摘要
由于田块破碎、灌区信息化水平不高、土壤墒情反演困难等原因,在我国开展较高精度灌溉面积遥感监测依然面临很多困难。基于GF-1较高空间分辨率卫星数据,通过光谱匹配方法像元尺度应用,并引入OTSU自适应阈值算法,构建了高分辨率灌溉面积遥感监测新方法。选择我国西北干旱半干旱区典型渠灌灌区即东雷二期抽黄灌区为研究区,对其2018年的主要粮食作物种植强度及其灌溉面积开展了遥感识别提取研究。结果表明,东雷二期抽黄灌区灌溉面积为81 571.58 hm^2,其中双季轮作(小麦与玉米轮作)灌溉面积为40 335.88 hm^2,单季小麦灌溉面积为15 276.94 hm^2,单季玉米灌溉面积为14 059.14 hm^2;各灌溉子系统灌溉面积由大到小排序依次是流曲、孙镇、兴镇、荆姚、刘集、蒲城和大荔;通过野外采样精度验证,结果总体精度为88.27%(Kappa系数为0.8308),与国际水管理研究所灌溉数据产品相比,能更有效识别小田块灌溉分布及建设用地信息,在作物种植强度及其灌溉面积分布方面更符合我国实际情况,可为干旱监测预警、灌溉面积监测、灌溉用水效益评估等提供技术保障。
Due to the fragmentation of the field,the low level of informatization in the irrigation area,and the difficulty in inversion of soil moisture,the remote sensing monitoring of high precision irrigation area in China still faces many difficulties.Based on the GF-1 satellite data with higher spatial resolution,a new remote sensing monitoring method for irrigation area with high-resolution was constructed through the application of spectral matching method in pixel scale and the introduction of OTSU adaptive threshold algorithm.A typical canal irrigation area,Donglei PhaseⅡIrrigation District pumping water from the Yellow River,in the arid and semi-arid in Northwest China,was selected as the research area,and the planting intensity and irrigation area of wheat and corn in 2018 were studied by remote sensing identification and extraction.The Donglei PhaseⅡIrrigation District is 81571.58 hm^2,of which the irrigation area of double-season rotation(wheat and corn rotation)is 40335.88 hm^2,the irrigated area of single-season wheat is15276.94 hm^2,and that of single-season maize is 14059.14 hm^2.The areas of each irrigation subsystem from large to small is listed as Liuqu,Sunzhen,Xingzhen,Jingyao,Liuji,Pucheng and Daxie.The accuracy is verified by field sampling,which show that the overall accuracy is 88.27%(Kappa=0.8308).Compared with the irrigation data products of the International Water Management Institute,it can more effectively identify the irrigation distribution and construction land information of small fields,and is more in line with the actual situation in China in terms of crop planting intensity and irrigation area distribution,which can provide technical support for drought monitoring and early warning,irrigated area monitoring,irrigation water benefit evaluation and so on.
作者
宋文龙
李萌
路京选
卢奕竹
史杨军
贺海川
SONG Wenlong;LI Meng;LU Jingxuan;LU Yizhu;SHI Yangjun;HE Haichuan(China Institute of Water Resources and Hydropower,Beijing 100038,China;Center of Flood and Drought Disaster Reduction of Ministry of Water Resources,Beijing 100038,China;Capital Normal University,Beijing 100048,China;Weinan Donglei Phase II Yellow River Engineering Administration,Weinan 714000,China)
出处
《水利学报》
EI
CSCD
北大核心
2019年第7期854-863,共10页
Journal of Hydraulic Engineering
基金
国家自然科学青年基金(51609259)
国家重点研发计划(2018YFC1508702,2016YFC0400106-2)
中国水利水电科学研究院专项(JZ0145B472016,JZ0145B862017)
水利部技术示范项目(SF-201703)
关键词
灌溉面积
种植强度
GF-1卫星数据
光谱匹配
OTSU
东雷二期抽黄灌区
irrigation area
cropping intensity
GF-1 satellite data
spectral matching technique
OTSU
Donglei PhaseⅡIrrigation District