Distribuciónes Con Límite Superior e Información no Sistemática Para El Análisis de Frecuencia de Crecidas Extremas. Aplicación al Río Turia, España

This paper proposes a methodology to the flood frequency analysis, which combines upper bounded distribution functions with available Non-Systematic information at the place of interest. Proposed distribution functions are the TDF, LN4 and EV4, which were used first in 90s for the analysis of hydrologic extremes. Principal objectives of the proposed methodology are to improve high return period quantile estimation and to propose the upper bound of the functions as the statistical estimator of PMF (Probable Maximum Flood). In order to reduce uncertainty of the estimators Non-Systematic information has been used incrementing in a high quantity the number of years with available information. Parameter estimation method used was the Maximum Likelihood which allow add this type of information in an easy way. The methodology was applied at the Turia river in Spain. This river has a long period with Non-Systematic information. Application results show that the EV4 distribution is a recommendable function for rivers with Mediterranean regime and that it is possible to obtain a statistical estimator of the PMF with an associated uncertainty.