Science et changements planétaires / Sécheresse
MENUDaily rainfall simulation in semiarid areas using the Principal Component Analysis approach Volume 18, issue 2, Avril-Mai-Juin 2007
Figures
- Key words: modelling, principal component analysis (PCA), rainfall, semiarid zone, simulation
- DOI : 10.1684/sec.2007.0077
- Page(s) : 97-105
- Published in: 2007
Generating synthetic series is particularly appropriate for the development of a simulation model for optimising the management of reservoir dams. Usually, simulation is used in desing and planning activities, in operational studies for water resources management schemes as well as in climate change studies. This article presents a method for generating long synthetic sequences of daily rainfall preserving the main statistical features of historical series. The simulation model developed is based on the principal components analysis method. It offers great advantages over those commonly used as it delivers an important variety of combinations of generation methods and allows us to get rid of the constraints imposed by other models. Indeed, the different tests performed on random variables (rainfall) simulated on a daily step basis, have yielded excellent results. The stationarity, the type of phenomenon, the truncation of observations, the length of the historical series, or the climatic regime, are in no way an obstacle to the application of the elaborated model.