Go to content

1. Summary

1.1 Background and purpose

The Nordic Working Group for Climate and Air (NKL) under the Nordic Council of Ministers has initiated this project with a specific tender, and Aarhus University, Department of Environmental Science (ENVS) has conducted the research project.
The purpose of the project is to gain a better understanding of the implications of the new WHO guidelines in a Nordic setting. The project also looked at how far Nordic countries currently are from complying with the new 2021-guidelines, and provide a foundation for assessment of measures in the Nordic countries to achieve the new recommended WHO levels. This understanding will also serve to provide a Nordic perspective on the new proposed EU Air Quality Directive from 2022 - both regarding realistic future limit values and to the regulatory approach. The project focuses on three main tasks (1) Projection of air quality for 2030 for Nordic countries and selected cities, (2) Sector specific contribution to air quality in 2030 and health effects and related external costs, and (3) Evaluation of air quality monitoring in Nordic countries based on measurements from 2021 with comparison to WHO guidelines and EU limit values for the Nordic countries as well as for the selected cities, and general implications of the new proposed EU Air Quality Directive for the Nordic countries.

1.2 Methodology and assumptions

The evaluation of predicted air quality in 2030 in relation to the new WHO guidelines from 2021 is focused on a selection of cities in the Nordic countries. Based on the number of inhabitants in the largest cities in the Nordic countries three cities in each country were selected to cover the largest cities, but also to have a good geographical coverage. However, for Iceland only one city is selected but including the greater metropolitan area of Reykjavík. The cities are:
  • Sweden: Stockholm, Göteborg and Malmö;
  • Denmark: København, Aarhus and Odense;
  • Finland: Helsinki, Tampere and Oulu;
  • Norway: Oslo, Bergen and Trondheim;
  • Iceland: Greater Reykjavík (here after just Reykjavík).
To be able to model air quality in 2030 for the Nordic countries as well as the contribution of emission sectors to air quality, Aarhus University has setup and applied the regional scale air quality model (DEHM) and the local scale urban air quality model (UBM) for the selected cities in the Nordic countries. Further, to address health impacts and related external costs we set up and applied the integrated modelling system – EVA (Economic Valuation of Air Pollution). Aarhus University has developed all these models.
Emissions for Europe and the rest of the Northern Hemisphere are based on emission data from the EMEP database with 0.1 x 0.1 degrees resolution and the ECLIPSE v6b database with 0.5 x 0.5 degrees resolution. Furthermore, global ships emissions are based on the STEAM model with 0.1x0.1 degrees spatial resolution and monthly time resolution.
Projected emissions for 2030 are drawn from existing international emission databases. For EU countries, emissions reflect national projections if targets are met in the NEC directive (National Emission Ceilings), otherwise they reflect reduction targets set out in the NEC directive for 2030 for the country. For non-EU-countries, the national projections are used.
Emissions for the Nordic countries are based on the emission dataset from the research project NordicWelfAir (funded by NordForsk) that for the first time established a high-resolution geographically distributed emission inventory of 1 km x 1 km for the Nordic countries for selected years from 1990 to 2014. The emission inventory for 2014 has been scaled to 2019 and 2030 based on the total of the country specific emission sectors and assuming a geographic distribution as in 2014.
The year 2019 serves as a reference year where model results are compared with measurements from monitoring stations in the selected cities to evaluate the performance of the air quality models.
Meteorological data are obtained from the Weather Research and Forecasting model (WRF), operated at Aarhus University.

1.3 Main conclusions and results

1.3.1 Evaluation of air quality monitoring in 2021

WHO AQ guidelines and the EU AQ Directive

WHO has tightened their air quality guidelines from 2005 to 2021 for the three pollutants that pose the largest health burden (PM2.5, NO2 and O3). The WHO Air Quality Guidelines (AQG) are much lower than in the present EU Air Quality Directive, and also lower than the proposed revised EU Air Quality Directive from October 2022. However, the proposed EU Air Quality Directive will eventually align with WHO guidelines for annual PM2.5 (5 µg/m3) and annual NO2 (10 µg/m3) for the average exposure concentration measured at urban background stations. This is to be achieved with continuous requirement for reduction in concentrations until the target is met. In 2030, a 25% reduction has to be met over a ten-year period, and the same the following years until the target is met.

