Retrieval of the Leaf Area Index from MODIS Top-of-Atmosphere Reflectance Data Using a Neural Network Supported by Simulation Data

The leaf area index (LAI), a key parameter used to characterize the structure and function of vegetation canopy, is crucial simulations carbon, nitrogen, water cycles Earth’s system. In this paper, neural network (NN) method coupled with canopy atmospheric radiative transfer (RT) proposed realize LAI retrieval without prior data support complex corrections. look-up table (LUT) top-of-atmosphere (TOA) reflectance associated input variables was simulated by 6S (6S simulation) based on top-of-canopy (TOC) LUT PROSAIL. This then train NN obtain inversion model. has been successfully applied MODIS L1B (MOD021KM), realized. estimated compared (MOD15A2H) using mid-latitude summer from 2000 2017 in DIRECT 2.0 ground database. experiments indicated that retrieved TOA (r = 0.7852, RMSE 0.5191) not much different TOC 0.8063, 0.7669), accuracy better than 0.7607, 0.8239), which proves feasibility method.