{ "id": "AIBD-0088", "title": "NHC blended Google DeepMind's AI ensemble into official Hurricane Melissa forecasts before its Category 5 landfall in Jamaica", "summary": "NHC blended Google DeepMind's AI ensemble into official Melissa forecasts from its first advisory on 21 October 2025; Cat 5 landfall in Jamaica 28 October 2025.", "date_observed": "2025-10-21", "date_basis": "unassessed", "date_precision": "unassessed", "benefit_status": null, "source_collection": null, "source_entry": null, "account": "The US National Hurricane Center (NHC) blended Google DeepMind's experimental AI hurricane model (ensemble mean GDMI) into official forecasts for Tropical Storm Melissa. It did so from its first advisory, 21 October 2025. Discussion 1 placed the official track between the DeepMind ensemble mean and the consensus aid HCCA. Discussion 2 described the track as a blend of the DeepMind mean, HCCA and TVCA. Discussion 8 (23 October) moved the track toward a blend of the DeepMind mean and HCCA. Discussion 10 blended the HCCA and GDMI aids. It set an intensity forecast in best agreement with GDMI after all 50 DeepMind members showed a major hurricane. On 28 October 2025 Melissa made landfall in southwestern Jamaica near New Hope with 185 mph winds and 892 mb pressure. IFRC reported 45 deaths and about 1.6 million people affected. It relayed a World Bank/IDB estimate of US$8.8 billion in physical damage. Its CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time.", "category": "protective response", "unexpected": false, "purpose": "Experimental AI tropical cyclone forecast model providing ensemble track and intensity guidance to NHC forecasters.", "system": "Google DeepMind AI hurricane forecast model, ensemble mean designated GDMI in NHC guidance", "model_version": null, "developer": "DeepMind", "deployer": "NOAA National Hurricane Center", "location": "New Hope, Jamaica", "domain": "Weather forecasting / tropical cyclones", "segment": "Public-sector meteorological service", "benefit_type": "Improved Task Accuracy", "taxonomy_note": "The AI model's contribution to official forecasts is read as forecast accuracy, with no error figure for this storm.", "registry_ids": {}, "ratings": [ { "group": "personal", "depth": null, "reason": "No cited source attributes an individual-level outcome to the AI guidance. Impacts are reported only at population level." }, { "group": "organizational", "depth": 2, "reason": "NHC's official track and intensity forecasts were materially shaped by GDMI (Discussions 8, 10, 13, 14). NHC's Science Operations Officer called the AI guidance very valuable for Melissa." }, { "group": "societal", "depth": 3, "reason": "Official forecasts for a Category 5 landfall affecting about 1.6 million people. A CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time. The review says its early actions were informed by forecast risks (NHC is not named). Material to many, but no counterfactual that the AI changed outcomes." } ], "ai_role": { "status": "assessed", "statement": "NHC forecasters state in Discussions 8, 10, 12, 13 and 14 that official track and intensity forecasts were blended or aligned with Google DeepMind's ensemble. The official forecast remains a human synthesis of several aids. So the AI contribution to outcomes in Jamaica is not isolated." }, "evidence": [ { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 1, 1100 AM EDT Tue Oct 21 2025, forecasters Cangialosi/Bucci. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 1", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml", "date": "2025-10-21", "written_by": "deployer", "quote": "The NHC official track forecast lies between the Google DeepMind ensemble mean track and the correct consensus aid, HCCA." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 2, 500 PM EDT Tue Oct 21 2025, forecaster Cangialosi. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 2", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml", "date": "2025-10-21", "written_by": "deployer", "quote": "This prediction is a little left of the previous forecast and remains a blend of the latest Google DeepMind ensemble mean, the corrected consensus aid HCCA, and simple consensus TVCA." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 8, 500 AM EDT Thu Oct 23 2025, forecaster Cangialosi. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 8", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.008.shtml", "date": "2025-10-23", "written_by": "deployer", "quote": "The NHC track forecast is shifted a little to the right of the previous one, trending toward a blend of the latest Google DeepMind ensemble mean and HCCA solution." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 10, 500 PM EDT Thu Oct 23 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 10", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.010.shtml", "date": "2025-10-23", "written_by": "deployer", "quote": "the latest NHC track was only nudged slightly southward from this morning, blending the reliable HCCA and GDMI aids" }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 13, 1100 AM EDT Fri Oct 24 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 13", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.013.shtml", "date": "2025-10-24", "written_by": "deployer", "quote": "Every single Google DeepMind member shows the system become a Category 4 hurricane or higher" }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 14, 500 PM EDT Fri Oct 24 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 14", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml", "date": "2025-10-24", "written_by": "deployer", "quote": "The NHC intensity forecast is on the high side of the guidance envelope, and closest too the Google DeepMind ensemble mean (GDMI)." }, { "kind": "report", "citation": "National Hurricane Center (2025). Hurricane Melissa Tropical Cyclone Update, 100 PM EDT Tue Oct 28 2025. NOAA/NWS.", "title": "Hurricane Melissa Tropical Cyclone Update (landfall)", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.update.10281700.shtml", "date": "2025-10-28", "written_by": "deployer", "quote": "CATEGORY 5 MELISSA MAKES LANDFALL IN JAMAICA...ONE OF THE MOST POWERFUL HURRICANE LANDFALLS ON RECORD IN THE ATLANTIC BASIN" }, { "kind": "press release", "citation": "National Weather Service (n.d.). AI in Hurricane Forecasting at the National Hurricane Center: A Q&A with NHC Science Operation Officer Wallace Hogsett. NOAA/NWS News Around NOAA.", "title": "AI in Hurricane Forecasting at the National Hurricane Center", "url": "https://www.weather.gov/news/261102-AI-Hurricane-Forecasting", "date": null, "written_by": "deployer", "quote": "The AI models honed in very early on the likely track and intensity and provided very valuable guidance to complement our traditional NWP guidance." }, { "kind": "report", "citation": "Cangialosi, J. P., Martinez, J., Reinhart, B., Papin, P. (2026). National Hurricane Center Forecast Verification Report: 2025 Hurricane Season. NOAA/NHC, 30 March 2026.", "title": "NHC Forecast Verification Report, 2025 Hurricane Season", "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "date": "2026-03-30", "written_by": "deployer", "quote": "the NHC track errors are notably lower than average for major hurricanes, including Humberto and Melissa in 2025" }, { "kind": "report", "citation": "Kelly, Larry A., José Álamo, and Aidan D. Mahoney. Hurricane Melissa (AL132025), 21–31 October 2025. National Hurricane Center Tropical Cyclone Report. Revised May 19, 2026, originally February 25, 2026. 16–17, Tables 5b and 6b.", "title": "Hurricane Melissa (AL132025): Tropical Cyclone Report", "url": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "date": "2026-02-25", "written_by": "deployer", "quote": "The NHC OFCL forecasts were comparable to GDMI through 96 h, but performed slightly better than GDMI at 120 h." } ], "replication": { "produced_by": [ "Experimental Google DeepMind ensemble guidance available in real time to NHC forecasters", "Forecasters choosing to blend GDMI with consensus aids (HCCA) in official track and intensity forecasts", "A slow-moving, rapidly intensifying hurricane where models disagreed on track relative to Jamaica" ], "must_hold": [ "AI ensemble output delivered on time and verified before operational use, as NHC describes", "Human forecasters retain the final official forecast and integrate conventional NWP guidance", "Forecast accuracy can be assessed against the observed track and intensity. NHC’s Tropical Cyclone Report, revised 19 May 2026, provides that comparison for Melissa" ], "echoes": [ { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 Atlantic season track forecasts", "model_version": null, "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" }, { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 eastern North Pacific track forecasts (GDMI beat official and consensus aids at 48–120 h)", "model_version": null, "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" } ] }, "paired_incident": { "database": "AI Incident Database", "identifier": "1546", "title": "Google Android Earthquake Alerts System Reportedly Underestimated Kahramanmaraş Earthquake and Failed to Send Urgent Warnings to Millions", "url": "https://incidentdatabase.ai/cite/1546", "severity": null, "group": "societal", "basis": "Google-built hazard early-warning system underestimated the Kahramanmaras earthquake and failed to warn millions; same developer family and public-warning task as DeepMind's Melissa