Distribution probability functions for models applied to hydraulic works in urban areas: LOG-GUMBEL vs LOG-PEARSON IIIRaúl

Key Points

Log-Gumbel provides safer estimates for long return periods, favoring safety in urban drainage systems.
Log-Pearson III consistently yields stable results despite data variability and is preferred for limited data designs.
The study utilizes data from the Almodóvar reservoir in Cádiz to illustrate the models' applications.
Choosing between Log-Gumbel and Log-Pearson III impacts hydraulic design outcomes, emphasizing safety versus stability concerns.

Abstract

Hydrometeorological models for hydraulic design rely on probability functions to model extreme rainfall. This study compares Log-Gumbel and Log-Pearson III functions using data from Almodóvar reservoir (Cádiz). Log-Gumbel, a variation of the Gumbel distribution, yields higher, safer estimates for long return periods (T), but is highly sensitive to data changes. Log-Pearson III, widely used and more stable, shows consistent results despite data variability. For designs with limited or variable data—common in hydraulic works—Log-Pearson III is preferred. However, for urban drainage where safety is prioritized over stability, Log-Gumbel offers a conservative alternative for high T values.

Read Full Paperexternally

Mark Helpful

Bookmark

Relay

View Full Paper