Summer Maize Growth Estimation Based on Near-Surface Multi-Source Data

Rapid and accurate crop chlorophyll content estimation the leaf area index (LAI) are both crucial for guiding field management improving yields. This paper proposes an monitoring method LAI soil plant analytical development (SPAD) values (which closely related to content; we use SPAD instead of relative content) based on fusion ground–air multi-source data. Firstly, in 2020 2021, collected unmanned aerial vehicle (UAV) multispectral data, ground hyperspectral UAV visible-light environmental cumulative temperature data multiple growth stages summer maize, respectively. Secondly, effective height (canopy model (CHM)), accumulation (growing degree days (GDD)), canopy vegetation (mainly spectral index) features maize were extracted, sensitive screened by correlation analysis. Then, single-source linear regression (MLR), partial least-squares (PLSR) random forest (RF) used construct inversion models. Finally, distribution prescription plots was generated trend two analyzed. The results as follows: (1) correlations between position red edge first-order differential value with all greater than 0.5. index, including a near-infrared band, above 0.75. accumulated 0.7. (2) models more made RF achieved highest accuracy In testing set, models’ R2 0.9315 0.7767; RMSE 0.4895 2.8387. (3) absolute error extraction result each map measured small. predicted less 0.4895. difference first increased then decreased advancement period, which line actual conditions. research indicate that proposed could effectively monitor parameters provide scientific basis management.