基于极化SAR的小麦倒伏灾害与长势监测研究

发布时间：2018-03-30 17:37

本文选题：简缩极化SAR　切入点：作物倒伏　出处：《西安科技大学》2017年硕士论文

【摘要】：农业作为国民经济的基础,直接影响我国乃至全球经济的稳健发展。为顺应时代发展的需求,国家进一步加大了在农业方面的投资与扶持力度,精准农业监测能够有效保障我国粮食安全,使我国农业由传统方式向信息化、精准化与现代化发展。遥感技术的发展为大范围作物长势监测提供了一种手段,传统的光学遥感在监测过程中易受到观测时客观条件影响,导致数据实效性较差。合成孔径雷达(Synthetic Aperture Radar,SAR)遥感技术能够实现全天候对地观测,且具有一定的穿透性,不受天气条件制约,同时对作物结构特征和水分十分敏感,在农业监测中具有独特优势。目前,雷达遥感在农业领域中应用比较广泛,因此对简缩极化SAR(Compact Polarimetric SAR)的农业应用研究需要做进一步的探讨。本文基于简缩极化SAR与全极化SAR数据完成了小麦倒伏的定性监测研究,并基于水云模型完成了小麦生物量的定量反演研究。本文以小麦为研究对象,以内蒙古自治区额尔古纳市上库力农场为研究区域,完成以下工作:(1)同步获取研究区域2013年5景Radarsat-2雷达数据和地面实测倒伏数据。基于全极化SAR数据模拟简缩极化SAR数据并对其进行极化分解,利用简缩极化SAR极化参数结合地面实测倒伏数据构建小麦倒伏监测模型,并进行模型验证,从定性角度分析了简缩极化SAR极化参数的后向散射特征对倒伏的响应规律;(2)同步获取研究区域2016年2景Radarsat-2雷达数据与地面实测小麦长势参数,提取全极化SAR极化参数的后向散射系数值,结合地面实测数据构建水云模型,完成小麦生物量反演,从定量角度分析了雷达极化特征对生物量的响应情况以及反演小麦生物量的效果,并与经验模型反演结果进行对比。得出以下结论:(1)在小麦抽穗灌浆期,出现倒伏的地块线极化比uL显著降低,而线极化度pL和极化度Dop呈现一定程度降低,且阈值分别为1.8、0.3和0.4附近时区分倒伏效果最佳,并与单极化SAR验证效果做对比,实验结果表明简缩极化SAR监测倒伏效果明显,该方法为大面积监测作物倒伏提供了依据;(2)全极化SAR极化参数的后向散射系数对干生物量响应明显,对比水云模型和经验模型发现,8月11日HH极化参数反演的决定系数R2分别为0.6414和0.8426,实验结果表明水云模型基本可以达到经验模型的反演效果,该方法可以为大范围作物生物量反演提供一种技术支撑。
[Abstract]:Agriculture, as the foundation of the national economy, directly affects the steady development of our country and even the global economy. In order to meet the needs of the development of the times, the state has further increased its investment and support in agriculture. Precision agriculture monitoring can effectively guarantee the food security of our country, make the agriculture of our country develop from the traditional way to information, precision and modernization. The development of remote sensing technology provides a means for the large-scale crop growth monitoring. Traditional optical remote sensing is easily affected by the objective conditions of observation in the process of monitoring, which results in poor effectiveness of data. Synthetic Aperture radar synthetic Aperture radar (SAR) remote sensing technology can realize all-weather earth observation, and it has a certain penetrability. Not subject to weather conditions and sensitive to crop structural characteristics and moisture, it has a unique advantage in agricultural monitoring. At present, radar remote sensing is widely used in the agricultural field. Therefore, it is necessary to further study the agricultural application of contracted polarized SAR(Compact Polarimetric SAR.The qualitative monitoring of wheat lodging is completed based on the data of contracted polarized SAR and fully polarized SAR. Based on the water cloud model, the quantitative inversion of wheat biomass has been completed. In this paper, wheat is taken as the research object, and Shangkuili Farm in Erguna City, Inner Mongolia Autonomous region, is taken as the research area. Complete the following work: 1) simultaneously acquire 5 Radarsat-2 radar data and ground measured lodging data in the study area. Based on the fully polarized SAR data, we simulate and decompose the reduced polarized SAR data. A wheat lodging monitoring model was constructed by using reduced polarization SAR polarization parameters and ground measured lodging data, and the model was verified. From the qualitative point of view, the response of backscattering characteristics of SAR polarization parameters to lodging is analyzed. (2) the Radarsat-2 radar data obtained from the study area in 2016 and the wheat growth parameters measured on the ground are obtained simultaneously. The backscattering coefficients of the polarimetric parameters of the fully polarized SAR were extracted, and the water cloud model was constructed by combining the measured data on the ground to complete the retrieval of wheat biomass. The response of radar polarization characteristics to biomass and the effect of inversion of wheat biomass were analyzed quantitatively, and the results were compared with the results of empirical model inversion. The linear polarization of the landmass with lying down is significantly lower than that of ULL, while the linear polarization and polarization degree Dop are decreased to some extent, and the effect of distinguishing lodging is the best when the threshold values are 1.80.3,0.4, respectively, and the results are compared with that of unipolar SAR. The experimental results show that the effect of simple polarized SAR on monitoring lodging is obvious, and this method can provide a basis for monitoring crop lodging in large area. The backscattering coefficient of polarization parameters of fully polarized SAR has obvious response to dry biomass. Comparing with the water cloud model and the empirical model, it is found that the determination coefficient R2 of the HH polarization parameter inversion on August 11 is 0.6414 and 0.8426, respectively. The experimental results show that the water cloud model can basically achieve the inversion effect of the empirical model. This method can provide a technical support for large scale crop biomass inversion.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S512.1;S127

【参考文献】