Paper/Poster/Presentation Title

Censored continuous simulation for flood estimation

Keywords

Flood modelling, flood probability, continuous simulation, censoring, censored continuous simulation

Start Date

27-6-2018 9:00 AM

End Date

27-6-2018 10:20 AM

Abstract

The understanding of flood risk relies on accurate estimation of flood exceedance probability. This is often a challenging task in estuarine regions, as estuarine floods are often caused by multiple processes, such as storm surge and riverine floods. Current available methods for estuarine flood estimation are either timing consuming (e.g. continuous simulation) or has to ignore the dynamic interaction between different flood drivers as a compromise to improve efficiency (e.g. the design variable method (Zheng et al. 2015)). This study proposes a censored continuous simulation approach, which draws advantages from both existing methods. The censored continuous simulation approach is based on the fact that floods are relatively rare events and therefore, flood data from most of the time series record are often not used for flood probability estimation. Therefore, these “no-flood” periods can be “censored out” in the simulation process. This censored continuous simulation approach has the advantages of continuous simulation, which considers the dynamic interaction between flood drivers. It also reduces the simulation time significantly compared to full continuous simulation by simulating mainly the “flood events” of interest. The application of this approach is demonstrated through a case study in Western Australia.

Stream and Session

C11: Integrated Methods and Tools for Flood Risk and Water Supply Management

COinS
 
Jun 27th, 9:00 AM Jun 27th, 10:20 AM

Censored continuous simulation for flood estimation

The understanding of flood risk relies on accurate estimation of flood exceedance probability. This is often a challenging task in estuarine regions, as estuarine floods are often caused by multiple processes, such as storm surge and riverine floods. Current available methods for estuarine flood estimation are either timing consuming (e.g. continuous simulation) or has to ignore the dynamic interaction between different flood drivers as a compromise to improve efficiency (e.g. the design variable method (Zheng et al. 2015)). This study proposes a censored continuous simulation approach, which draws advantages from both existing methods. The censored continuous simulation approach is based on the fact that floods are relatively rare events and therefore, flood data from most of the time series record are often not used for flood probability estimation. Therefore, these “no-flood” periods can be “censored out” in the simulation process. This censored continuous simulation approach has the advantages of continuous simulation, which considers the dynamic interaction between flood drivers. It also reduces the simulation time significantly compared to full continuous simulation by simulating mainly the “flood events” of interest. The application of this approach is demonstrated through a case study in Western Australia.