Review and Assessment of the Canada-U.S. Air Quality Agreement 2023

U.S. Disclaimer

To the extent this document mentions or discusses statutory or regulatory authority, it does so for informational purposes only. This document does not substitute for those statutes or regulations, and readers should consult the statutes or regulations to learn what they require. Neither this document, nor any part of it, is itself a rule or a regulation. Thus, it cannot change or impose any legally binding requirements on EPA, States, Tribes, the public, or the regulated community.

General Disclaimer

All information in this document is up to date as of July 19, 2023. Exceedance levels of Canadian Ambient Air Quality Standards (CAAQS) and National Ambient Air Quality Standards (“NAAQS”) may have been updated since the publication of this report.

Executive Summary

The Canada-U.S. Air Quality Agreement (AQA), signed in 1991, was originally designed to address transboundary contributions to acid rain caused by emissions of sulfur dioxide (SO₂) and oxides of nitrogen (NO_x). In 2000, the Agreement was amended to address the problem of transboundary ground-level ozone with the addition of commitments on volatile organic compounds (VOCs) and additional measures on NO_x.

By 2007, both countries had met their respective commitments under the Agreement. While the AQA is a remarkable example of what can be achieved through bilateral cooperation, commitments are now over 20 years old. Thus, under Article X, which calls for a comprehensive review and assessment of the Agreement and its implementation every five years unless otherwise agreed, Canada and the U.S. have jointly undertaken the work presented herein.

The objectives of this review and assessment are to:

• Review what the AQA has accomplished to date, including whether it is meeting its current objectives and whether emission reductions mandated by the Agreement have met the AQA objectives to: reduce the transboundary flow of air pollution, reduce acid deposition, reduce concentrations of ground-level ozone, and improve air quality in Canada and the U.S.;
• Assess whether the emissions reduction targets and measures included in Annex 1 (acid rain) and in Annex 3 (ozone) and the commitments in Annex 2 (scientific cooperation) remain appropriate for Canada/U.S. policy and science needs; and
• Examine whether new actions such as commitments and/or measures would be appropriate (e.g., for pollutants included under the Agreement and those not currently addressed, such as PM_2.5).

In the context of this review and assessment, the Parties shall consider such action as may be appropriate, including the modification of the AQA and/or the modification of existing policies, programs, and measures.

Acid Rain

The Acid Rain Annex (Annex 1) to the AQA sets out objectives for Canada and the U.S. to reduce emissions of SO₂ and NO_x that cause acid rain. Both countries have met their commitments to reduce SO₂ and NO_x emissions under the Agreement since 2007.

In 2020, Canada’s total SO₂ emissions were approximately 651,000 metric tons, a 78% reduction from Canada’s total SO₂ emissions of 3.0 million metric tons in 1990. Between 1990 and 2020, Canada’s total NO_X emissions also decreased by 36% (826 thousand metric tons). These emissions reductions have been achieved through programs including the 1985 Eastern Canada Acid Rain Control Program and Canada-wide Acid Rain Strategy for Post-2000.

In the U.S., between 1990 and 2020, SO₂ emissions have decreased by 92% from 23.1 million metric tons to 1.9 million metric tons, and NO_x emissions have decreased by 69%, from 25.5 million metric tons to 7.8 million metric tons. The Acid Rain Program (ARP) in the U.S. has dramatically cut power plant emissions of SO₂ and NO_x, reducing acid rain. Regulatory actions in the U.S. such as the Cross-State Air Pollution Rule (CSAPR) and its subsequent updates have also achieved large reductions in annual SO₂ and annual and summertime NO_x emissions from the power sector.

Continued and remarkable success in both countries in reducing pollutants contributing to acid deposition (SO₂ and NO_x) has led to recent signs of recovery. There are areas in both countries, most notably in eastern Canada, that are still recovering from the historic pollutant loadings and receiving acid deposition that may be in exceedance of current critical loads. Modeling suggests transboundary influence on total deposition, particularly in the less populated parts of northern Montana and the northern parts of the province of Ontario, where deposition is lower than in the northeastern U.S. Furthermore, deposition of reduced nitrogen (including NH₃ and NH₄⁺) has not decreased in recent decades, and increased deposition of reduced nitrogen has been observed in some areas.

Ozone

In 2000, the Ozone Annex (Annex 3) to the AQA set out commitments by Canada and the U.S. to reduce emissions of NO_x and VOCs that contribute to transboundary ozone pollution. These commitments apply to a defined region in both countries known as the Pollutant Emission Management Area (PEMA), which includes central and southern Ontario, Southern Quebec, 18 U.S. States and the District of Columbia.  At the time of the signing of the Annex, this PEMA was the area deemed the most critical for reducing transboundary ozone. These commitments aimed to help both countries attain their respective air quality goals, and to protect human health and the environment.

Canada and the U.S. have met their commitments in the Ozone Annex to reduce emissions of NO_x and VOCs from stationary and mobile sources and from solvents, paints, and consumer products. Canada’s national emissions of NO_x and VOCs have decreased by 36% and 49%, respectively, between 1990 and 2020. The U.S. national air emissions of NO_x and VOCs decreased by 70% and 48% respectively between 1990 and 2020.

Ambient ozone concentrations have also declined within the Canada-U.S. border region since the establishment of the Ozone Annex. Annual 4^th highest MDA8 ozone concentrations have decreased by more than 10 parts per billion (ppb) at many monitoring stations across Ontario, Quebec, and the Maritimes, and by as much as 20 ppb at some stations in the Great Lakes states and Ohio Valley, where ozone concentrations are highest.

Ozone also continues to have significant impacts on public health and agricultural production in the U.S. and Canada, despite progress under Annex 3 of the Agreement. Transport from the U.S. continues to contribute a large fraction of anthropogenic ozone in Canada, with the largest influence in the Windsor-Quebec corridor, as well as southwestern British Columbia, in the greater Vancouver and Victoria area, southern Alberta, the Greater Toronto-Hamilton area, and the Montreal area. Air monitoring stations in southern Ontario and southern Quebec continue to measure ozone concentrations which approach or exceed the CAAQS, and modeling projections suggest continued CAAQS exceedances in 2035. Modeling suggests that transboundary flow of ozone and its precursors from the U.S. to Canada contributes to a significant portion of health impacts in central and Atlantic Canada and is the dominant source of health impacts in Ontario, Quebec, Nova Scotia, New Brunswick, Newfoundland and Labrador, and Prince Edward Island. Transboundary flow from the U.S. into Canada is also estimated to contribute to reduced crop yield, particularly along the Windsor-Quebec City corridor.

Fine Particulate Matter

Although the Agreement does not include PM_2.5, emissions of some of the precursors of secondary PM_2.5 are addressed via actions to reduce NO_x, SO₂, and VOCs. However, direct emissions of primary PM_2.5 and NH₃ (a PM_2.5 precursor) are not addressed under the Agreement. From 1990 to 2020, Canada’s emissions of primary PM_2.5 decreased by 15%, having plateaued at approximately 1.5 million metric tons per year. U.S. national emissions of primary PM_2.5 decreased by 38% between 1990 and 2020, having gradually decreased until 2015, and then plateaued in recent years. The regional and multi-state programs that led to decreased ozone concentrations also reduced emissions of several chemical precursors to secondary PM_2.5 (NO_x, SO₂, and VOCs). However, emissions of NH₃ (another PM_2.5 precursor) have increased by 24% in Canada and by 25% in the U.S. from 1990 to 2020.

Adverse health impacts of PM_2.5 exposure are well documented and both countries continue to take action to address their respective emissions. In recent years, PM_2.5 concentrations are largest near urban areas and particularly in the Ohio Valley, Atlantic coast, and the Windsor-Quebec corridor, with observed concentrations for several stations in Canada approaching or exceeding the CAAQS. Although modeling projects that PM_2.5 concentrations will decrease by 2035, they are also projected to continue to exceed the CAAQS in some of Canada’s largest cities. The analysis presented in this review and assessment finds greater transport of PM_2.5 from the U.S. to Canada. Recent modeling and analysis also indicate that transboundary PM_2.5 increases morbidity and mortality in Canada and has a larger health impact than transboundary ozone. Modeling results support the conclusion that the majority of transboundary PM_2.5 impacts are within several hundred kilometers of the border and felt predominantly in the Michigan-Ontario and Quebec regions – with the largest impacts in the Detroit-Windsor area.

Scientific and Technical Cooperation

The Parties have strengthened their relationship through collaboration and science exchanges under the AQA. Since 1994, the Parties have notified each other of specific sources of air pollution within 100 miles of the Canada-U.S. border. The Canadian and U.S. governments share data through a range of programs and tools such as AirNow and the National Atmospheric Deposition Program (NADP), and maintain ongoing informal dialogue across a range of topics related to monitoring networks and measurement methods. Canada and the U.S. collaborate and share emission inventories, summaries, and analyses under several agreements and reports such as the AQA bi-annual Progress Report, Arctic Council, and the Convention on Long-Range Transboundary Air Pollution (LRTAP). In addition to meeting their respective pollution commitments, the Parties completed multiple pilot projects and joint modeling and analysis under the Ozone Annex. These efforts have expanded information sharing and knowledge of transboundary transport, to the benefit of each country.

Looking to the future, Subcommittee 2 (SC2) co-chairs held a series of science exchange workshops to share information, continue to build connections, and inform policy-related dialogue on current and emerging transboundary issues of concern. The Parties have accomplished a great deal under the AQA, continue to collaborate through a variety of projects and look for ways to continue working together in pursuit of shared goals to improve air quality conditions.

Overall Conclusion and Recommendation

The Canada-U.S. AQA is a model of successful bilateral cooperation resulting in significant improvement in the environment over its three-decade history. Overall findings of the AQA review and assessment indicate that important results have been achieved under the current AQA and both countries have fully met their obligations. However, transboundary air pollution continues to impact both countries from a health and environmental perspective. To continue to meet the objective of the AQA “to control transboundary air pollution between the two countries”, it is recommended that the Agreement be updated, including exploring new strategies to address emerging issues of concern not currently covered under the AQA.

List of Abbreviations/Acronyms

• ADAGIO: Atmospheric Deposition Analysis Generated from optimal Interpolation from Observations
• ANC: Acid neutralizing capacity
• APCA: Anthropogenic Precursor Culpability Assessment
• APEI: Air Pollutant Emissions Inventory (Canada)
• ARP: Acid Rain Program (U.S.)
• AQA: Canada-United States Air Quality Agreement
• AQBAT: Air Quality Benefits Assessment Tool (Health Canada)
• AQC: Canada-United States Air Quality Committee
• AQMEII4: Air Quality Model Evaluation International Initiative Phase 4
• AQMS: Air Quality Management System (Canada)
• AQS: Air Quality System (U.S.)
• AQVM2: Air Quality Valuation Model
• BAU: Business as usual
• BLIER: Base Level Industrial Emissions Requirements (Canada)
• CAAQS: Canadian Ambient Air Quality Standards
• CAIR: Clean Air Interstate Rule (U.S.)
• CAMx: Comprehensive Air Quality Modeling with Extensions
• CAPMoN: Canadian Air and Precipitation Monitoring Network
• CEM: Continuous Emissions Monitoring
• CIESIN: Center for International Earth Science Information Network
• CMAQ: Community Multiscale Air Quality Modeling System
• Comb: Combustion
• COPD: Chronic Obstructive Pulmonary Disease
• CSAPR: Cross-State Air Pollution Rule (U.S.)
• DOC: Dissolved organic carbon
• ECCC: Environment and Climate Change Canada
• EQUATES: EPA Air Quality Time Series Project
• Elec: Electric
• EVE: Electric Vehicles and the Environment
• GEM-MACH: Global Environmental Multiscale Model – Modeling Air Quality and Chemistry
• GRPE: UNECE Working Party on Pollution and Energy
• ICAO: International Civil Aviation Organization
• IMO: International Maritime Organization
• IRA: Inflation Reduction Act (U.S.)
• kg ha^-1 yr^-1: Kilograms per Hectare per Year
• LRTAP: Convention on Long Range Transboundary Air Pollution
• LTM: Long-term monitoring
• MATS: Mercury and Air Toxics Standards (U.S.)
• MDA8: Maximum daily 8-hour average
• MFG: Manufacturing
• Mm^-1: Inverse megameters
• MOOSE: Michigan-Ontario Ozone Source Experiment
• MOVES3: Motor Vehicle Emissions Simulator
• MSAPR: Multi-Sector Air Pollutants Regulations (Canada)
• μeq/L: Micro-equivalents per liter
• µg/m³: Micrograms per cubic meter
• µmol_c L^-1: Micro-mol charge equivalent per liter
• MWe: Megawatts electrical
• NAAMC: National Ambient Air Monitoring Conference (U.S.)
• NAAQS: National Ambient Air Quality Standards (U.S.)
• NADP: National Atmospheric Deposition Program (U.S.)
• NAPS: National Air Pollution Surveillance (Canada)
• NASA: National Aeronautics and Space Administration (U.S.)
• NEI: National Emission Inventory (U.S.)
• NH₃: Ammonia
• NH₄⁺: Ammonium
• NO₃^-: Nitrate
• NO_x: Nitrogen oxides
• NOAA: National Oceanic and Atmospheric Administration (U.S.)
• NSPS: New source performance standards (U.S.)
• nssSO₄^-2: Non-sea-salt sulfate
• OSAT: Ozone Source Apportionment Technology
• PA: Policy Assessment (U.S.)
• PA16: PA nominal present for base year 2016 (U.S.)
• PEMA: Pollutant Emission Management Area (AQA Ozone Annex)
• pH: Potential hydrogen
• PM: Particulate matter
• PM_2.5: Fine particulate matter; inhalable particles generally 2.5 micrometers in diameter or smaller
• PM₁₀: Inhalable particles generally 10 micrometers in diameter or smaller
• ppb: Parts per billion by volume
• PSAT: Particulate matter source attribution technology
• RAQDRS: Regional Air Quality Deterministic Reforecast System (Canada)
• RCU: Final Revised Cross State Air Pollution Rule Update (U.S.)
• RCU16: RCU modeling for year 2016 (U.S.)
• RCU23: RCU projections for year 2023 (U.S.)
• RCU28: RCU projections for year 2028 (U.S.)
• RHR: Regional Haze Rule (U.S.)
• RHR16: Regional Haze Rule modeling for 2016 (U.S.)
• RHR28: Regional Haze Rule modeling for 2028 (U.S.)
• RWC: Residential Wood Combustion
• SC1: AQC Subcommittee on Program Monitoring and Reporting
• SC2: AQC Subcommittee on Scientific Cooperation
• SIP: State Implementation Plan (U.S.)
• SO₂: Sulfur Dioxide
• SO₄^2-: Sulfate
• SSWC: Steady-State Water Chemistry model
• TDEP: Total deposition
• TEMPO: Tropospheric Emissions: Monitoring of Pollution
• TSD: Technical Support Documentation
• UNECE: United National Economic Commission for Europe
• UOG: Upstream Oil and Gas
• U.S.: United States of America
• U.S. EPA: United States Environmental Protection Agency
• Util: Utility
• VOC: Volatile Organic Compound
• WHO: World Health Organization
• WMO: World Meteorological Organization

List of Tables

• Table 2‑1. Acid Rain (Annex 1): Specific objectives concerning SO₂ and NO_x
• Table 3‑1. Ozone Annex (Annex 3): Specific objectives concerning ground-level ozone precursors
• Table 3‑2. National ambient air quality standards for ozone
• Table 4‑1. National ambient air quality standards for PM_2.5
• Table 5‑1. Summary of Annex 2 – Scientific and Technical Activities and Economic Research
• Table A-1. Description of simulations used in the U.S. analysis for ozone.
• Table A-2. Description of simulations used in the U.S. analysis for ozone.

