The AI model WeatherNext, developed by Google DeepMind and Google Research, predicted with 80% confidence five days before Hurricane Melissa made landfall...
The AI model WeatherNext, developed by Google DeepMind and Google Research, predicted with 80%
confidence five days before Hurricane Melissa made landfall in October 2025 that it would hit Jamaica with the strength of a Category 5 hurricane. According to a paper published in *Nature*, the model's cyclone predictions were unprecedentedly accurate, providing, on average, one more day of warning than existing models—that is, its three-day forecast accuracy is equivalent to the two-day level of current models. 🔗 Read the original via AIHOT · https://aihot.virxact.com/items/cmsk9yb97046zrow93i345be7
Google DeepMind 与 Google Research 开发的 AI 模型 WeatherNext，在 2025 年 10 月飓风 Melissa 登陆前 5 天，以 80%
的置信度预测其将以 5 级飓风强度袭击牙买加。
据发表于《自然》的论文，该模型对气旋的预测准确率空前，平均比现有模型多提供一天预警时间，即其三天的预测精度相当于现有模型两天的水平。
🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsk9yb97046zrow93i345be7
Google DeepMind 与 Google Research 开发的 WeatherNext 模型，被报道在 2025 年 10 月飓风 Melissa 登陆牙买加前五天，以 80% 置信度预测其将达到 5 级飓风强度。
材料称，相关研究已发表于《自然》。报道将 WeatherNext 的气旋预测表现与现有模型比较，称其平均可带来额外一天的预警时间，并以“三天预测精度相当于现有模型两天”为例说明。
Aioga 判断，这类评估若能在更多气旋案例和不同业务流程中保持一致，可能使 AI 预报模型从补充参考进一步成为预警决策的重要输入。值得关注的是，材料未说明具体评测范围与部署方式。
对预报机构而言，更长的有效预警窗口可能有助于提前安排风险研判与应急准备。但材料仅陈述模型预测准确率和预警时间的比较，未提供其实际运行场景、使用主体或灾害响应结果。 值得关注后续公开研究是否披露 WeatherNext 的测试数据集、比较基线、误差指标及跨地区表现，也应观察预报机构如何将模型输出与既有预报流程结合，以验证其业务价值。
The readable text on this page was extracted from the public source and organized with attribution, publication time and the original link. Copyright remains with the original author and publisher.
Ingestion channel: RSS · Source domain: arstechnica.com
Source: Ars Technica：AI（RSS）
Original link: Open original source
Aioga archive: Open intelligence page
Content record: summary-fallback · Updated: 2026-08-08T11:05:50.000Z
统一接入主流 AI 模型 API，为开发、测试与生产环境提供稳定调用入口。
立即访问 api.w173.com
Scan to open this articleAioga aggregates global AI updates and preserves source information for verification and citation.