Water Quality Prediction of Small Watershed Based on Wavelet Neural Network

In recent years, China has faced a very serious issue of water pollution, which has had a dreadful impact on the ecological environment and human health. Due to the rapid growth of industry and economy, water pollution around China's urban areas has received extensive attentions. Among them, small watershed pollution, which is difficult to sample real-time data, is particularly prominent. Therefore, it is extremely important to propose new, better and more reliable prediction models to accurately predict the water quality in these small watersheds. This paper selects the water quality data of small watersheds around Chongqing for the study to come up with a new wavelet neural network model of forecasting using small amount of data to predict the China Water Quality Index (WQI). The present study is aimed to improve the prediction results by minimizing the prediction errors of current machine learning algorithms by considering the main environmental pollutant in small watersheds as input. The results show that when there is a strong interaction and correlation between the water quality characteristic attribute and WQI, the MAPE of the wavelet neural network model training results will decrease. In addition, the geographical location is found to play an important role in the Chongqing WQI forecast.