AIR QUALITY MANAGEMENT IN CENTRAL ASIA © 2025 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Some rights reserved This work is a product of the staff of the World Bank. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, links/footnotes, and other information shown in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The citation of works authored by others does not mean the World Bank endorses the views expressed by those authors or the content of their works. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages the dissemination of its knowledge, this work may be reproduced, in whole or in part, for non-commercial purposes as long as full attribution to this work is given. Please cite the work as follows: “World Bank. 2024. Air Quality Management in Central Asia: Summary Report. ©World Bank.” All queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; email: pubrights@worldbank.org. Cover photo credits: (clockwise from top left) Uldis Laganovski (Tashkent, Uzbekistan), EyeEm (Astana, Kazakhstan), Vladimir M. (Ashgabat, Turkmenistan), Collab Media (Bishkek, the Kyrgyz Republic), Bakhrom Tursunov (Samarkand, Uzbekistan), and Andrey Y. (Dushanbe, Tajikistan). Cover design and layout: Vladimir Mirzoyev Air Quality Management in Central Asia Summary Report April 2025 Contents Acknowledgments.......................................................................................................................................................................................................vii Acronyms and Abbreviations.............................................................................................................................................................................viii Executive Summary......................................................................................................................................................................................................1 1. Introduction..................................................................................................................................................................................................................6 2. Air Quality in Central Asia..................................................................................................................................................................................8 2.1. Approach and scope of the analysis............................................................................................................................................................8 2.2. Pollution sources in Central Asia..................................................................................................................................................................8 2.3. Primary and precursor emissions of anthropogenic PM 2.5 ....................................................................................................... 10 2.4. Concentrations of PM 2.5 in ambient air.................................................................................................................................................. 13 2.5. The dispersion of PM 2.5 in the atmosphere......................................................................................................................................... 15 2.6. Population exposure to PM 2.5 and country-wide source apportionment......................................................................... 15 2.7. A focus on cities.....................................................................................................................................................................................................16 2.8. Improving air quality in Central Asian cities........................................................................................................................................ 19 3. The Air Quality Management System in Central Asian Countries.....................................................................................35 3.1. Legal and regulatory framework.................................................................................................................................................................36 3.2. Committed executive and nested planning....................................................................................................................................... 38 3.3. Horizontal and vertical coordination.......................................................................................................................................................39 3.4. Accountability and transparency.............................................................................................................................................................. 40 3.5. Recommendations for Improved AQM................................................................................................................................................. 44 4. Financing Air Quality Improvement.........................................................................................................................................................47 4.1. Cost of inaction..................................................................................................................................................................................................... 47 4.2. Challenges in financing air quality improvement............................................................................................................................ 48 4.3. Options for public funding and financing, including through concessional finance..................................................49 4.4. Incentives for private sector investments...........................................................................................................................................53 5. Conclusion: The Way Forward..................................................................................................................................................................... 58 References...................................................................................................................................................................................................................... 60 ANNEXES Annex I: The GAINS Model Tool......................................................................................................................................................................... 63 Annex II: Emission Estimates for Central Asian Countries for 2020..........................................................................................67 Annex III: Hourly PM 2.5 Concentrations Measured in Various Locations throughout Central Asia..........................69 Annex IV: Cost-effective Air Pollution Control Measures for the Various Emission Source Sectors....................70 Annex V: Air Quality and Ozone Standards in Central Asia, EU, and WHO Guidelines..................................................... 72 Annex VI: Financial Structure for Emissions Reduction-Linked Bond and for Sustainability-Linked Bond in Rwanda ............................................................................................................................................................................................................ 74 FIGURES Figure 1: Contributions of hours with PM 2.5 concentrations above 200 µg/m³ to annual mean concentrations of PM 2.5 , average of 2020–2022.........................................................................................................................................9 Figure 2: Contribution of natural sources (soil and desert dust) to annual mean PM 2.5 concentrations (µg/m³)................................................................................................................................................................................ 10 Figure 3: Per capita emissions of primary anthropogenic PM 2.5 emissions in 2020, by sector..................................... 11 Figure 4: Per capita emissions of PM 2.5 precursors in 2020, by sector........................................................................................ 11 iv BACK TO CONTENTS Air Quality Management in Central Asia Figure 5: Primary energy consumption by sector and fuel, 2020....................................................................................................12 Figure 6: Concentrations of PM 2.5 in 2020 (modeled - left) and 2022 (satellite - right) (µg/m³)....................................14 Figure 7: Concentrations of small particulate matter (PM 2.5) in ambient air in 2020 originating from the anthropogenic emissions of the various countries (µg/m³) ......................................................................................... 15 Figure 8: Origin of country-wide population-weighted PM 2.5 exposure in 2020 (µg/m³)..................................................16 Figure 9: Spatial origin of PM 2.5 exposure in the six cities, 2020 (2018 for Almaty) (µg/m³).......................................... 18 Figure 10: Sector contributions to PM 2.5 exposure in the six cities, 2020 (2018 for Almaty and Astana).............. 18 Figure 11: Contributions in 2020 to PM 2.5 from primary PM 2.5 emissions (by sector) and from secondary PM 2.5 formed in the atmosphere from the precursor emissions of SO 2 , NO x and NH 3....................................................... 19 Figure 12: Source apportionment of population-weighted PM 2.5 exposure in Bishkek in 2020 .................................23 Figure 13: Emission scenarios for Bishkek.................................................................................................................................................... 24 Figure 14: Sources of PM 2.5 exposure in Bishkek in 2020, in baseline projections for 2030 and 2040 baseline, and in the Global clean air scenario in 2040............................................................................................................. 24 Figure 15: The shares of the cost-effective exposure reduction potential in the city of Bishkek that can be achieved through measures in different source sectors within and outside Bishkek in 2040....................26 Figure 16: Source apportionment of population-weighted PM 2.5 exposure in Dushanbe in 2020............................. 27 Figure 17: Emission scenarios for Dushanbe................................................................................................................................................28 Figure 18: Sources of PM 2.5 exposure in Dushanbe in 2020, in the baseline projections for 2030 and 2040 baseline, and in the Global clean air scenario in 2040...........................................................................................................................28 Figure 19: The shares of the cost-effective exposure reduction potential in the city of Dushanbe that can be achieved through measures in the different source sectors within and outside Dushanbe in 2040...................... 30 Figure 20: Source apportionment of population-weighted PM 2.5 exposure in Tashkent in 2020............................... 31 Figure 21: Emission scenarios for Tashkent................................................................................................................................................. 31 Figure 22: Sources of PM 2.5 exposure in Tashkent in 2020, in the baseline projections for 2030 and 2040 baseline, and in the Global clean air scenario in 2040.............................................................................................32 Figure 23: The shares of the cost-effective exposure reduction potential in Tashkent that can be achieved through measures in the different source sectors within and outside Tashkent in 2040..............................................33 Figure A1: Information flow in the GAINS model analysis................................................................................................................... 63 Figure A2: Spatial densities of primary PM 2.5 emissions in 2020, by economic sector (kg/km2)............................... 65 Figure A3: Spatial densities of PM 2.5 precursor emissions in 2020, by economic sector (kg/km2)........................... 65 Figure A4: Hourly PM 2.5 concentrations measured at the US embassy in Dushanbe.........................................................69 Figure A5: Hourly PM 2.5 concentrations measured at the US embassy in Bishkek..............................................................69 Figure A6: Hourly PM 2.5 concentrations measured at the US embassy in Tashkent...........................................................69 Figure A7: A World Bank financial structure for incentivizing emission reductions............................................................ 74 Figure A8: Rwanda - World Bank Sustainability-Linked Bond structure for a national Development Bank........ 74 TABLES Table ES.1: PM 2.5 population-weighted exposure reduction potentials by 2040.....................................................................3 Table ES.2: Recommended actions for improving AQM in CA countries......................................................................................4 Table 1: Estimates of population-weighted exposure to PM 2.5 from windblown soil dust, annual mean concentrations (µg/m³)............................................................................................................................................................................................. 10 Table 2: Comparison of emission estimates of this study and the national estimates that were officially submitted to CLRTAP/EMEP................................................................................................................................................................................. 13 Air Quality Management in Central Asia BACK TO CONTENTS v Table 3: Annual mean concentrations of PM 2.5 measured at US embassies and consulates (µg/m³).......................14 Table 4: The potential exposure reductions in Bishkek from the cost-effective key measures of the 2040 clean air scenario..............................................................................................................................................................................25 Table 5: The potential exposure reductions in Dushanbe from the cost-effective key measures of the 2040 clean air scenario..............................................................................................................................................................................29 Table 6: The potential exposure reductions in Tashkent from the cost-effective key measures of the 2040 clean air scenario..............................................................................................................................................................................33 Table 7: Number of AQ monitoring stations in CA countries...............................................................................................................41 Table 8: Recommended actions for improving AQM in CA countries..........................................................................................45 Table 9: Annual premature deaths attributable to PM 2.5 pollution and associated economic costs in CA...........47 Table 10: Air pollution charges in CA, as % of GDP.................................................................................................................................. 54 Table A1: Emissions of primary PM 2.5 in 2020 (kilotons).......................................................................................................................67 Table A2: SO2 emissions in 2020 (kilotons)...................................................................................................................................................67 Table A3: NOx emissions in 2020 (kilotons).................................................................................................................................................67 Table A4: NH3 emissions in 2020 (kilotons)................................................................................................................................................. 68 Table A5: Cost-effective air pollution control measures for the various emission source sectors...........................70 Table A6: AQ standards in CA, EU, and WHO guidelines...................................................................................................................... 72 Table A7: Ozone standards in CA, EU, and WHO guidelines...............................................................................................................73 BOXES Box 1: Integrated AQM and GHG reduction assessment for Almaty and Astana: Integrated AQM and GHG reduction assessment for Almaty and Astana .....................................................................................................................20 Box 2: Improving urban greening in Bishkek................................................................................................................................................ 22 Box 3: A framework to assess AQM governance and institutional arrangements...............................................................35 Box 4: AQM legal and regulatory framework in the EU..........................................................................................................................37 Box 5: Public funding for AQ in selected countries................................................................................................................................. 50 Box 6: Designing good AQ programs................................................................................................................................................................ 51 Box 7: Revolving mechanism to support transition to clean heating in the Kyrgyz Republic....................................... 