IIT Guwahati的AI模型确定了东喜马拉雅山的492个新的冰川湖,以便更好地预测洪水风险。 IIT Guwahati's AI model identifies 492 new glacial lakes in the Eastern Himalayas to better predict flood risks.

IIT Guwahati开发了一个由AI驱动的框架,利用卫星数据和Bayesian神经网络,确定东喜马拉雅山492个潜在的冰川湖新地点,通过分析温坡、螺旋岩和退缩冰川等地形特征,改进冰川湖暴洪风险预测。 IIT Guwahati has developed an AI-driven framework using satellite data and Bayesian Neural Networks to identify 492 potential new glacial lake sites in the Eastern Himalayas, improving predictions of Glacial Lake Outburst Flood risks by analyzing terrain features like gentle slopes, cirques, and retreating glaciers. 该模型在《自然科学报告》中发表,通过纳入不确定性和复杂的地形,支持预警、基础设施规划和水管理,提高了准确性。 The model, published in Nature’s Scientific Reports, enhances accuracy by incorporating uncertainty and complex topography, supporting early warnings, infrastructure planning, and water management. 研究人员计划用光学数据和实地验证来完善模型,敦促与当局和社区合作,解决气候驱动的冰川退缩问题。 Researchers plan to refine it with moraine data and field validation, urging collaboration with authorities and communities to address climate-driven glacial retreat.