List of Figures

• Figure 2‑1. Canadian SO₂ emission trends for 1990-2020.
• Figure 2‑2. Canadian NO_x emission trends for 1990-2020.
• Figure 2‑3. U.S. SO₂ emission trends for 1990-2020.
• Figure 2‑4. U.S. NO_X emission trends for 1990-2020.
• Figure 2‑5. Annual nssSO₄^2- wet deposition for (a) 1990 and (b) 2019.
• Figure 2‑6. Annual wet NO₃^- deposition for (a) 1990 and (b) 2019.
• Figure 2‑7. Three-year average of total sulfur deposition in the U.S. for (a) 2000-2002 and (b) 2018-2020.
• Figure 2‑8. Three-year average of total nitrogen deposition in the U.S. for (a) 2000-2002 and (b) 2018-2020.
• Figure 2‑9. Influence of Canadian and U.S. emissions on annual sulfur deposition in transboundary region estimated from the U.S. EPA 2020 Ozone Policy Assessment simulations.
• Figure 2‑10. Influence of Canadian and U.S. emissions on annual nitrogen deposition in transboundary region estimated from the U.S. EPA 2020 Ozone Policy Assessment simulations.
• Figure 2‑11. Maps of Average Accumulated Exceedance (AAE) of natural and semi-mineral soils under the Global Environmental Multiscale Model – Modeling Air Quality and Chemistry (GEM-MACH) sulfur plus nitrogen deposition models
• Figure 2‑12. Mapped lake exceedance under GEM-MACH modeled deposition for (a) baseline 2015 and (b) BAU 2035, with changes to the exceedance shown in (c).
• Figure 2‑13. Lake and stream exceedances of estimated critical loads for total nitrogen and sulfur deposition
• Figure 3‑1. Map of Ozone Annex PEMA
• Figure 3‑2. Canadian VOC emission trends for 1990-2020
• Figure 3‑3. U.S. VOC emission trends for 1990-2020.
• Figure 3‑4. Ozone concentrations (three-year average of the annual 4^th highest MDA8 ozone concentration) along the Canada-U.S. border for (a) 2000-2002 and (b) 2018-2020. Concentrations are shown for the region within 500‑km of the Canada-U.S. border.
• Figure 3‑5. Ozone concentrations (five-year average of the annual 4^th highest MDA8 ozone concentration) along the Canada-U.S. border for (a) 1996-2000 and (b) 2016-2020.
• Figure 3‑6. Seasonal average of the MDA8 ozone concentrations from ECCC modeling scenarios for May-September for the base year (a) 2015 and for future projections (b) 2025 and (c) 2035.
• Figure 3‑7. Seasonal average of the MDA8 ozone concentrations from U.S. EPA modeling scenarios for May-September for the base year (a) PA16 and (b) RCU16, and for projections (c) RCU23 and (d) RCU28.
• Figure 3‑8. Influence of Canadian and U.S. emissions on daily maximum ozone concentrations from the ECCC model. Differences in modeled annual average daily maximum ozone concentrations for 2015 when (a) U.S. and (b) Canadian emissions are zeroed out, and for 2035 when (c) U.S. and (d) Canadian emissions are zeroed out.
• Figure 3‑9. Influence of Canadian and U.S. emissions (left y-axis) and ozone concentration (right y-axis) for the 2015 summer average of daily maximum 8-hr ozone concentrations in (a) Canadian and (b) U.S. cities from ECCC modeling.
• Figure 3‑10. Attribution of ozone for PA16 from (a) U.S. emissions and (b) Canadian emissions; for RCU23 from (c) U.S. emissions and (d) Canadian emissions; and for RCU28 from (e) U.S. emissions and (f) Canadian emissions. All results are MDA8 ozone concentrations averaged from May-Sept.
• Figure 3‑11. Ratio of ozone from U.S. to Canada from U.S. PA16 (a), RCU23, (b), and RCU28 (c) modeling scenarios.
• Figure 3‑12. ECCC modeled population-weighted May-Sept mean MDA8 ozone contributions (ppb) from sources (All, U.S., Canada) to populations within 500-km or the PEMA states, the U.S. portion, or the Canadian portion.
• Figure 3‑13. U.S. EPA modeled population-weighted May-Sept mean MDA8 ozone contributions (ppb) from sources (All, U.S., Canada) to populations within 500-km or the PEMA states, the U.S. portion, or the Canadian portion.
• Figure 3‑14. Ratio of contributions from U.S. sources to those from Canadian sources to total ozone-related health impacts in Canada (estimated as an economic value per year) by census division.
• Figure 4‑1. Canadian PM_2.5 Emission Trends. 1990-2020.
• Figure 4‑2. Canadian NH₃ Emission Trends. 1990-2020.
• Figure 4‑3. U.S. PM_2.5 emission trends from 1990-2020.
• Figure 4‑4.  U.S. NH₃ emission trends from 1990-2020.
• Figure 4‑5. PM_2.5 concentrations (98^th percentile of the daily averages) along the Canada-U.S. border for (a) 2001-2005 and (b) 2016-2020.
• Figure 4‑6. PM_2.5 concentrations (Annual average of the daily averages) along the Canada-U.S. border for (a) 2001-2005 and (b) 2016-2020.
• Figure 4‑7. Annual average PM_2.5 concentrations from ECCC modeling scenarios for the base year (a) 2015, and for future projections (b) 2025 and (c) 2035.
• Figure 4‑8. Annual average PM_2.5 concentrations from U.S. EPA modeling scenarios for the base year (a) PA16 and (b) RHR16, and for future projections (c) RHR28.
• Figure 4‑9. Influence of U.S. and Canadian emissions on annual average PM_2.5 concentrations from ECCC zero out modeling.
• Figure 4‑10. Influence of Canadian and U.S. emissions (left y-axis) and PM_2.5 concentrations (right y-axis) for the 2015 PM_2.5 annual average concentrations in (a) Canadian and (b) U.S. cities from ECCC modeling.
• Figure 4‑11. Influence of U.S. and Canadian emissions on annual average PM_2.5 concentrations from U.S. EPA zero out modeling. (a) U.S. and (b) Canadian attribution for PA16 modeling scenario; (c)
• Figure 4‑12. Ratio of PM_2.5 from the U.S. and Canada from U.S. PA16 (left) and RHR28 (right) modeling scenarios.
• Figure 4‑13. ECCC modeled population weighted annual mean PM_2.5 contributions (µg/m³) from sources (All, Canada, U.S.) to populations within 500-km or PEMA states, the U.S. portion, or the Canadian portion.
• Figure 4‑14. U.S. EPA modeled population weighted annual mean PM_2.5 contributions (µg/m³) from sources (All, Canada, U.S.) to populations within 500-km or PEMA states, the U.S. portion, or the Canadian portion.
• Figure 4‑15. Ratio of contributions from U.S. sources to those from Canadian sources to total PM_2.5-related health impacts in Canada (estimated as an economic value per year), by census division.
• Figure 4‑16. The relationship between daily PM_2.5 concentration and visual range reconstructed from particle speciation data stations across Canada for the years 2003-2015.
• Figure 4‑17. Chemical and source composition of regional haze in the UL Bend Park in Montana.
• Figure A-1. Canadian NO_x emissions for 2015, 2025, 2030 and 2035.
• Figure A-2. Canadian VOC emissions for 2015, 2025, 2030 and 2035.
• Figure A-3. Canadian primary PM_2.5 emissions for 2015, 2025, 2030 and 2035.
• Figure A-4. Yearly average daily maximum ozone concentrations from ECCC modeling for (a) BASE CASE 2015, (b) BASE minus Canada and U.S. 20% reduction 2015, (c) BASE minus Canada 20% reduction 2015, and (d) BASE minus U.S. 20% reduction 2015.
• Figure A-5. U.S. and Canada population map with the black outline of a proposed analysis zone for the agreement analysis.
• Figure B-1. MERRA-2 monthly average surface pressures and wind fields are shown for May, July, and September of 2016.
• Figure B-2. MERRA-2 monthly average surface pressures and wind fields are shown for January, March, and November of 2016. The winds are shown at 50-meters (50M), 850 hPa and 500 hPa that represent increasing altitude.

1 Introduction

1.1 Historical Context

In the 1970s, the forests and lakes of North America began to show damage from acid rain caused by emissions of sulfur dioxide (SO₂) and oxides of nitrogen (NO_x). Recognizing that the transboundary flow of these pollutants was an important contributor to acid rain, on March 13, 1991, Canada and the United States (U.S.) (also referred to as ‘the Parties’) signed the Canada-United States Air Quality Agreement (AQA; herein referred to as ‘the Agreement’ or ‘AQA’), a treaty-level agreement that commits both countries to reducing emissions and impacts of transboundary air pollution. The purpose of the Agreement is to establish “a practical and effective instrument to address shared concerns regarding transboundary air pollution with the objective “to control transboundary air pollution between the two countries” (Canada-United States Air Quality Agreement, 1991).

The main body of the Agreement addresses issues such as general objectives, roles and responsibilities of the Parties, methods for exchanging information, and undertaking assessment and consultation. It establishes the framework for the obligations of both countries to deal with transboundary air pollution.

The original AQA included two annexes:

• Specific Objectives Concerning SO₂ and NO_x (Annex 1) contains specific commitments for Canada and the U.S. to reduce nationwide emissions of the precursors of acid rain – SO₂ and NO_x.
• Scientific and Technical Activities and Economic Research (Annex 2) contains guidelines for collaboration in scientific, technical activities and economic research, monitoring activities, and the exchange of information related to air quality, acid deposition, and other areas of mutual interest.

In 2000, the AQA was amended to include Specific Objectives Concerning Ground-level Ozone Precursors (Annex 3) (Canada-US Air Quality Agreement: Ozone Annex, 2000), aiming to address the problem of transboundary ground-level ozone, a key component of smog. Smog is a term that describes a fog or haze combined with smoke and other atmospheric pollutants. Poor air quality due to smog is often associated with reduced visibility and increased incidences of respiratory-related illnesses such as asthma. Smog is composed of a mixture of air pollutants, but its two main components are ground-level ozone and particulate matter (PM).

In 2000, Canada and the U.S. also updated Annex 2 to include further guidelines on cooperation and information exchange related to emissions trading, outreach activities, and data sharing on ground-level ozone and its precursors.

1.2 Environmental Context

Transboundary air pollution refers to emissions of air pollutants that are released in one jurisdiction and then transported or moved by winds and weather systems into another. Transboundary flows include pollutants directly emitted into the air. (i.e., primary pollutants) and those that transform into different substances via a chemical reaction in the air (i.e., secondary pollutants). Many countries are both sources and receptors for transboundary air pollution (Kauffmann & Saffirio, 2020). The AQA was established to address transboundary air pollution between the two countries. Due to prevailing winds and large emissions sources, U.S. emissions of air pollutants can affect air quality in certain regions of Canada as shown by Canadian modeling scenarios. Transboundary transport of pollutants occurs across the length of the Canada-U.S. border but has a greater impact on air quality over southern Ontario, Quebec, and Atlantic Canada (AMC, 2021a).  However, recent episodes of wildfires in both countries have highlighted the potential for wildfire smoke to exert adverse effects across the length of the border (Albores et al., 2023; NOAA, 2023).

Air pollution is the most important environmental contributor to the global burden of disease, leading to an estimated 6 to 7 million premature deaths annually and large economic losses ($5.1 trillion U.S. dollars or 6.6% of the global world product) (UNEP, 2019).  In 2020 the Health Effects Institute reported that air pollution was the fourth leading risk factor for early death worldwide in 2019. The Institute for Health Metrics Evaluation has calculated that in 2019, air pollution worldwide contributed to an estimated 6.67 million premature deaths each year, and 213 million disability adjusted life years lost. In this analysis ambient fine particulate matter (PM_2.5) accounts for 4.14 million premature deaths, household (indoor) air pollution accounts for 2.31 million premature deaths, and ozone accounts for an estimated 365,000 premature deaths (Health Effects Institute, 2020).

Although Canada and the U.S.’s overall air quality is relatively good compared to that of other developed nations, several recent studies indicate that air pollution increases the risk of mortality even at low ambient concentrations (Brunekreef et al., 2021; Crouse et al., 2015; Pappin et al., 2019; Pinault et al., 2017). Researchers looked at the effects of low ambient concentrations of PM_2.5 in 68.5 million older Americans, finding an estimated 6% to 8% increased risk of mortality per 10 μg/m³ (annual average) of PM_2.5 in a low exposure sub-group (Dominici et al., 2022). Health Canada estimates that air pollution from human and natural sources in North America contributes to 15,300 premature deaths per year in Canada, as well as 2.7 million asthma symptom days, and 35 million acute respiratory symptom days per year, with a total economic cost approximately $120 billion Canadian dollars (Health Canada, 2021). The Global Burden of Disease Study estimates that air pollution contributes to 602,000 premature deaths in the U.S. Ten percent of those estimated deaths are caused by chronic obstructive pulmonary disease, 4% from lung cancer, and 3% from lower respiratory infections (Global Burden of Disease Collaborative Network, 2021).

Air emissions also have impact on the environment through deposition. Acid deposition (wet or dry deposition of acidic compounds) removes essential nutrients from soils via leaching and mobilizes toxic aluminum. This loss of nutrients negatively affects the health and growth of trees and depletes the capacity of soils to neutralize future loadings of acid deposition. As such, acid deposition can contribute to declining growth rates and increased death rates in trees and a reduction in biodiversity (e.g., Clark et al., 2019). Detailed descriptions of the impacts of acid deposition on aquatic and terrestrial ecosystems in North America are available in assessment documents from the U.S. (Burns et al., 2011) and Canada (Environment Canada, 2005).

Since establishing the AQA, Canada and the U.S. have both achieved significant reductions in emissions of SO₂ and NO_x, the two major pollutants leading to acidic deposition. Since 2000, the Parties also have made further progress in reducing emissions of NO_x and volatile organic compounds (VOCs) to address ground-level ozone in the Canada-U.S. border region^Footnote1 . In 2007, Canada and the U.S. achieved the emissions reduction targets laid out in both the acid rain and ozone annexes, and these emissions have continued to decrease in the subsequent years. These emissions reductions have led to lower levels of acid rain and ambient ground-level ozone, as well as decreased levels of ambient PM (ECCC & US EPA, 2023).

Over the past three decades, the global environmental context, including that specific to North America, has shifted substantially. For instance, the impacts of climate change are increasingly apparent worldwide, our understanding of climate change and its impacts has evolved, and climate change itself has accelerated. Climate change is linked to changes in air quality – including changes in ozone and PM concentrations – through higher temperatures, increasingly common slow-moving high-pressure weather systems, and more frequent extreme events related to rising temperatures, like wildfires (Health Canada, 2022b). In Canada and the U.S., the impacts of climate change are already being felt in many communities through more frequent and intense extreme weather and climate events. These events are expected to cause disruption and damage to infrastructure and property, and impede the rate of economic growth. Additionally, impacts on the health and well-being of the public, specifically vulnerable populations, remain a concern (USGCRP, 2018). There have also been significant changes to emissions sources. For example, there is evidence that the rapid growth in oil and gas extraction in the Bakken Formation is leading to an increase in transboundary transport of air pollutants in both directions between North Dakota, Montana, Alberta, and Saskatchewan (Prenni et al., 2016).

