Designing optimal spectral indices: A feasibility and proof of concept study

Abstract

Vegetation indices constitute a simple and convenient approach to extract useful information from satellite remote sensing data, provided they are designed to address the needs of specific applications and take advantage of the characteristics of particular instruments. Two factors motivate the development of better spectral indices at this time. The first one is the upcoming arrival of a new generation of advanced Earth observation sensors such as the Medium Resolution Imaging Spectrometer (MERIS) on Envisat, the VEGETATION instrument on the SPOT-4 platform, and GLI on ADEOS II, among others. The second is the recent publication of methodological papers on the design and evaluation of optimal spectral indices. The present contribution describes preliminary results obtained in the definition of a spectral index optimized to monitor the state of terrestrial vegetation, where the fraction of absorbed photosynthetically active radiation in plant canopies is considered the key observable physical process. The specifications of the MERIS instrument are used as an example, but the approach can be extended to other sensors. These results are encouraging and show the feasibility of defining optimal indices that exploit advanced characteristics of new instruments to meet the needs of specific applications.