Short-term Hourly Streamflow Prediction with Graph Convolutional GRU Networks
This work addresses flood prediction for communities at risk, but it is incremental as it builds on existing graph and recurrent neural network techniques for a specific study area.
The paper tackles short-term streamflow prediction for flood mitigation by developing a Graph Convolutional GRU model, which outperforms baseline methods like persistence and a Convolutional Bidirectional GRU network in predicting the next 36 hours of streamflow for a specific sensor location.
The frequency and impact of floods are expected to increase due to climate change. It is crucial to predict streamflow, consequently flooding, in order to prepare and mitigate its consequences in terms of property damage and fatalities. This paper presents a Graph Convolutional GRUs based model to predict the next 36 hours of streamflow for a sensor location using the upstream river network. As shown in experiment results, the model presented in this study provides better performance than the persistence baseline and a Convolutional Bidirectional GRU network for the selected study area in short-term streamflow prediction.