A unified SIF-based framework integrating daytime and nighttime transpiration for irrigated maize in semi-arid regions

Accurate estimation of crop transpiration is important for understanding water use in irrigated agroecosystems. However, most existing solar-induced chlorophyll fluorescence (SIF)-based transpiration models focus on daytime transpiration ( T d ) and do not explicitly represent nighttime transpiration ( T n ), limiting their ability to estimate continuous 24 h crop water use. This study developed an integrated 24 h SIF-based transpiration framework for estimating maize transpiration across daytime and nighttime periods in a semi-arid irrigation district. Using the 2023 eddy covariance fluxes and meteorological observations, the SIF–Gc and SIF–GPP approaches were first evaluated for T d estimation. The better-performing daytime model was then improved by incorporating a nonlinear volumetric soil water content term ( VWC 2 ) identified through sensitivity analysis. A T n model was further developed by incorporating longwave radiation, residual stomatal conductance, and CO 2 inhibition effects into the SIF–Gc method. The improved T d model and the new T n model were then combined into an integrated 24 h transpiration framework, with 2023 used for calibration and 2024–2025 for validation. The results indicated that the SIF–Gc approach performed better than the SIF–GPP approach for T d estimation, with R 2 of 0.72 and RMSE of 1.25 mm/d. After incorporating VWC 2 , the T d model improved consistently during 2023–2025, with the mean R 2 increasing to 0.84 and RMSE decreasing by 18.67% on average. The T n model also performed well, with R 2 values of 0.81–0.84. The integrated 24 h transpiration framework achieved robust performance for total transpiration ( T total ) during the 2023–2025 growing seasons, with R 2 values of 0.82–0.90. Overall, explicitly separating and integrating T d and T n improved the estimation of T total and provided a practical framework for continuous 24 h transpiration estimation in irrigated maize systems in semi-arid regions. • A SIF-based framework was developed for daytime and nighttime transpiration estimation. • A nonlinear soil moisture term improved daytime transpiration estimation across years. • Nighttime transpiration was modeled using Lw , residual Gc , and CO 2 inhibition. • The integrated framework enabled continuous 24 h transpiration estimation in irrigated maize.