POTENTIAL OF PHENOCAM COLOR INDICES IN PHENOLOGICAL ANALYSIS OF EVERGREEN FORESTS

Abstract

Vegetation phenology plays a key role in regulating ecosystem processes. Changes in phenology serve as sensitive indicators of climate change impacts on ecosystems. Bud and leaf development is an important indicator of plant phenology. Identifying the changes in bud and leaf development will help us to understand how ecosystem responds to environmental changes. This study uses a near surface remote sensing technique, utilizing phenocams to capture high temporal resolution images for studying the phenological events in evergreen forests. The extracted phenocam color indices such as GCC, RCC, VCI, and ExG help in tracking the growing season of bud and leaf phenology. The research considers phenocam data of black spruce stands in Simoncouche Research Station in Laurentides Wildlife Reserve of Quebec, Canada from 2017 to 2020. By analyzing the phenological parameters from the time series color indices and comparing them with ground observations, the study aims to unravel the potential of various phenocam color indices. It involves understanding the distinct capabilities of color indices in tracking phenology parameters. Our results show that GCC is the most effective index for SOS with a mean difference of 13.9 days and both RCC and GCC for tracking the EOS with 10.5 and 11.1 days respectively. ExG also showed a good correlation with field observations, while VCI performed lower in comparison. The integration of a white reflectance panel in PhenoCam setup proved crucial for normalizing images under varying illumination conditions, enhancing the accuracy of phenological assessments. Further GCC estimates improved to 0.9 day for SOS and 4.2 days for EOS with the inclusion of a reflectance panel. Additionally, field observations estimates demonstrated close alignment with EVI estimates than NDVI, emphasizing the potential of combining ground-based and remote sensing technologies for precise phenological monitoring. The research aims to contribute to the broader understanding of how specific phenocam indices and calibration of data influence the reliability of phenological studies in evergreen forest ecosystems.