forecasts." }, "headline": "NHC blends DeepMind's ensemble into Hurricane Melissa forecasts", "metric": null, "comparator": null, "later": { "status": "persisted", "date": "2026-03-30", "note": "NHC's 2025 verification report, five months after landfall, records low official track errors for Melissa and the DeepMind ensemble mean outperforming the official Atlantic track forecast at 12 to 72 hours.", "quote": "It has been seen in multiple cases during the past few years that the NHC track errors are notably lower than average for major hurricanes, including Humberto and Melissa in 2025.", "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" }, "value": null, "developer_country": "United Kingdom", "deployer_country": "United States", "unrated_because": null, "brief_notes": null, "brief": null, "revisions": [ { "date": "2026-10-20", "change": "Record drafted from the cited sources.", "source": null, "editor": "AIBD" }, { "date": "2026-10-20", "change": "Paired with AI Incident Database incident 1546 (medium confidence); severity not given by the source.", "source": "https://incidentdatabase.ai/cite/1546", "editor": "AIBD" }, { "date": "2026-10-20", "change": "Set the date observed to 21 October 2025 and reworded the summary and account because Discussions 1 and 2 show the Google DeepMind ensemble mean shaping the official track from the first advisory, added those two discussions as evidence, corrected the Discussion 14 title to “Tropical Storm Melissa” as its header reads, attributed the US$8.8 billion damage figure to the World Bank/IDB estimate relayed by IFRC, and tied the DREF's early actions to “forecast risks” as the review says rather than to NHC.", "source": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml; https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml; https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml; https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-dref-no-mdrjm005-operation-update-3; https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-imminent-dref-mdrjm005-dref-operational-review-report-april-may-2026", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Added what happened since from the cited sources.", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Benefit type set to Improved Task Accuracy under Trustible's full taxonomy of 23 types; it was Improved Knowledge Discovery among the six first used.", "source": "https://trustible.ai/post/a-framework-for-measuring-the-benefits-of-ai/", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Country of the developer and the deployer added from the organization's own site or a registry.", "source": "https://en.wikipedia.org/wiki/Google_DeepMind; https://www.nhc.noaa.gov/aboutintro.shtml", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Prose pass: a rating reason, account, the replication conditions, the system's part rewritten in shorter sentences with abbreviations expanded; facts, figures and quotations unchanged.", "source": null, "editor": "AIBD" }, { "date": "2026-10-21", "change": "Notes about which sources could be read were reworded for readers; facts, figures and quotations unchanged.", "source": null, "editor": "AIBD" }, { "date": "2026-10-22", "change": "Organization name standardized to DeepMind.", "source": null, "editor": "AIBD" }, { "date": "2026-10-24", "change": "Moved two IFRC reports to background source notes because they describe storm impacts and humanitarian action, without establishing the AI forecast’s contribution to those outcomes. Preserved their citations, excerpts, and the existing account and depths. Retained NHC evidence of forecast use and verification, attributed to the deployer rather than an independent AI-benefit evaluator.", "source": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml; https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "editor": "AIBD" }, { "date": "2026-10-25", "change": "Added NHC’s revised Hurricane Melissa report, replaced the obsolete statement that it was unpublished, and separated forecast verification from humanitarian background.", "source": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "editor": "AIBD" } ], "confidence": "claimed", "shade": "low", "claim_status": "The deployer's account", "claim_basis": "Sources: nine reports from the deployer and the deployer's press release. The cited evidence includes no independent source and no finding by a regulator or court.", "profile": { "case_id": "AIBD-0088", "catalog_version": "1.0.0", "source_fingerprint": "7494545dd8aeba975dc7387c52bc23ab138bc6a416517d50ebb709bf803d7efa", "collection_refs": [ { "collection_id": "aibd", "version": "snapshot-8bf7cfec9cb185aa" }, { "collection_id": "aibd", "version": "2026-10-25" }, { "collection_id": "viator", "version": "2026-10-25" }, { "collection_id": "adyt", "version": "2026-10-25" }, { "collection_id": "chanticleer", "version": "2026-10-25" } ], "source_ids": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009", "AIBD-0088-source-0010" ], "families": { "strategic assessments": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "record_ids": [] }, "evaluations": { "state": "value_present", "value": { "kind": "identifiers", "identifiers": [ "NHC-Melissa-AL132025-forecast-verification" ], "role": "Forecast verification for Hurricane Melissa" }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "record_ids": [ "NHC-Melissa-AL132025-forecast-verification" ] }, "monetary observations": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "record_ids": [] }, "institutional decisions": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "record_ids": [] } }, "fields": { "actors[]": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "developer": "DeepMind", "deployer": "NOAA National Hurricane Center" } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "developer", "deployer" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "actors[].ownership_control": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "ai_contribution": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "ai_role": { "status": "assessed", "statement": "NHC forecasters state in Discussions 8, 10, 12, 13 and 14 that official track and intensity forecasts were blended or aligned with Google DeepMind's ensemble. The official forecast remains a human synthesis of several aids. So the AI contribution to outcomes in Jamaica is not isolated." } } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "ai_role" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "ai_relationship": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "category": "protective response", "ai_role": { "status": "assessed", "statement": "NHC forecasters state in Discussions 8, 10, 12, 13 and 14 that official track and intensity forecasts were blended or aligned with Google DeepMind's ensemble. The official forecast remains a human synthesis of several aids. So the AI contribution to outcomes in Jamaica is not isolated." } } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "ai_role", "category" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "assessments[].alternatives": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "assessments[].assessor_date": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "assessments[].depth": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "ratings": [ { "group": "personal", "reason": "No cited source attributes an individual-level outcome to the AI guidance. Impacts are reported only at population level." }, { "group": "organizational", "depth": 2, "reason": "NHC's official track and intensity forecasts were materially shaped by GDMI (Discussions 8, 10, 13, 14). NHC's Science Operations Officer called the AI guidance very valuable for Melissa." }, { "group": "societal", "depth": 3, "reason": "Official forecasts for a Category 5 landfall affecting about 1.6 million people. A CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time. The review says its early actions were informed by forecast risks (NHC is not named). Material to many, but no counterfactual that the AI changed outcomes." } ] } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "ratings" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "assessments[].display_confidence": { "state": "value_present", "value": { "kind": "structured", "data": { "confidence": "low", "evidence_word": "claimed", "claim_status": "The deployer's account", "basis": "Sources: eight reports from the deployer and the deployer's press release. The cited evidence includes no independent source and no finding by a regulator or court.", "interpretation": "Existing display rule preserved; the richer outcome-specific confidence question still needs review." } }, "provenance": { "kind": "curatorial", "basis": "Recomputed from retained evidence using the existing display rule; source originality does not establish an independent evaluation of the benefit.", "record_fields": [ "evidence" ], "source_refs": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ] }, "review_state": "record_mapped", "coverage": "partial" }, "assessments[].rationale": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "ratings": [ { "group": "personal", "reason": "No cited source attributes an individual-level outcome to the AI guidance. Impacts are reported only at population level." }, { "group": "organizational", "depth": 2, "reason": "NHC's official track and intensity forecasts were materially shaped by GDMI (Discussions 8, 10, 13, 14). NHC's Science Operations Officer called the AI guidance