膜下滴灌条件下玉米冠层光截获特征和叶面积指数（LAI）遥感反演方法研究

It is still a vulanerable spot for agricultural water-conservation researches to quantitatively evaluate the effect of yield-increasing and water-saving at regional scale under mulched drip irrigation, which is also important for constituting the relevant national macro-strategies. The study will be carried out for discussing the relationship of “vertical distribution of light in maize canopy – the light interception characteristics – the canopy reflectivity – remote-sensing information extraction” under mulched drip irrigation. For the maize cultivation in the Northeast Plain, the effects of mulched drip irrigation on light interception and canopy reflectance will be studied, as well as the methods of estimating the applied area of mulched drip irrigation and LAI inversion by remote sensing. In order to achieve the above purposes, 2 years filed experiments will be conducted and mathematical models will be built for quantifying the effects of mulched drip irrigation on the maize growth, photosynthetically active radiation (PAR), light interception, canopy reflectance and vegetation spectrum. It will be explored that the relationship of maize canopy light interception rate, canopy structure, leaf angle and foliage area volume density under mulched drip irrigation. Then, the mathematical model for indicating the relationship between interception of PAR and cumulative increase of LAI will be built. Based on GF-1/GF-2 satellite, the maize cultivation area under mulched drip irrigation will be extracted by the hierarchical decision tree classification method. Furthermore, using PROSAIL model, the inversion model will be built for evaluating LAI of maize under mulched drip irrigation and the accuracy of the model will be verified by using the ground sample data in Heilongjiang/Liaoning Province. The expected results are valuable practically and of scientific significance for revealing the mechanism of water-saving and yield-increasing, development of remote sensing products and validation methods and guiding the production of maize under mulched drip irrigation.

从区域尺度科学定量评价膜下滴灌节水增产效应等方面的研究，依然是农业节水研究领域的薄弱环节，也是国家相关宏观战略政策制定的重要依据，本申请以膜下滴灌条件下“玉米冠层光垂直分布规律—冠层光截获特征—生育期内冠层反射率变化—遥感信息提取”的响应机制为研究主线，针对东北典型区，基于2年的田间试验和数值模拟，定量分析膜下滴灌对玉米生长、光截获量、冠层反射率及植被光谱等的影响机制，并探索光截获率与种植结构、叶倾角、叶面积体密度分布的响应关系，构建膜下滴灌条件下玉米冠层光截获率与向上累积LAI的数学模型；采用GF-1/GF-2遥感影像和分层决策树分类的方法对玉米膜下滴灌种植面积进行提取；结合PROSAIL模型，构建膜下滴灌条件下玉米LAI遥感反演模型，并进行区域应用和验证。预期研究成果对揭示膜下滴灌节水增产的内在机理，实现膜下滴灌遥感产品业务化生产，指导我国玉米膜下滴灌生产具有重要的实用价值和科学意义。

光截获能力和LAI是玉米生长的重要参数。本文针对东北典型区，基于田间试验和数值模拟，定量分析了膜下滴灌对玉米生长、光截获量、冠层光谱等参数的影响，结合PROSAIL模型，并探索玉米膜下滴灌种植面积提取和冠层LAI遥感反演方法。研究表明，膜下滴灌一定程度地降低了冠层和地面处的净辐射，并促进了株高和LAI的生长，冠层截获的净辐射与株高呈负相关关系，与LAI呈正相关关系，但膜下滴灌并未对冠层光利用率造成显著性影响。本文进一步构建了APAR和FPAR与LAI的线性回归模型，定量描述了膜下滴灌条件下玉米冠层光截获能力与LAI的相关关系。基于PROSAIL模型模拟和田间实测的光谱，结合GF-1和Sentinel-2卫星影像分析，当LAI较小时，膜下滴灌主要对可见光和短波红外波段产生影响，当LAI > 3时，膜下滴灌对玉米冠层反射率无明显影响。利用高分辨率遥感影像的可见光和红边波段，构建了均匀地表下膜下滴灌玉米种植面积提取模型。基于可见光、红边、短波红外波段，构建了辽宁省建平县膜下滴灌玉米冠层LAI的遥感反演模型，并结合当地实测数据进行了验证，验证精度较好。相关研究成果可为指导我国玉米膜下滴灌生产提供一定的科学参考。