Original Paper Landslides (2012) 9:1–11 DOI 10.1007/s10346-011-0280-x Received: 2 March 2011 Accepted: 5 July 2011 Published online: 24 July 2011 © Springer-Verlag 2011

Kazuo Konagai I Ahsan Sattar

Partial breaching of Hattian Bala Landslide Dam formed in the 8th October 2005 Kashmir Earthquake, Pakistan

Abstract The 8th October 2005 Kashmir Earthquake of magnitude 7.6 triggered a huge landslide 3.5km upstream of Hattian Bala town in the state of Azad Jammu Kashmir of Pakistan. The debris mass blocked two tributaries of the Karli branch of the Jhelum River and was breached on 9th February 2010. This debris dam provided us with a rare opportunity to keep careful and continuous eyes on its post-earthquake behavior especially as it was a serious threat to people living along the lower reaches of both the Karli and Jhelum Rivers. This paper describes postformation behaviors of the debris mass, breaching-inflicted changes of not only the debris mass but also both upstream and downstream reaches based upon laser-scanned images of landforms and Differential Global Positioning System survey results. Keywords Hattian dam . Breaching . Landslide dam . Flood . Dam-break . Drawdown Introduction The 8th October 2005 Kashmir Earthquake of magnitude 7.6 (USGS) triggered a huge landslide 3.5 km upstream of Hattian Bala town in the state of Azad Jammu Kashmir of Pakistan. The debris deposit blocked two tributaries of the Karli branch of the Jhelum River forming Karli Lake (Large Lake) and Tang Lake (Small Lake). This landslide dam (called the Hattian dam) became a major concern for the people living along the lower reaches of both the Karli and the Jhelum Rivers. Dunning et al. 2007 and Schneider 2008 discussed the geological features of the landslide dam and its surrounding mountain terrains and the potential hazards posed by the landslide dam. Sattar et al. 2010 discussed the potential risk of flooding following such landslide dam breaching and observed deformations / erosions that had been occurring to the debris deposit since the debris dam was formed. The dam was sustained over a period of about four years and four months, and finally the water of Karli Lake breached the northwestern part of the dam on 9th February 2010 following a continuous 5-day rainfall. Later on, the water of Tang Lake also breached the northeastern lobe of the dam in the devastating monsoon rains of July to August of 2010. This paper describes the findings obtained from the authors’ surveys conducted before and after the dam failure, and discusses possible failure scenarios based on these findings. Post-formation features of debris deposit The 65 million m3 Hattian landslide mass originating from the Dana hill ran down the valley slope and deposited its final volume of 85 million m3 along the Karli River, thus impounding the two lake waters of 62 million m3 for Karli Lake and 5 million m3 for Tang Lake (Fig. 1). A time line for the various incidences at Hattian Bala landslide dam is given in Fig. 2. Owing to the potential danger of water outbursts from the Karli Lake, the

