谷歌研究团队的最新论文在国际知名科学期刊《自然》（Nature）上发表，揭示了一项革命性的技术突破：人工智能模型能提前7天预测未被监测流域的洪水情况。该模型由Grey Nearing领导的洪水预测团队开发，他们利用全球5680个测量仪的数据进行训练，以预测那些没有直接测量设备的河流在预测期内的日径流。
这一创新模型在预测准确率上表现出色，与全球领先的洪水预警系统——全球洪水预警系统（GloFAS）进行了对比测试。结果显示，该AI模型在同日预测精度上与GloFAS相当，甚至在某些情况下更高。这标志着洪水预警技术的重大进步，尤其对于那些缺乏监测设施的地区，这一模型的预测能力将极大地提升洪水防范和应对的效率。
论文标题为“全球未监测流域的极端洪水预测”（Global prediction of extreme floods in ungauged watersheds），其发表在《自然》杂志上，预示着人工智能在环境科学和灾害管理领域的应用将更加广泛，为全球的洪水风险管理提供了新的工具和策略。这一进展对于提高全球洪水灾害的预警和应对能力具有里程碑式的意义。
英语如下：
**News Title:** “Google Research AI Breakthrough: Accurately Predicting Undetected River Floods 7 Days in Advance”
**Keywords:** Google AI, Flood Prediction, Nature Paper
**News Content:**
**Title:** Google Research Team Develops AI Model to Forecast Undocumented River Floods 7 Days Early
A groundbreaking study by Google’s research team has been published in the renowned scientific journal *Nature*, detailing an innovative AI model capable of predicting floods in unmonitored river basins up to 7 days ahead of time. The model, developed by the flood prediction team led by Grey Nearing, was trained using data from 5,680 global gauges to anticipate daily discharge in rivers without direct measurement devices.
The model’s predictive accuracy outshines, demonstrating impressive performance when compared to the world’s leading flood warning system, the Global Flood Awareness System (GloFAS). In same-day predictions, the AI model matches GloFAS’s accuracy and even surpasses it in certain instances. This represents a significant advancement in flood warning technology, particularly for regions with limited monitoring infrastructure, where the model’s predictive capabilities will greatly enhance flood preparedness and response efficiency.
Titled “Global prediction of extreme floods in ungauged watersheds,” the paper, published in *Nature*, signals a broader application of artificial intelligence in environmental science and disaster management. It introduces new tools and strategies for global flood risk management, marking a milestone in enhancing flood disaster warning and response capabilities worldwide.
【来源】https://mp.weixin.qq.com/s/GoOPqLtdYvPv3_no7GJUJQ
Views: 1