Key points are not available for this paper at this time.
Poor water quality is a serious problem in the world which threatens human health, ecosystems, plant and animal life. Prediction of surface water quality is a main concern in water resource and environmental systems.The lake Nokoue in Benin, the countrys main water body, is experiencing without equivox pollution. Usually, the water quality index is evaluated manually by complex mathematical formulas with major risks of errors. This study discusses the development and validation of a Random Forest and Artificial Neural Network (ANN) model in estimating water quality index (WQI) in the lake Nokoue.The two models have been developed and tested using data from 20 monitoring stations over a period of 12 months. The modeling data was divided into two sets. For the first set, RF and ANN were trained, tested and validated using 12 physical-parameters as input parameters. A detailed comparison of the overall performance showed that prediction of the random forest (RF) model was better than artificial neural networks with coefficient of correlation (R2)=0.98, root mean squared error (RMSE)=0.12, explained variance score (EVS)=0.98 and mean absolute error (MAE)= 0.14 at training phase while and at the validation phase their values are0.80, 0.19, 0.23, 0.74 respectively which demonstrates that RF is capable of estimating WQI with acceptable accuracy. This method simplifies the calculation of the WQI and reduce substantial efforts and time by optimizing the computations. This will help in taking appropriate preventive measures to control the water quality of lake Nokoue through associated chemical treatments.
Dabire et al. (Tue,) studied this question.