Close-range, heavy-duty hyperspectral imaging for tracking drought impacts using the PROCOSINE model. (March 2021)

• Record Type: Journal Article
• Authors: Proctor, Cameron; Dao, Phuong D.; He, Yuhong
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 263(2021)
• Publication Date: 2021-03
• Subjects: Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2021.107528
• Languages: English
• ISSNs: 0022-4073

Abstract

Highlights: Drought creates a mixed canopy of photosynthetic and non-photosynthetic vegetation. The spectral evolution was tracked by a novel close-range hyperspectral platform. PROCOSINE and PROSPECT model inversion were contrasted. The PROCOSINE model effectively tracked leaf decay due to drought. Abstract: Leaf response to stressors is heterogeneously distributed within the canopy, creating complex light-matter interactions that hinder translation of spectra into meaningful plant traits at larger scales. A deeper understanding of the linkages between leaf traits and canopy spectra has been achieved by a variety of close-range hyperspectral imaging systems. Yet the majority are scaled for small plants in pots, and thereby face numerous challenges in replicating field conditions for temporal drought monitoring. In monocots, drought alters foliar biochemistry and canopy structure, leading to the eventual senescence and decay of individual leaves, yet few radiative transfer models have been applied to track drought stress at the submillimeter scale in mixed canopies of photosynthetic and non-photosynthetic vegetation. Our objective was to compare the retrieved parameter set for PROCOSINE and PROSPECT radiative transfer model inversions utilizing two decay pigments. Although the average inversion values for both models indicated an increase in initial decay pigments after drought, the per-pixel inversion results for the PROCOSINE and PROSPECT inversions were unique (Mann-Whitney U-test p<0.05). Model retrievals for PROCOSINE were consistently higher for N, and lower for chlorophyll and carotenoids. In PROCOSINE, the Thatai and BSpec trended towards smaller values due to the increasing percentage of non-photosynthetic leaves, suggesting that mixed canopies are less prone to bidirectional reflectance. Regardless, the spatial-temporal trend of the inversion estimated parameter set was similar between both models, demonstrating the efficacy of the close-range imaging platform to monitor drought effects over time.