植物冠层内光照分布决定了植物物理过程与生态环境之间的交互关系, 是植物冠层叶绿素荧光动力学研究的重要基础。 光谱技术在构建植株冠层组分含量的预测模型中发挥了重要的作用, 而针对自由纺锤形苹果树冠层不同光照区域的叶绿素荧光性状研究文献报道较少。 因此, 以自由纺锤形苹果树为研究对象, 将苹果树冠层空间划分为5层, 每层划分为边长50 cm的立方体网格, 并测定各个网格空间的光照强度, 确定其光照分布情况。 以此为基础, 获取不同光照区域的光谱数据及对应叶绿素荧光性状参数值, 通过光谱一阶微分去除系统误差干扰, 确定红边区域（660～760 nm）的一阶微分光谱曲线, 利用BP神经网络建立红光范围（680～760 nm）内一阶微分光谱最大值与叶绿素荧光性状参数关系模型, 创新性地提出了苹果树冠层不同光照区域的叶绿素荧光性状计算方法。 并采用决定系数、 均方根误差、 平均绝对百分比误差等对模型进行有效性评价, 试验结果表明: 该叶绿素荧光计算方法的计算精度均在80%以上, 研究成果可为苹果树的整形修剪, 获取充足光照分布提供理论依据。

As the basis of plant canopy chlorophyll fluorescence kinetics, light distribution within the canopy determines the interaction relationship between plant physical processes and ecological environment. Spectroscopy technology plays a very important role in building a prediction model of component content to plant canopies. However, there is only limited number of reports about chlorophyll fluorescence properties of different light intensity areas to free spindle apple canopies. In this paper, with the free spindle apple tree as the research object, the canopy space of apple tree was divided into five layers, and six cube grids with 50cm length of side in each layer, and then the light distribution was determined through measuring the light intensity of each cube grids space. firstly, spectrum data and characters of chlorophyll fluorescence were obtained in the different light area; secondly, a differential spectrum curve in red area(680~760 nm) was determined through removing the interference of system error by a differential spectrum; thirdly, relationship model has been established innovatively between the maximum value in red area(680~760 nm) and the chlorophyll fluorescence characters, which has been used as calculation method of chlorophyll fluorescence characters in different light area to apple tree canopy. Fourthly, root mean square error, mean absolute percentage error, mean forecast error were adopted to evaluate the method. The test result shows that the accuracy of the method is all above 80%, which can be the theoretical basis for pruning and getting best light distribution to apple tree canopy.