relevant authorities decided to hastily excavate a spillway to secure the safety of the dam. The excavation of the spillway considerably reduced the total natural storage capacity of the Karli Lake from 86 to 62 million m3. As a part of studies for this mitigation measure, the discharges flowing into the lakes and seeping out at the toe of the landslide dam were measured by the Water and Power Development Authority (WAPDA) of Pakistan and published in WAPDA report 2007. The water levels of the lakes rose and the water of Tang Lake started overflowing in February 2006 where a relatively short, 130-m long spillway had been excavated until that time across the northeastern lobe of the landslide dam. Not long after, the water level of Tang lake gradually reduced by 7–8 m over a period of 2 months (from late April to early June 2006, Sattar et al. 2010) and later on reached a state of equilibrium between the inflow and seepage discharges and remained steady afterwards, leaving approximately 3.5 million m3 of water. The debris mass started to show signs of gradual deformations/erosions after the overflow started through the spillway. Figure 3 shows gradual changes that were occurring along the path of overflowing water from the Tang Lake. A 450-m long spillway was excavated in front of the Karli Lake whose water started overflowing in April 2007. The overflowing water from the lake was being confined within the rock-covered spillway walls over its short stretch of about 450 m. Beyond this stretch, the water flowed over the unprotected surface of the landslide dam with its discharge being decreased as it flowed down over the dam body. Waters were again found seeping out at various places at lower elevations. After confirming the successful performance of the rock-covered spillway of the Karli Lake, the monitoring responsibility of the site was handed over to the Azad Jammu Kashmir government. Afterwards, less attention was drawn towards the monitoring of the dam site as the spillway was performing well and no immediate threat was imminent. The authors group had been conducting measurements for quantifying the changes occurring in the landslide dam with the support of both the Japanese and Pakistani relevant organizations obtaining the grant-in-aid from the Japanese Ministry of Education, Culture, Sports, Science and Technology since the Japanese fiscal year of 2008 (MEXT project no. 20254003). Various techniques including Differential Global Positioning System (DGPS) measurements, laser scanning, etc, were employed to quantify the deformations. The most significant change observed over the surface of the landslide dam was a sudden detachment of an approximately 65,000 m3 soil mass from the downstream slope of the landslide dam (Figs. 4 and 5; Sattar et al. 2010). The eroded gulch appeared during heavy rain in 2009, and therefore showed the potential risk of breaching. The DGPS measurements conducted over a six months interval (June 2009–Novermber 2009) showed that the crest zone of the dam settled by the order of 100 mm and the zone near the head of the eroded gulch was Landslides 9 & (2012)

1

Original Paper

(a)

To Muzaffarabad

Pakistan

Jhelum River

Source area Dana Hill

Karli River Fig. 1b

Approx. location of active Balakot Bagh fault

Tang Lake catchment

Karli Lake

Tang Lake

Hattian dam

Karli Lake catchment

Quickbird (Date 27/10/2005) 0

500

(b)

1000m

Debris mass

Tang Lake (Small Lake) Karli Lake (Large Lake) Fig. 1 a Map of Hattian area showing the source area of Hattian landslide mass, Hattian landslide dam, Karli Lake, and Tang Lake with their catchment areas, approximate location of the active Balakot-Bagh fault (Nakata et al. 1991), and the main tributaries of the Karli River and Jhelum River. b Quick bird image (27 Oct 2005) of the Hattian dam

uplifted by a few millimeters (Fig. 6). Meanwhile, gradual changes were being observed along the northeastern lobe of the debris deposit where the water from the Tang Lake was seeping into the deposit of segregated large boulders and was emerging at some distances downstream. Over the course of time, it was suspected that pipes were being developed in the interior of the debris mass

2

Landslides 9 & (2012)

as indicated by the gradual surface deformations that were observed over the surface of the deposit (Fig. 3). Failure of Karli Lake After five days of incessant rains in early February 2010, the water of the Karli Lake breached the debris dam on 9th February 2010.

2005 Kashmir Earthquake, Landslide trigger

Tang Lake Breaching

Karli Lake Overtopping

Piping in front of Tang Lake Tang Lake Filling Tang Lake Overtopping

Gradual Toe Erosion

16-Mar-10

17-Sep-09

22-Sep-08

26-Mar-08

28-Sep-07

1-Apr-07

3-Oct-06

6-Apr-06

8-Oct-05

Excavation of spillway

21-Mar-09

GPS Measurements

Karli Lake Filling

Erosion Gully appeared

GPS Measurements Karli Lake Breaching

Fig. 2 Timeline of various events that occurred at Hattian Bala landslide dam site

This breaching drained about 36 million m3 of water from the lake and left 26 million m3 of water in the lake. The erosion stopped after the bedrock was exposed at the scoured bottom of the breach channel (Fig. 7a). Figure 8 compares the laser-scanned landform of the breached landslide dam (April 2010) with the one before the breaching. From this figure, about 7.78 million m3 of soil was estimated to have been eroded during the breaching event. The flushed debris mass was deposited along the Karli River (Fig. 9). The thickness of the debris deposit was estimated by measuring the top widths of the deposit at about 500 m regular intervals and superimposing them on the digital elevation model (DEM) of the Karli River Valley, which had been acquired before the breaching event. Figure 10 shows the Karli river profile before and after the breaching. A major portion of the eroded material is deposited over the approximately 2 km downstream stretch from the dam along Karli River and less significantly beyond the stretch. Figure 11 shows that precipitation estimated for the Hattian catchment with the data from Satellite “Tropical Rainfall Measuring Mission” (TRMM, 2011), over the period since the overtopping of Karli Lake water until the breaching. TRMM is a satellite mission lunched jointly by the Japan Aerospace Exploration Agency and National Aeronautics and Space Administration. 3B42 V6 TRMM, a product, with 0.25×0.25° spatial and three-hourly temporal resolutions, was used for rainfall estimate. The TRMM-

