Tracking forest biophysical properties with automated digital repeat photography: A fisheye perspective using digital hemispherical photography from below the canopy

Abstract

In recent years, digital repeat photography has received increasing attention as an inexpensive means by which seasonal changes in vegetation canopies can be monitored. Offering automation and an increased measurement frequency, colour indices derived from above-canopy digital repeat photography have proven a popular alternative to traditional observations of forest phenology. Nevertheless, previous work has demonstrated several features in time-series of colour indices that are unrelated to canopy structure, limiting their utility to track specific biophysical properties such as leaf area index (LAI). Whilst techniques such as digital cover photography and the use of radiometric sensors are better suited to this task, they are restricted by the need for careful calibration of above- and below-canopy reference sensors, ancillary information on canopy leaf angle distribution, and smaller measurement footprints. Using data collected at a deciduous broadleaf forest site in Southern England, we investigated a new method to derive continuous measurements of LAI, making use of automated digital hemispherical photography (DHP) from below the canopy. After applying simple data screening procedures, the LAI observations derived from our automated DHP system demonstrated very close agreement with those obtained from manually acquired DHP images, which were collected under optimal illumination conditions over the surrounding forest plot (r2 = 0.99, RMSE = 0.20, NRMSE = 13%). By combining our automated DHP system with an above-canopy time-lapse digital camera, we then investigated the relationship between the green chromatic coordinate (GCC) and LAI. Distinct hysteresis effects were observed, as were substantial differences between phenological transition dates derived from the GCC and LAI, particularly in the case of the onset of senescence. Our results indicate that phenological transition dates derived from colour indices cannot easily be linked to any one biophysical property. We recommend further investigation of the automated DHP approach, which provides time-series of LAI whose physical interpretation is straightforward, as an alternative to above-canopy digital repeat photography.