Minimizing measurement uncertainties of coniferous needle-leaf optical properties. Part II: Experimental setup and error analysis

Citation data:

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, ISSN: 1939-1404, Vol: 7, Issue: 2, Page: 406-420

Publication Year:
2014
Usage 150
Downloads 144
Abstract Views 6
Captures 18
Readers 18
Citations 6
Citation Indexes 6
Repository URL:
https://ro.uow.edu.au/smhpapers/2498
DOI:
10.1109/jstars.2013.2292817
Author(s):
Yanez-Rausell, Lucia; Malenovsky, Zbynek; Clevers, Jan G. P. W; Schaepman, Michael E
Publisher(s):
Institute of Electrical and Electronics Engineers (IEEE)
Tags:
Earth and Planetary Sciences; Conifers; gap fraction (GF); integrating sphere; leaf; needles; optical properties (OPs); reflectance; transmittance; Medicine and Health Sciences; Social and Behavioral Sciences
article description
We present uncertainties associated with the measurement of coniferous needle-leaf optical properties (OPs) with an integrating sphere using an optimized gap-fraction (GF) correction method, where GF refers to the air gaps appearing between the needles of a measured sample. We used an optically stable artificial material simulating needle leaves to investigate the potential effects of: 1) the sample holder carrying the needles during measurements and 2) multiple scattering in between the measured needles. Our optimization of integrating sphere port configurations using the sample holder showed an underestimation of the needle transmittance signal of at least 2% in flat needles and 4% in nonflat needles. If the needles have a nonflat cross section, multiple scattering of the photons during the GF measurement led to a GF overestimation. In addition, the multiple scattering of photons during the optical measurements caused less accurate performance of the GF-correction algorithms, which are based on the assumption of linear relationship between the nonGF-corrected signal and increasing GF, resulting in transmittance overestimation of nonflat needle samples. Overall, the final deviation achieved after optimizing the method is about 1% in reflectance and 6% in transmittance if the needles are flat, and if they are nonflat, the error increases to 4%-6% in reflectance and 10%-12% in transmittance. These results suggest that formulae for measurements and computation of coniferous needle OPs require modification that includes also the phenomenon of multiple scattering between the measured needles. © 2008-2012 IEEE.