Landslide Risk Management of Patong City, Phuket, … Risk Management of Patong City, Phuket,...
Transcript of Landslide Risk Management of Patong City, Phuket, … Risk Management of Patong City, Phuket,...
Landslide Risk Management of Patong City, Phuket, Thailand
S. Soralump1, D. Pungsuwan2, M. Chantasorn3 & N. Inmala4 1Geotechnical Specialist 2Researcher 3GIS Analyst 4Graduate Student Geotechnical Engineering Research and development center, Civil Engineering Department, Kasetsart University, Thailand
NM.S.I. Alambepola Asian Disaster Preparedness Center, Thailand
ABSTRACT: The development of Patong city, Thailand is occasionally caused the serious man-made disaster such as flooding and landslide. Geotechnical approach was introduced for producing landslide susceptibility map and landslide risk map. The new landslide causative factor was introduced which are the man made factors such as the affected area by road cutting, the area of prohibition by law and others. Critical API value, which is based on rainfall recorded data, was calculated and the appropriate warning method was adapted for local community based on their current communication scheme. The project is perhaps the first landslide risk mitigation project in Thailand.
Keywords: Landslide, Landslide Susceptibility Map, Landslide Risk Map, Landslide Management, Antecedent Precipitation Index
1 INTRODUCTON
Landslide is one of the natural hazards that af-fected Thailand. The direct economic lost due to landslide is calculated to be equal to 100 million Baht per year and the return period of large area landslide is once in every 3-5 years (Soralump 2010).
Patong beach, located in Phuket province, is one of the famous tourist destinations in Thailand. This beautiful beach is surrounded by the crescent moon shape mountain range. Since the city is ex-panding, the use of steep slope area on the moun-tain is unavoidable. This causes the disturbance to the environment which later on causing the land-slide. Asian Disaster Preparedness Center (ADPC) together with The Norwegian Geotechnical Insti-tute (NGI) had the responsibility for execution of Regional Capacity Enhancement for Landslide Im-pact Mitigation pro-gram (RECLAIM) which the funding was provided by the Royal Norwegian Embassy in Bangkok. Department of Mineral Re-sources (DMR) and Geotechnical Engineering Re-search and Development center (GERD), Kasetsart University were asked by ADPC to take responsi-bility for the implementation the project in which they decided to demonstrate the landslide mitigation in Patong city, Phuget province.
2 STUDY AREA
Patong Municipality is approximately 16 square kilometers in area and located on the west coast of Phuket Island. Population mostly consists of tour-ists and local Muslim communities. However, after the tsunami event in 2004 the development tends to move higher on the mountain. Improper change of the slope geometry and the land cover are the main causes of the landslide in the area (Figure 1) (Pungsuwan 2006).
Table 1 shows some landslide event records gathered by Patong municipality, it shows that most of the landslide has triggered by excessive rainfall.
Fig. 1 Na Nai roadside landslide on October 25, 2007
Table 1. Landslide records gathered by Patong Municipality. No. date location Triggering
factors
1 October 19, 2001
Various places
Heavy rainfall
2 October 21, 2003
50th anni-versary Rd.
Cut slope and heavy rainfall
3 October 14, 2004
Na Nai village
Blockage of drainage and heavy rainfall
4 October 14, 2004
Kalim village
Inappropriate drainage and heavy rainfall
5 October 25, 2006 Na Nai Rd.
Cut slope non protection and heavy rainfall
6 July 15, 2007
50th anni-versary Rd.
Construction failure
7 September 5, 2008
50th anni-versary Rd. Drainage
8 September 19, 2009
Kathu-Patong Rd. Heavy rainfall
3 GEOLOGICAL CHARACTERISTICS
Geological characteristics of Patong include forma-tion of granite in Cretaceous period and modern beach sediments in Quaternary period. Granite rock is found to be in moderately to highly weath-ered condition with various sets of joint and frac-ture. Left lateral strike-slip fault has found mostly with their strike lies between north-east to south-west direction. Department of mineral resources has interpreted the satellite image for the geologic structures of Patong as shown in Figure 2 (De-partment of Mineral Resources 2006).
Fig. 2 Geological characteristics, Patong (Department of Mineral Resources 2006).
