On the applicability of Bartlett-Lewis rectangular pulses models in the modeling of design storms at a point

Niko Verhoest, Peter A. Troch, François P. De Troch

Research output: Contribution to journalArticlepeer-review

87 Scopus citations


This paper discusses the applicability of three versions of the Bartlett-Lewis rectangular pulses model of temporal rainfall for the calculation of design storms. We consider the original model, the modified model (developed in order to improve the zero depth probability) and the modified gamma model. This last model was recently suggested by Onof and Wheater [Onof, C. and Wheater, H.S., J. Hydrol., 157 (1994) 177-195] to improve the representation of extreme rainfall events. The data set used here is 27 years of 10 min rainfall records observed at Uccle (Belgium) between 1967 and 1993. First, the performance of the three models to preserve first- and second-order moments is evaluated. Overall, the modified Bartlett-Lewis model fits best the observed time series. Then, the models are applied to extrapolation of data sets for extreme value frequency analysis of rainfall in order to improve estimates of design intensities for return periods up to 100 years. For the data used here the modified Bartlett-Lewis gamma model does not produce the expected result: an improved representation of extreme rainfall events is not observed. Finally, interstorm variability, important in design storm selection, is analyzed based on time distribution behavior during heavy storms. It is found that all three models represent reasonably well the internal storm structure, but that the average storm duration is underestimated.

Original languageEnglish (US)
Pages (from-to)108-120
Number of pages13
JournalJournal of Hydrology
Issue number1-4
StatePublished - Dec 1997


  • Bartlett-Lewis model
  • Design storms
  • Temporal rainfall

ASJC Scopus subject areas

  • Water Science and Technology


Dive into the research topics of 'On the applicability of Bartlett-Lewis rectangular pulses models in the modeling of design storms at a point'. Together they form a unique fingerprint.

Cite this