Attributable deaths in Austria due to ozone under different climate scenarios

Hanns Moshammer; Monika Mayer; Harald Rieder; Christian Schmidt; Birgit Bednar-Friedl; Peter Wallner; Hans-Peter Hutter

doi:10.1093/eurpub/ckae126

. 2024 Aug 10;34(5):1015–1020. doi: 10.1093/eurpub/ckae126

Attributable deaths in Austria due to ozone under different climate scenarios

Hanns Moshammer ^1,^2,^✉, Monika Mayer ³, Harald Rieder ⁴, Christian Schmidt ⁵, Birgit Bednar-Friedl ⁶, Peter Wallner ⁷, Hans-Peter Hutter ⁸

PMCID: PMC11430921 PMID: 39127464

Abstract

Tropospheric ozone is an air pollutant that poses a public health problem in Europe. Climate change could increase the formation of ozone. Applying past and predicted annual total (all-cause) mortality data and modeled daily ozone concentrations, we performed a nationwide health impact assessment estimating annual ozone-related (attributable) deaths in Austria. Different approaches were compared. Estimates were based on maximal 1-h averages of ozone. Until the decade from 2045 till 2055, more people will die in Austria because of the demographic trends. Therefore, more deaths will also be attributable to ozone. Higher greenhouse gas emission scenarios (e.g. Representative Concentration Pathway RCP8.5 compared to RCP2.6) will lead to more ozone-related deaths, mostly due to the national emission of ozone precursors (a difference of 250–340 cases per year, depending on the model), but to a lesser extent because of global climate change. Increases in attributable deaths will be affected mostly by national, not global mitigation measures. National emission reduction will certainly have a strong and beneficial effect on local atmospheric chemistry, air quality, and public health.

Introduction

Tropospheric ozone (O₃) is a cytotoxic oxidizing substance and a secondary air pollutant which is formed in the atmosphere. In detail, the ozone forming process is described by Mayer et al.:¹ “In surface air, O₃ is produced through a series of chemical reactions involving sunlight and O₃ precursors such as nitrogen oxides (NO_X: NO and NO₂), non-methane volatile organic compounds (NMVOCs), carbon monoxide (CO) and methane (CH₄). These precursors are emitted from anthropogenic (e.g. traffic or industry) and natural (e.g. soil, trees, lightning) sources. Ozone production rates depend strongly on the chemical environment (i.e. NO_X or VOC limitation) and ambient meteorological conditions (i.e. solar radiation, temperature, and humidity).”

In the EU-27 in 2021 22 000 deaths were attributable to ozone (short-term exposure to concentrations above 70 µg/m³), as the European Environment Agency states.² Climate change could increase, inter alia, the formation of secondary pollutants, such as ozone.³

Within the framework of a nationally funded project (ATtain-O₃⁴), we performed a health impact assessment in regard to ozone exposures under different climate change and emission scenarios and considered demographic projections. This raised a variety of questions like: which effect estimates should be used for the Austrian HIA? Which cut-off for the ozone effect estimates should be chosen? And: which ozone metric should be used?

Methods

Air pollution data

In short, ozone concentrations were obtained from simulations of a regional chemistry-transport model, WRF-Chem.^5–8 All simulations have been performed with a spatial resolution of 9 km × 9 km. Anthropogenic emissions of NOx, C3O, SO₂, PM, and NMVOCs have been taken for the baseline simulation from TNO MACC-III data.⁹ Following initial model sensitivity analyses and evaluations, Austrian traffic NOx emissions over the historical reference period have been increased on average by a factor of 1.5 to match model burdens to observed concentrations. Future NOx emissions were obtained by applying NOx reduction paths to these adjusted NOx emissions, with reductions specific to individual RCPs (Representative Concentration Pathways), as defined by the International Panel of Climate Change (IPCC). In addition to the 10-year time slice simulations spanning three RCPs with a radiative forcing of 2.6, 4.5, and 8.5 W/m² (RCP2.6, RCP4.5, and RCP8.5) over 2026–35 and 2046–55, a sensitivity simulation focusing on the effect of ambitious national emission reductions in analogy to RCP2.6 but global developments following RCP8.5 was performed. The RCP2.6 NOx emissions scenario is in good agreement with the national emissions estimate of the Environment Agency Austria.¹⁰

The modeling approach finally produced daily maximum 1- and 8-h concentrations averaged over each Austrian municipality. Since mortality data were only available on the district and not the municipality level, population-weighted averages per district had to be calculated based on 2019 population numbers.¹¹

Mortality data

For the historical period (2007–16), daily total (all-cause) death counts per district were available per age group, sex, and main cause of death. Already for previous studies,¹²^,¹³ these data had been purchased from Statistics Austria.

