Water quality simulation of Pardo river with incorporation of uncertainty analysis

Abstract

The main perspective of the present work is the mathematical simulation of dissolved oxygen and biochemical oxygen demand transformations in the Pardo River (Espírito Santo, Brazil), considering the incorporation of uncertainty analysis by the Monte Carlo method. For the Monte Carlo method application, hydrodynamic variables, Pardo River initial quality conditions, and effluents and water quality model kinetic constants were randomly generated with the aid of Normal and Uniform probability distributions. The use of each probability distribution involved performing 1,000 (one thousand) simulations of the dissolved oxygen and biochemical oxygen demand profiles. The simulation results indicated that, for the studied watercourse, the minimum dissolved oxygen concentrations varied predominantly between 6.0 and 6.5 mg.L^-1 and, in none of the simulations, did they present lower values to the environmental quality standard. biochemical oxygen demand concentrations, in turn, remained above the maximum environmental quality standard in the Pardo River largest intermediate portion, with maximum values between 16.0 and 17.7 mg.L^-1. The use of different probability distributions for the random generation of hydrodynamic variables, kinetic constants, and water quality concentrations parameters did not produce variations considered relevant when analyzing the frequency of critical concentrations of simulated water quality parameters.

Keywords:
mathematical modeling; Monte Carlo method; dissolved oxygen; biochemical oxygen demand; Pardo river