very valuable for Melissa." }, { "group": "societal", "depth": 3, "reason": "Official forecasts for a Category 5 landfall affecting about 1.6 million people. A CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time. The review says its early actions were informed by forecast risks (NHC is not named). Material to many, but no counterfactual that the AI changed outcomes." } ] } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "ratings" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial", "display": { "answer": "GDMI and regional models increased forecasters’ confidence in predicting rapid intensification. NHC’s later review found that GDMI track forecasts performed comparably to its official forecasts through 96 hours. No study isolates the AI’s effect on protective actions or losses." } }, "assessments[].rubric_version": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "assessments[].status_basis": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "assessments[].target": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "ratings": [ { "group": "personal", "reason": "No cited source attributes an individual-level outcome to the AI guidance. Impacts are reported only at population level." }, { "group": "organizational", "depth": 2, "reason": "NHC's official track and intensity forecasts were materially shaped by GDMI (Discussions 8, 10, 13, 14). NHC's Science Operations Officer called the AI guidance very valuable for Melissa." }, { "group": "societal", "depth": 3, "reason": "Official forecasts for a Category 5 landfall affecting about 1.6 million people. A CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time. The review says its early actions were informed by forecast risks (NHC is not named). Material to many, but no counterfactual that the AI changed outcomes." } ] } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "ratings" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "availability": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "brief": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "case_kind": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "case_links[]": { "state": "value_present", "value": { "kind": "structured", "data": { "answer": "Google-built hazard early-warning system underestimated the Kahramanmaras earthquake and failed to warn millions; same developer family and public-warning task as DeepMind's Melissa forecasts.", "links": [ { "database": "AI Incident Database", "identifier": "1546", "title": "Google Android Earthquake Alerts System Reportedly Underestimated Kahramanmaraş Earthquake and Failed to Send Urgent Warnings to Millions", "url": "https://incidentdatabase.ai/cite/1546", "group": "societal", "basis": "Google-built hazard early-warning system underestimated the Kahramanmaras earthquake and failed to warn millions; same developer family and public-warning task as DeepMind's Melissa forecasts.", "relation": "comparable_harm" } ] } }, "provenance": { "kind": "legacy_record", "basis": "Existing paired_incident classification; no new source review.", "record_fields": [ "paired_incident" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "claims[].evidence_links": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "claims[].evidence_strength": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "claims[].missingness": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "claims[].proposition": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "cross_border_flows[]": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "decisions[].action_outcome": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "institutional decisions" }, "decisions[].grounds": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "institutional decisions" }, "decisions[].institution_task": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "institutional decisions" }, "decisions[].success_criterion": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "institutional decisions" }, "deployment_stage": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "economics[].amount": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].costs[]": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].counterfactual": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].currency_price_basis": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].gross_net": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].measure_type": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].period_unit": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].realization": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].recipient_payer": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].return_payback": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "economics[].valuation_method": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "monetary observations" }, "evaluations[].ai_uplift": { "state": "value_present", "value": { "kind": "text", "text": "NHC credits GDMI and regional models jointly with increasing forecasters’ confidence at unusually long lead times. It does not isolate GDMI’s effect on the official forecast." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "NHC credits GDMI and regional models jointly with increasing forecasters’ confidence at unusually long lead times. It does not isolate GDMI’s effect on the official forecast." } }, "evaluations[].comparator": { "state": "value_present", "value": { "kind": "text", "text": "NHC’s official forecasts, conventional model guidance, and climatology-persistence forecasts for the same storm." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "NHC’s official forecasts, conventional model guidance, and climatology-persistence forecasts for the same storm." } }, "evaluations[].design": { "state": "value_present", "value": { "kind": "text", "text": "Retrospective verification of operational forecasts against the storm’s observed track and intensity." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "Retrospective verification of operational forecasts against the storm’s observed track and intensity." } }, "evaluations[].effect": { "state": "value_present", "value": { "kind": "text", "text": "GDMI’s track forecasts were comparable to NHC’s official forecasts through 96 hours. NHC performed slightly better at 120 hours and had lower intensity errors than every guidance model through 96 hours." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "GDMI’s track forecasts were comparable to NHC’s official forecasts through 96 hours. NHC performed slightly better at 120 hours and had lower intensity errors than every guidance model through 96 hours." } }, "evaluations[].exposure": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "evaluations[].id": { "state": "value_present", "value": { "kind": "identifiers", "identifiers": [ "NHC-Melissa-AL132025-forecast-verification" ], "role": "Forecast verification for Hurricane Melissa" }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "NHC’s forecast verification for Hurricane Melissa (AL132025)." } }, "evaluations[].measurement_method": { "state": "value_present", "value": { "kind": "text", "text": "NHC compared forecasts at matching lead times with its final estimates of the storm’s position and intensity. It compared official forecasts with model guidance on a common sample." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "NHC compared forecasts at matching lead times with its final estimates of the storm’s position and intensity. It compared official forecasts with model guidance on a common sample." } }, "evaluations[].metric": { "state": "value_present", "value": { "kind": "text", "text": "Track forecast error, intensity forecast error, and forecast lead time." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "Track forecast error, intensity forecast error, and forecast lead time." } }, "evaluations[].reproduction_materials": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "evaluations[].setting_sample": { "state": "value_present", "value": { "kind": "text", "text": "Hurricane Melissa in the Atlantic, 21–31 October 2025. The report compares forecasts at lead times through 120 hours." