Roberto Villalobos Herrera
Events-based approach for representation of rainfall events.
Email: r.villalobos-herrera2@ncl.ac.u
Project title
Generation of stochastic design rainfall ensembles using an events-based analysis of sub-hourly rainfall
Supervisors
Project description
Rainfall data is a key input to drainage infrastructure design and evaluation methods. In many places, outdated methods and a limited representation of rainfall events limit this rainfall input.
Design rainfall options available for engineers are limited to simple design storms or to continuous simulation methods of long, stochastically-generated rainfall time series. These approaches have important limitations.
Design storm methods are usually limited to few hyetograph shapes. They also lack the variability observed in rainfall events.
Continuous simulation methods improve the representation of variability. This is at increased computational costs in the impact simulation stage. Analysis is usually constrained to hydrological models.
This project seeks to bridge the gap between these two approaches.
Roberto will use an events-based approach to analyse observed variability in rainfall event magnitude, duration and hyetograph shape. This is different from conventional rainfall frequency analysis. Event duration isn’t prescribed to fixed intervals.
Roberto will use observed event characteristics to develop a stochastic rainfall event ensemble generator. This will preserve the observed variability in extreme rainfall events.
He will also develop and test options to perturb the generator to account for plausible impacts of climate change.
The project will provide users with a design rainfall alternative with improved representation of rainfall event variability. This can be used as input for advanced urban drainage simulation software.