• Applied Sobol’ global sensitivity analysis to quantify APSIM-Wheat grain yield sensitivity using 26 years (2000–2025) of daily climate data under farmer-managed irrigation. • Identified canopy establishment and senescence traits (initialₜpla, frₗfₛenᵣate, nodeₛenᵣate) as the dominant controls on yield variability under the tested non-limiting water and nutrient conditions, providing a clear calibration priority. • Found weak and non-significant climate modulation of annual sensitivity indices for the modeled wheat system, indicating broadly stable sensitivity rankings across years under irrigated, non-water-stressed conditions. • Revealed strong parameter interactions (first-order sensitivity indices accounting for 0. 8), followed by frₗfₛenᵣate and nodeₛenᵣate (STi =0. 7–0. 8). photopₛens and grainsₚergramₛtem showed moderate sensitivity (STi =0. 6–0. 7), whereas ttₑndₒfⱼuvenile and maxgrainₛize were lower (STi =0. 55–0. 6) ; yᵣue was least sensitive (STi =0. 45). Regressing annual STi against rainfall, maximum/minimum temperature, and solar radiation indicated weak and statistically non-significant relationships (| r |≤0. 37, p ≥ 0. 07). This stability occurred despite substantial interannual climate variability over the 25-year period, suggesting that parameter sensitivities were largely robust to interannual climate fluctuations, primarily because farmer irrigation consistently buffered water-stress variability. Moreover, Si / STi <0. 5 for all parameters highlights that yield responses were dominated by parameter interactions rather than single-parameter effects. Overall, in warm-temperate, semi-humid monsoon wheat systems, yield calibration should prioritize canopy development and senescence parameters over climate-specific retuning, providing practical guidance for efficient and transferable APSIM-Wheat cultivar calibration under climate variability.
Wang et al. (Sat,) studied this question.