Pore pressure as a trigger of shallow landslides in the ...The area has a history of shallow...
1
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION For more information: References: Pore pressure as a trigger of shallow landslides in the Western Ghats Introduction Pore pressure is considered an important trigger of landslides. Continuous monitoring of pore pressure at locations of landslides is limited by the fact that the initiation locations of landslides are almost never known prior to the occur- rence of the events. This is especially true in the case of shallow landslides such as debris flows. Researchers circumvent this constrain by moni- toring pore pressure in laboratory scaled models, known active and dormant landslides or hydro- logically and topographically similar terrain. This paper presents some initial interpretations of the observations of pore pressure from a network of piezometers coupled with the rainfall-intensity measurements from an automated weather station in the upper Tikovil River basin of the Western Ghats mountain chain of Kerala, India. Study Area The study area comprises Aruvikkal catchment, a 9.5 km2 sub basin of the Tikovil River, a tributary of the Meenachil River (Figure 1). The area is underlain by Precambrian charnockites, which weathers very slowly, leading to rather shallow sandy soils over a thin layer of sparolite inter- leaved by lithomargic clay. The predominant land use is rubber (Hevea Brasiliensis) plantations, which have to be cleared after 20 years, thus exposing the land to the high intensity rainfall. The area has a history of shallow landslides and consequent debris flows that originate in topo- graphical hollows. On 6th October 1993 the re- gion experienced 100 debris flows, the triggering 2 day cumulated rainfall of which was 315 mm; the experimental catchment experienced over 11 landslides on that day. On 22 nd June 2007 the catchment and surroundings experienced about 20 landslides, most of which were shallow land- slides such as debris flows, with the catchment experiencing two soil slips and a debris flow on the day (Figure 2). Figure 1: Study Area – Aruvikkal Basin Materials and Methods Two micro catchments of the Aruvikkal catch- ment were instrumented with 15 piezometers in the period June to September 2007, as part of an ongoing research on establishing the physical basis of slope hydrology and slope stability of the region. Three piezometers of the 15 were de- ployed before the onset of the monsoon. The pi- ezometers used were of the type Keller DCX-22 AA which measures and records ground water levels using a two sensor technology. The sub- mersible depth sensor measures the water level. Barometric pressure variations are measured and compensated with the built-in waterproof air pressure sensor which is mounted in the elec- tronics housing at the top of the borehole (www. keller-druck.com). Of the 15 piezometers, 4 were installed in hollows, with at least 0.7 m of over- burden thickness, 7 were installed in terraces with rubber plantations, 2 were installed in other landuse types and 2 were installed at drainage outlets. A low cost wireless automated weather station (AWS) branded Vantage Pro2 (www. davisnet.com), capable of measuring all standard weather variables and solar radiation was estab- lished at Meladukkam CSI church (see Figure 1) located in the center of the area at an altitude of 660 m. The logger of the AWS was programmed to record data at an interval of 15 minutes, and could store data for about 45 days given this interval. The data from the 3 piezometers that were installed before June 2007 were analyzed in combination with the AWS data. Figure 2: Initiation and run out zones of the debris flow (mass movement 1 in Figure 1) Results and Discussion During the south west monsoon season of 2007 Aruvikkal catchment received 3123 mm rainfall which was 16.3% less than the seasonal LPA rainfall of the area. On the day of the events, the region received 199.8 mm rainfall and the 2 day cumulated rainfall for that day was 278.4 mm. The 2 day cumulated rainfall explains the relatively lesser landslide activity in the region on 22 nd June when compared to that of the 6th October 1993 events. Figure 3 shows