The massive landslide created a high dam across the Sun Koshi River. A river gauging station of the Department of Hydrology and Meteorology (DHM) at Pachuwarghat downstream of the landslide dam showed a rapid decline in water flow three hours after the landslide, after which the flow of water completely stopped for approximately 12 hours (Figure 1). An inflow of about 160 m3/sec of water quickly created a large lake behind the dam. Within 13 hours the newly formed lake – which rapidly grew to a volume of an estimated 7 million cubic metres – extended about 3 km upstream, completely submerging the 2.6 MW Sanima Hydropower station. Had Nepal’s security forces not taken timely action to release some of the stored water through controlled explosions, the backwater would have extended further upstream and caused great damage in Barabise, the nearest upstream town. However, the risk that the dam will breach still remains, bringing with it the threat of a catastrophic flood. The Home Ministry has declared the area a ‘flood crisis zone’, and has issued a warning to communities downstream, with many vulnerable villages being evacuated.
It cannot be predicted when and how the landslide dam will erode and how the stored water will be released. However, it is probable that the Arniko Highway, a major trade link between China and Nepal with exchange that stands at nearly at NPR 38 million per day (nearly USD 400,000), will remain blocked for a long time. This could mean serious medium-term impacts for Nepal. Damage from the landslide has already interrupted power supply from several hydropower plants in the valley, including the Sun Koshi and Bhote Koshi power plants, contributing to the country’s scheduled power cuts.
This is not the first time the Sun Koshi valley has experienced a lethal flood, and this is certainly not the last time. Like many places around the Hindu Kush Himalayan region, the Sun Koshi’s weak geological formation and steep topography combined with frequent intense rainfall events and the increasing impact of climate change makes it prone to different types of water-induced hazards, including landslides.
The valley is also vulnerable to the outburst of growing glacial lakes located in the northern part of the catchment in the Tibet Autonomous Region of China. In 1981, the Zhangzangbo glacial lake located 20 km upstream from the Nepal-China border breached; the resulting flood caused extensive damage in Nepal from the border down to the village of Dolalghat.
Although we cannot control natural hazards like landslides and floods, there are many things that can be done to minimize their adverse impact on lives, livelihoods, and valuable infrastructure. More efforts to map landslide risks are needed, and much more frequent monitoring of potential landslide sites is necessary. Both will help in designing mitigation measures and reducing risks.
In hindsight, photo documentation from 2013 shows a number of scars along the mountain slope in Jure. If there had been an appropriate monitoring mechanism in place, measures could have been taken to raise awareness about the potential of a larger land slip. While the exact timing and size of landslides are difficult to predict, potential landslide areas can be mapped relatively accurately and the approximate size of the potential landslide can be calculated.
Over the last 30 years, the Sun Koshi valley has experienced three major floods. In 1982, a landslide dam outburst flood (LDOF) in the Balephi River, a tributary of the Sun Koshi, killed 97 people. Another flash flood event in 1987 killed 98 people, and a 1996 flood swept away Larcha village, killing 54 people.
Landslides and other natural disasters are also common outside of the Sun Koshi valley. Nepal alone experienced 13 large landslide events between 1967 and 2010, the most recent being the 2010 Madi landslide in Central Nepal. Recent natural disasters across the rest of the Hindu Kush Himalayan region include the Swat valley flash flood and Attabad landslide disaster in Pakistan in 2010, the disastrous 2012 Seti flash flood in Nepal, the Uttarakhand disaster in July 2013, and the landslide in Badakhshan, Afghanistan in May this year.
Knowledge and information from past disasters can also support disaster risk management. Regular monitoring of hydrological and meteorological variables generates valuable information that can be fed into hydrological models. These models can be used to provide information about areas at risk of inundation during a flood event, including for potential glacial lake outburst floods. This type of analysis can be used in zoning river corridors and preparing land use plans.
While zoning and land use planning are essential elements of risk management, if they are not properly implemented, these efforts are futile. Despite an entire village being washed away by the 1996 Larcha flood, a village has been resettled in the exact same location. With these settlements constructed along the flood plain, many households remain at risk of being destroyed in future floods. Even commercial enterprises are taking calculated risks, with a mini-hydropower project now constructed in Larcha. In areas with significant commercial activity like Khadichaur, the construction of settlements along the flood plain has increased in the past decade.
Across the Hindu Kush Himalayan region, improper or insufficient planning of infrastructure and settlements has put unnecessary lives and investments in harm’s way. It is believed that unregulated and haphazard development is partially to blame for the severity of the 2013 Uttarakhand disaster. During such events infrastructure may also create additional risks, for example when stored water is released from hydropower reservoirs into already full river channels.
Experience from the 1981 GLOF event in the Sun Koshi valley has shown the value of proper planning. Following the 1981 flood, more than USD 3 million was spent to rehabilitate the Arniko Highway. During this process, the 27 km stretch of road damaged by the flood was raised 20-30 m above its previous position, and the three destroyed bridges were replaced with arc structures, both of which reduced the potential losses of infrastructure during future floods.
Because of the transboundary nature of rivers in the Hindu Kush Himalayan region, events of a large magnitude often impact more than one country. The Government of Nepal has informed the Government of India about the potential threat of flooding should there be a sudden outburst of the temporary lake formed behind the landslide dam. Recognizing the risk for communities downstream, all of the gates of the Sapta Koshi barrage, which is under the control of the Government of Bihar, were opened. The Bihar government has sounded a flood alert in eight districts and have begun the evacuation of nearly 200,000 people living along the Koshi embankment in India.
The governments of China and India have already offered technical support to the Government of Nepal in its response to the Sun Koshi event. China has good experience in managing mountain hazards, including a landslide following the 2008 Wenchuan earthquake that blocked a valley. Their technical expertise would prove invaluable in managing the massive landslide in the Sun Koshi valley. However, this cooperation should be extended beyond this particular event to long-term transboundary collaboration in managing risks, including in regular monitoring and assessment of potential risks and the implementation of early warning systems.
The scale of the Sun Koshi landslide is beyond the capacity of local communities to manage alone. However, national governments must promote the central role of communities in landslide risk management, including preparedness, adaptation, and mitigation. This is especially true in remote areas where limited access can delay the national disaster response efforts.
Consultation with local communities and the use of indigenous knowledge is crucial, particularly in the case of landslides. Use of indigenous knowledge in scientific and technical risk assessments can strengthen the resilience of communities, help communities take decisions informed by their own knowledge, and, when combined with scientific data, correct their own misperceptions about potential risks. This will help communities translate risk perceptions into enhanced preparedness for landslides and other hazards.