Characterizing spatiotemporal variations of forest canopy gaps using aerial laser scanning data

Abstract

Spatiotemporal variations of forest canopy gaps are essential indicators for both forest dynamic monitoring by capturing their size and location variations and forest ecosystem management for tree seedling regeneration through altering the radiation regime within and under forest canopies. To differentiate the effects of forest canopy gaps on forest three-dimensional (3-D) structure and direct incoming solar radiation, we proposed two novel concepts named physical forest canopy gap (PCG) and physiological forest canopy gap (POCG). Using aerial laser scanning (ALS) data, this study developed ALS-based methods of extracting and characterizing spatiotemporal variations of the PCG and POCG in a natural homogeneous forested area. The ALS-based PCG and POCG results were validated using visual-based interpretation and field-measured data, respectively. The overall accuracy was all over 0.80 and Kappa coefficients all over 0.70 for PCG extraction. The coefficient of determination between the ALS-based POCG area and field-based one was 0.86 (n = 16, p < 0.01), and the root mean square error was 16.28 m2. Spatiotemporal variations of forest canopy gaps included variations of their sizes and locations and the sunlit areas within them. This study provided a solid foundation for forest dynamic monitoring and forest ecosystem management.