A comprehensive assessment on the disastrious characteristics of landslide.

Wang, Jeffrey ; Guangya, Kong ; Wenyi, Yang 等

This paper has presented a method by using the tool of weight functions to assess the disastrous characteristics of landslides. This method allows us to consider the key factors which affect landslides to make a comprehensive assessment quantitatively the possibility of disastrous landslides. The method has been used in the feasibility study of the Three Gorges dam feasibility study to assess the slope stability of the dam and the disastrous characteristics of potential landslides.

GENERAL

One of the problems associated with the assessment of landslides in hydraulic dam area is the prediction of disastrous characteristics of landslides during the process of water level rising. Upon the landslide occurs under any circumstance, the sliding speed of landslide is one of the key effects on the impact to the safety of dam and surroundings. The current method on prediction of landslide speed is very primitive, which are only based on site observation and preliminary judgment. The actual prediction on the possibility of high speed landslide is actually impossible, as the factors which affect the movement of landslide is too many to judge if without a systemically assessment.

SALIENT FACTORS AFFECTING LANDSLIDE SPEED

A landslide could be disastrous to surrounding if it is located in urban area, or near a dam, or other sensitive buildings. To classify the disastrous level of a landslide, one sensible way is to use its sliding speed including maximum speed and average speed of landslide (Table 1).

It is prudent to consider a landslide as a system rather than an isolated case. There are many factors, which affects the occurring, sliding of a landslide. It is hardly to combine all those factors to conclude the disastrous level of a landslide. However, if we analysis all factors including geology, geometry, environment, rain, human factor etc., it is possible to summaries the salient factors which affect the sliding speed of a landslide as the followings: 1) volume of landslide (v); 2) slope of slide path ([alpha]) ; 3) relative height of shear-out position (h); 4) transitional angle of sliding path ([beta]); 5) magnitude of earthquake (a); 6) rain factor (w); 7) deformation of landslide body (u); 8) ratio of length of a landslide to length of path ([K.sub.L]).

Where the rain factor (w), transitional angle of sliding path ([beta]) and ratio of length of a landslide to length of path ([K.sub.L]) are defined as (Fig.1):

(1) w = W/[W.sub.b] (where W is water content and [W.sub.b] is the saturated water content)

(2)[beta] = ([[alpha].sub.1] - [[alpha].sub.2])

(3) [K.sub.L] = [L.sub.2] + [L.aub.3]/[L.sab.1]

[FIGURE 1 OMITTED]

WEIGHT FUNCTION OF SALIENT FACTORS

The contribution of the salient factors which affect the sliding speed of landslide could be quantitatively represented by three weight functions, which are corresponding to high speed, medium and low speed, respectively.

The standard trends of three weight functions for all salient factors are shown in Fig. 2, in which Fig. 2 (a) shows the weight function of the particular salient factor attributive to high speed, whereas Figs.2 (b) and (c) to medium and low speed landslides, respectively. However, the weight functions of transitional angle of sliding path ([beta]) are different from others, which are shown in Fig. 3.

[FIGURES 2-3 OMITTED]

The characteristic values [F.sub.ij] of each salient factor could be obtained from overall analysis of the historical records of different types of landslides (Table 2). On the basis of the above weight function, it is possible to calculate the value of combined coefficient [[eta].sub.ij], which reflects the possibility of each case, i.e. high sliding speed, medium sliding speed or low sliding speed.

(4). [[sigma].sub.ij] [8.summation over (k = 1)] W.sub.kj] ([d.sub.ik])[[eta].sub.kj]

Where, i =1,2,3, representing the cases, i.e. high sliding speed, medium sliding speed or low sliding speed; Wkj is value of white function of each salient factor, j=1,2,3 ...8, representing each salient factor; [d.sub.ik] is actual values of salient factors of the particular landslide to be assessed.

[[eta].sub.kj] = [F.sub.kj]/[8.summation over (i = 1)]Fij, Where j=1,2,3, representing the cases, i.e. high sliding speed, medium

sliding speed or low sliding speed; [F.sub.ij] refer to the characteristic values of each salient as shown in Fig. 2.