the 2 day cumulative rainfall and maximum daily rainfall intensity recorded by the AWS during the monsoon seasons of 2007. Figure 4 shows the hourly rainfall and the hourly maximum rain intensity from 12.00 pm on 21st June to 12.30 pm on 22 nd June. However, such absolute values are only of indicative relevance unless substanti- ated by long term observations. In situations where long term temporal landslide inventory is unavailable, physically based dynamic models of slope hydrology coupled with slope stability is an ideal alternative for hazard quantification. The pore pressure responses of the catchment also follow the general trend of the rainfall. Table 1 provides the topographic and land use charac- teristics of the locations of the three piezometers deployed before the critical date. It also provides the maximum, average of the maximum daily and standard deviation of the maximum daily pore pressure observations of the three piezom- eters during the south west and north east mon- soon seasons. It could be seen from the pore pressure data that the hydrological response of the sub surface started between 8.00 and 9.00 pm on 21 st June. From the time of the initial hy- drological response of the pores at around 8.00 pm to the beginning of the critical time (6.00 am, 22 nd June) of the landslide events 9 hrs later, the pore pressure values remained high. Figure 5 shows the pore pressure response at each piezometer during a continuous rain spell from 21 st June 12.00 pm to 23 rd June 12.30 pm. A close observation of the continuous re- sponse pattern of pressure waves brings out the transient nature of the hydrological response processes. Piezometers 1 (P1) and 3 (P3) behave similarly in their response to the fluctuations of the rainfall, while piezometer 2 (P2) shows a delayed response. P3 shows much more fluctua- tions than P1 which may be attributed to the local characteristics of the location such as satu- rated hydraulic conductivity, soil depth and slope which results in the sudden drainage of excess pore water. 0 50 100 150 200 250 300 5/28 6/22 7/17 8/11 9/5 9/30 10/25 11/19 12/14 Days Cumulated Rainfall (mm) 0 50 100 150 200 250 300 Rain intensity (mm/hr) Intensity 2 days cumulated Figure 3: 2 days cumulated rainfall and maxi- mum daily rainfall intensity of Aruvikkal catch- ment during the south west and north east monsoon seasons (note the highest 2 days cumulative rainfall on 22-06-2007) Table 1: Location characteristics, maximum pore pressure response, average of the maxi- mum daily and standard deviation of the maxi- mum daily pore pressure observations of the three piezometers during the south west and north east monsoon seasons Pore pressure (kPa) South West North East P A (m) S (º) SD (m) KSat (m/hr) LU/LC Max Avg Std Max Avg Std 1 873.7 46.7 2.6 0.11 MC 8.61 1.31 1.92 5.14 0.29 0.81 2 320.5 24.9 1.3 0.14 R 7.46 0.29 1.11 1.38 0.14 0.24 3 915.8 31.7 1.9 0.13 MF 12.33 4.47 3.49 10.29 1.14 2.59 P – Piezometers; A – Altitude; S – Slope; SD – Soil Depth; Ksat – Saturated hydraulic conductivity; LU/LC – land use/land cover (MC – Mixed crops, MF – Mixed Fallow, R – Rubber). Max – Maximum, Avg – Average, Std – Standard deviation 0 5 10 15 20 25 30 12:30 PM 2:30 PM 4:30 PM 6:30 PM 8:30 PM 10:30 PM 12:30 AM 2:30 AM 4:30 AM 6:30 AM 8:30 AM 10:30 AM 12:30 PM 2:30 PM 4:30 PM 6:30 PM 8:30 PM 10:30 PM 12:30 AM 2:30 AM 4:30 AM 6:30 AM 8:30 AM 10:30 AM 12:30 PM Time Rainfall (mm) 0 20 40 60 80 100 120 140 160 180 200 Rain intensity (mm/hr) Rainfall (mm) Rain intensity (mm/hr) Figure 4: Hourly rainfall and the hourly maxi- mum rain intensity from 21 st June 2007, 12.00 pm to 23 rd June 2007, 01.00 pm (note the three rainfall intensity peaks and the anteced- ent rainfall prior to the critical period) 0 2 4 6 8 10 12 14 21 - 11:31 AM 21 - 4:19 PM 21 - 9:07 PM 22 - 1:55 AM 22 - 6:43 AM 22 - 11:31 AM 22 - 4:19 PM 22 - 9:07 PM 23 - 1:55 AM 23 - 6:43 AM 23 - 11:31 AM Time Pore Pressure (kPa) 0 5 10 15 20 25 30 Rainfall (mm) Piezometer 1 Piezometer 2 Piezometer 3 Rainfall Figure 5: Pore pressure response at each pi- ezometer from 21 st June 2007, 12.00 pm to 23 rd June 2007, 01.00 pm Conclusion Preliminary analysis of rainfall and pore pressure data for the south-west and north east