Journal of Conference Abstracts

An Empirical Statistical Model Using Relationships Between Catchment Characteristics and Nitrate Concentration in Freshwaters to Identify Sensitivity to Nitrogen Breakthrough

M. Kernan¹ (mkernan@geog.ucl.ac.uk), T. E. H. Allott¹ (tallott@geog.ucl.ac.ucl) &
R. Harriman² (HARRIMANR@marlab.ac.uk)

¹ Environmental Change Research Centre, Dept. of Geography, University College London,
26 Bedford Way, London, WC2H0AP, U.K.

² Freshwater Fisheries Laboratory, SOAFD, Faskally, Pitlochry, U.K.

Much of the work carried out on critical loads calculation and mapping has hitherto emphasised the role of S deposition in the acidification of surface waters. Statistical models have been produced quantifying the relationships between S critical loads and catchment characteristics (Hall et al., 1995; Kernan, 1995). The Critical Loads Advisory Group (CLAG) Freshwaters subgroup is currently focusing on the role of nitrogen (N) within the context of total acidity. Additionally, consideration is now been given to a catchment scale approach. The shift towards N means that, in terms of critical loads, surface waters cannot be assumed to be in steady state with incoming acid deposition. The complexities of N cycling dynamics within a catchment are such that N breakthrough, and subsequent acidification, are extremely difficult to predict. Dynamic models such as MAGICwand (Ferrier et al., 1995) seek to model N fluxes through catchment systems and offer a process based approach to predicting freshwater acidity and N breakthrough. However, these models require substantial amounts of high resolution data for parameterisation and, currently, can only be applied to catchments where these data exist. There is a need for intermediate level models (e.g the First Order Acidity Balance (FAB); Henriksen et al., 1993) which can assess the role of N leaching in surface water acidification on a regional basis. The success of such models will depend on the extent to which natural data sets can be used as surrogates for N processes. This paper presents the results of an empirical statistical analysis which seeks to identify those catchment attributes available from national data sets which can best predict N breakthrough.

The model was calibrated in North Wales in an area of high (mapped) N deposition. Each standing water body (76 sites) has been sampled in a 20 x 20 km grid square for water chemistry in summer (maximum N uptake) and in winter (minimum N uptake). The catchments vary in terms of soil type, geology, altitude, land cover and lake/ catchment ratio. Catchment boundaries have been digitised and, using GIS, a variety of attributes have been quantified from available digital data sets.

The spatial variation of freshwater NO₃^- concentration within the grid square is presented together with multivariate statistical analysis indicating those catchment characteristics which most explain this variation. This allows sites which are sensitive to N breakthrough to be defined and predicted within stated confidence limits. Sensitivity to N deposition is mapped at a catchment scale. The results are discussed in terms of the parameterisation of the FAB critical loads model.

References

Ferrier, R.F., Jenkins, A., Cosby, B.J., Helliwell, R.C., Wright, R.F., & Bulger, A.J., Water, Air and Soil Pollut., 85, 707-712 (1995).

Hall, J., Wright, S.M., Sparks, T.H., Ullyet, J., Allott, T.E.H. & Hornung, M., Water, Air and Soil Pollut., 85, 2443-2448 (1995).

Henriksen, A., Forsius, M., Kämäri, J., Posch, M. & Wilander, A., Exceedance of Critical Loads for Lakes in Sweden, Norway and Finland: Reduction Requirements for Nitrogen and Sulphur Deposition. Report 32/1993, NIVA (1993).

Kernan, M., Water, Air and Soil Pollut., 85, 2479-2484 (1995).