Calibration of the CLAIR model by using landsat 8 surface reflectance higher-level data and MODIS leaf area index products

Abstract

This study proposes a method for the calibration of the semiempirical CLAIR model, a simplified reflectance model used to estimate Leaf Area Index (LAI) from optical data. The procedure can be applied in case of lacking of both LAI field measurements and surface reflectance data by exploiting free of charge data as the novel high level Landsat 8 Operational Land Imager Surface Reflectance (OLISR) product and the MODIS LAI (MCD15A3H level 4 product). This last dataset was used as LAI reference within an iterative procedure based on the resampling, at the MODIS pixel size, of LAI estimated from OLISR data. The procedure generated LAI information consistent with the MCD15A3H LAI estimation. Lastly, the method was tested and statistically assessed in a territory characterized by an extremely heterogeneous and fragmented landscape (irrigation district “Sinistra Ofanto”) located in the Apulia Region (Italy).