Services on Demand
Journal
Article
Indicators
Cited by SciELO
Access statistics
Related links
Similars in SciELO
Share
Tecnología y ciencias del agua
On-line version ISSN 2007-2422
Abstract
ALVAREZ-OLGUIN, Gabriela; MARTINEZ-RAMIREZ, Saúl and LICONA-MORAN, Brenda I. G.. Predicción de lluvias máximas para la república mexicana mediante modelos probabilísticos no estacionarios. Tecnol. cienc. agua [online]. 2020, vol.11, n.4, pp.179-214. Epub June 10, 2024. ISSN 2007-2422. https://doi.org/10.24850/j-tyca-2020-04-06.
The prediction of maximum rainfall is the basis of the design of hydraulic structures for flood mitigation. This prediction is traditionally made from frequency analysis methods that consist of studying past events in order to define the probabilities of future occurrences. However, because climatic variability causes changes in the mean and variance of rainfall time series, design events are unreliable if they are estimated from techniques valid for stationary conditions. For Mexico, there is evidence that rainfall patterns are changing, consequently for predictions to be reliable, it is necessary to apply methods that contemplate changes in the statistical characteristics of the data over time. Therefore, the aim of this study was to estimate annual maximum 24-hour rainfall events associated with different return periods, through non-stationary probabilistic models. We analyzed 769 series to which were applied the Pettitt, Mann-Kendall and Ensemble Empirical Mode Decomposition tests to verify the seasonality. Different probabilistic models were proposed, in which parameters of the Lognormal, Gamma, Gumbel, Weibull and Generalized Extreme Value distribution have as covariates to time and Pacific Decadal Oscillation index. The results indicated that for the non-stationary series the proposed models represent the variability of the data better than conventional stationary models.
Keywords : Frequency analysis; extreme hydro-meteorological events; Pacific Decadal Oscillation; change point and trend and climatic variability.