Sensitivity of crop model predictions to entire meteorological and soil input datasets highlights vulnerability to drought

Crop growth models are increasingly used as part of research into areas such as climate change and
bioenergy, so it is particularly important to understand the effects of environmental inputs on model
results. Rather than investigating the effects of separate input parameters, we assess results obtained
from a crop growth model using a selection of entire meteorological and soil input datasets, since these
define modelled conditions. Yields are found to vary significantly only where the combination of inputs
makes the crop vulnerable to drought, rather than being especially sensitive to any single input. Results
highlight the significance of soil water parameters, which are likely to become increasingly critical in
areas affected by climate change. Differences between datasets demonstrate the need to consider the
dataset-dependence of parameterised model terms, both for model validation and predictions based on
alternative datasets.