Simulations Used in Geotechnical Practice: Part 2 - Comparing Normal and Non-normal PDF in Sampling

Monte Carlo (MC) and Latin Hypercube (LH) sampling were compared in the Part 1 companion paper. The simulation results are compared to show that the LH better represents low probability outcomes by forcing the sampling of the simulation to include events in the tails. At a high number of simulation iterations both provide similar outputs, but at low simulation iterations the LH is more reliable. However, both the MC and LH sampling suffer from impractical values at low or high probability events, when using the normal probability density function (PDF). The normal PDF is commonly used in statistical modelling. Non-normal PDFs often represent the best fit PDF when a goodness of fit test is carried out. The errors associated with using the common normal PDF are shown with the above-mentioned simulation models. This best fit PDF applies whether simulation models are used or even with simple "what if" sensitivity models in traditional spreadsheet analysis.

Keywords: Simulation, Probability density functions, characteristic value, normal PDF, non normal PDF.