引用本文:崔京路,毛克彪※,陈日清,曹萌萌,袁紫晋,唐世浩.基于高分辨率遥感影像的农作物灾损评估研究[J].中国农业信息,2018,30(6):63-70
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 24次   下载 54 本文二维码信息
码上扫一扫!
分享到: 微信 更多
基于高分辨率遥感影像的农作物灾损评估研究
崔京路1, 毛克彪※2,3, 陈日清1, 曹萌萌2, 袁紫晋2, 唐世浩4
1.福建农林大学计算机与信息学院,福州350002;2.中国农业科学院农业资源与农业区划研究所,北京100081;3.湖南农业大学资源环境学院,长沙410128;4.国家气象卫星中心,北京100081
摘要:
【目的】2016 年湖北省部分地区遭受洪涝灾害,导致该省农作物产量下降。利用遥 感手段进行灾损评估研究可快速、准确获取灾后农作物的受灾程度与减产量,给农民提供 实时、准确的参考数据,以减少农民的经济损失。【方法】基于2015 年7 月与2016 年7 月 的灾前灾后Landsat-8 影像,利用面向对象分类方法和分层抽样方法对湖北省荆州市与荆门 市部分地区的农作物进行受灾面积提取,统计农作物受灾面积总量;对灾前灾后农作物的 NDVI 值进行对比,根据灾前灾后的NDVI 差值进行受灾等级划分,以此得到农作物的受灾 等级。【结果】(1)由连续暴雨引起的洪涝灾害导致研究区农作物都有不同程度的受损,采 用了面向对象分类方法精度指标的Kappa 系数为0.849 5,以此分类统计出的受灾农作物总 面积为635.838 km2;(2)受灾严重的区域主要是依傍河岸和地势低洼处。荆州市的弥市镇 虎渡河附近的农作物受灾等级最高,受灾最为严重,几乎完全被洪水淹没;而荆州市马良镇 的杨家湾因为地势低以及对农作物的灾害预防措施做得不够,也造成大量的农作物被淹没, 甚至出现了绝收现象。【结论】对于农作物灾后监测,有关部门要借助遥感手段及时向农民 提供数据支持,将农民的经济损失降到最低,最大程度地保障农民的生活。
关键词:  NDVI 差值;灾损评估;受灾面积;受灾等级
DOI:10.12105/j.issn.1672-0423.20180606
分类号:
基金项目:国家重点研发计划课题“高时空分辨率多源卫星遥感气象灾害产品融合技术”(2018YFC1506502),国家 自然科学基金项目“基于遥感研究气候变化背景下农业旱灾时空变化对粮食生产影响(41571427)
Crop damage assessment based on high resolutionremote sensing imagery
Cui Jinglu1, Mao Kebiao※2,3, Chen Riqing1, Cao Mengmeng2, Yuan Zijin2, Tang Shihao4
1.School of Computer and Information,Fujian Agriculture and Forestry University,Fuzhou 350002,China;2.Institute of Agricultural Resources and Regional Planning,Chinese Academy of Agricultural Sciences,Beijing 100081,China;3.College of Resources & Environment,Hunan Agricultural University,Changsha 410128,China;4.National Satellite Meteorological Center,Beijing 100081,China
Abstract:
[Purpose]In 2016,parts of Hubei Province suffered from floods,which led to a decline in crop yields throughout the province. Using remote sensing method to conduct damage assessment research can quickly and accurately obtain the degree of disaster of post-disaster crops and yield reduction in order to provide real-time and accurate reference data to farmers and to reduce farmers’ economic losses.[Method]Based on the Landsat8 images of the pre-disaster disasters in July 2015 and July 2016,and the crops in parts of Jingzhou and Jingmen in Hubei Province,this paper uses object-oriented classification method to extract the affected area as the simulated true value,compares the NDVI value of the crops before and after the disaster,and divides the disaster level according to the difference of NDVI after the disaster so as to obtain the disaster level of the crop.[ Result]The results show that:(1) The floods caused by continuous heavy rain have caused different degrees of damage to the crops in the study area. The Kappa coefficient of the accuracy index after the object classification is 0.8495. The total area of crops is 635.838 square kilometers; (2) The areas affected by the disaster are mainly crops that depend on the banks of the river and lowlying areas. The crops near the Hudu River in Mishi Town,Jingzhou City have the highest level of disasters,and they are the most severely affected,almost completely flooded.The Yangjiawan in Maliang Town,Jingzhou City is not enough because of the low terrain and disaster prevention measures for crops,which also causes a large number of crops to be submerged,and even a phenomenon of rejection.[Conclusion]For post-disaster monitoring of crops,relevant departments should provide timely data support to farmers by means of remote sensing,so as to minimize their economic losses and guarantee their lives to the greatest extent.
Key words:  NDVI difference;damage assessment;disaster area;disaster level