graphcast

POI Radio monitors subgraph data integrity in real time using Graphcast SDK

Gossip about Subgraphs with other Graph Protocol Indexers

POC for the Graphcast sdk, as well as a single Radio example

Project Airstorm is an AI-powered weather visualization tool for AFGSC aviation, integrating GraphCast AI for real-time, high-accuracy forecasts. Built with React, Flask, AWS, and Google APIs, it enhances mission planning with extreme weather alerts, predictive analytics, and seamless integration into military operations. 🚀

CLI for Graphcast

Simplified, reproducible reimplementation of DeepMind's GraphCast weather model using graph neural networks (JAX · Haiku · Jraph). Runs on a single GPU, with honest negative-results reporting and detailed failure-mode analysis of a heavily scaled-down forecasting model.

AI+学科交叉教学案例集：涵盖生物、化学、物理、数学、天文、文史等19个实验

Monitor gossip on Graphcast

Explainable AI analysis of DeepMind's GraphCast weather forecasting model — applied feature occlusion, SHAP, permutation feature importance, and integrated gradients to Storm Zeynep (B.Sc. Thesis, HTW Berlin).

Weather data service that integrates EUMETSAT historical satellite data with Google GraphCast and AIFS forecasts through the Model Context Protocol (MCP)l for intelligent weather prediction for LLM models

Operational AIFS, GraphCast and enhanced hourly PanguWeather forecasting system with HPC orchestration, GRIB processing and GeoJSON/MBTiles visualization

Do AI weather models refuse to forecast unseen extremes? GraphCast, Pangu and FuXi on every 2m temperature record of 2020. No hard ceiling.

Subgraph API for querying Graphcast IDs

A small, open check of how AI weather forecasts (GraphCast, Pangu-Weather, FuXi, GenCast, ECMWF AIFS) compare with ECMWF IFS once turned into real wind power in 15 European countries.