The main sources of uncertainty in our results are due to (1)
model errors and (2) uncertainties in emissions. Uncertainties in
model simulations could arise from lack or insufficient parameterization
of physical processes, finite resolutions, limited input
data, etc. The model performance is therefore continuously verified
by comparison with observations (Section 2.1). Note that the
impact of model uncertainties is limited by focusing on yearly
average concentrations as the end-result: while hourly concentrations
can differ substantial from observations, the yearly average is
much more accurate. However, since we use the same model setup
and meteorology for all the simulations, the uncertainties of the
relative differences reported are mainly driven by the uncertainties
in emissions. For O3 and NOx, additional uncertainties are related to
the ability of the model to simulate the nonlinear chemistry of
these species, and are difficult to estimate. The emission data
include uncertainties in activity data (5e10% for the present
emissions, EMEP/EEA, 2013) and in emission factors (up to 20%,
EMEP/EEA, 2009). The future emissions depend on economic
growth, penetration of new technologies and many other autonomic
developments that are difficult to predict with certainty.
These uncertainties in emissions will have less impact on the
relative differences between scenarios than on each simulation;
also, the differences between future scenarios will be less affected
by uncertainties (since they share the same activity data) than the
differences between present and future scenarios.