Short-term Streamflow and Flood Forecasting based on Graph Convolutional Recurrent Neural Network and Residual Error Learning

TL;DR

This paper introduces a novel graph convolutional recurrent neural network with residual error learning for short-term streamflow and flood forecasting, significantly improving accuracy within a 1-6 hour window to aid flood risk mitigation.

Contribution

It presents a new spatiotemporal neural network model that effectively addresses data errors from rating curves, enhancing short-term flood forecasting accuracy.

Findings

Outperforms traditional models in 1-6 hour forecasts

Residual error learning improves prediction accuracy

Provides a reliable tool for flood risk mitigation

Abstract

Accurate short-term streamflow and flood forecasting are critical for mitigating river flood impacts, especially given the increasing climate variability. Machine learning-based streamflow forecasting relies on large streamflow datasets derived from rating curves. Uncertainties in rating curve modeling could introduce errors to the streamflow data and affect the forecasting accuracy. This study proposes a streamflow forecasting method that addresses these data errors, enhancing the accuracy of river flood forecasting and flood modeling, thereby reducing flood-related risk. A convolutional recurrent neural network is used to capture spatiotemporal patterns, coupled with residual error learning and forecasting. The neural network outperforms commonly used forecasting models over 1-6 hours of forecasting horizons, and the residual error learners can further correct the residual errors. This provides a more reliable tool for river flood forecasting and climate adaptation in this critical 1-6 hour time window for flood risk mitigation efforts.