A calibration protocol for soil-crop models

Abstract

Process-based soil-crop models are widely used in agronomic research. They are the tools of choice for evaluating climate change impact on crop production and for testing adaptation and mitigation strategies. Multi-model simulation studies show a wide diversity of results among models given the same inputs, implying that simulation results are very uncertain. A large part of the variability results from uncertainty in parameter values due to different approaches to model calibration. This study proposes an innovative calibration protocol that should improve model predictions compared to common practices and that should reduce variability between modeling groups. The two major innovations concern the treatment of multiple output variables and the choice of parameters to estimate, both of which are here based on standard statistical procedure adapted to the specificities of soil-crop models. The protocol performed well in an artificial-data test with relatively many measured variables and estimated parameters. The protocol is formulated so as to be applicable to a wide range of models and data sets. If widely adopted, it could substantially reduce model error and inter-model variability.