Model inversion procedure for retrieving wheat biophysical variables from hyperspectral measurements

The paper deals with the estimation of biophysical variables in the frame of a precision farming project. N-trials were performed on wheat crops for 2 years, providing a calibration and a validation dataset. Airborne radiometric measurements were acquired over the trials. Biophysical variables were retrieved from inversion of the radiometric signal through radiative transfer models. To minimize the number of solutions and constrain the inversion, a priori knowledge about the parameters was introduced in the cost function. Using this technique, green leaf area index and leaf chlorophyll content were estimated with a 10% accuracy.