Evaluación de diferentes metodologías de predicción del coeficiente de resistencia al flujo de ríos de montaña considerando su incertidumbre

Abstract

The resistance coefficient is a very important parameter in the one-dimensional simulation of water flow. This parameter can be estimated using various methodologies, which can be equations, data tables and photographs. Like any other estimate, the prediction of the resistance coefficient is subject to an error or uncertainty due to various factors. In the present work, the performance of the different estimation methodologies of the Manning resistance coefficient was evaluated when considering a percentage variation of the coefficient. The data used for this purpose pertain to 3 morphologies of the Quinuas mountain river (cascade, step-pool y plane-bed). The performance was evaluated by comparing the water depth measured in various sections with the water depths obtained in the different one-dimensional models of the HEC-RAS software. The results indicate that the nondimensional geometry equations better estimate the resistance coefficient. In turn, in the cascade and step-pool morphologies, the nondimensional geometry equations have reached a maximum possible efficiency, therefore, in these cases, varying the estimated coefficient of resistance only worsens the results or a minor improvement is obtained. The remaining methodologies show in most cases that they underestimate the flow resistance, especially in flows of low magnitude. However, some methodologies under certain conditions show similar results to the nondimensional geometry equations. Finally, it was found that, in the cascade morphology, the performance of the methodologies is less sensitive to the variation of the resistance coefficient compared to the performance in the plane-bed morphology