For future mortality data, the demographic forecast of Statistics Austria¹⁴ was used. This demographic forecast is based on the extrapolation of current trends. Among others, it presents estimated annual deaths (all causes combined) per district for every 5 years until 2075.

Effect estimates

We decided to use the arithmetic means of the summer and winter estimates of the random-effect model controlled for PM10 according to the European Gryparis et al.’s study¹⁵ (Table 1). The point estimates of the fixed-effect models were similar, although with a smaller confidence interval.

Table 1.

Effect estimates per season according to Gryparis et al.,¹⁵ adjusted for PM10, random-effect model: percent change per 10 µg/m³, point estimates, and 95% confidence interval

	Summer	Winter	Arithmetic mean
1-h average	0.27 (0.10 to 0.47)	0.21 (−0.07 to 0.48)	0.24 (0.02 to 0.47)
8-h average	0.27 (0.08 to 0.49)	0.22 (−0.08 to 0.51)	0.24 (0.00 to 0.50)

Open in a new tab

The arithmetic mean was calculated from the seasonal values and used in the HIA.

Procedure

Calculations were performed using the HIA short-term spreadsheet provided by the APHEKOM project.¹⁶ This project has set the standards for HIA about 10 years ago in Europe.¹⁷ The spreadsheet allows the importation of daily data of pollutants concentrations for up to 3 years, the importation of annual case numbers (e.g. total number of deaths), and the definition of effect estimates and the cut-off concentration.

We set cut-off levels according to the HRAPIE proposal¹⁸ to 35 ppb (= 70 µg/m³) and 10 ppb.

Predictions of annual deaths per district¹⁴ were only available for every fifth year. Averaging across three consecutive years would reduce the chance of year-by-year variation in meteorology and mortality. Therefore, the option in the tool to lump together 3 years was used (2007–09, 2010–12, 2013–15 for the historical period, for the future periods, total deaths for 2025 were paired with ozone estimates from 2026 to 2028, for 2030 with 2029–31, and for 2035 with 2032–35, and likewise for the 2046–55 decennium).

In order to distinguish between the effects of ozone and demography, historical ozone levels were first applied to historical and future populations (Fig. 1), and next historical and future ozone levels were applied to predicted future populations (Figs 2 and 3).

Figure 1. — Number of deaths attributable to ozone assuming ozone levels of the years 2007–09 until 2013–15, effect of demographic trends. See text for details.

Figure 2. — Number of deaths attributable to ozone assuming ozone levels under different emission scenarios, using predicted number of total deaths for the years 2025, 2030, and 2035. See text for details.

Figure 3. — Number of deaths attributable to ozone assuming ozone levels under different emission scenarios, predicted number of total deaths for the years 2045, 2050, and 2055. See text for details.

Results

Demographic trends and ozone concentrations

Austria, in 2019, had 8 894 380 inhabitants. It consists of nine federal countries with a population between 134 872 (Vorarlberg) and 1 908 104 (Vienna). Each federal country consists of 4 (Vorarlberg) to 24 (Lower Austria) districts. Districts are either single (larger) towns or several (smaller) municipalities. The latter situation is more prevalent in less densely populated rural areas where districts can span a quite large and diverse area. For example, the district of Vöcklabruck (Upper Austria) consists of 52 or Innsbruck Land (Tyrol) even of 65 municipalities dispersed over lower and higher altitudes.

Between 2007 and 2016, in Austria 73 994–83 073 deaths occurred each year. According to the demographic forecast,¹⁴ the Austrian population will continue to grow and get older on average. For 2050 till 2060, a total population of 9 677 467–9 763 789 and annual deaths of 102 936–105 994 are predicted.

The spatial patterns of O₃ concentrations remained fairly stable across the years. This means that a district or a municipality with a higher (average) concentration in one year tended to have higher concentrations also in the other years. The correlation coefficient for annual averages between the years 2007 and 2016 was between 0.8986 and 0.9954 when analyzing the concentrations per municipality. During the historical period, there was some fluctuation between years, but no consistent temporal trend. Under future higher emission scenarios, there was a trend to higher ozone concentrations. More or less unchanged ozone concentrations were predicted under low-emission scenarios.

Attributable deaths under different climate change scenarios

Over the years, with an increasing and aging population, also attributable deaths will increase. This is demonstrated in Fig. 1, which depicts attributable deaths assuming the predicted demographic changes, but historical climate scenarios and hence the same ozone concentrations as in the historical period. Scenarios were calculated using maximum 1-h average concentrations and effect estimates.