ANALYSIS OF OCCURRED LANDSLIDS BY WEIGHT FUNCTION METHOD

A few of large scale landslide which occurred in China were assessed by using the above described method. Table 3 shows the values of all salient factors, which affect the sliding speed, and Table 4 shows the results of assessment and the comparison between the assessed results and actual conditions. As can be seen, the results of assessment of each occurred landslides show a good agreement with the actual conditions, which in return show the validity of the proposed method. It should take a note that when calculating the combined coefficient [[eta].sub.ij], the values of slope of slide path ([alpha]) and transitional angle of sliding path ([beta]) should be converted to 10 times of radian to minimize the magnitude effect.

APPLICATION OF WEIGHT FUNCTION METHOD TO THREE GORGES DAM PROJECTS

Three Gorges dam project has about 686km long slope from Miao River to Jiangjin. It has been found that there are about total 277 number of landslides or potential landslide along the slope, with average frequency of one per 2 to 3km long river slope. As the catastrophic landslide (high speed landslide) could induce overflow of dam, or damage of the dam structure, it is substantial to investigate the sliding mode of each landslide and potential landslide, which would provide rational measures to prevent any possible high speed landslide along the river after the built-up of the dam.

The above described weight function method has been adopted for a total 13 number of large scale landslides (volume more than 10million cubic meters). The values of all salient factors, which affect the sliding speed, are surveyed and shown in Table 5, and the results of assessment and the comparison between the assessed results and actual condition are shown in Table 6. As indicated in Table 6, most of landslide or potential landslide, if reactivated after the built-up of the dam, the slide speed is medium to low except Huangnashi landslide, which could be strengthened before the dam construction stage. This conclusion is of substantial to the overall evaluation of slope stability along the river for the feasibility study of Three Gorge dam project.

REFERENCES

Hsu, K.J. (1969). "Role of cohesive strength in the mechanics of overthrust faulting and landsliding", Geol. Soc. Amer. Bull , Vol. 80, 927-957.

Cruden, D.M. (1980). "Anatomy of landslide", Canada Geotech., Vol. 17,. No.2.295-299 378

JEFFREY WANG

Tritech Consultants Pte Ltd, Singapore

KONG GUANGYA, YANG WENYI

Tiandi Science and Technology Co. Ltd, China

Table 1. Classification of landslide according to sliding speed.

 Average Maximum
 sliding sliding disastrous Engineering
 speed speed level solutions

High >10 m/sec. >20 m/sec. High Serious precautions,
speed close monitoring.
landslide Strengthen if
 possible

Medium 5 to 10 10 to 20 Average Close monitoring.
speed m/sec. m/sec. Strengthen if
landslide possible.

Low speed <5 m/sec. >20 m/sec. Possibly Monitoring.
landslide low

Table 2. Index values of weigh function of each salient factors
Affecting sliding speed.

 Factors [F.sub.a1] [F.sub.a2]

1 volume of landslide (v) 7 --
2 slope of slide path ([alpha]) 30[degrees] --
3 relative height of shear-out
 position (h)
4 transitional angle of sliding path 10[degrees] 30[degrees]
 ([beta])
5 magnitude of earthquake (a) 7[degrees] --
6 rain factor (w); 1 --
7 deformation of landslide body (u); 1m --
8 ratio of length of a landslide to 1.5 --
 length of path ([K.sub.L]).

 Factors [F.sub.b1] [F.sub.b2]

1 volume of landslide (v) 6 12
2 slope of slide path ([alpha]) 20[degrees] 40[degrees]
3 relative height of shear-out 5m 10m
 position (h)
4 transitional angle of sliding path 20[degrees] 40[degrees]
 ([beta])
5 magnitude of earthquake (a) 6[degrees] 12[degrees]
6 rain factor (w); 0.8 1.6
7 deformation of landslide body (u); 0.5m 1m
8 ratio of length of a landslide to 1 2
 length of path ([K.sub.L]).

 Factors [F.sub.c1] [F.sub.c2]

1 volume of landslide (v) 5 12
2 slope of slide path ([alpha]) 10[degrees] 30[degrees]
3 relative height of shear-out 0.2m 5m
 position (h)
4 transitional angle of sliding path 30[degrees] --
 ([beta])
5 magnitude of earthquake (a) 5[degrees] 12[degrees]
6 rain factor (w); 0.5 1
7 deformation of landslide body (u); 0.2m 1m
8 ratio of length of a landslide to 0.5 1.5
 length of path ([K.sub.L]).

Table 3. Values of all salient factors which affect the sliding
speed of some occurred typical landslides in China.