data-based estimate shows a several-days continuous precipitation in early February immediately before the breaching of the Karli Lake. This precipitation is relatively moderate as compared to the other significant precipitations that the area had previously suffered. The estimate shows a peak value of 32 mm/day on 8th February, 1 day before breaching. However, only a part of this precipitation may have flowed into the lake because the slopes surrounding the Hattian Dam were all covered with snow, as shown in Fig. 12, when the breaching occurred. Since the greater part of February 2010 precipitation may have fallen as snow, there was a question if the discharge through the spillway of Karli Lake was sufficient to trigger the breaching of the deteriorated dam body. Given no hydraulic data measured during the breaching process, past records were reviewed to discuss the possible water depth at the spillway. As was explained in the previous section, WAPDA (2007) had been measuring the discharge into the Karli Lake since November, 2005 till June, 2006. During this period of time, several discharge peaks in winter (Fig. 13a) were compared with those estimated from TRMM data. A number of authors compared precipitations estimated from TRMM data with raingauge measurements. Their results are confined to particular sites and are not necessarily representative of results from other regions. As for the Himalayan regions, Yatagai and Kawamoto (2008) compared the Precipitation Radar (PR) data obtained from TRMM with daily precipitation records in Nepal, Bangladesh, Bhutan, Pakistan, India, Myanmar, and China using a dense network of rain gauges and concluded that (1) PR data from TRMM systematically underestimated precipitation by 28% to 38% in summer (July–September), (2) significant correlations were found for all months, but it was relatively low in winter, and (3) the relationship was investigated in different altitude zones, and PR underestimated rain gauges for most levels, except for February (250–1,000 m), March (0–1,000 m), and April (0– 1,500 m). Focusing on the vicinity of the Hattian landslide dam, there are two sets of gauge data at Garhi Duptta, about 20 km northwest of Hattian, and Muzaffarabad (Table 1). Daily precipitations of 87 and 75 mm/day on 9th February 2010 at Garhi Duptta and Muzaffarabad, respectively, are much larger than 46 and 44 mm/day estimated for these two locations from TRMM data (Table 2), indicating that even in February, it is likely that PR underestimate rain gauges at Hattian and its vicinity. As the information states, it was tentatively assumed that PR underestimated rain gauges at Hattian by 50% to discuss the maximum credible increase in the water level at the spillway. With the

Breach channel in front of Tang lake

15th November 2006

28th June 2008

8th Dec 2010

Fig. 3 View of the path of water flowing from the Tang Lake. The surface change can be observed between November 2006 (left) and June 2008 (middle). While the right most figure shows the breach channel formed in front of Tang Lake. Photo points are approximately N 34.1420°, E 73.7347° for the left and middle and N 34.1427°, E 73.7326° for the right

Landslides 9 & (2012)

3

Original Paper Karli Lake Hattian Landslide Dam

Spillway

Eroded portion from downstream face

Fig. 4 Front view of the landslide dam. Rim of the eroded region on the downstream face of the dam is highlighted. Photo points are approximately N 34.13733°, E 73.72792°

