11:00 〜 13:00
[MIS15-P02] 極楽平における高山植生の最近10年間の紅葉フェノロジーと発色の変化
キーワード：定点カメラ、画像解析、RGB、カラーインデックス、合成画像
Autumn leaf coloring provides valuable local tourist business, and the information of its phenology is important for tourists and local economy. Autumn leaf phenology is highly sensitive to climate, and a useful indicator for climate change. In recent years, however, changes in autumn phenology and the color brightness have reported around the world. Although the impacts of climate change are stronger in alpine zone, the information about autumn phenology is limited in the area due to difficulties in field observations. Therefore, we have monitored alpine ecosystems using automated digital time-lapse cameras for more than ten years. In this study, we detected the changes in autumn leaf coloring and phenology of alpine vegetation by means of image analysis based on red, green, and blue (RGB) pixel values of the repeat photography.
The study site locates at Gokuraku-daira at c.a. 2650m a.s.l. in central Japanese Alps in Nagano prefecture. The dominant alpine plants are Pinus pumila (ever green), Betula ermanii (yellow leaf), Sorbus matsumurana (red leaf). The images were taken every hour with a high-resolution digital camera (Nikon D7100). The RGB pixel values were derived from jpeg format image files during every mid-September and mid-October for 10 years from 2012 to 2021. And the autumn peak color and peak day were calculated using four color indexes, such as red chromatic coordinate (Rcc), green-red vegetation index (GRVI), red-green ratio (RGR), and excess red (ExR). The yearly images for autumn peak color were composited from the RGB pixel values at the time of maximum index value for each pixel.
As results, the most sensitive index for autumn color for overall both yellow and red leaves was ExR among the four indexes. Although GRVI and RGR were sensitive to red leaves, they were not effective for yellow leaves and showed the maximum values after leaf fall. Whichever index showed higher brightness in 2012, 2013, and 2014, and the lowest brightness in 2016. Most of the deciduous leaves fell before the color changed to red or yellow in the year. The degradation of leaf color brightness was found after 2016. The autumn peak was earlier in 2015, 2018, and 2021, and later in 2012, 2013 and 2019. Although later trend of autumn phenology has been reported in some place, later trend was not recognized at this site. The relationships between the autumn phenology and its color brightness and meteorological factors, such as temperature, precipitation, solar radiation, wind speed, and snowmelt time were discussed.
Further research and monitoring using repeat photography at many sites would help better understanding of the impacts of climate change on alpine vegetation.
The study site locates at Gokuraku-daira at c.a. 2650m a.s.l. in central Japanese Alps in Nagano prefecture. The dominant alpine plants are Pinus pumila (ever green), Betula ermanii (yellow leaf), Sorbus matsumurana (red leaf). The images were taken every hour with a high-resolution digital camera (Nikon D7100). The RGB pixel values were derived from jpeg format image files during every mid-September and mid-October for 10 years from 2012 to 2021. And the autumn peak color and peak day were calculated using four color indexes, such as red chromatic coordinate (Rcc), green-red vegetation index (GRVI), red-green ratio (RGR), and excess red (ExR). The yearly images for autumn peak color were composited from the RGB pixel values at the time of maximum index value for each pixel.
As results, the most sensitive index for autumn color for overall both yellow and red leaves was ExR among the four indexes. Although GRVI and RGR were sensitive to red leaves, they were not effective for yellow leaves and showed the maximum values after leaf fall. Whichever index showed higher brightness in 2012, 2013, and 2014, and the lowest brightness in 2016. Most of the deciduous leaves fell before the color changed to red or yellow in the year. The degradation of leaf color brightness was found after 2016. The autumn peak was earlier in 2015, 2018, and 2021, and later in 2012, 2013 and 2019. Although later trend of autumn phenology has been reported in some place, later trend was not recognized at this site. The relationships between the autumn phenology and its color brightness and meteorological factors, such as temperature, precipitation, solar radiation, wind speed, and snowmelt time were discussed.
Further research and monitoring using repeat photography at many sites would help better understanding of the impacts of climate change on alpine vegetation.