An advanced method to apply multiple rainfall thresholds for urban flood warnings

Research output: Contribution to journalArticlepeer-review

21 Citations (Scopus)


Issuing warning information to the public when rainfall exceeds given thresholds is a simple and widely-used method to minimize flood risk; however, this method lacks sophistication when compared with hydrodynamic simulation. In this study, an advanced methodology is proposed to improve the warning effectiveness of the rainfall threshold method for urban areas through deterministic-stochastic modeling, without sacrificing simplicity and efficiency. With regards to flooding mechanisms, rainfall thresholds of different durations are divided into two groups accounting for flooding caused by drainage overload and disastrous runoff, which help in grading the warning level in terms of emergency and severity when the two are observed together. A flood warning is then classified into four levels distinguished by green, yellow, orange, and red lights in ascending order of priority that indicate the required measures, from standby, flood defense, evacuation to rescue, respectively. The proposed methodology is tested according to 22 historical events in the last 10 years for 252 urbanized townships in Taiwan. The results show satisfactory accuracy in predicting the occurrence and timing of flooding, with a logical warning time series for taking progressive measures. For systems with multiple rainfall thresholds already in place, the methodology can be used to ensure better application of rainfall thresholds in urban flood warnings.

Original languageEnglish
Pages (from-to)6056-6078
Number of pages23
JournalWater (Switzerland)
Issue number11
Publication statusPublished - 2015

All Science Journal Classification (ASJC) codes

  • Biochemistry
  • Geography, Planning and Development
  • Aquatic Science
  • Water Science and Technology

Fingerprint Dive into the research topics of 'An advanced method to apply multiple rainfall thresholds for urban flood warnings'. Together they form a unique fingerprint.

Cite this