Figure 2 compares the health impact of ozone concentrations for the historical years 2007–09 with concentrations for 2026–28 assuming an RCP4.5 and an RCP8.5 scenario with the demographic predictions for the year 2025; 2010–12 with 2029–31 concentrations for the year 2030 population; and concentrations 2013–15 with 2032–34 for demographic predictions for the year 2035 (Fig. 2). Similarly, for the decade around the year 2050, Fig. 3 pairs 2007–09 and 2046–48 concentrations with the year 2045, 2010–12, and 2049–51 concentrations with the year 2050, and 2013–15 and 2052–54 concentrations with the year 2055 (Fig. 3). Thus, it shows the comparison between the historical climate and emission scenario and the RCP4.5 and RCP8.5 scenarios. For the decade around the year 2050, for one representative year (ozone data from 2051, mortality data from 2050), also the health impact of a global business as usual climate path (RCP8.5) and national low emissions (in accordance with RCP2.6) is depicted in Fig. 3.

Higher ozone levels were modeled with higher global emission rates leading to higher numbers of attributable deaths. In the decade around the year 2030 (Fig. 2), the effect of climate change or of different emission scenarios was higher for a cut-off at 20 µg/m³ than for the usual cut-off at 70 µg/m³. For example, the RCP8.5 in comparison to current climate and ozone levels, caused an increase between 58 and 198 deaths per year based on a cut-off of 20 µg/m³, compared to −13 to 100 deaths per year based on the 70 µg/m³ cut-off.

Differences between the scenarios were less pronounced in the decade around the year 2050 (Fig. 3). Still, the RCP8.5 scenario led to the highest ozone levels and to the highest number of attributable deaths. Local emission reduction (only assessed for the year 2050/51) led to a substantial reduction of ozone and attributable deaths, even assuming a global high-emission scenario. In these models, using the 70 µg/m³ cut-off, low national emissions would reduce the number of annual deaths by about 250, and using the 20 µg/m³ cut-off the reduction would even amount to about 340 cases.

Discussion

This health impact assessment dealt with ozone-attributable deaths in Austria under different climate scenarios. Due to the demographic trend, an increase of deaths and therefore also of attributable deaths is to be expected. In the future, higher emission scenarios, especially RCP8.5 in the decade around 2050, will lead to considerably larger numbers of attributable deaths compared to a low-emission scenario. For that effect, national emissions of ozone precursor substances are far more relevant than global emissions and global climate change.

We used effect estimates from a European study,¹⁵ and as metric we used the average maximum 1-h concentration as in the APHENA study on the effects of ozone in 125 cities in Europe and North America.¹⁹ APHENA indicates differences in atmospheric chemical processes between the two continents and therefore supports the use of effect estimates from European studies for ozone HIAs in Europe. This is likely the reason why the HRAPIE risk coefficients¹⁸ are based on data from 32 European cities of the APHENA study.

In a previous time-series analysis of data from Vienna spanning the years 1991 till 2009,²⁰ we found that within-day variability of ozone and same-day peak levels (hourly maximum) showed the best correlation with daily deaths. We suggested that peak levels of ozone not only act directly by causing oxidative stress in the airways, but that atmospheric conditions leading to a faster increase in ozone concentrations and thus to higher peak levels also are detrimental to health. Peak levels of ozone and within-day variation in that sense would also serve as proxy indicators of a more complex pollution mixture.

HIAs usually report the confidence interval inherited from the effect estimate alone. This is not precisely correct. In addition to the uncertainty stemming from the original study, there is the uncertainty from applying this estimate to a different setting. For example, the paper by Gryparis et al.¹⁵ clearly demonstrates substantial heterogeneity between cities. The concentration of a single air pollutant in an epidemiological study always also represents part of a much more complex pollution mixture that is bound to differ between regions in its composition. Next, there is the uncertainty regarding the exposure assessment which in the current HIA is based on atmospheric models. Lastly, there is the uncertainty regarding future demographic developments. All these uncertainties if not more should be included in the final estimate of attributable deaths. Therefore, we decided not to report confidence intervals as they would be biased or at least imperfect anyway. It suffices to remember that numbers of attributable deaths only provide an order of magnitude or tendencies, and are not exact calculations.

Internally the APHEKOM tool assumes that on every day of the year the same number of people has died, calculates for every day the attributable numbers and adds these numbers up over the whole year. In the case of ozone with higher concentrations usually in summer, when daily mortality is lower than in winter, this leads to an overestimation of ozone impact. Due to the erroneously even distribution of deaths across the year, half of the deaths would be counted in summer and because of the higher ozone levels a higher number of attributable cases would be estimated. This is not just a short-coming of the APHEKOM tool. Also the WHO HIA tool AirQ+²¹ faces the same problem. Daily deaths are necessary for time-series analyses. For a HIA often only annual case numbers are available or, as in the case of future scenarios, are the only data that are meaningful. As we have shown for Vienna,²² although some seasonality still exists for mortality, this seasonal variation has much decreased in the last decennia and is likely to decrease further with increasing frequency and intensity of heat-waves in summer.