55 vi BACK TO CONTENTS Air Quality Management in Central Asia Acknowledgments This report was produced by a core World Bank team led by Elena Strukova Golub (Senior Environmental Economist). The team included Inobat Allobergenova (Natural Resources Specialist), Kirtan Sahoo (Senior Climate Change Specialist), Wei-Jen Leow (Senior Environmental Finance Specialist), Markus Ammann (Senior Consultant), Vasil Borislavov Zlatev (Senior Environmental Consultant), and Wolfgang Schöpp (Senior Consultant). The team is grateful to Nagaraja Rao Harshadeep (Lead Environmental Specialist) and the World Bank’s Disruptive KIDS (Knowledge, Information & Data Services), for their support with data processing, and to Sabrina Zechmeister (Analyst) and Marius Karolinski (Consultant) for technical support. Nigara Abate (Senior Communications and Knowledge Management Specialist) prepared this report for publication. This assessment was produced under the overall guidance of Tatiana Proskuryakova (Regional Director  for Central Asia, World Bank), Sanjay Srivastava (Regional Manager, Environment Department, Europe and Central Asia Region, World Bank), and Sameh Naguib Wahba (Regional Director, Sustainable Development, Europe and Central Asia Region, World Bank). The report was produced in collaboration with the respective Ministries of Ecology and Natural Resources and Environmental Protection Committees from Central Asian countries. The team also expresses profound appreciation to Prof. Jay Turner (McKelvey School of Engineering, Washington University) and to our colleagues from the United Nations Environment Programme for their expert support and knowledge sharing. The team is grateful to Urvashi Narain (Program Leader, World Bank), Pedro Arizti (Senior Public Sector Specialist, World Bank), Szilvia Doczi (Senior Energy Economist, World Bank) and Julie Rosenberg (Senior Economist, SCADR) for valuable comments and guidance on this report and to the many other contributors who reviewed, provided comments, edited, and designed this report. The team is also grateful to Linh Van Nguyen (Senior Program Assistant) and Lisa Fonick Haworth (Senior Program Assistant) for their project management support. This report was supported by multi-donor trust funds from PROGREEN and the Climate Support Facility (CSF). Air Quality Management in Central Asia BACK TO CONTENTS vii Acronyms and Abbreviations AQ Air Quality Integrated Environmental IEP Permit AQM Air Quality Management Institute for Applied Systems ASW Ash and Slag Waste IIASA Analysis BAT Best Available Techniques KPI Key Performance Indicator CA Central Asia/Central Asian Maximum Allowed Comprehensive Air Quality MAC CAMx Concentration Model with extensions MoF Ministry of Finance Convention on Long-rage CLRTAP Nationally Determined Transboundary Air Pollution NDC Contribution CHPP Combined Heat and Power Plant O&M Operation and Maintenance ELV Emission Limit Value Organisation for Economic European Monitoring and OECD Co-operation and Development EMEP Evaluation Programme of the Participating Financial LRTAP Convention PFI Intermediary EQI Environmental Quality Indicator Fine particulate matter with an ESCO Energy Service Company PM 2.5 aerodynamic diameter of less EU European Union than 2.5 µg/m³ Development Finance Fine particulate matter with an DFI Institution PM 10 aerodynamic diameter of less than 10 µg/m³ DP Development Partner PPP Public-Private Partnership FMI Finnish Meteorological Institute SDS Sand and Dust Storms Greenhouse Gas - Air Pollution GAINS Interactions and Synergies Model SFH Single-Family Home GBD Global Burden of Disease SMEs Small and Medium Enterprises GDP Gross Domestic Product UN Convention to Combat UNCCD Desertification Atmospheric chemistry model GEOS- driven by meteorological United States Environmental US EPA Chem input from the Goddard Earth Protection Agency Observing System VOC Volatile Organic Compound GHG Greenhouse Gas VSL Value of Statistical Life IEA International Energy Agency WHO World Health Organization viii BACK TO CONTENTS Air Quality Management in Central Asia Executive Summary This report aims to enhance the understanding of the priorities, needs, and solutions for improving air quality (AQ) in Central Asia (CA) through local action and regional collaboration. In particular, the report focuses on the key components of holistic air quality management (AQM): (a) evidence-based analytics to identify the main sources of air pollution in CA, (b) application of modern tools to assess the impact of cost-effective measures to improve AQ, (c) assessment of the institutional and governance setup for AQM in CA and recommendations to strengthen it, and (d) approaches to financing AQ improvement. Given the lack of comprehensive systematic and validated emission inventories of all PM 2.5 precursor emissions, the technical assessment employs the regional emission inventory of the Greenhouse Gas - Air Pollution Interactions and Synergies (GAINS) model. Input data were updated for this study based on recent energy statistics and relevant national surveys. This report addresses emissions and the regional transboundary flows of pollution between Kazakhstan, the Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan. Subsequently, the resulting PM 2.5 concentrations in ambient air throughout CA were computed with the atmospheric chemistry and transport calculations of the GAINS model. Employing the source apportionment results of the GAINS model, the analysis then examines the contributions to PM 2.5 population exposure. The report also presents source apportionment analyses for important air pollution hot spots in CA: Dushanbe (Tajikistan), Bishkek (the Kyrgyz Republic), Tashkent (Uzbekistan), Samarkand (Uzbekistan), Astana and Almaty (Kazakhstan). Key findings from the technical assessment of air quality in Central Asia Poor AQ in CA is predominantly a problem in urban areas, where PM 2.5 1 concentrations are typically 6 to 12 times above the World Health Organization (WHO) guideline of 5 µg/m 3 . Annual average PM 2.5 concentrations in the studied urban areas of Dushanbe, Bishkek, Tashkent, Samarkand, Almaty, and Astana are clearly above the WHO guideline value. PM 2.5 exposure in CA cities originates from various emission sources in different economic sectors. Investing in AQ improvements has multiple benefits. It reduces healthcare costs, increases productivity, and prevents premature deaths while also enhancing cognitive development in children and mobilizing government revenue for green innovation. Improved AQ boosts agricultural yields by reducing pollution-related crop damage and enhances urban livability by improving public health, reducing emissions, and fostering sustainable development. These comprehensive benefits highlight the significant economic advantages of cleaner air, contributing to higher quality of life, economic competitiveness, and stronger local economies. The health costs of PM 2.5 ambient air pollution in CA are estimated to range between US$15.2 and US$21.7 billion per year, which is equivalent to 3–5 percent of regional gross domestic product (GDP) in 2022. In addition to health costs, air pollution affects cities’ development in terms of potentially reduced competitiveness, lower levels of investment, and, overall, lower quality of life. Understanding the sources that contribute to population exposure to PM 2.5 and implementing measures to reduce air pollution can thus save lives and at the same time deliver economic and developmental benefits. 1 Particulate matter with an aerodynamic diameter of less than 2.5 µg/m³. Air Quality Management in Central Asia BACK TO CONTENTS 1 In many CA cities, the contribution of soil and desert dust to PM 2.5 concentrations is more significant than in most other regions in the world. Soil and desert dust accounts for 20–50 percent of total exposure in CA cities even on an annual basis. On the flip side, this also means that 50–80 percent of total exposure in the cities is caused by anthropogenic emissions, which can be controlled through dedicated policy interventions. Effective efforts to manage urban AQ need to address emission sources in the entire airshed that extends outside the immediate city jurisdictions. Although cities emerge as the major pollution hot spots in CA, only a limited share of PM 2.5 exposure within the city boundaries (typically 10–50 percent) originates from local emission sources. In addition, secondary PM 2.5 formed from PM 2.5 precursor emissions such as sulfur dioxide (SO 2 ), nitrogen oxides (NO x ), and ammonia (NH 3 ) accounts for 10–50 percent of the exposure from anthropogenic sources (excluding natural soil dust). Since the chemical formation of secondary PM 2.5 in the atmosphere takes some time, the importance of secondary PM 2.5 increases with distance. Thus, controls of the precursor emissions across the entire city airsheds will be critical for reducing PM 2.5 exposure in CA cities. Effective reductions of secondary PM 2.5 need to address emissions of SO 2 (mainly from coal combustion), NO x (predominantly from mobile sources), and NH 3 (coming from agriculture). These gaseous emissions react in the atmosphere with NH 3 (emitted from livestock manure management and fertilizer application) forming ammonium sulfate and ammonium nitrate, the main components of secondary inorganic PM 2.5 . The largest anthropogenic contribution to PM 2.5 concentrations in all cities in the study comes from solid fuel combustion for heating in the residential sector. Emissions originate from a small share of households in suburban areas and from rural households in the surrounding regions. PM 2.5 exposure from mobile sources is about 10 percent of total exposure in each city. Primary PM 2.5 emissions from urban vehicles account only for about half of this share, while the rest consists partly of secondary PM 2.5 (formed from NO x emissions) and/or is transported into the city from surrounding areas. The assessment of cost-effective measures to improve AQ in Bishkek, Dushanbe, and Tashkent by 2040 reveals that achieving WHO interim targets is possible, but there is a need for context- specific approaches to urban AQM. While cost-effective measures targeting the urban areas of Bishkek and Dushanbe will deliver the bulk of AQ benefits by 2040, interventions beyond the urban borders of Tashkent show the largest potential to achieve AQ improvements. In all the cities studied, except Tashkent, the most cost-effective interventions beyond the urban borders involve reducing emissions from residential heating, whereas in Tashkent, interventions to reduce industrial pollution come up as the most cost-effective. In Tashkent and Bishkek, road transport measures at the urban level are prioritized from a cost-effectiveness perspective (see Table ES.1). Thus, interventions in these sectors could be prioritized for immediate actions. The report provides a summary of the most cost-effective measures for each of the studied cities to guide implementation actions. As replacement of polluting passenger vehicles and space heating appliances at scale takes time and resources, authorities might consider interventions in other sectors that could provide air quality benefits quicker. Even though, industry is not the main source of PM 2.5 pollution in all studied cities, authorities should also consider cleaner industrial production policies by strengthening the permitting process, reducing the emission limit values (ELVs) for large industries and power plants and incentivizing cleaner production methods. Another area where authorities might have more resources and control over the implementation of emission reduction measures quickly is public transport. Authorities could secure funding for cleaner public transport and for measures to improve the attractiveness of public transport such as optimizing public transport routes and timetables and providing faster commute for public transport users. 2 BACK TO CONTENTS Air Quality Management in Central Asia Table ES.1: PM 2.5 population-weighted exposure reduction potentials by 2040 Bishkek Dushanbe Tashkent PM 2.5 population-weighted exposure reduction potential of urban measures Share of total reduction potential 58% 76% 27% Key sector delivering the reductions Road transport Residential heating Road transport PM 2.5 population-weighted exposure reduction potential of outside (urban) measures Share of total reduction potential 42% 24% 73% Key sector delivering the reductions Residential heating Residential heating Industry Projected PM 2.5 population-weighted annual exposure after implementing all cost-effective measures in 2040, in µg/m 3 22 27 24 Source: Own elaboration, IIASA, World Bank 2023a, World Bank 2023b, World Bank 2023d. Cities in CA can improve urban AQ only to a limited extent on their own. Efforts to achieve international AQ standards need to involve measures in surrounding regions and sometimes even in other countries. This requires a completely new, airshed-based governance approach to AQM, geared toward cooperation between the jurisdictions of different regions and a clear distribution of responsibilities across different governance levels, that is, from the national level over provinces to individual cities. Air quality management system in Central Asian countries Key components of the AQM system in CA countries need strengthening to enable effective governance for improved AQ. Given the shared historical development of CA, AQM systems in the different CA countries share some common challenges: ለ AQM in CA is primarily top-down driven. Local AQ plans or strategies are absent, and hence local authorities’ role in strategic AQM planning is limited. ለ The airshed approach to AQM that considers impacts on air quality of all major emission sources, irrespective of the sources’ location with respect to administrative divisions, is not applied consistently in CA. Therefore, the governance framework for AQM needs to be updated to adopt the airshed approach and to ensure cross-sector coordination and long-term commitment. ለ The efficiency of the permitting system in CA is questionable as emission limit values (ELVs) are generally set at a high level and consider the historical level of emissions of enterprise. ለ The setting of AQ standards in CA countries uses an outdated approach that sets maximum allowed concentrations (MACs) for over 600 pollutants, which shifts the focus away from the main substances that cause the largest health impacts and does not allow for proper assessment of improvements in AQ. ለ AQ monitoring in CA is still dominated by manual sampling and subsequent laboratory analysis of the samples as well as limited use of mobile AQ stations. ለ Structuring of emission inventories in CA is non-consistent and follows outdated approaches which generally focus only on industrial and mobile sources (transport). Thus, emission inventories omit an important source of urban air pollution such as residential heating. ለ The capacity for AQ modeling in CA is limited. Air Quality Management in Central Asia BACK TO CONTENTS 3 Strengthening the institutional, legal, and policy framework for AQM; updating AQ standards and approaches to emission inventories; expanding the AQ monitoring and modeling capabilities; increasing stakeholder awareness and engagement; and supporting regional cooperation in AQM are shared priorities for CA countries. Table ES.2 outlines recommendations for actions applicable to the region while recognizing that countries within the region might have advanced more on certain actions listed in the table. Table ES.2: Recommended actions for improving AQM in CA countries Institutional, legal, and policy framework ለ Strengthen governmental roles, responsibilities, and structures to support an effective AQM system. ለ Establish interministerial AQ Coordination Committees. ለ Strengthen strategic AQM planning on national and local levels adopting the airshed approach. ለ Review and update sectoral legislation for the key emitting sectors. AQ standards ለ Reassess the list of pollutants for which AQ standards are adopted. ለ Update AQ standards in line with WHO recommendations. Technical components of AQM ለ Expand and upgrade AQ monitoring. ለ Strengthen technical capabilities of the AQ monitoring networks. ለ Update and strengthen the emission inventory systems to meet international best practices. ለ Strengthen capacities for AQ modeling. Communication ለ Strengthen stakeholder engagement. ለ Support awareness raising and education. Regional cooperation ለ Set up a platform/mechanism for information and knowledge exchange. ለ Share the implementation of some AQM activities on a regional scale. ለ Agree on actions to reduce transboundary air pollution. Source: Original compilation. Financing air quality improvement The mechanisms to finance emission reduction measures depend on the characteristics of the emission source in question. For instance, large point sources such as industrial enterprises have established revenue streams. Although, in most cases, emission reduction measures entail high capital costs, those costs can be passed on to the consumers or internalized within their financial structures. On the other hand, implementing emission reduction measures for small, decentralized sources such as non-district residential heating with large social implications, lack the capacity to generate significant revenue streams and absorb extra costs; hence, public sector involvement in the financing of such measures is required. Pollution charges, while a common tool in environmental policy, are often insufficient as an incentive for polluters to invest in pollution control. From the start of its post-Soviet era, CA countries have implemented pollution charges to incentivize polluters to reduce their emissions and to fund environmental cleanup and protection initiatives. Though effective in theory, pollution charge systems in their current form have not been an effective tool. Comparative studies have found that 4 BACK TO CONTENTS Air Quality Management in Central Asia ELVs in CA, particularly for air pollutants from large combustion plants, were more lenient than in many Organisation for Economic Co-operation and Development (OECD) countries (OECD 2019). A more effective system would involve shifting the focus of the charges from penalizing noncompliance to re-incentivizing compliance (for example, through a system of rebates) and avoiding undue overlap between policy instruments, such as a fuel excise based on the content of carbon as a pollutant. In addition to or instead of pollution charges, most advanced AQM systems (for example, in European Union [EU] and the US) adopt legislation that obliges enterprises to implement, as a minimum, benchmark levels of resource efficiency and clean industrial production. Public sector funding remains a key part of financing AQ improvement. While the main areas funded by the public sector relate to staffing of AQ departments and agencies, as well as for environmental monitoring, public sector funding plays a key role in financing emission reduction measures for decentralized pollution sources (for example, non-district residential heating and cooking) that lack strong revenue streams to support market-based approaches for financing such measures. In those cases, public sector funding can also attract and support private participation in the co-funding of emission reduction measures. In addition, fiscal reforms such as the elimination and/or repurposing of subsidies could provide additional funding for AQ improvement measures and could influence businesses and households to make choices that are better aligned with the AQ improvement agenda. Private sector funding and utilization of new and innovative financing mechanisms are essential for achieving substantial emission reductions, especially when there is a need for large-scale technology investments. Clear regulations that unequivocally point to a move toward cleaner production and de-risking mechanisms such as guarantees, concessional loans, and blended finance are needed to support private investment in AQ improvement measures. Public-private partnerships (PPPs) are another potential instrument that can be utilized to lower the financial risk of reducing emissions for private investments into infrastructure projects. New and innovative funding mechanisms include green outcome bonds, result-based payments, sustainability-linked debts— all which link payments or debt conditions on the achievement of predefined targets for emission reductions or other environmental outcomes. Integrating AQ financing opportunities with climate change mitigation funding can also support more active private sector participation in the funding of emission reduction measures. There is a strong alignment between the climate change mitigation agenda and the local AQ agenda. Moreover, development finance institutions (DFIs) are increasingly supporting integrated projects and actively promoting new models for AQ financing. A strategic approach to financing that combines policy alignment, tailored financial tools, and private sector engagement is needed to support AQ improvement in CA. Essential components of sustainable financing for AQ improvement include integrating AQM investments with climate action to allow access to broader funding opportunities, creating supportive environments through policy reforms and robust institutions, and employing customized financial mechanisms that address specific pollution sources to attract private investment and manage risks. Air Quality Management in Central Asia BACK TO CONTENTS 5 1. Introduction Over the last decade, Central Asian (CA) countries 80 percent of total damage costs of air pollution experienced increasing levels of ambient air (De Bruyn 2020). Notably, PM 2.5 in ambient air pollution with a growing number of days marked originates not only from primary PM 2.5 emissions by extremely high pollution concentrations, (for example, soot and dust). A significant fraction most noticeable in urban areas. Reported annual (up to 50  percent) is formed in the atmosphere mean concentrations of fine particulate matter through chemical reactions from so-called PM 2.5 (PM 2.5 ) are 6–12 times higher than the guideline 2 precursor emissions, that is, sulfur dioxide (SO 2 ), value of the World Health Organization (WHO) of 5 nitrogen oxides (NO x ), ammonia (NH 3 ), and non- µg/m³ (WHO 2021). Air pollution, and particularly methane volatile organic compounds (VOCs). PM 2.5 pollution, leads to economic losses due Due to their small size, PM 2.5 and its precursor to its adverse health impacts, including links emissions remain in the atmosphere for about to premature death. The latest (2021) Global one week and are transported over significant Burden of Disease (GBD) global study3 estimates distances during this time, typically up to 1,000 that over 65,000 premature deaths in CA could km. Thus, at any given location, PM 2.5 in ambient be attributed to ambient PM 2.5 pollution annually. air originates from a large area beyond the The health costs of PM 2.5 ambient air pollution in boundaries of cities, districts, and provinces, and CA range between US$15.2 and US$21.7 billion in many cases even from other countries. per year (see Chapter 4). In addition to the health Unlike many other regions in the world, satellites costs, air pollution affects other parameters and global model calculations suggest that CA such as cities’ competitiveness, quality of life, as has large spatial differences in concentrations well as attractiveness to investors and to tourists of fine particulate matter (PM 2.5) between cities alike. 