TRMM estimated precipitation in February in 2006, a runoff analysis was conducted using the two-dimensional runoff model by O’Brien et al. (1993). The assumptions for the runoff analysis are 1. Square grid of 150 m on a side is used. 2. Uniform Manning’s roughness coefficient of 0.15 was assumed for the entire terrain with the greater part of the catchment area covered with vegetation and debris exposed along stream paths as per Hydrologic Engineering Center manual. 3. Neither infiltration into the ground nor interception by the vegetation is assumed for all Flo-2d analysis. There were several peaks of discharges estimated for February 2007 (Fig. 13b), among them; the minimum overall loss of 50%, was reached on February 26th. The loss can be due to infiltration, evaporation, interception etc. If the greater part of precipitation has fallen as snow, the loss would have been much larger as was shown in the other peaks in February, 2007. Assuming that this minimum percentage of loss was realized in the rains/snows preceding the breaching event, a relatively moderate peak discharge of 10 m3/s at the spillway was estimated to have been reached on the morning of 9th February 2010. This relatively moderate discharge is not likely to cause a large scale

Fig. 5 Contour plot of the erosion gully developed between November 2008 and June 2009. After Sattar et al. 2010

4

Landslides 9 & (2012)

Fig. 6 Settlements of debris deposit. After Sattar et al. 2010. Universal Transverse Mercator coordinate system is used

breaching erosion. Above all else, the debris dam had been surviving in much heavier rainfalls before its failure. The possibility that lakefront landslides may have caused an increase in the lake water level and/or generated tsunami waves across the lake, and eventually increased the spillway discharge needs to be discussed. A large landslide mass was observed on the right bank slope of the Karli lake (Fig. 14), and there were two secondary landslides that originated from the debris mass (Fig. 15). However, as for the right bank slope (Fig. 13), witnesses said that the landslide movement started after the drawdown of the lake, and continued for two weeks. A photo of the lakefront part of the right bank slope (Fig. 16b) taken on the next day of the breaching shows that the water line that appeared as the clear boundary between the snow-covered onshore slope and dark brown exposed slope does not exhibit any visible change from the one in Fig. 16a taken on August 20th 2008. The affected area is about 1,000 m long in a run-out direction and 800 m wide across. The landslide destroyed about 174 houses and displaced 1,000 residents. As for the secondary landslides with their uppermost scars at about the same level as the lake water level before the drawdown, their volumes (<100,000 m3) do not seem to be substantially large enough to cause significant water displacement in the lake. Moreover, whether these slides happened before the breaching event or after the event was not witnessed. As shown above, a sudden lake drawdown can disturb the equilibrium conditions in the interior of a lakefront slope in such a way that the pore pressure gradient in the slope increases, leading to the increase in seepage forces, thus boosting the risk of the slope failure. Needless to say, the lake drawdown rate must have been largely affected by the down-cutting rate of the downward erosion of the debris dam. Therefore, an attempt was made to estimate the down-cutting rate from flood marks remaining along Karli River. The flood wave from the breached landslide dam scoured vegetation on slopes along the Karli valley and damaged about 24 houses near the confluence of the Karli and Jhelum Rivers. The

Fig. 7 a Front view of Hattian landslide dam, before and after failure. The exposed bed rock is visible under of the breach channel. Photo points are N 34.1562°, E 73.74496°. b View of Karli lake and breach channel, 1 day after breaching of dam. Snow covering the surrounding slopes of the dam. Photo by Dr. Kausar. Photo points are approximately N 34.13733°, E 73.72792°

(a) November, 2009

April, 2010

(b)

flood level was estimated by measuring the heights of the flood marks. Over some distance downstream of the dam, the flood marks were not visible, or rather thick debris deposited as shown in Fig. 10. Figure 17 shows the measured flow depths along the Karli River along with the flow velocities estimated from superelevations of the measured mud lines according to Johnson and Rodine (1984). The outflow in the Karli valley is simulated using the twodimensional flood routing model by O’Brien et al. (1993). The outflow hydrograph for the down-cutting rate of 25 m/h with peak outflow discharge of 5,500 m3/s (Fig. 18) turned out to be the best fit to the measured flow depths and flow velocities along the Karli River in Fig. 17. Figure 19 shows the simulated flow depth and flow velocities corresponding to the inversely determined outflow hydrograph.

