CHARACTERISTICS OF VEGETATION INDICES FROM VISIBLE IMAGES FOR GROUND-BASED PHENOLOGY OBSERVATION

摘要

Ground-based phenology observations using digital cameras are being carried out to understand the effects of abnormal weather and climate variability on vegetation. When using commercial digital cameras in phenology observations, camera settings such as white balance and the difference of camera models and illumination conditions influence the RGB digital numbers of the images. Vegetation indices used in phenology observations that are calculated from digital numbers of images, must also be put in issue. In this study, the effectiveness of 4 vegetation indices calculated from color-corrected visual images for detecting the change of leaves and flowers was investigated by comparing under different weather conditions and within the change of season. The influence of bi-directional reflectance could be reduced by using images that were taken when the direction of the sun was around 90° to the direction of the field of view of camera. Also, it is possible to detect the timing of the phenological events by using the vegetation indices, but the influence of weather remains as a problem.