a birds-eye view. The geometry of the retaining wall can also be seen. The extent of the slump is approximately 140 meters by 170 meters, with a total relief of 42 meters. The head scarp is approximately 8 meters high. The color data is acquired by the Texel Camera (TM) capability of USU's custom built LIDAR system.

ROCK SLIDE INVESTIGATION NOVEMBER 20, 2004

A field investigation and LIDAR survey of a translational rock slide were completed on November 20, 2004. This rockslide is one of many that occurred adjacent to one another along a section of Highway 291 east of Ojiya and west of Yamakoshi Village. The rockslide shear plane is remarkably planar and occurred along sandstone bedding planes with a dip-slope orientation. It is this dip-slope geometry along this section of highway that likely led to the multiple failures in the area. The LIDAR survey was conducted from one strategically positioned tripod setup as shown in Figs. 19 and 20. The rockslide is approximately 70 meters long and 40 meters wide and varied from a depth of 1 meters on the left flank (looking upslope) to approximately 5 meters on the right flank. The uniform slope is 24.6°. The map data presented in the section was prepared using ESRI ArcGIS GIS software.

Fig. 18. Aerial photograph of landslides in east Ojiya (37.3294N, 138.8259E). The two rock-block slides were triggered on Highway 291, close to the location where the highway exits the mountains at the Shinano River. This area is also just south of the White Rock landslide. Both slides occurred along bedding planes on dip slopes dipping about 25^o west. LIDAR survey and subsequent photos are from slide on the left hand side of this photo. (Aerial photo courtesy of ORIS http://www.oris.co.jp/jishin2004/h16jisin_top.htm).

Fig. 19. In this photo, the landslide block is in the right foreground; the notch in the ridge in the upper center of the photo was created by the landslide. The shiny surface in the center of the photo is the bedding-plane shear surface.

Fig. 20. Photo showing LIDAR instrument setup on the left flank of the rockslide. Note the planarity of the shear plane and the height of the side scarp on the on the far side.

Fig. 21. View laterally across the shear surface to the left of the block slide. The left margin is defined by a smooth vertical surface about 5 m high.

Fig. 22. Shear surface of block slide looking toward left slide margin (man for scale in right-center of photo). The surface is very planar and is well exposed except for some small rubble left behind the main slide block.

Fig. 23. Photo illustrating the almost perfect planarity of the shear surface.

Fig. 24. Upper face of the main slide block, view looking downslope. Most of the trees are vertical, which indicates pure translation on a planar surface.

Fig. 25. Evidence of previous block sliding is exposed in the left margin of the slide. Intact rock is visible upslope (upper left), but colluvium has filled in downslope where an earlier block moved downslope.

Fig. 26. The head of the rockslide occurs on a ridge.

Fig. 27. Colored LIDAR data showing the profile of the rockslide, in perspective. Note the location of the highway shown at the foot of the slope.

Fig. 28. Contour map of the rockslide derived from the LIDAR survey. This data were all collected from one tripod setup as shown. The map color represents elevation and is sliced into 1 meter intervals. Note the location of the cross-section and profile that are plotted below.

Fig. 29. Rockslide profile showing linear shear surface. Note the location of trees and rock debris carried to the foot of the slope adjacent to the highway.

Fig. 30. Rockslide cross-section showing the asymmetry of the rockslide depth. The original ground surface is assumed to be a linear projection of the adjacent ground surface.