Potential of Hyperion EO-1 hyperspectral data for wheat crop chlorophyll content estimation

Abstract

Chlorophyll content is an essential biochemical parameter to track the main developmental stages and yield of cereals relevant for agriculture. In this perspective, several spectral chlorophyll indices have been developed to estimate chlorophyll content at both the leaf and canopy levels using remote sensing data and considering different crop types. The application of these chlorophyll indices under agricultural field conditions has not been rigorously tested and validated for wheat crops. The objective of this study is to investigate the relationship between a wide range of spectral chlorophyll indices and chlorophyll content of wheat crop using hyperspectral data acquired with the Hyperion Earth Observing 1 (EO-1) sensor and ground and laboratory measurements for validation purposes. The Hyperion data and ground measurements were acquired on 30 June 2002 over two agricultural fields near Indian Head, Saskatchewan, Canada. The image data were corrected for a spatial shift between the visible near-infrared (VNIR) and short-wave infrared (SWIR) detectors, destriped, and noise reduced. The data were then transformed to surface reflectance, corrected for sensor smile (1–3 nm shift in the VNIR and SWIR), and postprocessed to remove residual errors. The ground measurements included the leaf chlorophyll in arbitrary units using the soil-plant analyses development 502 (SPAD-502) meter and leaf chlorophyll content estimated from chemical laboratory analysis. The SPAD-502 measurements were correlated with laboratory-extracted leaf chlorophyll content to establish two calibration equations for the computation of chlorophyll ab (Chl ab) and chlorophyll a (Chl a) content, with a coefficient of determination (R2) of 0.72 and 0.69 and a root mean square error (RMSE) of 3.53 and 1.94 µg/cm2, respectively. The chlorophyll indices were derived from the Hyperion data and validated against those derived from a subset of the converted SPAD-502 measurements. The normalized difference pigment index (NDPI) showed the best results for wheat chlorophyll content estimation using the Hyperion data against those derived from the converted SPAD-502 measurements, with an index of agreement (D) of 0.66 and an RMSE of 2.89 µg/cm2. The performance of the NDPI at the wheat canopy level makes it a suitable tool to calibrate biophysical process models used in agriculture.