314
J. Bieser et al.
from agriculture and on-road transport these are the species with the highest expected
difference in the SMOKE-EU emission dataset. For CO and ozone, however we find
only a comparably small impact of the emission dataset. Especially for CO this result
is surprising as the SMOKE-EU dataset uses a distinct spatial and temporal disaggregation of the emissions based on heating demand compared to fixed annual profiles
for the MACC/EDGAR dataset. As expected, the impact of emissions increases near
sources. Thus, with increasing CTM resolution it will become more important to
correctly model emission hot-spots such as traffic and residential emissions [10].
Finally, this study is only based on two CTMs and the findings need to be verified
by a larger model ensemble.
Question and Answers
Questioneer: C. Geels
Q: You talked about changed emissions—but actually it was “only” the temporal
profiles?
A: This is a very important question indeed, thus I will answer it in detail:
(1) For emission from shipping and on-road transport we use bottom-up models
that result in independently calculated annual total emissions. Moreover, we
gain high resolution spatial distribution data. This means the exact location of
each ship at each time step and vehicle road densities which are based on vehicle
counting and traffic demand models. Finally, we use dynamic emission factors
depending on vehicle type, fuel, and temperature. This has, for example, a large
impact on the NO/NO 2 split.
(2) For emission from residential heating we calculate individual temporal profiles
for each grid cell based on a heating demand formula. We then normalize only
the annual total emissions which leads to a spatial redistribution of emissions
to cold regions with a different distribution for each day.
(3) For the agricultural emissions we use a similar approach for emissions from
animal husbandry and fertilizer application. Here, the redistribution is based
on temperature and wind speed. But also legal aspects such as national and
European legislation on fertilizer application are implemented into the model.
Moreover, we use a plant growth model that calculates additional NH 3 emissions
which are not covered by the national totals.
Q: For some components like NH 3 the annual emission might change from year to
year just because of temperature changes. So warm year more NH 3 emissions—do
you think we should include that.
A: We already do that to a certain extend but I think NH 3 emissions exhibit the highest
uncertainty and variability of the standard (criteria) pollutants. Especially because
of the impact on secondary inorganic particle conversion I would strongly support
the development of more sophisticated emission models for agricultural emissions.
J. Bieser et al.
from agriculture and on-road transport these are the species with the highest expected
difference in the SMOKE-EU emission dataset. For CO and ozone, however we find
only a comparably small impact of the emission dataset. Especially for CO this result
is surprising as the SMOKE-EU dataset uses a distinct spatial and temporal disaggregation of the emissions based on heating demand compared to fixed annual profiles
for the MACC/EDGAR dataset. As expected, the impact of emissions increases near
sources. Thus, with increasing CTM resolution it will become more important to
correctly model emission hot-spots such as traffic and residential emissions [10].
Finally, this study is only based on two CTMs and the findings need to be verified
by a larger model ensemble.
Question and Answers
Questioneer: C. Geels
Q: You talked about changed emissions—but actually it was “only” the temporal
profiles?
A: This is a very important question indeed, thus I will answer it in detail:
(1) For emission from shipping and on-road transport we use bottom-up models
that result in independently calculated annual total emissions. Moreover, we
gain high resolution spatial distribution data. This means the exact location of
each ship at each time step and vehicle road densities which are based on vehicle
counting and traffic demand models. Finally, we use dynamic emission factors
depending on vehicle type, fuel, and temperature. This has, for example, a large
impact on the NO/NO 2 split.
(2) For emission from residential heating we calculate individual temporal profiles
for each grid cell based on a heating demand formula. We then normalize only
the annual total emissions which leads to a spatial redistribution of emissions
to cold regions with a different distribution for each day.
(3) For the agricultural emissions we use a similar approach for emissions from
animal husbandry and fertilizer application. Here, the redistribution is based
on temperature and wind speed. But also legal aspects such as national and
European legislation on fertilizer application are implemented into the model.
Moreover, we use a plant growth model that calculates additional NH 3 emissions
which are not covered by the national totals.
Q: For some components like NH 3 the annual emission might change from year to
year just because of temperature changes. So warm year more NH 3 emissions—do
you think we should include that.
A: We already do that to a certain extend but I think NH 3 emissions exhibit the highest
uncertainty and variability of the standard (criteria) pollutants. Especially because
of the impact on secondary inorganic particle conversion I would strongly support
the development of more sophisticated emission models for agricultural emissions.