Failure of Tang Lake The monsoon rains in 2010 caused a mega flood in Pakistan and the flood is considered as the worst in record of the region. Intensive rainfall was observed in Muzaffarabad and surrounding areas, including Hattian Bala. Continuous rainfall reportedly caused extensive landslides along the Neelum valley, and to a lesser extent, along the Jhelum valley. The Tang Lake (smaller lake) breached during the days of the most intensive monsoon rains (27st July to 1st August), however the exact day of breaching remains unknown. A peak rainfall rate of 105 mm/day was estimated for 30th July from the TRMM data (Fig. 20). Clearly this peak is the highest compared to the others that have occurred over the Hattian catchment since April 2007 (Fig. 11). Moreover, the actual rain-gauge precipitation could have been higher than the TRMM-data-based estimate as discussed

Fig. 8 Contour plot of Hattian dam before the failure generated from 5-m resolution DEM and contour map of the breached dam generated by laser scanning of the Karli lake breach channel

Landslides 9 & (2012)

5

Original Paper February, 2010

mm/day

9Feb-

7Feb-

25

4-Jan-10

4-Oct-09

4-Jul-09

4-Apr-09

4-Jan-09

4-Oct-08

4-Jul-08

4-Apr-08

4-Jan-08

4-Oct-07

4-Jul-07

0 4-Apr-07

Debris deposit

50

4-Jan-07

Karli River

3Feb-

75

40 30 20 10 0 5Feb-

mm/day

Breached Hattian dam

Fig. 11 TRMM-data-based precipitation estimate for Hattian catchment from overtopping (April, 2007) till failure (February, 2010) Fig. 9 Partially breached Hattian dam and debris deposit along the Karli River. Photo taken on 6th June 2010 at N 34.156567°, E 73.744833°

above. In the breaching process, the water from Tang Lake eroded the northeastern end of the landslide lobe, which had been covered with boulders segregated up at the time of deposit, and approximately 15 m depth of the lake water was drained leaving 1.9 million m3 water in the lake (Fig. 21). As mentioned earlier, piping is considered to have been developed in the interior of the debris deposit in front of Tang Lake, and it is inferred that the breaching process initiated as a backward erosion piping. The down-cutting rate of the Tang Lake seems to have been slow because the boulders which were present along the water path were carried downstream without extensive crushing as was observed in the event of the Karli Lake failure. The slow downcutting rate of Tang Lake can be attributed to the armoring effect

Fig. 10 Karli River bed profile before and after the breaching of Hattian dam. GPS point 12 marked on right figure is at the toe of the landslide dam and point 41 at the junction of Jhelum River and Karli River

6

Landslides 9 & (2012)

provided by the surface boulder layer. No clear flood water marks were found for this event probably because the discharge of the flood water was much smaller than that from the Karli Lake, and actually many residents at Hattian Bala knew that Tang Lake had failed only after the authors reported it. Discussion It is clear that Karli Lake failed immediately after a runoff from the rain/snow fall in early February 2010. However a question remains about how this moderate precipitation may have caused the breaching, as the debris dam had survived much heavier rainfalls before its failure, and actually, the lake water mark that appeared as a lower horizontal bound of the lightly snow-dusted lakefront slope did not show any clear sign of the increase in the lake water level. A possible mechanism that caused the breaching

Table 1 Rain gauge data for Garhi Dupatta and Muzaffarabad City, for the Month of February 2010

Rain gauge data (mm/day) February 2010

5

6

7

8

9

10

Garhi Dupatta

0

25

22

49

87

17

Muzaffarabad

0

35

27

40

75

26

Data acquired from Pakistan Metrological Department. Source: http://www.pakmet.com.pk/

Fig. 12 View of Hattian dam after breaching, showing the snow covering the surrounding slopes of the catchment area. Photo from Dr. Allahbukhsh Kausar. Photo taken on 10th February 2010 at approximately N 34.14612°, E 73.74196°

can be explained considering the weak weathering resistance of the debris material. The debris material originated from the Murree formation of Miocene-aged sedimentary rocks composed of reddish mudstones, shales and grayish sandstones. The effect of weathering on the mudstones had been observed at various locations over the dam body, where boulders of mudstone were slaked and reduced to fragments (Fig. 22). Reddish water marks, which were found from time to time along paths of waters seeping out of the dam body, clearly indicated that finer substances had been washed out due to internal seepage flows. Kiyota et al. (2011) performed a series of direct shear tests on the mudstone samples taken from the Hattian landslide dam. Keeping normal and shear loading stresses constant to examine creeping behavior of a sample, he observed that as soon as the dried specimen was soaked up with water in the middle of the test, it exhibited a sudden increase in the deformation rate. The Hattian area experiences extreme weather conditions during summer and winter along with the monsoon season with heavy rainfalls, Discharges through Large and Small Lakes WAPDA Report August 2006 6

3

D is c harge, m /s

(a)

26th Feb.

