Optimising nitrogen nutrition index (NNI) for maize cultivation with controlled release fertilizer treatments guided by UAV remote sensing technology

Purpose Efficient nitrogen (N) management is essential for improving maize yield and resource use efficiency. The nitrogen nutrition index (NNI) is a common tool to assess crop N status. However, its response under controlled-release fertilizer (CRF) management is still not clear. This study evaluated how different N fertilizer types affect maize NNI. It also tested the use of UAV multispectral data for accurate N monitoring. Methods A two-year field experiment with five N treatments was conducted to evaluate the effects of fertilizer type on the maize’s CNDC and the NNI at the V6, VT and R2 growth stages. CNDCs and NNI models were developed from leaf area index (LAI) and aboveground biomass (AGB) and compared using Akaike (AIC) and Bayesian (BIC) information criteria. A gradient boosting decision tree (GBDT) model was built from UAV multispectral data. Results CRFs significantly improved N distribution uniformity and model accuracy. CRF had stronger effects on LAI than on AGB, but NNI based on AGB better reflected actual N balance. Based on accumulated N deficit conditions characterised using AIC and BIC information criteria, a fertilisation range of 150–225 kg ha− 1 was determined. Furthermore, a CNDC model established using measured NNI_AGB was implemented to utilise unmanned aerial vehicles for predicting NNI. The GBDT model based on UAV multispectral data and by screening sensitive variables through variance inflation factors (VIF) achieved the highest accuracy and stability (R2 = 0.88, RMSE = 0.04, nRMSE = 4.98%). Conclusion CRF management combined with UAV-based monitoring offers an effective way to diagnose maize N status. This approach supports precise N application and sustainable agriculture.