With future climate scenarios, higher ozone levels will also occur in spring and fall. Effect estimates based on summer or “peak season” (6 months) ozone levels alone will therefore not suffice. In our study, we used the arithmetic means of the summer and winter effect estimates.

In its 2020 report presenting figures for the year 2018,²³ EEA estimates for Austria 6100 premature deaths due to PM2.5, 790 due to NO₂, and only 420 due to O₃. Since 2022 EEA has used the concentration–response functions from the WHO Air Quality Guidelines.²⁴ In the country fact sheet from November 2023 the numbers are as follows: PM2.5: 3200 attributable deaths, NO₂: 830, and O₃: 470.²⁵ These figures clearly convey the impression that O₃ is much less important than PM2.5. Because of at least two reasons, this impression is biased: (i) For O₃ effect estimates, a cut-off of 70 µg/m³ is chosen, while for PM2.5, no cut-off or (since 2022) 5 µg/m³ is assumed. When we used a cut-off of 20 µg/m³ for ozone, numbers were much higher. (ii) While effect estimates for other pollutants are based on long-term effects, ozone effects are calculated for short-term effects usually. This seems still to be the case with the latest EEA estimates.²⁵

Certainly, there are fewer studies on O₃ effects than on the health impact and mortality risk of particulate matter. But as for particulate matter, existing studies do not indicate a clear threshold below which no adverse health effects occur. As with particulate matter, effect estimates at very low concentrations are unclear because of a lack of data. At least an additional cut-off of 20 µg/m³ should be considered.¹⁸

Huangfu and Atkinson²⁶ gave the cohort studies on ozone-related mortality a “low” to “moderate” rating, based on too few studies, inconsistent results between studies, and a lack of analyses of the shape of the dose-response curve. The evidence was upgraded to “moderate” by WHO²⁴ in the 2021 global air quality guidelines. Still, also these guideline values for chronic O₃ exposure are based on studies examining ozone concentrations during peak months. This might be sufficient for the definition of guideline values, but for estimating changes in health impact due to climate change that will rather prolong the ozone season than increase average summer values, these studies are not very useful. There is an urgent need for European cohort studies examining different exposure metrics for O₃. As shown with other pollutants, acute effects of day-to-day variation in exposures are only the tip of the iceberg, compared to long-term effects that capture health outcome differences between populations exposed to different annual concentration averages.

This is not the first HIA estimating attributable deaths due to climate-change-induced changes in ozone concentrations.³ These studies were usually done on a larger scale providing global or European projections. They often focused on the impacts of climate change leaving everything else equal.²⁷ Assuming constant demography is fine, when the aim of the study is only to demonstrate the impact of climate change. But climate change models depend on assumptions concerning (local and global) emissions. Neglecting the effect of changes in emissions on ozone concentrations by only focusing on changes in meteorological parameters will certainly not provide the full picture. Besides, while global or European models in general assume higher ozone concentrations with rising temperatures, this trend is much more pronounced in Southern Europe and even reversed in Northern Europe.²⁷ This is mostly due to changing precipitation patterns and these changes differ much more depending on the climate model applied than the changes in temperature. Depending on the climate model, the predictions for ozone concentrations especially for the alpine region and therefore also for Austria differed widely in the European assessment.²⁷

The exposure data for this HIA were derived from chemical-meteorological models. Different models were applied that, in general, produced similar concentration estimates that in the historical period (2007–16) were overall in good accordance with monitoring data. Still, especially in and near urban centers, the original models tended to underestimate concentrations and necessitated a rule-of-thump adaptation of NO_X emission estimates. Better emission data are therefore still necessary.

With regard to O₃ concentrations, (local) emission reduction is more effective than (global) climate change mitigation: this is already true for average O₃ concentrations, but even more so for peak concentrations. This is why an HIA using 70 µg/m³ as a cut-off value, estimates similar differences in attributable deaths between a high-emission scenario (RCP8.5) and a low-emission scenario (RCP2.6) as when using 20 µg/m³ as the cut-off. The mean emission scenario (RCP4.5), in the near future (around the year 2030), compared to current ozone concentrations, will lead to similar or even lower numbers of attributable deaths when a cut-off of 70 µg/m³ is considered and to higher numbers than today with a cut-off of 20 µg/m³. This is mostly due to the fact that in spring and fall concentrations that are currently still mostly below 20 µg/m³ will in future exceed this cut-off, while peak levels especially in summer will tend to decline.¹^,²⁸ As ozone peak concentrations (or even within-day variability of concentrations)²⁰ might be more health-predictive than average concentrations, this decline in peak concentrations is especially important and noteworthy. Because local emission reduction has only limited overall impact on global climate change, additional local benefits must be highlighted in order to strengthen the resolve for climate change mitigation action.