Present air quality in Nordic countries and WHO guidelines

A comparison between measurement data from 2021 (extracted from the EEA database European Air Quality Portal) and the former and new WHO AQG, has been carried out for each of the Nordic countries where the maximum value at any measurement station within the country is compared with the WHO guidelines. This provides an overview of exceedances of the new WHO guidelines in the Nordic countries. Furthermore, it also describes how the exceedances have changed between the former and new WHO guidelines.
The health impacts of air pollution are by far the largest for long-term exposure to PM2.5 then followed by NO2, both as annual means, and then exposure to elevated levels of ozone. All five Nordic countries exceed the 2021 WHO AQG for annual means of PM2.5 and NO2, and also ozone (8h peak season) based on the highest measured values in 2021 (except Iceland for peak ozone as data is not available). More exceedances were observed in 2021 compared with 2005 as the WHO guidelines were tightened from 2005 to 2021.

Present air quality in selected cities

An analysis was carried out based on available measurements of NO2 and PM2.5 in 2021 from rural, urban background and street stations in the selected cities and the new WHO guidelines and the proposed EU Air Quality Directive. This analysis gives an indication of the concentration contribution of the cities (difference between urban background and rural concentrations) and further the contributions of hotspots (difference between street concentrations and urban background concentrations). For each of the selected cities the one station with the highest concentration measured, is representing that particular city.
In general, the measurements of NO2 and PM2.5 in 2021 are following the expected concentration pattern with the highest values represented by the traffic stations and the lowest seen at rural background stations with suburban/urban background in-between.
In all the selected cities in the five Nordic countries, annual NO2 and PM2.5 in 2021 are well below the annual limit values (40 μg/m3 and 25 μg/m3, respectively) of the current EU Air Quality Directive. In relation to the newly proposed EU Air Quality Directive limits of 20 μg/m3 and 10 μg/m3 for NO2 and PM2.5, respectively, 9 out of 13 traffic stations are exceeding the proposed annual limit value for NO2 and only one for PM2.5. None of the urban or rural background stations are exceeding the annual limit values of NO2 and PM2.5 of the newly proposed directive.
When it comes to the new 2021 WHO guidelines for annual NO2 (10 μg/m3), all the traffic stations show higher concentrations than the guideline levels. The same applies to annual PM2.5 (5 μg/m3). For the urban background stations, 6 out of 12 are exceeding the NO2 guideline levels. For PM2.5, 8 out of the 10 are exceeding the guideline levels. For NO2 measured in the rural background, all the stations with measurements are well below the guideline levels. For PM2.5, however, for the rural background stations 3 out of 9 with measurements are exceeding the guideline levels.

General implications of proposed AQ Directive

The overall implications for the Nordic countries of the proposed new EU Air Quality Directive have been analysed.
Measurements
In relation to the present EU Air Quality Directive and the general improvements in the air quality in the Nordic countries over the past 10 years, a comprehensive reorganization would be expected to take place of the air monitoring programmes with cheaper methods and fewer measuring points. However, the new proposal for the EU Air Quality Directive radically changes this possibility, since the tightening of the limit values is accompanied by more stringent assessment thresholds used to determine the number of measuring points and the requirements for the measurement methods.
It is further a requirement to measure ultrafine particles (UFP) and the size distribution of UFP.
There are also increased requirements for documentation of spatial representativeness of measurements and design of the measurement program.
Establishment of supersites
The new proposal for the EU Air Quality Directive requires the establishment of supersites with the aim to increase knowledge about particle pollution at EU level. Supersites require the measurement of a large number of new particle components, of which a requirement to measure the oxidative potential of PM is something completely new in the context of air quality monitoring.
New requirements for average exposure concentrations
The proposed new EU Air Quality Directive introduces a strengthening of the requirements for reducing the average exposure concentration for PM2.5 based on measured urban background concentrations and introduces similar requirements for NO2 that gradually will align with the WHO AQG.
Requirement for air quality modelling
The new proposal for the EU Air Quality Directive includes model calculations as an obligatory element where concentrations of pollutants exceed limit values, or target values. Furthermore, short-term air quality forecasts shall be carried out. Additionally, designated reference model institutions have to be appointed and participate in periodical model reviews and international model intercomparison exercises.
Public information on actual air quality and an Air Quality Index
The proposal for a new air quality directive imposes significantly stricter requirements on information on current air quality including obligatory hourly updates. It is also a requirement to establish an air quality index, and provide short-term air quality forecasts based on modelling.

