Remote sensing of grassland chlorophyll content: Assessing the spatial-temporal performance of spectral indices

The retrieval of chlorophyll content at a range of spatial-temporal scales is central for monitoring vegetation physiology, productivity and health. However, estimating chlorophyll measurements from the leaf to canopy level still remains a challenge over years, especially for heterogeneous landscapes. This study focuses on the estimation of chlorophyll content between years based on the combination of lab-based leaf hyperspectral data, field-based canopy hyperspectral data, and SPOT-5 multispectral imagery. A comprehensive list of 144 published spectral indices including broad- and narrow-band were calculated and their performances are evaluated in order to determine a general spectral index that could be used to estimate vegetation chlorophyll contents between two consecutive years at leaf and canopy levels. At the leaf and field-based scales, narrowband indices utilizing wavelengths from the red to red edge region had the strongest relationships with chlorophyll content for both field seasons. At the space-based multispectral level, the Simple Ratio (SR) index had the best predictive performance for canopy level chlorophyll content. A chlorophyll map was generated from SPOT-5 derived SR values with a map accuracy of 63.6%.