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "Hurricane Melissa in the Atlantic, 21–31 October 2025. The report compares forecasts at lead times through 120 hours." } }, "evaluations[].uncertainty_limitations": { "state": "value_present", "value": { "kind": "text", "text": "The sources establish that forecasters used the guidance. They do not establish that it changed protective actions, reduced losses, or prevented deaths." }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "The sources establish that forecasters used the guidance. They do not establish that it changed protective actions, reduced losses, or prevented deaths." } }, "external_references[]": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "paired_incident": { "database": "AI Incident Database", "identifier": "1546", "title": "Google Android Earthquake Alerts System Reportedly Underestimated Kahramanmaraş Earthquake and Failed to Send Urgent Warnings to Millions", "url": "https://incidentdatabase.ai/cite/1546", "group": "societal", "basis": "Google-built hazard early-warning system underestimated the Kahramanmaras earthquake and failed to warn millions; same developer family and public-warning task as DeepMind's Melissa forecasts." } } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "paired_incident" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "follow_up[]": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "later": { "status": "persisted", "date": "2026-03-30", "note": "NHC's 2025 verification report, five months after landfall, records low official track errors for Melissa and the DeepMind ensemble mean outperforming the official Atlantic track forecast at 12 to 72 hours.", "quote": "It has been seen in multiple cases during the past few years that the NHC track errors are notably lower than average for major hurricanes, including Humberto and Melissa in 2025.", "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" } } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "later" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial", "display": { "answer": "NHC published its Tropical Cyclone Report on 25 February 2026 and revised it on 19 May. The report evaluates the forecasts against the storm’s observed track and intensity." } }, "human_oversight": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "id": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "id": "AIBD-0088" } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "id" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "infrastructure_dependencies[]": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "outcomes[].beneficiaries": { "state": "value_present", "value": { "kind": "text", "text": "National Hurricane Center forecasters. People in Melissa’s path were the intended beneficiaries of the forecasts." }, "provenance": { "kind": "legacy_record", "basis": "Moved the specific information already present in the entry into its applicable question. No new source review is implied.", "record_fields": [ "account", "ratings" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial", "display": { "answer": "National Hurricane Center forecasters. People in Melissa’s path were the intended beneficiaries of the forecasts." } }, "outcomes[].beneficiary_count": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "display": { "answer": "No count of people who benefited from the AI guidance is established." } }, "outcomes[].benefit_type": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "benefit_type": "Improved Task Accuracy", "taxonomy_note": "The AI model's contribution to official forecasts is read as forecast accuracy, with no error figure for this storm." } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "benefit_type", "taxonomy_note" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "outcomes[].claim": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "account": "The US National Hurricane Center (NHC) blended Google DeepMind's experimental AI hurricane model (ensemble mean GDMI) into official forecasts for Tropical Storm Melissa. It did so from its first advisory, 21 October 2025. Discussion 1 placed the official track between the DeepMind ensemble mean and the consensus aid HCCA. Discussion 2 described the track as a blend of the DeepMind mean, HCCA and TVCA. Discussion 8 (23 October) moved the track toward a blend of the DeepMind mean and HCCA. Discussion 10 blended the HCCA and GDMI aids. It set an intensity forecast in best agreement with GDMI after all 50 DeepMind members showed a major hurricane. On 28 October 2025 Melissa made landfall in southwestern Jamaica near New Hope with 185 mph winds and 892 mb pressure. IFRC reported 45 deaths and about 1.6 million people affected. It relayed a World Bank/IDB estimate of US$8.8 billion in physical damage. Its CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time.", "benefit_type": "Improved Task Accuracy" } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "account", "benefit_type" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial", "display": { "answer": "National Hurricane Center forecasters used Google DeepMind’s guidance in their official forecasts for Hurricane Melissa." } }, "outcomes[].distribution": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "outcomes[].duration": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "outcomes[].id": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "outcomes[].stakeholder_group": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "ratings": [ { "group": "personal", "reason": "No cited source attributes an individual-level outcome to the AI guidance. Impacts are reported only at population level." }, { "group": "organizational", "depth": 2, "reason": "NHC's official track and intensity forecasts were materially shaped by GDMI (Discussions 8, 10, 13, 14). NHC's Science Operations Officer called the AI guidance very valuable for Melissa." }, { "group": "societal", "depth": 3, "reason": "Official forecasts for a Category 5 landfall affecting about 1.6 million people. A CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time. The review says its early actions were informed by forecast risks (NHC is not named). Material to many, but no counterfactual that the AI changed outcomes." } ] } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "ratings" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "outcomes[].status": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "outcomes[].tradeoffs": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "outcomes[].unexpected": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "unexpected": false, "purpose": "Experimental AI tropical cyclone forecast model providing ensemble track and intensity guidance to NHC forecasters." } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "unexpected", "purpose" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "places[]": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "location": "New Hope, Jamaica", "developer_country": "United Kingdom", "deployer_country": "United States" } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "location", "developer_country", "deployer_country" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "purpose": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "purpose": "Experimental AI tropical cyclone forecast model providing ensemble track and intensity guidance to NHC forecasters." } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "purpose" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "record_dates": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "date_observed": "2025-10-21", "revisions": [ { "date": "2026-10-20", "change": "Record drafted from the cited sources.", "editor": "AIBD" }, { "date": "2026-10-20", "change": "Paired with AI Incident Database incident 1546 (medium confidence); severity not given by the source.", "source": "https://incidentdatabase.ai/cite/1546", "editor": "AIBD" }, { "date": "2026-10-20", "change": "Set the date observed to 21 October 2025 and reworded the summary and account because Discussions 1 and 2 show the Google DeepMind ensemble mean shaping the official track from the first advisory, added those two discussions as evidence, corrected the Discussion 14 title to “Tropical Storm Melissa” as its header reads, attributed the US$8.8 billion damage figure to the World Bank/IDB estimate relayed by IFRC, and tied the DREF's early actions to “forecast risks” as the review says rather than to NHC.", "source": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml; https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml; https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml; https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-dref-no-mdrjm005-operation-update-3; https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-imminent-dref-mdrjm005-dref-operational-review-report-april-may-2026", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Added what happened since from the cited sources.", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Benefit type set to Improved Task Accuracy under Trustible's full taxonomy of 23 types; it was Improved Knowledge Discovery among the six first used.", "source": "https://trustible.ai/post/a-framework-for-measuring-the-benefits-of-ai/", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Country of the developer and the deployer added from the organization's own site or a registry.", "source": "https://en.wikipedia.org/wiki/Google_DeepMind; https://www.nhc.noaa.gov/aboutintro.shtml", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Prose pass: a rating reason, account, the replication conditions, the system's part rewritten in shorter sentences with abbreviations expanded; facts, figures and quotations unchanged.", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Notes about which sources could be read were reworded for readers; facts, figures and quotations unchanged.", "editor": "AIBD" }, { "date": "2026-10-22", "change": "Organization name standardized to DeepMind.", "editor": "AIBD" }, { "date": "2026-10-24", "change": "Moved two IFRC reports to background source notes because they describe storm impacts and humanitarian action, without establishing the AI forecast’s contribution to those outcomes. Preserved their citations, excerpts, and the existing account and depths. Retained NHC evidence of forecast use and verification, attributed to the deployer rather than an independent AI-benefit evaluator.", "source": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml; https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "editor": "AIBD" }, { "date": "2026-10-25", "change": "Added NHC’s revised Hurricane Melissa report, replaced the obsolete statement that it was unpublished, and separated forecast verification from humanitarian