A numerical approach to uncertainty propagation in which values are drawn at random from each input's assigned probability distribution, pushed through the measurement model, and the resulting output values are collected into a distribution. A run of a million trials is the usual target, enough to pin the ends of a 95 % coverage interval to one or two significant digits. It handles nonlinear models, asymmetric outputs and dominant non-normal inputs that the first-order propagation law gets wrong.
Ran the Monte Carlo method with a million trials and it matches the GUM budget to the second digit, so the simple budget stands.
Use Monte Carlo as a cross-check on a propagation budget; if the two agree, keep the simpler budget for the paperwork.
Also heard as
- MCM
- Monte Carlo simulation
- propagation of distributions
Related terms
Law of propagation of uncertainty
The first-order formula that combines input standard uncertainties, weighted by sensitivity coefficients, into the output uncertainty.
Probability density function, the curve describing how likely each possible value of a quantity is.
Coverage interval
The interval around a result that contains the measurand's plausible values with a stated probability.
Measurement model
The equation linking the output quantity you report to all the input quantities that feed into it.
GUM
Guide to the Expression of Uncertainty in Measurement, the core reference method for evaluating measurement uncertainty.