A combined model based on secondary decomposition and the optimized support vector machine algorithm for regional rainfall forecasting

Abstract

The rainfall series exhibits uncertainty and non-stationarity. Improving the accuracy of rainfall prediction is of significant importance for flood prevention and mitigation. This study proposes a hybrid model and applies it to rainfall forecasting in the eastern region of Hubei Province. The proposed method first uses variational mode decomposition and improved complete ensemble empirical mode decomposition with adaptive noise to reduce high-frequency noise components. Then, particle swarm optimization and support vector machine are used for training and forecasting. Compared with other models, the prediction model after noise reduction shows better performance than the model without secondary decomposition, with results that are closer to the actual values. The proposed hybrid model outperforms other models, with the predicted trend more closely aligning with the actual data, and the value of R² of predictions for individual cities reaches 0.96. This study not only provides an efficient method for rainfall forecasting but also holds significant importance for understanding and addressing climate change.