background.", "source": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "editor": "AIBD" } ] } }, "provenance": { "kind": "curatorial", "basis": "Moved two IFRC reports to background source notes because they describe storm impacts and humanitarian action, without establishing the AI forecast’s contribution to those outcomes. Preserved their citations, excerpts, and the existing account and depths. Retained NHC evidence of forecast use and verification, attributed to the deployer rather than an independent AI-benefit evaluator.", "record_fields": [ "revisions" ], "source_refs": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-24" }, "review_state": "record_mapped", "coverage": "partial" }, "replication.attempts[]": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "replication": { "produced_by": [ "Experimental Google DeepMind ensemble guidance available in real time to NHC forecasters", "Forecasters choosing to blend GDMI with consensus aids (HCCA) in official track and intensity forecasts", "A slow-moving, rapidly intensifying hurricane where models disagreed on track relative to Jamaica" ], "must_hold": [ "AI ensemble output delivered on time and verified before operational use, as NHC describes", "Human forecasters retain the final official forecast and integrate conventional NWP guidance", "Forecast accuracy can be assessed against the observed track and intensity. NHC’s Tropical Cyclone Report, revised 19 May 2026, provides that comparison for Melissa" ], "echoes": [ { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 Atlantic season track forecasts", "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" }, { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 eastern North Pacific track forecasts (GDMI beat official and consensus aids at 48–120 h)", "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" } ] } } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "replication" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "replication.conditions": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "replication": { "produced_by": [ "Experimental Google DeepMind ensemble guidance available in real time to NHC forecasters", "Forecasters choosing to blend GDMI with consensus aids (HCCA) in official track and intensity forecasts", "A slow-moving, rapidly intensifying hurricane where models disagreed on track relative to Jamaica" ], "must_hold": [ "AI ensemble output delivered on time and verified before operational use, as NHC describes", "Human forecasters retain the final official forecast and integrate conventional NWP guidance", "Forecast accuracy can be assessed against the observed track and intensity. NHC’s Tropical Cyclone Report, revised 19 May 2026, provides that comparison for Melissa" ], "echoes": [ { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 Atlantic season track forecasts", "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" }, { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 eastern North Pacific track forecasts (GDMI beat official and consensus aids at 48–120 h)", "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" } ] } } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "replication" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial", "display": { "answer": "Forecasters need timely model output, checks before operational use, and conventional guidance for comparison. Later observations are needed to assess forecast accuracy." } }, "revision_history[]": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "revisions": [ { "date": "2026-10-20", "change": "Record drafted from the cited sources.", "editor": "AIBD" }, { "date": "2026-10-20", "change": "Paired with AI Incident Database incident 1546 (medium confidence); severity not given by the source.", "source": "https://incidentdatabase.ai/cite/1546", "editor": "AIBD" }, { "date": "2026-10-20", "change": "Set the date observed to 21 October 2025 and reworded the summary and account because Discussions 1 and 2 show the Google DeepMind ensemble mean shaping the official track from the first advisory, added those two discussions as evidence, corrected the Discussion 14 title to “Tropical Storm Melissa” as its header reads, attributed the US$8.8 billion damage figure to the World Bank/IDB estimate relayed by IFRC, and tied the DREF's early actions to “forecast risks” as the review says rather than to NHC.", "source": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml; https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml; https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml; https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-dref-no-mdrjm005-operation-update-3; https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-imminent-dref-mdrjm005-dref-operational-review-report-april-may-2026", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Added what happened since from the cited sources.", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Benefit type set to Improved Task Accuracy under Trustible's full taxonomy of 23 types; it was Improved Knowledge Discovery among the six first used.", "source": "https://trustible.ai/post/a-framework-for-measuring-the-benefits-of-ai/", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Country of the developer and the deployer added from the organization's own site or a registry.", "source": "https://en.wikipedia.org/wiki/Google_DeepMind; https://www.nhc.noaa.gov/aboutintro.shtml", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Prose pass: a rating reason, account, the replication conditions, the system's part rewritten in shorter sentences with abbreviations expanded; facts, figures and quotations unchanged.", "editor": "AIBD" }, { "date": "2026-10-21", "change": "Notes about which sources could be read were reworded for readers; facts, figures and quotations unchanged.", "editor": "AIBD" }, { "date": "2026-10-22", "change": "Organization name standardized to DeepMind.", "editor": "AIBD" }, { "date": "2026-10-24", "change": "Moved two IFRC reports to background source notes because they describe storm impacts and humanitarian action, without establishing the AI forecast’s contribution to those outcomes. Preserved their citations, excerpts, and the existing account and depths. Retained NHC evidence of forecast use and verification, attributed to the deployer rather than an independent AI-benefit evaluator.", "source": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml; https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "editor": "AIBD" }, { "date": "2026-10-25", "change": "Added NHC’s revised Hurricane Melissa report, replaced the obsolete statement that it was unpublished, and separated forecast verification from humanitarian background.", "source": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "editor": "AIBD" } ] } }, "provenance": { "kind": "curatorial", "basis": "Moved two IFRC reports to background source notes because they describe storm impacts and humanitarian action, without establishing the AI forecast’s contribution to those outcomes. Preserved their citations, excerpts, and the existing account and depths. Retained NHC evidence of forecast use and verification, attributed to the deployer rather than an independent AI-benefit evaluator.", "record_fields": [ "revisions" ], "source_refs": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-24" }, "review_state": "record_mapped", "coverage": "partial" }, "sector": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "domain": "Weather forecasting / tropical cyclones", "segment": "Public-sector meteorological service" } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "domain", "segment" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "sources[].access_version": { "state": "value_present", "value": { "kind": "structured", "data": { "reviewed_on": "2026-10-25", "versions_and_access": [ { "url": "https://www.weather.gov/news/261102-AI-Hurricane-Forecasting", "access": "Full original page read on October 25, 2026. Q&A 7 discusses Melissa, and Q&A 2 describes NHC’s development partnership with DeepMind.", "metadata": { "DC.date.created": "June 24th 2026 4:15 PM" }, "date_treatment": "Do not infer a publication date from the URL or treat creation metadata as original publication. The original March 30, 2026 verification report already links to this URL on printed p. 7. The future-date suspicion is resolved and should not appear on the website." }, { "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "access": "Full original PDF downloaded successfully (HTTP 200) on October 25, 2026. Title page dated March 30, 2026. Printed p. 7 links to the NHC interview, and the report provides seasonal forecast comparisons.", "supersedes_access_note": "The earlier direct-PDF 403 limitation is now resolved for this document." }, { "source_ref": "AIBD-0088-source-0010", "source_url": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "publication_date": "2026-02-25", "version_reviewed_date": "2026-05-19", "status": "Full original PDF downloaded successfully (HTTP 200) and relevant text read on October 25, 2026. HTTP Last-Modified: May 19, 2026.", "version_note": "The May 19 title-page note says that this version revised fatality totals and damage estimates based on updated government information. Preserve February 25 as the original report date and May 19 as the reviewed revision." } ], "limitations": [ "The separately cited IFRC operational review and update retain their earlier access limits. The November 2025 IFRC operational strategy was accessible." ] } }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "The NHC interview, 2025 verification report, and revised Melissa report were accessible in full. Earlier access limits remain attached to the individual IFRC reports." } }, "sources[].citation_locator": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "evidence": [ { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 1, 1100 AM EDT Tue Oct 21 2025, forecasters Cangialosi/Bucci. