Evaluation of re-aeration equations for river Ghataprabha, Karnataka, India and development of refined equation.

Stream re-aeration is an extremely important component to enhance the self-purification capacity of streams. To estimate the dissolved oxygen (DO) present in the river, estimation of re-aeration coefficient is mandatory. Normally, the re-aeration coefficient is expressed as a function of several stream variables, such as mean stream velocity, shear stress velocity, bed slope, flow depth and Froude number. Many empirical equations have been developed in the last years. In this work, 13 most popular empirical re-aeration equations, used for re-aeration prediction, have been tested for their applicability in Ghataprabha River system, Karnataka, India, at various locations. Extensive field data were collected during the period March 2008 to February 2009 from seven different sites located in the river to observe re-aeration coefficient using mass balance approach. The performance of re-aeration equations have been evaluated using various error estimations, namely, the standard error (SE), mean multiplicative error (MME), normalized mean error (NME) and correlation statistics. The results show that the predictive equation developed by Jha et al. (Refinement of predictive re-aeration equations for a typical Indian river. Hydrological Process. 2001;15(6):1047-1060), for a typical Indian river, yielded the best agreement with the values of SE, MME, NME and correlation coefficient r. Furthermore, a refined predictive equation has been developed for river Ghataprabha using least-squares algorithm that minimizes the error estimates.