【作者】 韩秀珍；

【导师】 马建文；

【作者基本信息】 中国科学院研究生院（遥感应用研究所） ， 地图学与地理信息系统， 2003， 博士

【摘要】 我国是农业大国，蝗灾同水灾、旱灾并列为农业三大灾害。蝗灾不发则已，一发则赤地千里，不可收拾。近年来，全球气候变暖、连续干旱，我国14个省200多个县，同时爆发蝗虫灾害，面积超过200多万公顷，对我国农业生产造成严重威胁。因此，2002年2月国家计委发布《实施应用高技术控制我国蝗灾产业化专项的公告》。中科院遥感所，农业部全国农业技术推广中心，河北省计委联合开始在环渤海湾地区研究东亚飞蝗灾害，同时中科院遥感所开始了环渤海湾东亚飞蝗灾害的遥感监测机理和方法研究，本论文就是在该项研究的基础上完成的，主要成果体现在以下方面： (1)提出了对基于东亚飞蝗灾害建立全生育周期综合遥感监测新模式 当前国际上运行的遥感虫害(包括蝗虫灾害)监测系统一般是灾情后发挥作用。本研究提出了针对东亚飞蝗灾害建立全生育周期综合遥感监测新模式，将灾后监测扩展为孵化期、发育期、成虫期三个阶段监测，即：应用遥感数据对蝗虫栖息地生境特征及蝗害发生不同阶段进行多尺度多时相连续监测和分析，揭示生境特征变化与蝗虫种群发生、发育之间的关系机理，利用多元数据融合和数量分析方法建立蝗虫种群密度-生境指标-地面光谱-遥感影像之间的定量模型。从遥感层面为研究蝗害发生趋势和防治途径提供科学方法和依据。 (2)东亚飞蝗生育环境要素的遥感探测机理和方法研究 通过野外实地探测蝗虫孵化期地表水热条件，建立遥感反演水热参数与实测参数之间的关联模型，建立野外实测东亚飞蝗生长发育期环境指标、地面光谱、遥感影像提取参数之间的相关关系；对比研究植被要素对飞蝗生境的响应；开展了野外同步实验，验证遥感反演生物物理参数的可靠性，估算了不同尺度1000m，250m和30m)遥感数据反演参数的精度；在蝗区大尺度范围内利用了低空间分辨率、高时间分辨率的MODIS数据，对东亚飞蝗孵化的水热条件和植被要素进行了实时动态监测；在黄骅、天津大港蝗区利用高分辨率Landsat-5／TM、Landsat-7／ETM数据进行东亚飞蝗的栖境条件监测。 (3)应用3S手段将多源数据融合进行蝗虫灾害监测的综合分析 在对蝗虫生物性分析数据、地面实测环境要素的统计数据、遥感反演的水热条件、植被要素、多年气候数据、历史蝗灾记录、蝗害发生时有关数据进行集成和分析的基础上，提供孵化出土爆发地、蝗虫密集生长区域、迁移方向、灾情评价以及来年孽生地预测:实现蝗虫发生量与灾害预测，并生成蝗灾发生程度和分布区域的空间分布图。(4)本研究将成为“35+C+W”(遥感、地理信息系统、全球定位系统、通讯、网络)集成技术建成的东亚飞蝗监测系统的核心内容。更多还原

【Abstract】 The major economy of our country is agriculture. The locust hazard is three main agriculture hazards after floods and droughts. Once it breaks out it is covered on kilometers of lands in red color and very difficult to control. In recent years, because of the global warming and continuous dry climate East Asian Migratory Locust hazards broke out in 160 counties of 12 provinces in China involving 2,000,000 hectares agriculture lands. In February of 2002 National Planning Council announced a project of application of the high tech-knowledge to control the East Asian Migratory Locust hazards. Institute of remote sensing application (IRSA) of China Academy Sciences, Agriculture tech-knowledge promote center of China Agriculture and Planning Council of Hebei province organized together to cooperate to start studying the methods of using remote sensing to monitor East Asian Migratory Locust hazards. IRSA went ahead to carry out remote sensing new model for monitoring the locusts. This dissertation presents the results of the study. There are some merits of the dissertation as following,(1) The new model of building remote sensing monitoring system for the East Asian migrating locust based on the nature circle of locust breeding stagesCurrently, the remote sensing monitor system of insect pest hazards brings into play after hazards breaking out. In the dissertation a new model to monitor the locusts in active way, in which the locusts breeding process is divided into three stages there are hatching, growing and adult in correspond with there are some differences of monitoring parameters and vegetation conditions (biology-physicsfeatures) within the stages. Remote sensing can retrieve surface temperature and moisture at hatching stage; biomass lose at locust growing stage; locust grouping and moving at adult stage.(2) The biology and physical mechanism test and retrieval methods of remote sensing data for East Asian migratory locustEstablish the correlation models among locust breakout and breeding environment indicators, field spectral features and remote sensing retrievals; comparison of the response of retrieved parameters and vegetation factors; synchronism test was carried out to test the reliability of remote sensing retrieved soil moisture and temperature, vegetation conditions and assessment of the accuracy of retrieved parameters with different scale such as 1000m. 250m and 30m resolution remote sensing data; The MODIS data was used in large scale which contribute high temporal in formation. In Dagang district of Tianjin city and Huanghua county TM/ETM data were used for monitor locust breeding conditions.(3) The integrated analysis application of Remote Sensing, GIS and GPS (3S) The locust biology and physical condition data, field testing data ofenvironment factors, soil moisture and temperature data, vegetation data retrieved from remote sensing, climate data as well as history record of locust hazards were syntactically analyzed as a results to provide locust hatching ground, locust assemble areas, migrating directions as well as prediction of the breed areas for next year. In this way the locust hazards time comes through and spatial distribution can be mapped.(4) The research results will be the kernel tech-knowledge and major procedures for "RS+C+W" Migrate Locust of East Asian monitoring system.更多还原