谷歌DeepMind发布的AI天气预报模型GraphCast，在全球预测未来10天天气状况方面首次超越传统方法。该模型在1380项指标中，在90%的指标上优于ECMWF的系统，标志着天气预报领域出现了拐点。GraphCast AI在谷歌TPU v4云计算机上运行，1分钟内就可以生成10天内的天气预报。
根据《科学》杂志上的论文，GraphCast AI在预测全球未来10天的天气状况方面表现出色。此外，该模型在90%的指标上优于ECMWF的系统，这是天气预报领域的一个重要突破。GraphCast AI的准确度创历史新高，成为天气预报领域的拐点。
新闻翻译：
News Title: AI Weather Forecasting Model GraphCast Surpasses Traditional Methods, Achieves Historical High Accuracy
Keywords: AI, Weather Forecasting, Model, Accuracy, Turning Point
News Content:
Google DeepMind’s AI weather forecasting model, GraphCast, has achieved a historical high in accuracy by surpassing traditional methods in predicting weather conditions for the next 10 days. According to a paper published in the journal Science, GraphCast AI demonstrated outstanding performance in predicting global weather conditions. Additionally, the model outperformed ECMWF’s system by 90% in 1380 indicators, which marks a significant breakthrough in the field of weather forecasting. With an accuracy rate of 90% and above, GraphCast AI has set a new benchmark for weather forecasting.
【来源】https://new.qq.com/rain/a/20231115A00Q6C00/
Views: 1