{"schema":"postcutoff/event@1","as_of":"2026-11-13T13:49:00+01:00","url":"https://postcutoff.com/e/2024-07-22-neuralgcm/","md":"https://postcutoff.com/e/2024-07-22-neuralgcm/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"2024-07-22-neuralgcm","date":"2024-07-22","date_precision":"day","short_title":"NeuralGCM: Google's hybrid physics-ML atmosphere model matches top weather forecasts and runs decades-long climate simulations","deck":null,"takeaway":"In Nature (Kochkov et al., 22 July 2024) Google introduced NeuralGCM.","category":"science","category_label":"Science & math","importance":3,"confidence":"high","status":{"key":"confirmed","labels":["Result confirmed"]},"sources":[{"n":1,"title":"Nature: Neural general circulation models for weather and climate","url":"https://www.nature.com/articles/s41586-024-07744-y","type":"paper","group":"primary","domain":"nature.com"},{"n":2,"title":"arXiv 2311.07222","url":"https://arxiv.org/abs/2311.07222","type":"paper","group":"primary","domain":"arxiv.org"},{"n":3,"title":"Google Research: NeuralGCM harnesses AI to better simulate long-range global precipitation","url":"https://research.google/blog/neuralgcm-harnesses-ai-to-better-simulate-long-range-global-precipitation/","type":"official","group":"primary","domain":"research.google"}],"official":3,"filed":"2026-09-29","updated":"2026-09-29","orgs":["Google Research","ECMWF","MIT","Harvard"],"title":"NeuralGCM: Google's hybrid physics-ML atmosphere model matches top weather forecasts and runs decades-long climate simulations","summary":"In Nature (Kochkov et al., 22 July 2024) Google introduced NeuralGCM. It pairs a differentiable spectral dynamical core with neural-network physics parameterisations trained end-to-end. It was competitive with ECMWF for 1–15-day forecasts, reproduced four decades of observed temperatures in AMIP-style runs, and needed 3–5 orders of magnitude less compute than conventional models.","key_facts":["Paper: 'Neural general circulation models for weather and climate', Nature 632, 1060–1066 (2024); arXiv 2311.07222","Hybrid: physics-based dynamical core + learned column physics, trained end-to-end through the solver","Runs at 8–40× coarser horizontal resolution than ECMWF IFS and global cloud-resolving models, giving 3–5 orders of magnitude compute savings","Stable multi-decade climate simulations, unlike pure-ML weather emulators at the time"],"key_numbers":[],"tags":["climate","weather","hybrid-model","jax","differentiable-physics"],"science":{"field":"climate-weather","subfield":"atmospheric modelling","problem":"Fast, accurate general circulation models for both weather and climate","result":"Hybrid differentiable GCM competitive with ECMWF on medium-range forecasts and able to run decades-long climate simulations at a fraction of the cost.","open_since":"","ai_system":["NeuralGCM"],"human_role":"Human-led research","verification":{"key":"peer-review","label":"Peer-reviewed","text":"Peer-reviewed in Nature"},"status":"confirmed","shock":""},"body_md":"## What happened\n\nUnlike GraphCast-style end-to-end emulators, NeuralGCM kept a numerical dynamical core and learned only the unresolved physics. That made it stable enough for climate-length runs.\n\n## Why it matters\n\nIt showed that ML can reach climate modelling, not only weather forecasting, and made differentiable hybrid GCMs a serious research direction.","disputed":[],"related":[{"id":"2024-12-04-gencast-ensemble-weather","url":"https://postcutoff.com/e/2024-12-04-gencast-ensemble-weather/","date":"2024-12-04","date_precision":"day","short_title":"GenCast: diffusion-based ensemble forecast beats ECMWF's ENS on 97% of targets","deck":null,"takeaway":"GenCast (Nature, Dec 2024) is a diffusion model producing probabilistic 15-day ensemble forecasts.","category":"science","category_label":"Science & math","importance":3,"confidence":"high","status":{"key":"confirmed","labels":["Result confirmed"]},"sources":2,"official":2,"filed":"2026-09-29","updated":"2026-09-29","orgs":["Google DeepMind"]},{"id":"2023-11-14-graphcast-weather","url":"https://postcutoff.com/e/2023-11-14-graphcast-weather/","date":"2023-11-14","date_precision":"day","short_title":"GraphCast: ML weather model beats the world's best physics-based 10-day forecast on 90% of targets","deck":null,"takeaway":"It triggered the rapid move of AI weather models into operational forecasting worldwide.","category":"science","category_label":"Science & math","importance":4,"confidence":"high","status":{"key":"confirmed","labels":["Result confirmed"]},"sources":3,"official":3,"filed":"2026-09-29","updated":"2026-09-29","orgs":["Google DeepMind"]}],"people":[],"posts":[],"videos":[],"models":[],"changes":[{"date":"2026-09-29","type":"filed","text":"Created"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-09-29","run":null,"sources_read":null,"updated":"2026-09-29","human_review":null,"version":{"date":"2026-10-04"}},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":null,"in_training_data":true},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":null,"in_training_data":true},{"model_id":"gemini-4-hydrogen","name":"Gemini 4 Hydrogen","cutoff":"2026-03","days_after":null,"in_training_data":true},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":null,"in_training_data":true}],"short_url":null}