Field-scale evaluation of satellite-derived vegetation indices and image timing for in-season nitrogen management in corn

Introduction: L.) rely on grain yield data, limiting in-season decision-making for nitrogen (N) management. Vegetation indices (VIs) derived from satellite imagery can serve as proxies for grain yield and help estimate in-season AONR, enabling timely sidedress applications. This study aimed to (1) quantify how VI-yield relationships vary during the vegetative period across fields with different tillage systems and crop residue; (2) determine whether VI-N response curves can be used to estimate AONR (AONRvi); and (3) assess the accuracy of AONRvi relative to yield-based AONR (AONRy). Methods: Three rainfed on-farm field trials with contrasting tillage systems and four to five N rates were conducted in Indiana in 2021. PlanetScope imagery (3-m resolution) was used to calculate 16 VIs (8 NIR-based and 8 RGB-based) across multiple growth stages. Linear regressions between yield and VIs were used to assess strength of their relationship, followed by VI-N response curves to estimate AONRvi. Results and Discussion: ≤ 0.31) and varied across fields, with fewer significant relationships under higher crop residue conditions. Of the VI-N response curves evaluated, 17% met selection criteria for estimating AONRvi, with lower proportions observed in higher-residue systems. Mid-vegetative period imagery (V10-V11) produced the smallest deviations from AONRy for NIR-based indices, although no single VI consistently performed best across fields and timings. These results indicate that 3-m satellite imagery has potential to detect crop N response under commercial field conditions, but its reliability for estimating in-season AONR depends on management context, image timing, and spectral domain.