Both the U.S. and Canada are taking action to address new environmental challenges through new policy initiatives designed to decrease air pollution and mitigate climate change. In 2022, the U.S. announced its largest investment in combating climate change, the Inflation Reduction Act (IRA). IRA investments – along with additional investments within the 2021 Bipartisan Infrastructure Law and 2022 CHIPS and Science Act – accelerate the transition to a clean energy economy, address environmental injustice, reduce renewable energy costs, spur innovation, and are expected to reduce carbon emissions by roughly 40% by 2030. The U.S. is also advancing new regulatory efforts to address a myriad of air pollutants such as new requirements designed to address ozone transport through the Good Neighbor Rule (US EPA, 2023b). In 2022, the Government of Canada released the 2030 Emissions Reduction Plan 2030 Emissions Reduction Plan, which provides a roadmap to reach its climate commitments, such as reducing national greenhouse gas emissions by 40 to 45% below 2005 levels by 2030 under the Paris Agreement, and achieving net-zero emissions by 2050. The Government of Canada is currently developing the Clean Electricity Regulations that will help drive progress towards a net-zero electricity grid by 2035. Further, the Government of Canada has proposed amendments to regulations^{Footnote 2}  which require manufacturers and importers to meet specified annual targets of zero-emission vehicles. Canada has also amended the Canadian Environmental Protection Act, 1999 (Canada, 2023) to recognize that every individual in Canada has a right to a healthy environment, which includes consideration of the principle of environmental justice. Many of these landmark policies are not captured in the projections included in this report, but will, directly or indirectly, impact emissions across the economies of both countries.

1.3 Objectives of the Review and Assessment

Article X of the AQA calls for Canada and the U.S. to conduct a comprehensive review and assessment of the Agreement and its implementation every five years unless otherwise agreed upon. The last review and assessment was completed in 2012, and was included as a section in the 2012 biennial AQA progress report (Environment Canada & US EPA, 2012). The 2012 review recommended that consideration be given to: streamlining the reporting process under the AQA; expanding the scope of the Agreement to address transboundary PM; and addressing transboundary air quality issues in the western border area, if the science demonstrated that there were issues of concern in this area. The ongoing threat to human health, the environment, and the economy posed by air quality issues, particularly in the context of ongoing global warming with its potential to intensify air quality issues, further motivated the need for a review and assessment at the time.

In November 2020, the Canada-U.S. Air Quality Committee (AQC), which oversees implementation of the AQA, agreed to undertake an exercise to define the scope and content of a potential review and assessment of the AQA. This exercise focused on reviewing current Annexes covering acid rain and ground-level ozone and evaluating transboundary impacts from PM_2.5, which is not currently part of the AQA. After defining the scope of an AQA review and assessment in the spring of 2021, the AQC Subcommittee on Program Monitoring and Reporting (SC1), in consultation with the AQC Subcommittee on Scientific Cooperation (SC2), finalized plans to undertake a new review of the AQA, its effectiveness, and potential gaps.

The objectives of this review and assessment are to:

• Review what the AQA has accomplished to date, including whether it is meeting its current objectives and whether emission reductions mandated by the Agreement have met the AQA objectives to: reduce the transboundary flow of air pollution, reduce acid deposition, reduce concentrations of ground-level ozone, and improve air quality in Canada and the U.S.;
• Assess whether the emissions reduction targets and measures included in Annex 1 and in Annex 3 and the commitments in Annex 2 remain appropriate for Canada/U.S. policy and science needs; and
• Examine whether new actions such as commitments and/or measures would be appropriate (e.g., for pollutants included under the Agreement and those not currently addressed, such as PM_2.5).

1.4 Approach for the Review and Assessment

To assess the effectiveness of actions under Annex 1 and Annex 3, and to inform future work under the AQA, emissions, monitoring, and modeling data were evaluated. Emissions data from the Canadian Air Pollutant Emissions Inventory (APEI) and the U.S. National Emissions Inventory (NEI) were used to evaluate historical changes since the AQA came into effect. Note that the most recent year for which data are available for this report, the year 2020, was marked by the COVID-19 pandemic, which coincided with changes in emissions for many pollutants. Wet and dry deposition were measured by the Canadian Air and Precipitation Monitoring Network (CAPMoN), the Alberta Precipitation Quality Monitoring Program, and the U. S. National Atmospheric Deposition Program (NADP). Ozone and PM_2.5 were measured by the Canadian National Air Pollution Surveillance Program (NAPS) and U.S. air monitoring networks included in the Air Quality System (AQS). The monitoring data were used to evaluate historical trends and current concentrations and deposition. Air quality and deposition modeling were conducted to estimate future concentrations and deposition, and to attribute the relative influence of Canadian and U.S. emissions at a given location. Four modeling datasets were considered, including new modeling performed by Environment and Climate Change Canada (ECCC) and modeling from previous U.S. Environmental Protection Agency (U.S. EPA) studies^{Footnote 3} , as described in Appendix A. Instead of performing new joint modeling for this report, both countries agreed to use existing modeling if appropriate, and to perform new separate modeling runs as needed. The U.S. and ECCC model runs are not directly comparable, as they consider different emissions inventories and time periods. However, the modeling runs can be considered together qualitatively to gain confidence in the results. The modeling runs do not include wildfire emissions, as the focus of the current AQA is on the transboundary impact of emission sources that can be directly addressed through targeted measures by each country.

The ECCC air quality modeling output was also used with health and agricultural models to estimate the impacts of transboundary pollution.

The results of these evaluations are presented for Annex 1 commitments on acid rain (Section 2), Annex 3 commitments on ozone (Section 3), as well as for PM_2.5 (Section 4). A review of Annex 2 on scientific and technical cooperation (Section 5) was also conducted. Per Article X of the Agreement, in the context of this review and assessment, the Parties shall consider such action as may be appropriate, including the modification of the AQA and/or the modification of existing policies, programs, and measures. Key findings (Section 6) and recommendations for further collaboration (Section 6.3) are also outlined.

2 Acid Rain

What is acid rain: Acid deposition is the removal of acidic compounds from the atmosphere by the process of wet deposition (precipitation and fog) and dry deposition (transfer of gases and particles to the earth’s surface). Wet deposition is more commonly known as acid rain. Emissions of SO₂ and NO_x from power plants, transportation, industries, and other sources, react in the atmosphere with oxidants to form various acidic compounds, notably sulfuric acid and nitric acid. These acidic compounds can react with ammonia (NH₃) to form secondary inorganic species, such as particle sulfate (SO₄^2-) and particle nitrate (NO₃-), which are also major components of PM_2.5 (Section 4). Once deposited (in either gas or particle form) to surfaces, the acidic compounds harm aquatic and terrestrial ecosystems (particularly forests) and damage surfaces of buildings or other man-made structures.  

Acid rain in the AQA: Annex 1, the Acid Rain Annex to the AQA (see Table 2‑1) established commitments by Canada and the U.S. to reduce emissions of SO₂ and NO_x, the primary precursors to acid rain, from stationary and mobile sources. Both Canada and the U.S. have met their commitments under the Acid Rain Annex, as described in Section 2.1. Specific details on when commitments were achieved over the past three decades can be found in the biennial progress reports.  Between 1990 and 2020, Canada’s total emissions of SO₂ and NO_x decreased by 78% and 36%, respectively. The U.S. total emissions reductions for the same timeframe for SO₂ and NO_x were 93% and 70%, respectively. Reductions in SO₂ and NO_x emissions in both Canada and the U.S. since 1990 have led to major decreases in the wet deposition of SO₄^2- and NO₃- over the eastern half of the two countries. Implementation of various regulatory and non-regulatory actions for more than two decades in Canada has significantly reduced emissions of SO₂ and NO_x and ambient concentrations. Similar measures, especially regulatory programs in the electric power sector, have significantly reduced emissions of SO₂ and NO_x and ambient concentrations in the U.S.

Table 2‑1. Acid Rain (Annex 1): Specific objectives concerning SO₂ and NO_x^{Footnote 4}

2.1 Effect of Emissions Reduction Strategies on Acidifying Pollutants

2.1.1 Canada

Under the Acid Rain Annex (Annex 1), Canada agreed to a permanent national SO₂ emissions cap of 3.2 million metric tons per year by 2000. Canada also agreed to reduce annual NO_x emissions from power plants, major combustion sources, and metal smelting operations by 100 thousand metric tons below the forecasted level of 970 thousand metric tons by 2000. Canada met these commitments through efforts undertaken by the federal government and eastern provinces as part of the 1985 Eastern Canada Acid Rain Control Program and Canada-wide Acid Rain Strategy for Post-2000, which was adopted in 1998. The 1985 Eastern Canada Acid Rain Control Program committed Canada to cap total SO₂ emissions in the seven provinces from Manitoba eastward at 2.3 million metric tons by 1994. Canada met this cap in 1993. Under the AQA, this cap was extended to cover the period 1994-1999.

The Canada-wide Acid Rain Strategy for Post-2000 put in place a framework for addressing the issues related to acid rain with the goal of ensuring that the deposition of acidifying pollutants does not further deteriorate the environment in eastern Canada and that new acid rain problems do not occur elsewhere in Canada. As part of this Strategy, the provinces of Ontario, Quebec, New Brunswick, and Nova Scotia committed to an additional 50% reduction in their SO₂ emissions beyond their 1985 Eastern Canada Acid Rain Program targets by 2010 (by 2015 for Ontario). All four provinces have met the stricter SO₂ emissions targets established under the Strategy. Canada continues to develop measures to reduce emissions that contribute to acid rain and smog. In 2016, Canada published the Multi-sector Air Pollutants Regulations (MSAPR), which includes limits on SO₂ emissions from cement manufacturing facilities. These regulations established Canada’s first mandatory national air pollutant emission standards for major industrial facilities, and are a key element of Canada’s Air Quality Management System (AQMS).

As of 2020, Canada’s total SO₂ emissions were approximately 651 thousand metric tons, about 80% below the national cap of 3.2 million metric tons. Overall, SO₂ emissions decreased by 78% (2.4 million metric tons) between 1990 and 2020 (Figure 2‑1). Reductions in emissions from the Ore and Mineral industries, and in particular the Non-Ferrous Refining and Smelting Industry sector, were the largest driver of this downward trend, particularly in the early 1990s, and again from 2008 to 2020. The decrease in SO₂ emissions since 2008 can be attributed to the preparation and implementation of pollution prevention plans by facilities, the installation of new technology or processes at facilities, the closure of four major smelters in Manitoba, Ontario, Quebec and New Brunswick, and facilities achieving Base Level Industrial Emissions Requirements (BLIERs) through environmental performance agreements (ECCC, 2017, 2018). Emissions from Electric Power Generation (Utilities) decreased significantly from 2005 to 2020, primarily owing to the closure of, or improvements to, generating stations burning heavy fuel oil. Improvements consisted of installing pollution control equipment or switching to low sulfur heavy fuel oil. Furthermore, Coal-fired electric power generation saw an important SO₂ emission decrease of 19% (37 thousand metric tons) between 2019 and 2020, attributed to a decrease in coal consumption.

Canada has also met its 2000 commitment to reduce NO_x emissions from power plants, major combustion sources, and metal smelting operations by 100 thousand metric tons below the forecasted level of 970 thousand metric tons (i.e., cap set at 870 thousand metric tons). Recent measures in Canada’s 2016 MSAPRs further limit NO_x emissions from industrial boilers, heaters, stationary gaseous fuel-fired engines, and cement manufacturing facilities.

Figure 2‑1. Canadian SO₂ emission trends for 1990-2020.

Figure 2-2. Canadian NO_x emission trends for 1990-2020.

Between 1990 and 2020, Canada’s total NO_x emissions decreased by 36% (826 thousand metric tons) (Figure 2‑2) (ECCC, 2022). The most significant changes in NO_x emissions include a decrease of 47% (607 thousand metric tons) from Transportation and Mobile Equipment, a decrease of 61% (156 thousand metric tons) from Electric Power Generation, and an increase of 30% (103 thousand metric tons) from the Oil and Gas industry. A more comprehensive discussion of SO₂ and NO_x emissions in Canada can be found in Canada’s APEI Report for 2022 (ECCC, 2022).

2.1.2 United States

The U.S. met its commitments to reduce SO₂ and NO_x under the Acid Rain Annex (Annex 1). The national Acid Rain Program (ARP) has dramatically cut power plant emissions of SO₂ and NO_x, reducing acid rain as well as secondary formation of PM_2.5. Further reductions in power plant pollution have been achieved by state and U.S. EPA efforts to cut interstate air pollution, which also helped downwind states meet health-based air quality standards for fine particles and ozone. The Clean Air Interstate Rule (CAIR) achieved large reductions in power plant annual SO₂ and NO_x emissions, as well as additional summertime NO_x reductions beyond those required by the 1998 NO_x State Implementation Plan (SIP) Call. In 2015, CAIR was replaced by the Cross-State Air Pollution Rule (CSAPR). In addition, the Mercury and Air Toxics Standards (MATS) rule, which went into effect in April 2015, achieved substantial SO₂ emissions reductions as an additional benefit to air toxics emissions reductions from the power sector. These reductions occurred while the demand for electricity increased and were the result of continued increases in efficiency, installation of state-of-the-art pollution controls, and the switch to lower emitting fuels. These regulatory programs along with economic forces, contributed to a decrease in SO₂ and NO_x emissions by 93% and 70%, respectively, between 1990 and 2020. In addition to control programs mentioned above, the achievement of greater efficiency in energy production and the use of lower emitting fuels have reduced emissions.

The Clean Air Act requires that when new industrial facilities are designed and built, good pollution control must be part of the design. In areas not meeting the NAAQS, to avoid making pollution worse, new and modified large plants and factories must meet the lowest achievable emission rate and obtain offsetting emissions reductions from other sources. In areas that meet the NAAQS, new and modified large plants and factories must apply the best available control technology, considering cost and other factors, and avoid causing significant degradation of air quality or visibility impairment in national parks. For example, natural gas fired units reduce NO_x emissions through flue gas recirculation and low NO_x burners, reducing NO_x by 60 to 90%.

In the transportation sector, which includes highway and non-road vehicles, U.S. EPA has required significant reductions over the years in emissions from new motor vehicles and non-road engines through standards that require a combination of cleaner engine technologies and cleaner fuels. Dating back to the mid-1970s/early 1980s when U.S. EPA established the very first emission standards for new motor vehicles, each decade since has brought forward new more stringent emission standards and cleaner fuels for the fleet. To highlight a few accomplishments, during the 1980s U.S. EPA established Inspection and Maintenance programs for motor vehicles and finalized regulations to remove lead from gasoline. During the 1990s U.S. EPA imposed limits on diesel fuel sulfur content and finalized new emission standards for diesel engines used in construction and agricultural equipment. In the early 2000s U.S. EPA finalized regulations for small non-road handheld engines such as trimmers and chainsaws and also finalized regulations to reduce air toxics from mobile sources. As the most recent example, in December 2021, U.S. EPA finalized standards for light-duty vehicles (US EPA, 2021b) that will reduce emissions, while bolstering energy security and encouraging manufacturing innovation. The U.S. EPA also recently finalized new heavy-duty engine and vehicle standards (US EPA, 2022b) in December 2022. The rule sets stronger emission standards to further reduce air pollution from heavy-duty vehicles and engines starting in model year 2027. The final program includes new, more stringent emissions standards that cover a wider range of heavy-duty engine operating conditions compared to today's standards, and it requires these more stringent emissions standards to be met for a longer period of time when these engines operate on the road.