While the dispersion models originally had a resolution of 9 km × 9 km, concentrations were first averaged across municipalities, but mortality data were only available at the district levels. Therefore, concentrations were further averaged (population number weighted) per district. When O₃, all else considered equal, does increase mortality rates or reduces life expectancy, then averaging over larger areas leads to an underestimation of attributable numbers of deaths. Indeed, this could be demonstrated by further averaging district concentrations across federal countries (data not shown). The effect was not large, but consistent across groups of years and across federal countries.

The demographic model predicts further within-country migration from alpine and rural districts into the urban centers. Overall, average O₃ levels in urban centers are lower than in the rural and alpine districts. Therefore, for temporal comparisons of the health effects of O₃, it is even more important to clarify which metric of O₃, average or peak levels, is the better predictor of health effects.

In conclusion, the health impacts of ozone are likely underestimated. Nevertheless, the changes in effect due to concentration changes can still be estimated with sufficient accuracy. Climate change will likely lead to increasing concentrations of ozone in Austria. Scenarios with a higher emission path (RCP8.5) will lead to a more pronounced increase in attributable deaths. Still, the changes might be small compared to increases due to demographic changes.

Increases in attributable deaths will be affected mostly by national, not global mitigation measures. National emission reduction will certainly have a strong and beneficial effect on local atmospheric chemistry, air quality, and public health.

Contributor Information

Hanns Moshammer, Department of Environmental Health, Center for Public Health, Medical University Vienna, Vienna, Austria; Department of Hygiene, Medical University of Karakalpakstan, Nukus, Uzbekistan.

Monika Mayer, Institute of Meteorology and Climatology, Department of Water, Atmosphere, and Environment (WAU), BOKU University, Vienna, Austria.

Harald Rieder, Institute of Meteorology and Climatology, Department of Water, Atmosphere, and Environment (WAU), BOKU University, Vienna, Austria.

Christian Schmidt, Institute of Meteorology and Climatology, Department of Water, Atmosphere, and Environment (WAU), BOKU University, Vienna, Austria.

Birgit Bednar-Friedl, Department of Economics, University of Graz, Graz, Austria.

Peter Wallner, Department of Environmental Health, Center for Public Health, Medical University Vienna, Vienna, Austria.

Hans-Peter Hutter, Department of Environmental Health, Center for Public Health, Medical University Vienna, Vienna, Austria.

Conflict of interest

None declared.

Funding

This research was funded by the Austrian “Klima- und Energy-Fonds” [Climate and Energy Funds], grant number KR18AC0K14686.

Data availability

The mortality data underlying this article are publicly available and the sources were provided. A paper describing the models calculating the ozone concentration is under preparation. We must refer to this paper regarding ozone data.

Key points.

Health impact of ground-level ozone is still underrepresented in the public debate.
With climate change, i.e. higher temperatures and longer sunny periods, ozone levels will likely increase in the future.
Ozone levels will especially increase in spring and fall, while past effect estimates mostly focused on the effect of summer time ozone levels.
Local emissions of ozone precursors are predicted to be more relevant for local ground-level ozone concentrations than the impact of global climate change.
If Austria follows a high-emission instead of a low-emission pathway, this would cause 250–340 more premature annual deaths due to ozone exposure, independently from global emission trends.