4 Therefore, failure to take measures on air and rural areas. In general, CA countries are pollution incurs both health and economic costs sparsely populated, resulting in comparably low at the local and national levels. spatial emission densities in rural areas despite At the same time, there is only an incomplete sometimes high emissions in industrial centers. understanding of the nature of the pollution However, air pollution in urban agglomerations problem in CA, with very limited monitoring of air reaches considerable levels, often attributed quality (AQ) — mainly in major cities — and without to emissions from mobile sources. In addition, systematic knowledge of the sources of pollution. windblown dust and sea salt originating from As a legacy of the Soviet Union, the air quality arid regions cause an increasing frequency of management (AQM) systems of CA countries regional episodes with extremely high PM 2.5 used to record the quantities of emissions of concentrations. Moreover, the proximity of major several dozen air polluting substances from cities in the border regions of Tajikistan, the some key sources, while little emphasis was Kyrgyz Republic, Turkmenistan, Uzbekistan, and placed on the completeness and validation of Kazakhstan — all well within the typical transport reported emissions and their relevance to public distance of PM 2.5 in the atmosphere — is likely health. In contrast, worldwide scientific evidence leading to transboundary flows of pollution, reveals PM 2.5 as the air pollutant with the largest which would require regional approaches for health impacts (IHME 2019), causing more than effective AQM. 2 Fine particulate matter with an aerodynamic diameter of less than 2.5 µg/m³. 3 Global Burden of Disease Study 2021 (GBD 2021) Data Resources | GHDx. 4 https://documents1.worldbank.org/curated/en/115861550150961022/pdf/WPS8740.pdf. 6 BACK TO CONTENTS Air Quality Management in Central Asia Although the understanding of the nature of the improve AQ, (c) assessment of the institutional air pollution problem in CA is still imperfect, there setup for AQM in CA and recommendations to are strong indications that conditions are very strengthen it, and (d) approaches to financing different to those in other countries and regions AQ improvement. The report provides a in the world, in terms of emission densities, the comparative analysis of AQM in CA. It identifies relative importance of the various source sectors regional similarities while also highlighting that contribute to population exposure, the extent differences in the emission sources’ intensities to which transboundary sources contribute, and as well as the potential cost-effective measures the importance of natural sources. Thus, generic to improve AQ in selected urban pollution recipes that worked well in other countries hotspots. to improve AQ might not be effective in CA. Chapter 2 summarizes the findings from the However, methods have been developed and are analytical assessment of AQ in CA, including now available to identify the major sources that emission estimates, spatial distribution of PM 2.5 cause deteriorated AQ in a region and to define pollution, and key sources of PM 2.5 population priority sectors and interventions for cost- exposure. Chapter 2 then focuses the analysis effective improvements of AQ. on city level by establishing a baseline and 2040 To support a more holistic understanding clean air scenarios for three major cities in the of AQM, this report aims to enhance the region. The clean air scenario suggests cost- understanding of the priorities, needs, and effective measures that can lead to improved AQ solutions for improving AQ through local action meeting international standards in the studied and regional collaboration and identifying cost- cities. Chapter 3 analyzes the institutional setup effective measures to enhance AQ. In particular, of AQM in CA countries considering the main the report focuses on the key components AQM building blocks. Chapter 3 ends with a set of holistic AQM: (a) evidence-based analytics of recommendations to strengthen AQM in CA. to identify the main sources of air pollution in Chapter 4 considers approaches to finance AQ CA, (b) application of modern tools to assess improvement measures and explores potential the impact of cost-effective measures to innovative financing methods for AQ funding. Air Quality Management in Central Asia BACK TO CONTENTS 7 2. Air Quality in Central Asia 2.1. Approach and scope of the analysis To enhance the understanding of the priorities from a common starting point (i.e. base year) and needs for improving AQ in CA through local across CA countries and to utilize the most action and regional collaboration, this chapter detailed energy statistics available at the identifies the main sources responsible for the start of this study, input data for the 2021 exposure of the population to harmful PM 2.5 in review of the Gothenburg Protocol were used. ambient air. Average annual PM 2.5 exposure is The data were further refined for this study the most meaningful metric to assess the health based on recent energy statistics and relevant impact of air pollution requiring continuous national surveys to reflect changes since the monitoring data of adequate annual coverage 2021 review. Subsequently, the resulting PM 2.5 for CA cities. Yet, continuous automatic AQ concentrations in ambient air throughout CA monitoring in CA is limited and varies across were computed with the atmospheric chemistry CA countries. U.S. Embassies in CA conduct AQ and transport calculations of the GAINS model monitoring using reference U.S. Environmental and compared against the harmonized and Protection Agency (EPA) methods. Therefore, to validated ambient AQ observations at the US standardize the assessment of air pollution in CA embassies in the region and available local cities, data from the U.S. Embassy automatic AQ measurements. Annex I provides detailed monitoring stations were used in this study. information on the GAINS model. To conduct a source apportionment, the analysis Employing the source apportionment results of assesses the impact of PM 2.5 pollution from the GAINS model, the analysis then examines emission-generating activities, the resulting the contributions to PM 2.5 population exposure. emissions of all PM 2.5 precursor substances, First, the analysis addresses the regional to concentrations of PM 2.5 in ambient air due situation, quantifying the contributions of to the atmospheric chemistry and transport. the various emission source sectors to the Computed concentrations are then compared overall population exposure in each country. against available observations. The validated Subsequently, the analysis zooms into the quantification of the pathway of pollution enables main cities, exploring the contributions from a source apportionment analysis that reveals, for emissions within the city, from the surrounding the various countries and for the main cities, the areas within the same country, from other spatial and sectoral contributions of the various countries, and from natural sources. emission sources to PM 2.5 population exposure. This report addresses emissions and the Given the lack of comprehensive systematic regional transboundary flows of pollution and validated emission inventories of all between Kazakhstan, the Kyrgyz Republic, PM 2.5 precursor emissions, this assessment Tajikistan, Turkmenistan, and Uzbekistan. It employs the regional emission inventory of the then addresses the situation in urban pollution Greenhouse Gas - Air Pollution Interactions and hot spots, presenting source apportionment Synergies (GAINS) model (Amann et al. 2011), analyses for Dushanbe (Tajikistan), Bishkek the central analytical tool of the Convention (the Kyrgyz Republic), Tashkent (Uzbekistan), on Long-range Transboundary Air Pollution Samarkand (Uzbekistan), Almaty (Kazakhstan), (CLRTAP). To allow a comparative analysis and Astana (Kazakhstan). 8 BACK TO CONTENTS Air Quality Management in Central Asia 2.2. Pollution sources in Central Asia Soil and desert dust CA countries are highly susceptible and Figure 1: Contributions of hours with PM 2.5 vulnerable to the risk of sand and dust storms concentrations above 200 µg/m³ to annual (SDS). The ‘dust belt’, the primary and continuous mean concentrations of PM 2.5 , average of 2020–2022 source of storms in CA, stretches from the western deserts to the southern deserts, 60 including the Caspian Sea deserts, the Kyzyl- Annual mean concentration (µg/m3) 50 Kum, Aralkum deserts (Aral Sea region), and Southern Balkhash deserts (Middleton 2017). 40 According to the UN Convention to Combat Desertification (UNCCD), the frequency of dust 30 storms in Tajikistan has surged more than tenfold over the past 30 years. Human-induced factors 20 influencing or contributing to the increased occurrence of SDS include unsustainable land 10 use practices, excessive water extraction, soil salinization, deforestation, and unsustainable 0 agriculture and farming methods leading to Dushanbe Bishkek Tashkent Almaty soil erosion (UNEP 2017). The degradation Peak hours > 200 µg/m 3 or elimination of vegetation cover, primarily Rest of the year through wildfires and overgrazing, has been Source: Original elaboration. identified as the primary process triggering the formation of dust storms in the pre-Aral and Betpak-Dala regions (Nobakht, Shahgedanova, in the atmosphere would require chemically and White 2021) 50–100°E. speciated AQ monitoring extending over long periods with subsequent chemical analysis of The impacts of dust storms, especially in the the measured components. summer, are evident in peak hourly PM 2.5 con- centrations (that is, concentrations above 200 Unfortunately, such measurements are unavail- µg/m³) measured in CA cities (see Annex III for able in CA. As an alternative, estimates of the details). source strengths of such dust storms derived from satellite imagery (Nobakht, Shahgedano- Such episodic events can be easily detected by va, and White 2021) 50–100°E combined with continuous AQ monitors at specific locations. regional atmospheric dispersion calculations For the monitoring data obtained at the US (Ge et al. 2021; McDuffie et al. 2021) ecosys- embassies, hours with PM 2.5 concentrations tems, and the climate. Of concern are high above a threshold of 200 µg/m³ contribute concentrations and\ndeposition of reactive ni- between less than 1 percent (in Almaty) and trogen (Nr have projected the regional contribu- 13 percent (in Bishkek) (see Figure 1) to the tions made by soil dust to PM 2.5 concentrations total annual mean concentrations, the relevant throughout the year. An example of a regional metric for public health concerns. pattern of PM 2.5 concentrations in CA comput- However, such extreme peak events constitute ed with the European Monitoring and Evaluation only the most visible effect of windblown sand Programme of the LRTAP Convention (EMEP) and desert dust, which also prevails at lower global atmospheric dispersion model (Simpson wind speeds throughout the year. Observations et al. 2012) is shown in Figure 2. of the long-term concentrations of soil dust Air Quality Management in Central Asia BACK TO CONTENTS 9 Figure 2: Contribution of natural sources (soil and desert dust) to annual mean PM 2.5 concentrations (µg/m³) 46 50 45 44 40 35 42 30 25 40 20 15 38 10 5 36 0 55 60 65 70 75 80 Source: Original GAINS analysis based on the EMEP Eulerian dispersion model. Available modeling studies employ different soil and desert dust) to population-weighted methodologies and input data. Despite some PM 2.5 exposure in CA (Table 1). Overall, soil and variations in results, there is good agreement desert dust accounts for 20–50 percent of to- on the magnitude of the overall contribu- tal exposure in CA cities as discussed in Sec- tions of natural sources (that is, sea salt and tion 2.7. Table 1: Estimates of population-weighted exposure to PM 2.5 from windblown soil dust, annual mean concentrations (µg/m³) EMEP model estimate GEOS-Chem 5 model estimate used for this study as employed in McDuffie et al. 2021 Kazakhstan 5.5 8.2 Kyrgyz Republic 8.6 9.5 Tajikistan 14.7 13.6 Turkmenistan 24.3 17.9 Uzbekistan 11.2 15.1 Source: Original analysis, based on McDuffie et al. (2021); Simpson et al. (2012). 2.3. Primary and precursor emissions of anthropogenic PM 2.5 The emission inventories developed for this re- representative of the emission characteristics port were obtained through combining recent and applied pollution controls in the region. The energy, transport, industrial, and agricultural sta- structured emission inventories for this study in- tistics for CA countries (for example, FAO 2023; dicate large differences between the CA coun- IEA 2023). The emission factors of the GAINS tries in emission densities and the composition database used in the emission calculations are of sources of PM 2.5 precursor emissions (see 5 Atmospheric chemistry model driven by meteorological input from the Goddard Earth Observing System. 10 BACK TO CONTENTS Air Quality Management in Central Asia Annex II). Note that the inventories shown in this largest amount of PM 2.5 emissions. However, section refer to total national emissions and are mobile sources make the largest contributions not applicable to individual cities. to NO x emissions. 6 Coal combustion (in the residential sector and for power generation) The residential sector constitutes the largest is the major source of SO 2 emissions, and source of primary PM 2.5 emissions in all countries agricultural activities (livestock farming and except Turkmenistan, followed by agricultural fertilizer application) dominate NH 3 emissions. activities, that is, crop burning and soil dust Apart from Kazakhstan, per capita emissions are (Figure 3). This finding is in clear contrast to like those in the European Union (EU) (Figure 3 the widespread belief in the region that mobile and Figure 4). sources (diesel vehicles) would contribute the Figure 3: Per capita emissions of primary anthropogenic PM 2.5 emissions in 2020, by sector 14 12 10 kg/person 8 6 4 2 0 Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 China India Power sector Industrial combustion Industrial processes Residential Mobile sources Agriculture Waste Source: Original GAINS analysis. Figure 4: Per capita emissions of PM 2.5 precursors in 2020, by sector 50 SO2 NOx NH3 40 30 kg/person 20 10 0 Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 China India Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 China India Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 China India Power sector Industrial combustion Industrial processes Residential Mobile sources Agriculture Waste Source: Original GAINS analysis. 6 Based on the available statistics, per capita emissions from mobile sources in Turkmenistan emerge as much higher than in other countries, although further validation seems warranted. Air Quality Management in Central Asia BACK TO CONTENTS 11 Figure 5: Primary energy consumption by sector and fuel, 2020 The differences in the sectoral compositions of PM 2.5 precursor KAZAKHSTAN 4000 emissions (Figure 4) are rooted in the specific fuel consumption 3000 structures of the various 2000 countries (as reported in the PJ International Energy Agency 1000 [IEA] energy statistics). For 0 instance, energy consumption in Power Industry Residential Mobile Sum Turkmenistan and Uzbekistan is KYRGYZ REPUBLIC dominated by natural gas, while 200 coal and, especially, biomass (in 150 the residential sector) account for about half of the energy 100 PJ consumption in the Kyrgyz 50 Republic and Tajikistan (Figure 5). 0 The latter fuels are major sources Power Industry Residential Mobile Sum of primary PM 2.5 emissions, and combustion of oil products by TAJIKISTAN 250 mobile sources is a prominent source of NO x emissions. 200 150 Annex I further provides an PJ 100 overview of spatial emission densities estimated throughout 50 CA, a region characterized by 0 Power Industry Residential Mobile Sum very low population density and thus also relatively lower TURKMENISTAN spatial emission intensities 1000 compared to other regions of 800 the world. This approach allowed 600 to approximate emissions in CA PJ 400 (Table 2). 200 The emission figures presented 0 above have been derived from Power Industry Residential Mobile Sum the databases of the GAINS UZBEKISTAN model, using the methodol- 2500 ogy adopted by the CLRTAP. 