Figure 2‑3 shows trends in U.S. SO₂ emissions. The greatest reduction in SO₂ emissions between 1990 and 2020 is from stationary fuel combustion (95%). In 2020, electric utility fuel combustion contributed 49% of the total SO₂ emissions (1.7 million short tons). The transportation sector experienced a 97% decrease in total SO₂ emissions (42,000 short tons).

Figure 2‑4 shows U.S. NO_x emission trends. Total NO_x emissions from electric utility fuel combustion decreased 87% from 1990 to 2020. Petroleum and Related Industries noted an increase in NO_x emissions to a peak in 2012 but have decreased 28% since then. Transportation NO_x emissions have decreased 70% from 1990 to 2020. The methods used to compute all emissions categories were updated starting with the year 2002. A different version of U.S. EPA’s Motor Vehicle Emissions Simulator (MOVES3) was used starting 2001 for highway vehicles, emission factors, and activity data accounts for the increase in emissions from 2001 to 2002.

Figure 2-3. U.S. SO₂ emission trends for 1990-2020.

Figure 2-4. U.S. NO_x emission trends for 1990-2020.

2.2 Current Levels and Trends in Acid Deposition

Wet deposition of SO₄^2- and NO₃^- is measured by precipitation chemistry monitoring networks in Canada and the U.S. The measurement data, presented in kilograms per hectare per year (kg ha^-1 yr^-1), are the basis for binational spatial wet deposition maps.

Figure 2‑5 shows the spatial patterns of annual wet SO₄^2- deposition of non-sea-salt sulfate (nssSO₄^2-), in 1990 and 2019 along with point values at sites in less densely measured regions. Non-sea-salt sulfate is the measured SO₄^2- with the contribution from sea salt SO₄^2- removed (WMO, 2004) for sites within 100 km of an ocean. The interpolation for 2019 was extended over southwestern Canada because of the increased site density in this region compared to 1990. The lower Great Lakes region received the highest wet deposition of SO₄^2- early in the 30-year period, while recently the maximum has shifted to the Mississippi valley. In 1990, SO₄^2- deposition exceeded 26 kg nssSO₄^2- ha^‑1 over a large area of eastern North America. In 2019, only a small area in southern Louisiana exceeded 10 kg nssSO₄^2- ha^‑1. It is noteworthy that while SO₄^2- deposition decreases were most pronounced in the east, all areas of the domain exhibited decreased impacts.

Figure 2‑6 shows the patterns of wet NO₃^- deposition in 1990 and 2019, along with point values at sites in less densely measured regions. Similar to SO₄^2-, the lower Great Lakes region consistently received the highest wet deposition of NO₃^- early in the 30-year period. NO₃^- deposition exceeded 19 kg NO₃^- ha^‑1 in many parts of the northeastern U.S. and southern Ontario and Quebec in 1990. In 2019, NO₃^- deposition was less than 13 kg NO₃^- ha^‑1 throughout North America. The steep declines in NO₃^- wet deposition after the year 2000 are due to major NO_x emission reductions in both countries. Similarly to SO₄^2-, most areas of the continent experienced benefits with regards to deposition of NO₃^-.

Figure 2‑5. Annual nssSO₄^2- wet deposition for (a) 1990 and (b) 2019.

Data sources: CAPMoN, the Alberta Precipitation Quality Monitoring Program, and the U.S. NADP.

Figure 2‑6. Annual wet NO₃^- deposition for (a) 1990 and (b) 2019.

Data sources: CAPMoN, the Alberta Precipitation Quality Monitoring Program, and the U.S. NADP.

These results are consistent with other recent studies, which have found that reductions in SO₂ and NO_x emissions in both Canada and the U.S. between 1990 and 2019 have led to decreases in the wet deposition of SO₄^2- and NO₃^- over the eastern half of both countries. Wet deposition of SO₄^2- and NO₃^- at monitoring stations in eastern Canadian and eastern U.S. decreased by 69% and 46%, respectively, from 1989 to 2016, with the decline in NO₃^- wet deposition occurring primarily after 2000 (Feng et al., 2021), corresponding to the substantial declines in NO_x emissions commencing in 2000. Similar decreasing trends in wet SO₄^2- and NO₃^- were observed in the U.S. Mid-Atlantic, Midwest and Northeast based on precipitation samples collected as far back as 1981 to as recently as 2019 (Baldigo et al., 2021; Burns et al., 2021; Isil et al., 2022; Likens et al., 2021; McHale et al., 2021).

Total deposition is made up of wet deposition plus dry deposition. Dry deposition is typically cost-intensive to measure and typically is more subject to uncertainties than wet deposition (Walker, Bell, et al., 2019). In order to estimate dry and subsequently total deposition, measurements of ambient concentrations are paired with modeled dry deposition velocities in measurement model fusion techniques (Fu et al., 2022), including the Total Deposition (TDep) method (Schwede & Lear, 2014) for the U.S. and the Atmospheric Deposition Analysis Generated from optimal Interpolation from Observations (ADAGIO) method (Robichaud et al., 2020) for Canada.

Dry deposition in 2018-2020 comprised 33% of U.S. total sulfur deposition on average across the U.S., which decreased from 40% from 2000 to 2002. There are areas with notable high dry sulfur deposition (exceeding 60%) along the Canada-U.S. border (northwestern and northeastern Washington state, northeastern Montana and northwestern North Dakota). Based on data from 15 CAPMoN sites collected between 2000 and 2018 (Cheng et al., 2022), dry deposition contributes to 11 to 55% and wet from Cheng et al., 2022, the reduction in total sulfur deposition at eastern Canadian sites from 2000–2002 to 2016-2018 was 70%. There are limited data on the total nitrogen deposition budget in Canada as CAPMoN does not routinely measure NO₂, other oxidized forms of nitrogen, or NH₃ in ambient air for subsequent estimation of dry deposition fluxes. Maps of total deposition (wet plus dry) are shown for the U.S. for sulfur deposition in Figure 2‑7 and nitrogen deposition in Figure 2‑8. Note that similar maps of total deposition are not yet available for Canada because of ongoing development and evaluation of a model-measurement fusion deposition product. The reduction in total sulfur deposition (wet plus dry) in the eastern U.S. has been of similar magnitude to that of wet deposition with an overall average reduction of 81% from 2000–2002 to 2018–2020 (Figure 2‑7). All areas of the eastern U.S. have shown significant improvement in wet SO₄^2- deposition, with an overall 70% reduction from 2000–2002 to 2018–2020. Between 2000–2002 and 2018–2020, the Northeast and Mid-Atlantic experienced the largest reductions in wet SO₄^2- deposition, 77% and 74% reduction, respectively. Reductions in total nitrogen deposition recorded since the early 1990s have been less pronounced than those for sulfur. The most recent Canadian and U.S. studies report that the decline in total deposition since 2010 has been slower than that of the previous decade, particularly for nitrogen (Benish et al., 2022; Cheng et al., 2022).

The increasing importance of reduced nitrogen (NH₃ and ammonium - NH₄⁺) to total nitrogen deposition has been reported across both countries. In Canada, wet deposition of inorganic nitrogen (NO₃^- plus NH₄⁺) decreased on average by 25% at eastern Canadian sites from 2000–2002 to 2016–2018. However, this was entirely due to changes in NO₃^-, as no trends were observed in wet NH₄⁺deposition at the majority of the sites from 2000 to 2018 (Cheng et al., 2022).  

Figure 2‑7. Three-year average of total sulfur deposition in the U.S. for (a) 2000-2002 and (b) 2018-2020.^{Footnote 7}

Figure 2‑8. Three-year average of total nitrogen deposition in the U.S. for (a) 2000-2002 and (b) 2018-2020.^{Footnote 8}

In the U.S., wet deposition of inorganic nitrogen decreased an average of 19% in the Mid-Atlantic and 32% in the Northeast but increased by 17% and 9% in the North and South Central regions from 2000–2002 to 2018–2020. Increases in wet deposition of inorganic nitrogen in the North Central and South Central regions are attributed to 44% and 34% increases in wet deposition of reduced nitrogen (NH₄⁺), respectively, between 2000 and 2020. Considering total (wet + dry) deposition, decreases in oxidized nitrogen (NO_x) have generally been greater than increases in reduced nitrogen (NH_x) deposition. Total oxidized nitrogen deposition decreased 57 % in the east, which total deposition of reduced nitrogen increased by an average of 38 % in the east from 2000–2002 to 2018–2020 (NADP, 2023). Overall, long-term trends in atmospheric deposition of NH₄⁺ and NH₃ in Canada and the U.S. were either observed to fluctuate with no statistically significant trend (Burns et al., 2021; Cheng et al., 2022; Feng et al., 2021; Likens et al., 2021; Zhang et al., 2018) or were increasing (Benish et al., 2022; McHale et al., 2021). The increasing contribution of reduced nitrogen to acidification, eutrophication, and air quality impacts is a topic of growing interest to scientists (Walker, Beachley, et al., 2019) and policy makers (Kanter et al., 2020).

2.3 Influences of Transboundary Flow

The U.S. 2020 Ozone Policy Assessment (PDF)^{Footnote 9}  (Ozone PA) (US EPA, 2020d) encompassed simulations that estimated nitrogen and sulfur deposition. These simulations used zero-out for attribution, which lends itself to a similar contribution analysis of deposition. However, the simulations were not evaluated against deposition observations and have not been fused with observations like the total deposition products in the maps shown above (US EPA, 2020d). Qualitatively, Figure 2‑9 and Figure 2‑10 show that contributions from each country are largest near their own population centers. In the less populated parts of northern Montana and the northern parts of the province of Ontario, there are larger contributions due to transboundary impacts. Compared to the Northeastern regions, these areas have relatively low total deposition in both the raw model results and, for Montana, in the fused maps above.

Figure 2‑9. Influence of Canadian and U.S. emissions on annual sulfur deposition in transboundary region estimated from the U.S. EPA 2020 Ozone Policy Assessment simulations. The panels show (a) the total deposition, (b) total deposition from U.S. emissions sources only, (c) total deposition from Canadian sources only, and (d) the ratio total deposition from U.S. over Canadian emissions sources. Modeled deposition is shown for the region within 500-km of the Canada-U.S. border.

Note that Ozone PA modeling results utilized different datasets for the U.S. and Canada. There are sharp transitions in land-use classifications between grassland and farmland, particularly in the southern Alberta and northern Montana region. This sharp shift in classification could impact modeled levels of nitrogen deposition in particular, considering the relatively large levels of nitrogen associated with cropland. Further verification of land-use classifications and their representation in modeled outcomes could be pursued by both countries in future work.

Figure 2‑10. Influence of Canadian and U.S. emissions on annual nitrogen deposition in transboundary region estimated from the U.S. EPA 2020 Ozone Policy Assessment simulations. The panels show (a) the total deposition, (b) total deposition from U.S. emissions sources only, (c) total deposition from Canadian sources only, and (d) the ratio of total deposition from U.S. over Canadian emissions sources. Modeled deposition is shown for the region within 500-km of the Canada-U.S. border.

2.4 Aquatic Acidification and Eutrophication

Deposition of total nitrogen and sulfur to water bodies can cause or contribute to aquatic acidification and eutrophication which can affect the health of the aquatic ecosystem. The amount of a pollutant that is estimated to cause a biological effect or change in an ecological system is called a critical load. Critical loads for aquatic systems are generally developed using models that take into account historical pollution impacts and geological conditions, both of which affect leaching of total nitrogen and sulfur out of the aquatic system. Ecosystems that are less responsive to acidic pollution have high critical loads, while sensitive ecosystems have low critical loads. An ecosystem is said to be in exceedance of its critical load when acid deposition crosses the critical loads threshold. In both the U.S. and Canada, sustained emissions reductions and deposition rates outlined in Sections 2.1 and 2.2 have resulted in broad reductions in critical load exceedances. However, some persistent challenges remain in both countries where current sulfur and nitrogen loadings from 2019-2021 still exceed levels required for recovery of some lakes and streams. In the U.S., critical loads are used for analysis purpose only, whereas in Canada, critical loads are used to inform emissions policy development.

While deposition from air-related emissions sources is an important contributor of total nitrogen and total sulfur, it is generally not the dominant source of total nitrogen and total sulfur to water bodies that are located within or close to agricultural areas. For example, synthetic fertilizer is estimated to be the single largest source of human-caused total nitrogen inputs (US EPA, 2020b). Geologic sources in the soil of the watershed, generally in the form of SO₄^-2, release total sulfur into the water and can be important contributors, especially in areas that have experienced historically high total sulfur deposition (US EPA, 2020b).

2.4.1 Canada

Critical loads of acidity were established for lakes across Canada using the Steady-State Water Chemistry model (SSWC) (Henriksen & Posch, 2001) and setting a threshold based on the acid neutralizing capacity (ANC). The ANC threshold was usually set at a value related to the dissolved organic carbon (DOC) concentration in the lake water^{Footnote 10}  or, in the absence of DOC data, at the commonly used value of 40 µmol_c L^-1 (Henriksen et al., 2002). The critical load for a region was set at a level to protect 95% of lakes in that area from the harmful effects of acid deposition. Critical loads of acidity for natural and semi-natural mineral soils (i.e., excluding organic and agricultural soils) were recently mapped using the Simple Mass Balance model (Posch et al., 2015), which considers inputs from deposition and base cation weathering as well as outputs such as leaching and harvesting. A site-specific critical base cation to aluminum ratio representing a 5% root or biomass growth restriction was set to protect sensitive trees and vegetation.

The resulting aquatic and terrestrial critical loads of acidity can be used to estimate exceedances to identify areas of concern. For example, in a recent study (Cheng et al., 2022), terrestrial critical loads of acidity were estimated surrounding 14 stations and aquatic critical loads were estimated around five sites in Canada, which represent a range of soil and surface water conditions. In the early 2000s, acid deposition exceeded critical loads at approximately one third of the sites, but since 2012 has decreased below these thresholds.

Model estimates of nitrogen and sulfur deposition provide improved spatial coverage and chemical coverage (e.g., deposition of non-measured species) and the ability to look at the impact of future emission projections on acid deposition and resulting exceedances of critical loads. The model scenarios described in Appendix A present estimated deposition of total nitrogen and total sulfur. Figure 2‑11 compares the terrestrial sulfur plus nitrogen critical load exceedances for the base year (2015) to the projected exceedances for the 2035 business as usual (BAU) scenario. Lake critical load exceedances (Figure 2‑12) were estimated using sulfur deposition only, since the SSWC model does not take into account sinks of nitrogen in a catchment or lake and should be considered the “best case scenario” for surface waters.

As seen in Figure 2‑11 and Figure 2‑12, large areas of northern Alberta and Saskatchewan are particularly sensitive to acid deposition and are in exceedance of their critical loads both in the base year, 2015, and future scenario for 2035.

