OneScience-Group/pysteps

pysteps

Model Introduction

pysteps is an open-source framework for probabilistic precipitation nowcasting. This reproduction focuses on optical flow, cascade decomposition, AR(2), and STEPS ensemble generation.

Paper: Pysteps: an open-source Python library for probabilistic precipitation nowcasting (v1.0)
https://doi.org/10.5194/gmd-12-4185-2019

Model Description

The method was proposed by teams from the Finnish Meteorological Institute, MeteoSwiss, ETH Zurich, Colorado State University, and collaborators. It estimates motion, cascade, and autoregressive parameters online from five-minute radar sequences collected in several countries. It supports one- to three-hour probabilistic precipitation nowcasting and ensemble uncertainty analysis.

Usage Instructions

hf download OneScience-Group/pysteps --local-dir ./pysteps
cd pysteps

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

STEPS has no offline gradient training. Inference generates a finite ensemble with shape [24,12,128,128], and evaluation reports RMSE and spread.

Trained Weights

No weights are bundled under weight/. pysteps estimates parameters online and does not use pretrained neural-network weights; the official software is available at https://github.com/pySTEPS/pysteps.

Citation and License

This repository is an independent engineering reproduction of the public pysteps specifications.

The original paper is licensed under CC BY 4.0 and the official pysteps software under BSD-3-Clause; the paper, software, and radar data retain their respective terms.