monsoon season of 2007 leads to the following infer- ences: • A combination of prolonged and high intensity rainfall and corresponding pore pressure variations is the main trigger of landslides in the region. • Thampi et al., (1998) suggested a minimum rainfall of 300 mm in two days as needed to produce landslides in the region. In light of this research continuous rain spell of 9 hours of the order of 195 mm is sufficient to cause landslides. However, it is too far fetched to say that all rain spells lasting 9 hours amounting to 195 mm will cause landslides, as a rainfall characterization analyzed against corresponding temporal landslide inventory is pending for the region. • Rainfall thresholds also needs to be coupled with the continuous monitoring of pore pressure at representative hollows to be useful in an early warning system for landslides as not all rain spells with the same quantity produces landslides. • Pore pressure measurements in the hollows indicate that persistence of high pore pressure in shallow soils for over 9 hours triggered the mass movements of 2007 in the catchment. • More specifically, it may be the persistence of critical conditions and not the maximum pore pressure attained at the locations which lead to the slope failures on 22 nd June 2007. However, a generalization of the pore pressure thresholds needs long term data or physically based dynamic modelling efforts to be proven useful. Mr. Sekhar Lukose Kuriakose (UNU-ITC), Email: [email protected] Prof. Dr. V.G Jetten (ITC), Email: [email protected] Dr. C.J van Westen (UNU-ITC), Email: [email protected] Dr. L.P.H van Beek (UU), Email: [email protected] Ir. G. Sankar (CESS), Email: [email protected] Thampi, P.K., Mathai, J., Sankar, G. and Sidharthan, S., 1998. Evaluation study in terms of landslide mitigation in parts of Western Ghats, Kerala. Research report submitted to the Ministry of Agriculture, Government of India. Centre for Earth Science Studies, Government of Kerala, Thiruvananthapuram, India. Critical Period Critical Period
Transcript of Pore pressure as a trigger of shallow landslides in the ...The area has a history of shallow...
I N T E R N A T I O N A L I N S T I T U T E F O R G E O - I N F O R M A T I O N S C I E N C E A N D E A R T H O B S E R V A T I O N
For more information: References:
Pore pressure as a trigger of shallow landslides in the Western Ghats
IntroductionPore pressure is considered an important trigger of landslides. Continuous monitoring of pore pressure at locations of landslides is limited by the fact that the initiation locations of landslides are almost never known prior to the occur-rence of the events. This is especially true in the case of shallow landslides such as debris flows. Researchers circumvent this constrain by moni-toring pore pressure in laboratory scaled models, known active and dormant landslides or hydro-logically and topographically similar terrain. This paper presents some initial interpretations of the observations of pore pressure from a network of piezometers coupled with the rainfall-intensity measurements from an automated weather station in the upper Tikovil River basin of the Western Ghats mountain chain of Kerala, India.
Study AreaThe study area comprises Aruvikkal catchment, a 9.5 km2 sub basin of the Tikovil River, a tributary of the Meenachil River (Figure 1). The area is underlain by Precambrian charnockites, which weathers very slowly, leading to rather shallow sandy soils over a thin layer of sparolite inter-leaved by lithomargic clay. The predominant land use is rubber (Hevea Brasiliensis) plantations, which have to be cleared after 20 years, thus exposing the land to the high intensity rainfall. The area has a history of shallow landslides and consequent debris flows that originate in topo-graphical hollows. On 6th October 1993 the re-gion experienced 100 debris flows, the triggering 2 day cumulated rainfall of which was 315 mm; the experimental catchment experienced over 11 landslides on that day. On 22nd June 2007 the catchment and surroundings experienced about 20 landslides, most of which were shallow land-slides such as debris flows, with the catchment experiencing two soil slips and a debris flow on the day (Figure 2).
