Evaluación de los resultados de diferentes funciones Likelihoods de la metodología GLUE aplicadas en la obtención de coeficientes de rugosidad para ríos de montaña

Abstract

This thesis work compares the behavior of different likelihood functions of the GLUE methodology in the classification of roughness coefficients of mountain rivers (three river morphologies: Cascade, Plane Bed and Step-pool) obtained from models. Bibliographic research on likelihood functions used to obtain roughness coefficients on rivers was carried out, and then these functions were used to classify a set of modeled data using as comparison parameters values measured in the field. The results showed an almost similar behavior between likelihood functions with the exception of some cases that tend to validate roughness coefficients that differ considerably with the values measured in the field. Finally, it can be seen that a likelihood function, depending on the parameters that are incorporated, can be as simple as the researcher wishes, making the complex function excessively particular to the river being analyzed, limiting its applicability to other rivers with similar characteristics.