Introducing a new adaptive look-up table subset selection method for leaf chlorophyll and carotenoids retrieval in broadleaved forests

Plant functional traits have been extensively studied using remote sensing. In this study, we compared three approaches for retrieving leaf chlorophyll and carotenoid content in the Lanžhot forest, Czech Republic. For the statistical-based approach, we used field data for both training (calibration portion) and validating (test portion) the models. The physical-based approach utilized a large, simulated look-up table as the training dataset. Lastly, we introduced a new hybrid approach, where a subset of the look-up table was selected based on performance on the calibration portion of the field data. All methods were applied to four Harmonised Landsat Sentinel-2 product images across contrasting phenological phases of vegetation growth. Although the statistical-based approach showed the best results, the hybrid approach presents a promising alternative with potential for future improvement.