Figure 1: Study Area – Aruvikkal Basin
Materials and MethodsTwo micro catchments of the Aruvikkal catch-ment were instrumented with 15 piezometers in the period June to September 2007, as part of an ongoing research on establishing the physical basis of slope hydrology and slope stability of the region. Three piezometers of the 15 were de-ployed before the onset of the monsoon. The pi-ezometers used were of the type Keller DCX-22 AA which measures and records ground water levels using a two sensor technology. The sub-mersible depth sensor measures the water level. Barometric pressure variations are measured and
compensated with the built-in waterproof air pressure sensor which is mounted in the elec-tronics housing at the top of the borehole (www.keller-druck.com). Of the 15 piezometers, 4 were installed in hollows, with at least 0.7 m of over-burden thickness, 7 were installed in terraces with rubber plantations, 2 were installed in other landuse types and 2 were installed at drainage outlets. A low cost wireless automated weather station (AWS) branded Vantage Pro2 (www.davisnet.com), capable of measuring all standard weather variables and solar radiation was estab-lished at Meladukkam CSI church (see Figure 1) located in the center of the area at an altitude of 660 m. The logger of the AWS was programmed to record data at an interval of 15 minutes, and could store data for about 45 days given this interval. The data from the 3 piezometers that were installed before June 2007 were analyzed in combination with the AWS data.
Figure 2: Initiation and run out zones of the debris flow (mass movement 1 in Figure 1)
Results and DiscussionDuring the south west monsoon season of 2007 Aruvikkal catchment received 3123 mm rainfall which was 16.3% less than the seasonal LPA rainfall of the area. On the day of the events, the region received 199.8 mm rainfall and the 2 day cumulated rainfall for that day was 278.4 mm. The 2 day cumulated rainfall explains the relatively lesser landslide activity in the region on 22nd June when compared to that of the 6th October 1993 events. Figure 3 shows the 2 day cumulative rainfall and maximum daily rainfall intensity recorded by the AWS during the monsoon seasons of 2007. Figure 4 shows the hourly rainfall and the hourly maximum rain intensity from 12.00 pm on 21st June to 12.30 pm on 22nd June. However, such absolute values
are only of indicative relevance unless substanti-ated by long term observations. In situations where long term temporal landslide inventory is unavailable, physically based dynamic models of slope hydrology coupled with slope stability is an ideal alternative for hazard quantification.
The pore pressure responses of the catchment also follow the general trend of the rainfall. Table 1 provides the topographic and land use charac-teristics of the locations of the three piezometers deployed before the critical date. It also provides the maximum, average of the maximum daily and standard deviation of the maximum daily pore pressure observations of the three piezom-eters during the south west and north east mon-soon seasons. It could be seen from the pore pressure data that the hydrological response of the sub surface started between 8.00 and 9.00 pm on 21st June. From the time of the initial hy-drological response of the pores at around 8.00 pm to the beginning of the critical time (6.00 am, 22nd June) of the landslide events 9 hrs later, the pore pressure values remained high.
Figure 5 shows the pore pressure response at each piezometer during a continuous rain spell from 21st June 12.00 pm to 23rd June 12.30 pm. A close observation of the continuous re-sponse pattern of pressure waves brings out the transient nature of the hydrological response processes. Piezometers 1 (P1) and 3 (P3) behave similarly in their response to the fluctuations of the rainfall, while piezometer 2 (P2) shows a delayed response. P3 shows much more fluctua-tions than P1 which may be attributed to the local characteristics of the location such as satu-rated hydraulic conductivity, soil depth and slope which results in the sudden drainage of excess pore water.