1.3.2 Regional concentrations compared with the WHO AQG

Nordic emissions in 2019 and 2030

SNAP7 (road transport) contributes the most to NOx emissions in 2019 but decreases towards 2030 where other sectors also contribute significantly. SNAP2 (residential combustion=wood stoves) is an important contributor to PM2.5 emissions, except for Iceland. The combined sectors of SNAP1 (energy), 3 (industrial combustion) and 4 (industrial processes) are the emission sectors that contribute the most to SOx, except for Iceland where it is the combined sector of SNAP5 (extraction etc.), 6 (solvents) and 9 (waste) where SOx from geothermal energy and power production contributes by far the most allocated to SNAP5. The majority of SOx in Iceland is hydrogen sulfide (H2S).
There is a downward trend for emissions for all countries and pollutants from 2019 to 2030.

Regional concentrations in 2019 and 2030

The DEHM model was used to model regional background concentrations in 2019 and 2030 in the model domain of the Nordic countries in order to compare with the WHO AQG from 2021.
For annual NO2, there is a gradient from south to north with higher concentrations in Denmark and Southern Sweden. It is clear that the concentrations in most areas of the domain are lower than the WHO guideline (10 µg/m3), except for a few big cities. Ship emissions are also shown to cause elevated concentrations along ship lines, and at big harbors. There is a general decrease in NO2 concentrations from 2019 to 2030. Five of the selected cities are exceeding the WHO guideline in 2019: København, Trondheim, Stockholm, Oslo and Reykjavík, and only Reykjavík in 2030.
For annual PM2.5, the concentration gradient also shows a decrease from south to north in the domain. There is also a decrease in concentrations from 2019 to 2030. There are significant exceedances of the WHO guidelines (5 µg/m3) in 2019 for Denmark and Southern Sweden, however, these exceedances are much smaller in 2030. Four of the selected cities exceeding the WHO guideline in 2019 were located in Denmark (København, Aarhus, Odense) and Southern Sweden (Malmö) which is reduced to three in 2030 in the same countries, as concentrations in Aarhus no longer exceed the guideline in 2030. In line with other regional scale models, the DEHM model tends to underestimate the mass concentration of PM2.5 due to the so-called “mass-closure problem”.  Probably part of the "missing mass" in the model is water in the particles, which is measured, but not modelled. Also processes and sources that are not fully described in the regional model such as re-suspended dust and mineral dust (to a lesser extent) could be contributing to the measurements, but not included completely in the model.
For peak ozone, the WHO guideline (60 µg/m3) is exceeded in most part of the domain both in 2019 and 2030. Peak ozone concentrations show a very slight decrease towards 2030, especially for Iceland and Finland. Concentrations of ozone are not higher in big cities, but actually lower compared with rural areas, due to the titration effects of NOx where NOx emissions from e.g. traffic convert ozone to NO2. All selected cities are exceeding the WHO guideline in both 2019 and 2030.
A similar analysis as above has been carried out for the pollutants PM10, CO and SO2. All selected cities are below the WHO guideline in 2019 and 2030 for annual PM10 (15 µg/m3) and peak PM10 (45 µg/m3 as annual 99th percentile of 24h-mean). All selected cities are below the WHO guideline for CO in 2019 and 2030 (4 mg/m3 as annual 99th percentile of 24h-mean). All selected cities are below the WHO guideline for peak concentration of SO2 in both 2019 and 2030 (40 µg/m3 as annual 99th percentile of 24h-mean).
As there is uncertainty on model results, there is also uncertainty associated with the above assessment when modelled concentrations are compared with a threshold and stated to be either above or below, especially when predicted concentrations are close to the thresholds.

