Freshwater Umbrella Research

Critical Loads Screening and Validation (CLAG2)

April 1994 — March 1995

Project Summary

The Critical Loads Screening and Validation (CLAG2) project followed on from the original CLAG Fresh waters Sub-Group project of the same name, which ran from April 1990 through March 1994 as part of the original Freshwater Umbrella research programme.

Critical loads maps for sulphur for UK fresh waters have been developed and published using empirical models applied to a dataset of water chemistry collected from sites across the country. This report explains the approach adopted for this critical loads mapping exercise, and describes research undertaken to screen and validate the maps and to aid their interpretation.

Two empirical models have been used for mapping freshwater critical loads in the UK:

Both these models require data on water chemistry, acidic deposition and rainfall/runoff for their application.

In contrast to the random sampling programmes undertaken in Scandinavia, with the production of summary maps of critical loads distributions, the approach in the UK has been to produce national maps of the lowest critical loads and greatest exceedances for each mapping unit in order to highlight regional patterns of vulnerability and damage through acidification. The standard mapping units used were 10 km x 10 km squares in sensitive areas and 20 km x 20 km squares in the least sensitive regions.

In order to select sites for sampling and mapping, protocols for the identification of the sites most sensitive to acidification were developed using sensitivity maps (derived from geology, soils and land-use data) and altitude data. Lakes were selected in preference to streams where possible, and 1573 sites were sampled in order to create a national mapping dataset of water chemistry from which to calculate and map critical loads.

The first stage in the site screening process included the identification of rogue sites where either the water chemistry indicated that reliable critical load or exceedance values were unlikely, or where the location of exceeded sites in relation to the freshwater sensitivity maps and sulphur deposition maps did not fit the expected patterns. 22 rogue sites for possible replacement were identified at this stage of screening and their potential substitutes were sampled.

The second aspect of site screening was to ensure that sites had been selected on a uniform basis and that the most sensitive sites within each grid square according to the sensitivity maps were being used for mapping. 139 sites were found to be not in the most sensitive class on the sensitivity maps, and 58 second choice sites for which more appropriate replacements existed were identified. In all these cases, potential substitute sites were sampled.

To determine which new sites were to replace the original mapping sites, certain criteria were applied. Non-rogue sites with smaller critical loads than the original sites in both SSWC and diatom models were automatically substituted into the mapping dataset. As a result of the site screening exercise, 136 substitutions were made into the mapping dataset. The number of sites showing critical load exceedance increased by over 10% in the new, final mapping dataset.

Since only one site per grid square was used in the mapping dataset, a within-square variability study was carried out to assess how critical loads and exceedances varied between sites in each grid square. A 10% random sub-sample was taken from the most exceeded squares, stratified by exceedance class. Each of the 311 lake sites in these squares was sampled and critical loads and exceedances calculated. In one third of squares the most sensitive site had been correctly chosen for mapping using the diatom model for sulphur, and in a further third of cases a site with a critical load in the same class as the most sensitive site had been selected. For the SSWC model, the selected site was in the most sensitive class within the grid square in only one half of cases. Therefore the mapping protocols adopted mean that there is some overestimation of critical loads compared with the most sensitive sites.

In some mapping squares, streams were sampled because lake sites were absent or deemed inappropriate for critical loads maps. A comparison of critical loads in lake inflow streams and lake outflows at 33 sites found that in half of cases the lake outflow critical load was the smaller of the two. There was no evidence that streams are more sensitive overall, and the preference given to lake sites for mapping appears to be well founded given the greater temporal stability of their chemistry.

Overall it is concluded that the final sulphur critical loads maps for fresh waters using the empirical diatom and SSWC models provide robust tools for policy makers in identifying the most sensitive and most adversely affected areas of the UK in terms of sulphur deposition. However, there are some issues regarding the applicability of the models in specific regions (e.g. the Pennines, north-west Scotland) which will necessitate further study.

Selected publications
  • Battarbee R.W., Allott T.E.H, Bull K., Christie A.E.G., Curtis C.J., Flower R.J., Hall J., Harriman R., Jenkins A., Juggins S., Kreiser A., Metcalfe S.E., Ormerod S.J. and Patrick S.T. (1995) CLAG Freshwaters, critical loads of acid deposition for United Kingdom freshwaters. Department of the Environment, London.
  • Curtis C.J., Allott T.E.H, Battarbee R.W., Harriman R. and Hall J. (1996) Screening and validation of the empirical maps for UK freshwater critical loads. ECRC Research Paper No. 12, ECRC, University College London, London, UK 121pp.