Remote Estimation of Crop Chlorophyll Content by Means of High‐Spectral‐Resolution Reflectance Techniques

Abstract

The chlorophyll content (Chl) of plants is an important variable in assessing their physiological status and photosynthetic performance. High‐spectral‐resolution reflectance measurements have shown a powerful capacity to detect the subtle variations in spectral absorption features related to changes in Chl. This study proposed a simple technique based on an empirical linear relationship between Chl and the normalized area over reflectance curve (NAOC) index to remotely estimate the spatial distribution of Chl. The NAOC is based on the calculation of the Chl‐sensitive spectral interval obtained by hyperspectral measurements. The data sets used came from four recent European Space Agency campaigns that aimed to study crop properties from space: SPARC (Spain, 2003–2004), AgriSAR (Germany, 2006), CEFLES2 (France, 2007), and Sen3Exp (Spain, 2009). The following steps were undertaken to investigate the proposed relationship: (i) calibrating the SPAD‐502 chlorophyll meter, (ii) collecting spatially distributed Chl data across various crops during the Sen3Exp campaign by using the calibrated SPAD‐502 m, and (iii) fitting a NAOC with Chl field data and tuning it using data from the other campaigns so that a broad variety of crops, climatic conditions, and soil types was covered. The proposed technique was applied to multitemporal hyperspectral images during June 2009. The Chl maps were generated with a RMSE accuracy of 10 μg/cm2. Mapping their differences revealed the spatial distribution of greening up and senescing across crop fields. This new technique shows great potential for remotely monitoring the physiological status of crops.