---
title: "Ryan Keisler Open-Sources the 2022 Model That Sparked ML Weather Forecasting"
subtitle: "Ryan Keisler open-sources the graph neural network from his influential 2022 paper, delivering 10-day global weather forecasts in under a minute."
url: "https://alphasignal.ai/news/ryan-keisler-open-sources-the-2022-model-that-sparked-ml-weather-forecasting"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-05-04T02:42:56.000Z"
updated: "2026-08-22T04:01:45.455Z"
type: "Repo"
topics:
  - "Open Source"
  - "Llms"
subtopics:
  - "Vision Language"
upvotes: 1453
source_url: "https://github.com/rkeisler/keisler-2022"
comments: 0
truncated: true
---

# Ryan Keisler Open-Sources the 2022 Model That Sparked ML Weather Forecasting

> Ryan Keisler open-sources the graph neural network from his influential 2022 paper, delivering 10-day global weather forecasts in under a minute.

The 2022 paper that helped kick off the machine learning weather forecasting wave finally has an official code release. Ryan Keisler [open-sourced the model](https://github.com/rkeisler/keisler-2022) from [Forecasting Global Weather with Graph Neural Networks](https://arxiv.org/abs/2202.07575), giving anyone with a laptop the ability to run global atmospheric forecasts that were, until recently, the domain of national supercomputing centers.

The release is intentionally lightweight. A 10-day global forecast runs in under a minute on a GPU, roughly two minutes on an 8-vCPU machine, and the entire pipeline is a single Python script with a JAX backend. It ships under an MIT license with scripts for evaluation, sensitivity analysis, and a Hurricane Sandy case study.

## How the model actually works

The system learns to step the current 3D atmospheric state forward by six hours, and multiple steps chain together to produce skillful forecasts several days out. Under the hood, it is a three-stage graph neural network:

* **Encoder** maps ERA5 latitude and longitude fields onto an H3 hexagonal mesh (Uber's spherical grid system, which avoids the pole distortion of a rectangular lat/lon grid).
* **Processor** runs 9 rounds of message passing on that mesh, where each hexagonal cell exchanges information with its neighbors.

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [rkeisler/keisler-2022 GitHub repository](https://github.com/rkeisler/keisler-2022)
- [Forecasting Global Weather with Graph Neural Networks (arXiv)](https://arxiv.org/abs/2202.07575)
- [openclimatefix/graph_weather](https://github.com/openclimatefix/graph_weather)
- [Forecasting Global Weather with Graph Neural Networks (Semantic Scholar)](https://www.semanticscholar.org/paper/Forecasting-Global-Weather-with-Graph-Neural-Keisler/80d7b9180299ed1954dbc3acde4ad4efa8974e0a)