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 1", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml", "date": "2025-10-21", "written_by": "deployer", "quote": "The NHC official track forecast lies between the Google DeepMind ensemble mean track and the correct consensus aid, HCCA." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 2, 500 PM EDT Tue Oct 21 2025, forecaster Cangialosi. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 2", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml", "date": "2025-10-21", "written_by": "deployer", "quote": "This prediction is a little left of the previous forecast and remains a blend of the latest Google DeepMind ensemble mean, the corrected consensus aid HCCA, and simple consensus TVCA." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 8, 500 AM EDT Thu Oct 23 2025, forecaster Cangialosi. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 8", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.008.shtml", "date": "2025-10-23", "written_by": "deployer", "quote": "The NHC track forecast is shifted a little to the right of the previous one, trending toward a blend of the latest Google DeepMind ensemble mean and HCCA solution." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 10, 500 PM EDT Thu Oct 23 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 10", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.010.shtml", "date": "2025-10-23", "written_by": "deployer", "quote": "the latest NHC track was only nudged slightly southward from this morning, blending the reliable HCCA and GDMI aids" }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 13, 1100 AM EDT Fri Oct 24 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 13", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.013.shtml", "date": "2025-10-24", "written_by": "deployer", "quote": "Every single Google DeepMind member shows the system become a Category 4 hurricane or higher" }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 14, 500 PM EDT Fri Oct 24 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 14", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml", "date": "2025-10-24", "written_by": "deployer", "quote": "The NHC intensity forecast is on the high side of the guidance envelope, and closest too the Google DeepMind ensemble mean (GDMI)." }, { "kind": "report", "citation": "National Hurricane Center (2025). Hurricane Melissa Tropical Cyclone Update, 100 PM EDT Tue Oct 28 2025. NOAA/NWS.", "title": "Hurricane Melissa Tropical Cyclone Update (landfall)", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.update.10281700.shtml", "date": "2025-10-28", "written_by": "deployer", "quote": "CATEGORY 5 MELISSA MAKES LANDFALL IN JAMAICA...ONE OF THE MOST POWERFUL HURRICANE LANDFALLS ON RECORD IN THE ATLANTIC BASIN" }, { "kind": "press release", "citation": "National Weather Service (n.d.). AI in Hurricane Forecasting at the National Hurricane Center: A Q&A with NHC Science Operation Officer Wallace Hogsett. NOAA/NWS News Around NOAA.", "title": "AI in Hurricane Forecasting at the National Hurricane Center", "url": "https://www.weather.gov/news/261102-AI-Hurricane-Forecasting", "written_by": "deployer", "quote": "The AI models honed in very early on the likely track and intensity and provided very valuable guidance to complement our traditional NWP guidance." }, { "kind": "report", "citation": "Cangialosi, J. P., Martinez, J., Reinhart, B., Papin, P. (2026). National Hurricane Center Forecast Verification Report: 2025 Hurricane Season. NOAA/NHC, 30 March 2026.", "title": "NHC Forecast Verification Report, 2025 Hurricane Season", "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "date": "2026-03-30", "written_by": "deployer", "quote": "the NHC track errors are notably lower than average for major hurricanes, including Humberto and Melissa in 2025" }, { "kind": "report", "citation": "Kelly, Larry A., José Álamo, and Aidan D. Mahoney. Hurricane Melissa (AL132025), 21–31 October 2025. National Hurricane Center Tropical Cyclone Report. Revised May 19, 2026, originally February 25, 2026. 16–17, Tables 5b and 6b.", "title": "Hurricane Melissa (AL132025): Tropical Cyclone Report", "url": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "date": "2026-02-25", "written_by": "deployer", "quote": "The NHC OFCL forecasts were comparable to GDMI through 96 h, but performed slightly better than GDMI at 120 h." } ] } }, "provenance": { "kind": "source_review", "basis": "Moved two IFRC reports to background source notes because they describe storm impacts and humanitarian action, without establishing the AI forecast’s contribution to those outcomes. Preserved their citations, excerpts, and the existing account and depths. Retained NHC evidence of forecast use and verification, attributed to the deployer rather than an independent AI-benefit evaluator.", "record_fields": [ "evidence" ], "source_refs": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-24" }, "review_state": "record_mapped", "coverage": "partial" }, "sources[].id": { "state": "value_present", "value": { "kind": "identifiers", "identifiers": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009", "AIBD-0088-source-0010" ], "role": "Retained sources" }, "provenance": { "kind": "curatorial", "basis": "Synchronized source identifiers.", "record_fields": [], "source_refs": [] }, "review_state": "record_mapped", "coverage": "complete" }, "sources[].language": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "sources[].relationship": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "evidence": [ { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 1, 1100 AM EDT Tue Oct 21 2025, forecasters Cangialosi/Bucci. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 1", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml", "date": "2025-10-21", "written_by": "deployer", "quote": "The NHC official track forecast lies between the Google DeepMind ensemble mean track and the correct consensus aid, HCCA." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 2, 500 PM EDT Tue Oct 21 2025, forecaster Cangialosi. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 2", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml", "date": "2025-10-21", "written_by": "deployer", "quote": "This prediction is a little left of the previous forecast and remains a blend of the latest Google DeepMind ensemble mean, the corrected consensus aid HCCA, and simple consensus TVCA." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 8, 500 AM EDT Thu Oct 23 2025, forecaster Cangialosi. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 8", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.008.shtml", "date": "2025-10-23", "written_by": "deployer", "quote": "The NHC track forecast is shifted a little to the right of the previous one, trending toward a blend of the latest Google DeepMind ensemble mean and HCCA solution." }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 10, 500 PM EDT Thu Oct 23 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 10", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.010.shtml", "date": "2025-10-23", "written_by": "deployer", "quote": "the latest NHC track was only nudged slightly southward from this morning, blending the reliable HCCA and GDMI aids" }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 13, 1100 AM EDT Fri Oct 24 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 13", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.013.shtml", "date": "2025-10-24", "written_by": "deployer", "quote": "Every single Google DeepMind member shows the system become a Category 4 hurricane or higher" }, { "kind": "report", "citation": "National Hurricane Center (2025). Tropical Storm Melissa Discussion Number 14, 500 PM EDT Fri Oct 24 2025, forecaster Papin. NOAA/NWS.", "title": "Tropical Storm Melissa Discussion Number 14", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml", "date": "2025-10-24", "written_by": "deployer", "quote": "The NHC intensity forecast is on the high side of the guidance envelope, and closest too the Google DeepMind ensemble mean (GDMI)." }, { "kind": "report", "citation": "National Hurricane Center (2025). Hurricane Melissa Tropical Cyclone Update, 100 PM EDT Tue Oct 28 2025. NOAA/NWS.", "title": "Hurricane Melissa Tropical Cyclone Update (landfall)", "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.update.10281700.shtml", "date": "2025-10-28", "written_by": "deployer", "quote": "CATEGORY 5 MELISSA MAKES LANDFALL IN JAMAICA...ONE OF THE MOST POWERFUL HURRICANE LANDFALLS ON RECORD IN THE ATLANTIC BASIN" }, { "kind": "press release", "citation": "National Weather Service (n.d.). AI in Hurricane Forecasting at the National Hurricane Center: A Q&A with NHC Science Operation Officer Wallace Hogsett. NOAA/NWS News Around NOAA.", "title": "AI in Hurricane Forecasting at the National Hurricane Center", "url": "https://www.weather.gov/news/261102-AI-Hurricane-Forecasting", "written_by": "deployer", "quote": "The AI models honed in very early on the likely track and intensity and provided very valuable guidance to complement our traditional NWP guidance." }, { "kind": "report", "citation": "Cangialosi, J. P., Martinez, J., Reinhart, B., Papin, P. (2026). National Hurricane Center Forecast Verification Report: 2025 Hurricane Season. NOAA/NHC, 30 March 2026.", "title": "NHC Forecast Verification Report, 2025 Hurricane Season", "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "date": "2026-03-30", "written_by": "deployer", "quote": "the NHC track errors are notably lower than average for major hurricanes, including Humberto and Melissa in 2025" }, { "kind": "report", "citation": "Kelly, Larry A., José Álamo, and Aidan D. Mahoney. Hurricane Melissa (AL132025), 21–31 October 2025. National Hurricane Center Tropical Cyclone Report. Revised May 19, 2026, originally February 25, 2026. 