# NeuralHydrology -- Interpreting LSTMs in Hydrology

## TL;DR

This paper explores how LSTM networks can be interpreted in hydrology, demonstrating that trained models internally learn to represent hydrological processes consistent with domain knowledge, thus enhancing interpretability in environmental applications.

## Contribution

It introduces methods to interpret LSTM internals in rainfall-runoff forecasting, linking neural network representations to hydrological system understanding.

## Key findings

- LSTMs can model hydrological reservoirs and storages.
- Interpreted internal states align with hydrological concepts.
- Models successfully applied to different catchments.

## Abstract

Despite the huge success of Long Short-Term Memory networks, their applications in environmental sciences are scarce. We argue that one reason is the difficulty to interpret the internals of trained networks. In this study, we look at the application of LSTMs for rainfall-runoff forecasting, one of the central tasks in the field of hydrology, in which the river discharge has to be predicted from meteorological observations. LSTMs are particularly well-suited for this problem since memory cells can represent dynamic reservoirs and storages, which are essential components in state-space modelling approaches of the hydrological system. On basis of two different catchments, one with snow influence and one without, we demonstrate how the trained model can be analyzed and interpreted. In the process, we show that the network internally learns to represent patterns that are consistent with our qualitative understanding of the hydrological system.