Estimation of the transverse dispersion coefficient for two-dimensional models of mixing in natural streams

Abstract

Existing equations to predict the transverse dispersion coefficient for the two-dimensional model of mixing and transport in natural streams use the channel aspect ratio and roughness factor as well as the sinuosity of the channel as the controlling parameters, and the performance of the equation varies largely according to the range of each parameter that was studied. To this end, this study suggests the criteria to select the proper method to calculate the transverse dispersion coefficient according to the availability and range of each parameter. The strengths and limitations of existing methodologies were compared against observed data acquired from tracer experiments conducted in natural streams. Further, a flow chart was proposed with the criteria to select a suitable method under specific hydraulic and geometric conditions of the stream.The results of the classification flow chart showed that, at the first step, in the cases where secondary currents data were available for natural streams, the theoretical equation by Baek and Seo (2011) could be used to estimate the transverse dispersion coefficient. At the second step, in the case of large value of P, i.e., P>0.04, the equation by Baek and Seo (2013) was suitable to estimate the transverse dispersion coefficient, while in the case of a small value of P, equations by Yotukura and Sayer (1976) and Baek and Seo (2013), could be used with little differences. At the third step, for the narrow streams with W/h<50, those proposed by Bansal (1971) and Deng et al. (2001) were preferable to estimate the transverse dispersion coefficient for two-dimensional mixing models. In wide streams with W/h>50, the results of Jeon et al. (2007) showed much better agreement with the observed values than the others.