Correlation of turbidity and suspended sediment concentration in natural water flow using alternative data of water plant treatment, case of study in the upper Jundiaí river, Brazil.

Abstract

Measurements of suspended sediment discharge are greatly significant and represent a basic need for the initial assessment of the effects caused by this process and for the subsequent formulation of management actions. Also, considering that obtaining direct hydrosedimentological data is not a trivial and low-cost task due mainly to the special and temporal variability of the phenomenon, substitute technologies are well employed for monitoring watersheds. Among these are turbidity measurements, described in many papers. An adaptation of this technique was designed and performed in this study for the verification and secondary use of turbidimetric data carried out by a drinking water treatment company at the closing point of the upper part of the Jundiaí river watershed, in the city of Campo Limpo Paulista, São Paulo state. Direct measurements with traditional samplers and valuations of particle size distribution throughout the hydrological year were used to establish empirical correlations between the suspended sediments concentration and turbidity. The definition of the most appropriate adjustment was carried out based on the statistical analysis of the linear regression parameters (R2, Value-p, MSPE, and PRESS) and residue statistics, thus adopting the model with piecewise regression by turbidimetric ranges (250 NTU), which in turn, presented low overestimation of the SSC for high values of turbidity. The definition of the historical series of suspended sediment discharge (SSD)with hourly temporal resolution for the study site validated the main objective of reusing secondary data with hydro-sedimentological purposes, highlighting the plausible applications for different places where there are insufficient resources for the installation and use of in situ turbidity equipment.