Inflow large Lake

4

Inflow Small lake

2

Seepage from Landslide dam

0 3-Jan-06

(b)

23-Jan-06

12-Feb-06

4-Mar-06

24-Mar-06

TRMM data of Karli Catchment

m m /da y

30 20

setting ideal conditions for wetting and drying cycles on the slaking process. Signs of a deteriorating dam body were being observed with the gradual surface deformations and later on with the sudden detachment of a significant amount of soil from the downstream face of the dam, as discussed above. In February 2010, a relatively moderate precipitation extending over a period of 5 days in the Hattian catchment followed dry and clear days in February, and thus may have accelerated the slaking process in a way that even moderate waters flowing over and seeping through the debris mass had a significant effect in accelerating backward erosion. Though this scenario is qualitatively consistent with the observed phenomena, further investigation of the exposed wall of the landslide dam is essential, and is now planned. Conclusions The water of Karli lake (Large lake) of Hattian dam breached the northwestern part of the debris deposit on 9th February 2010 after five days of incessant rains/snowfalls, and the water of Tang lake (Small Lake) breached the northeastern lobe of debris deposit during the monsoon rains of July–August, 2010. Until the failure of Karli Lake, the landslide dam had been a serious threat to people living along the lower reach of the Jhelum River, and periodic observations and measurements of changes occurring in the landslide mass had been conducted since the authors’ project started in 2008. The landslide mass started to show clearer signs of gradual deformations/erosions after the overflow started through the spillway, which was excavated hastily to secure the safety of the dam against possible failure of the Karli Lake. Evidence of weathering (slaking) of the mudstones of Murree formation was observed over the boulder-strewn landslide dam. The most significant change observed in the dam body before the failure of Karli Lake was a sudden detachment of an approximately 65,000 m3 soil mass from the downstream slope of the landslide dam during a heavy rainfall in 2009. TRMM-data-based rainfall estimate over the Hattian catchment area indicates that the precipitation of February 2010 that caused the breaching was less significant than the other heavy rainfalls that the area had

Table 2 TRMM-data-based precipitation estimate for Garhi Dupatta and Muzaffarabad City, for the Month of February 2010

10

TRMM data (mm/day) 0 3-Jan-06

23-Jan-06

12-Feb-06

4-Mar-06

24-Mar-06

Fig. 13 a Measured discharges into the Karli Lake, Tang Lake and seepage through the body of Hattian dam (photo from WAPDA 2007). b TRMM-data-based rainfall estimate for the same period of time as for (a)

February 2010

5

6

7

8

9

10

Garhi Dupatta

1

5

7

15

46

8

Muzaffarabad

2

6

56

12

44

9

Landslides 9 & (2012)

7

Original Paper Fig. 14 a Photo of the landslide on the right bank of the Karli lake. Photo points are approximately N 34.13554°, E 73.71191°. Contour map generated by laser scanning of the landslide mass. b Longitudinal section of the landslide mass

(a)

A’

A

Karli Lake

N

A’

N34.12257

A

E73.70959

200m

Elevation, [m]

(b) 1700

A

A’

1200 0

experienced since the overflowing from Karli Lake started in April 2007. Though it was suspected that a lakefront slope failure has caused tsunami wave across the lake, a photo of the lakefront part of the right bank slope taken on the next day of the breaching showed that the water line that clearly appeared as the lower horizontal bound of the snow-dusted onshore slope did not exhibit any visible change in the lake water level. It was remarkable that the drawdown of the lake triggered a large landslide on the right bank of the lake, and the mass movement continued for two weeks. This coherent landslide mass was about 1,000 m long in run-out direction and 800 m wide