2000 This methodology applies to a 1500 comprehensive set of emission PJ 1000 sources based on internationally 500 available activity statistics and 0 representative emission factors Power Industry Residential Mobile Sum that reflect applied emission Coal Biomass Liquid Gas Renewable Other controls. However, comparisons with local emission inventories Source: Original GAINS analysis based on IEA 2023. 12 BACK TO CONTENTS Air Quality Management in Central Asia are hampered, among others, by a limited avail- the CLRTAP with obligations to submit com- ability of complete national inventories. Only Ka- plete emission inventories for the key pollutants zakhstan and the Kyrgyz Republic have ratified according to international standards (Table 2). Table 2: Comparison of emission estimates of this study and the national estimates that were officially submitted to CLRTAP/EMEP PM 2.5 SO 2 NO x Last Last Last Estimate Estimate Estimate officially officially officially for 2020 for 2020 for 2020 reported reported reported for this for this for this figure to figure to figure to study study study LRTAP LRTAP LRTAP Kazakhstan 241.1 230.9 847.9 1575.0 667.1 646.2 Kyrgyz Republic a 27.2 8.7 43.4 15.3 56.8 56.0 Tajikistan 54.5 — 50.0 — 59.4 — Turkmenistan 19.0 — 153.9 — 241.5 — Uzbekistan 71.6 — 80.3 — 414.8 — Source: Original compilation. Note: Data from the 2021 review of the Gothenburg Protocol a. The latest reported numbers of the Kyrgyz Republic refer to 2017. The GAINS estimates for Kazakhstan in this recent reviews for Kazakhstan and the Kyrgyz report are in close agreement with Kazakhstan’s Republic concluded that while the inventory nationally reported totals for PM 2.5 and NO x methodologies are generally in line with the as well as for NO x of the Kyrgyz Republic. 7 EMEP EEA Emission Inventory Guidebook, However, there are striking disagreements the submitted inventories were incomplete as for the other pollutants, likely related to the several important emission sources were not observations made by the CLRTAP in-depth included. Also, several internal inconsistencies reviews of the national inventories, which were noted, among others, in reported time have been conducted by independent teams series and the use of more advanced Tier 2 of international experts. In particular, the methods for key categories. 2.4. Concentrations of PM 2 .5 in ambient air There is only a limited number of monitoring Bank reports for the individual cities (World stations in CA, mainly located in big cities, Bank 2023a, 2023b). At the US embassies, that provide quality-controlled measurements the highest concentrations were measured of PM 2.5 concentrations in ambient air with in Dushanbe and Tashkent (55 µg/m³ and 40 sufficient time coverage to be representative µg/m³, respectively), clearly above the WHO of the full year. Typical levels of urban PM 2.5 in interim target 1 for PM 2.5 of 35 µg/m³. Observed CA in the period from 2020–2022 are shown levels in Bishkek and Almaty were around this in Table 3, measured at the US embassies interim target, but generally exceeding AQ and consulates following international quality limit values (for instance, the Kyrgyz Republic’s assurance practices. 8 Variations within cities annual average limit value is 25 µg/m³, while the and comparisons with national monitoring updated annual average limit value in the EU is data are discussed in the respective World 10 µg/m³). 7 Reports are available on https://www.ceip.at/review-of-emission-inventories/technical-review-reports. 8 https://gispub.epa.gov/airnow/index.html?tab=3. Air Quality Management in Central Asia BACK TO CONTENTS 13 Table 3: Annual mean concentrations of PM 2.5 measured at US embassies and consulates (µg/m³) 2020 2021 2022 Average Almaty (Kazakhstan) 36.4 35.1 28.6 33.4 Bishkek (Kyrgyz Republic) 33.2 35.8 28.4 32.5 Dushanbe (Tajikistan) 47.4 62.9 53.5 54.6 Ashgabat (Turkmenistan) 21.6 21.3 16.8 19.2 Tashkent (Uzbekistan) 37.5 45.3 38.1 40.3 Source: http://airnow.gov, downloaded 12/2023. Unlike in other countries and world regions (for Given the absence of reliable AQ monitoring data example, Europe, China, and India), satellite in rural areas, the dispersion calculations of the imagery and global model calculations suggest GAINS model have been used to obtain a spatial considerable spatial diversity of PM 2.5 levels picture of the PM 2.5 concentrations prevailing within CA, with significantly lower concentrations across CA. Results have been aggregated in the sparsely populated rural regions compared to annual mean concentrations as the most to urban areas. relevant metric associated with public health impacts. Annual mean concentrations of PM 2.5 Figure 6: Concentrations of PM 2.5 in ambient air are computed at a 10 km x 10 km in 2020 (modeled - top) spatial resolution distinguishing the release and 2022 (satellite - bottom) (µg/m³) heights of the different emission sources. 44 50 Computed annual mean concentrations of PM 2.5 43 45 employing the PM 2.5 precursor emissions of 2020 42 40 (see Figure 4) and a global estimate of natural 35 41 soil dust emissions (see Section 3.2) are shown 30 40 in Figure 6. There are large differences in PM 2.5 LAT_01 25 39 concentrations across CA, ranging from below 20 38 5  µg/m³ to more than 50  µg/m³. Generally, the 15 highest levels occur in the deserts of western 37 10 5 Turkmenistan and at hot spots in and around the 36 0 major urban areas. Such hot spots are especially 35 concentrated in the airshed where Kazakhstan, 66 68 70 72 74 76 78 LON_01 the Kyrgyz Republic, Uzbekistan, and Tajikistan share common borders and where major cities of >50 all countries are located (Figure 6). In the following 35 figure, modeled PM 2.5 concentrations (top panel) roughly correspond to the satellite-derived PM 2.5 25 concentrations (bottom panel) where PM 2.5 15 levels are estimated by integrating data from satellites, simulations, and monitoring sources. 9 10 <5 9 As described in Shen et al. (2024), satellite data are merged with simulation data, accounting for their respective uncertainties, to generate estimates that account for most of the variance observed in ground-based PM 2.5 measurements. Source: Original GAINS analysis and https://sites.wustl.edu/acag/ Subsequently, information from ambient PM 2.5 monitoring is datasets/surface-pm2-5/#V6.GL.02. incorporated through a statistical fusion process. 14 BACK TO CONTENTS Air Quality Management in Central Asia Observed ground level observations in cities Figure 7: Concentrations of small are generally well resembled by model results, particulate matter (PM 2.5) in ambient air in while they tend to be underestimated in the 2020 originating from the anthropogenic emissions of the various countries (µg/m³) interpretation of remote sensing data. On the CONTRIBUTION OF KAZAKHSTAN other hand, satellite estimates reveal additional 46 areas with high concentrations of soil dust, 44 which are not fully captured by the modelling 42 approach based on available land use data. 40 The paucity of PM 2.5 measurements, especially in the 38 rural areas of CA, hinders a comprehensive validation of model results. A comparison of computed 36 annual mean concentrations of PM 2.5 with the few 55 60 65 70 75 80 CONTRIBUTION OF KYRGYZ REPUBLIC observations from urban monitoring stations shows 46 a remarkably good agreement, given the prevailing 44 uncertainties in the available emission inventories. 42 2.5. The dispersion of PM 2.5 in the 40 atmosphere 38 36 The geophysical approach of the atmospheric 55 60 65 70 75 80 dispersion model employed by GAINS makes it CONTRIBUTION OF TAJIKISTAN possible to track the fate of emissions emerging 46 from specific sources and thereby to quantify 44 their contributions to total PM 2.5 concentrations 42 in ambient air in the defined area. Calculations 40 consider the spatial and temporal distributions of 38 the various emission sources, their release heights, 36 meteorological conditions, the chemical processes 55 60 65 70 75 80 that produce secondary PM 2.5 , the inflow of pollution CONTRIBUTION OF TURKMENISTAN from outside sources, and the background load from 46 windblown soil dust. 44 The contributions to PM 2.5 concentrations throug- 42 hout CA that are made by the anthropogenic 40 emissions from the various countries are shown in 38 Figure 7. Naturally, the contributions are the highest 36 around the emission hot spots but the atmospheric 55 60 65 70 75 80 transport of PM 2.5 and its precursor emissions CONTRIBUTION OF UZBEKISTAN spreads beyond cities and administrative borders, 46 thereby emphasizing the need for an airshed 44 approach to AQM. 42 Unlike in many other regions in the world, natural 40 soil and desert dust constitutes an important 38 contribution to total PM 2.5 concentrations and 36 are the dominant source of PM 2.5 in some areas. 55 60 65 70 75 80 This is particularly true for western Turkmenistan and a large area in the east as well as central Kazakhstan 0.01 0.05 0.10 0.50 1.00 5.00 10.00 15.00 20.00 25.00 30.00 (see Figure 6 and the related discussion). Source: Original GAINS analysis. Air Quality Management in Central Asia BACK TO CONTENTS 15 2.6. Population exposure to PM 2.5 and country-wide source apportionment While air pollution maps provide valuable in- large differences in the absolute levels and rel- sights into the spatial diversity of the origins of ative shares of the contributions from soil and PM 2.5 concentrations in CA, aggregated metrics desert dust between the countries. In Turkmen- that summarize contributions in a larger region or istan, soil and desert dust contributes about 24 at specific locations are more informative to ad- µg/m³ to mean PM 2.5 exposure, that is, more than dress the burden of air pollution on public health 80 percent. Anthropogenic emissions generat- and to maximize health benefits from pollution ed in Turkmenistan account for only 2.4 µg/m³ control interventions. For this purpose, mean (8 percent) to exposure in the same country. In population exposure to PM 2.5 has been found as contrast, in the Kyrgyz Republic only 6 µg/m³ (30 a relevant metric. This section focuses on pop- percent of total exposure) originates from soil ulation exposure 10 in each country, considering and desert dust. the inflow of pollution from outside areas. Emissions from anthropogenic sources that There are strong transboundary fluxes of pollu- can be directly influenced by domestic policies tion in the central area where Kazakhstan, the account for between 8 percent (Turkmenistan) Kyrgyz Republic, Tajikistan, and Uzbekistan are and 65 percent (Kazakhstan) of total exposure, bordering each other. Figure 8 illustrates that while the inflow of anthropogenic pollution the transboundary transport of PM 2.5 between from other countries ranges between 5 percent the CA countries also affects mean population in Kazakhstan and 22  percent in the Kyrgyz exposure at national levels. Notably, there are Republic (Figure 8). Figure 8: Origin of country-wide population-weighted PM 2.5 exposure in 2020 (µg/m³) 100% 80% 60% 40% 20% 0% Kazakhstan Kyrgyzstan Tajikistan Turkmenistan Uzbekistan Soil dust Other countries Same country Source: Original GAINS analysis. 2.7. A focus on cities Given the large diversity in PM 2.5 concentrations Tashkent, Samarkand, and Almaty) with a total across CA and its high degree of urbanization, population of about 7 million people are located understanding the origin of country-wide PM 2.5 within the typical transport distance of PM 2.5 exposure is of limited relevance for pollution hot (<1,000 km) so that the inflow of pollution spots—namely, big cities where a major share into cities—even from other countries—could of population in CA lives. Furthermore, several possibly constitute an important share of total of the major CA cities (Bishkek, Dushanbe, exposure (Figure 8). 10 Population exposure is computed as the sum over the whole country of the products of local PM 2.5 concentrations and the number of people affected by this concentration (using a 10 km × 10 km spatial resolution). 16 BACK TO CONTENTS Air Quality Management in Central Asia 2.7.1. Sources of current PM 2.5 exposure sources. 11 Thus, effective efforts to manage ur- in Central Asian cities ban AQ need to address emission sources in the This assessment focuses on three cities in CA, entire airshed, which extends outside the imme- that is, Bishkek in the Kyrgyz Republic, Dushan- diate city jurisdictions, especially into the sur- be in Tajikistan, and Tashkent in Uzbekistan. The rounding provinces and regions. Depending on origin of population-weighted PM 2.5 exposure in the location, the airsheds can also include other those cities is analyzed employing the source countries. apportionment analyses of the GAINS model. The report’s findings regarding the sources of Detailed source apportionments for these cities PM 2.5 exposure could be summarized as follows: and their implications for AQM are discussed in detail in Sections 2.8.3 to 2.8.5. Earlier analyses ለ PM 2.5 exposure originates from various emis- for Almaty and Astana (Kazakhstan) are present- sion sources in different economic sectors ed in World Bank (2021) and World Bank (2022a) (Figure 10). and summarized in Box 1. ለ In all cities, the largest single anthropogenic Despite important differences between the contribution is caused by solid fuel combus- cities due to local conditions, several common- tion for heating in the residential sector, from alities provide general guidance for AQM in CA a small share of households in suburban areas cities. and from rural households in the surround- ለ Annual average PM 2.5 concentrations in the ing regions. On a per household basis, PM 2.5 urban areas of Dushanbe, Bishkek, Tashkent, emissions from coal and wood fired stoves Samarkand, Almaty, and Astana are clearly are 4,000–5,000 times higher than from above the WHO guideline value of 5 µg/m³. natural gas heating. The impact of stoves and boilers in urban households on AQ in the city ለ In many of the CA cities, soil and desert dust is enhanced by their low-emission heights. contributes much more to PM 2.5 concentra- tions than in most other regions in the world. ለ PM 2.5 exposure from mobile sources rang- The largest share of soil dust is transported es between 2 µg/m³ and 5.5 µg/m³, that is, from the large deserts in the west, most vis- about 10  percent of total exposure in each ibly during episodic pollution events with ex- city. However, primary PM 2.5 emissions from tremely high PM 2.5 concentrations. However, urban vehicles account only for about half soil and desert dust accounts for 20–50 per- of this share, while the rest of the exposure cent of total exposure in the cities even on an from mobile sources consists partly of sec- annual basis. On the flip side, this also means ondary PM 2.5 (formed from NO x emissions) that 50–80 percent of total exposure in the and/or transport emissions from surround- cities is caused by anthropogenic emissions, ing areas. which can be controlled through dedicated ለ The controlled or uncontrolled open burning policy interventions. of municipal waste emerges as an underap- Although cities emerge as the major pollution preciated source of PM 2.5 exposure in urban hot spots in CA, only a limited share of PM 2.5 areas and can in some cities constitute a exposure within the city boundaries (typically significant fraction of the contribution from 10–50 percent) originates from local emission urban sources. 11 In Kazakhstan, the isolated locations of the major cities lead to higher local contributions. Air Quality Management in Central Asia BACK TO CONTENTS 17 Figure 9: Spatial origin of PM 2.5 exposure in the six cities, 2020 (2018 for Almaty) (µg/m³) 70 60 50 40 µg/m3 30 20 10 0 Dushanbe Bishkek Tashkent Samarkand Almaty Astana Soil dust From other countries From same country From city Source: Original GAINS analysis. Figure 10: Sector contributions to PM 2.5 exposure in the six cities, 2020 (2018 for Almaty and Astana) 70 60 50 40 µg/m3 30 20 10 0 Dushanbe Bishkek Tashkent Samarkand Almaty Astana Soil dust Power and industry Residential Mobile sources Municipal waste Agriculture Source: Original GAINS analysis. ለ Secondary PM 2.5 formed from PM 2.5 react in the atmosphere with NH 3 (emitted precursor emissions such as SO 2 , NO x , from livestock manure management and fer- and NH 3 accounts for 10–50 percent of tilizer application) forming ammonium sulfate the exposure from anthropogenic sources and ammonium nitrate, the main components (excluding natural soil dust), see Figure of secondary inorganic PM 2.5 . 11. Thus, a sole focus of pollution control ለ Since the chemical formation of secondary measures on primary PM 2.5 emissions will PM 2.5 in the atmosphere takes some time, leave this share unaffected. the importance of secondary PM 2.5 increases ለ Effective reductions of secondary PM 2.5 need with distance. Thus, for reducing PM 2.5 to address emissions of SO 2 (mainly from coal exposure in the CA cities, controls of the combustion) and NO x (predominantly from precursor emissions across the entire city mobile sources). These gaseous emissions airsheds will be critical. 18 BACK TO CONTENTS Air Quality Management in Central Asia Figure 11: Contributions in 2020 to PM 2.5 from primary PM 2.5 emissions (by sector) and from secondary PM 2.5 formed in the atmosphere from the precursor emissions of SO2 , NOx and NH3 70 60 50 40 µg/m3 30 20 10 0 Dushanbe Bishkek Tashkent Samarkand Almaty Astana Soil dust Power and industry Residential Mobile sources Municipal waste Agriculture Secondary Source: Original GAINS analysis. 