[H-TT17] Environmental Remote Sensing

[HTT17-P11] Estimation of Solar Radiation in Forests using Drone LiDAR Data in Mt. Karasawa Plantation Forest, Tochigi Prefecture, Japan

*Shiori Takamura¹, Yuichi Onda², YUPAN ZHANG ³, ASAHI HASHIMOTO³, Hiroaki Kato², Takashi Gomi⁴, CHEN WEI CHIU⁵ (1.Univercity of Tsukuba, 2.Center for Research on Isotopes and Environmental Dynamics, University of Tsukuba, 3.Graduate School of Science and Technology, University of Tsukuba, 4.Univercity of Nagoya, 5.Tokyo University of Agriculture and Technology)

Keywords:solar raditation, forest regeneration, LiDAR, remote sensing

The percentage of forests to Japan's national land area is very high, at more than 70%. Plantation forests account for about 40% of this total. However, the forestry industry in Japan has not stopped declining since the peak of Japan's rapid economic growth, and plantation forests are not adequately managed. The impact of such degraded plantation forests on the forest ecosystem is immeasurable, as the canopy closes in, the understory vegetation does not grow on the forest floor, and evapotranspiration from the canopy increases. In recent years, efforts have been made to improve the light environment in Japanese forests through intensive thinning, but evaluation methods have not yet been established.
In this study, we analyzed the estimation of solar radiation in forests in order to clarify the actual changes in the light environment in forests due to thinning. Traditionally, the amount of solar radiation in the forest and the degree of canopy openness have been evaluated by image analysis of hemispherical photogrphy taken with a fisheye lens. However, hemispherical photogrphy have limitations in estimation accuracy because they cannot correctly represent the actual canopy structure due to low resolution and lens distortion.
Here, we used drone LiDAR data to estimate the amount of solar radiation in the forest, which is a three-dimensional high-density point cloud data that can represent the forest structure in detail. The data was collected efficiently by using a drone.
A hemispherical photographic photogrammetric model was created to estimate the amount of solar radiation in the forest. The LiDAR data, which is originally expressed in Cartesian coordinates, was transformed into polar coordinates, and the angular perspective was converted to a downward angle similar to that of a hemispherical photogrphy. The light regime function of the Hemisfer image analysis software was used to estimate the amount of solar radiation in the forest based on the image data from the hemispherical photogrphy and the trajectory of the sun.
Furthermore, noting the large influence of direct sunlight on solar radiation, the canopy aperture was calculated based on the angle of direct sunlight, i.e., the position of the sun's altitude and solar azimuth, to correct the estimated solar radiation in the forest. This operation allows us to express the light transmittance in the forest as canopy aperture and to clarify the relationship with the actual solar radiation in the forest.
The study site was a cypress plantation forest located in Mt.Karasawa, Sano City, Tochigi Prefecture. The forest was intensively thinned with a thinning rate of 50% in 2011. The slope is south-facing with a slope of approximately 30 degrees, and 25 solar radiation meters were set up in the forest in a grid pattern with 1 m intervals to measure the actual amount of solar radiation in the forest.
As a result, while the estimated solar radiation values in the forest based on hemispherical photogrphy alone did not capture the detailed solar radiation changes in the measured values, the use of drone LIDAR data and the consideration of the degree of canopy opening successfully reproduced solar radiation changes that were closer to the measured values. Therefore, the combined use of LiDAR data has the potential for more accurate estimation and quantification of the light environment in forests, which is expected to promote forest management.