{"data":{"event":{"id":"361cd2f2-188b-450c-9e67-e68d36323a3a","slug":"gencast-ensemble-weather","title":"GenCast beats ECMWF ENS on most 15-day ensemble weather targets","short_summary":"DeepMind’s GenCast, published in Nature on 4 December 2024, generated 0.25° 15-day ensemble forecasts that outperformed ECMWF ENS on 97.2 percent of 1,320 evaluated targets.","full_description":"GenCast is a diffusion-based probabilistic weather model trained on ERA5. The Nature paper reported better ensemble skill than ENS, including extremes, tropical-cyclone tracks, and wind-power applications, with a single 15-day member in about eight minutes on a TPU. Code and weights were released. Evaluation is on 2019 holdout data, not a multi-year operational campaign.","ledger_type":"benefit","primary_domain_id":"99fd8a40-a540-41d3-af98-57c1709ee422","event_status":"verified","event_date":"2024-12-04T00:00:00.000Z","discovery_date":"2024-12-04T00:00:00.000Z","first_published_date":"2024-12-04T00:00:00.000Z","last_reviewed_date":"2024-12-04T00:00:00.000Z","geographic_scope":"Global atmosphere","affected_population":"Ensemble-forecast users and energy-weather researchers","base_impact_tier":3,"base_score":"10.00","attribution_multiplier":"0.9000","evidence_multiplier":"0.7000","realization_multiplier":"0.6500","durability_multiplier":"0.6000","current_event_score":"2.457000","confidence_level":"high","score_explanation":"First widely reported ML ensemble to beat ENS on a broad scorecard (tier 3). Peer review. Realization is open model plus later productization, not full agency replacement. Medium durability.","methodology_version_id":"d7881163-fb23-4e71-8625-bac1a8662c0f","original_methodology_version_id":"d7881163-fb23-4e71-8625-bac1a8662c0f","published_at":"2024-12-04T00:00:00.000Z","created_at":"2026-09-13T20:47:20.075Z","updated_at":"2026-09-13T20:47:20.075Z","flags":[],"domain_name":"Climate & Environment","domain_slug":"climate-environment","methodology_version":"0.1"},"contributions":[{"id":"725f21e5-1958-44a5-ac34-94972e52a807","event_id":"361cd2f2-188b-450c-9e67-e68d36323a3a","model_id":"fbf3aff9-3e52-4068-b50d-f1e779187d7f","role_description":"Diffusion ensemble weather model compared with ECMWF ENS.","attribution_multiplier":"0.9000","credit_share":"1.0000","contribution_score":"2.457000","attribution_rationale":"The published GenCast system is the artifact.","attribution_confidence":"medium","first_used_date":"2024-12-04T00:00:00.000Z","model_version_if_known":null,"review_status":"approved","created_at":"2026-09-13T20:47:20.075Z","model_slug":"gencast","model_name":"GenCast","identity_class":"named_release","is_internal":false,"family_name":"GraphCast","family_slug":"google-deepmind-graphcast","organization_name":"Google DeepMind","organization_slug":"google-deepmind"}],"sources":[{"id":"586d9179-24bd-4e6b-9cf1-c72d926b1d5f","event_id":"361cd2f2-188b-450c-9e67-e68d36323a3a","url":"https://www.nature.com/articles/s41586-024-08252-9","canonical_url":"https://www.nature.com/articles/s41586-024-08252-9","source_type":"peer_reviewed","publisher":"Nature","author":null,"publication_date":"2024-12-04T00:00:00.000Z","retrieved_at":"2026-09-13T20:47:20.075Z","title":"Probabilistic weather forecasting with machine learning","excerpt":null,"content_hash":null,"source_reliability_class":"high","is_primary_source":true,"is_independent":true,"is_peer_reviewed":true,"archived_url":null,"created_at":"2026-09-13T20:47:20.075Z"},{"id":"bc39139a-bfe5-42f0-8ad8-392428b8ef90","event_id":"361cd2f2-188b-450c-9e67-e68d36323a3a","url":"https://deepmind.google/discover/blog/gencast-predicts-weather-and-the-risks-of-extreme-conditions-with-sota-accuracy/","canonical_url":"https://deepmind.google/discover/blog/gencast-predicts-weather-and-the-risks-of-extreme-conditions-with-sota-accuracy/","source_type":"developer_announcement","publisher":"Google DeepMind","author":null,"publication_date":"2024-12-04T00:00:00.000Z","retrieved_at":"2026-09-13T20:47:20.075Z","title":"GenCast predicts weather and the risks of extreme conditions","excerpt":null,"content_hash":null,"source_reliability_class":"low","is_primary_source":true,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-13T20:47:20.075Z"}],"claims":[{"id":"3d00fbd1-736a-4957-a490-b9f125da4eb1","event_id":"361cd2f2-188b-450c-9e67-e68d36323a3a","claim_text":"GenCast outperformed ENS on 97.2 percent of 1,320 evaluation targets in the paper.","claim_type":"outcome","claim_status":"supported","confidence_score":null,"created_at":"2026-09-13T20:47:20.075Z","updated_at":"2026-09-13T20:47:20.075Z"}],"revisions":[{"id":"2de17efc-3677-4afd-9dfe-1fd439b53f2c","event_id":"361cd2f2-188b-450c-9e67-e68d36323a3a","model_id":null,"previous_score":"0.000000","new_score":"2.457000","previous_factors":{},"new_factors":{"evidence":0.7,"base_score":10,"durability":0.6,"attribution":0.9,"realization":0.65},"change_reason":"Initial adjudicated seed score under methodology 0.1.","trigger_type":"seed","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-13T20:47:20.075Z"}],"secondary":[{"name":"Energy","slug":"energy"},{"name":"Physics","slug":"physics"}]},"methodology_version":"0.1"}