Journal of Conference Abstracts

Investigating the Consequences of Uncertainties in the BC/Al Dose Response Function in an Environmental Risk Perspective, Using the Critical Load Concept

Andreas Barkman (Andreas.Barkman@chemeng.lth.se),
M. Alveteg (Mattias.Alveteg@chemeng.lth.se),
P. Warfvinge (Per.Warfvinge@chemeng.lth.se) & H. Sverdrup (Harald.Sverdrup@chemeng.lth.se)

Department of Chemical Engineering 2, P.O. Box 124, Lund University, Sweden.

The molar base cation to inorganic aluminium ratio in soil solution is frequently used within the Convention on Long-Range Transboundary Air Pollution (CLRTAP) as the chemical criteria for the assessment of critical loads of acidifying compounds for terrestrial ecosystems. The critical limit is usually set to unity. Although there is a rather strong correlation between adverse effects on forest ecosystems and the BC/Al ratio, the scientific support is not strong for setting a distinct limit on the BC/Al ratio under which no adverse effects occur. A critical load as determined by a critical limit of the BC/Al ratio can thus be viewed upon as an ecological risk assessment. The value of the critical limit implicitly defines an acceptable risk of adverse effects.

Critical loads assessments are iterative in nature and utilised chemical criteria and critical limits are always chosen from the perspective of present knowledge. However present knowledge is always impaired by a range of uncertainties, which may question or alter the actual usefulness of the critical loads concept and the biogeochemical models utilised for its assessment.

The objective of this study was to investigate the effects of uncertainties in the BC/Al dose response on calculated ecological responses. Uncertainties in the BC/Al dose response function were treated as an uncertainty in the half saturation constant in the valence unspecific ion exchange equation commonly used for evaluating the BC/Al dose response. Methodological uncertainties are thus not considered.

The regionalised version of the soil chemistry model PROFILE was used to assess the steady-state chemistry and the anticipated ecological response for Norway Spruce (Picea Abies) for 300 sites from the Swedish National Forest Inventory. Sites were chosen to represent the national distribution of calculated critical loads and exceedances, as well as for the geographical and geological regions of Sweden. The combined effect of uncertainties in the dose-response function and uncertainties in model input were estimated by means of Monte-Carlo simulations, with 500 runs per site, using an ordinary random sampling technique.

Calculated BC/Al ratios in soil solution for each site ranged between 25-40 % of the best estimate. Available data on the BC/Al dose response produced a skewed distribution of possible half-saturation constants, ranging from 1 x 10^-5 to 0.8. A weighted average BC/Al ratio was calculated for each site using layer thickness and total rooting depth.

The overlap between percentiles of anticipated effects, limits the possibility of identifying risk classes for which the effects are statistically different. Thus, a high requirement of high probability of sites being systematically different reduces the possibility of identifying distinct classes.

Consequently, the usefulness of the BC/Al ratio as a measurement endpoint, may therefore be limited in some fields of applications and to some users if uncertainties are included. This is particularly true for applications on the county and municipality level in Sweden where identification of risk areas is emphasised. Dynamic models may improve the possibility of resolving the categorisation of risk classes. The spread of the anticipated response is, however, not likely to decrease. Dynamic forecasts of foreseen adverse effects may therefore still be constrained by uncertainties.