重大自然灾害房屋倒塌程度高分辨率遥感识别方法:以舟曲特大泥石流灾害为例

Mapping on Building Collapse Degree Caused by Large Natural Disaster Using Remote Sensing:A Case Study on Zhouqu Large Flood Debris Flow Disaster

重大自然灾害引起的房屋倒塌程度的快速制图,对灾害应急救援和灾后损失评估意义重大。针对目前利用高空间分辨率遥感数据开展房屋倒塌程度调查中存在的主要问题,如房屋倒塌分类标准不统一、解译标准缺乏等,在考虑人员和经济损失状况、灾害救助、恢复重建难度等方面的基础上,建立了高空间分辨率遥感数据支持下的重大自然灾害房屋倒塌程度的分类体系,分为未倒塌、轻度倒塌、中度倒塌、重度倒塌、完全倒塌5级。系统描述了房屋倒塌程度遥感解译规则,建立了遥感解译标志,并提出了一种基于瓦砾信息开展房屋倒塌程度的遥感识别方法。最后,以舟曲特大山洪泥石流灾害为例,开展了房屋倒塌程度遥感制图研究。结果表明,利用此方法,房屋倒塌程度识别精度达到92.73%,完全能够满足自然灾害应急救援和灾后损失评估的需求。该方法体系为重大自然灾害应急救援和灾后损失评估提供了科学支撑。

It is significant for disaster emergency rescue and disaster loss evaluation to study on rapid mapping of building collapse caused by major natural disasters.First,the major issues on building collapse degree mapping by high spatial resolution remote sensing technique were analyzed,such as different classification standard and lack of interpretation standard.Second,based on several conditions,the classification scheme of building collapse by remote sensing was developed（i.e.,undamaged,slightly collapsed,moderately collapsed,severely collapsed,and fully collapsed）.Thirdly,the interpretation signs were developed.Finally,Zhouqu large flood debris flow disaster was selected,and the mapping of building collapse degree was generated.The results show that the accuracy of recognition of building collapse is up to 92.37%.This can meet with the requirement of emergency rescue.These results can present the scientific support for emergency rescue and disaster loss assessment on major natural disasters.