1.3.3 Urban background concentrations in selected cities

Observed versus model concentrations in 2019

Observed concentrations at urban background stations were compared with modelled urban background concentrations based on DEHM/UBM for 2019.
As expected the model results for the urban background concentrations of PM2.5 and NO2 with UBM are generally higher than the DEHM results due to the higher resolution of UBM of 1 km x 1 km compared with the resolution of DEHM of 5.6 km x 5.6 km. Consequently the model results for O3 are lower with UBM than with DEHM due to previously described reactions with NOx.
The comparison between observations of NO2 at urban background measuring stations and UBM model results in 2019 shows that the model predicts a higher range in concentration levels than shown in the observations and the model tends to overestimate the concentration. The comparison for O3 shows that the model tends to underestimate, which is expected if NO2 is overestimated. The comparison is best for PM2.5 where the UBM modelled concentration levels are in agreement with the observed levels. Taking the mass-closure problem mentioned above into account, this means that the model most like overestimates the PM2.5 concentrations.
The UBM model results show descreasing urban background concentrations from 2019 to 2030, which can be explained by the decreasing urban emissions. Part of the decrease in urban background concentrations is also due to decreasing regional concentrations modelled with DEHM serving as input to UBM from the regional background, due to decreasing emissions from the country in which the city is located as well and from abroad.

Geographic distribution of modelled concentrations over selected cities in 2019 and 2030

The geographic distribution of concentrations in the cities are depending on population density and associated emission density, large road transport corridors as well as ship traffic. Concentrations of NO2 and PM2.5 are decreasing from 2019 to 2030 whereas concentrations of O3 are slightly increasing.

Estimates of urban background concentrations in 2030 versus WHO AQG

Emission projections and model results are subject to uncertainties, and since the UBM model also generally overestimates NO2 concentrations, caution should be taken when comparing modelled concentrations with WHO AQG. Therefore, an estimate of compliance with the WHO AQG in 2030 is based on current observation levels and modelled changes from 2019 to 2030 with focus on the capital cities.
Based on this analysis for NO2 it is likely that the urban background stations in København, Stockholm, Helsinki and Reykjavík will be under the WHO AQG of 10 µg/m3 in 2030. For PM2.5 København may be slightly over the WHO AQG of 5 µg/m3 in 2030 whereas other cities are likely to be below the guideline.
Concentrations in all selected cities are exceeding the WHO guideline in both 2019 (observations, except Reykjavík with no data) and 2030 (model results) for peak ozone. Modelled peak ozone concentrations show a very slight decrease towards 2030.

