**谷歌研究团队开发AI模型，可提前7天预测未监测河流洪水**
近日，国际知名科学期刊《自然》（Nature）发表了一篇来自Google Research洪水预测团队的最新论文，揭示了一项人工智能技术在洪水预警领域的突破。Grey Nearing领导的团队成功开发出一款AI模型，该模型能提前7天预测那些目前未设置水位测量仪的河流可能出现的洪水情况。
据了解，这款AI模型在训练过程中，利用了全球5680个已有的水位测量仪数据，以学习和理解河流流量的变化模式。在预测未监测流域的日径流方面，该模型展现出了强大的预测能力。在对比测试中，研究人员将AI模型的预测结果与目前全球领先的短期和长期洪水预警系统——全球洪水预警系统（GloFAS）进行了对比，发现其在同日预测的准确率上至少与GloFAS相当，甚至在某些情况下更高。
这项名为“全球未监测流域的极端洪水预测”（Global prediction of extreme floods in ungauged watersheds）的研究，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备的地区，将极大地提高洪水灾害的预防和应对能力。未来，这一技术有望在全球范围内广泛应用于洪水风险管理，为保护生命财产安全提供更及时、更精准的预警支持。
英语如下：
**News Title:** “Google Research AI Model Breaks Record: Accurately Predicting Undetected River Floods Seven Days in Advance”
**Keywords:** Google AI, Flood Prediction, Nature Paper
**News Content:**
A research team from Google has developed an AI model capable of forecasting floods in unmonitored rivers up to seven days ahead, as detailed in a recent paper published in the renowned scientific journal _Nature_. The study, led by Grey Nearing, highlights a breakthrough in flood warning systems using artificial intelligence.
The AI model was trained using data from 5,680 existing water level gauges globally, allowing it to learn and understand the patterns of river flow. Demonstrating strong predictive capabilities, the model can anticipate daily discharge in ungauged basins. In comparative tests, the model’s predictions were compared with those of the Global Flood Awareness System (GloFAS), the world’s leading short- and long-term flood warning system, and found to be at least as accurate on the same day, and in some instances, even more so.
Titled “Global prediction of extreme floods in ungauged watersheds,” this research represents a significant advancement in flood warning technology, particularly for regions with limited monitoring infrastructure. It has the potential to greatly enhance flood preparedness and response capabilities. In the future, this technology is expected to be widely applied for flood risk management worldwide, providing more timely and precise early warnings to safeguard lives and property.
【来源】https://mp.weixin.qq.com/s/GoOPqLtdYvPv3_no7GJUJQ