Figure 2‑11. Maps of Average Accumulated Exceedance (AAE) of natural and semi-mineral soils under the Global Environmental Multiscale Model – Modeling Air Quality and Chemistry (GEM-MACH) sulfur plus nitrogen deposition models for (a) baseline 2015 and (b) BAU 2035, along with (c) reductions (negative values) or increases (positive values) in AAE between 2015 and 2035. Critical loads of acidity were estimated using the Simple Mass Balance model (UNECE, 2004).

Data sources: Soil data sourced from the OpenLandMap project as well as the Canadian Soil Information System. Runoff estimates from Reinds et al., 2015. Forest data sourced from the National Forest Inventory and supporting information from Pardo et al., 2005; Paré et al., 2013. Landcover data from North American Land Cover Change Monitoring System (2010).

Figure 2‑12. Mapped lake exceedance under GEM-MACH modeled deposition for (a) baseline 2015 and (b) BAU 2035, with changes to the exceedance shown in (c). Lake critical loads of acidity were estimated with the SSWC model using sulfur deposition only. Values above 0 are considered in exceedance of their critical loads.  Critical loads were summed to the 5^th percentile, to provide protection for 95% of the ecosystems in each 42 km grid (Jeffries et al., 2010).

Eastern Canada and southwestern British Columbia also show critical load exceedances for both soils and surface waters, though the mineral soil exceedances are significantly smaller in area under the projected future emissions. These latter regions are significantly impacted by contributions from transboundary transport of total nitrogen and total sulfur deposition, as shown in Figure 2‑9 and Figure 2‑10. Under the 2035 BAU scenario, exceedance of soil critical loads was broadly predicted to decline in magnitude and area except in a few regions near point sources; a smaller reduction in S exceedance in central and eastern Canada is seen in the lake maps.

2.4.2 United States

Figure 2‑13 shows that U.S. emissions reductions achieved between 2000 and 2021 have contributed to broad surface water improvements and increased aquatic ecosystem protection across the five Long-Term Monitoring (LTM) regions along the Appalachian Mountains. These emissions reductions are expected to continue to contribute to improvements in coming years.

In support of the analysis shown in Figure 2‑13, the U.S. collected surface water samples from 7,869 lakes and streams along acid-sensitive regions of the Appalachian Mountains and some adjoining Northern coastal plain regions through a number of water quality monitoring programs. Critical loads information were obtained from the National Critical Load Database (NCLDv3.2.1), a repository of critical load data for the U.S. (CLAD, 2022; Lynch et al., 2022). Aquatic critical loads were determined using a host of methods from the SSWC model (Henriksen & Posch, 2001) to a regional regression model (McDonnell et al., 2012; Sullivan et al., 2012). Critical load exceedances were determined using total deposition estimate of total nitrogen and sulfur deposition for the period of 2000-2002 and 2019-2021 (US EPA, 2021a).

The analysis shown in Figure 2‑13 focuses on aquatic biological resources in acid-sensitive regions in the eastern U.S. Lake and stream waters having an ANC^{Footnote 11}  – a key indicator of aquatic ecosystem recovery from acidification – value greater than 50 μeq/L.

Figure 2‑13. Lake and stream exceedances of estimated critical loads for total nitrogen and sulfur deposition, for (a) sites that exceeded the critical load in 2000-2002 but did not exceed the critical load in 2019–2021 and (b) sites that still exceeded the critical load in 2019-2021.

An ANC of 50 is often used as a target in Critical Loads for freshwater waterbodies^{Footnote 12} . The aquatic critical load represents the amount of combined sulfur and nitrogen that could be deposited annually to a lake or stream and its watershed and still support a moderately healthy aquatic ecosystem. Critical loads exceedances showed a decreasing trend from 2000-2002 to 2019-2021. The percentage of lakes and streams exceeding critical loads fell from 38% to 5.8%.

2.5 Recovery

Recent studies of long-term trends in water and soil chemistry provide indicators of ecosystem recovery in both countries. Signs of acid deposition impacts on soils include the release of free aluminum, which is toxic to plants and animals, and increased potential hydrogen (pH). Forest soils in eastern Canada and the northeastern U.S. showed some evidence of decreased free aluminum and increased pH at most sites sampled (Lawrence et al., 2015), suggesting that forest soils are undergoing recovery from the effects of acidic deposition. In response to acid deposition reductions, SO₄^2- concentrations significantly declined, and pH generally increased, in surface waters in Atlantic Canada, Ontario, Quebec, the northeastern U.S., New York, and Virginia (e.g., Baldigo et al., 2021; Houle et al., 2022; Marty et al., 2021; Nelson et al., 2021; Redden et al., 2021; Scanlon et al., 2021; Watmough & Eimers, 2020; Webster et al., 2021). However, the recovery is incomplete. For example, very acid-sensitive lakes in Nova Scotia did not show evidence of recovery up to 2007 (Clair et al., 2011), have only recently started to recover (Redden et al., 2021), and pH and calcium values are still below recommended thresholds to protect aquatic life (Houle et al., 2022). In other locations, ANC, an indicator of recovery, has continued to decrease (e.g., Scanlon et al., 2021), or toxic aluminum continued to increase (Redden et al., 2021). The recovery of lakes is impacted by cumulative effects of acidic deposition on catchment soils (e.g., Eng & Scanlon, 2021; Hazlett et al., 2020; Scanlon et al., 2021), and may need decades to reach target pH and ANC (Whitfield et al., 2007). Sustained efforts to maintain or further reduce levels of acid deposition throughout the U.S. and Canada would allow these sensitive areas time to recover.

2.6 Summary

The Acid Rain Annex (Annex 1) to the AQA sets out objectives for Canada and the U.S. to reduce emissions of SO₂ and NO_x that cause acid rain. Both countries have met their commitments to reduce SO₂ and NO_x emissions under the Agreement since 2007.

In 2020, Canada’s total SO₂ emissions were approximately 651,000 metric tons, a 78% reduction from Canada’s total SO₂ emissions of 3.0 million metric tons in 1990. Between 1990 and 2020, Canada’s total NO_X emissions also decreased by 36% (826 thousand metric tons). These emissions reductions have been achieved through programs including the 1985 Eastern Canada Acid Rain Control Program and Canada-wide Acid Rain Strategy for Post-2000.

In the U.S., between 1990 and 2020, SO₂ emissions have decreased by 93% from 23.1 million metric tons to 1.9 million metric tons, and NO_x emissions have decreased by 70%, from 25.5 million metric tons to 7.8 million metric tons. The ARP in the U.S. has dramatically cut power plant emissions of SO₂ and NO_x, reducing acid rain. Regulatory actions in the U.S. such as CSAPR and its subsequent updates have also achieved large reductions in annual SO₂ and annual and summertime NO_x emissions from the power sector. Further reductions, not yet reflected in cited modeling, are expected following recent regulatory actions, such as the Good Neighbor Plan (US EPA, 2023b), and upcoming implementation of the provisions contained in the Inflation Reduction Act.

SO₄^2- and NO₃^- wet deposition have decreased by approximately 70% and 50%, respectively, in response to SO₂ and NO_x emission reductions in both Canada and the U.S. Long-term water quality surveys show conditions improving from historical acidification for watersheds in some regions, while delayed recovery has been observed in others.

While important progress has been made, further reductions in total nitrogen and sulfur inputs to U.S. and Canadian water bodies, including those from air-related sources, could further reduce aquatic acidification and speed up the recovery of aquatic and terrestrial ecosystems. Zero-out modeling suggests transboundary influence on total deposition, particularly in the less populated parts of northern Montana and the northern parts of the province of Ontario, where deposition is lower than in the northeastern U.S. Many water bodies in Canada as well as some in the U.S. are still exposed to total nitrogen and sulfur that exceed the capacity of soils and surface waters to neutralize the acidic contributions. The continued exceedances and modeling projections indicate that aquatic acidification remains an environmental concern. Additionally, emissions of ammonia have not experienced the sharp decline that has been seen with SO₂ and NO_x emissions. In fact, in some areas (i.e., agricultural regions) reduced nitrogen emissions have increased. The role of nitrogen in contributing to acidification and eutrophication is an issue of great interest to scientists (Walker, Beachley, et al., 2019) and policy makers (Kanter et al., 2020).

3 Ground-level Ozone

What is ground-level ozone: Ground-level ozone is a colorless and highly irritating air pollutant and greenhouse gas that forms just above the earth’s surface. As outlined in Section 1.2, ozone is harmful to human health and ecosystems and reduces crop yields. Ozone is not emitted directly into the air but rather is formed primarily by chemical reactions between NO_x and VOCs  in the presence of sunlight. NO_x and VOCs are emitted by combustion sources such as vehicles and power plants. VOCs are also given off by solvents, cleaners, and paints, as well as natural (biogenic) sources.

Ground-level ozone in the AQA: Annex 3, the Ozone Annex, contains commitments by Canada and the U.S. to control and reduce emissions of NO_X and VOCs, key precursors to ground-level ozone. Both Canada and the U.S. have met their commitments, as described in Section 3.2. These commitments apply to a defined region in both countries known as the Pollutant Emission Management Area (PEMA), which at the time of the signing of the Annex, was the area deemed the most critical for reducing transboundary ozone. The PEMA includes central and southern Ontario, southern Quebec, 18 U.S. states, and the District of Columbia (Figure 3‑1). The objective of the Ozone Annex is to help both countries achieve their respective air quality standards for ozone, that is, the then-current Canada-wide Standard for ozone (later replaced by the Canadian Ambient Air Quality Standard (CAAQS) for Ozone in 2013) and the U.S. ozone NAAQS. See Table 3‑1. for a summary of the commitments under the Ozone Annex.

Figure 3‑1. Map of Ozone Annex PEMA

Table 3‑1. Ozone Annex (Annex 3): Specific objectives concerning ground-level ozone precursors^{Footnote 13}  

3.1 Air Quality Standards and Guidelines

Both Canada and the U.S. have established air quality standards to protect human health and the environment.

For Canada, CAAQS were established as objectives under the Canadian Environmental Protection Act, 1999, and are a key component of Canada’s AQMS. The CAAQS for ozone were selected on the basis of providing a specific level of improvement to population exposure (and the related improvement to population health) when the targets are met. Future more stringent targets recognize the non-threshold nature of the health effects of ozone (i.e., adverse effects occur even at low concentrations) and the AQMS principle of continuous improvement. Further, while the CAAQS are primarily health-based, they also explicitly recognize that at these levels, environmental effects may also be exerted by ozone. The CAAQS are underpinned by four air quality management levels, where each level requires progressively more rigorous management action by provincial and territorial jurisdictions as the air quality in a given air zone or area approaches or exceeds the CAAQS (CCME, 2021). Certain Canadian jurisdictions have also established air quality criteria/objectives of their own. Under AQMS, provincial and territorial governments are required to provide annual reports on air quality for each of their air zones. These reports include the actual metric values and achievement status of the CAAQS for each CAAQS reporting station and air zone, as well as associated management actions. The CAAQS are not legally binding or enforceable, meaning that there are no mechanisms (financial or otherwise) imparted by the federal government that require provinces to achieve the CAAQS.

The U.S. NAAQS (US EPA, 2023e) are set by the U.S. EPA in accordance with the U.S. Clean Air Act, which requires the U.S. EPA to establish standards for those pollutants for which air quality criteria have been issued under section 108 of the U.S. Clean Air Act, which are also referred to as criteria pollutants. The NAAQS are generally implemented through the development of state implementation plans, through which states are required to attain and maintain the level of the NAAQS, and which must be submitted to the U.S. EPA for review and approval. Section 109 of the U.S. Clean Air Act directs the EPA Administrator to set two types of NAAQS, “primary” and “secondary” standards, for criteria pollutants. Under section 109(b)(1) of the Act, the primary NAAQS are health-based, and are defined as those standards “the attainment and maintenance of which in the judgment of the Administrator, based on such [air quality] criteria and allowing an adequate margin of safety, are requisite to protect the public health.” Under section 109(b)(2), secondary standards are welfare based, and must “specify a level of air quality the attainment and maintenance of which, in the judgment of the Administrator, based on such criteria, is requisite to protect the public welfare from any known or anticipated adverse effects associated with the presence of [the] pollutant in the ambient air.” Under section 302(h) of the Act, effects on welfare include, but are not limited to “effects on soils, water, crops, vegetation, manmade materials, animals, wildlife, weather, visibility, and climate, damage to and deterioration of property, and hazards to transportation, as well as effects on economic values and on personal comfort and well-being.” (US EPA, 2023e).

Table 3‑2 summarizes the current CAAQS for ozone and the U.S. ozone NAAQS. Each standard is defined in terms of the chemical species or mixture to be measured, an averaging time period, a “numerical value” (level), and a “metric” (the statistical form of the numerical standard). For ozone, the CAAQS and U.S. NAAQS are the annual 4^th-highest daily maximum 8-hour concentration (MDA8), averaged over three years. The CAAQS numerical values for ozone are becoming more stringent over time, with different standards for 2015, 2020, and 2025.

Table 3‑2. National ambient air quality standards for ozone.

3.2 Effect of Emissions Reduction Strategies on Ground-Level Ozone Precursors

3.2.1 Canada

Under the Ozone Annex, Canada committed to establishing more stringent NO_x and VOC emissions standards for vehicles and engines, limiting the sulfur content in fuels, establishing annual NO_x emissions caps for fossil fuel power plants in southern Ontario and southern Quebec, and establishing regulations to reduce emissions of VOCs. Canada has met its Ozone Annex commitments. Canada has implemented and continues to implement a series of regulations to align Canadian emission standards for vehicles, engines, and fuels with corresponding standards in the U.S. These include regulations that set emission performance standards for on-road (On-Road Vehicle and Engine Emission Regulations, SOR/2003-2) and off-road vehicles (Off-Road Small Spark-Ignition Engine Emission Regulations, SOR/2003-355) and off-road mobile and stationary engines (see the most recent Progress Report for further details (ECCC & US EPA, 2023)), as well as regulations to limit sulfur and benzene in gasoline and sulfur in diesel. Both Ontario and Quebec have met their NO_x limits for the electricity sector.

Canada has also put in place various regulations to address VOC emissions, such as regulations to reduce VOCs from dry cleaning, solvent degreasing, automotive refinishing products, and architectural coatings. Recent regulations establish maximum VOC concentrations and emissions for the manufacture and import of over 130 categories and sub-categories of products. Canada has published a notice of intent to renew the federal agenda on the reduction of emissions of VOCs from consumer and commercial products for the period of 2022 to 2030. Canada has also established requirements to limit VOC emissions from industrial facilities, as well as regulations to reduce emissions of methane and certain VOCs from the upstream oil and gas sector.

The addition of the Ozone Annex in 2000 resulted in further decreases in Canada’s NO_x emissions, as shown in Figure 2‑2. Strategies for reduction of NO_x emissions were included under the Acid Rain Annex, but NO_x emissions in Canada continued to increase by 485 thousand metric tons (21%) from 1990 to 1999. From 2000 to 2020, after the Ozone Annex was included in the AQA, NO_x emissions in Canada decreased by 1.3 million metric tons, with the largest decreases in the transportation (900 thousand metric tons), electric power generation (226 thousand metric tons), and manufacturing (111 thousand metric tons) sectors.

