Inductive Predictions of Extreme Hydrologic Events in The Wabash River Watershed
This enables prediction of extreme hydrologic events in new geographical areas using a model trained on watershed data, addressing water resource management challenges.
The authors developed a machine learning method using bidirectional LSTM with timestep reduction to predict extreme hydrologic events like droughts from hydrological and meteorological data in the Wabash River Watershed, achieving faster training than complex attention networks without sacrificing accuracy and demonstrating spatially-inductive predictions in unobserved locations.
We present a machine learning method to predict extreme hydrologic events from spatially and temporally varying hydrological and meteorological data. We used a timestep reduction technique to reduce the computational and memory requirements and trained a bidirection LSTM network to predict soil water and stream flow from time series data observed and simulated over eighty years in the Wabash River Watershed. We show that our simple model can be trained much faster than complex attention networks such as GeoMAN without sacrificing accuracy. Based on the predicted values of soil water and stream flow, we predict the occurrence and severity of extreme hydrologic events such as droughts. We also demonstrate that extreme events can be predicted in geographical locations separate from locations observed during the training process. This spatially-inductive setting enables us to predict extreme events in other areas in the US and other parts of the world using our model trained with the Wabash Basin data.