Dıscrete-tıme dynamıc water qualıty ındex model ın coastal water

BACKGROUND AND OBJECTIVES: It is important to develop dynamic water quality index software that reflected accurately the state of enclosed coastal water quality. This study explored water quality index model software including the third-order and daily based discrete-time transfer function in Simulink-MATLAB environment to predict the past and future water quality index changes versus discrete-time by using the data measured approximately once a month. METHODS: A modelling software for daily based discrete-time water quality index was developed to evaluate the pollution level in enclosed coastal water bodies affected by marinas. Measurements were done at three different stations near marina entrances in Bucak, Kaş, and Fethiye Bays located at the south western Mediterranean coast of Turkey. The computed water quality index values and the sampled indicators data defined in terms of the deviation variables were used to identify the proposed third-order transfer function parameters. The proposed software is applicable for past and future estimates, where inputs may include some missing measurements. The input data are interpolated to estimate daily based inputs by using the developed model in the Simulink-MATLAB environment. For model verifications, monthly measured water quality parameters are used. FINDINGS: The software including the daily based discrete-time transfer function and the input sources was successfully applied to predict past and future water quality index changes with 4.2 percent, 4.3 percent, and 7.1 percent of the absolute maximum errors respectively in Fethiye, Kaş, and Bucak stations. In three stations studied, seasonal comparison of the enclosed coastal water quality showed that the quality in winter (72±2) is lower than the one (82±8) in other seasons. The past and future daily predictions of water quality index changes versus discrete-time were realized successfully by using the proposed software and the data measured approximately once a month. CONCLUSION: By determining similar transfer functions and selecting some adequate indicators, the software proposed can be adapted for quality assessment in other enclosed water bodies.