Uncertainty analysis in a water quality mathematical model applied to the Ribeirão do Ouro, Araraquara, SP, Brazil

Abstract

Mathematical models for surface water quality constitute an important tool to assist decision-making used to prevent and to design water resources degradation restoration measures. However, there are uncertainties in the determination of parameters and input variables of the models that can significantly alter the expected results. If these uncertainties are simply ignored, wrong decisions can be taken, based on results of a low confidence level model. In this context, the aim of this study was to analyze the uncertainty, based on the use of a stochastic method, of a deterministic model of surface water quality applied to Ribeirão do Ouro, located in the Araraquara municipality, Brazil. The method of Analysis of First Order Errors was used and the results were interpreted taking into consideration the water quality standards recommended by CONAMA 357/2005 resolution to generate surface water quality maps for the watershed. Significant influence of human activities on the quality of surface water was observed in more than 7 km along the river indicating environmental violations by the polluter. Considering the uncertainty components, the stochastic method has proved to be relevant to the management of water resources, allowing the presentation of results in terms of probability of occurrence.