{"resourceId":"colorado-state-university-ai-hail-forecasting-stormscope-20261026","versions":[{"version":"external-2e25ae3c3ae5fad97d0af82934cabc4c9b4abb89134d561506e6eaa463c43947","resource":{"id":"colorado-state-university-ai-hail-forecasting-stormscope-20261026","title":"Colorado State University (Oct 26): researchers test NVIDIA-based Hail-StormScope nowcasts alongside CSU-MLP outlooks, warning the approaches are not yet tested together","organization":"Colorado State University","sector":"Higher education and public weather research","geography":"Colorado, United States","publishedAt":"2026-10-26 (CSU news release)","publicationDate":"2026-10-26","eventDate":"2026-10-26","sourceName":"Colorado State University News and Media Relations","sourceLabel":"University news release","sourceUrl":"https://newsmediarelations.colostate.edu/2026/10/26/csu-uses-artificial-intelligence-to-improve-hail-forecasting/","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["infrastructure","operating-model"],"finding":"CSU says its researchers are developing AI hail forecasting on two horizons: Hail-StormScope, a specialised version of NVIDIA Earth-2's StormScope architecture trained and deployed in a project a CSU researcher led with NVIDIA, for 2 to 3 hours ahead, and the CSU-MLP machine-learning system for 1 to 8 days ahead. The work sits within NSF's ASCEND programme (up to $160 million over 10 years), notes the team's models are accurate for areas of about 1.3 km and plans Vaisala RM60 hail sensors from spring 2027 for verification. Researchers have not decided whether to merge or keep the two systems separate, and a researcher cautions the approaches have not been tested extensively or together.","sledRelevance":"Relevant to state emergency management, transportation and agriculture agencies and to public universities doing applied weather research; it shows a public university building on an NVIDIA model with explicit verification plans.","evidence":"University news release quoting researchers; describes plans and early development, not verified forecast skill.","architectureImplications":"Combines a vendor-originated nowcasting model with a university-developed medium-range system; verification depends on new ground sensors arriving in 2027.","governanceImplications":"Operational use by forecasters or emergency managers would need skill verification against observed hail before adoption.","securityPrivacyImplications":"No specific security issues raised; public weather data.","caveats":"Release from the institution doing the work. No skill scores for hail are reported. The release says the hail model came from a project led with NVIDIA but gives no agreement terms or NVIDIA funding details.","streamIds":["nvidia"],"roles":{"sales":"Interpretation: CSU's work is a credible example of a public university building on NVIDIA's StormScope model for a public-interest problem, but it is early research. With state emergency management or transportation agencies, the useful framing is to follow CSU's verification plan rather than offer a product. Any conversation should acknowledge that researchers themselves say the approaches have not been tested extensively or together, and that hail verification sensors arrive only in spring 2027.","engineering":"Interpretation: the system pairs a fine-tuned StormScope nowcaster with the CSU-MLP medium-range model with models CSU describes as accurate to about 1.3 km. The StormScope paper does not evaluate hail and works at 6 km, so the hail adaptation and higher resolution are new work. Engineers advising agencies should track how CSU verifies hail skill with the RM60 sensors and whether the two systems are combined, before treating outputs as operational guidance.","delivery":"Interpretation: there is nothing to deliver to agencies yet. A sensible near-term step is a watching brief: track CSU's published verification, sensor deployment from spring 2027 and the merge decision. If an agency later pilots the forecasts, proposed acceptance criteria should compare hail nowcasts with observed hail reports and existing forecaster guidance over a full season, and should set out who acts on the forecasts."},"roleDetails":{"sales":{"customerProblem":"State agencies need earlier, more precise hail warnings to protect people, crops and infrastructure.","stakeholders":["State emergency management","State transportation department","Agriculture agencies","University atmospheric science departments"],"discoveryQuestions":["How do you use hail forecasts today, and at what lead time?","Would you participate in verification of experimental nowcasts?"],"valueHypothesis":"Interpretation: AI nowcasts could extend useful hail lead time; skill is not yet verified.","potentialEngagement":"Watching brief and possible verification partnership, not a product sale.","unsupportedClaims":["Any measured hail forecast skill","Operational readiness"]},"engineering":{"fit":"Research and verification settings with radar and satellite data access.","architectureIntegration":"StormScope-based nowcaster for 2 to 3 hours, CSU-MLP for 1 to 8 days, verified with ground hail sensors.","prerequisites":["GPU compute for training and inference","Radar and satellite data feeds","Hail sensor network"],"constraints":["Not tested extensively or together","Sensors arrive spring 2027","Base model not evaluated on hail"],"security":"Public data; standard research-computing controls apply.","proofOfValue":["Season-long comparison of nowcasts with observed hail","Comparison with existing forecaster guidance"]},"delivery":{"work":"Track verification results and sensor deployment; scope a pilot only after skill is published.","dependencies":["CSU publications","NSF ASCEND timeline","Sensor deployment"],"ownership":"Agency forecasting or emergency management lead would own any pilot.","skillsAdoption":"Forecasters need guidance on interpreting AI nowcasts.","governanceCheckpoints":["Review published verification before pilot","Decide on merged or separate systems"],"acceptanceCriteria":["Proposed: hail nowcasts compared with observed reports over a full season","Proposed: comparison with existing guidance documented"],"risks":["Adopting before verification","Resolution and hail adaptation untested"]}},"retrievedAt":"2026-10-27T02:56:52Z","enrichedAt":"2026-10-27T03:07:35Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Forecasters will need training to interpret AI hail nowcasts alongside existing guidance.","procurementImplications":"Agencies interested in hail nowcasts should wait for published verification results rather than procure on the basis of this announcement.","operatingModelImplications":"Open question whether short-range and medium-range systems will be merged or run separately.","sourceVerification":{"openedUrl":"https://newsmediarelations.colostate.edu/2026/10/26/csu-uses-artificial-intelligence-to-improve-hail-forecasting/","referenceExcerpt":"These are new technologies and AI approaches that have not been tested extensively or together yet","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}