Manning roughness coefficient for Paracatu river, Brazil

Revista Brasileira de Engenharia Agrícola e Ambiental

Lyra, Gustavo B.³  Cecílio, Roberto A.²  Zanetti, Sidney S.⁴  Lyra, Guilherme B.¹ 

[1] UFAL, Maceió;UFES, Alegre;UFRRJ, Seropédica;UFES, São Mateus

关键词: Water flow;    open channel;    artificial neural networks    INTRODUÇÃO Para atender às necessidades crescentes de uso da água;    torna-se necessária a caracterização hidrológica dos cursos d'água;    estudo este importante para a estimativa de sua disponibilidade hídrica;    do seu potencial energético;    do controle de inundações e do dimensionamento de obras hidráulicas (Tucci;    2001);    haja vista que só assim é possível implementar um sistema de gestão dos recursos hídricos em bases técnico-científicas;    desta forma;    é oportuno um acompanhamento contínuo das variáveis mais importantes que descrevem o comportamento hidrológico dos rios (Martoni & Lessa;    1999b);   

DOI  :  10.1590/S1415-43662010000400001

学科分类：农业科学（综合）

来源: Universidade Federal de Campina Grande * Centro de Ciencias e Tecnologia

【 摘 要 】

According to open channel draining theory, the Manning's roughness coefficient is one of most important parameters to describe surface flow. One of the difficulties in applications of Manning's equation is the definition of its roughness coefficient in rivers and channels. Thus, the aims of this paper are to estimate the Manning's roughness coefficient for the periods of minimum and maximum mean monthly flow in some parts of Paracatu river and to propose a model based on artificial neural networks to estimate the roughness coefficient. The coefficient was determined in function of channel geometrical characteristics (wetted area, hydraulic radius and slope of the channel) and of the outflow series of six fluviometric stations of Paracatu river. Long-term series of 21 years (1976 - 1996) of outflow were used. The roughness coefficient does not show any tendency (up or down)as a function of the dry or overflow periods only. The characteristics of margin and watercourse of river influenced directly the roughness coefficient data. The model based on neural network showed satisfactory performance, which allow us to estimate the roughness coefficient as a function of the quota, slope and hydraulic radius of the river.

【 授权许可】

CC BY-NC