目的 运用雷达遥感影像数据提取水体信息的方法,快速有效地识别因洪水导致钉螺扩散的血吸虫病传播潜在风险区域。方法 采集安徽省当涂县2020年5月15日汛前和7月16日洪水高峰期间两景哨兵一号（Sentinel-1A）卫星合成孔径雷达（SAR）影像数据,采用双极阈值分割方法,构建洪水淹没区域提取模型。收集安徽省2016年钉螺调查数据,建立当涂县钉螺分布空间数据库,随机选取75%和25%钉螺分布数据模型训练数据集和验证数据集,采用归一化植被指数阈值判别方法,构建钉螺孳生点生态环境特征提取模型,识别洪灾后血吸虫病传播潜在风险区域。结果 当涂县汛前水体面积约为228.07 km²,洪水高峰期水体面积约为259.26 km²,洪灾导致的水体面积扩大约13.67%。判别洪灾后钉螺潜在扩散地主要分布在当涂县原有钉螺环境相毗邻的江心乡、石臼湖西岸湖阳、大陇、塘南和乌溪乡等地,以及河道相连的内陆田地等区域。对比验证集钉螺实际分布与模型分析结果,模型准确率为82%。结论 基于Sentinel-1A雷达影像提取水体信息并用于识别洪灾后血吸虫病传播潜在风险区域是可行的。洪灾后钉螺在水淹区域存在扩散孳生风险,需在重点区域开展有针对性的监测和防控工作。

Objective To use the image data of radar remote sensing for extracting water body information to perform rapid and effective identification of schistosomiasis transmission risk areas due to spreading of snails resulted from flooding.Methods Binary map-sets of Sentinel-1A satellite synthetic aperture radar （SAR） image data for Dangtu County of Anhui Province on May 15, 2020 before the flood and on July 16 at flooding peak period, were collected. A bipolar threshold segmentation method was used to estimate the area covered by flood. The data of snail survey in Anhui Province in 2016 were collected to generate a snail spacial distribution database of Dangtu County. The areas with 75% and 25% snail distribution were randomly selected as the training and verifying map-sets. The threshold discrimination method of normalized vegetation index was applied to extract eco-environmental characteristics of snail habitats and to identify schistosomiasis transmission risk areas after a flood.Results Of Dangtu County, the water body surface area before flood was about 228.07 km², and 259.26 km² in the flood peak period, showing an area expansion of about 13.6%. Based on the identified potential spreading areas of snail habitants after flooding were mainly distributed in where adjacent to the previously recorded snail-infested areas including Jiangxin township, Huyang in the west bank of Shijiu Lake, Dalong, Tangnan and Wuxi townships, as well as in the inland fields connecting rivers. The contrast of the snail actual distribution on the verification map-set with the model analysis indicated an accuracy rate of 82%.Conclusion It is feasible to identify schistosomiasis transmission risk areas after flooding using Sentinel-1A satellite SAR image data for extraction of water body information. There is a risk of spread of snail breeding in flooded areas, thus, it is imperative to strengthen surveillance and control that relevant to the target areas.