2.8. Improving air quality in Central Asian cities Ranging between 30 µg/m³ and 60 µg/m³, the need to be involved in comprehensive AQM plans current PM 2.5 exposure levels in the three cities for the three cities. The analysis (a) is based on of Bishkek, Dushanbe, and Tashkent are well the source apportionments for 2020 shown in above international AQ standards such as the the preceding sections, (b) examines the like- WHO guideline of 5  µg/m³, the updated EU limit ly changes in PM 2.5 precursor emissions with- value for PM 2.5 of 10 µg/m³, and even the highest out further AQ policy interventions in 2030 and WHO interim target 1 of 35 µg/m³. As shown 2040, (c) estimates impacts on PM 2.5 exposure elsewhere, such high exposure causes significant in the cities, and (d) identifies the key sectors for health impacts with considerable economic which measures within the jurisdictions of local consequences (World Bank 2022b). and national authorities could deliver the largest AQ improvements. This analysis considers the At the same time, rapid economic growth dynamics of socioeconomic development and together with fast urbanization and the how they will change the relative importance prevailing lenient pollution controls will further of the various economic sectors on population deteriorate AQ in CA cities in the coming exposure in the future. It also takes full account decades unless effective countermeasures of how policies and measures that have been re- are taken. This section examines how PM 2.5 cently decided will unfold in the future. However, levels in Dushanbe, Bishkek, and Tashkent are a robust economic appraisal of the various op- likely to develop by 2040 and identifies the tions would require a comparative assessment emission source sectors within and around the of the costs of all available measures within a city cities for which effective pollution controls will and in the entire airshed as well as their cost-ef- be most critical if international AQ standards fectiveness for improving PM 2.5 levels in the city, were to be met. The cost-effectiveness of AQ which is beyond the scope of this analysis. Also, policy interventions in Almaty and Astana in there are important interactions between AQ Kazakhstan has been explored with the GAINS measures and greenhouse gas (GHG) mitigation model in an earlier report (World Bank 2022a) efforts that could deliver substantial co-benefits and is summarized in Box 1 below. or possibly result in undesired trade-offs. These The analysis employs the GAINS model (see An- have been explored in detail in the earlier studies nex I) to determine priority sectors for AQM that for Almaty and Astana (World Bank 2022a). Air Quality Management in Central Asia BACK TO CONTENTS 19 Box 1: Integrated AQM and GHG reduction assessment for Almaty and Astana: Integrated AQM and GHG reduction assessment for Almaty and Astana The study Clean Air and Cool Planet - Integrated Air Quality Management and Greenhouse Gas Reduction for Almaty and Astana (at the time of the study called Nur-Sultan) highlights how potential synergies between AQ improvement and GHG reduction measures can be enhanced in a cost-effective manner. To identify and maximize those synergies and assess the measures’ cost-effectiveness, the study applies two extensions to the Greenhouse GAINS model—GAINS-City and GAINS-Policy. The study demonstrated that the most cost-effective measures to improve AQ in Almaty and Astana are not the same as the most cost-effective measures to mitigate climate change. The largest low-cost PM 2.5 exposure reduction in Almaty and Astana comes from improving efficiency of coal-fired heating systems in buildings, whereas the largest low-cost climate mitigation is offered by (a) a switch from coal to natural gas heating in detached houses and (b) replacement of coal-fired combined heat and power plants (CHPPs) with new combined cycle gas turbines. The study estimated that total upfront investments in low-emission residential and commercial space heating of EUR 43 million and EUR 91 million for Astana and Almaty, respectively, are needed for a 10-year period to achieve PM 2.5 concentrations in line with WHO’s Interim Target 2. The study showed that a scenario integrating the most cost-effective air pollution and climate change mitigation measures minimizes the costs of achieving significant improvements on both environmental fronts. For both cities, integrated cost-effective measures include (a) an accelerated replacement of existing coal stoves and boilers in detached houses and commercial buildings with gas heating devices or, where technical or social reasons make this impossible, with new, more efficient and/or cleaner heating appliances and systems; (b) refurbishment of the district heating distribution networks to reduce losses; and (c) alignment of thermal codes of new buildings with those of the EU. For centralized heat and power generation, replacement of coal-fired CHPPs, existing heat-only boilers, and electricity imports having higher carbon dioxide (CO 2 ) intensities with new combined cycle gas turbines (with heat storage facilities) and possibly biomass-fired CHPPs emerges as cost-effective. This also facilitates penetration of electric light-duty vehicles. The system costs of implementing such an integrated scenario were compared against the costs in (a) a baseline scenario where no additional measures are adopted, (b) an AQ- driven scenario that achieves PM 2.5 targets but not GHG ones, and (c) a climate mitigation driven scenario that reduces GHG emissions to a level compatible with Kazakhstan’s NDC commitments but does not achieve PM 2.5 standards. The study estimated that system costs of implementing all scenarios were lower than the baseline scenario with costs of the integrated AQ and climate change mitigation scenario being EUR 15 million per year higher than the AQ- driven scenario. However, the study found that implementing some fiscal instruments, such as a carbon tax, would further reduce system costs of the integrated scenario. 2.8.1. A baseline scenario toward 2040 Campagne, and Jooste 2019; World Bank 2022c, While only sparse quantitative information about 2023c, 2024b, 2025b). At the same time, accel- future socioeconomic trends in individual cities erating urbanization in CA will concentrate pop- is available, business-as-usual projections sug- ulation and economic growth in the urban areas. gest continued yet moderating economic growth Given the general lack of quantitative projec- for CA countries as in reference scenarios for re- tions for the cities, the baseline scenario for this cent Country Climate and Development Reports report derives future levels of emission-generat- in Kazakhstan, Uzbekistan, and Tajikistan (Burns, ing activities in the urban areas by applying pub- 20 BACK TO CONTENTS Air Quality Management in Central Asia lished national trends in per capita income and host of proven measures are available that could sectoral energy consumption to the population deliver significant improvements to public health and GDP growth of the cities considered. In par- through cleaner air. To this end, this assessment ticular, this report employs the national trends examines the AQ impacts of a set of cost-ef- for GDP and sectoral energy consumption of the fective measures for Bishkek, Dushanbe, and stated policies scenario published by the IEA in Tashkent, as derived in the forthcoming World its World Energy Outlook 2021 (IEA 2021). Ur- Bank report on ‘Accelerating Access to Clean banization trends are taken from the UN popu- Air’ (World Bank 2025a). This flagship report es- lation projections. 12 Because of the importance tablishes a visionary target for 2040: to halve— of the residential sector for AQ in the CA cities, in each of the 12 world regions considered—the the national trends in fuel consumption for space number of people exposed in 2020 to PM 2.5 con- heating in households have been replaced by centrations of more than 25 µg/m³. Two groups more detailed and city-specific information that of cost-effective measures are considered in the has been collected for this assessment regard- global report: (a) measures that form critical parts ing the building stock, its energy efficiency, and of climate and sustainable development policies, the shares of different fuels for space heating. and which deliver significant co-benefits on AQ Equally, national trends for the power and heat and (b) measures traditionally considered in con- sector have been replaced by specific expansion ventional AQM practices, such as end-of-pipe plans for centralized heat and power generation. 13 emission controls. The list of cost-effective air The projected baseline levels of emission- pollution control measures for the various emis- generating activities have been combined with sion source sectors is presented in Annex IV. future trends in emission factors for the various Due to the current uncertainties about the near- PM 2.5 precursor emissions that reflect the term prospects of energy policies (among others, ongoing penetration of already decided emission related to the role of natural gas) in CA countries, control legislation in the various countries. especially their implications for the three cities For CA, the analysis considers cyclones and under consideration in the near future, the 1-field electrostatic precipitators to control clean air scenario presented in this assessment PM emissions from solid fuel combustion in focuses on the combination of climate/energy large power plants and industrial boilers as well measures and conventional AQM options, as as for fugitive emissions in the cement and listed in Annex IV. Some of those AQM measures steel industry. Notably, no enforced policies overlap with decarbonization policies. While the on emission controls for road vehicles are benefits of further energy and carbon policies considered in the baseline. are not assessed in this study, they have been The resulting emission projections of PM 2.5 assessed evaluated previously for Almaty and precursor substances are then fed into the Astana, see Box 1 above (World Bank 2022a). atmospheric dispersion model to estimate The analysis presented in this report explores baseline concentrations of PM 2.5 in the for all cities a common set of measures that ambient air of the cities and to determine the has been adjusted to the local conditions based contributions from the various emission sources on available local data. However, further work to population exposure. with more detailed statistical information will be necessary to refine estimates for individual 2.8.2. The 2040 clean air scenario cities. Although the basic set of measures While the baseline provides a benchmark of fu- is similar in each CA country (Annex IV), the ture AQ without further policy interventions, a combination of measures and the sequence of 12 https://population.un.org/wpp/. 13 https://www.wri.org/research/global-database-power-plants. Air Quality Management in Central Asia BACK TO CONTENTS 21 their application is different. Urbanization levels m³ was caused by solid fuel combustion in the and economic growth outlooks, which affect residential sector. Of this, half originated from changes in exposure to PM 2.5 pollution, vary detached houses in the suburban areas of the significantly across different locations. These city and the other half transported into the city variations are influenced by geophysical and from the surroundings. This large contribution economic conditions, fuel mix, fuel subsidies, of residential sources occurred despite the high cost structures, the uptake of green incentives, share (about 46 percent) of households con- and the penetration of related technologies. nected to the district heating network. How- Each of these factors determines which sectors ever, according to the Bishkek Household Sur- will be targeted by AQ measures in each country. vey, many of the households not connected to Governments in CA often start reducing pollution district heating used coal as the main heating fuel (20  percent traditional coal-fired stoves from industry or the power sector, focusing and 37 percent simple coal-fired boilers [World on large polluters. Institutional barriers could Bank 2020a]). Road transport constituted the postpone the implementation of measures in second largest source, accounting for about certain sectors, for instance, transport requires 26 percent of total exposure. Roughly two- involvement from the Ministry of Internal Affairs, thirds of it are caused by light-duty vehicles. while agriculture and residential sectors need private investors and changes in consumer The spatial and sectoral origins of population behavior. Additionally, the marginal cost of exposure in Bishkek, illustrated by the so-called implementing pollution reduction measures source apportionment diagrams, show that varies between countries. about 30 percent of total PM 2.5 in ambient air in Bishkek consists of secondary PM 2.5 formed Consequently, sequencing and combination of in the atmosphere from gaseous precursor measures to improve AQ is not uniform across emissions (mainly SO 2 , NO x , and NH 3 ). As the countries. The tailored approach ensures that chemical processes that form secondary PM 2.5 AQ measures effectively address the specific take some time, most of secondary PM 2.5 is challenges faced by each nation. The sections transported from outside into the city and will below present customized packages of AQ im- therefore remain unaffected by conventional provement measures for Bishkek, Dushanbe, AQM measures that address primary PM 2.5 and Tashkent. emissions at the city scale. 2.8.3. Outlook for Bishkek (Kyrgyz Republic) Box 2 describes the approach that Bishkek is In 2020, about one-third of total popula- taking to mitigate the effects of PM 2.5 transported tion-weighted PM 2.5 exposure of about 34 µg/ from outside into the city. Box 2: Improving urban greening in Bishkek Improving urban greening is one of the components in the Kyrgyz Republic Air Quality Improvement Project, which is backed by a US$50 million in concessional loan from the World Bank’s International Development Association (IDA), spread over 50 years with a 10-year grace period. The urban greening component represents US$7.1 million of the total project. The design of the component was informed by a technical assessment 14 that allowed the identification of the main directions from which transboundary transfer of PM 2.5 occurs in Bishkek. This was coupled with collaborative work with local institutions responsible for greening to assess the current state of urban greening and the potential solutions to improve it. The urban greening interventions have the following subcomponents: 14 World Bank Document. 22 BACK TO CONTENTS Air Quality Management in Central Asia Box 2: Improving urban greening in Bishkek ለ Supporting measures to preserve and expand urban green cover in Bishkek, including creating green belts to mitigate the impacts of windblown dust: about 10 km of green belts/ corridors with an estimated green area of about 13 ha will be created in the project along selected roads in the city. ለ Constructing an irrigation system to ensure sustainability of urban green in Bishkek: To ensure sustainability of the newly created green belt(s) and supply of adequate irrigation water for the underserved areas, this subcomponent will support installation of about 15– 20 borewells, including rehabilitation of two existing borewells and associated irrigation infrastructure (including water-efficient and climate-resilient solutions). The project will also finance rehabilitation of two to three monitoring wells in different parts of the city to support hydrogeology expedition in improving groundwater monitoring and analysis. ለ Strengthening institutional capacity of the municipal greening and irrigation agencies in Bishkek: This subcomponent will finance equipment, tools, and technical support to the municipal enterprises in Bishkek responsible for greening and irrigation for better irrigation and urban tree management as well as provide expert support on urban planning with a focus on greenery and irrigation. Figure 12: Source apportionment of population-weighted PM 2.5 exposure in Bishkek in 2020 Total sector contributions from all PM2.5 precursor emissions Contributions of primary and secondary PM2.5 40 40 30 30 µg/m3 µg/m3 20 20 10 10 0 0 Natural From other From same From Total Natural From other From same From Total sources countries country the city sources countries country the city Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles Road dust Other agriculture Crop burning Municipal waste Observations Source: Original GAINS analysis. Total sector contributions from all PM2.5 precursor emissions Contributions of primary and secondary PM2.5 2040, For60 the analysis assumes a population Without 60 further AQM interventions, this socio- growth in Bishkek’s metropolitan area from economic growth combined with the structural 50 50 about 1.0  million in 2020 to 3.3 million people changes cause Bishkek’s primary PM 2.5 emis- in 2040, that is, by more than 50 percent. With sions to increase by 30 percent between 2020 40 40 an annual increase of per capita income of 2.9 and 2040, SO 2 by about 10 percent, and NO x by percent, urban GDP would rise by 173 percent. percent (Figure 13). Considering the likely 6030 µg/m3 µg/m3 30 Heat demand from households is assumed emission changes in the other areas in Bishkek’s to increase 20 by 54  percent, with equal relative airshed, 20 these urban trends in precursor emis- increases of all fuels assumed in the baseline. sions will then result in PM 2.5 exposure in the city 10 10 0 0 Natural Air Quality Management From other in Central From Asia same From Total Natural From other From same BACK TOFrom CONTENTS Total 23 sources countries country the city sources countries country the city Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles to increase from about 33 µg/m³ in 2020 to 38 µg/ implementation of the measures that emerge m³ in 2030 and more than 45 µg/m³ in 2040, well as cost-effective for attaining the 2040 Global above the WHO interim target 1 of 35 µg/m³ (Fig- clean air targets (see Table 4) would cut primary ure 14). The largest increase is anticipated from PM 2.5 emissions by about 45 percent, NO x emis- road transport sources (notably from light-duty sions by about 60 percent, and SO 2 emissions by vehicles), which in 2040 will account for 36 per- about 80 percent (Figure 13). This would reduce cent of total exposure. Only a slightly lower share population-weighted exposure to PM 2.5 in Bish- (32  percent of the total) is linked to residential kek to about 22  µg/m³ (the ‘2040 Global clean heating with solid fuels, especially from house- air scenario in Figure 14). Such an annual average holds in the surrounding areas. concentration means that it is likely that even However, a variety of proven measures are avail- the maximum concentrations within the city will able that could bring PM 2.5 exposure in Bishkek in drop below 25 µg/m³, the WHO interim target 2 line with international AQ standards. To this end, and Kyrgyz Republic’s annual average limit value. Figure 13: Emission scenarios for Bishkek 200% 160% Relative to 2020 120% 80% 40% 0% 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air PM 2.5 SO 2 NO x BC Power generation Industry Residential Transport Other Source: Original GAINS analysis. 180% Figure 14: Sources of PM 2.5 exposure in Bishkek in 2020, in baseline projections for 2030 and 2040 160% and in the Global clean air scenario in 2040 baseline, 50 140% 40 120% µg/m 3 to 2020 100% 30 Relative 80% 20 60% 40% 10 20% 0 0% From From From From From From From From outside the city outside the city outside the city outside the city 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2030 Baseline 2040 Baseline 2040 Global clean air PM2.5 SO2 NOx BC Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles Power generation Industry Residential Transport Other Road dust Other agriculture Crop burning Municipal waste Source: Original GAINS analysis. 