1. DPRI, KYOTO UNIV.
Comments on the report of the
2011 Unmyong-san disaster
Hiroshi FUKUOKA
Associate Professor
Research Center on Landslides
Disaster Prevention Research Institute
Kyoto University

2. DEBRIS SLIDE – DEBRIS FLOW
EXPERIMENT IN JAPAN, 2003
by giving artificial rainfall of 80 mm/h x
5 hours to natural slope
Artificial landslide experiment using rainfall simulator
(front view) (Forestry and Forest Product Research Institute)

3. Artificial landslide experiment using rainfall simulator
(side view) (Forestry and Forest Product Research Institute)

4. Comments on the “6. Concluding Remarks”
• 1. Widely distributed thick colluvial deposits and shallow ground
water table which frequently appears suggests high risk.
• 2. (recovery work at Dukwooam source area is not clear)
• 3. Topography and other geomaterial condition on the border of Air
Force before the disaster is still unknown, then their responsibility is
not clear. (Problem is not the # of slides.)
• 4. Check dam and other large-scale structured countermeasure
construction needs more than a year. Complete coverage needs
gigantic costs (unrealistic). Damage to the properties in 2011 could
NOT be avoided, but early warning could have helped more lives.
• 5. Reservoirs sometimes contributed to longer debris flow runout.
• 6. Ground motion due to underground blastings never caused slides,
however, may have contributed to more underground cracks.
Groundwater veins, pipings may have been affected, and long-term
slope stability could be affected, too.

5. Present risk of artificial fill on the
border of Air Force Base

6. Air force base border

7. Hyeongchon
reservoir site

8. Temple site

9. Conduit of Temple site

10. Proposed slit
work against
upcoming
debris flows

11. Comment (1)
• Extra-ordinary intense rainfall of 120-years
return period is the primary triggering factor of
the 2011 disaster.
• Frequency of those extreme weather events in
limited area (a few km2) is apparently increasing
in most countries. Could expect more events in
the future.
• Structured countermeasure for those events
everywhere cost so high and not realistic.

12. Numerous landslides
in Shobara city,
Hiroshima Prefecture,
Japan, induced by
extreme rainfall in
July 2010. This
disaster is localized
in a small area of
about 5 km x 3 km.

13. Debris flow distribution in Shobara city
(Asia Air Survey, Co.)

14. A number of landslides were induced by the
intense rainfall

15. A panoramic photo of the head part of the
Ohto river. Shallow landslides took place
and covered almost 360 degrees.
C

16. Debris flow distribution in Hofu city
interpreted by airphotos

17. Hourly precipitation and 10 minutes precipitation as well as cumulative in
Hofu and Yamaguchi city

18. Calculated return period of hourly, 3-hours, 6-
hours, and 24 hours precipitation
Yamaguchi city: return period of max 6 hour rain = 601,7 years.
Hofu city: max 6-hours rain: 245,9 years.
The key factor was extraordinary large 6-hours precipitation

19. Comment (2)
• Micro-landslide scars could be extracted by
detailed interpretation of LiDAR-based map
combined with appropriate image processing.
• Precursor depression could NOT be extracted
by present sensors yet, except ground-based
precise sensors.
• Comparison of before and after the disaster can
give volume estimates of initiation, erosion and
deposition.

20. Airborne laser scanner penetrating
forests to detect old landslide scars
Example : July
2009 Hofu city
debris flow disaster
area, western Japan
GPS Satellites
Pulse laser penetrates to
Helicopter the ground surface
hidden under the forest
Reference pt.

21. Comment (3)
• Initiation mechanism: penetration of rain
water + sliding surface liquefaction,
otherwise unlimited displacement can not
be expected. ß geotechnical
characterisitcs
• Evolving mechanism: Fluidizing or
scraping of torrent deposits by undrained
impacting loading by slide body.

22. Air photo of a slide-debris flow in Hiroshima (1999.7.14 Sassa）and its plan map
（after Yokota et al., 1999）

23. The source volume is about 300 m3,
the deposits is 20 times larger.

24. Ring-
Ring-shear test for reproduction of landslide movement
Simulating
the sliding
surface
Normal axis Normal
stress stress
res s
Shear st
Sliding Sample
surface 24
Un-drained shear box

25. A slide- debris flow disaster in Izumi city, 1997

26. 体積減少
Volume decrease せん断ゾーン
Shear zone
すべり面における粒子破砕とすべり面液状化
Grain crushing takes place in the shear zone, volume tends to
decrease, then excess pore pressure is generated up to almost
same level as the normal stress (liquefaction). This condition is
named as “Sliding-Surface Liquefaction”
Illustration of sliding surface liquefaction
（Sassa, 1996 and Sassa et al.1996）

27. Landslide Ring-Shear Simulator to reproduce high speed
shearing at a sliding surface
（Normal Stress 200 kPa , Shear Velocity 200cm/sec）
Grain crushing causes muddy water and sliding surface liquefaction

28. Sliding Surface Liquefaction can be compared to….
3rd floor
※Rooms are filled with water and undrained
2nd floor condition is maintained
1st floor
Big reduction of apparent friction
High water pressure is generated

29. Model of the landslide-triggered debris flow (Sassa et al. 1997)

30. Illustration of moving landslide mass along a torrent
(after Sassa et al 2004a)

31. Results of a test to
simulate undrained
loading to the tuff
breccia deposits in
Minamata landslide
(BD = 0.89) (after
Sassa et al 2004a).

32. Thank you for your
attention !!!