16–17, Tables 5b and 6b.", "title": "Hurricane Melissa (AL132025): Tropical Cyclone Report", "url": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "date": "2026-02-25", "written_by": "deployer", "quote": "The NHC OFCL forecasts were comparable to GDMI through 96 h, but performed slightly better than GDMI at 120 h." } ] } }, "provenance": { "kind": "source_review", "basis": "Moved two IFRC reports to background source notes because they describe storm impacts and humanitarian action, without establishing the AI forecast’s contribution to those outcomes. Preserved their citations, excerpts, and the existing account and depths. Retained NHC evidence of forecast use and verification, attributed to the deployer rather than an independent AI-benefit evaluator.", "record_fields": [ "evidence" ], "source_refs": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-24" }, "review_state": "record_mapped", "coverage": "partial" }, "sources[].upstream_sources": { "state": "value_present", "value": { "kind": "structured", "data": { "origins": [ { "source_refs": [ "AIBD-0088-source-0001" ], "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml", "title": "Tropical Storm Melissa Discussion Number 1", "relationship": "Original operational advisory establishing that DeepMind ensemble guidance contributed to the official forecast from the first advisory." }, { "source_refs": [ "AIBD-0088-source-0002" ], "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.002.shtml", "title": "Tropical Storm Melissa Discussion Number 2", "relationship": "Original operational advisory recording the blend of DeepMind and other guidance." }, { "source_refs": [ "AIBD-0088-source-0003" ], "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.008.shtml", "title": "Tropical Storm Melissa Discussion Number 8", "relationship": "Original operational advisory recording a later adjustment toward DeepMind and HCCA guidance." }, { "source_refs": [ "AIBD-0088-source-0004" ], "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.010.shtml", "title": "Tropical Storm Melissa Discussion Number 10", "relationship": "Original operational advisory recording track and intensity guidance. It does not isolate downstream public benefits." }, { "source_refs": [ "AIBD-0088-source-0005" ], "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.013.shtml", "title": "Tropical Storm Melissa Discussion Number 13", "relationship": "Original operational advisory recording the ensemble’s Category 4-or-higher signal." }, { "source_refs": [ "AIBD-0088-source-0006" ], "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.014.shtml", "title": "Tropical Storm Melissa Discussion Number 14", "relationship": "Original operational advisory recording the intensity forecast’s agreement with GDMI." }, { "source_refs": [ "AIBD-0088-source-0007" ], "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.update.10281700.shtml", "title": "Hurricane Melissa Tropical Cyclone Update (landfall)", "relationship": "Original landfall update establishing storm context, not the AI system’s effect on public outcomes." }, { "source_refs": [ "AIBD-0088-source-0008" ], "url": "https://www.weather.gov/news/261102-AI-Hurricane-Forecasting", "title": "AI in Hurricane Forecasting at the National Hurricane Center", "relationship": "Original interview with NHC’s Science Operations Officer. NHC deployed the model and partnered in its development; this is not an independent evaluation." }, { "source_refs": [ "AIBD-0088-source-0009" ], "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "title": "NHC Forecast Verification Report, 2025 Hurricane Season", "relationship": "Original seasonal verification by NHC and a cooperating researcher. It evaluates forecasting performance, but it is not an independent evaluation of AI-attributable humanitarian outcomes." }, { "url": "https://www.nhc.noaa.gov/data/tcr/AL132025_Melissa.pdf", "title": "Hurricane Melissa (AL132025): Tropical Cyclone Report", "relationship": "Original retrospective verification by the agency that used the guidance. The report compares track and intensity forecasts for Melissa and explains how GDMI and regional models contributed to forecasters’ confidence. It does not isolate the AI model’s effect on humanitarian outcomes.", "locator": "Printed pp. 16–17, Track Forecast and Intensity Forecast, Tables 5b and 6b.", "original_publication_date": "2026-02-25", "version_reviewed_date": "2026-05-19", "version_note": "The May 19 title-page note says that this version revised fatality totals and damage estimates based on updated government information. Preserve February 25 as the original report date and May 19 as the reviewed revision.", "access": "Full original PDF downloaded successfully (HTTP 200) and relevant text read on October 25, 2026. HTTP Last-Modified: May 19, 2026.", "source_refs": [ "AIBD-0088-source-0010" ] } ], "related_material": [ { "url": "https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-imminent-dref-mdrjm005-dref-operational-review-report-april-may-2026", "title": "Jamaica - Hurricane Melissa Imminent DREF (MDRJM005) Operational Review Report", "published_on": "2026-08-07", "relationship": "Humanitarian and storm context preserved from the existing case. It does not establish an AI-attributable outcome and is not counted as independent AI-benefit evaluation.", "preserved_evidence": { "kind": "report", "citation": "IFRC (2026). Jamaica - Hurricane Melissa Imminent DREF (MDRJM005), DREF Operational Review Report, April - May 2026. ReliefWeb, 7 August 2026.", "title": "Jamaica - Hurricane Melissa Imminent DREF (MDRJM005) Operational Review Report", "url": "https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-imminent-dref-mdrjm005-dref-operational-review-report-april-may-2026", "date": "2026-08-07", "written_by": "independent", "quote": "Rapid approval and pre-landfall mobilisation, enabled by prior training, pre-positioned stock, and strong logistics systems, allowed JRC to respond quickly." }, "access": "The original ReliefWeb URL returned HTTP 403 during this narrow review. Existing citations and quotations are preserved; no new full-text verification is claimed." }, { "url": "https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-dref-no-mdrjm005-operation-update-3", "title": "Jamaica | Hurricane Melissa DREF MDRJM005 Operation update #3", "published_on": "2026-07-07", "relationship": "Humanitarian and storm context preserved from the existing case. It does not establish an AI-attributable outcome and is not counted as independent AI-benefit evaluation.", "preserved_evidence": { "kind": "report", "citation": "IFRC (2026). Jamaica | Hurricane Melissa DREF No. MDRJM005 Operation update #3. ReliefWeb, 7 July 2026.", "title": "Jamaica | Hurricane Melissa DREF MDRJM005 Operation update #3", "url": "https://reliefweb.int/report/jamaica/jamaica-hurricane-melissa-dref-no-mdrjm005-operation-update-3", "date": "2026-07-07", "written_by": "independent", "quote": "winds of nearly 300 km (185 miles) per hour" }, "access": "The original ReliefWeb URL returned HTTP 403 during this narrow review. Existing citations and quotations are preserved; no new full-text verification is claimed." }, { "url": "https://go-api.ifrc.org/api/downloadfile/93656/MDRJM005OS", "title": "Jamaica | Hurricane Melissa: Emergency Appeal Operational Strategy (MDRJM005)", "date": "2025-11-11", "locator": "p. 2, timeline; p. 3, situation overview; p. 10, targeting", "finding": "The timeline records the CHF80,000 allocation on October 23, 2025. The targeting section attributes the estimated 1.6 million people affected to analysis of UNOSAT wind forecasts and demographic data.", "relationship": "Original IFRC background on anticipatory aid and exposure. This report does not attribute those actions or the affected-population estimate to Google DeepMind.", "access": "Full original PDF downloaded successfully (HTTP 200) and relevant sections read on October 25, 2026. The IFRC emergency page dates this operational strategy November 11, 2025.", "recommendation": "Keep as related background rather than independent evidence of AI-created benefit. Preserve existing IFRC related material and its separate access status." } ], "claim_lineage": [ { "claim": "The first official Melissa forecast used DeepMind ensemble guidance.", "source_ref": "AIBD-0088-source-0001", "locator": "Track discussion, paragraph beginning “The system was moving”", "access": "Original advisory text read." }, { "claim": "GDMI performed well in NHC’s 2025 verification.", "source_ref": "AIBD-0088-source-0009", "locator": "Abstract and Atlantic track-verification discussion", "access": "Full original PDF inspected on 25 October 2026.", "limitation": "Seasonal