References

1. Mayer M, Schreier S, Spangl W et al. An analysis of 30 years of surface ozone concentrations in Austria: Temporal evolution, changes in precursor emissions and chemical regimes, temperature dependence, and lessons for the future. Environ Sci Atmos 2022;2:601–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. European Environment Agency EEA. Harm to Human Health From Air Pollution in Europe: Burden of Disease 2023. https://www.eea.europa.eu/publications/harm-to-human-health-from-air-pollution (25 April 2024, date last accessed).
3. Orru H, Ebi KL, Forsberg B. The interplay of climate change and air pollution on health. Curr Environ Health Rep 2017;4:504–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. ATain-O3. Evaluating the Effects of Climate Warming and Precursor Emission Changes on the Attainment of the Austrian Ozone Standard. Publizierbarer Endbericht im Rahmen von ACRP (Austrian Climate Research Program): Climate and Energy Fund, Vienna, Austria, 2023.
5. Grell GA, Peckham SE, Schmitz R et al. Fully coupled ‘online’ chemistry in the WRF model. Atmos Environ 2005;39:6957–76. [Google Scholar]
6. Fast JD, Gustafson WI Jr, Easter RC et al. Evolution of ozone, particulates, and aerosol direct forcing in an urban area using a new fully-coupled meteorology, chemistry, and aerosol model. J Geophys Res 2006;111:D21305. [Google Scholar]
7. Peckham S, Grell GA, McKeen SA et al. WRF-Chem version 3.3 user’s guide. NOAA Technical Memo 2011;98. [Google Scholar]
8. Powers JG, Klemp JB, Skamarock WC et al. The weather research and forecasting model: overview, system efforts, and future directions. Bull Am Meteorol Soci 2017;98:1717–37. [Google Scholar]
9. Kuenen JJP, Visschedijk AJH, Jozwicka M et al. TNO-MACC_II emission inventory: a multi-year (2003–2009) consistent high-resolution European emission inventory for air quality modelling. Atmos Chem Phys 2014;14:10963–76. [Google Scholar]
10. Anderl M, Haider S, Krutzler T et al. Air Emission Projection 2021 for 2020, 2025 and 2030. Vienna, Austria: Environment Agency Austria, 2021. REPORT REP-0769.
11. Statistics Austria. Gemeindeergebnisse der Abgestimmten Erwerbsstatistik und Arbeitsstättenzählung ab, 2011. (Gebietsstand jeweils zum Stichtag 31.10.). https://www.data.gv.at/katalog/dataset/stat_gemeindeergebnisse-der-abgestimmten-erwerbsstatistik-und-arbeitsstattenzahlung-ab-20-31-10#resources (25 March 2024, date last accessed).
12. Poteser M, Moshammer H. Daylight saving time transitions: impact on total mortality. Int J Environ Res Public Health 2020;17:1611. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Moshammer H, Poteser M, Kundi M et al. Nitrogen-dioxide remains a valid air quality indicator. Int J Environ Res Public Health 2020;17:3733. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Statistics Austria. Kleinräumige Bevölkerungsprognose für Österreich 2018 bis 2040 mit einer Projektion bis 2060 und Modellfortschreibung bis 2075 (ÖROK-Prognose) [Small-Scale Population Forecast For Austria From 2018 to 2040 with a Projection up to 2060 and Model update up to 2075 (ÖROK Forecast)]. January 2019. https://www.oerok.gv.at/fileadmin/user_upload/Bilder/2.Reiter-Raum_u._Region/2.Daten_und_Grundlagen/Bevoelkerungsprognosen/Prognose_2018/Bericht_BevPrognose_2018.pdf (25 March 2024, date last accessed).
15. Gryparis A, Forsberg B, Katsouyanni K et al. Acute effects of ozone on mortality from the “air pollution and Health: a European approach”; project. Am J Respir Crit Care Med 2004;170:1080–7. [DOI] [PubMed] [Google Scholar]
16.APHEKOM publications, HIA Tools. HIA Tool Short Term (rev. Sept 2013) and HIA Tool Long Term (rev. Jan 2016). http://aphekom.org/web/aphekom.org/publications (25 March 2024, date last accessed).
17. Perez L, Declercq C, Iñiguez C et al. Chronic burden of near-roadway traffic pollution in 10 European cities (APHEKOM network). Eur Respir J 2013;42:594–605. [DOI] [PubMed] [Google Scholar]
18.WHO Europe. Health Risks of Air Pollution in Europe—HRAPIE Project. Recommendations for Concentration–Response Functions for Cost–Benefit Analysis of Particulate Matter, Ozone and Nitrogen Dioxide. Copenhagen, 2013. https://iris.who.int/handle/10665/153692?show=full (24 March 2024, date last accessed).
19. Katsouyanni K, Samet JM, Anderson HR et al. HEI Health Review Committee. Air pollution and health: a European and North American approach (APHENA). Res Rep Health Eff Inst 2009;142:5–90. [PubMed] [Google Scholar]
20. Moshammer H, Hutter HP, Kundi M. Which metric of ambient ozone to predict daily mortality? Atmos Environ 2013;65:171–6. [Google Scholar]
21.WHO Europe. AirQ+: Software Tool For Health Risk Assessment of Air Pollution. https://www.who.int/europe/tools-and-toolkits/airq—software-tool-for-health-risk-assessment-of-air-pollution (25 March 2024, date last accessed).
22. Weitensfelder L, Moshammer H. Evidence of adaptation to increasing temperatures. Int J Environ Res Public Health 2019;17:97. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. EEA. Air Quality in Europe 2020. https://www.eea.europa.eu/publications/air-quality-in-europe-2020-report (24 March 2024, date last accessed).
24. WHO. WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide. Geneva: WHO, 2021. [PubMed]
25. EEA. Austria—Air Pollution Country Fact Sheet. https://www.eea.europa.eu/themes/air/country-fact-sheets/2023-country-fact-sheets/austria-air-pollution-country (25 March 2024, date last accessed).
26. Huangfu P, Atkinson R. Long-term exposure to NO₂ and O₃ and all-cause and respiratory mortality: a systematic review and meta-analysis. Environ Int 2020;144:105998. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Orru H, Andersson C, Ebi KL et al. Impact of climate change on ozone-related mortality and morbidity in Europe. Eur Respir J 2013;41:285–94. [DOI] [PubMed] [Google Scholar]
28. Rieder HE, Fiore AM, Clifton OE et al. Combining model projections with site-level observations to estimate changes in distributions and seasonality of ozone in surface air over the U.S.A. Atmos Environ 2018;193:302–15. 10.1016/j.atmosenv.2018.07.042. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Key points.

