Monte-Carlo sensitivity
FISPACT-II uses a Monte-Carlo approach to sensitivity analysis. A series S of inventory calculations are performed with the set of I independent variables chosen from distributions with means and standard deviations, which are the input cross sections and/or the half-lives (note only cross section variation is available in Release 3.00). These runs produce a set of J dependent variables, which are outputs of the code: either nuclide inventories or some radiological quantities.
The implementation of the scheme uses the SENSITIVITY keyword to initialise the collecting of data within the main inventory calculation. The keyword ZERO causes the series of S runs with different independent variables to be undertaken to compute, process and output the set of dependent variables. The default distribution is taken to be log-normal, but can be modified using the MCSAMPLE keyword. Any sequence of irradiation pulses, changes in cross section, etc. that are possible with FISPACT-II can be used in the sensitivity calculations. The code performs the base calculation with full output and then repeats S times the sequence of steps with different sets. The resluts of the base calculation are not included in the sensitvity calculation.
Sensitivity calculations provide both uncertainty and sensitivity output. Summary uncertainty output of the independent and dependent means and their standard deviations are sent to the output file.
The sensitivty of each dependent quantity on the independent variables is assessed using the Pearson product-moment correlation coefficient, r_ij.
The magnitude of r_ij is less or equal to one and a value near one indicates near linear correlation. Values less than one are typically expected for decays depleting the nuclide in question, while positive for those which produce it.
FISPACT-II writes tables of means, standard deviations and correlation coefficients to output, and then writes the raw data of independent and dependent variables to the sens file for post-processing by the user.
