Canopy Structure Diurnal Mediation of the Photochemical Reflectance Index in Three U.S. Coastal Plain Forest Stands Observed using Airborne Hyperspectral and Lidar Imaging

Using airborne hyperspectral and lidar data, we evaluated the role of canopy structure in the changing diurnal light environment of three forests stands and associated changes in a key vegetation functional indicator, the Photochemical Reflectance Index (PRI). PRI has been used in prior studies as a surrogate for Light Use Efficiency (LUE) which is the ratio of Gross Primary Productivity (GPP) and absorbed photosynthetically active radiation. Canopy structure controls the interception of direct and diffuse sunlight in different parts of a canopy as illumination changes through the course of the day. To assess PRI as a function of illumination direction and forest canopy structure, we used G-LiHT data acquired nine times from mid-morning to late afternoon over two days, collected in mid-October at Parker Tract in eastern North Carolina during the 2013 FLEX-US Airborne Campaign. 420 m x 420 m plots were evaluated in forests with differing structures and compositions: (1) a 20-year-old thinned loblolly pine stand in North-South oriented rows with deciduous understory, (2) an older loblolly pine stand with East-West rows, and (3) a senescing, mixed deciduous broadleaf forest. Sunlit, mixed illumination and diffusely illuminated shadowed parts of the canopy were classified based on the reflectance distributions of 2 m pixel panchromatic images produced by averaging G-LiHT visible bands. Per-pixel PRI was computed for the three illumination classes and the pixel height was determined using a lidar-derived Canopy Surface Model. Stand-averaged PRI for the sunlit canopy is substantially lower than the shadowed components at all times of day, suggesting LUE is reduced by excess illumination. Mid-day PRI is lower than the morning and afternoon values, especially for the shadowed pixels. PRI for the illumination classes show no variation as a function of height, except mid-day for the younger pine sunlit understory which has lower PRI values than the pine canopy. These late growing seasons results may not be representative of structural control of forest function during months with higher productivity. Therefore, further study of hyperspectral measures of forest function and changing diurnal illumination, mediated by structure, is warranted that extends throughout the growing season.