Between 1990 and 2020, national VOC emissions in Canada decreased by 49% (1.4 million metric tons) (Figure 3‑2). For example, increasingly stringent regulations on spark-ignition engines lead to a nation‑wide decrease of 657 thousand metric tons in VOC emissions from off-road gasoline, liquefied petroleum gas, and natural gas vehicles and equipment.

Figure 3‑2. Canadian VOC emission trends for 1990-2020.

3.2.2 United States

Under the Ozone Annex, the U.S. committed to implement existing U.S. vehicle, non-road engine, and fuel quality rules to achieve both VOC and NO_x reductions. In addition, the U.S. committed to implementing existing rules for control of emissions from stationary sources of hazardous air pollutants and control of VOCs from consumer and commercial products, architectural coatings, and automobile repair coatings, and to continue existing U.S. new source performance standards (NSPS) to achieve VOC and NO_x reductions from new sources. These actions have led to the U.S. meeting its commitments under the Ozone Annex

Emissions of ozone precursors have been reduced due to several programs. Reductions in power plant emissions have been achieved through multi-state NO_x budget programs designed to address interstate transport of air pollution contributing to ozone, first by twelve New England and Mid-Atlantic states and the District of Columbia, and later for much of the eastern U.S. (the NO_x SIP Call Regulations). The subsequent CAIR achieved large reductions in power plant annual SO₂ and NO_x emissions beyond those required by the NO_x SIP Call. In response to litigation, CAIR was replaced by the CSAPR. CSAPR was later updated and revised to further reduce summertime NO_x emissions from power plants in the eastern U.S. and help downwind states to meet ozone standards. In March 2023, the U.S. adopted a new set of NO_x control requirements for power plants and industrial sources under the “Good Neighbor Rule” (US EPA, 2023b).

The U.S. has required significant reductions over the years in NO_x and VOC emissions from new motor vehicles and non-road engines – such as those used in construction, agriculture, industry, trains, and marine vessels – through standards that require a combination of cleaner engine technologies and cleaner fuels, such as those listed in Section 2.1.2.

U.S. EPA continues to implement and update NSPS to achieve VOC and NO_x reductions from new and modified sources. Reductions of NO_x emissions are also being achieved through rules on solid waste incineration units and guidelines that impact new and existing incineration units. Programs and rules such as those mentioned above have contributed significantly to NO_x and VOC reductions, enhancing public health and environmental protections regionally and for local communities.

As shown in Figure 2‑4, there has been an overall trend of emissions reduction for NO_x. A similar trend can be seen for VOCs, as shown in Figure 3‑3. The decrease in emissions from 1990 to 2020 for NO_x is 70% and for VOCs is 48%. The greatest NO_x reductions are found in electric utility fuel combustion. Transportation recorded a greater than 80% reduction in VOC emissions from 1990 to 2020. While there is an overall decrease in VOC emissions, petroleum and related industries noted an increase in emissions from 611 thousand short tons in 1990 to a peak of 3.1 million short tons in 2015, a pattern similar to that of NO_x. In 2020, VOC emissions were 13% less than at the 2015 peak. Emissions estimation methods and reporting for sources in oil and gas production have also improved significantly in recent years.

Figure 3‑3. U.S. VOC emission trends for 1990-2020.

3.3 Ambient Concentrations of Ozone

Ambient ozone concentrations were analyzed to assess the changes in ozone over the course of the AQA and to evaluate current ambient concentrations. Figure 3‑4 shows interpolated (by kriging) maps of average annual 4^th highest MDA8 ozone concentrations within 500 km of the Canada-U.S. border for 2000-2002 and 2018-2020. Figure 3‑4a shows ozone concentrations in the border region for 2000-2002, around the time of signing of the Ozone Annex. The highest ozone concentrations were generally located in the Great Lakes-St. Lawrence region, the upper midwestern and northeastern U.S., southern Ontario, southern Quebec, and the southern Maritimes. Parts of southern Ontario and southern Quebec had annual 4^th highest MDA8 ozone concentration over 80 ppb and 70 ppb, respectively, exceeding the 2000 Canada-wide standard for ozone of 65 ppb. Various areas of nonattainment in U.S. Great Lakes States also exceeded the 1997 NAAQS of 80 ppb, reaching 3-year averages of annual 4^th highest MDA8 ozone concentration of 90 ppb in some locations. The lowest ozone concentrations were observed in the western side of the U.S.-Canadian border region.

Ozone concentrations have decreased significantly throughout the border region since the Ozone Annex was signed. Figure 3‑4b, illustrating ozone concentrations in the border region for 2018-2020, shows that ozone concentrations are much decreased compared to 2000-2002. However, higher ozone concentrations still occur near the Great Lakes and along the U.S. eastern coast. The lowest values are generally found in western and eastern Canada. Concentrations are generally higher in, and downwind of, urban areas. Although ozone concentrations have decreased in southern Ontario, southern Quebec, and Great Lake States, areas continue to have ozone concentrations exceeding national standards in both the U.S. and Canada in these regions at the time of this report.

Figure 3‑4. Ozone concentrations (three-year average of the annual 4^th highest MDA8 ozone concentration) along the Canada-U.S. border for (a) 2000-2002 and (b) 2018-2020. Concentrations are shown for the region within 500‑km of the Canada-U.S. border.

Ozone monitoring data for individual stations located within 500 km of the Canada-U.S. border for the period just prior to when the Ozone Annex was signed, 1996-2000 (Figure 3‑5a), and for the most recent data years, 2016-2020 (Figure 3‑5b), are shown in Figure 3‑5. Ozone concentrations are shown as a 5-year average of the annual 4^th-highest MDA8 ozone concentration. This statistical form is similar to metrics for both the CAAQS and U.S. NAAQS, but it is extended over a longer averaging period to account for variability introduced by wildfires and other factors. The concentrations were not adjusted for transboundary flow or exceptional events, including wildfires. Wildfires have a significant influence on ozone at some stations, as described below.

Nearly all the Canadian and U.S. stations east of the Manitoba-Ontario border have recorded lower ozone concentrations since 2000, by more than 10 ppb for many stations across Ontario, Quebec, and the Maritimes, and by as much as 20 ppb at some stations in the Great Lakes states and Ohio Valley. While some ozone concentrations measured over the past two decades have been associated with cool, rainy summers, the decreases from high MDA8 ozone concentrations are mainly due to air quality regulatory programs undertaken by both countries. A slight rise in ozone concentrations from 1996-2000 to 2016-2020 is observed at some stations in Alberta, Saskatchewan and Washington State. This may be related to impacts from a number of larger wildland fires (Canadian Interagency Forest Fire Centre, 2023; National Interagency Fire Center, 2023).  

Overall, net ozone concentration decreases highlight the success of air quality regulations in the transport-relevant analysis region. However, despite decreasing trends, 2016-2020 ozone concentrations continue to be elevated in southeastern and southern Ontario, as well as some stations near the Mid-Atlantic and New England coasts. Some locations downwind of the Greater Toronto Area recorded higher ozone concentrations, likely a result of increased NO_x emissions from sources in this large metropolitan area. In Canada, many of the stations within the PEMA continue to measure ozone concentrations that approach or exceed the CAAQS.

Figure 3‑5. Ozone concentrations (five-year average of the annual 4^th highest MDA8 ozone concentration) along the Canada-U.S. border for (a) 1996-2000 and (b) 2016-2020.

3.4 Projected Changes in Ozone

Current and projected ozone concentrations from recent ECCC and U.S. EPA modeling efforts were used to estimate how ozone concentrations are projected to change in the future. The modeling is described in further detail in Appendix A. The ECCC modeling was performed using GEM-MACH, a chemical transport model. Modeling scenarios are shown for base year 2015 as well as for projected BAU scenarios for 2025 and 2035. The projected years include regulations that will be enforced in the future. The U.S. EPA modeling data include results from the 2020 Ozone PA (US EPA, 2020d) and the Final Revised Cross-State Air Pollution Rule Update (RCU) (US EPA, 2020a).^{Footnote 17}  The EPA modeling uses the Community Multiscale Air Quality Model (CMAQ) in the PA modeling and Comprehensive Air Quality Model with Extensions (CAMx) for the RCU modeling. Future year scenarios include activity growth and emission controls associated with “on the books” regulations in the U.S. – for example the effect of fleet turnover. In the RCU modeling Canadian emissions were developed from the 2015 base year and 2023 and 2028 future projection data provided to U.S. EPA by ECCC. More details on the inventories in the RCU project can be found in the emissions modeling technical support documentation (TSD) (US EPA, 2021c).

The ECCC and U.S. EPA modeling results are shown for the seasonal average of the MDA8 ozone concentration for May through September during the future year simulations. The modeling was performed using 2019 meteorology (ECCC) or 2016 meteorology (U.S. EPA), and by varying the projected emissions for different future years. Hence, the projected changes in the future are due entirely to changes in emissions of precursors, and the subsequent transportation, transformation, and fate of chemical constituents. Note that the ECCC and U.S. EPA modeled different reference and future years and the modeling uses different years for their meteorology and different emissions datasets, so the outputs are not directly comparable. However, the modeling projections can be considered together qualitatively to gain more confidence in projected changes in ozone concentrations.

The ECCC modeling results are shown in Figure 3‑6 for the base year 2015, and for two future scenarios – 2025 and 2035. The modeling for the 2015 base year is broadly consistent with the monitoring data shown in Section 3.3, with the highest ozone concentrations in the Great Lakes – St. Lawrence region and southern Maritimes. Similarly in the U.S., the highest ozone concentrations were estimated to the northwest and south of Lake Erie, corresponding to the urban centers of Detroit (Michigan) and Cleveland (Ohio). Somewhat elevated ozone concentrations are also captured by the model for parts of Alberta and southern British Columbia. Between 2015 and 2025, there is a substantial decrease in the modeled ozone concentrations, with the biggest changes in the U.S. and southern Ontario. These results are consistent with projected decreases in emissions of precursors (NO_X and VOCs) due to continued actions in Canada and the U.S. (see Section 3.2). Between 2025 and 2035, there is little projected change in emissions, and ozone concentrations remain fairly stable. Despite the projected decreases in emissions and ozone concentrations between 2015 and 2035, projected annual 4^th highest value of the MDA8 ozone concentrations for 2035 (not shown here) exceed the 2025 CAAQS of 60 ppb, in areas in southern Ontario, with the highest concentrations along the Ontario-Michigan border region.

The U.S. EPA modeling results are shown in Figure 3‑7 for the 2016 base year, and for projections to 2023 and 2028. The PA and RCU baseline modeling for 2016 are broadly consistent with the monitoring data shown in Section 3.3 and the ECCC 2015 baseline modeling results. Modeled ozone concentrations are higher in the Ohio Valley region (Illinois, Indiana, Ohio, West Virginia) and southern Ontario, and lower in other areas. Notably, the south-to-north gradient of ozone is most pronounced along the eastern U.S. and Canada. The PA nominal present for base year 2016 (PA16) modeling has a less pronounced gradient than the RCU16. This is likely due to a combination of factors. The PA16 used an earlier version of emissions provided by ECCC, which had large emissions in Alberta associated with oil production. The RCU used a newer version where the Alberta emissions were revised downward. PA16 used the Community Multiscale Air Quality Modeling System (CMAQ) and RCU16 used the Comprehensive Air Quality Model with Extensions (CAMx). Though many fundamentals are shared, each model has its own implementation of chemistry, aerosols, transport, and deposition. In addition, the U.S. emission inventories are different versions of the 2016 modeling platform (i.e., versions 2016fe and 2016fh). The magnitude of these differences is useful for understanding the range of reasonable results.

By comparing the three RCU plots in Figure 3‑7, projected changes in ozone concentrations can be inferred. Ozone concentrations are projected to decrease between 2016 and 2023, with the largest decreases within the Ohio Valley and northeastern U.S. Between 2023 and 2028, ozone concentrations remain fairly steady. This is broadly consistent with the ECCC modeling results. These plots highlight the influence of projected decreases in emissions from both the U.S. and Canada, but more simulations are necessary to quantify and attribute the total contributions. Based on the RCU modeling, it is expected that locations in the U.S. that are within 500 km of the Canadian border will likely attain the ozone NAAQS threshold of 70 ppb before 2028. Note that the modeling does not consider the impact of climate change and increasing temperatures, likely to have an impact on ozone formation.

Figure 3‑6. Seasonal average of the MDA8 ozone concentrations from ECCC modeling scenarios for May-September for the base year (a) 2015 and for future projections (b) 2025 and (c) 2035.

Figure 3‑7. Seasonal average of the MDA8 ozone concentrations from U.S. EPA modeling scenarios for May-September for the base year (a) PA16 and (b) RCU16, and for projections (c) RCU23 and (d) RCU28.

3.5 Influence of Transboundary Flow

Ozone is known to have a long atmospheric lifetime ranging from 8 days to 120 days (Seinfeld & Pandis, 2006) that allows for long-distance transport. Previous air quality modeling by the Ontario Ministry of Environment and Climate Change^{Footnote 18}  (MOECC, 2018) found large transboundary contributions to ozone near the Canada-U.S. border in southwestern Ontario.

Appendix B shows a meteorological analysis of winds along the Canada-U.S. border. For ozone, we focus on the summertime when ozone production is high. During this time, aloft winds are generally west-to-east. Near the surface, the flows take on more of a south-to-north component particularly between Detroit and Windsor. This change in flow leads to transport from the U.S. to Canada being favorable during the ozone season.

To estimate the influence of emissions from Canada and the U.S., additional modeling runs were performed. The ECCC and U.S. EPA Ozone PA modeling runs were repeated, with anthropogenic emissions for each country set to zero (i.e., zeroed out). To visualize the influence of the emissions from each country, a difference map can be calculated by subtracting the zeroed-out result from the base case (which includes the emissions from both countries). Where a difference map is positive, the base case has a higher value, meaning that the zeroed-out country has an influence in those regions. Note that for the U.S. EPA modeling, a zero-out of Canada only was not performed. Instead, both Canada and Mexico were simultaneously removed. The two countries are far enough apart and transport patterns are sufficiently different that the Canada (or Mexico) contribution can generally be distinguished by location. The zero-out modeling runs are a first order method of determining the impact of one set of emissions against another. As an intermediate step, 20% reduction runs for the 2015 emissions year were also performed with the ECCC model to better understand these results. The U.S. EPA RCU modeling uses the Ozone Source Apportionment Technology (OSAT) method to infer contributions from Canada and the U.S.

These modeling results are interpreted qualitatively due to the non-linear nature of ozone production. While NO_x is a precursor for ozone under many conditions, in highly polluted regions it can titrate ozone leading to very low ozone concentrations, for example, in city centers or regions of low ventilation. In such cases, local ozone concentrations can be lower than background, and a reduction in NO_x can lead to an increase in ozone since it is no longer being titrated. An example can be found in southern Quebec near the Canada-U.S. border, for which a decrease of 20% of either Canadian or U.S. emissions in the ECCC modeling (see Appendix A) results in a slight increase in ozone. In addition, ozone can be formed by either NO_x or VOCs, such that in regions with lower NO_x levels, ozone production can be controlled by NO_x and vice versa. This factor adds another level of nonlinearity into ozone production. Since the zero-out scenarios are extreme cases and not realistic, quantitative interpretation of the zero-out modeling results is not presented here. The zero-out modeling can however be used to broadly demonstrate the relative influence of transboundary flow, such that qualitative interpretation of the results is presented here, with an indication of whether transboundary flow makes a large or small contribution to total concentrations.