Fig. 15 a Photo of the secondary landslides triggered from the debris deposit and GPS coordinates of photo point N 34.13277°, E 73.69878°. b Contour map generated by laser scanning of the secondary slides

500 1000 Distance, [m]

1500

across. The landslide destroyed 174 houses and displaced about 1,000 residents, making it the larger damage than the one directly caused by flooding. The lake drawdown rate must have been largely affected by the down-cutting rate of the breach channel. The outflow hydrograph with peak outflow discharge of 5,500 m3/s turned out to be the best fit for the flow depths and flow velocities estimated from mud marks remaining along the Karli River. As contrasted with the Karli lake failure, the Tang Lake failure did not produce any significant flood wave; indeed, many residents at Hattian Bala knew that Tang Lake had failed only after the authors reported it. This is probably because the

(a)

(b) Hattian Dam

Water level mark before breaching Karli Lake (after drawdown) N34.13371° E73.7253°

8

Landslides 9 & (2012)

100m

Discharge, [cumecs]

(a)

6000 4000 2000 0 0

2

4

6

Time, [hrs] 20/08/2008

Water level

Fig. 18 Inversely analyzed breach outflow hydrograph

Toe part of the landslide mass before breaching

(b)

Photo from Dr. Allahbukhsh Kausar 10/02/2010

Water level

Toe part of the landslide mass after breaching

Fig. 16 Right bank slope a on 20 Aug 2008, before the breaching event and b 1 day after the breaching event showing snow-covered slope and mark of the water level of Karli Lake before breaching event

total amount of the emptied water (1.6 million m3) was much smaller than that from the Karli Lake and the lobe of the debris mass that had been stopping the water of Tang Lake was thickly covered with segregated boulders. The potential flood hazard for the downstream reach along Karli Valley is now rather limited because of the fact that a bedrock has been exposed on the scoured bottom of the breach

channel which is not likely to be eroded any further, however, the Hattian Bala landslide dam and its vicinity is not yet completely out of danger. At the Hattian landslide dam, landslides (triggered by earthquake or heavy rainfall or otherwise) from the exposed bare walls of the breach channel can clog up the channel at any time and increase the lake volume. Soil masses continually detached from the exposed wall of the Hattian landslide dam will be fed into the stream causing some visible changes in river bed elevations. The lakefront landslide on the right bank of Karli Lake, whose movement started after a sudden drawdown of the lake water, has emerged as another hazardous location. The authors should finally add that this right bank landslide is to be taken as a warning story about potential risks of landslides, which can be induced by sudden reservoir drawdown. Currently, an imminent threat is being reported about the Atabad landslide dam (36.308001°, 74.821529°) that was formed on 4th January 2010 at 80 km upstream of Gilgit city, creating a 22-km-long lake that inundated several villages and submerged 5 km of the Karakoram Highway. Careful and continuous eyes are to be kept on any suspicious movements of not only landslide masses but also lakefront slopes.

Fig. 17 a Flood inundation heights along the Karli River. b Houses affected by the debris originating from the Hattian dam. c Flow velocities estimated from super-elevations of the measured mud marks near the junction of Karli River and Jhelum River

Landslides 9 & (2012)

9

Original Paper

Fig 22 A boulder of mudstone weathered (slaking) to form pile of soil over the body of the Hattian dam

150

Hattian catchment TRMM estimate for July - August 2010

100 50 1-Aug

31st July

30th July

29th July

28th July

0 27th July

Rainfall, [mm]

Fig. 19 a Maximum flow depth and b maximum flow velocities for the outflow hydrograph corresponding to 25 m/h down-cutting rate, having peak outflow discharge of 5,500 m3/s; 10-m computation grids extracted from 5-m resolution post-earthquake DEM data

Fig. 20 TRMM-data-based precipitation estimate of Hattian catchment during the monsoon rains (27th July to 1st August 2010)

