Leaf Area Index Retrieval From SPARC Data: Assessment Of Radiative Transfer Model Inversion

Leaf Area Index (LAI) is a key parameter for many biophysical and climatic models. In the field of interest of our research group, an accurate LAI estimation is needed for modelling crop water requirements for precision farming and agricultural resource management applications. The objective of this study is to assess the accuracy of LAI retrieval from EO data by means of a radiative transfer
model inversion technique. To this aim multi†angular CHRIS/PROBA data, from SPARC 2003 and 2004 campaigns, has been employed in the inversion of PROSPECT†SAILH (P†SH) model by using a numerical optimisation technique based on the Marquardt†Levenberg (M†L) algorithm. From the same data set, the closer to nadir reflectance in the red and near†infrared bands has been selected in order to estimate LAI by using an empirical approach based on the CLAIR model. Such estimated LAI has been thus employed as prior information in the P†SH model. LAI values retrieved with this combined approach have been estimated with good accuracy for some type of crops (e.g. R2 = 0.80, RMSE=0.51 m2m−2 for Alfalfa canopies). Ongoing and future work includes further improvements of the M†L optimisation method and the implementation of a different optimisation method based on Genetic Algorithm GA.