A Monte Carlo method for calculating Bayesian uncertainties in internal dosimetry.

Area of Science:

• Computational physics
• Bayesian statistics
• Environmental monitoring

Background:

• Bayesian analysis is crucial for interpreting complex monitoring data.
• Accurate parameter estimation and uncertainty quantification are essential in scientific modeling.

Purpose of the Study:

• To introduce a novel Weighted Likelihood Monte-Carlo sampling method (WeLMoS) for Bayesian analysis.
• To develop and validate a comparative Markov chain Monte Carlo (MCMC) method.
• To assess the capabilities of these methods for parameter estimation and uncertainty quantification in monitoring data.

Main Methods:

• Developed WeLMoS, a Monte Carlo method sampling parameters and weighting by likelihood.
• Developed MCMC using the Metropolis algorithm for direct posterior sampling.
• Evaluated methods using simulated plutonium nitrate aerosol exposure data.

Main Results:

• Both WeLMoS and MCMC methods successfully calculated uncertainty on internal dose.
• The methods provided probability distributions for model parameter values.
• Demonstrated ability to estimate best-fit parameters and associated uncertainties.

Conclusions:

• WeLMoS and MCMC are powerful tools for Bayesian analysis of monitoring data.
• These methods enable robust parameter estimation and uncertainty quantification.
• The methodology is applicable to real-world data, such as determining lung solubility parameters.