Climate Change: Defining Reference Conditions and Restoring Freshwater Ecosystems - Climate Change Impacts on Freshwater Ecosystems - page 215

The Round Loch of Glenhead

6.4

MAGIC model pH (Gothenburg Protocol)
Diatom-inferred pH
reconstructions (3 cores, 3 training sets)
pH of analogue sites (1800)
Weighted mean (1800)
Mean annual pH 1988-2000

6.2

6.0

5.8

5.6

5.4

5.2

5.0

4.8

4.6

1790 1810 1830 1850 1870 1890 1910 1930 1950 1970 1990 2010 2030

Figure 9.5 Comparison of pH reconstruction outputs and annual measured pH.
Chronology of diatom inferred pH according to SWAP (The Surface Water Acidification
Programme), UK and EDDI models (fine lines) for 210Pb dated samples from three
sediment cores (RLGH 81, RLGH 3 and K05). The RMSEP of the SWAP, UK and EDDI
training sets are 0.38, 0.31 and 0.25 pH units respectively. Modern annual pH of nine
lakes providing the strongest biological analogues for a pre-acidification (c. 1800)
sediment sample (open triangles) and the weighted average of these (filled triangle).
MAGIC model pH reconstruction (open circles) and mean annual average pH for the
period 1988e2000 and the year 1979 (open squares). (From Battarbee et al . 2005.)

to have been higher in the past when sulphur deposition was lower. Although
not enough is known about the relationship between sulphur deposition and
soil water DOC concentration for DOC to be incorporated into the MAGIC
model, evidence is growing that the relationship is important (e.g. Monteith
et al . 2007).
With regard to pH reconstruction using sediment records, uncertainty could
still be reduced by increasing the size and representativeness of the calibration
data-sets and by continued verification against observations from monitoring
programmes. Overall, the finding that different methods to establish a reference
condition were congruent, and, importantly, that model predictions agreed with
contemporary monitoring data are encouraging. This study clearly demonstrates
the usefulness of hindcasting methods for establishing the reference condition of
lakes where no present-day spatial analogues exist. However, the modelling and
contemporary time-series also indicate substantial variability on decadal and
yearly time scales, trends that are clearly the result of human-generated changes
in surface water quality. More knowledge is needed concerning how climate
change, with both relatively short- (interannual) and long-term (shifts in baseline
conditions) variability, affects our ability to detect degradation and recovery.