4 SLOPE STABILITY ANALYSIS
The strength characteristics of residual soil from 12 locations in the study area were used for the stability analysis. The strength parameters of re-sidual soil were obtained by performing the Multi-Stage Direct Shear Test together with the data gathering from the previous studies (Soralump et al. 2007). The drained direct shear tests were done to the soaked (almost saturated) samples. During such test, in order to ensure the drained behavior, the pressure sensor which capable of measuring both positive and negative pore pressure was em-bedded in the top cap to monitor the change in pore water-air pressure. It was found that the ex-cess pore pressure only exist shortly during con-solidation stage, however there was no evidence of positive pore pressure in the shearing stage. This can be concluded that the soil samples were sheared under fully drained condition. The strength parameter values (c’,φ’) shown that the cohesion value has tendency of increasing when the degree of saturation is getting higher (Fig.3).
Slope stability analyses were done using KUs-lope computer program developed by GERD (Is-aroranit 2001). The geometry of the mountain slope was studied in order to select the appropriate cross section for the analysis.
The analyses were done by modeling various slope angles from 14 to 40 degrees for both natural (one bench) and cut slope. Since the strength pa-rameters vary through the degree of saturation (Figure 3) and also there are some variation of val-ues among them, therefore the slope analysis were done by using various pairs of strength parameters (c’,φ’) along the left boundary line as shown in Figure 3. In order to select a pair that gives the cor-relation of the lowest calculated FS for various slope angles. The results are concluded in Fig 4 which shows the correlation as such using the se-lected strength parameters. The results also show that the factor of safety values of the cut slope is generally lower than the natural slope, which is expected to be. Furthermore, it was found that the natural slope angle that the slope cutting might trigger the landslide is found to be 17.1 degree in which corresponding to FS equal to 1.3. This slope angle is generally gentle than the current regulation which prohibits the slope modification that of greater than 19.3 degree (35%). Therefore, it is ad-vised that any construction cutting (1H:2V) done to the natural slope with 17.1 degree angle or more, the proper slope analysis and the counter measures need to be considered in Patong
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Friction, degree
Coh
esio
n, t
/m2
New Data
Previous Data91.85
94.99
91.34
94.18
86.51
85.01
86.96
89.77
63.89%
62.16%
86.19%
100
Cohesion 0.168 0.170 0.200 0.250 0.300 0.350 0.410
Degree 34.5 33.0 29.6 27.0 25.2 24.8 26.1 Fig. 3 Boundary of strength parameters used for the analyses
0.500.600.700.800.901.001.101.201.301.401.501.601.701.801.902.002.102.202.302.402.502.602.702.802.903.00
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42Natural Slope Angle (degree)
Fact
or o
f Saf
ety
Normal Slope cohesion = 0.35 degree = 24.8Cut Slope 2 : 1 cohesion = 0.17 degree = 33Cut Slope 3 : 1 cohesion = 0.17 degree = 33
Fig. 4 Factor of safety-slope relationship of cut and natural slope.
5 LANDSLIDE SUSCEPTIBILITY MAP
The landslide susceptibility area in Patong was analyzed using weighting factor method. 8 major factors listed below were considered in the analysis including the safety factor calculated from geo-technical method explained in the previous section.
1. Soil slope stability 2. Rock slope stability 3. Lineament zone 4. Rock type (Later neglected) 5. Distance from road 6. Elevation 7. Land use and land cover 8. Surface drainage zone
The factor related to the rock slope stability which are the rock fractures and slope face orienta-tion were neglected since the detail analysis found that there is less likely that the rock slope instabil-ity will be occurred. Furthermore, since there is only one rock type in Patong, the factor of rock type was also neglected. Therefore, the numerical rating for landslide potential is then used 6 related factors. The area of 5x5 square meters grid cells have been employed for the analysis by GIS pro-gram. The weight-rating value of each factor was determined in each square grid cell of each deriva-tive map. Table 2 shows the weighting value of each factor, calculating from the weighted matrix based on expert assessments. Finally the scores of weight-rating in each 5x5 square meters grid cell were obtained from the summation of weight-rating values of each derivative map. The levels of score were classified to obtain the landslide sus-ceptibility level of each grid and finally obtain the susceptibility map. Table 2. Weights and rating values used for the analyses.