Health impact of ground-level ozone is still underrepresented in the public debate.
With climate change, i.e. higher temperatures and longer sunny periods, ozone levels will likely increase in the future.
Ozone levels will especially increase in spring and fall, while past effect estimates mostly focused on the effect of summer time ozone levels.
Local emissions of ozone precursors are predicted to be more relevant for local ground-level ozone concentrations than the impact of global climate change.
If Austria follows a high-emission instead of a low-emission pathway, this would cause 250–340 more premature annual deaths due to ozone exposure, independently from global emission trends.

[ckae126-B1] 1. Mayer M, Schreier S, Spangl W et al. An analysis of 30 years of surface ozone concentrations in Austria: Temporal evolution, changes in precursor emissions and chemical regimes, temperature dependence, and lessons for the future. Environ Sci Atmos 2022;2:601–15. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ckae126-B2] 2. European Environment Agency EEA. Harm to Human Health From Air Pollution in Europe: Burden of Disease 2023. https://www.eea.europa.eu/publications/harm-to-human-health-from-air-pollution (25 April 2024, date last accessed).

[ckae126-B3] 3. Orru H, Ebi KL, Forsberg B. The interplay of climate change and air pollution on health. Curr Environ Health Rep 2017;4:504–13. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ckae126-B4] 4. ATain-O3. Evaluating the Effects of Climate Warming and Precursor Emission Changes on the Attainment of the Austrian Ozone Standard. Publizierbarer Endbericht im Rahmen von ACRP (Austrian Climate Research Program): Climate and Energy Fund, Vienna, Austria, 2023.

[ckae126-B5] 5. Grell GA, Peckham SE, Schmitz R et al. Fully coupled ‘online’ chemistry in the WRF model. Atmos Environ 2005;39:6957–76. [Google Scholar]

[ckae126-B6] 6. Fast JD, Gustafson WI Jr, Easter RC et al. Evolution of ozone, particulates, and aerosol direct forcing in an urban area using a new fully-coupled meteorology, chemistry, and aerosol model. J Geophys Res 2006;111:D21305. [Google Scholar]

[ckae126-B7] 7. Peckham S, Grell GA, McKeen SA et al. WRF-Chem version 3.3 user’s guide. NOAA Technical Memo 2011;98. [Google Scholar]

[ckae126-B8] 8. Powers JG, Klemp JB, Skamarock WC et al. The weather research and forecasting model: overview, system efforts, and future directions. Bull Am Meteorol Soci 2017;98:1717–37. [Google Scholar]

[ckae126-B9] 9. Kuenen JJP, Visschedijk AJH, Jozwicka M et al. TNO-MACC_II emission inventory: a multi-year (2003–2009) consistent high-resolution European emission inventory for air quality modelling. Atmos Chem Phys 2014;14:10963–76. [Google Scholar]

[ckae126-B10] 10. Anderl M, Haider S, Krutzler T et al. Air Emission Projection 2021 for 2020, 2025 and 2030. Vienna, Austria: Environment Agency Austria, 2021. REPORT REP-0769.

[ckae126-B11] 11. Statistics Austria. Gemeindeergebnisse der Abgestimmten Erwerbsstatistik und Arbeitsstättenzählung ab, 2011. (Gebietsstand jeweils zum Stichtag 31.10.). https://www.data.gv.at/katalog/dataset/stat_gemeindeergebnisse-der-abgestimmten-erwerbsstatistik-und-arbeitsstattenzahlung-ab-20-31-10#resources (25 March 2024, date last accessed).

