Estimation of wheat LAI in different leaf layers through LiDAR canopy vertical light distribution parameters and UAV vegetation indices

Different leaf layer area index (LAI) is a key indicator that describes the progress of crop canopies and photosynthetic potential in crops. Presently, remote sensing-based methods mainly focus on estimating total LAI of the canopy, with less research on estimating LAI of different leaf layers within the canopy. A new method for estimating LAI was proposed based on laser radar estimation of canopy vertical light distribution (CVLD), combined with multi-spectral (MS) unmanned aerial vehicle acquisition of vegetation indices (VIs). We constructed different canopy structures of winter wheat by combining different plant types varieties, planting densities and nitrogen application levels. The VIs and LAI values of different leaf layers at 0 day after anthesis (0 DAA), 20 DAA and 30 DAA and canopy vertical point cloud distribution (CVPCD) and CVLD data were measured. Firstly, the correlation between CVLD and LAI at different leaf layers was analyzed, then a model for estimating LAI based on measured CVLD and VIs at each stage was established. Finally, using CVPCD to estimate CVLD accurately enabled the construction of a secondary estimation model for LAI at different leaf layers. The results showed that (1) the model based on the measured CVLD and VIs could accurately estimate LAI of each leaf layer. The coefficient of determination (R²) of the model was between 0.77-0.96. (2) The correlation coefficient between CVPCD and CVLD ranged from 0.58 to 0.87 and using different features of CVPCD could accurately estimate CVLD. (3) The R², RMSE, and MAE of the LAI estimation models for each leaf layer based on the predicted CVLD combined with VIs ranged from 0.36 to 0.92, 0.06 to 0.76, and 0.02 to 0.60 respectively. This method achieved efficient non-destructive estimation of LAI in various leaf layers in winter wheat plants while providing a new perspective for studying CVLD.