Effects of different irrigation regimes on soil moisture availability evaluated by CSM-CERES-Maize model under semi-arid condition

Crop models are useful tools for evaluation of management factors for any possible productivity improvement under water-deficit conditions. Such applications require an accurate simulation of the soil water balance. The main objective of this study was to evaluate the performance of CSM-CERES-Maize model for simulating soil moisture under different irrigation levels of silage-maize. This experiment was conducted in growing seasons of 2003 and 2004. Treatments were four irrigation levels (two deficit-irrigation levels at 0.7 and 0.85 soil moisture depletion (SMD), a full irrigation (SMD) and an over-irrigation treatment (1.13 SMD), indicated by W1, W2, W3 and W4, respectively). Soil moisture was measured on a daily basis in different layers of the soil profile. In the first year, gravimetric sampling method and in the second year a neutron probe were used for measuring soil moisture. Simulated soil moisture was compared with measured field values for each individual soil layer. Results indicated that root mean square error (RMSE) of the model-predicted soil moisture for different treatments, depending on depth, was 0.8-13.6%. Systematic error and the index of agreement of the model in estimating total water in 60 cm soil profile was 0.8-2.00 cm. The greatest error in estimating soil moisture always happened for top layer of the soil profile. Based on the results, it can be concluded that CSM-CERES-Maize model is able to simulate soil moisture content for wide range of soil conditions and irrigation regimes.