Acknowledgments This paper summarizes one of the outcomes of the MEXT Research Project, “Scientific surveys for long-lasting geotechnical problems caused by large earthquakes and their implementations for rational rehabilitation strategies,” Konagai K. Leader of the project, 2008 Grant-in-aid for scientific research (A) no. 20254003, Ministry of Education, Culture, Sports, Science and Technology (MEXT). The authors would like to express their sincere thanks to officers and experts at the Geological Survey of Pakistan, the State Earthquake research and Rehabilitation Authority, the Development Authority Muzaffarabad, and Earthquake Research and Rehabilitation Authority for their collaboration and help during the field surveys and devotion towards these long-lasting issues and their mitigation strategies. The authors are also thankful to officers and experts at Japan Embassy in Islamabad, Islamabad Office of Japan International Cooperation Agency, and Islamabad Office of Tobishima Corporation not only for their willingness to support the authors’ surveys but also for providing the team with every possible logistic convenience. Special thanks are due to Mr. Takaaki Ikeda at Tobishima Research Institute of Technology, Dr. Takashi Kiyota and Mr. Zaheer A. Kazmi, Association Professor and Ph.D. candidate, respectively, at the Institute of Industrial Science, University of Tokyo, who have devoted much of their energies to field surveys and analyzing the results.

References

Tang Lake 15m Approx. Breach channel

Fig. 21 Tang lake after breaching (photo taken on 6th Dec 2010) GPS coordinates N 34.13951°, E 73.73631°

10

Landslides 9 & (2012)

Dunning SA, Mitchell WA, Rosser NJ, Petley DN (2007) The Hattian Bala rock avalanche and associated landslides triggered by the Kashmir Earthquake of 8th October 2005. Eng Geol 93(3–4):130–144. doi:10.1016/j.enggeo.2007.07.00 Johnson AM, Rodine JR (1984) Debris Flow. In: Brunsden D, Prior DB (eds) Slope Instability. Wiley, Chichester, pp 257–361 Kiyota T, Konagai K, Sattar A, Kazmi ZA, Okuno D, Ikeda T (2011) Breaching failure of a huge landslide dam formed by 2005 Kashmir earthquake. Soils and Foundations. (accepted for publication in Vol. 51, No. 6) Nakata T, Tsutsumi H, Khan SH and Lawrence RD (1991) Active faults of Pakistan, map sheets and inventories. Research Center for Regional Geography, Hiroshima University, Special Publication 21, pp. 1–141 O’Brien JS, Julien PY, Fullerton WT (1993) Two-dimensional water flood and mudflow simulation. J Hydr Engrg 119(2):244–261. doi:10.1061/(ASCE)0733-9429(1993)119:2 (244)

Sattar A, Konagai K, Kiyota T, Ikeda T, Johansson J (2010) Measurement of debris mass changes and assessment of the dam-break flood potential of earthquake-triggered Hattian landslide dam. Landslides J. doi:10.1007/s10346-010-0241-9 Schneider JF (2008) Seismically reactivated Hattian slide in Kashmir, Northern Pakistan. J Seismol. doi:10.1007/s10950-008-9103-5 Tropical rainfall measuring Mission (TRMM) A joint mission between NASA and the Japan Aerospace Exploration Agency (JAXA). Available from http://trmm.gsfc.nasa. gov/. Accessed 06 June 2011 WAPDA Report (2007) Study of Hattian Ballah landslide, Potential hazards of land sliding and mitigation measures at Hattian Ballah and other earthquake hit areas, National Engineering Services Pakistan limited (NESPAK) and Geological Survey of Pakistan (GSP), Water and Power Development Authority (WAPDA), 2006, Government of Pakistan Yatagai A., Kawamoto H. (2008) Quantitative estimation of orographic precipitation over the Himalayas by using TRMM/PR and a dense network of rain gauges. In: Remote

Sensing and Modeling of the Atmosphere, Oceans, and Interactions II, doi: 10.1117/ 12.811943

K. Konagai ()) : A. Sattar Institute of Industrial Science, University of Tokyo, Tokyo, Japan e-mail: [email protected] URL: http://shake.iis.u-tokyo.ac.jp/home A. Sattar e-mail: [email protected]

Landslides 9 & (2012)

11