Modelling Catchment-Scale Responses to Climate Change - Climate Change Impacts on Freshwater Ecosystems - page 257

0.35

0.3

0.25 Monthly mean

Mean plus 1 StDev
Mean minus 1 StDev
Upper 95% confidence limit
Lower 95% confidence limit

0.2

0.15

0.1

0.05

0

Month

Figure 10.6 Soluble reactive phosphorus concentrations simulated by the INCA-P
model for the 2050s in the River Lambourn, together with uncertainty bounds.

used to fit confidence intervals to model predictions. The tool can also be used
for sensitivity analysis using Kolmogorov-Smirnoff statistics to produce a list of
input parameters ranked in order of their importance in distinguishing good
from bad simulations. It is being systematically applied to simulation results
produced in the Euro-limpacs project from catchments across Europe in order to
assess patterns in overall uncertainty, but this activity was not complete at the
time of writing.

Conclusions

The ultimate goal of the catchment modelling in the Euro-limpacs project was to
provide a pan-European assessment of the likely impacts of climate change on
the flow regimes and water quality of river systems representative of key climate
types across Europe. For a comprehensive assessment of water quality, a range of
water quality parameters needed to be considered. Those considered were
nitrogen, sediment, phosphorus, indicators of acidification (such as pH, alkalinity
and acid neutralizing capacity), carbon and mercury. We are still some way from
the ultimate goal, but we have made significant progress in three areas.
First, new catchment-scale models have been developed and existing ones
improved leading to an improved capability in understanding how different
factors and processes are integrated in response to climate or land-use management
changes. Each of the INCA family of models has been harmonized in the sense
that the versions for each of the different water quality indicators have the same
structural representation of the key hydrological stores and pathways. The only
differences between the models are the inputs and the biogeochemical cycle
incorporated for each water quality measure. Thus, a toolkit of models now
exists for a range of key pollutants with a common representation of catchment
structure and hydrology.