required, including characterization of the emitted
particles, investigation of health impacts, and the
development of methods of prevention and mitigation. The importance of brake wear may diminish in the future as regenerative braking systems in
hybrid and electric vehicles partially or fully
replace traditional mechanical ones. However,
this is not currently taken into account in emission
calculations and projections.
More detailed information on PM emissions
from rail and shipping (especially inland shipping) is also required, including emission factors
for different engines and fuel type combinations,
in terms of size distribution, particle numbers, and
chemical speciation [34]. Few studies have examined PM levels in the ambient air alongside railway lines. These often run very close to housing,
and may contribute to local air quality problems,
especially where large numbers of diesel trains are
in use. PM concentrations could also be elevated
near shunting yards and depots. In addition, there
is little information on PM concentrations in stations where diesel trains are left running at idle
and the dispersion of pollution is restricted.
Understanding NO x Emissions In order to
understand atmospheric NO 2 it is important to
understand NO x emissions, and it is possible that
the latter are being underestimated by models. For
example, there is evidence to suggest that NO x
emissions from Euro V heavy-duty vehicles are
significantly higher than the limit at type approval
[110]. Vehicles equipped with SCR have very low
emission levels during rural and motorway driving, but not under the low engine loads associated
with urban or driving. The reason for this is that
the SCR system needs to attain a certain temperature for efficient NO x reduction. The real-world
performance of SCR should therefore be investigated. Furthermore, more extensive information
on f-NO 2 values for a wider range of vehicles is
required.
Improving Emission Models Accuracy, reliability, consistency, and credibility are prerequisites of emission estimates. However,
comparisons between the results from different
emission models have highlighted substantial differences. There is therefore a need for emission
modelling approaches to be reevaluated, and for
improvements to be made. This requires a better
understanding of the circumstances under which
emission models provide inaccurate predictions.
For example, some modellers believe that the
exhaust emission factors for road vehicles are
unsuitable for certain road types and traffic conditions. Particular concerns have been expressed
about the validity of the emission factors for
so-called congested traffic. The term “congestion”
also needs to be defined in a manner that is meaningful in terms of emissions.
Improving Activity Data While obtaining primary activity data for road transport by direct
measurement remains the main approach for
short-term modelling exercises, it tends to be a
rather expensive activity, and would be unfeasible
where an emission estimate is required for a large
geographical area. Traffic assignment models and
microsimulation tools are increasingly being used
to provide the relevant data, and communications
technology such as Intelligent Transport Systems
and Advanced Transport Telematics enable transport networks to be monitored in real time. It
could well be possible to adapt a number of
models and new technologies to contribute appropriate information on fleets and traffic activity.
In shipping the use of data from AIS and
CEMS has a great deal of potential, and even
more so if it can also be extended to cover inland
shipping. Data from AIS have already been incorporated into at least one model [167]. However,
some work is required to validate emission estimates obtained using AIS data [34]. Research is
also required on the movements of empty ships, as
these are often excluded from emission
estimates [34].
Model Validation Model validation is an important area for improvement. The literature on the
validation of transport emission models is quite
limited, and validation studies are often only partial in the sense that they address just one type of
road, type of vehicle, or time period. The understanding of the uncertainties in emission models
and the main factors affecting prediction errors is
therefore poor [168]. Emission models should
quantify prediction errors. The models in common
Air Quality, Surface Transportation Impacts on
77
particles, investigation of health impacts, and the
development of methods of prevention and mitigation. The importance of brake wear may diminish in the future as regenerative braking systems in
hybrid and electric vehicles partially or fully
replace traditional mechanical ones. However,
this is not currently taken into account in emission
calculations and projections.
More detailed information on PM emissions
from rail and shipping (especially inland shipping) is also required, including emission factors
for different engines and fuel type combinations,
in terms of size distribution, particle numbers, and
chemical speciation [34]. Few studies have examined PM levels in the ambient air alongside railway lines. These often run very close to housing,
and may contribute to local air quality problems,
especially where large numbers of diesel trains are
in use. PM concentrations could also be elevated
near shunting yards and depots. In addition, there
is little information on PM concentrations in stations where diesel trains are left running at idle
and the dispersion of pollution is restricted.
Understanding NO x Emissions In order to
understand atmospheric NO 2 it is important to
understand NO x emissions, and it is possible that
the latter are being underestimated by models. For
example, there is evidence to suggest that NO x
emissions from Euro V heavy-duty vehicles are
significantly higher than the limit at type approval
[110]. Vehicles equipped with SCR have very low
emission levels during rural and motorway driving, but not under the low engine loads associated
with urban or driving. The reason for this is that
the SCR system needs to attain a certain temperature for efficient NO x reduction. The real-world
performance of SCR should therefore be investigated. Furthermore, more extensive information
on f-NO 2 values for a wider range of vehicles is
required.
Improving Emission Models Accuracy, reliability, consistency, and credibility are prerequisites of emission estimates. However,
comparisons between the results from different
emission models have highlighted substantial differences. There is therefore a need for emission
modelling approaches to be reevaluated, and for
improvements to be made. This requires a better
understanding of the circumstances under which
emission models provide inaccurate predictions.
For example, some modellers believe that the
exhaust emission factors for road vehicles are
unsuitable for certain road types and traffic conditions. Particular concerns have been expressed
about the validity of the emission factors for
so-called congested traffic. The term “congestion”
also needs to be defined in a manner that is meaningful in terms of emissions.
Improving Activity Data While obtaining primary activity data for road transport by direct
measurement remains the main approach for
short-term modelling exercises, it tends to be a
rather expensive activity, and would be unfeasible
where an emission estimate is required for a large
geographical area. Traffic assignment models and
microsimulation tools are increasingly being used
to provide the relevant data, and communications
technology such as Intelligent Transport Systems
and Advanced Transport Telematics enable transport networks to be monitored in real time. It
could well be possible to adapt a number of
models and new technologies to contribute appropriate information on fleets and traffic activity.
In shipping the use of data from AIS and
CEMS has a great deal of potential, and even
more so if it can also be extended to cover inland
shipping. Data from AIS have already been incorporated into at least one model [167]. However,
some work is required to validate emission estimates obtained using AIS data [34]. Research is
also required on the movements of empty ships, as
these are often excluded from emission
estimates [34].
Model Validation Model validation is an important area for improvement. The literature on the
validation of transport emission models is quite
limited, and validation studies are often only partial in the sense that they address just one type of
road, type of vehicle, or time period. The understanding of the uncertainties in emission models
and the main factors affecting prediction errors is
therefore poor [168]. Emission models should
quantify prediction errors. The models in common
Air Quality, Surface Transportation Impacts on
77