[ckae126-B12] 12. Poteser M, Moshammer H. Daylight saving time transitions: impact on total mortality. Int J Environ Res Public Health 2020;17:1611. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ckae126-B13] 13. Moshammer H, Poteser M, Kundi M et al. Nitrogen-dioxide remains a valid air quality indicator. Int J Environ Res Public Health 2020;17:3733. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ckae126-B14] 14. Statistics Austria. Kleinräumige Bevölkerungsprognose für Österreich 2018 bis 2040 mit einer Projektion bis 2060 und Modellfortschreibung bis 2075 (ÖROK-Prognose) [Small-Scale Population Forecast For Austria From 2018 to 2040 with a Projection up to 2060 and Model update up to 2075 (ÖROK Forecast)]. January 2019. https://www.oerok.gv.at/fileadmin/user_upload/Bilder/2.Reiter-Raum_u._Region/2.Daten_und_Grundlagen/Bevoelkerungsprognosen/Prognose_2018/Bericht_BevPrognose_2018.pdf (25 March 2024, date last accessed).

[ckae126-B15] 15. Gryparis A, Forsberg B, Katsouyanni K et al. Acute effects of ozone on mortality from the “air pollution and Health: a European approach”; project. Am J Respir Crit Care Med 2004;170:1080–7. [DOI] [PubMed] [Google Scholar]

[ckae126-B16] 16.APHEKOM publications, HIA Tools. HIA Tool Short Term (rev. Sept 2013) and HIA Tool Long Term (rev. Jan 2016). http://aphekom.org/web/aphekom.org/publications (25 March 2024, date last accessed).

[ckae126-B17] 17. Perez L, Declercq C, Iñiguez C et al. Chronic burden of near-roadway traffic pollution in 10 European cities (APHEKOM network). Eur Respir J 2013;42:594–605. [DOI] [PubMed] [Google Scholar]

[ckae126-B18] 18.WHO Europe. Health Risks of Air Pollution in Europe—HRAPIE Project. Recommendations for Concentration–Response Functions for Cost–Benefit Analysis of Particulate Matter, Ozone and Nitrogen Dioxide. Copenhagen, 2013. https://iris.who.int/handle/10665/153692?show=full (24 March 2024, date last accessed).

[ckae126-B19] 19. Katsouyanni K, Samet JM, Anderson HR et al. HEI Health Review Committee. Air pollution and health: a European and North American approach (APHENA). Res Rep Health Eff Inst 2009;142:5–90. [PubMed] [Google Scholar]

[ckae126-B20] 20. Moshammer H, Hutter HP, Kundi M. Which metric of ambient ozone to predict daily mortality? Atmos Environ 2013;65:171–6. [Google Scholar]

[ckae126-B21] 21.WHO Europe. AirQ+: Software Tool For Health Risk Assessment of Air Pollution. https://www.who.int/europe/tools-and-toolkits/airq—software-tool-for-health-risk-assessment-of-air-pollution (25 March 2024, date last accessed).

[ckae126-B22] 22. Weitensfelder L, Moshammer H. Evidence of adaptation to increasing temperatures. Int J Environ Res Public Health 2019;17:97. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ckae126-B23] 23. EEA. Air Quality in Europe 2020. https://www.eea.europa.eu/publications/air-quality-in-europe-2020-report (24 March 2024, date last accessed).

[ckae126-B24] 24. WHO. WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide. Geneva: WHO, 2021. [PubMed]

[ckae126-B25] 25. EEA. Austria—Air Pollution Country Fact Sheet. https://www.eea.europa.eu/themes/air/country-fact-sheets/2023-country-fact-sheets/austria-air-pollution-country (25 March 2024, date last accessed).

[ckae126-B26] 26. Huangfu P, Atkinson R. Long-term exposure to NO₂ and O₃ and all-cause and respiratory mortality: a systematic review and meta-analysis. Environ Int 2020;144:105998. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ckae126-B27] 27. Orru H, Andersson C, Ebi KL et al. Impact of climate change on ozone-related mortality and morbidity in Europe. Eur Respir J 2013;41:285–94. [DOI] [PubMed] [Google Scholar]

[ckae126-B28] 28. Rieder HE, Fiore AM, Clifton OE et al. Combining model projections with site-level observations to estimate changes in distributions and seasonality of ozone in surface air over the U.S.A. Atmos Environ 2018;193:302–15. 10.1016/j.atmosenv.2018.07.042. [DOI] [Google Scholar]

PERMALINK

Attributable deaths in Austria due to ozone under different climate scenarios

Hanns Moshammer

Monika Mayer

Harald Rieder

Christian Schmidt

Birgit Bednar-Friedl

Peter Wallner

Hans-Peter Hutter

Abstract

Introduction