Estimation of leaf area index from PROBA/CHRIS hyperspectral multi-angular data

Abstract

Leaf Area Index (LAI) is a key structural and functional biophysical variable of the vegetated surfaces which is important in quantifying evapotranspiration rates and the energy exchange of terrestrial vegetation. Remote sensing offers a method of providing estimates of LAI through the analysis of the Bidirectional Reflectance Distribution Function (BRDF), an angular-dependent surface response. High-resolution, multi-angular and hyperspectral image data from PROBA/CHRIS (Project On-Board Autonomy/ Compact High Resolution Imaging Spectrometer) are used to estimate LAI. The retrieval of LAI is accomplished using the 1D turbid-medium canopy reflectance model, SAIL, coupled with the leaf reflectance model, PROSPECT REDUX. Look-up-tables are generated using scene-specific parameters required to invert the physically based model. Two experiments are performed to examine the contribution of multispectral versus hyperspectral reflectances (nadir direction) and single-look versus multi-look hyperspectral reflectances in deriving the LAI. Image data of the calibration/validation site at Chilbolton, Hampshire, UK are used for the inversion. In addition, ground measurements of LAI are compared with the retrieved LAI estimates. Retrieved LAI estimates using various spectral and directional sampling suggest that the spectro-directional reflectances from CHRIS provides more accurate results than their lower-resolution counterparts such as single-look and multispectral reflectances.

More information

Identifiers

Catalogue record

Export record