Advanced Early Warning Systems Protect Lives and Livelihoods in Nepal

October 12th, 2017

An innovative flood early warning system grants life-saving hours of preparation to some of the world’s most vulnerable people. In July, 2016 this system alerted 13,000 people to impending floods 3 hours before the traditional warning system was triggered.

Nepal is one of the one most flood-exposed countries in the world, and the people who live in the Karnali River basin are also among the most vulnerable. More of the world’s poor live in this area than in any other regional river basin. Poverty plays a significant role in vulnerability and resilience to natural hazards, as those with fewer resources are less able to protect themselves from flooding, and are also impacted to a greater degree by the loss and damage caused by floods.

With so much at stake, time is crucial for families to protect themselves, their homes and their livelihoods. The current early warning system in place measures water levels upstream of the lower part of the river basin, using real-time measurements to trigger a warning, sent via SMS messages, for communities downstream when a pre-determined threshold is crossed.

This approach is simple, sustainable and saves lives, but its effectiveness is constrained by the limited lead time it provides. Typically, communities have only two to three hours, making it difficult to save assets, livestock and tools, and presenting challenges for more vulnerable groups such as people with disabilities, pregnant women, the elderly and children.

With little time to prepare, families are placed at serious risk when flooding occurs


Practical Action’s solution

With the support of the Zurich Flood Alliance Programme, we worked with Dr Paul Smith of waternumbers[1] and Nepal’s Department of Hydrology and Meteorology (DHM) to pilot a new approach which would increase the lead time and the detail of information provided by the early warning system. This approach is based on probabilistic forecasting, presenting the likelihood of water levels crossing danger and warning thresholds, rather than simply reporting what is happening as it is happening. This information is communicated visually to key decision-making personnel in the DHM. Initial results indicate that this approach could provide up to five additional hours of early warning, and the DHM has now adopted this model and is rolling it out to all major river basins across Nepal.

Proven success

During the 2016 monsoon in western Nepal, we were able to see the impact of this new system. When rainfall stations recorded heavy rainfall on 26 July, the forecast generated by the probabilistic system at 8:00 am indicated that the water level would rise above the danger threshold within the next three hours. Based on this, the DHM issued a flood advisory, and over 13,000 at-risk people received SMS messages alerting them to the risk. The danger threshold was crossed at 11:00 am, with water levels reaching 5.4 metres, and peaking at 8.15 metres at 10:00 pm.

The longer warning lead time from probabilistic forecasts was significant in minimizing the risk to lives and livelihoods as communities gained extra time to prepare, evacuate, and respond.

Find out more

For more detailed information, please refer to the full article by Paul J. Smith, Sarah Brown and Sumit Dugar, ‘Community-based early warning systems for flood risk mitigation in Nepal’ published in the Journal of Natural Hazards and Earth System Sciences.

[1] A consultancy committed to assisting the management of risks from natural hazards using sustainable, innovative and tailored modelling and analytical solutions

Leave a reply