Weight Value
Rating Value
Parameter Parameter Description
Rating
(1-5)
1. Factor safety and slope angle relationship
1.875 A. F.S.≤ 1.3 (≥26 de-gree)
B. 1.3 < F.S. ≤ 1.5 (22≤ slope< 26 degree)
C. 1.5 < F.S. ≤ 1.8 (18≤ slope< 22 degree)
D. F.S. >1.8 (< 18 de-gree)
5
3.66
2.33
1
2. Lineament zone
1.625 A. Area inside ineament zone
B. Area outside linea-ment zone
5
1
3. Distance from road
1 A. Area inside road zone
B. Area outside road zone
5
1
4. Elevation
1 A. >80 m B. 40-80 m C. 0-40 m (Not include
slope <10 Degree)
1 3 5
5. Surface drainage
1 A. Area inside surface drainage zone
B. Area outside surface drainage zone
5
1
6. Land use and land cover
1 A. Agriculture area B. Urban and built-up
area C. Other deforestation D. Forest area
5 3.66
2.33
1
Left Boundary Line
The results of the analysis are shown in Fig 5. It clearly shows that the actual event locations are mostly located within the moderate to high land-slide susceptibility area obtained from the analysis. Considering this, the percentage of map accuracy is determined to be 78.2 percent. Furthermore, Ta-ble 3 shows the suggestion for land development act, proposed to Patong Municipal which will be used as a guideline for the future enforcement. This is rather important since most of the landslide in Patong occurred because of the improper prac-tice in design and construction in the slope area.
Fig. 5 Landslide susceptibility map (ADPC, GERD and DMR, 2008, Soralump, 2010).
Table 3. Recommendations for slope treatment in vari-ous zones of landslide susceptibilities.
Landslide susceptibility Levels Action High Medium Low Very
Low Geotechnical engineer re-quired
√ √
Geologist required
√
Land cover control
√ √
Drainage management
√ √
Control of the slope angle for cut slope
√ √ √ √
6 LANDSLIDE RISK MAP
Since risk is a function of hazard and consequence. Therefore, in order to develop risk mapping, the consequence area from landslide need to be esti-mated. Building locations and their pertinent data were obtained from taxation map provided by Pathong City. The hazard levels were classified based on landslide susceptibility map produced as discussed earlier. The affected area at the toe slope area was estimated to be equal to the height of the slope above the toe based on Finlay et al. (1999). Fig. 6 shows the boundary of affected area at the toe of the slope. The buildings were classified into levels based on their vulnerability from landslide. The number of population at risk (PAR) is esti-mated from the census data in the taxation map.
Fig. 6 Landslide risk map in Patong (part).
7 API MAP
Landslide disaster management will be incomplete if lack of the warning system. The past studies found that the accumulated rainfall (or short period rainfall history) and current precipitation has great influence in rainfall-triggered landslide (Soralump 2010). This is consistent with the use of the API concept (Antecedent Precipitation Index). The API represents, presumably, the moisture of the soil at any time using the values measured by rain gauge. The critical API is determined by calculating the critical moisture content in the soil layer that will trigger the failure of soil layer. Factor of safety was analyzed by infinite slope stability method with the soil strength parameters of c' = 0.35 ksc., φ′ = 24.8o (Figure 5), the unit weight of soil ( d
γ ) = 1.41 t/m3 , Gs = 2.65, void ratio (e) = 0.89, degree
of saturation (Sr) = 93% and the porosity of the soil (n) = 0.471. These soil properties were ob-tained from the site investigation stage explained earlier. The calculation was done according to equation 1 and the critical thickness is determined from Equation 2. The result of critical thickness based on slope angle are shown in Figure 7. The critical API map is then produced as shown in Fig-ure 8.
cri
u HcrSF
γβββφβ
′−′
+−=cossin
1
tan
tan)]tan1(1[.. 2 (1)
When F.S. = factor of safety for infinite slope; ru = pore pressure ratio Hu γ/ ; β = Slope angle; γ = unit weight of sliding mass; Hcri = critical depth (m.)
crcrrcr TSnAPI ⋅⋅=,
(2) when crAPI = critical API (mm.); n = porosity;
crrS,
= percent critical of saturation; crT = critical depth (m.)
RELATIONSHIP BETWEEN SLOPE AND THICKNESS
0
5
10
15
20
25
30
35
40
45
0 10 20 30 40 50 60 70 80 90CRITICAL SLOPE
CRI
TIC
AL
THIC
KNE
SS,
m Sr = 93%
Fig. 7 Relationship between the critical thickness and slope angle.
Fig. 8 Critical API map of Patong (ADPC, GERD and DMR 2008, Soralump 2010).
