OneScience-Group/RainRunoff-LSTM

RainRunoff-LSTM

Model Introduction

RainRunoff-LSTM predicts daily streamflow from long meteorological histories. It supports basin-specific, regional, and regional-pretraining experiments.

Paper: Rainfall–runoff modelling using Long Short-Term Memory (LSTM) networks
https://doi.org/10.5194/hess-22-6005-2018

Model Description

The model was proposed by researchers at the University of Natural Resources and Life Sciences Vienna. It was trained with CAMELS Daymet precipitation, minimum and maximum temperature, shortwave radiation, vapor pressure, and streamflow. It supports daily rainfall-runoff simulation, regional sharing, and basin fine-tuning.

Usage Instructions

hf download OneScience-Group/RainRunoff-LSTM --local-dir ./RainRunoff-LSTM
cd RainRunoff-LSTM

Environment Dependencies

Hardware Requirements

• A GPU or DCU is recommended.
• A CPU can be used for connectivity validation with the default small-sample configuration.
• DCU users should install DTK 25.04.2 or a compatible OneScience-recommended version first.

DCU Environment

# Activate DTK and Conda first
conda create -n onescience311 python=3.11 -y
conda activate onescience311
pip install onescience[earth-dcu] -i http://mirrors.onescience.ai:3141/pypi/simple/ --trusted-host mirrors.onescience.ai

GPU Environment

# Activate Conda first
conda create -n onescience311 python=3.11 -y libstdcxx-ng=12 libgcc-ng=12 gcc_linux-64=12 gxx_linux-64=12
conda activate onescience311
pip install onescience[earth-gpu] -i http://mirrors.onescience.ai:3141/pypi/simple/ --trusted-host mirrors.onescience.ai

python scripts/fake_data.py
python scripts/train.py
torchrun --standalone --nproc_per_node=2 scripts/train.py
python scripts/inference.py
python scripts/result.py

Trained Weights

No weights are bundled under weight/. The paper does not publish a directly loadable official pretrained checkpoint.

Citation and License

This repository is an independent engineering reproduction of the public RainRunoff-LSTM specifications.

The original paper is licensed under CC BY 4.0; related code and CAMELS data retain their respective terms.