24 80 BACK TO CONTENTS Air Quality Management in Central Asia 70 200% In total, implementation of these measures differences in the technical opportunities and would reduce PM 2.5 exposure in Bishkek from the costs of the various measures between sectors. baseline level by about 24 µg/m³ (Table 4). About For instance, the measures in the transport sec- 42 percent of this potential can be achieved tor may show the largest reduction potential for through emission controls for road transport Bishkek (see Table 4), but the sequencing of their in the city, 9 percent through controls for resi- implementation can be done in various orders. dential heating with solid fuels, and 5 percent Cost-effective road transport measures mainly through improved waste management practices involve introduction of emission standards and in the city (Figure 15). In total, measures taken improved inspection and maintenance programs, within the city boundaries deliver 58 percent of which have lower costs and shorter implementa- the total potential. The remaining 42  percent is tion periods than a large-scale replacement of linked to emission reductions in the surround- solid fuel appliances for residential heating with ings, notably for residential heating (14  percent cleaner alternatives. At the same time, measures of the total potential), power generation and road encouraging clean residential heating remain the transport (8 percent each), and industry (6  per- ones that would improve AQ the most, and hence cent). Note that the relative magnitude of these their design and implementation should be pri- potentials differs from the overall contributions oritized. Residential heating measures represent of the various sectors that have been identified the second largest cost-effective key measure in the source apportionment (Figure 15) due to for reducing emissions (see Table 4). Table 4: The potential exposure reductions in Bishkek from the cost-effective key measures of the 2040 clean air scenario Exposure reduction potential (µg/m³ ) Source Share Cost-effective measures From From sector (%) urban outside Total sources sources Electric vehicles (buses, cars, two-wheelers), Euro 6 emission standards for new light- Road duty cars, Euro VI standards for new heavy- 9.8 1.9 11.7 50 transport duty vehicles, effective inspection and maintenance programs with enforced repair or retirement of broken vehicles Full connection to grid-based heating Residential systems (district heat, heat-only boilers), 2.1 3.3 5.4 23 heating replacement of remaining solid fuel stoves and boilers with cleaner heating methods Power plants Flue gas desulfurization for large coal boilers 0.3 1.9 2.2 9 Flue gas desulfurization and high-efficiency Industry PM filters for large boilers and fugitive 0.3 1.3 1.6 7 emissions Municipal Collection and sorting of waste, composting/ waste 1.1 1.1 5 recycling/managed landfill of residual waste management Ban on open burning of agriculture residue, Crop burning 0.0 0.6 0.6 2 energetic use of crop residue Livestock Low-nitrogen feed and efficient manure 0.9 0.9 4 farming management at large industrial farms Sum 13.6 9.9 23.5 100 Source: Original compilation. Air Quality Management in Central Asia BACK TO CONTENTS 25 Figure 15: The shares of the cost-effective exposure reduction potential in the city of Bishkek that can be achieved through measures in different source sectors within and outside Bishkek in 2040 1% 1% 5% 8% 9% Road transport Residential heating Ou tsid Powerplants 14% Industry e meas ures 58% Crop burning Livestock ures 42% Road transport meas Residential heating 8% ban Powerplants 42 Industry Ur % MSW 6% 4% 2% The total potential amounts to 24 µg/m³. Source : Original GAINS analysis. The clean air scenario presented in this assess- be was caused by solid fuel combustion in the ment explores the scope for measures 4% that are residential sector, 75 percent by buildings in typically part of the portfolios of AQ managers suburban areas within the city boundaries, and % 76 % Road transport 41 (for example, end-of-pipe emission controls) the rest transported into the city from the sur- 15 s ure Residential heating % and do not overlap with measures in other policy (Figure 16). This large contribution roundingsPowerplants eas Urban m areas (for example, energy policies). As shown Industry occurs despite the high share of buildings in the Crop burning elsewhere (for example, in the forthcoming city heated with district heat, natural gas ,and Livestock 7% World Bank flagship report on ‘Accelerating Ac- electricity (about 50 percent of the households cess to Clean Air’), interventions aimed at other were connected to district heat, 17 percent used Road transport Residential heating policy objectives (for example, climate and en- natural gas, and 8 percent used electricity for Industry % 8% ergy policies) can have significant co-benefits space heating). Only about 26 percent of house- 24 Solid Waste Management re s on PM 2.5 precursor emissions and thereby lead holds—mainly detached houses in suburban su a side to lower population exposure to PM 2.5 as4% areas—relied on coal (simple coal stoves and ea m 18% effect (however, these are beyond the scope i 4% boilers) (IEA 2022; World Bank 2024a; UNICEF de % 1% ts Ou 1 of this analysis). At the same time, some of the 2023). In contrast, only 15 percent of the expo- AQM measures of the clean air scenario will also sure originated from road transport and 12 per- affect global temperature increase, although cent from municipal waste management. the absolute impact of action in Bishkek will 15% About 15 percent of total PM 2.5 in ambient air in O the be small. Most notably, compared to 2020, Dushanbe consists of secondary PM 2.5 formed in ut measures in the clean air scenario will also lead sid 7% the atmosphere from gaseous precursor emis- em to a 55 percent decline in black carbon emis- sions (mainly SO 2 , NO x , and NH 3 ). As the chemical 13 Road transport ea % sions, a potent short-lived climate pollutant with sur Residential heating reactions that form secondary PM 2.5 take some 5% es 7 strong positive radiative forcing (Figure 15). Powerplants of secondary PM 2.5 is transported from time, mostIndustry 3% 2.8.4. Outlook for Dushanbe (Tajikistan) outside into the Crop city and will therefore remain un- burning Urban m eas Livestock In 2020, 42 percent of total population-weight- affected by conventional AQM measures that ad- ed PM 2.5 exposure of about 57 µg/m³ in Dushan- dress primary PM 2.5 emissions at the city scale. Road transport 21% Residential heating u re % MSW 22 s2 7% 26 BACK TO CONTENTS Air Quality Management in Central Asia 8% 8% Natural From other From same From Total Natural From other From same From Total sources countries country the city sources countries country the city Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles Road dust Other agriculture Crop burning Municipal waste Observations Figure 16: Source apportionment of population-weighted PM 2.5 exposure in Dushanbe in 2020 Total sector contributions from all PM2.5 precursor emissions Contributions of primary and secondary PM2.5 60 60 50 50 40 40 µg/m3 µg/m3 30 30 20 20 10 10 0 0 Natural From other From same From Total Natural From other From same From Total sources countries country the city sources countries country the city Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles Road dust Other agriculture Crop burning Municipal waste Observations Source : Original GAINS analysis. Total sector contributions from all PM2.5 precursor emissions Contributions of primary and secondary PM2.5 50 50 For 2040, the analysis assumes a population anticipated from road transport sources (notably growth 40 in Dushanbe’s metropolitan area from from 40 light-duty vehicles), which would then about 1.6  million in 2020 to 3.3 million people account for 21 percent of total exposure. In the 2040, that is, by about 80 percent. With in 30 business-as-usual 30 baseline, total exposure will µg/m3 µg/m3 an annual increase of per capita income of remain dominated by residential heating with 20 20 2.9 percent, urban GDP would rise by 215 solid fuels (39 percent). percent. Considering ongoing energy efficiency 10 However, 10 a variety of proven measures are improvements, heat demand is assumed to grow available that could bring PM 2.5 exposure in 0 percent (World Bank 2024a). by 15 0 Dushanbe in line with international AQ standards. Natural From other From same From Total Natural From other From same From Total furthercountries Withoutsources AQM interventions, country the the city baseline end, implementation To this sources countries country of the measures the city Soil to trends lead a 50 percent dust increase Secondary PM of primary that emerge Powerplants as cost-effective for attaining the Industrial combustion PM 2.5 in Dushanbe between Residential Industrial processes 2020 and 2040 2040. SO 2 Heavy & commercial Global duty diesel clean vehicles air targets Light (see Table 5) would duty vehicles Road dust and NO x emissions Other agriculture are expected to grow by 33 Crop burning cut primary PM 2.5 emissions by moreObservations Municipal waste than 80 percent and 70 percent, respectively (Figure 17). percent and SO 2 and NO x emissions by 60–70 Considering the likely emission changes in other percent (see Figure 17). This would bring PM 2.5 areas in Dushanbe’s airshed, these urban trends exposure in Dushanbe close to international in precursor emissions will then result in PM 2.5 AQ standards. Additionally, these measures exposure in the city increasing from about 57 µg/ could reduce population exposure to PM 2.5 in m³ in 2020 to 68 µg/m³ in 2030 and 77 µg/m³ Dushanbe to about 27 µg/m³ (the ‘2040 Global in 2040, all well above the WHO interim target 1 clean air scenario’ in Figure 18), which is only of 35 µg/m³ (Figure 17). The largest increase is slightly higher than the WHO interim target. Air Quality Management in Central Asia BACK TO CONTENTS 27 0% 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air PM 2.5 SO 2 NO x BC Power generation Industry Residential Transport Other Figure 17: Emission scenarios for Dushanbe 180% 160% 140% 50 120% Relative to 2020 40 100% 30 80% 3 µg/m 60% 20 40% 10 20% 0% 0 From From From From From From From From 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air outside the city outside the city outside the city outside the city 2020 PM2.5 2030 Baseline SO2 Baseline 2040 NO x 2040 Global BC clean air Soil dust Secondary PM Powerplants Industrial combustion Power generation Industry Residential Transport Other Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles Road dust Other agriculture Crop burning Municipal waste Source: Original GAINS analysis. Figure 18: Sources of PM 2.5 exposure in Dushanbe in 2020, in the baseline projections for 2030 and 2040 baseline, and in the Global clean air scenario in 2040 80 200% 70 160% 60 µg/m 3 to 2020 120% 50 Relative 40 80% 30 40% 20 0% 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 10 PM2.5 SO2 NOx BC 0 Fromgeneration Power From Industry From Residential From From OtherFrom Transport From From outside the city outside the city outside the city outside the city 2020 2030 Baseline 2040 Baseline 2040 Global clean air Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles Road dust Other agriculture Crop burning Municipal waste Source: Original GAINS analysis. In total, implementation of these measures in 5). About 41 percent of this potential can 2040 can reduce PM 2.5 exposure in Dushanbe be achieved through emission controls for 60 the baseline level by 51 µg/m³ (Table from residential heating with solid fuels, 18 percent 50 28 BACK TO CONTENTS Air Quality Management in Central Asia 40 /m 3 30 through controls of vehicle emissions in the and costs of the various measures between city, and 15  percent through improved waste sectors. In the case of Dushanbe, acting on management practices in the city (Figure 19). In residential heating will be of utmost importance total, measures taken within the city boundaries to reduce emissions in a cost-effective manner would deliver 76 percent of the total potential. (see Table 5). Compared to Bishkek, road The remainder is linked to emission reductions transport plays a much smaller role in a quickly in the surroundings, notably for residential expanding and renovating Dushanbe. Thus, heating (8  percent of the total potential) and priority actions for Dushanbe will primarily road transport (7 percent). Note that the relative lie in further expanding the existing district magnitude of these potentials differs from the heating network and further reducing the use of overall contributions of the various sectors that solid fuels for heating. The second most cost- have been identified in the source apportionment effective emission reduction measure lies in the (Figure 16: Source apportionment of population- road transport sector, in which actions such as weighted PM 2.5 exposure in Dushanbe in 2020) a further expansion of road transport standards due to differences in the technical opportunities and of electric vehicles will be beneficial. Table 5: The potential exposure reductions in Dushanbe from the cost-effective key measures of the 2040 clean air scenario Exposure reduction potential (µg/m³ ) Source Share Cost-effective measures From From sector (%) urban outside Total sources sources Electric vehicles (buses, cars, and two- wheelers), Euro 6 emission standards for new Road light-duty cars, Euro VI standards for new 9.4 3.5 12.8 25 transport heavy-duty vehicles, effective inspection and maintenance programs with enforced repair or retirement of broken vehicles Full connection to grid-based heating systems; Residential replacement of remaining conventional solid 20.7 4.2 25.0 49 heating fuel stoves and boilers with cleaner heating methods Power plants Flue gas desulfurization for large coal boilers 0.0 2.1 2.1 4 Flue gas desulfurization and high-efficiency PM Industry filters for large boilers and fugitive emissions, 1.9 1.8 3.7 7 especially for cement industry Municipal Collection and sorting of waste, composting/ waste 6.2 6.2 12 recycling/managed landfill of residual waste management Ban on open burning of agriculture residue; Crop burning 0.4 0.4 1 energetic use of crop residue Livestock Low-nitrogen feed and efficient manure 0.6 0.6 1 farming management at large industrial farms Sum 38.2 12.6 50.8 100 Source: Original compilation. Air Quality Management in Central Asia BACK TO CONTENTS 29 2% Residential heating m 8% ban Powerplants 42 Industry Ur % MSW 6% 2 % 4% Figure 19: The shares of the cost-effective exposure reduction potential in the city of Dushanbe that can be achieved through measures in the different source sectors within and outside Dushanbe in 2040 4% % 76 % Road transport 41 15 s ure Residential heating % eas Powerplants Urban m Industry Crop burning Livestock 7% Road transport Residential heating Industry % 8% 24 Solid Waste Management re s su 4% ea m 18% 4% i de 1% 1 % ts The total potential amounts to 50 µg/m³. Ou Source: Original GAINS analysis. The clean air scenario presented in this report Notably, only 20 percent of total PM 2.5 exposure 15% measures that are explores the scope for originated from emissions within the city, while O typically part of the portfolios of AQ managers about 40 percent was transported into the city ut sid 7 % (for example, end-of-pipe emission controls) and from other regions in Uzbekistan and 25 percent em 13 Road transport do not cause major interference with other policy from other countries. About 18 percent consists ea % sur Residential heating areas (for example, energy policies). of windblown soil dust. 5% es 7 Powerplants 3% At the same time, some of the AQM measures Compared to other cities in CA, heating with Industry Crop burning of the clean air scenario will also affect global solid fuels within the city made only a negligible Urban m eas Livestock temperature increase, although the absolute contribution to PM 2.5 exposure in Tashkent, ow- impact of action in Dushanbe will be small. dominance ing to theRoad transport of natural gas as the main 21% Residential heating heating fuel. However, residential solid fuel use u re % Most notably, compared to 2020, the measures MSW 22 s2 in the surrounding areas contributed about 18 7% in the clean air scenario will also lead to a 75 percent to PM 2.5 exposure in the city. percent decline in black carbon emissions, a It is noteworthy that more than 40 percent of potent short-lived8climate % pollutant with strong 8% total PM 2.5 in ambient air in Tashkent consists positive radiative forcing (Figure 19). of secondary PM 2.5 formed in the atmosphere 2.8.5. Outlook for Tashkent (Uzbekistan) from gaseous precursor emissions (mainly SO 2 , In 2020, residential heating, road transport, NO x , and NH 3 ). As the chemical processes that and industry contributed almost equal shares form secondary PM 2.5 take some time, most of (between 14 percent and 18 percent) to the total secondary PM 2.5 is transported from outside into population-weighted PM 2.5 exposure of 43 µg/ the city and will therefore remain unaffected by m³ in Tashkent (Figure 21). conventional AQM measures that address pri- mary PM 2.5 emissions at the city scale. 30 BACK TO CONTENTS Air Quality Management in Central Asia Natural From other From same From Total Natural From other From same From Total Re sources countries country the city sources countries country the city 40% Soil dust Secondary PM Powerplants Industrial combustion 0% Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles 2020 Road dust2040 BL Clean Other 2020 2040 BL Clean Air agriculture Air burning Crop 2020 2040 BL Clean Air Municipal waste 2040 BL 2020 Clean Air Observations PM 2.5 SO 2 NO x BC Figure 20: Source apportionment of population-weighted PM 2.5 exposure in Tashkent in 2020 Power generation Industry Residential Transport Other Total sector contributions from all PM2.5 precursor emissions Contributions of primary and secondary PM2.5 50 50 180% 40 40 160% 30 30 µg/m3 µg/m3 140% 20 120% 20 Relative to 2020 100% 10 10 80% 0 0 Natural From other From same From Total Natural From other From same From Total 60% sources countries country the city sources countries country the city 40% Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles 20% Road dust Other agriculture Crop burning Municipal waste Observations 0% Source: Original GAINS analysis. 