Urban flood prediction and prevention: a checklist for cities

Flood prediction starts with accurate field observation; flood prevention starts with a system that turns predictions into action. Seven things for cities to check.

Concept illustration of AI flood risk analysis: a color heatmap of flood risk over a city district, with low roads and an underpass shown in red

Flood prediction · check first

  • A map of past flood locations: underpasses, semi-basements, low-lying roads
  • Field rainfall observation near risk points (every minute)
  • Zone-level flood risk analysis combined with terrain and drainage data
  • A monitoring screen where staff see predictions immediately

Flood prevention · turn it into action

  • Staged alerts matched to institutional thresholds (advisory, warning)
  • Vulnerable residents matched with evacuation support staff in advance
  • Agreed criteria for underpass closures and facility checks

Prediction vs prevention

Flood predictionFlood prevention
GoalKnow when and where risk will riseAct before damage happens
Key dataField rainfall, terrain and drainageResidents, staff and facilities
OutputRisk level by zoneAlerts, dispatch, closures, evacuation complete

Often missed

  • City-average rainfall misses the hot spots of local downpours
  • Alerts without clear roles—who does what—lead to late action
  • Platform alerts do not replace official warnings

Sources

  1. Early Warnings for All — United Nations, 2022
  2. Sendai Framework for Disaster Risk Reduction 2015–2030 — UNDRR, 2015
  3. Climate Change 2021: The Physical Science Basis — IPCC Working Group I (AR6), 2021