Anil KumarMalik, Anurag2018-10-082018-10-082014-06http://krishikosh.egranth.ac.in/handle/1/5810075002Correct estimation of suspended sediment load carried by a river is very important for water resources management, channel navigability, reservoir filling, fish habitat, river aesthetics and scientific interests. In this study, the Co-active Neuro-Fuzzy Inference System (CANFIS), Multiple Linear Regression (MLR), and Sediment Rating Curve (SRC) techniques were applied for estimating the daily suspended sediment concentration at Tekra site on Pranhita River, which is a major tributary of Godavari River basin in Andhra Pradesh, India. The daily streamflow and suspended sediment concentration data of monsoon season from June 23, 2000 to November 11, 2002 were used for training/calibration, and from June 23, 2003 to November 11, 2003 were used for testing/validation of models. The Neurosolutions for Excel Release 5.0 software was used for designing the CANFIS networks based on Gaussian membership function, Takagi-Sugeno-Kang (TSK) fuzzy model, hyperbolic tangent activation function, and Delta-Bar-Delta learning algorithm for training and testing of the models. The performance of CANFIS was compared with MLR and SRC techniques using statistical functions such as root mean square error (RMSE), coefficient of efficiency (CE) and correlation coefficient (r). The results indicate that the CANFIS performs superior to the MLR and SRC models, while MLR performs better than SRC model in estimating daily suspended sediment load. The best CANFIS and MLR models reveal that the suspended sediment of the present day depends on current and previous one day’s streamflow as well as sediment concentration of previous two days at the Tekra station on Pranhita River of Godavari basin in India.ennullThesis