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air For 2040, the analysis PM2.5 assumes a population SO2 NOx AQM interventions, Without further BCthe baseline growth in Tashkent’s metropolitan area from trends lead to a 37 percent increase of primary Power generation Industry Residential Transport Other about 2.5  million in 2020 to 3.3 million people PM 2.5 in Tashkent between 2020 and 2040. SO 2 in 2040, that is, by about 30 percent. With an and NO x emissions grow by about 30 percent annual increase of per capita income of 2.9 and 70 percent, respectively (Figure 21). percent, urban GDP would rise by 130 percent. Figure 21: Emission scenarios for Tashkent 200% 160% Relative to 2020 120% 80% 40% 0% 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air 2020 2040 BL Clean Air PM2.5 SO2 NOx BC Power generation Industry Residential Transport Other Source: Original GAINS analysis. Considering the likely emission changes in the from road transport sources (notably from light- other areas in Tashkent’s airshed, these changes duty vehicles), which would then account for 25 in urban emissions will then lead PM 2.5 exposure percent of total exposure. in the city to increase from about 43 µg/m³ in However, a variety of proven measures are avail- 2020 to 47 µg/m³ in 2030 and 52 µg/m³ in 2040, able that could bring PM 2.5 exposure in Tash- all well above the WHO interim target 1 of 35 µg/ kent in line with international AQ standards. To m³ (Figure 22). The largest increase is anticipated this end, implementation of the measures that Air Quality Management in Central Asia BACK TO CONTENTS 31 10 0 From From From From From From From From outside the city outside the city outside the city outside the city 2020 2030 Baseline 2040 Baseline 2040 Global clean air emerge as cost-effective for attaining the 2040 in line with international AQ standards. Addition- Soil dust Global clean 6) PM Secondary air targets (see Table would cut Powerplants ally, these measures could Industrial reducecombustion population Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles primary PM 2.5 emissions by about 25 percent and exposure to PM 2.5 in Tashkent to about 24 µg/m³ Road dust Other agriculture Crop burning Municipal waste SO 2 and NO x emissions by 55–65 percent (Figure (the ‘2040 Global clean air scenario’ in Figure 22), 21). This would bring PM 2.5 exposure in Tashkent which is slightly below the WHO interim target 2. Figure 22: Sources of PM 2.5 exposure in Tashkent in 2020, in the baseline projections for 2030 and 2040 baseline, and in the Global clean air scenario in 2040 60 50 40 µg/m 3 30 20 10 0 From From From From From From From From outside the city outside the city outside the city outside the city 2020 2030 Baseline 2040 Baseline 2040 Global clean air Soil dust Secondary PM Powerplants Industrial combustion Industrial processes Residential & commercial Heavy duty diesel vehicles Light duty vehicles Road dust Other agriculture Crop burning Municipal waste Source: Original GAINS analysis. In total, implementation of these measures of the various measures between sectors. In the would reduce PM 2.5 exposure in Tashkent from case of Tashkent, the most cost-effective emis- the baseline level by 28 µg/m³ (Table 6). About sion reduction potential lies in the sectors of 27 percent of the potential is linked to measures road transport and industrial production with 28 within Tashkent—21 percent to emission controls percent and 22 percent, respectively. But there for road vehicles and 5 percent to improved are institutional barriers for immediate imple- practices for municipal waste management. mentation of measures in the transport sector. However, the much larger potential (73 percent) In the short run, Tashkent may benefit from and is related to measures in the surroundings. About further develop its ongoing efforts as the city 22 percent of the total potential for AQ improve- is currently developing mechanisms to regulate ments in Tashkent relates to controls of industri- industry and is similarly looking into introducing al emissions outside the city, 15 percent to the specific regulations in transport. use of solid fuels in the residential sector, and The reduction potential in the residential sector 13 percent to power plant emissions (Figure 23). in Tashkent is comparably smaller than in the Note that the relative magnitude of these poten- other two cities examined in this report. Howev- tials differs from the overall contributions of the er, there is a reduction potential in commercial various sectors that have been identified in the buildings, for example, through introduction of source apportionment (Figure 21) due to differ- cleaner heating for commercial buildings and in- ences in the technical opportunities and costs stallations (that is, greenhouses) near the city. 32 BACK TO CONTENTS Air Quality Management in Central Asia 1% 1% 5% 8% 9% Table 6: The potential exposure reductions in Tashkent from the cost-effective Road transport key measures of the 2040 clean air scenario Residential heating Ou tsid Powerplants Exposure reduction 14% Industry potential (µg/m³ ) e meas Share ures 58% Crop burning Source sector Cost-effective measures From From Livestock (%) urban outside Total ures 42% sources sources Road transport meas Electric vehicles (buses, cars, two-wheelers),Residential heating 8% Euro 6 emission standards for new light- ban Powerplants duty cars, Euro VI standards for new heavy- Industry 42 r Road transport 5.9 1.9 7.9 28 U % duty vehicles, effective inspection and MSW 6% maintenance programs with enforced repair 2% % broken of or retirement 4 vehicles Residential and Full connection to grid-based heating commercial systems, installation of filters, replacement (for example, 0.0 4.4 4.4 16 of remaining conventional solid fuel stoves greenhouses) and boilers with cleaner heating alternatives heating 4% Power plants Flue gas desulfurization for large coal boilers 0.3 3.7 3.9 14 % 76 Flue gas desulfurization and high-efficiency % Road transport 41 15 s ure Industry PM filters for large boilers and fugitive 0.1 Residential heating 6.1 6.2 22 % eas Powerplants emissions Urban m Industry Municipal waste Collection and sorting of waste; composting/ 1.4 Crop burning 1.4 5 management recycling/managed landfill of residual waste Livestock 7% Ban on open burning of agriculture residue, Crop burning 2.2 2.2 8 energetic use of crop residue Road transport Low-nitrogen feed and efficient manure Residential heating Livestock Industry 2.2 2.2 8 % farming management at large industrial farms 8% 24 Solid Waste Management Sum 7.7 20.5 28.2 100 re s su Source: Original compilation. 4% ea m 18% 4% i de 1 u % 1% ts Figure 23: The shares of the cost-effectiveOexposure reduction potential in Tashkent that can be achieved through measures in the different source sectors within and outside Tashkent in 2040 15% O ut sid 7% em 13 Road transport ea % sur Residential heating 5% es 7 Powerplants 3% Industry Crop burning Urban m eas Livestock Road transport 21% Residential heating u re % MSW 22 s2 7% 8% 8% The total potential amounts to 28 µg/m³. Source : Original GAINS analysis. Air Quality Management in Central Asia BACK TO CONTENTS 33 The clean air scenario presented in this report scenario will also affect global temperature in- explores the scope for measures that are typi- crease, although the absolute impact of action in cally part of the portfolios of AQ managers (for Tashkent will be small. Most notably, compared example, end-of-pipe emission controls) and do to 2020, the measures in the clean air scenar- not cause major interference with other policy io will deliver a 40 percent cut in black carbon areas (for example, energy policies). At the same emissions, a potent short-lived climate pollutant time, some of the AQM measures of the clean air with strong positive radiative forcing (Figure 23). 34 BACK TO CONTENTS Air Quality Management in Central Asia 3. The Air Quality Management System in Central Asian Countries There is no universal definition of a strong, institutional attributes: (a) legal and regulatory effective, and efficient AQM system; however, framework, (b) committed executive, (c) evidence has shown that aside from sector- nested planning, (d) horizontal and vertical specific aspects, other governance and coordination, and (e) accountability and institutional cross-cutting mechanisms and transparency. The assessment of AQM processes play a critical role. The methodological systems in CA countries presented in the next approach to assess governance and institutional sections provides a high-level evaluation of arrangements in this report is based on the those governance and institutional attributes, emerging framework that the World Bank has with a larger focus on the legal and regulatory used in recent regional studies. framework and on the accountability and transparency aspects. Country-specific AQM Emerging evidence from countries with various roadmaps and assessments provide more detail maturity levels for airshed planning and on the AQM and governance systems in the management suggests that recommendations individual countries. 15 can be built on the following governance and Box 3: A framework to assess AQM governance and institutional arrangements The framework identifies formal institutions and actual behaviours. It favours organizational and institutional ‘function’ over ‘form’ and reviews both formal and informal mechanisms. The framework consists of 16 components organized around five attributes and a set of guidance questions to carry out the assessment. It has been developed specifically to distil lessons for policy makers, keeping in view the distinct challenges that the country faces (for example, population and land size, geographic diversity, particularities of public administration, type of federation, and general social dynamics). The framework can also be adapted and applied in other country settings, and an attempt has been made to make the attributes as mutually exclusive as possible. The framework is built on the following governance and institutional attributes: Legal and regulatory framework. Series of laws, acts, and regulations for defining the mandate for effective AQM, setting the country’s AQ standards, assigning the required institutional roles and responsibilities, establishing compliance, creating reporting and enforcement mechanisms, addressing transboundary air pollution, and adhering to international commitments. Committed executive. Strength of the executive’s commitment and pledge to the AQM agenda revealed by the existence of an explicit, clear, and publicly available vision and strategy, properly backed by enough resources, policy instruments and incentives, and capabilities in place to provide assurance that it will be effectively implemented and meet its targets. 15 1) Air Quality Assessment for Tashkent and the Roadmap for Air Quality Management Improvement in Uzbekistan, 2) Air Quality Management in Tajikistan, 3) Air Quality Analysis for Bishkek, and 4) Clean Air and Cool Planet Volume II : Integrated Air Quality Management and Greenhouse Gas Reduction for Almaty and Nur-Sultan. Air Quality Management in Central Asia BACK TO CONTENTS 35 Box 3: A framework to assess AQM governance and institutional arrangements Nested planning. Modalities and planning procedures for administering the AQM system vertically across different airshed zones and state and national administrative boundaries Horizontal and vertical coordination. Existing functional arrangements to coordinate AQM stakeholders across sectors (horizontal) and between different levels of government (vertical). This attribute includes a description of the membership, functions, management, and effectiveness of the coordination mechanisms. Accountability and transparency. Mechanisms to disclose information, track and evaluate progress, promote public and private sector participation, and hold institutions to account through adequate evidence/databases, information disclosure, and established channels for recourse. Also, description of the existing instruments for holding governments accountable when national or subnational jurisdictions are in nonattainment of AQ standards. Sources: World Bank 2025a, 2025b. 3.1. Legal and regulatory framework Developed AQM systems have legal frameworks environment to sectoral ministries in most CA in place, comprising national standards for countries occurred relatively recently. Most ambient AQ and sector pollution control countries in CA have AQ directorates in the (emissions and technology standards), with line ministries/committee, but since, in some a clear mandate to protect the health of the cases, those have been established recently, population in line with WHO guidelines. More they lack experience in policy making and generally, the legal and regulatory framework in policy development. On the other hand, national those AQM systems makes a clear distribution hydrometeorological services (Hydromets) of roles and responsibilities between the play a crucial role in technical aspects of AQM central government and subnational authorities. such as AQ monitoring, compiling of emission Legal frameworks establish a combination of inventories, and in some cases modeling. policy tools for AQM, including command-and- Given the shared historical development of CA, control and market-based instruments and AQM systems in the different CA countries exhibit enforcement tools (covering noncompliant common features. AQ is regulated through jurisdictions, companies, or citizens and using dedicated AQ laws—in most cases adopted in several specific instruments). They tend to the mid-1990s and updated through the years. include specific legislation on the use of market- Since 2021, the primary regulatory act in the based instruments, such as environmental taxes field of environmental protection, including AQ, and pollution charges. In some cases, the legal in Kazakhstan is the Environmental Code. Other framework also integrates climate and air policies CA countries are also considering the adoption by setting common objectives, targets, and of Environmental Codes. In addition to primary tools. Moreover, the legal framework establishes AQ-specific laws, AQ objectives, measures, and reduction targets for key pollutants, and AQ action plans are specified in national strategies, standards are regularly reviewed and revised development concepts, and action plans. based on current country-specific scientific However, currently no CA country has adopted research and data. a national AQM strategy. Only two CA countries AQ in CA countries is governed by sectoral (the Kyrgyz Republic and Kazakhstan) have ministries of environment/natural resources, ratified the UNECE Convention on Long-Range apart from Tajikistan where the governing body Transboundary Air Pollution (CLRTAP) and is the Committee on Environmental Protection. hence, have joined international cooperation The evolution from state committees on on AQM. 36 BACK TO CONTENTS Air Quality Management in Central Asia CA countries generally regulate emissions from in line with best international practice. Enterpris- stationary and mobile sources. Larger industries es that have obtained the IEP are exempt from and power and heat generation facilities, com- payments, provided their emissions are below monly defined as category I enterprises in CA the designated ELVs, whereas fees and charges countries, are mainly regulated through emission for enterprises that have not adopted BAT are limit values (ELVs). Emission fees and charges progressively increasing. are paid by polluters based on the amount of The efficiency of the ELV system in CA has been emissions that exceed the enterprises’ ELVs, questioned as ELVs are generally set at a high level but in many cases the payments are not suffi- and consider the historical level of emissions of cient to incentivize companies to invest in clean- enterprises. In addition, the level of ELVs is also er production practices or apply best available related to the nationally determined maximum techniques (BAT). In the case of Kazakhstan, the allowed concentrations (MACs) of air pollutants last revision of the Environmental Code in 2021 that are also generally higher than international brought amendments to the fee structure based guidelines. ELVs play an important role in AQM as on the significance of the negative environmental they not only set regulatory limits on emissions impact of the enterprise and the Integrated Envi- and establish the foundation for emission fees ronmental Permit (IEP), which is based on BAT. and charges but can also encourage the adoption Thus, Kazakhstan became the first CA country to of cleaner production practices. mandate the use of BAT for certain enterprises 16 Box 4: AQM legal and regulatory framework in the EU In EU countries, national legal frameworks usually transpose EU ambient air and emissions ceilings directives, and in some cases set more stringent standards for certain pollutants. More generally, the legal and regulatory framework in EU countries makes a clear distribution of roles and responsibilities between the central government and subnational authorities. Legal frameworks establish a combination of policy tools for AQM, including command-and-control and market-based instruments and enforcement tools (covering noncompliant jurisdictions, companies, or citizens and use several specific instruments). They tend to include specific legislation on the use of market-based instruments, such as environmental taxes and pollution charges. In some cases, the legal framework also integrates climate and air policies by setting common objectives, targets, and tools. 3.1.1. Air quality standards with pollutants in ambient air. Such an approach The setting of AQ standards in CA countries to the setting of AQ standards shifts the focus uses an outdated approach that goes back to the away from the main substances that cause the Soviet era. CA countries usually set two types largest health impacts, does not allow for the of AQ standards in the form of MACs—a one- proper assessment of improvements in AQ, and time MAC and a daily average MAC. In addition, does not incorporate the latest scientific findings. MACs are set for a large number of pollutants— The global scientific community widely agrees in most CA countries for over 600 pollutants. that the health impact of air pollution is driven The pollutant list for which MACs are set mixes by a handful of pollutants, 17 most importantly by pollutants commonly found in industrial settings 16 Newly commissioned category I enterprises are obliged to adopt BAT, whereas there are different periods for BAT application for existing category I enterprises in specific industries. 17 The WHO sets AQ guideline levels for particulate matter (PM 2.5 and PM 10 ), ozone (O 3 ), nitrogen dioxide (NO 2 ), sulfur dioxide (SO 2 ), and carbon monoxide (CO). Air Quality Management in Central Asia BACK TO CONTENTS 37 PM 2.5 , and that chronic exposure, as opposed to im target 2 (25 µg/m 3). Overall, MACs for key pol- a one-time exposure, to air pollution is the main lutants in CA countries are generally between two culprit for adverse health effects. This is why and six times higher than WHO guidelines. For most international AQ standards (for example, in instance, the PM 1018 daily average MAC in Uzbeki- the US and EU) generally adopt hourly average, stan is more than six times higher than the WHO daily average, and annual average concentration guideline (see Annex V: Air Quality and Ozone Stan- limit values to assess AQ improvements and the dards in Central Asia, EU, and WHO Guidelines). air pollution exposure impact on the population. In Part of the reason why some MACs in CA, addition to those, allowed number of exceedances especially for the coarse fraction of PM, are set at of the hourly and daily average limit values per a high level is because the CA region experiences year is also set. In the absence of such targets on at times severe SDS that make a major the number of exceedances, these limit values contribution to PM concentrations. However, are largely meaningless as they fail to address the instead of setting high MACs for PM 10 and PM 2.5 , chronic nature of exposure to air pollution. CA countries could follow the approach utilized in Even though CA countries set MACs for over 600 the EU and the US, which allows for discounting pollutants, MACs for the key pollutant from a the days with large natural dust contributions if it health perspective — PM 2.5 — are missing or have can be determined with sufficient certainty that been missing until recently. Most CA countries high pollutant concentrations are due to SDS. (Kazakhstan, the Kyrgyz Republic, Tajikistan, and Uzbekistan) now have legislated MACs for PM 2.5 , In addition to MACs, CA countries use AQ indexes. with the Kyrgyz Republic, Tajikistan, and Uzbeki- However, AQ indexes in CA do not include key stan establishing an annual average MAC in line air pollutants such as PM 2.5 in their calculation. with international best practices. The annual av- Some CA countries (for example, Kazakhstan erage PM 2.5 MACs in Uzbekistan correspond to and the Kyrgyz Republic) are updating their AQ WHO’s PM 2.5 interim target 1 (35 µg/m ), whereas 3 indexes to include key air pollutants from a health the annual average PM 2.5 MACs in the Kyrgyz Re- perspective and to adjust the index calculation public and Tajikistan correspond to WHO’s inter- formula to international best practices. 