Research and analysis

R91 and NAME III models: an intercomparison

This report (HPA-CRCE-029) compares the predictions of the HPA application of the ‘R91’ model with those of the UK Met Office NAME III model.


HPA-CRCE-029: intercomparison of the ‘R91’ Gaussian plume model and the UK Met Office’s Lagrangian particle NAME III model in the context of a short-duration release

Request an accessible format.
If you use assistive technology (such as a screen reader) and need a version of this document in a more accessible format, please email Please tell us what format you need. It will help us if you say what assistive technology you use.


This report compares the predictions of the Health Protection Agency (HPA) application of the ‘R91’ model with those of the UK Met Office NAME III model. The study considers a simplified application of NAME and R91 to enable a fair model comparison.

The comparison is centred upon analysis of model output generated from a single baseline run for a short duration release of the type often considered in emergency response assessments. Subsequent model runs are performed, scoping a range of model scenarios and commonly modified model input parameters. Differences in the predictions of the two models are investigated and explained. The quantitative assessment of differences in the baseline model output is used as part of a qualitative assessment of observed differences across a range of model runs and their associated output.

There is a disparity (of up to a factor of approximately 3) between time-integrated activity concentrations in air derived using NAME and those derived using R91, most notably in the near-field. R91 is more conservative in its approach, and estimates made by R91 are typically greater than those made by NAME. The primary sources of the observed differences between the model outputs are identified as:

  • cross-wind spread of the plume
  • vertical spread of the plume
  • wind-driven advection of the plume
Published 1 November 2011