The development of predictive tools for recreational water quality based on rainfall levels: case study of Cachoeira das Emas beach (SP)

ABSTRACT

This article presents the results of a case study conducted on Cachoeira das Emas beach, at Mogi-Guaçu River, in Pirassununga (SP). Using precipitation levels as explanatory variables, simple statistical models based on linear regression were developed. Positive correlation between the concentration of faecal indicators in water and accumulated rainfall in 24, 48 and 72 hours, prior to sampling on the river, were observed. Data sets used for calibration (2009-2011 period) and validation (period of 2012) of the predictive tools were originally acquired from distinct monitoring networks: rainfall levels data by the Departamento de Águas e Energia Elétrica (DAEE) and recreational water quality data by the Companhia Ambiental do Estado de São Paulo (CETESB). Due to different goals of these monitoring networks, not every rainfall event during the study period was related to a concentration value of fecal indicator in the water, and vice versa, so the regressions obtained were based on a low quantity of observations. The results indicate that there is a potential for applying the predictive tool, but it requires monitoring the model parameters in conjunct, in order to obtain the amount of data necessary to calibration and validation. Ideally, the use of this recreational water management tool would be during a certain period of the year, when there are more people at the beach site, on locations where the main source of pollution is diffuse, resulting from runoff drainage. It would also be possible to add other environmental variables to the models, such as flow and turbidity, provided that they are monitored with water quality conditions.

Keywords:
management; water quality; recreation; modeling