0
50
100
150
200
250
300
5/28
6/22
7/17
8/11
9/5
9/30
10/25
11/19
12/14
Days
Cum
ulat
ed R
ainf
all (
mm
)
0
50
100
150
200
250
300
Rai
n in
tens
ity (m
m/h
r)
Intensity 2 days cumulated
Figure 3: 2 days cumulated rainfall and maxi-mum daily rainfall intensity of Aruvikkal catch-ment during the south west and north east monsoon seasons (note the highest 2 days cumulative rainfall on 22-06-2007)
Table 1: Location characteristics, maximum pore pressure response, average of the maxi-mum daily and standard deviation of the maxi-mum daily pore pressure observations of the three piezometers during the south west and north east monsoon seasons
Pore pressure (kPa) South West North East P A (m) S (º)
SD (m)
KSat (m/hr)
LU/LC Max Avg Std Max Avg Std
1 873.7 46.7 2.6 0.11 MC 8.61 1.31 1.92 5.14 0.29 0.81 2 320.5 24.9 1.3 0.14 R 7.46 0.29 1.11 1.38 0.14 0.24 3 915.8 31.7 1.9 0.13 MF 12.33 4.47 3.49 10.29 1.14 2.59 P – Piezometers; A – Altitude; S – Slope; SD – Soil Depth; Ksat – Saturated hydraulic conductivity; LU/LC – land use/land cover (MC – Mixed crops, MF – Mixed Fallow, R – Rubber). Max – Maximum, Avg – Average, Std – Standard deviation
0
5
10
15
20
25
30
12:30 PM2:30 PM4:30 PM6:30 PM8:30 PM10:30 PM12:30 A
M2:30 A
M4:30 A
M6:30 A
M8:30 A
M10:30 A
M12:30 PM2:30 PM4:30 PM6:30 PM8:30 PM10:30 PM12:30 A
M2:30 A
M4:30 A
M6:30 A
M8:30 A
M10:30 A
M12:30 PM
Time
Rai
nfal
l (m
m)
020406080100120140160180200
Rai
n in
tens
ity (m
m/h
r)
Rainfall (mm) Rain intensity (mm/hr)
Figure 4: Hourly rainfall and the hourly maxi-mum rain intensity from 21st June 2007, 12.00 pm to 23rd June 2007, 01.00 pm (note the three rainfall intensity peaks and the anteced-ent rainfall prior to the critical period)
0
2
4
6
8
10
12
14
21 - 11:31 AM
21 - 4:19 PM
21 - 9:07 PM
22 - 1:55 AM
22 - 6:43 AM
22 - 11:31 AM
22 - 4:19 PM
22 - 9:07 PM
23 - 1:55 AM
23 - 6:43 AM
23 - 11:31 AM
Time
Pore
Pre
ssur
e (k
Pa)
0
5
10
15
20
25
30
Rai
nfal
l (m
m)
Piezometer 1 Piezometer 2 Piezometer 3 Rainfall
Figure 5: Pore pressure response at each pi-ezometer from 21st June 2007, 12.00 pm to 23rd June 2007, 01.00 pm
ConclusionPreliminary analysis of rainfall and pore pressure data for the south-west and north east monsoon season of 2007 leads to the following infer-ences:• Acombinationofprolongedandhigh intensity rainfall and corresponding pore pressure variations is the main trigger of landslides in the region.• Thampietal.,(1998)suggestedaminimum rainfall of 300 mm in two days as needed to produce landslides in the region. In light of this research continuous rain spell of 9 hours of the order of 195 mm is sufficient to cause landslides. However, it is too far fetched to say that all rain spells lasting 9 hours amounting to 195 mm will cause landslides, as a rainfall characterization analyzed against corresponding temporal landslide inventory is pending for the region.• Rainfallthresholdsalsoneedstobecoupled with the continuous monitoring of pore pressure at representative hollows to be useful in an early warning system for landslides as not all rain spells with the same quantity produces landslides.• Porepressuremeasurementsinthehollows indicate that persistence of high pore pressure in shallow soils for over 9 hours triggered the mass movements of 2007 in the catchment.• Morespecifically,itmaybethepersistence of critical conditions and not the maximum pore pressure attained at the locations which lead to the slope failures on 22nd June 2007. However, a generalization of the pore pressure thresholds needs long term data or physically based dynamic modelling efforts to be proven useful.
Mr. Sekhar Lukose Kuriakose (UNU-ITC), Email: [email protected]
Prof. Dr. V.G Jetten (ITC), Email: [email protected]
Dr. C.J van Westen (UNU-ITC), Email: [email protected]
Dr. L.P.H van Beek (UU), Email: [email protected]
Ir. G. Sankar (CESS), Email: [email protected]
Thampi, P.K., Mathai, J., Sankar, G. and Sidharthan, S., 1998. Evaluation study in terms of landslide mitigation in parts of Western Ghats, Kerala. Research report submitted to the Ministry of Agriculture, Government of India. Centre for Earth Science Studies, Government of Kerala, Thiruvananthapuram, India.
Critical Period
Critical Period
