31 A High-Resolution National Emission Inventory …
203
31.4 Conclusions
The improvements in gridding implemented for PM 2.5 emissions originated from
machinery had a substantial relative impact on the modeled concentrations and the
resulting population exposure. Although the improvement considered less than half
of the machinery activities (i.e. excluding agriculture, forestry and industry machinery), it decreased the predicted population exposure caused by all machinery by 38%,
calculated over the population of Finland in 2015. Similarly, more drastic overestimation of health impacts due to flawed use of population-based proxy has been
shown also for PM 2.5 emissions originated from residential wood combustion [4].
This paper has shown that it is necessary to use more realistic spatial proxies than
solely the population density for evaluating the emissions originated from different
emission categories.
Questions and answers
Questioner 1: Richard Menard
Question 1: Do you think we can get a quantification of the emission uncertainty as
this would be important for example for top-down emissions?
Answer 1: Quantified uncertainty estimates for emission inventories widely exist
concerning activity data and emission factor uncertainties in different emission source
sectors and different pollutants (e.g. [11, 12]). However, the impact of assumptions
for the spatial disaggregation of emissions on uncertainties is often lacking (e.g.
[14]). When, e.g., a country-level top-down emission inventory is used as a basis for
a city-level air quality assessment modelling, the method of spatial disaggregation
of emissions within the country might increase the emission uncertainty for the area
in question considerably. Thus, the Question 1 in the context of gridded emission
inventories is highly relevant and important. Estimates on the uncertainties of gridded
emission inventories have been presented (e.g. [10]), however, methods to quantify
the uncertainties arising from spatial disaggregation are still being developed by
various inventory reserch teams.
Questioner 2: Volker Matthias
Question 2: How do you know how much wood is consumed in the countryside and
how much in cities [in residential wood stoves]?
A2: The question refers to the gridding procedure for residential wood combustion
emissions used in the Finnish FRES model that uses house locations as a basis for
gridding and takes into account differences in average wood use per household in
different types of houses in different sizes of settlements [13]. The data on average
wood use per household are based on questionnaires.
Acknowledgements This work has been funded by Academy of Finland in the project Environmental impact assessment of airborne particulate matter: the effects of abatement and management
strategies (BATMAN) and NordForsk under the Nordic Programme on Health and Welfare Project
#75007: Understanding the link between air pollution and distribution of related health impacts
and welfare in the Nordic countries (NordicWelfAir).
203
31.4 Conclusions
The improvements in gridding implemented for PM 2.5 emissions originated from
machinery had a substantial relative impact on the modeled concentrations and the
resulting population exposure. Although the improvement considered less than half
of the machinery activities (i.e. excluding agriculture, forestry and industry machinery), it decreased the predicted population exposure caused by all machinery by 38%,
calculated over the population of Finland in 2015. Similarly, more drastic overestimation of health impacts due to flawed use of population-based proxy has been
shown also for PM 2.5 emissions originated from residential wood combustion [4].
This paper has shown that it is necessary to use more realistic spatial proxies than
solely the population density for evaluating the emissions originated from different
emission categories.
Questions and answers
Questioner 1: Richard Menard
Question 1: Do you think we can get a quantification of the emission uncertainty as
this would be important for example for top-down emissions?
Answer 1: Quantified uncertainty estimates for emission inventories widely exist
concerning activity data and emission factor uncertainties in different emission source
sectors and different pollutants (e.g. [11, 12]). However, the impact of assumptions
for the spatial disaggregation of emissions on uncertainties is often lacking (e.g.
[14]). When, e.g., a country-level top-down emission inventory is used as a basis for
a city-level air quality assessment modelling, the method of spatial disaggregation
of emissions within the country might increase the emission uncertainty for the area
in question considerably. Thus, the Question 1 in the context of gridded emission
inventories is highly relevant and important. Estimates on the uncertainties of gridded
emission inventories have been presented (e.g. [10]), however, methods to quantify
the uncertainties arising from spatial disaggregation are still being developed by
various inventory reserch teams.
Questioner 2: Volker Matthias
Question 2: How do you know how much wood is consumed in the countryside and
how much in cities [in residential wood stoves]?
A2: The question refers to the gridding procedure for residential wood combustion
emissions used in the Finnish FRES model that uses house locations as a basis for
gridding and takes into account differences in average wood use per household in
different types of houses in different sizes of settlements [13]. The data on average
wood use per household are based on questionnaires.
Acknowledgements This work has been funded by Academy of Finland in the project Environmental impact assessment of airborne particulate matter: the effects of abatement and management
strategies (BATMAN) and NordForsk under the Nordic Programme on Health and Welfare Project
#75007: Understanding the link between air pollution and distribution of related health impacts
and welfare in the Nordic countries (NordicWelfAir).