Volume no :11, Issue no: 1, March (2014)

VISUALIZING AND INSPECTING SENSITIVITIES IN LARGE MODELS USING DOT CHARTS AND STAR PLOTS, WITH APPLICATION TO COMPARTMENTAL MODELS

Author's: Suzan GazioÄŸlu
Pages: [1] - [30]
Received Date: January 7, 2014
Submitted by:

Abstract

Sensitivity analysis (SA) is a procedure for determining and quantifying the change in model behaviour as model input factors change. When working with models which consist of a large number of model inputs, identifying the most influential factors is of great interest. Even though some graphical tools are used to summarize results from SA, this approach is still rather new. Creating graphs that can help us visualize the results from the analysis can be very useful. In this study, two modern graphical techniques, namely, star plots and dot charts, which can be used for SA, are discussed. These visual summary tools help us to identify the most influential input factors at a glance. To demonstrate their use in SA, we utilize two different compartmental models of the global carbon cycle.

Keywords

computer models, global carbon cycle, graphical sensitivity analysis, visualization techniques.