Dixit et al. (2026) Integrating SMART principles in flood early warning system design in the Himalayas

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2026
• Date: 2026-03-10
• Authors: Sudhanshu Dixit, Sumit Sen, Tahmina Yasmin, Kieran Khamis, Debashish Sen, Wouter Buytaert, David M. Hannah
• DOI: 10.5194/nhess-26-1251-2026

Research Groups

• Centre of Excellence in Disaster Mitigation and Management, Indian Institute of Technology Roorkee, Roorkee, Uttarakhand, India
• School of Geography, Earth & Environmental Sciences, University of Birmingham, Birmingham, UK
• People’s Science Institute, Dehradun, India
• Department of Civil and Environmental Engineering, Imperial College London, London, UK

Short Summary

This study integrates SMART principles with low-cost, real-time hydrometeorological monitoring to design an urban flood early warning system (EWS) in the data-scarce Lesser Himalayas. It demonstrates how community-engaged monitoring captures crucial spatiotemporal rainfall variability and watershed dynamics, which are poorly represented by secondary datasets, providing foundational insights for effective, community-centered EWS implementation.

Objective

• To demonstrate the benefits of low-cost, real-time monitoring systems and the importance of fostering community engagement in flood risk management.
• To integrate the spatiotemporal variability and dynamic nature of watershed characteristics into flood EWS.

Study Configuration

• Spatial Scale: Bindal River watershed, Lesser Himalayas, Uttarakhand, India, covering 44.4 km² with elevations ranging from 539 m to 997 m.
• Temporal Scale: Data collection commenced in September 2022. Analysis periods included late monsoon 2022, non-monsoon 2023, and monsoon 2023, with an extended dataset (August 2024–October 2025) for variability persistence.

Methodology and Data

• Models used: Maximum Likelihood (ML) supervised classification tool for Land Use/Land Cover (LULC) mapping; threshold-based approach for flood alerts.
• Data sources:
  • In-situ: 3 LiDAR-based water-level sensors (data every 5 min), 4 tipping bucket rain gauges (data every 15 min).
  • Satellite: Sentinel-2 images (for LULC), GPM (IMERG) Version 06.
  • Reanalysis: ERA5 global atmospheric reanalysis data.
  • Community interaction: Participatory Rural Appraisal (PRA) exercises and Focus Group Discussions (FGDs).

Main Results

• Community engagement revealed a decrease in river baseflow, an increase in flood frequency and intensity over the last two decades, and identified riverbed encroachment as a major factor in heightened flood intensity.
• Significant spatial variability in rainfall was observed, with a 187 mm difference in cumulative monthly rainfall between rain gauges 8.24 km apart in September 2022, and higher elevations consistently receiving more rainfall.
• Cross-correlation between rain gauges declined rapidly with distance (e.g., from 0.83 to 0.24 over 2.7 to 8.24 km in non-monsoon), indicating high spatial variability, particularly during monsoon seasons.
• Secondary datasets (GPM-IMERG and ERA5) inadequately captured rainfall heterogeneity; ERA5 underestimated rainfall (e.g., >40% discrepancy in monsoon), while GPM overestimated it, both showing significant biases (PBIAS: GPM -43.57, ERA5 100.16).
• Rainfall storm movement was consistently southwestward, with an approximate 15 min lag observed between the highest elevation rain gauge (RG1) and downstream urban areas (RG3).
• Watershed response times to rainfall events ranged from 15 minutes to 2 hours 30 minutes, narrowing to 15-45 minutes during heavy and very heavy rainfall.
• Flood alerts varied from watch to warning levels even during heavy rainfall events, highlighting the complex interplay of rainfall patterns and watershed dynamics in determining flood severity.

Contributions

• Demonstrates the critical role of integrating low-cost, real-time hydrometeorological monitoring with community-centric SMART principles for effective flood early warning system (EWS) design in data-scarce, complex urban mountainous regions.
• Provides practical evidence and foundational hydrometeorological insights for the implementation of community-centered urban flood EWS in the Himalayan region.
• Highlights the limitations of widely used secondary rainfall datasets (GPM-IMERG and ERA5) in accurately capturing the high spatiotemporal rainfall variability essential for flash flood prediction in complex terrains.
• Emphasizes the necessity of dynamic, locally-determined flood thresholds, informed by community knowledge, to account for anthropogenic changes in river channel morphology.
• Characterizes specific watershed dynamics, including rainfall storm movement direction and lag times, and the rapid hydrological response of small mountainous watersheds.

Funding

• Natural Environment Research Council (NERC COP26 A&R, Project Scoping Call-2021COPA&R31Hannah)
• Prime Minister’s Research Fellowship (PMRF, grant no. 2802448)

Citation

@article{Dixit2026Integrating,
  author = {Dixit, Sudhanshu and Sen, Sumit and Yasmin, Tahmina and Khamis, Kieran and Sen, Debashish and Buytaert, Wouter and Hannah, David M.},
  title = {Integrating SMART principles in flood early warning system design in the Himalayas},
  journal = {Natural hazards and earth system sciences},
  year = {2026},
  doi = {10.5194/nhess-26-1251-2026},
  url = {https://doi.org/10.5194/nhess-26-1251-2026}
}