Local and global sensitivity analysis of methane production relevant parameters in the AusBeef rumen model of beef production

Abstract

Sensitivity analysis of mechanistic models allows for identification of influential model parameters as well as evaluation of model behaviour respective to current understanding of the biological processes involved in the model. Local sensitivity analysis, which looks at the effect of changing one parameter at a time on model outputs, is important for identifying internal parameters to which the model is highly sensitive for a given space of input and output. Global sensitivity analysis, which varies multiple parameters at a time across a given space, is important for identifying parameter interactions and downstream effects, as well as checking the mechanistic validity of the model with known biological response patterns.