Evaluating the Relative Contribution of Photosystems I and II for Leaf Nitrogen Estimation Using Fractional Depth of Fraunhofer Lines and SIF Derived From Sub-Nanometer Airborne Hyperspectral Imagery

• Main Authors: Belwalkar, A., Poblete, T., Hornero, A., Zarco-Tejada, P.J.
• Format: Conference Proceeding
• Language: English
• Subjects: Airborne; Estimation; Fraunhofer lines; GPR; Hyperspectral; Image resolution; leaf Nitrogen; Nitrogen; Physiology; Predictive models; Sensors; SIF; sub-nanometer; Vegetation mapping
• ISSN: 2153-7003
• DOI: 10.1109/IGARSS52108.2023.10281680

Integrating far-red solar-induced chlorophyll fluorescence (SIF 760 ) and leaf biochemical constituents (primarily leaf chlorophyll content (C a+b )) has recently been demonstrated to improve the estimation of leaf nitrogen (N) concentration from airborne and spaceborne hyperspectral imagery in homogenous and heterogeneous crop canopies. The advent of sub-nanometer resolution imagers capable of detecting narrow solar Fraunhofer lines (FLs) has enabled a novel opportunity to investigate the prospect of leaf N estimation using individual FLs in addition to SIF 760 and C a+b traits. This study seeks to determine whether incorporating distinct FL depth derived from sub-nanometer airborne hyperspectral imagery could improve leaf N estimates. A sub-nanometer hyperspectral imager with ≤0.2 nm full-width at half-maximum (FWHM) resolution was flown in tandem with a narrow-band hyperspectral imager with 5.8 nm FWHM over a winter wheat field. Plots were fertilized with variable concentrations of nitrogen to enable nutrient variability. Regression models utilizing Gaussian process regression (GPR) were built with different permutations of SIF, C a+b and depths of individual FLs for estimating leaf N concentration. Laboratory-determined leaf N estimates were obtained by destructive sampling. Results show that GPR models incorporating the depth of distinct Fraunhofer lines as predictor variables performed better than the benchmark model constructed using C a+b and SIF 760 alone. The best leaf N-estimation model built with FLs from the red and far-red regions (C a+b , FL 682.97 nm , FL 757.002 nm ) yielded an R 2 of 0.71, outperforming the standard approach used in previous works (C a+b , SIF 760 ) (R 2 = 0.56).