The impact of vertical heterogeneity on remote sensing of water content in maize

Abstract:

Water stress can severely damage vegetation health and impact crop productivity. Remote sensing is used to estimate canopy water content for indicating water stress. Previous studies assumed that canopy water content has homogeneous vertical profile, and the variability of water distribution was trivial to influence canopy reflectance because top leaf which were dominantly reflected from the top leaf. However, recent studies have challenged these assumptions by identifying vertical heterogeneity in plant traits such as chlorophyll content. Therefore, the objectives of this research are to investigate the vertical profile of water content in maize and to examine its impacts on remote sensing based estimation. We measured leaf equivalent water thickness (EWT, mg/mm²) from the top, ear, and bottom heights in maize at experimental sites near Mead, Nebraska from May to August 2019. We also measured the leaf area index (LAI), leaf reflectance (380-2500 nm with the ASD spectrometer Field-Spec), and airborne hyperspectral images (400-1000nm and 1-meter resolution with AISA imagers). We calculated spectral indices including NDVI (R₈₂₀-R₆₇₀/ R₈₂₀+R₆₇₀), WI (R₉₇₀/R₉₀₀), mND705 (R₇₅₀-R₇₀₅/ R₇₅₀+R₇₀₅-2R₄₄₅) and mSR705 (R₇₅₀-R₄₄₅/ R₇₀₅+R₄₄₅) from leaf reflectance and examined their correlations with leaf EWT. We also examined these correlations between airborne spectra and canopy EWT at different height (leaf EWT * LAI). At the leaf level, the result showed that average ear leaf EWT (0.09mg/mm²) was higher than average top (0.08mg/mm²) and lower than bottom (0.103 mg/mm²) leaf EWT. At the canopy level, the results showed that 1) four indices had insignificant correlation with EWT when maize have lower water content, 2) vertical heterogeneity didn’t influence estimation with mSR705 and mND705 and 3) vertical heterogeneity had impacts on estimation with WI and better correlation with ear leaf canopy EWT. We concluded that impacts of vertical heterogeneity on EWT estimation varied by applied indices and spatial scale, and suggested to apply mSR705 in canopy EWT estimation and avoid using WI to minimize the impacts of water vertical heterogeneity.