Re-tuning the Walker-Kasting global carbon cycle box model using a parameter sensitivity analysis

Abstract

introduction: The Walker-Kasting global carbon cycle box model is a simple representation of the earth system used to study climatic events. This model has a high number of parameters whose sensitivity must be tested in order to better understand which of them dominate the behaviour of the model. In this study, we perform a parameter sensitivity analysis. moreover, we use these results to re-tune the model to preindustrial conditions using a quantitative criterion. We then compare our results to those determined by Walker-Kasting. Methods: We achieved the parameter sensitivity analysis by calculating, for each parameter, an index that measures the impact of a change in the initial parameter value on the equilibrium solutions. The most sensitive parameters were determined and then tuned in the model by comparing the model equilibrium solutions to a set of 32 experimental values. results: We found that nine of the tuning parameters were sensitive to a change to their initial value. Furthermore, we discovered that 5 of these parameter values were identical to those determined by Walker-Kasting, thus affirming their work. discussion: a sensitivity analysis is interesting to perform because it allows the users of a model to more fully comprehend the way in which the model reacts to changes in its parameters. sensitivity analysis is fundamental in the tuning of a model (for example, to a particular period in the earth’s history) since it allows researchers to consider only the most important parameters.