Surface soil moisture and land surface temperature as constraints for multi-objective calibration of a dual-source SVAT model over semi-arid irrigated winter wheat

Accurate estimation of turbulent fluxes, particularly sensible and latent heat fluxes, is essential for understanding land–atmosphere interactions and improving water and energy management in agricultural. This study applies ICARE-SVAT model over four winter wheat fields in the semi-arid Tensift Basin (Morocco) to estimate surface energy fluxes. Model calibration was performed at a reference field using two multi-objective algorithms (NSGA-II and MOSCEM-UA) under three calibration scenarios:(1) SSM-constrained, (2) LST-constrained, and (3) combined SSM+LST-constrained. The parameters adjusted during optimization were the minimum stomatal resistance rsc,min, the roughness ratio (z0/z0h), and the two empirical coefficients of soil resistance to evaporation ( ARSS and BRSS). Model Performance was evaluated against observed SSM, LST, heat fluxes and ETa partitioning. Results show that both algorithms showed robust and consistent performance, with comparable accuracies (R² and RMSE). Best simulations achieved, RMSE (R²) of 0.05 m³ .m−3 (0.85) for SSM and 1.38 °C (0.89) for LST. Flux estimations achieved RMSE (R²) values of 41 W.m⁻² (0.85) for LE, 32 W.m⁻² (0.82) for H, 21 W.m⁻² (0.87) for G, and 40 W.m⁻² (0.96) for Rn. The model also reproduced evapotranspiration partitioning accurately, with high R² (0.78–0.90) and low RMSE (0.15–0.40 mm.day⁻¹) for both transpiration (T) and soil evaporation (E). Validation over the three additional wheat fields confirmed the model’s robustness and transferability. Results suggest that SSM-only calibration is suitable when soil moisture data are frequent and accurate, LST-only calibration when energy balance accuracy is prioritized, and the combined SSM+LST configuration when both variables are available, as it offers the most balanced performance across all fluxes. Overall, integrating SSM and LST into SVAT modeling offers strong potential for improving water and energy management in semi-arid regions.