基于高光谱及红外热图像的棉花水分胁迫诊断机理与模型研究

Water stress has become a major international agricultural disaster at present and it is especially important to make real-time, fast and nondestructive monitoring for crop water status. There are some problems of crop water status monitoring using hyper-spectral remote sensing and infrared temperature measurement technology, such as low monitoring precision, difficult quantification of irrigation and so on. This project will use the cotton as the research object and combine hyperspectrum with infrared thermal imaging technology to make a research on the hyperspectral response characteristic under different water status of cotton and then build the hyperspectral estimation model of plant and leaf water content with high precision and good general applicability based on the wavelet analysis method. In addition, it will have a discussion on the relationship between spectral reflectivity and crop water stress index (CWSI) of cotton and then make the CWSI estimation model based on the hyperspectrum. The soil water content estimation model will also be worked out, which is based on CWSI and soil water correction coefficient (Ks) and also based on the hyperspectrum. The research results will quantify the time and amount of irrigation demand by the hyper-spectral estimation. It is crucial to guide the irrigation scientifically and rationally and economize water resource. Furthermore, the results could provide the theory basis and technical support for the development of precision agriculture.

目前，水分胁迫已经成为世界性的重大农业灾害，实时、快速、无损的监测作物水分状况显得尤为重要。针对当前高光谱遥感与红外测温技术监测作物水分状况普遍存在精度不高，灌溉量难以定量表达等问题，本课题以棉花为研究对象，通过高光谱与红外热成像技术相结合的方法，研究不同水分状况棉花高光谱响应特征，建立精度高、普适性更好的基于小波分析的棉花植株含水量、叶片含水量高光谱估算模型；研究光谱反射率与棉花水分胁迫指数（CWSI）之间的关系，建立基于高光谱的棉花CWSI估算模型；研究CWSI与土壤含水量之间的关系，建立基于CWSI与土壤水分修正系数Ks的土壤含水量估算模型和基于高光谱的土壤含水量估算模型。研究成果将实现棉花灌溉时间及灌水量的高光谱定量估算，对于科学合理的指导棉花灌溉，节约水资源具有重要意义，同时为发展"精准农业"提供理论依据和技术支撑。

当前气候多变，旱涝不均造成的水分胁迫时刻威胁着我国棉花产量的安全。因此，利用遥感技术实时、快速、无损、大面积监测棉花水分状况，对指导农业生产具有重大的意义。而目前遥感监测作物水分胁迫普遍存在精度不高，灌溉量难以定量表达等问题。按照课题申请书的研究内容及研究计划，本项目利用田间试验和理论分析相结合的方法，开展了不同水分状况高光谱响应特征研究以及棉花水分胁迫高光谱估算模型研究。取得了以下主要成果：得到了不同生育期、不同植株水分含量下棉花冠层高光谱特征及其红边特征，获得了不同水分处理条件下棉花冠层高光谱特征及其红边特征的变化规律，建立了基于植被指数、高光谱特征参数、红边参数、小波能量系数及小波分形维数的棉花植株水分含量估算模型，构建了基于植被指数、平稳小波分析的CWSI估算模型，从而为高光谱遥感定量估算棉花水分胁迫状况提供了理论依据，实现棉花水分状况的高光谱遥感监测。. 目前本研究已发表学术论文10篇，其中SCI收录 2篇，EI收录 1篇。