Main content area

Estimating canopy gap fraction and diffuse light interception in 3D maize canopy using hierarchical hemispheres

Wen, Weiliang, Guo, Xinyu, Li, Baojun, Wang, Chuanyu, Wang, Yongjian, Yu, Zetao, Wu, Sheng, Fan, Jiangchuan, Gu, Shenghao, Lu, Xianju
Agricultural and forest meteorology 2019 v.276-277 pp. 107594
algorithms, butterflies, canopy gaps, corn, cultivars, geometry, latitude, longitude, models, photographs, photosynthesis, trees
Studying the detailed organization of plant canopy structure allows a better understanding of functional processes in functional-structural plant models (FSPM). Canopy gap fraction (CGF) is an important indicator describing the canopy structure and affects the way plants capture light to perform photosynthesis, especially the interception of diffuse light. Though efforts have been made to improve the accuracy and efficiency of CGF measurement and estimation, current technologies are usually position limited. Thus, this work developed new virtual methods for computing CGF and diffuse light interception in the 3D space of plant canopies. Five hierarchical hemispheres, containing 15, 40, 360, 1,440, and 5760 triangle patches, respectively, were constructed by applying Sqrt-3 and Butterfly subdivision schemes on an original icosahedron. Compared with traditional hemisphere division strategies using solid angles or crossed arcs of latitude and longitude, the proposed hierarchical hemispheres provide more resolution choices for different accuracy demands. Most of the patches on the hemispheres are regular triangles with similar sizes, which improves the CGF and diffuse light calculation accuracy using the Turtle model. Acceleration mechanism was built when calculating the detection of plant facets in the canopy using the tree relationship between adjacent hemispheres. Two geometric models and six maize canopy geometric models with cultivar and density differences were constructed using measured data to validate the approach. The maximum CGF error was 1.39% for the two geometric models. The average error of the four derived CGFs from hemisphere photographs was 4.23%. The diffuse light distribution correlation coefficients R2 were 0.96, 0.98, and 0.90 for three different canopy densities. Our algorithm provides multi-scale hemisphere choices for CGF and diffuse light interception simulation. The study also paves the way for further investigation into plant canopy structure analysis and simulation of light dynamics in plant canopies.