Predicting Roughness Coefficient for Natural Mountain River Channel at High Stage Based on Field Data

A large number of data based on field measurements from 91 hydrology stations in Sichuan Province, China were collected for the purpose of evaluating four factors affecting Manning's roughness coefficient (n). These data show that water depth, surface slope, bed material size and cross-section shape of channel all affect the roughness coefficient. Although the relationship between roughness and depth is highly variable at low stage, it tends to be stable at high stage. Equations for n as a function of surface slope J are developed individually for different type of natural mountain rivers in Sichuan Province by using multiple regression analysis based on field measurements. A value of 65 of bed roughness coefficient A is suggested in this paper, which greatly exceeds the range given by previous studies and a new tabular range of n and surface slope is given in this paper based on field measurements. Given information of surface slope at high stage, one can apply the presented equations or refer to the table to predict the roughness coefficient for different characteristics of channels. This study can provide practical guidelines in properly determining the roughness coefficients for similar natural mountain river channels.