Hybrid ARIMA and GRNN Model for Accurate Short-Term Water Level Prediction

To more effectively predict water levels, a hybrid method combining time-series analysis and artificial neural networks was proposed. The water level time series were sampled at regular intervals and forecast using the Auto-regressive Integrated Moving Average (ARIMA) model. The results from the ARIMA model were then utilized as inputs for the Generalized Regression Neural Network (GRNN) to conduct accurate forecasting. The forecasting results demonstrate that this hybrid model can effectively predict short-term water level trends with stable prediction capability and high accuracy.