1.3.4 Sector contributions to air quality in capitals in 2030

Information for each of the Nordic capital cities on the contribution to air quality from different emission sectors given as the contribution from the city, from the country of the city (disregarding emissions from the city) and sources from abroad is provided based on model results. The analysis focuses on 2030 to illustrate the potential benefits of regulation of the different emission sectors in the future. Furthermore, the focus is on PM2.5 and NO2 as they are the largest contributors to health effects.
The contribution to air quality of PM2.5 in 2030 for the Nordic capital cities ranges from 10% to 26% from city emissions, 16% to 23% from country emissions and 51% to 76% from emissions originating abroad. The similar numbers for NO2 are 18% to 35%, 18% to 42% and 23% to 55%, respectively.
København stands out with the highest contribution from abroad and relative low contributions from city and country emissions to air quality of PM2.5 in this case due to its location close to Central Europe.
In the comparison, Reykjavík has a relatively large PM2.5 contribution from abroad despite its location in the North Atlantic Ocean. However, the absolute PM2.5 concentrations are the lowest compared with the other cities. Possible explanations could be the influence of ship emissions, long-range transport and sea salt. Moreover, Reykjavík and Iceland have a relatively small population further adding to the relative importance of emissions from abroad.
The three largest sectors contributing to urban background concentrations of PM2.5 in 2030 from city emissions are road transport (SNAP7), residential wood combustion (SNAP2) and the combined sector of energy, industrial combustion and industrial processes (SNAP134), except for Reykjavík where the contribution from residential wood combustion is insignificant and off-road (SNAP8) plays a larger role due to the fishing fleet. For NO2 the three largest sectors are road transport (SNAP7), the combined sector of energy, industrial combustion and industrial processes (SNAP134) and off-road (SNAP8) but all other sectors with combustion emissions also contribute.
The largest sectors contributing to urban background concentrations of PM2.5 in 2030 from country emissions are residential wood combustion (SNAP2), road transport (SNAP7), the combined sector of energy, industrial combustion and industrial processes (SNAP134) and off-road (SNAP8) but also agriculture (SNAP10). For NO2 the largest sectors are off-road (SNAP8), road transport (SNAP7), and the combined sector of energy, industrial combustion and industrial processes (SNAP134).
The contribution from abroad has not been broken down in emission sectors as it is computationally very demanding and time consuming to analyse results.

1.3.5 Sector contributions to health effects and external costs in capitals in 2030

A summary of the number of predicted premature deaths in the Nordic capital cities due to air pollution is given in Table 1.1.
Key factors in determination of premature mortality are air quality levels of PM2.5, NO2 and O3 and the number of inhabitants exposed, which is also evident from Table 1.1. It is also seen that although populations are expected to grow from 2019 to 2030, premature deaths are predicted to decrease as air quality improves except for Iceland where number of premature deaths is the same in 2019 and 2030 as a combined effect of a relatively high population growth and decreasing pollution levels. Data for 2030 for Oslo are missing due to erroneous geographic distribution of the emissions from oil and gas production for 2030 allocated to land areas.
Table 1.1. Number of premature deaths in capital cities in 2019 and 2030.
 
Area (km2)
Inhabitants in 2019
Inhabitants in 2030
2019
2030
Difference
København
95
594,679
610,810
410
310
-22%
Stockholm
207
1,064,033
1,154,401
620
510
-19%
Helsinki
195
687,693
687,865
390
330
-15%
Reykavik
173
226,661
264,756
39
39
1%
Oslo
262
664,000
697,526
450
n.a.
n.a.
The distribution of premature deaths caused by emissions from the city, from the country and from abroad for the Nordic capital cities is shown in Table 1.2.
The contribution from the city emissions range from 15% to 43%, from emissions in the rest of the country from 18%-27% and from emissions from abroad from 30% to 60%. Reykjavík has the highest contribution from the city and lowest from abroad. The opposite picture is seen for København with the lowest contribution from the city and highest from abroad.
 
From city
From country
From abroad
Total
København
15%
23%
63%
100%
Stockholm
35%
18%
47%
100%
Helsinki
25%
21%
53%
100%
Reykavík
43%
27%
30%
100%
Oslo
n.a.
n.a.
n.a.
n.a.
Table 1.2. Percentage of premature deaths in Nordic capital cities in 2030 distributed on emission contributions from city to city, from country and from abroad
 
2019
2030
Difference
København
1.1
0.9
-21%
Stockholm
1.6
1.3
-20%
Helsinki
1.2
0.9
-16%
Reykavík
  0.14
  0.14
-2%
Oslo
1.2
n.a.
n.a.
Table 1.3. Total external costs for air pollution in capitals in 2019 and 2030 (Billion EUR).
The external costs follow the premature mortality as the costs of morbidity plays a minor role, and hence the distribution on emission sectors and on the contribution from city, country and abroad also follows the number of premature deaths presented earlier.
Nordic Council of Ministers I www.norden.org/publications I pub@norden.org