ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume V-3-2021
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-3-2021, 211–217, 2021
https://doi.org/10.5194/isprs-annals-V-3-2021-211-2021
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-3-2021, 211–217, 2021
https://doi.org/10.5194/isprs-annals-V-3-2021-211-2021

  17 Jun 2021

17 Jun 2021

ESTIMATION OF OPTIMAL CROWN COVERAGE AND CANOPY SHAPE FOR SHADOW ESTIMATION ON TROPICAL MOIST BROADLEAF FOREST

T. Fujiwara and W. Takeuchi T. Fujiwara and W. Takeuchi
  • Institute of Industrial Science, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo, Japan

Keywords: Structural parameter, Sentinel-2, Virtual forest, Reflectance simulation, Light use efficiency

Abstract. Shadow fraction is essential for improving the estimation of gross primary production, but it is difficult to be observed by satellite due to the diurnal variations. Therefore, it is necessary to estimate the 3D model with physical parameters by simulating virtual forest reflectance. In this study, we aim to estimate the optimal combination of canopy shape and Crown Coverage (CC) through simulating virtual forests reflectance. First, satellite-derived Tree Height (TH) and CC for virtual forests were compared with the ones obtained by Canopy Hight Model (CHM). Second, virtual forests with different CC and canopy shapes were created, and the reflectance and shadow fraction were simulated. The canopy shape used were cylinder, ellipsoid, half-ellipsoid, and inverted half-ellipsoid. Finally, the simulated reflectance and shadow fraction were validated with Sentinel-2 reflectance and shadow fraction from voxel model. Our results show that the mean TH is 15 ± 2 m, and the CC was increased from 10% to 60% in 10% intervals. TH and CC obtained from the satellite had the Root Mean Square Error (RMSE) of 5m and 40%. Ellipsoid with 20% CC shows the lowest RMSE and the smallest discrepancy for shadow fractions at the same sun position. However, other combinations were more accurate in estimating mean daily shadow fraction. This would be caused by only one image adopted in validation, which could be improved by using multi-season images in the future.