 Landslide v [alpha] H [beta]
 (x 1000 m3) [??] M [??]
1 No. 2 3800 30-45 0 <10
 landslide
2 Longyang 80000 30 0 3-5
3 Mefang 80000 24 0 0
4 Chana 160000 40 0 0
6 Chaxi 46800 30-45 0 0
7 Shaleshan 60000 23 0 0

 Landslide a w u [K.sub.L]
 [??] m
1 No. 2 7 0.5 <1 <0.5
 landslide 1 >1
2 Longyang 7 >1.5
3 Mefang 7 0.5 0 1
4 Chana 7 0.5 >1 >1.5
6 Chaxi 7 0.5 >1 >1.5
7 Shaleshan 7 0.5 1.5 >1.5

Table 4. Assessment results of occurred landslide with weight functions
and comparison to actual conditions.

 Landslides Assessed results

1 No. 2 slide Low sliding speed
2 Longyang High sliding speed
3 Mefang Low to Medium sliding
 speed
4 Chana High sliding speed
6 Chaxi High sliding speed
7 Shaleshan High sliding speed

 Landslides Actual conditions

1 No. 2 slide Low sliding speed, less catastrophic
2 Longyang High sliding speed, catastrophic
3 Mefang Low sliding speed, less catastrophic
4 Chana High sliding speed, catastrophic
6 Chaxi High sliding speed, catastrophic
7 Shaleshan High sliding speed, catastrophic

Table 5. Values of all salient factors of typical large-scale
landslides in Three Gorges.

 [alpha] H [beta]

 Landslides (x 1000 [degrees] M [degrees]
 m3)
1 Baiyiyan 21220 20 0 20
2 Qiancaotuo 16250 20 0 20
3 Baihuanpin 91120 50 0 50
4 Sandenzi 15000 40 0 40
5 Xinwu 46800 24 0 5
6 Oh-Tang 27000 40 0 15
7 Gaojiazui 27000 14 0 0
8 Baota 27000 10 0 0
9 Jipazi 27000 10 0 0
10 Xichen 27000 10 0 0
11 Jiuxianpin 27000 15 0 0
12 Huangnashi 27000 30 0 0
13 Fanjiaping 27000 19.4 0 0

 a w u [K.sub.L]

 Landslides [degrees] m

1 Baiyiyan 6 >1 0 0.7
2 Qiancaotuo 6 >1 0 0.7
3 Baihuanpin 6 >1 0 0.7
4 Sandenzi 6 >1 0 0.7
5 Xinwu 6 >1 0 0.7
6 Oh-Tang 6 >1 0 0.7
7 Gaojiazui 6 >1 0 0.7
8 Baota 6 >1 0 0.7
9 Jipazi 6 >1 0 0.7
10 Xichen 6 >1 0 0.7
11 Jiuxianpin 6 >1 0 0.7
12 Huangnashi 6 >1 0 0.7
13 Fanjiaping 6 >1 0 0.7

Table 6. Assessment results of occurred landslide with weight functions
and comparison to actual conditions.

 Estimated values of combined
 coefficient [[eta].sub.ij]

 [[eta].sub.1j] [[eta].sub.2j] [[eta].sub.3j]
 (high (medium (low
j Landslides speed) speed) speed)

 1 Baiyiyan 0.4496 0.652 0.6232
 2 Qiancaotuo 0.4496 0.653 0.6232
 3 Baihuanpin 0.462 0.490 0.6371
 4 Sandenzi 0.461 0.5064 0.6359
 5 Xinwu 0.460 0.5759 0.5153
 6 Oh-Tang 0.4781 0.5569 0.561
 7 Gaojiazui 0.4467 0.5573 0.5233
 8 Baota 0.438 0.4534 0.518
 9 Jipazi 0.438 0.542 0.535
10 Xichen 0.438 0.538 0.531
11 Jiuxianpin 0.449 0.5484 0.5060
12 Huangnashi 0.615 0.546 0.410
13 Fanjiaping 0.459 0.562 0.491

 Assessment
 results of
 estimated
j Landslides sliding speed

 1 Baiyiyan medium
 2 Qiancaotuo medium
 3 Baihuanpin low
 4 Sandenzi low
 5 Xinwu medium
 6 Oh-Tang low
 7 Gaojiazui medium
 8 Baota low
 9 Jipazi medium
10 Xichen medium
11 Jiuxianpin medium
12 Huangnashi high
13 Fanjiaping medium