Evaluation of Sentinel-2 and RapidEye for Retrieval of LAI in a Saltmarsh using Radiative transfer model

A new era in the retrieval of plant traits has started by emerging the new satellites such as the Copernicus Sentinel families. Among these traits, leaf area index (LAI) is a key indicator of vegetation growth and an essential variable in biodiversity studies. Numerous literature has shown that radiative transfer approach has been a successful method to retrieve LAI from remote sensing data. However, suitability and adaptability of this approach largely depend on the type of remote sensing data, and the ecosystem studied. In this regard, the retrieval of leaf area index in a saltmarsh ecosystem is examined in this study using Sentinel-2 and RapidEye data through inversion of PROSAILH radiative transfer model. Field measurements of LAI and a number of other plant traits were obtained during two succeeding field campaigns in July 2015 and 2016 in the saltmarsh of Schiermonnikoog, the Netherlands. Sentinel-2 (2016) and RapidEye (2015) data were acquired concurrent to the time of field campaigns. The broadly employed PROSAILH model was inverted using a look-up table (LUT) which contained 500, 000 records. Different scenarios of band combinations, as well as different solutions, were considered to obtain the LAI estimates. The R2 and RMSE between measured and estimated LAI were used then to evaluate the retrieval accuracy. The removal of dead materials from the measured LAI improved the estimation accuracies. Our results showed that generally the LAI retrieved using the Sentinel-2 data had higher accuracy compared to RapidEye data. In particular, the SWIR bands of Sentinel were modeled best using the PROSAILH. Leaf area index was best retrieved using the NIR and SWIR bands of Sentinel-2 (R2=0.56, RMSE=1.7). Our results highlight the importance of proper parametrization of radiative transfer models and capacity of Sentinel-2 data, with impending high-quality global observation aptitude, for retrieval of plant traits at a global scale.