Figure 3‑8a shows the ECCC modeled annual average of daily maximum ozone concentration differences between the 2015 base case and the corresponding scenario without U.S. emissions, which can be used to estimate the influence of the U.S. emissions. The region that is most affected in Canada is southern Ontario, where there is a large gradient with concentration differences increasing from north to south. Smaller influences of U.S. emissions are also observed in southern Quebec, New Brunswick, Nova Scotia, and southern British Columbia. The influence of U.S. anthropogenic emissions on ambient ozone concentrations over the U.S. is significant, with the largest influence seen in the northeastern U.S. extending toward the south of Lake Superior, as well as around the Seattle area.

Figure 3‑8. Influence of Canadian and U.S. emissions on daily maximum ozone concentrations from the ECCC model. Differences in modeled annual average daily maximum ozone concentrations for 2015 when (a) U.S. and (b) Canadian emissions are zeroed out, and for 2035 when (c) U.S. and (d) Canadian emissions are zeroed out.

Figure 3‑8b shows the ECCC modeled annual average of daily maximum ozone concentration differences between the 2015 base case and the corresponding scenario without Canadian emissions, which can be used to estimate the influence of Canadian emissions on both the U.S. and Canada. The positive values indicate a contribution from Canadian anthropogenic emissions. There is a strong influence of Canadian emissions on ozone concentrations in Canada, with the largest influence in southwestern British Columbia, in the greater Vancouver and Victoria area, southern Alberta, the Greater Toronto and Hamilton Area, and the Montreal area. In southern Ontario, the influence of U.S. emissions on ambient ozone concentrations (Figure 3‑8a) is larger than the influence of Canadian emissions (Figure 3‑8b).

ECCC modeling projections to 2035 were used to estimate future influences of emissions from the U.S. (Figure 3‑8c) and Canada (Figure 3‑8d). Concentrations of ozone are expected to decrease between 2015 and 2035, as described in Section 3.4, due to projected decreases in emissions. However, because the same 2019 meteorology is used for the 2015 and 2035 modeling, the impact of transboundary flow is similar in 2035 compared to the 2015 base year, and projected decreases are due to changes in emission sources. The modeling projections do not include the impact of climate change on meteorology.

Figure 3‑9 shows the influence of Canadian and U.S. emissions on the seasonal average MDA8 ozone concentrations for May to September in 27 Canadian (Figure 3‑9a) and U.S. (Figure 3‑9b) cities located within 500 km of the Canada/U.S. border for the 2015 base year. For a given city, modeling data from the grid cell containing the latitude and longitude of the city were used.

Figure 3‑9. Influence of Canadian and U.S. emissions (left y-axis) and ozone concentration (right y-axis) for the 2015 summer average of daily maximum 8-hr ozone concentrations in (a) Canadian and (b) U.S. cities from ECCC modeling.

Note that the influences from Canada and the U.S. do not add to 100% because ozone concentrations are also affected by other factors, such as global background concentrations. The response in Canadian cities’ ozone concentrations to U.S. emissions varies from less than 5% of the summer MDA8 ozone concentration in Calgary, where the modeled ozone concentration is approximately 35 ppb, to approximately 40% of Windsor’s summer MDA8 ozone concentration, where the modeled ozone concentration is approximately 50 ppb. For more than half of the Canadian cities shown, the response to the U.S. emissions is larger than the response to the Canadian emissions, particularly for the cities located closer to the Canada-U.S. border. Across all Canadian cities listed, the majority of ozone concentrations are influenced by factors beyond U.S. and Canadian emissions, such as global background ozone levels. The response in U.S. cities’ ozone concentrations to Canadian emissions varies from less than 2% in Washington D.C., where the summer MDA8 ozone concentration is approximately 65 ppb to up to approximately 10% in Rochester, New York where the summer MDA8 ozone concentration is approximately 40-45 ppb. For all the U.S. cities shown, a much larger response is attributed to U.S. emissions than Canadian.

The U.S. EPA modeling of influences from U.S. and Canadian emissions are shown in Figure 3‑10 for the MDA8 ozone concentration averaged from May to September. The U.S. EPA modeling includes two projects with varying levels of detail. The Ozone PA simulations include a 2016 simulation and zero out attribution from coarse hemispheric (108-km) and fine regional simulations (12-km). The RCU project included a 2016 simulation and projections to the future for 2023 and 2028. Only the RCU figure simulations (2023 and 2028) included source apportionment-based attribution. For more details, see Appendix A. The U.S. zero out modeling results are shown as absolute MDA8 ozone concentrations from May to September and are not directly comparable to the ECCC modeling differences in modeled annual average daily maximum ozone concentrations. However, the broad patterns in the transboundary transport can be compared between the modeling runs.

Figure 3‑10 shows the ozone concentrations attributed to emissions from the U.S. for the 2016 base case. The U.S. contribution is largest in the Ohio Valley and the northeastern U.S. In all years, the largest contribution of the U.S. ozone concentrations to Canada occurs in the Windsor-Quebec Corridor, with smaller contributions observed in southern Quebec, and southern British Columbia. Some smaller contributions can also be seen in the 2016 base case in New Brunswick and Nova Scotia, but these areas are outside the geographic scope of the RCU modeling. This is broadly consistent with areas of largest influence observed in the ECCC zero out modeling results. Areas designated nonattainment of the 2015 ozone NAAQS that are within 500 km of the border have relatively large U.S. local contributions (Figure 3‑10a), and low contributions from Canada (Figure 3‑10b).

RCU23 and RCU28 modeling was used to project the U.S. and Canada contributions to ozone concentration for future years (Figure 3‑10, panels c-f). Smaller contributions are found in future years, which is consistent with projected emissions reductions (see Section 3.4). However, the PA16 used a different model and a different attribution technology, which prevents a quantitative comparison between the attributions in 2016 and 2023. The differences in attribution between 2023 and 2028, both for RCU modeling, show continued reductions in ozone. Furthermore, the patterns in influence of transboundary flow are consistent between 2016 and future modeled years, with transboundary influence from the U.S. into southwestern Ontario evident for all modeling years. This is broadly consistent with the findings of the ECCC modeling.

Differences in Canadian contributions to ozone concentrations are observed in Alberta between the PA16 (Figure 3‑10b) and the RCU23 and RCU28 modeling runs (Figure 3‑10d,f). As noted in Appendix A, the RCU23 modeling was performed for the 12 km domain with boundary conditions from a 36 km domain. The larger domain did not track attribution, so boundary conditions in the 12 km domain cannot distinguish the Canadian contribution. As a result, Canada’s emissions outside the domain are categorized as part of the “boundary conditions” rather than as attributed to Canada. As a result, the RCU23 and RCU28 values near the domain edge may underestimate Canada’s contribution. Again, the difference between PA16 and RCU23 reflect model and methodology differences described above in addition to the emission year. Even so, a decreasing trend from 2016 to 2023 and 2028 is observed. Note that the same color scale has been used for the U.S. and the Canadian contributions, which allows for comparison.

Figure 3‑10. Attribution of ozone for PA16 from (a) U.S. emissions and (b) Canadian emissions; for RCU23 from (c) U.S. emissions and (d) Canadian emissions; and for RCU28 from (e) U.S. emissions and (f) Canadian emissions. All results are MDA8 ozone concentrations averaged from May-Sept.

Figure 3‑11 shows the ratio of U.S. and Canada contributions from the U.S. EPA modeling PA16, RCU23 and RCU28 modeling runs. These results can be used to identify areas where the contributions from the U.S. and Canada are relatively similar (light red, white and light blue) or where one country’s contributions dominate. Red indicates areas where U.S. contributions are larger than Canada’s and blue indicates the reverse. Transport is clearly important when the U.S. contribution in Canada is comparable to the Canadian contribution, and vice versa.

Figure 3‑11. Ratio of ozone from U.S. to Canada from U.S. PA16 (a), RCU23, (b), and RCU28 (c) modeling scenarios.

In the U.S., the figures show that Canadian ozone contributions are most often less than a quarter of U.S. contributions. Contributions from Canada can be closer to half of U.S. contributions in a narrow band along the northern edge of Montana, North Dakota, New York, Vermont, New Hampshire, and Maine. The maximum ozone contribution of Canada in the U.S. is in the northwest corner of Washington. These influences can also be seen in Figure 3-10.

In Canada, the U.S. ozone contributions are frequently larger than Canadian contributions. This is most often the case in less populated areas. For example, U.S. contributions can be four times the Canadian contributions in the western portions of Ontario. In the more populated parts of Ontario, however, the U.S. contribution is two to two-thirds times the Canadian contribution. One notable area of exception is Windsor and the entirety of Essex County, where larger levels of transboundary influence can be seen due to the area’s close proximity to the border and U.S. metro areas. Although emissions from the U.S. influence ozone concentrations in Canada, it is noted that local emissions and local formation of ozone in some areas of eastern Canada play a significant role in elevated ozone concentrations close to or above the CAAQS. Furthermore, the Canadian Smog Science Assessment (Environment Canada, 2012) indicated that the intercontinental transport of pollutants from Asia into North America can occur through the winter and early spring, contributing to background concentrations^{Footnote 19}  at the surface and leading to increases in ambient air pollution across Canada and the U.S., from the mid-latitudes extending into the Arctic. Intercontinental transport of ozone and precursors (e.g., from Asian emissions) is explicitly captured in the U.S. EPA modeling at the hemispheric scale that provides hourly boundary conditions to the finer scale modeling. The ECCC modeling implicitly includes long-range transport via climatological boundary conditions. The climatological boundary conditions represent concentrations of ozone and precursors that flow in from outside the domain, and therefore includes trans-pacific transport (e.g., from Asia) in an average sense, but would not include episodic behavior associated with long-range transport events.

3.6 Health Impacts

Short-term exposure to ozone can cause a broad range of respiratory effects depending on the level of exposure. These health effects range from small, transient and/or reversible changes in lung function, airway inflammation and respiratory symptoms (e.g., coughing, scratchy throat, and pain when taking a deep breath), to more serious health outcomes such as emergency department visits and hospital admissions (US EPA, 2020c). In addition to the above effects, Health Canada concluded that acute exposure to ozone likely causes an increased risk of total non-accidental and cardiopulmonary mortality (Health Canada, 2013). Studies of short-term exposures to high levels of ozone also report a likely association with metabolic effects (US EPA, 2020c). Long-term ozone exposure is suggested to be linked to aggravation of lung diseases such as asthma and chronic obstructive pulmonary disease (COPD) and may lead to asthma development (US EPA, 2020c). Studies in locations with elevated concentrations also report associations of long-term exposure to ozone with deaths from respiratory causes. Certain populations, such as people with pre-existing respiratory conditions (e.g., asthma), children, older adults, outdoor workers, people carrying certain gene variants (e.g., antioxidant enzyme and inflammatory mediator variants), and people with reduced intake of certain nutrients (e.g., vitamins C and E) are at higher risk for health effects (US EPA, 2020c). Health Canada has concluded that there is a lack of evidence of a threshold below which there is no risk to population health (Health Canada, 2013). In Canada, the health burden of above-background ozone in 2016 was estimated to be 4,100 premature deaths annually plus large numbers of adverse respiratory outcomes, with an economic cost of $30 billion per year (2016 Canadian dollars) (Health Canada, 2021).

Population-weighted averages are used to estimate the ambient concentrations that a population is, on average, exposed to. As such, population-weighted concentrations are a good indicator of the direction and magnitude of health impacts. ECCC and U.S. EPA zero-out modeling runs were used to estimate contributions of U.S. and Canadian sources to the population-weighted average May-Sept mean MDA8 ozone concentration. This analysis uses the Gridded World Population dataset version 4 for population estimates on a 2.5 arcminute grid. Ozone concentration data were interpolated to the grid with the population data to calculate the population weighted average. Contributions to population-weighted average concentrations are reported in Figure 3‑12 and Figure 3‑13. Each panel represents a different population: (a) all people within the 500 km buffer or a PEMA state, (b) the U.S. population within that area, and (c) the Canadian population within that area. Each source definition represents a subset of emission sources: “All” represents the total emissions sources from Canada and the U.S., “U.S.” represents only the anthropogenic sources within the U.S., and “Canada” represents only the anthropogenic sources within Canada. For the U.S. EPA modeling, the “Canada” label includes emissions from both Canada and Mexico. However, Mexican contributions are small enough to neglect near the Canadian border, and therefore this is labeled as Canada.

Ozone population-weighted averages from the ECCC modeling are shown in Figure 3‑12. For all populations within the 500‑km buffer on either side of the border or a PEMA state (Figure 3‑12a), larger contributions to population-weighted average ozone are found from U.S. emissions, and smaller contributions from Canadian emissions^{Footnote 20} . This is expected because the population in the U.S. is larger and has a concomitant larger impact. Similar contributions are observed for U.S. populations within the 500‑km buffer or a PEMA state (Figure 3‑12b). For populations in Canada within the 500-km buffer (Figure 3‑12c), emissions from Canada and the U.S. have similar influences on population-weighted average ozone concentrations. Total population-weighted ozone concentrations (“All” in Figure 3‑12a) are lower in the Canadian portion of the 500-km buffer than in the U.S. portion. Population weighted ozone concentrations are projected to decrease in both Canada and the U.S. between 2015 and 2025, primarily due to a drop in contributions from U.S. sources. Between 2025 and 2035, there is little change in projected population-weighted ozone concentrations.

The U.S. EPA modeling of population-weighted averages is shown in Figure 3‑13 and is broadly consistent with the ECCC modeling. Within 500 km of the border on the U.S. side (Figure 3‑13b), total 2016 average MDA8 ozone concentration is projected to decrease by approximately 10% by 2028. The Canadian contribution is generally small, as is the magnitude of decreases between years. Trends cannot be quantitatively determined by direct comparison of PA16 and RCU23 modeling results for U.S. and Canada-specific contributions. However, decreased emissions in the U.S. shown in the “All” category between RCU16 and RCU23 can largely be seen as coming from U.S. sources, given the relatively marginal Canadian contributions in this area. Within 500 km of the border on the Canadian side (Figure 3‑13c), total 2016 MDA8 ozone concentration decreases by 5% in 2028. For population weighted values, the U.S. and Canadian contributions are nearly equal in 2016. Based on the RCU modeling, the contribution magnitudes from the U.S. in Canada are decreasing more rapidly from 2023 to 2028 than the contributions from Canada. The U.S. contribution decreases are larger in magnitude on the U.S. side of the border than on the Canadian side, though similar in percentages.

Figure 3‑12. ECCC modeled population-weighted May-Sept mean MDA8 ozone contributions (ppb) from sources (All, U.S., Canada) to populations within 500-km or the PEMA states, the U.S. portion, or the Canadian portion.

Figure 3‑13. U.S. EPA modeled population-weighted May-Sept mean MDA8 ozone contributions (ppb) from sources (All, U.S., Canada) to populations within 500-km or the PEMA states, the U.S. portion, or the Canadian portion.

To attribute ozone health impacts in Canada to U.S. and Canadian sources, Health Canada applied the Air Quality Benefits Assessment Tool (AQBAT v3) (Judek et al., 2019) to data produced by the ECCC zero-out modeling runs for 2015, as in Pappin et al., (2024). Since the modeling runs should be interpreted qualitatively, as described in Section 3.5, the health impact analysis should also be interpreted accordingly. Figure 3‑14 maps the estimated ratios of source contributions from the U.S. to those from Canada to total ozone health impacts by Canadian census division.

