RSV scripts and spreadsheets
This script package is intended to be used as supplementary material for the manuscript: 'Relative sensitivity value (RSV): a metric for measuring input parameter influence in LCA modeling' by Akul Bhatt and Bassim Abbassi. This script package can be used to calculate, plot and export RSV for any LCA model. The MATLAB script 'RSVscript.m' is the main file used to calculate RSVs. It inputs the parameter space X using the script 'LCA_import_param.m' and runs the LCA model script 'livestock_LCA_model.m' to determine the impacts I for each local perturbation, from which the OAT sensitivity space is filled and RSVs are calculated. The script 'LCA_import_param.m' imports the parameter variables and their associated baseline values as a structure X into the workspace. The script 'livestock_LCA_model.m' uses the structure X to calculate the relevant impact scores and stores into an array I. Based on the parameter structure X and the impact array I (both stored in the workspace), the script 'RSVscript.m' calculates RSV for each parameter, life cycle phase and impact category. The script also exports the RSVs to the spreadsheet LCA Model 'LCA Model Input-Output.xlsx'. Note that an LCA practitioner can replace the current livestock LCA model script with their own LCA model script within 'RSVscript.m' to determine RSV for their own model. The only requirements are that variables X and I be used to represent the input parameter structure and output impact array in the workspace, respectively. The script 'OAT_plot.m' plots the local sensitivity of all input parameters to impacts across all life cycle phases and impact categories. This can be useful in visually verifying model linearity.
Steps to reproduce
Simply run the script 'RSVscript.m' in MATLAB to calculate and output the RSV results in the worksheet 'RSV Output' of spreadsheet 'LCA Model Input-Output.xlsx'. The optional script 'OAT_plot.m' may be ran to plot the local sensitivity of the parameter space X. Input parameters accepted by the LCA model ('livestock_LCA_model.m') can be modified in the worksheet 'Parameter Inputs' before running the RSV script.