A new spectral index for estimation of wheat canopy chlorophyll density: considering background interference and view zenith angle effect

Remote sensing (RS) estimation of chlorophyll density serves as an effective measure to assess crop nitrogen (N) nutrition status and guide precision N fertilizer management. Thro mu gh multi-angular RS, this study aims to improve the estimation accuracy of chlorophyll density by reducing the disturbance of mixed background (soil and non-photosynthetic vegetation), and to explore the solutions to minimizing the influence of view zenith angles (VZAs). Wheat canopy multi-angular hyperspectral data (- 60 degrees, - 45 degrees, - 30 degrees, 0 degrees, 30 degrees, 45 degrees, 60 degrees) were systematically collected thro mu gh three-years of field experiments. A soil non-photosynthetic background and angle insensitive vegetation index ( SAIVI = ( ( rho(750))(-1) -( rho(860))(-1)) - ( ( rho(765))(-1) -( rho(860))(-1)) ) / ( ( rho(750))(-1) -( rho(860))(-1)) + ( ( rho(765))(-1) -( rho(860))(-1)) ) ) was proposed for inversion of chlorophyll density. Furthermore, SAIVI, along with another 11 vegetation indices (VIs), were evaluated for their performance in estimating three chlorophyll parameters, namely chlorophyll concentration (CC), canopy chlorophyll density based on leaf area (CCCL) and canopy chlorophyll density based on fresh weight (CCCW). The results indicated that SAIVI had strong stability in restraining distractor (mixed background of soil and non-photosynthetic vegetation). For inversion of CC, CCCL and CCCW, backward VZAs showed higher accuracy than vertical angle. The new proposed SAIVI performed best for estimating CCCL and CCCW with an optimal VZA of - 30 degrees, and the corresponding R-2 and RRMSE of 0.76 and 0.77, 14.5% and 26.6%, respectively.