Abstract: Leaf Area Index (LAI) is a crucial parameter affecting crop photosynthesis and reflecting crop growth status. Timely and accurate monitoring of maize LAI is vital for enhancing crop productivity. The purpose of this research aims to explore the potential of integrating UAV spectral information, thermal information, and canopy morphological parameters to improve the accuracy of maize LAI estimation at various growth stages. Moreover, we attempt to create LAI inversion maps for summer maize based on the optimal estimation model, facilitating precise water and nitrogen management. Firstly, images of maize canopies at multiple growth stages under different water and nitrogen treatments were collected using multispectral and thermal infrared sensors carried by UAVs and simultaneous measurements of maize growth parameters (LAI and crop height ) during two consecutive years of field experiments in 2022-2023. Secondly, LAI estimation models were developed using canopy spectral data, thermal infrared data, canopy morphological parameters, and their combinations, which included Partial Least Squares Regression (PLS), Backpropagation Neural Network (BP), and Random Forest (RF) using various machine learning algorithms. Finally, LAI inversion maps at the in-situ scale were created based on the optimal RF estimation model. It was found that maize LAI responded significantly to water treatments and to water nitrogen treatments, which are suitable for monitoring crop growth. Under the N6 (N6W2) treatment, the LAI ranged from 0.61 to 4.92 (0.89 to 5.94); under the N1 (N1W0) treatment, the LAI ranged from 0.49 to 4.37 (0.65 to 4.08). Vegetation index (VIs), normalised relative temperature (NRCT), and morphological parameters (CM) of the maize canopy showed stable correlations with LAI, however, there are limitations in relying on a single source of information to monitor the leaf area index of maize at multiple growth stages, where the LAI estimation model based on spectral information had R² between 0.36 and 0.61, and RMSE between 0.10 and 0.57. LAI estimation models based on thermal infrared information had R² between 0.25 and 0.49 and RMSE between 0.11 and 0.62. LAI estimation models based on canopy morphology (plant height or plant height+canopy coverage) had R² between 0.34 (0.50) and 0.58 (0.67) and RMSE between 0.06 (0.08) and 0.54 (0.51). In contrast, integrating multiple sources of data (including canopy spectra, thermal infrared data, and canopy morphological parameters) significantly improved the accuracy of LAI estimates, especially at later stages of maize growth, avoiding underestimation that could result from a single source of information. Compared with NRCT+CM and VIs+CM, R² was improved by 15.19% to 19.01% and 3.87% to 12.57%, and RMSE was reduced by 24.93% to 28.13% and 9.51% to 19.94%, respectively. Among the three machine learning models, the RF model showed the highest estimation accuracy with R² = 0.814～0.867, RMSE = 0.065～0.276, and MAE = 0.048～0.207. The inverse maps of in situ-scale LAI using the RF model accurately reflected the water and nitrogen status of the crop. The ranges of LAI for the different growth stages under N1 treatment were 0.36～4.97 under the N1 treatment; the range of LAI under the N6 treatment was 0.54～5.64. The range of LAI at different growth stages under the W0 treatment was 0.31～4.73, and the range of LAI at different growth stages under the W3 treatment was 0.39～5.84. The present study provides a feasible technological pathway for monitoring the growth status of the crop based on the unmanned platform as well as optimising the water and nitrogen management with high application potential. high potential for application. This study provides a feasible technical pathway for UAV-based monitoring of crop growth and the optimization of water and nitrogen management, with high application potential.