Assessment and Prediction of Water Quality in Shrimp Culture

Abstract

Diagnosis of water quality can be assessed through toxicity tests and by analyzing the negative effects of water quality parameters. Such diagnosis provides an indicator of the condition of the shrimp ecosystem. In this research, three models for water quality diagnosis were proposed: first, a model for immediate assessment, second, a model for historical evaluation and third, a model for prediction of water quality. We evaluated water quality parameters by their negative ecological impacts using a new index (gamma). A fuzzy inference system assessed each harmful situation presented in the pond, providing a final index of water quality: IWQI for immediate measurements and HWQI for a set of measurements. An autoregressive model (AR) was used for predicting values of each water quality parameter; these values were assessed by the proposed fuzzy inference system, establishing a predictive water quality index (PWQI). Comparisons of our models against classical models in the literature show good performance of the algorithms proposed in this work.