10 Modelling the Concentration of Ammonia and Exceedance …
63
in areas dominated by pig and poultry emissions, where emissions reductions were
highest, measurements showed a decrease in NH 3 concentrations.
Questioner Name: Stefano Alessandrini
Q: Do you simulate ammonia emissions one by one? How do you deal with background substances reacting with ammonia in the model? Why don’t you perform the
bias correction site by site? The bias correction might be site specific?
A: Ammonia emissions are distributed spatially by livestock category as well as
fertiliser and non-agricultural emissions. See below regarding the chemical reaction
scheme. There could be a justification to use a spatially variable correction to the
modelled ammonia concentrations, based on geography and climatology or indeed
dominant livestock category, though this hasn’t been checked However the bias for
any two model-measurement points is strongly influenced by the precise location
of the measurement site and the highly variable local ammonia concentration at a
model sub-grid resolution. The bias for all sets of paired model-measurement points
was therefore used collectively to obtain an overall statistically significant bias.
Questioner Name: Ted Russell
Q: In the U.S. ammonia emissions have increased along with a decrease in acidic
precursors, and ammonia levels have stayed relatively constant, similar to your code.
If one looks at the time scale analysis, the deposition of ammonia is rather fast, faster
than the lifetime of ammonia sulphate. This suggests that the ammonia gas levels
would be in a pseudo state between emissions and deposition, buffering any major
change in ammonia gas concentration changes. Does your model have a thermodynamic aerosol module and how does it show ammonia gas changes when sulphate
and nitrate is reduced?
A: In the model H 2 SO 4 and HNO 3 are generated by dry and aqueous phase oxidation reactions. NH 3 reacts instantaneously to form ammonium sulphate in a one-way
reaction whilst the reaction between HNO 3 and NH 3 to form ammonium nitrate is
represented as a two way equilibrium reaction. The sensitivity of ammonia concentration and deposition to changes in NO x and SO 2 emissions has not been calculated
specifically. However on a national scale both the model and measurements broadly
suggest that a reduction of 20% in NH 3 emissions corresponds to no change in NH 3
concentrations whilst SO 2 and NO x emissions have decreased significantly during
this period.
Acknowledgements This work was funded by the UK Natural Environment Research Council and
the UK Department for Environment, Food and Rural Affairs.
63
in areas dominated by pig and poultry emissions, where emissions reductions were
highest, measurements showed a decrease in NH 3 concentrations.
Questioner Name: Stefano Alessandrini
Q: Do you simulate ammonia emissions one by one? How do you deal with background substances reacting with ammonia in the model? Why don’t you perform the
bias correction site by site? The bias correction might be site specific?
A: Ammonia emissions are distributed spatially by livestock category as well as
fertiliser and non-agricultural emissions. See below regarding the chemical reaction
scheme. There could be a justification to use a spatially variable correction to the
modelled ammonia concentrations, based on geography and climatology or indeed
dominant livestock category, though this hasn’t been checked However the bias for
any two model-measurement points is strongly influenced by the precise location
of the measurement site and the highly variable local ammonia concentration at a
model sub-grid resolution. The bias for all sets of paired model-measurement points
was therefore used collectively to obtain an overall statistically significant bias.
Questioner Name: Ted Russell
Q: In the U.S. ammonia emissions have increased along with a decrease in acidic
precursors, and ammonia levels have stayed relatively constant, similar to your code.
If one looks at the time scale analysis, the deposition of ammonia is rather fast, faster
than the lifetime of ammonia sulphate. This suggests that the ammonia gas levels
would be in a pseudo state between emissions and deposition, buffering any major
change in ammonia gas concentration changes. Does your model have a thermodynamic aerosol module and how does it show ammonia gas changes when sulphate
and nitrate is reduced?
A: In the model H 2 SO 4 and HNO 3 are generated by dry and aqueous phase oxidation reactions. NH 3 reacts instantaneously to form ammonium sulphate in a one-way
reaction whilst the reaction between HNO 3 and NH 3 to form ammonium nitrate is
represented as a two way equilibrium reaction. The sensitivity of ammonia concentration and deposition to changes in NO x and SO 2 emissions has not been calculated
specifically. However on a national scale both the model and measurements broadly
suggest that a reduction of 20% in NH 3 emissions corresponds to no change in NH 3
concentrations whilst SO 2 and NO x emissions have decreased significantly during
this period.
Acknowledgements This work was funded by the UK Natural Environment Research Council and
the UK Department for Environment, Food and Rural Affairs.