ALMANAC model parameters sensitivity analysis by Sobol algorithm

Sobol algorithm is one of the most representative global sensitivity analysis methods. Parameters sensitivity analysis is a necessary step for crop model localization and regionalization. We used Sobol algorithm to analyze the parameters sensitivity of ALMANAC model, including crop parameters, soil parameters, meteorological parameters and field operating parameters. The result of the study showed that harvest index, potential radiation use efficiency, base temperature, maximum crop height, and potential heat units are the high sensitive parameters in model localization process; in the regionalization process, meteorological parameters and field operating parameters are more sensitive than soil parameters. After the two steps of sensitivity analysis, we summarized that heat was the most important factor in ALMANAC model.