model accuracy is not a measure of AI-attributable lives saved or of the AI contribution to each humanitarian action." } ], "access_audit": [ { "url": "https://www.nhc.noaa.gov/archive/2025/al13/al132025.discus.001.shtml", "access": "Original advisory text read through the indexed official page after the direct request returned HTTP 403." }, { "url": "https://www.weather.gov/news/261102-AI-Hurricane-Forecasting", "access": "Full original page read on October 25, 2026. Q&A 7 discusses Melissa, and Q&A 2 describes NHC’s development partnership with DeepMind.", "access_history": [ { "access": "Original NHC interview text read; it identifies NHC’s development partnership with Google DeepMind.", "reviewed_on": "2026-10-24" } ], "reviewed_on": "2026-10-25" }, { "url": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf", "access": "Full original PDF downloaded successfully (HTTP 200) on October 25, 2026. Title page dated March 30, 2026. Printed p. 7 links to the NHC interview, and the report provides seasonal forecast comparisons.", "access_history": [ { "access": "Indexed original report abstract and relevant text read; direct PDF request returned HTTP 403.", "reviewed_on": "2026-10-24" } ], "reviewed_on": "2026-10-25" } ], "limitations": [ "NHC is the operator and a development partner, so its verification is primary but not an independent evaluation of its own use of AI.", "Track and intensity comparisons do not measure lives saved, damage avoided, or whether IFRC acted earlier because of GDMI.", "The seasonal comparison must not be substituted for a Melissa-only result." ], "ordering_basis": "Keep the contemporaneous NHC account of AI use first, followed by retained operational documentation and later verification. Humanitarian background remains available in the source notes.", "lead_source_ref": "AIBD-0088-source-0001", "retained_source_order": [ "AIBD-0088-source-0001", "AIBD-0088-source-0002", "AIBD-0088-source-0003", "AIBD-0088-source-0004", "AIBD-0088-source-0005", "AIBD-0088-source-0006", "AIBD-0088-source-0007", "AIBD-0088-source-0008", "AIBD-0088-source-0009", "AIBD-0088-source-0010" ], "scope": "Operational use and retrospective verification of forecasts, with humanitarian material retained separately as background.", "reviewed_on": "2026-10-25" } }, "provenance": { "kind": "source_review", "basis": "Read the original NHC Melissa report, revised 19 May 2026, pp. 16–17 and Tables 5b/6b, together with the original NHC interview and seasonal verification report.", "record_fields": [], "source_refs": [ "AIBD-0088-source-0010", "AIBD-0088-source-0008", "AIBD-0088-source-0009" ], "reviewed_on": "2026-10-25" }, "review_state": "reviewed", "coverage": "partial", "display": { "answer": "NHC’s operational advisories document its use of DeepMind guidance. Its later verification reports assess forecast accuracy. IFRC reports provide humanitarian context without isolating AI’s effects." } }, "strategic_assessments[]": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial", "family": "strategic assessments" }, "summary": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "summary": "NHC blended Google DeepMind's AI ensemble into official Melissa forecasts from its first advisory on 21 October 2025; Cat 5 landfall in Jamaica 28 October 2025.", "account": "The US National Hurricane Center (NHC) blended Google DeepMind's experimental AI hurricane model (ensemble mean GDMI) into official forecasts for Tropical Storm Melissa. It did so from its first advisory, 21 October 2025. Discussion 1 placed the official track between the DeepMind ensemble mean and the consensus aid HCCA. Discussion 2 described the track as a blend of the DeepMind mean, HCCA and TVCA. Discussion 8 (23 October) moved the track toward a blend of the DeepMind mean and HCCA. Discussion 10 blended the HCCA and GDMI aids. It set an intensity forecast in best agreement with GDMI after all 50 DeepMind members showed a major hurricane. On 28 October 2025 Melissa made landfall in southwestern Jamaica near New Hope with 185 mph winds and 892 mb pressure. IFRC reported 45 deaths and about 1.6 million people affected. It relayed a World Bank/IDB estimate of US$8.8 billion in physical damage. Its CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time." } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "summary", "account" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "systems[].autonomy_permissions": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "systems[].component_role": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "systems[].deployment_configuration": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "systems[].inputs_outputs": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "systems[].model_name": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "systems[].model_version": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "systems[].product_name": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "system": "Google DeepMind AI hurricane forecast model, ensemble mean designated GDMI in NHC guidance" } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "system" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "systems[].provider": { "state": "value_present", "value": { "kind": "structured", "data": { "application_developer": "DeepMind" } }, "provenance": { "kind": "legacy_record", "basis": "Mapped from the existing case description without a new external source review.", "record_fields": [ "developer" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "taxonomy_crosswalks[]": { "state": "not_assessed", "review_state": "not_assessed", "coverage": "partial" }, "title": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "title": "NHC blended Google DeepMind's AI ensemble into official Hurricane Melissa forecasts before its Category 5 landfall in Jamaica", "headline": "NHC blends DeepMind's ensemble into Hurricane Melissa forecasts" } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "title", "headline" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial" }, "version": { "state": "value_present", "value": { "kind": "identifiers", "identifiers": [ "7494545dd8aeba975dc7387c52bc23ab138bc6a416517d50ebb709bf803d7efa" ], "role": "SHA-256 of normalized original case data" }, "provenance": { "kind": "curatorial", "basis": "Recomputed the content fingerprint after the source-attribution correction.", "record_fields": [ "evidence", "revisions" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "complete" }, "workflow": { "state": "value_present", "value": { "kind": "record_fields", "fields": { "account": "The US National Hurricane Center (NHC) blended Google DeepMind's experimental AI hurricane model (ensemble mean GDMI) into official forecasts for Tropical Storm Melissa. It did so from its first advisory, 21 October 2025. Discussion 1 placed the official track between the DeepMind ensemble mean and the consensus aid HCCA. Discussion 2 described the track as a blend of the DeepMind mean, HCCA and TVCA. Discussion 8 (23 October) moved the track toward a blend of the DeepMind mean and HCCA. Discussion 10 blended the HCCA and GDMI aids. It set an intensity forecast in best agreement with GDMI after all 50 DeepMind members showed a major hurricane. On 28 October 2025 Melissa made landfall in southwestern Jamaica near New Hope with 185 mph winds and 892 mb pressure. IFRC reported 45 deaths and about 1.6 million people affected. It relayed a World Bank/IDB estimate of US$8.8 billion in physical damage. Its CHF 80,000 Imminent DREF was approved on 23 October with a five-day lead time.", "replication": { "produced_by": [ "Experimental Google DeepMind ensemble guidance available in real time to NHC forecasters", "Forecasters choosing to blend GDMI with consensus aids (HCCA) in official track and intensity forecasts", "A slow-moving, rapidly intensifying hurricane where models disagreed on track relative to Jamaica" ], "must_hold": [ "AI ensemble output delivered on time and verified before operational use, as NHC describes", "Human forecasters retain the final official forecast and integrate conventional NWP guidance", "Forecast accuracy can be assessed against the observed track and intensity. NHC’s Tropical Cyclone Report, revised 19 May 2026, provides that comparison for Melissa" ], "echoes": [ { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 Atlantic season track forecasts", "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" }, { "date": "2026-03-30", "setting": "NHC homogeneous verification, 2025 eastern North Pacific track forecasts (GDMI beat official and consensus aids at 48–120 h)", "outcome": "recurred", "source": "https://www.nhc.noaa.gov/verification/pdfs/Verification_2025.pdf" } ] } } }, "provenance": { "kind": "legacy_record", "basis": "Preserved from the existing AIBD record; external sources were not rechecked.", "record_fields": [ "account", "replication" ], "source_refs": [] }, "review_state": "record_mapped", "coverage": "partial", "display": { "answer": "Forecasters received the experimental DeepMind ensemble in real time and combined it with conventional forecasts. They retained responsibility for the official track and intensity forecasts." } } }, "summary": { "total": 91, "addressed": 91, "known": 43, "reviewed": 11, "record_mapped": 32, "not_assessed": 48, "complete": false, "by_state": { "value_present": 43, "not_reported": 0, "unresolved": 0, "not_assessed": 48, "not_applicable": 0, "withheld": 0, "source_unavailable": 0 } } } }