Development of a masking device to exclude contaminated reflection during tower-based measurements of spectral reflectance from a vegetation canopy
- Authors
- Ide R., Hirose Y., Oguma H., Saigusa N.
- Journal
- Agricultural and Forest Meteorology, 223, 141-150
- DOI
- 10.1016/j.agrformet.2016.03.010
- Abstract
-
Plant canopy spectral reflectance is measured at many observation towers to estimate the canopy’s biophysical and physiological features. When a wide-angle downward-facing spectrometer is used for tower-based measurements of reflected irradiance, it also captures reflections from the tower and near-horizon light from the sky within the field of view, leading to incorrect canopy reflectance estimates. However, the effects of this contamination have not been quantified. Therefore, we developed a retractable masking device to allow more accurate measurement of canopy reflectance by excluding the contamination. We continuously observed canopy reflectance of a Japanese larch (Larix kaempferi) forest from a flux tower by using the device alternately in masked and unmasked modes. In this paper, we describe the design of the masking device and demonstrate how tower contamination affected the measured canopy reflectance. We found that the canopy reflectance in the visible spectral region was overestimated by a maximum of more than 3 times the actual value due to the tower contamination, but that the near-infrared reflectance was affected much less. The contamination effects were substantial and complex, and the magnitude varied depending on the wavelength, solar zenith and azimuth angles, weather conditions, and canopy phenological status. Consequently, vegetation indices calculated from the spectral data were greatly affected. For example, the contaminated Normalized Difference Vegetation Index was 8–22% lower than that obtained with the masking device. We conclude that to accurately measure canopy reflectance via wide-angle spectroradiometers or photosynthetically active radiation sensors, a masking device similar to the one described in this paper is strongly recommended.