The aim of the present research work was the analysis of the within-field crop yield variability in a study field using model simulations. The analysis showed that (1) the insufficient parameterization of the water flow model and (2) the simplified modeling of root growth could be the crucial considerations that mainly lead to the insufficient simulation of the crop growth variability under highly heterogeneous soil conditions. In order to conveniently parameterize the water flow model for the different soils, an approach that combines the use of data from infiltration experiments and of pedo-transfer functions was applied. In addition, based on an existing root growth model and the root length density observation (RLD) in the field, a new root model has been developed and linked to the CERES crop growth model. The comparisons of measured and simulated soil water contents, RLDs, and yields of maize and wheat demonstrated the suitability of the approach to simulate the observed crop yield variability.
«