An Encoder-Decoder LSTM model for un-gauged flood prediction – Hydrometeorology and Surface Processes

An Encoder-Decoder Long Short-Term Memory neural network is developed and tested for a 6-hour lead-time runoff prediction employed in un-gauged catchments.

A schematic illustration of the ED-LSTM.

Input composition and prediction process of the ED-LSTM structure for one event.

References

Zhang, Y., Ragettli, S., Molnar, P., Fink, O., Peleg, N., Generalization of an Encoder-Decoder LSTM model for flood prediction in ungauged catchments. Journal of Hydrology, 641(B), 128577.

Code

https://github.com/yikuizh/edlstm_flood_prediction