Performance of chlorophyll prediction indices for Eragrostis tef at Sentinel-2 MSI and Landsat-8 OLI spectral resolutions

Abstract

As remotely sensed data becomes more readily available around the world, satellites such as Landsat-8 and Sentinel-2 have great potential to support precision agriculture. Sensors with high spectral and spatial resolutions are particularly optimal for limited land resource farmers to improve land management. The objective of this short communication is to assess the performance of Sentinel-2 and Landsat-8 multispectral bands for chlorophyll prediction using indices that were originally developed using imaging spectroscopy/hyperspectral data. Remotely sensed chlorophyll content measures are often utilized as a proxy of plant health. Performance of a group of chlorophyll prediction indices is tested for tef (Eragrostis tef), an endemic grass crop native to Ethiopia that forms a major component of Ethiopian diets and is grown by limited land resource farmers. Hyperspectral reflectance data captured in situ at the canopy level were convolved into bands approximating Landsat-8 and Sentinel-2 sensors, and a suite of chlorophyll prediction indices were computed and regressed against chlorophyll content. Results show that simple pigment indices employing wavelengths corresponding to the blue and ultra-blue bands performed best for predicting chlorophyll. The red-edge index computed using the Sentinel-2 bands also performed well. These findings suggest that publicly available, multispectral imagery can potentially substitute for hyperspectral data in chlorophyll prediction indices, thereby improving the accessibility of precision agriculture methods.