Mapping LAI and chlorophyll content from at-sensor APEX data using a Bayesian optimisation of a coupled canopy-atmosphere model

Abstract

This contribution proposes a methodological approach based on a coupled canopy-atmosphere radiative transfer model and a Bayesian optimization algorithm, which allows the use of a priori data in the retrieval. This approach was used to estimate LAI and leaf chlorophyll content (Cab) in the agricultural test site Oensingen, Switzerland, from at-sensor radiance data of the new airborne APEX imaging spectrometer. The Bayesian optimization allowed having up to 7 free variables in the optimization. The obtained maps of estimated LAI and Cab values at the field level show a good agreement with our expectations in terms of the values themselves, but also their variation range and spread.