8 WARNING SYSTEM USING RAINFALL DATA
In order to apply the critical API method for land-slide warning, it is required to install automatic rain gauge to measure the precipitation in the study area. The rainfall data is used for calculating the API value at time t (APIt) (Equation 3) (Soralump and Thowiwat 2010). Fig.9 to 11 shows the com-parison of using the 3 days accumulated rainfall data and the API value for crating the critical en-velop. It can be seen that using 3 days accumulated rainfall does not give the pre-warning in some case. However, using API value seems to be more appropriate. However, as seen in Fig.11, it shows that a landslide event occurred before the calcu-lated critical API envelop, this is because of the critical API value was calculated based on assump-tion of natural landslide. However, the failure is considered to be the man-made landslide. Further-more, the Alert-Alarm-Action criteria were set as Shown in Figure 12 for practical purpose.
tttt PAPIKAPI +×= − )( 1 (3)
when APIt = API at time ‘t’ (mm.); APIt-1 = API at time ‘t-1’ (mm); Pt = Precipitation at time ‘t’ (mm.); K = recession constant (K =< 1.0 and usu-ally 0.85-0.98)
Critical rainfall envelope (3-days)
000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000050
7060
40
1500
70
0010
0000000000
25
000000
6050
20
5040
000
70
4020
4050
70
020
0000000000020
40
70
0000000000
110
0020
40 40
00000000010
5030
100
20000
30
00
30
100 100
70
3020
10500
3040
30
020
1005
30
000000000000000000000000000000000000000000000000000000000000000000000000000004.50.25000320.25
0.25500000.500000000000000000000002000000006.250.2500002.50.2500000000000074.2530011.5
01.25000001000000.25005.2500010
37.75
0.250.25018
37
0.254.75
34.25
41.5000000000000008.25
31.7545.5
130.250030.254.7511.25
56.25
0
41
00000010.75
00.25000000004
31.5
0.251.54.512.75218.25
000007.500000001.59.25
36.25
00005.5000000000.50.25000000.25216
000000.25000.250.25001018.25
1.759.751.250.50000003.25
36.75
0.50.2500000
25
0.7512.2512.7521.75
0000.2502029.75
40.25
00.7504.75
55.75
011.5
54
0003.5
127.25
200.25
41.75
11
85
34.5
6.50.750.521
0.250000000005002.7521.5
00000000000000000000000000000000000.2500000000000000000000000
50
100
150
200
250
300
0 50 100 150 200 250 300 350 400
3-days accumulated rainfall (mm.)
Dai
ly R
ainf
all (
mm
.)
2007(Tarawadee)2007(Sai Mon)2007(Tai Trang)200420032008
Oct 14,2004
Fig. 9 3 days accumulated rainfall and daily rainfall re-lationship.
Antecedance Precipitation Index ( APIcritical )
000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000050
7060
40
1500
70
0010
0000000000
25
000000
6050
20
5040
000
70
4020
4050
70
020
0000000000020
40
70
0000000000
110
002040 40
00000000010
5030
100
20000
30
00
30
100 100
70
3020
10500
3040
30
02010
05
30
000000000000000000000000000000000000000000000000000000000000000000000000000004.50.25000320.25
0.25500000.500000000000000000000002000000006.250.2500002.50.250000000000007 4.2530011.5
01.25000001000000.25005.2500010
37.75
0.250.25018
37
0.254.75
34.25
41.5000000000000008.25
31.7545.5
130.250030.254.7511.25
56.25
0
41
00000010.7500.25000000004
31.5
0.251.54.512.752
18.25000007.500000001.59.25
36.25
00005.5000000000.50.25000000.25216
000000.25000.250.25001018.25
1.759.751.250.50000003.25
36.75
0.50.2500000
25
0.7512.2512.75
21.750000.250202
9.75
40.25
00.7504.75
55.75
011.5
54
0003.5
127.25
200.25
41.75
11
85
34.5
6.50.750.5210.250000000005002.75
21.500000000000000000000000000000000000.2500000000000000000000000
50
100
150
200
250
300
0 50 100 150 200 250 300 350 400
Antecedance Precipitation Index API(t-1) (mm.)
Dai
ly R
ainf
all (
mm
.)
2007(Tarawadee)2007(Sai Mon)2007(Tri Trang)20042003API critical/day line2008
Oct 14,2004
Fig. 10 API value and daily rainfall (Soralump, 2010).
APIcr (mm.)
