Developing a SMART flood early warning system for a mountain watershed: experiences from the Lesser Himalayas

In the current context of climate change, urban areas in the Himalayas frequently experience flash floods. During high-intensity rainfall events in the catchments, due to hilly terrain and steep slopes, headwater streams cause flash floods and destroy life and property downstream. Increased encroachment along riverbanks and unplanned urban settlements expose financially distressed communities to the elevated risk of floods. This requires developing a reliable warning/alert system to ensure better preparedness for flood hazards and improve disaster resilience. Adequate hydrometeorological monitoring is a key element of such a system to generate knowledge on catchment/watershed characteristics as part of a broader disaster mitigation framework to reduce flood risk. 

The Bindal river in Dehradun (the capital city of Uttarakhand state in India) lies in the Doon valley on the foothills of the Himalayas, having a significant elevation difference of 450m with an area of 44.4 km2. The downstream settlements of the Bindal river experience flash floods during the monsoon season. Utilizing a SMART approach (developing shared understanding, monitoring, and awareness of the associated risks for preplanning response actions on time), this study aims to leverage and test a low-cost sensor network to provide information of hydrological variability and runoff response in the Bindal catchment. The SMART sensor network consists of 3 LiDAR river water level sensors and 4 tipping-bucket rain gauges at 15-minute intervals. The observed data showcases a substantial variability at both spatial and temporal scales within the small catchment of the Bindal river. The correlation coefficient (p value<0.05) between the rainfall observations at different stations varied from 0.82 to 0.20, with distance between their locations varying from 2.74 to 8.24km. The difference in total monthly rainfall recorded in two rain gauges 8.24 km apart in September is 187 mm. Additionally, the preliminary data suggests urban settlements in the downstream receive heavy rainfall within a short duration, while upper-catchment regions receive low-intensity rainfall for a longer duration. Future work will focus on developing a correlation between rainfall intensity and streamflow to define Intensity-Duration (ID) thresholds for early warning of flash floods. Similar systems in mountain landscapes with long-term rainfall and discharge data can contribute to establishing effective and low-cost flood warning systems for vulnerable riverine communities, particularly in developing countries.