The EA is concerned about the use of Critical Load Models to set pollution targets for aquatic and terrestrial ecosystems. There is considerable uncertainty associated with the CL Models and the drivers for these such as atmospheric pollution models and basic data on key chemical processes and deposition rates. The project was therefore set up to assess the sources of uncertainty and errors likely to occur when using Critical load models.

The Critical Loads Model provides a chemical mass balance of the inputs and outputs for any ecosystem. It is used as part of the Transboundary Pollution Control UN/ECE protocols to set national guidelines for atmospheric pollution control. The project was funded by the EA to assess uncertainty in Critical Load Models. It was decided to study a coniferous forested site near Liphook in Hampshire (Lat 51º 04' N.; Long. 0º 47' W.; OS Grid Reference 485700, 129900). This was a research site with a lot of site-specific, measured information. The Critical loads Model was set up and a detailed Monte Carlo analysis performed using the software package Crystal Ball (Decisioneering UK Ltd) which operates as an add-on to Microsoft Excel. Crystal Ball runs very fast and is thus ideal for investigating the effects of different Monte Carlo parameters and techniques. A comprehensive analysis of uncertainty was undertaken It was shown that that uncertainty limits would be wider for parameters derived from national datasets rather than measurements at research sites.

See EA report for details
Abbott, J., Hayman, G., Vincent, K., Metcalfe, S., Dore, T., Skeffington, R. A., Whitehead, P. G., Whyatt, D., Passant, N., and Woodfield, M. (2003) Uncertainty in acid deposition modelling and critical load assessments R&D Technical Report TR4-083(5)/1, Environment Agency (England and Wales), Bristol, UK.