Antecedance Precipitation Index ( APIcritical )
3.24.4500.3000000000000000000000000000000000000000000000000000000
27.3515.6514.85
26.8
015.05
1.2
25.459.450000
10.70.950.60.40
9.1522.818.55
00
47.9
0.10.713.352.454.857.65
20.752.95
14.9557.2
17.350
31.8518.9521.7
53.3
0.350.217.350.457.59.55
03.23.411.154.2511.28
32.2
2.352.10000.53.39.10.20.350.450.153.85
55.55
19.6
0.300
32.1
10.60.24.20.970
16.116.757.4001.35
11.8
67.8
25.65
4.05004.23.34.3000000.700004.8
52.25
28.15
00.051.9000.14.85.850.94.850.21000009.4
0.61.8
98.4
0.200
65.2
42.6
0004.40.4000
20.434.2
77.6
0.2020.60.60000000032.2
25.2
0.62.401.40000
22.6
0.44.811.2013.2
87
17.402.8
24
94
31.8
7.47.816
1012.2
22
0.200
21.4 15.46.43.41.8
12.422.4
0.4
22.2
00015.4
5.6
40.6
0.400.8
23.25.64.24.20.40
31.823.6
52
1310.8
38
14.41.2000.400.4
47.2
18.424
0018.6
0.2
47.8
27
49.8
156.800
63
11.240.20.40.200.21.8
11.8
34.822.2
3.40.2
65.4
12
58.6
0.2000000000000000000000.8100012
0.41.80.40000008.2
0.2000000
101.4
110.2
0
50
100
150
200
250
300
0 50 100 150 200 250 300 350 400
Antecedance Precipitation Index API(t-1) (mm.)
Daily
Rai
nfal
l (m
m.)
KALIM 2009API critical/day line
Sept. 19, 2009
Fig. 11 API graph showing the landslide event in 2009.
Antecedence Precipitation Index ( APIcritical )
0
50
100
150
200
250
300
0 50 100 150 200 250 300 350 400
Antecedence Precipitation Index API(t-1) (มม.)
Dai
ly R
ainf
all (
mm
.) mm./dayrainfall 2010
Fig. 12 Alert Alarm Action criteria.
9 CONCLUSIONS
1. Based on verified landslide susceptibility map, it was found that landslide mostly occurred in the lineaments zone.
2. Using API value rather than 3-day accumu-lated rainfall is much the warnings time.
3. The actual triggering API value seems to be less than the calculated critical API. This is be-cause the geotechnical model used for calculating the critical API is not considered the man-made factors.
ACKNOWLEDGMENTS
The authors would like to thank Norwegian gov-ernment and Patong Municipality.
REFERENCES
Asian Disaster Preparedness Center, ADPC; Department of Mineral Resources, DMR and Geotechnical Engineering Research and Development Center, GERD (2008). Land-slide Mitigation Demonstration Project for Patong City: Carried act as a Part of the Asian Program for Regional
Capacity Enhancement for Landslide Impact Mitigation (RECLAIM II) Bangkok, Thailand.
Department of Mineral Resources, DMR (2006). A Study of Prevention and Mitigation Landslide. Final Report, Bang-kok, Thailand.
Finlay, P.J., Mostyn, G.R., Fell, R. (1999). Landslide risk as-sessment: prediction of travel distance. Can. Geotech. J. 36, 556–562.
Isaroranit, R. (2001). Development for Slope Stability Pro-gram by Generalized Limit Equilibrium. M.S. Thesis, Ka-satsart University, Bangkok, Thailand.
Pungsuwan, D. (2006). Evaluation of Landslide Sensitive Area for Slope Development in Phuket. M.S. Thesis, Ka-satsart University, Bangkok, Thailand.
Soralump, S. (2010). Rainfall-Triggered Landslide: from re-search to mitigation practice in Thailand. The 17th South-east Asian Geotechnical Conference. May 10-13, 2010, Taipei, Taiwan.
Soralump, S., Thowiwat, W. and Mairaing, W. (2007). Shear strength testing of soil using for warning of heavy rainfall-induced landslide. Proceeding of 12th National Confer-ence on Civil Engineering. Phisanuklok, Thailand
Soralump, S. and W. Thowiwat (2010). Critical API Model for Landslide Warning. The Fifteenth National Conven-tion on Civil Engineering (NCCE15), 12-14 May 2010. Ubon Ratchathani, Thailand