Abstract
Monitoring and quantification of sunlight in the forest is important for silviculture and management as well as an input for modeling, among others, forest regeneration, species composition, soil condition or micro-climate. For consistency, otherwise specified, the term sunlight, referring to solar radiation below canopy within the 400-700 nm wavelength, will be used throughout this paper. This study is presenting application of airborne light detection and ranging (LiDAR) technology in modeling sunlight conditions in the managed mixed stands of Bialowieia Forest (BF). During the summer of 2015, mean daily sunlight in the form of photosynthetically active radiation (PAR) from 100 random circular plots of BF were measured using hemispherical photography (HP). This technique uses photos to capture canopy structures and indirectly estimates understory solar radiation. A total of 5 acquisitions per plot (center and cardinal points) were taken before averaging. Various LiDAR related metrics were generated from point clouds of airborne laser scanning (ALS) acquired on the same area. These metrics were used as predictors in regression models with the direct, diffuse and global radiation. Ratio-based metrics were highly correlated across all components of sunlight. Laser penetration index (LPI) alone was able to predict effectively the amount of sunlight (R-2 = 0.78 for diffuse). A multiple regression improved the model (R-2 = 0.83) with up to 4 predictors. Other contributing metrics included in the model were: mean height of all returns, the median height of returns below diameter at breast height (DBH) and the percentage of the first returns. Results from the research confirmed that LiDAR can be a suitable tool for modeling solar radiation at various levels and producing continuous information across large forested areas with complicated structure and species composition.
NSTL
Elsevier