Simulation of point and non-point source pollution in mahanadi river system lying in odisha, India

摘要: Assessment of point and non-point source pollution in a river system plays an important role for proper water resources management/utilization/protection, reducing environmental/health degradation, suitable waste load allocation decision-making monitoring networks. It has been observed that most the have utilized disposal municipal industrial wastes since early days addition to influx pollution. To address non-linearity, subjectivity, transfer transformation rule pollutants complexity cause-effect relationships between quality variables status, development use model is utmost importance. Since concentration constituents reliant on quantity flow, entry pollutants, reaction kinetics, etc., it essential supervise mathematical models predicting variables. Owing random discharge from various sources not only rendered such bodies eutrophic but also their advantageous uses as supply, irrigation, recharge ground water, recreation habitat flora fauna adversely affected. Oxygen-demanding substances are major contaminants domestic wastewater. The main indicators which deals with oxygen conditions Biochemical demand (BOD) dissolved (DO).To manage natural subjected pollutant inputs; one must be able predict degradation results inputs. another imperative variable responsible increasing stream/river. Recognizing magnitude assessing system, copious studies intended at understanding processes controlling nutrient concentration, fluxes systems quantification loads rivers proficient past. In present study, attempts made different Mahanadi lying Odisha, establish parameters values test the iv applicability Spatio-temporal conditions. Different data namely, discharge, BOD, DO, temperature, pH, turbidity, electrical conductivity, nitrate a Most commonly used BOD-DO simulate reaches Odisha deoxygenation coefficient (k1) reaeration rate (k2) established. Various empirical equations estimating were modified equation derived. Further, Multi-layer Perceptron (MLP) neural network techniques was estimate analysis terms BOD DO developed using collected upstream downstream stations Odisha. accuracy performance training, validation prediction seasonal tested. Another For recognizing importance nutrients (nitrate ortho-phosphate) non simulation, analytical entering estimated. To validity generalized ANN model, statistical errors, root mean square error (RMSE), multiplicative errors (MME), correlation (R) used.