Figure 3‑14. Ratio of contributions from U.S. sources to those from Canadian sources to total ozone-related health impacts in Canada (estimated as an economic value per year) by census division.

Areas in orange and yellow suggest that U.S. sources contribute more to local health impacts than Canadian sources, while areas in green suggest larger contributions from Canadian sources. The transboundary flow of ozone and its precursors from the U.S. to Canada contributes to a significant portion of health impacts in Central and Atlantic Canada and British Columbia and is the dominant source of health impacts overall in Ontario, Quebec, Manitoba, Nova Scotia, New Brunswick, Newfoundland and Labrador, and Prince Edward Island (Pappin et al., 2024). Some large areas of Quebec and Newfoundland and Labrador see comparable contributions from Canadian and U.S. sources, though many (but not all) are relatively unpopulated. In addition, one-third of the total contribution of health impacts from transboundary ozone occurs in Canadian provinces where U.S. contributions are less than Canadian contributions (ratios < 1). Transboundary ozone health impacts mostly occur in Ontario and Quebec within 300 km of the Canada-U.S. border. Health impacts in Canada attributable to transboundary ozone are projected to decline from 2015 to 2025, and increase from 2025 to 2035 due in part to an increasing number of Canadians susceptible to adverse health effects as a result of ageing, and population growth due to higher immigration (Pappin et al., 2024). Note that these results do not consider the impact of climate change on meteorology and the resulting impacts on air quality and health.

3.7 Environmental Impacts

There is a causal relationship between ozone exposure and impacts on vegetation (US EPA, 2020c). Ozone is taken up by plants through pores, or stomates, in their leaves and can cause direct physical damage resulting in premature plant aging, reduced photosynthesis, reduced plant reproduction, visible foliar symptoms (e.g., discoloring, necrosis, etc.), reduced uptake of carbon dioxide, increased vulnerability to pest attacks, increased tree mortality, reduced primary productivity, and more (US EPA, 2020c). These changes at the individual plant level can lead to reductions in growth and yield of vegetation, reduction in carbon storage capacity and ecosystem changes as plants that are more resistant to ozone can become dominant. The extent of damage to the plant depends on the ambient ozone concentration, the vegetation’s ozone detoxification capacity, the size of the stomata opening (which is impacted by time of day), the moisture content of the soil and meteorological conditions. Ozone can also reduce the capacity of forest ecosystems to provide ecological services such as air filtration and carbon sequestration. These deleterious impacts on vegetation may translate into reduced crop yields (US EPA, 2020c).

To estimate the impacts of transboundary ozone flows on Canadian crop production, ECCC used the agricultural module of the Air Quality Valuation Model (AQVM2) to estimate the change in sales revenue for the agricultural sector for 19 different types of crops in Canada associated with different levels of ground-level ozone. First, modeled ozone concentrations produced by GEM-MACH are fed into exposure-response functions for both the baseline and emission reduction scenarios to estimate yield reductions. Next, these yields are multiplied by the hectares of crops seeded in year 2016 and the crop prices. The process is repeated for each crop type in each of the 68 Census Agricultural Regions in Canada, and aggregated provincially to estimate the potential change in sales revenues for crop producers.

The estimated ozone-related crop yield reductions attributable to Canadian and transboundary emissions are of similar magnitude. Such an outcome can be expected because ozone and ozone precursors (such as NO_x) can travel over larger distances from transboundary sources into Canadian territory. Due to the extensiveness of farmlands in Alberta, Saskatchewan and Manitoba and the significant ozone precursor emissions originating from Alberta, the Prairies’ share of the yield reductions from Canadian emissions is larger. However, estimates of crop yield reductions stemming from transboundary emissions are relatively higher along the Windsor-Quebec City Corridor. It is worth cautioning that the crop yield reductions are not directly proportional to the change in ambient air quality as they also depend on the cultivated area and its proximity to emission sources, so areas experiencing the highest ozone concentrations may not always experience the highest crop yield reductions.

3.8 Methane as an Ozone Precursor

Methane emissions in Canada and the U.S. also contribute to global tropospheric ozone production. Because methane has a long lifetime in the atmosphere, methane concentrations are well-mixed, and a decrease in tropospheric methane emissions anywhere eventually leads to a decrease in tropospheric ozone concentrations everywhere. However, ground-level ozone concentrations, particularly in urban areas, can depend on local and regional chemistry. Impacts of local methane enhancements on ozone production needs further exploration. Several studies have developed approaches for quantifying the effects of methane emission reductions; the most recent is the Global Methane Assessment (UNEP, 2021). The Global Methane Assessment evaluates benefits of a 10 Mt methane emission reduction, which include fewer ozone-related respiratory deaths (428^{Footnote 21}  in the U.S. and 35^{Footnote 22}  in Canada) and fewer asthma-related emergency department visits (1,500 in the U.S. and 224 in Canada)^{Footnote 23} . Other benefits include fewer ozone-related cardiovascular deaths, increased crop production, and less ecological damage to natural vegetation. Although methane emissions near the U.S.-Canada border would not disproportionately impact transboundary conditions due to its high rate of global mixing, it plays a role as a contributor to ozone production in the U.S., Canada.

3.9 Summary

In 2000, the Ozone Annex (Annex 3) to the AQA set out commitments by Canada and the U.S. to reduce emissions of NO_x and VOCs that contribute to transboundary ozone pollution. These commitments aimed to help both countries attain their respective air quality goals, and to protect human health and the environment. At the time the Annex was introduced, the long-term goal was to ensure that ambient concentrations of ozone did not exceed the respective ozone air quality standards in each country.

Canada and the U.S. met their commitments in the Ozone Annex to reduce emissions of NO_x and VOCs from stationary and mobile sources and from solvents, paints, and consumer products. Canada’s national emissions of NO_x and VOCs have decreased by 36% and 49%, respectively, between 1990 and 2020. The U.S. national air emissions of NO_x and VOCs decreased by 70% and 48% respectively between 1990 and 2020.^{Footnote 24}

Ambient ozone concentrations have also declined within the Canada-U.S. border region since the establishment of the Ozone Annex. Annual 4^th highest MDA8 ozone concentrations have decreased by more than 10 ppb at many monitoring stations across Ontario, Quebec, and the Maritimes, and by as much as 20 ppb at some stations in the Great Lakes states and Ohio Valley, where ozone concentrations are highest. Regulatory programs and non-regulatory programs designed to meet emissions reduction commitments in the Ozone Annex, as well as programs designed to meet other goals, regulatory or statutory requirements for Canada and the U.S. individually, have contributed to the reduction in ozone concentrations.

Despite the progress made, areas in southern Ontario and southern Quebec continue to have higher ozone concentrations which approach or exceed the current 2020 ozone CAAQS of 62 ppb. Modeling projections suggest that ozone concentrations are expected to decrease in future years, as new regulations targeted for emission reductions take effect. However, southern Ontario is projected to have 4^th highest MDA8 ozone concentrations that will continue to exceed the 2025 ozone CAAQS (60 ppb) in 2035, based on ECCC air quality modeling.

Transport in the northeastern U.S. and Ohio Valley is favorable for predominantly U.S.-to-Canada transport, and transport from the U.S. continues to contribute a large fraction of anthropogenic ozone in Canada. U.S. EPA and ECCC modeling results of U.S. and Canadian emissions influences are broadly consistent – the largest influence continues to be in the Windsor-Quebec Corridor with emissions from the U.S. also affecting ozone concentrations in southwestern British Columbia, in the greater Vancouver and Victoria area, southern Alberta, the Greater Toronto-Hamilton area, and the Montreal area for both current emissions and future projected emissions. The impact of U.S. emissions on ozone concentrations in Canadian cities within 500 km of the border ranges from less than 5% in Calgary up to approximately 40% in Windsor based on ECCC modeling results. Canada-to-U.S. ozone transport can also be seen to a lesser extent within a narrow band along the border in northern Montana, North Dakota, New York, Vermont, New Hampshire, and Maine with the highest area of influence in northwest Washington.  

Ozone continues to have significant impacts on public health and agricultural production in the U.S. and Canada. Transboundary transport is seen in both countries, but has a larger influence in Canada, leading to greater impacts on air quality and health in Canada from U.S. air pollutant emissions than in the U.S. as a result of Canadian air pollutant emissions. Modeling results predict that transboundary flow of ozone and its precursors from the U.S. to Canada contributes to a significant portion of health impacts in central and Atlantic Canada and is the dominant source of health impacts in Ontario, Quebec, Nova Scotia, New Brunswick, Newfoundland and Labrador, and Prince Edward Island. Analysis also suggests that transboundary flow from the U.S. into Canada also causes crop yield reductions, particularly along the Windsor-Quebec City Corridor.

4 Fine Particulate Matter

What is PM_2.5: PM is a general term used to describe a mixture of solid particles and liquid droplets suspended in the air. PM is characterized according to size, and includes coarse PM (PM₁₀), which consists of particles with diameters that are 10 microns (mm) in diameter and smaller, and fine PM (PM_2.5), which consists of particles with diameters 2.5 mm or smaller. This assessment focuses on PM_2.5, due to its substantial health burden and because this size fraction can remain suspended in the air for several days to weeks, can be transported by winds over large distances, and thus is subject to atmospheric transboundary transport in North America. PM_2.5 directly emitted to the atmosphere (also called primary PM_2.5) originates from sources such as dust, sea spray, or combustion sources. Secondary PM_2.5 is formed via chemical reactions in the atmosphere involving precursor gases originating from a wide variety of transportation, combustion and industrial sources. Key PM_2.5 precursor gases include NH₃, SO₂, NO_x, and VOCs.

Previous bilateral science assessments on PM_2.5 under the AQA: Canada and the U.S. have previously explored adding an annex to the AQA to address transboundary PM_2.5. The two countries agreed to work on the necessary scientific, technical, and regulatory foundations required to consider adding a PM annex to the AQA, including a suggestion to update the Canada-U.S. Transboundary Particulate Matter Assessment that was completed in 2004. The 2004 assessment (PDF) (US EPA & Environment Canada, 2004) found that PM concentrations varied significantly over geographic regions, and that transboundary transport of PM and PM precursors can be significant enough to compromise attainment of national standards in both countries (US EPA & Environment Canada, 2004). In 2011, Canada and the U.S. created the Regulatory Cooperation Council to better align the regulatory approaches of the two countries, where possible.  One of the environment-related initiatives required the two countries “to consider the expansion of the Canada-U.S. AQA to address transboundary PM, the air pollutant most commonly associated with premature mortality, based on comparable regulatory regimes in the two countries” (Regulatory Cooperation Council, 2011).

The 2013 Transboundary Particulate Matter Science Assessment (PDF) was completed by Canada and the U.S. and concluded that “because of the important health and environmental effects associated with PM_2.5, it would be beneficial for both countries to track progress and exchange information relevant to achieving PM_2.5-related emissions reductions, air quality improvement, and program implementation” (ECCC & US EPA, 2016). It further stated that “there would be value in addressing PM_2.5 in some manner under the Agreement.” As such, Canada and the U.S. agreed to include PM_2.5 as part of this review and assessment. This review and assessment reflects the evolution in the scientific understanding of PM_2.5-related impacts on human health, advances in modeling tools used to estimate impacts, and regulations implemented in both countries since 2013.

4.1 Air Quality Standards and Guidelines

As described in Section 3.1, both Canada and the U.S. have established air quality standards to protect human and ecosystem health. Table 4‑1 provides a comparison of the most current CAAQS and the U.S. NAAQS for PM_2.5. The metrics (or statistical forms) of the Canadian and U.S. standards are identical, although the U.S. has primary (health-based) and secondary (welfare-based) standards for PM_2.5.

Table 4‑1. National ambient air quality standards for PM_2.5

4.2 Emission Trends in Primary PM_2.5 and PM_2.5 Precursors

Ambient concentrations of PM_2.5 are affected by both direct emissions of primary PM_2.5 and emissions of precursors, which can lead to the formation of secondary PM_2.5. In many locations within the U.S. and Canada, secondary PM accounts for the majority of PM_2.5 mass, much of which is derived at least in part from anthropogenic precursors. In Canada, the long-range transport of ambient NH₃ contributes to increased PM_2.5 concentrations, particularly in regions where local NH₃ emissions are low, such as southern Ontario and southern Quebec (Environment Canada, 2009). The sections below show trends in direct emissions of primary PM_2.5 and NH₃ as a PM_2.5 precursor. Emissions of other PM_2.5 precursors — NO_x, SO₂, and VOCs – (see Section 2.1 and Section 3.1) have all decreased significantly in both Canada and the U.S. Though PM_2.5 is not covered under the AQA, the U.S. and Canada continue to take domestic actions to reduce PM_2.5 and its precursors.

4.2.1 Canada

Figure 4‑1 shows emissions of primary PM_2.5 in Canada from anthropogenic sources (excluding wildfires). In 2020, approximately 1.4 million metric tons of primary PM_2.5 were emitted in Canada (ECCC, 2022). Dust sources accounted for 62% of total PM_2.5 emissions, with the most dominant dust sources being construction operations (35% of total PM_2.5 emissions) and unpaved roads (25% of total PM_2.5 emissions). Agriculture was the next largest contributor and accounted for 24% of total PM_2.5 emissions, most of which were attributed to crop production.

Commercial/residential/institutional sources accounted for an additional 7% of total PM_2.5 emissions in 2020, with home firewood burning (6%) being the most important contributor. From 1990 to 2020, Canada’s national emissions of primary PM_2.5 decreased 15% (250 thousand metric tons) (ECCC, 2022).

Figure 4‑1. Canadian PM_2.5 Emission Trends. 1990-2020.

As noted above, emissions from wildfires are not included in Canada’s APEI, with the exception of prescribed burning which is included in the “Fire (Excluding Wildfires)” source category. Wildfires are, however, a significant source of primary PM_2.5 in Canada. Using wildfire seasons from 2013 to 2016 as a reference, the total PM_2.5 emitted from wildfires ranges from 790,000 to 1,700,000 metric tons (Munoz-Alpizar et al., 2017). Therefore, wildfire emissions are on the same order of magnitude as total primary PM_2.5 emissions from anthropogenic sources (~1,200,000 to 1,700,000 metric tons).

Emissions of NH₃ in Canada from 1990 to 2020 are shown in Figure 4‑2. In 2020, approximately 487 thousand metric tons of NH₃ were released. NH₃ emissions originated primarily from agriculture, which accounted for 94% (457 thousand metric tons) of total emissions. From 1990 to 2020, NH₃ emissions increased by 24% (93 thousand metric tons), in contrast to the general downward trends of other air pollutant emissions (e.g., NO_x, SO₂, and VOCs) in Canada. NH₃ emissions increased until 2004 and the largely plateaued with some year-to-year fluctuations. Thie increasing trend between 1990 and 2004 was driven by emissions from animal production and the increasing use of inorganic nitrogen fertilizers in crop production. Animal production, which accounts for the majority of NH₃ emissions, steadily increased from 1990 to 2005, followed by a decrease from 2006 to 2011, and has since declined slowly. Emissions from crop production, however, have been steadily increasing since 2006, and now account for 38% of NH₃ emissions.

Figure 4-2. Canadian NH3 emission trends. 1990-2020.