Affiliation:
1. Colorado School of Mines Golden CO USA
2. Institute of Geo‐Hydroinformatics Hamburg University of Technology Hamburg Germany
Abstract
AbstractLand subsidence, referring to the vertical sinking of land surface, is a significant geohazard posing serious risks to security of infrastructure, natural resources, built environment, and businesses in numerous places worldwide. Using deep learning approaches combined with more than 46,000 subsidence data, we predicted global land subsidence based on 23 environmental parameters. The generated global map of land subsidence covers historically documented and new subsiding areas. We estimate that more than 6.3 million square kilometers of the global land is influenced by significant subsidence rates. That includes 231,000 square kilometers of urban and dense settlement areas and a population of nearly 2 billion. The model revealed a positive correlation between the intensity of groundwater abstraction and the subsidence rate. Our results offer new insights regarding potential hotspots of land subsidence and provide the information required to devise necessary action plans and develop effective policies to mitigate this growing challenge worldwide.
Publisher
American Geophysical Union (AGU)
Cited by
4 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献