3.2. Committed executive and nested planning High-level government officials in advanced AQM cialized financing institutions supporting green systems usually champion air and climate policies investments. Generally, the executive institutions and have clear, long-term agendas to achieve in charge of AQM have the authorizing environ- goals and targets on both fronts. Across sector ment to influence and convene sectoral depart- ministries, AQ concerns are integrated into sector ments to implement AQ programs and plans. strategies, and there are policies and programs Nesting in countries with advanced AQM systems in place to incentivize the adoption of cleaner takes place between national and subnational technologies and processes. More generally, the AQM planning instruments. Nesting between institutional setup responsible for AQM compris- planning and development instruments, as well es multiple institutions playing complementary as coordination to address local transboundary roles in policy making, regulation, enforcement, pollution, is required by law. Also, planning funding, data, and information generation. There instruments for AQM identify specific measures is adequate institutional, technical, and financial and have access to national and international capacity, especially in regulatory agencies, to funding for their implementation. Developed enable the department to perform its mandated AQM systems define airsheds and use them as the functions. Also, the central government has spe- key units of analysis and planning as the airshed 18 Fine particulate matter with an aerodynamic diameter of less than 10 µg/m³. 38 BACK TO CONTENTS Air Quality Management in Central Asia approach considers all impacts on air quality and hence, local authorities’ role in strategic in a given settlement without the constraints AQM planning is limited. CA capitals and at that strictly adhering to administrative divisions times larger cities have a special status of places. An airshed approach ensures that all ‘cities of republican significance’. As such, some major sources of air pollution impacting a given CA countries have developed AQ action plans settlement are accounted for irrespective of and roadmaps for these cities that originate their administrative location – e.g. in a different from the government rather than from local municipality/region/country. authorities—for instance, the AQ action plan Generally, countries follow a formal process for Bishkek (Kyrgyz Republic) and the roadmaps to classify airsheds that are out of compliance, for improving AQ for Almaty and Astana which in turn, subjects these airsheds to specific (Kazakhstan). Nevertheless, local authorities management and monitoring requirements are prescribed roles in the implementation of by the federal or national government – in those top-down AQ action plans and roadmaps. the EU, non-compliance with AQ limit values In addition, local authorities in Kazakhstan have triggers the process of development, adoption, to report on the implementation of specific implementation and monitoring of airshed- Environmental Quality Indicators (EQIs) that specific AQM plans. The AQM planning processes include targets related to AQ. Regular, systematic tend to be informed by independent, timely, and monitoring of the implementation of such action quality scientific research and data. Planning plans, roadmaps, and EQIs, however, is often is backed by source apportionment, emissions spread across institutions and does not provide inventories, and health impact estimates. detailed information on how objectives are met and which policies and measures have proven to In general, AQM in CA is primarily top-down be effective, practical, and cost-efficient. driven. Local AQ plans or strategies are absent, 3.3. Horizontal and vertical coordination Advanced AQM systems generally include a bodies at the central executive level (housed at mechanism to coordinate policies and actions a cabinet office) having autonomy tend to be the across sectors and administrative divisions. This most effective. mechanism has a strong mandate and political AQM is cross-sectoral in nature as different eco- leadership to oversee progress, coordinating AQ nomic sectors and activities, including outside plans and gathering scientific knowledge and the remits of environmental authorities, contrib- data at the subnational level. The coordination ute to air pollution. The main sectors responsible body can carry out horizontal and vertical for air pollutants’ emissions in CA are energy, in- coordination and has some level of fiscal and dustry, transport, and agriculture. However, AQ administrative autonomy and decision-making coordination councils or interministerial bodies, power. The functions include monitoring and where representatives of sectoral authorities can advising on plans, coordinating actions, and formally meet, discuss, and make common deci- pooling scientific research. In some cases, it sions on AQM matters, are lacking in CA coun- can enforce AQM regulations and even resolve tries. It is not unusual for CA countries to adopt disputes between government actors. Often, policies and measures targeted at improving these bodies have wide participation, including AQ via Presidential Decrees rather than specific from the scientific community, the private sectoral legislation—for instance, Uzbekistan re- sector, and civil society, and they generally stricted the sale and use of vehicles lower than disclose all data related to their activities and Euro 4 by 2028 through Presidential Decree 338 plans and progress on the goals. Lessons learned from September 24, 2024. from international practice confirm that formal Air Quality Management in Central Asia BACK TO CONTENTS 39 3.4. Accountability and transparency Advanced AQM systems have robust AQ ለ Laboratory assessments: The concentra- monitoring and emissions information system. tions of certain pollutants, such as benzo(a) AQ monitoring networks are well-established; pyrene, VOCs, and heavy metals, are de- have good coverage; and integrate national, termined in AQ laboratories as they are not regional, and local measurement stations. directly measured by the AQ monitoring Emission inventories are built on detailed activity stations. Therefore, AQ laboratories are an data for the main sources of air pollution and use indispensable part of a complete AQ moni- the latest emission inventory methodologies. toring network. Often, AQ and pollution information systems are ለ Calibration of instruments: Periodic integrated and easily accessible by the public (usually annual) calibration of the (for example, web-based platforms that gather instruments used in the AQ monitoring monitoring data to inform the public and the stations and the AQ laboratories is needed government on AQ in management zones in real to ensure the consistency and reliability of time). Civil society and individual citizens have AQ monitoring results. Failure to conduct access to AQ information and tend to participate periodic calibration of the monitoring in policy making, while national legislation equipment and instruments can lead to enables them to file cases against governments unreliable and, at times, counterintuitive for a lack of compliance with air policy regulation. AQ monitoring results. Calibration of instruments is usually managed and carried 3.4.1. Air quality monitoring out in a central reference AQ laboratory. AQ monitoring is a continuous and systematic Legislation in most CA countries provides for process that requires certain technical a minimum number of AQ monitoring stations infrastructure and institutional arrangements based on population. For instance, the Kyrgyz to be in place. AQ monitoring includes four main Republic legislates that at least one AQ activities. monitoring station is required per 100,000 ለ Monitoring air pollution: Automatic, contin- inhabitants. However, some major cities in CA uous AQ monitoring is the standard in devel- lack AQ monitoring altogether and have neither oped AQ monitoring networks. Recent devel- an upgraded manual nor an automatic AQ station. opments in sensor technology have allowed AQ monitoring in CA is still dominated by manual for lower-cost sensors to be incorporated sampling and subsequent laboratory analysis of into AQ monitoring networks in addition to the samples as well as limited use of mobile AQ the reference grade AQ monitoring stations. stations. Manual sampling is generally carried Mobile AQ stations, on the other hand, are out three times per day while automatic stations deployed to monitor AQ at a particular loca- provide continuous monitoring and are thus the tion of interest, to identify the AQ impacts established reference method for AQ monitoring from a naturally occurring event or an acci- globally. Kazakhstan’s AQ monitoring network dent and to monitor AQ at a location with no is the only case in CA where the network is other AQ monitoring. dominated by automatic AQ monitoring stations ለ Data communication and storage: The AQ instead of manual ones. Data from automatic stations and low-cost sensors need to com- AQ monitoring in CA is also available from municate and send data to a central data- Unites States Environmental Protection Agency base, from which the authority responsible (EPA) reference grade stations installed at US for AQ monitoring can share the data with embassies in the region. The US embassies different institutions and communicate the stations are installed in six cities in the region data to the public. and monitor PM 2.5 and O 3 in the case of Tashkent. 40 BACK TO CONTENTS Air Quality Management in Central Asia The low number of automatic AQ stations in to ensure that the other main activities involved in most CA countries also means that monitoring AQ monitoring are also present. Capacity needs of some key pollutants (for example, PM 2.5 and to be established to maintain and calibrate the O 3 ) is either inadequate or lacking altogether. expanding number of AQ monitoring stations. Nevertheless, automatic AQ monitoring in CA Communication, data transfer, and storage countries is expanding with some countries 19 capabilities also need to be improved to manage currently investing or securing funds to invest the increased AQ monitoring data flow. Moreover, in such stations. In addition to installing new AQ laboratories need to be upgraded to provide automatic AQ monitoring stations, it is important a broader scope of assessments. Table 7: Number of AQ monitoring stations in CA countries National AQ monitoring network US embassy AQ Country Total AQ stations Automatic AQ stations monitoring sites Kazakhstan 170 130 2 Kyrgyz Republic 15 1 1 Tajikistan 16 1 1 Turkmenistan 17 — 1 Uzbekistan 63 20 1 Sources : Kazakhstan: KazHydromet; Kyrgyz Republic: KyrgyzHydromet; Tajikistan: World Bank; Turkmenistan: UNECE Turkmenistan Environmental Performance Review 2012; Uzbekistan: UzHydromet; US embassies AQ monitoring: https://gispub.epa.gov/airnow/index.html?tab=3. Best practice AQ monitoring networks also con- ለ Rural background: Generally, providing an sider the category of AQ monitoring stations, indication of the natural background con- that is, the context of where the stations are centrations of pollutants. Understanding the placed. Ideally, a robust AQ monitoring network naturally occurring background concentra- would include stations from each category. De- tions is essential for analyzing measures to pending on the category of AQ monitoring sta- reduce pollution in urban hot spots. tions, there are specific siting requirements for ለ Ecosystem monitoring stations: Assessing the stations such as distance from the main the impacts of air pollution on ecosystems. source and air flow around the station. AQ moni- toring stations are generally classified as follows: Historically, AQ monitoring in CA has focused on monitoring the impacts of industrial activities on ለ Industrial: Assessing the impacts on AQ of AQ, and hence, existing AQ monitoring stations particular industries. are usually placed near an important industri- ለ Traffic oriented: Assessing the impact on al source. Therefore, there is a need to diversify AQ of transport. the categories of AQ monitoring stations so that ለ Urban background: Assessing the impact the AQ monitoring networks in CA include all the on AQ of different urban activities such different categories described above. Moreover, as residential heating, construction, and since most of the existing AQ monitoring stations commerce. Urban background stations in CA were established during the Soviet era, it are usually the preferred category of AQ is necessary to not only upgrade the monitoring monitoring stations to assess the impact of equipment and the infrastructure but also to re- residential heating on local AQ. view the location of the stations. Urban develop- 19 The Kyrgyz Republic is currently implementing an Air Quality Improvement Project, supported by the World Bank, and will install 11 new automatic AQ stations. In addition, Uzbekistan has proclaimed targets for rapid expansion of the number of automatic AQ monitoring stations. Air Quality Management in Central Asia BACK TO CONTENTS 41 ment in some places might suggest that AQ mon- Thus, emission inventories omit an important itoring stations be relocated to new sites that are source of urban air pollution such as residential more relevant to current population exposure. heating. In addition, it is often the case that Best practice AQ monitoring networks provide for official emissions reporting, for instance, periodic assessments of the overall AQ monitor- through statistical offices, provides information ing network, which include analyses of the equip- on the total emissions of all pollutants rather ment and infrastructure upgrade needs as well than on emissions of specific pollutants. as the adequacy of current stations’ locations. Reporting total pollutant emissions does not provide adequate information for designing Developing an AQ monitoring network based on policies and measures to reduce emissions as reference grade automatic AQ monitoring sta- emission sources have different contributions to tions requires finances as the reference grade the emissions of specific pollutants. Therefore, instruments are costly and require periodic cal- the methodologies and approach to structuring ibration. However, numerous lower-cost sensors emission inventories in CA need to be updated are available on the market. AQ monitoring sen- considering the latest scientific findings on sor technology has been rapidly developing so emission factors, tiered approach to calculating it is worth considering establishing an AQ mon- emissions, and so on. itoring network that incorporates both reference Most CA countries regularly report national grade AQ monitoring stations and lower-cost GHG emissions, but this is not the case with sensors. Lower-cost sensors could be especial- air pollutant emissions. Only Kazakhstan and ly useful for assessing patterns in air pollution the Kyrgyz Republic have ratified CLRTAP and and providing information that can aid decisions provide national air pollutant emission inventories about where to place reference grade AQ moni- to the convention. However, even these toring stations. Some CA countries such as Ka- countries do not provide national air pollutant zakhstan and the Kyrgyz Republic use low-cost emission inventories every year. Moreover, AQ sensors and incorporate the information from CLRTAP expert in-depth reviews of the national those sensors in their AQ data communication. air pollutant emission inventories submitted by Most of the low-cost AQ sensors in CA were in- Kazakhstan and the Kyrgyz Republic concluded stalled through projects so the continued use of that emissions have not been calculated for the sensors once the projects end is uncertain. some key emission categories, and there were 3.4.2. Emission inventories inconsistencies in the time series used for emission estimations. 20 Together with AQ monitoring, emission inventories form an important pillar of the Part of the issues in regularly structuring overall AQM system. While AQ monitoring emission inventories in some CA countries is the lack of proper institutional setup for performing provides information about the level of air this task. Air pollutant emission inventories in CA pollution, emission inventories can identify the are usually compiled by national Hydromets or main sources of air pollution and assist source statistical offices. Sometimes the compilation of apportionment which is crucial for AQM. Once emission inventories is outsourced. Therefore, the main sources of air pollution are known, in addition to updating the emission inventories’ policies and measures to reduce emissions from methodologies and processes, there is a need those sources can be put in place. for clear definition of which institutions are Structuring of emission inventories in CA follows responsible for data provision on the one hand, outdated approaches and generally focuses only and the estimation of the national emission on industrial and mobile sources (transport). inventories on the other. 20 Reports are available on https://www.ceip.at/review-of-emission-inventories/technical-review-reports. 42 BACK TO CONTENTS Air Quality Management in Central Asia In addition to national air pollutant emission different emission sources to concentrations is inventories, the best practice AQM systems one of the main parameters used for prioritizing provide for the structuring of local (for example, policies and measures to reduce air pollution. city or region) emission inventories for pollution Another main use of AQ modeling is related hot spots. The national-level emissions reporting to forecasting. AQ models can be used to might not be representative of the individual forecast expected air pollution levels based on contexts of air pollution hot spots, and hence, meteorological information, 21 and thus inform downscaling national emissions to local level the population to take measures to reduce its might be misleading in some cases. Similarly, exposure to projected air pollution episodes, the local emission inventory of a capital city for instance. Moreover, such forecasting might not be illustrative for the AQ context of could trigger the implementation of policies a smaller city or an industrial city, for instance. and measures to reduce emissions from Therefore, local emission inventories are needed certain activities due to expected high levels to establish the key sources of air pollution in of air pollution. AQ modeling can also provide important AQ hot spots. information on the expected impact of a policy Local emission inventories are generally not or measure to reduce emissions from a given available in CA. There have been some attempts source—for example, the impact from switching to structure local emission inventories in away from heating with coal or mandating certain Kazakhstan through the work on local EQIs, but emission standards for vehicles. those have the same drawbacks as described The capacity for AQ modeling in CA is limited. above focusing on industrial and transport Some Hydromets, most notably, Kazakhstan’s emissions, using outdated methodologies, and Hydromet, use the System for Integrated lacking sufficient granular data for the emission Modelling of Atmospheric composition (SILAM) estimations. An approach to developing model 22 developed by the Finnish Meteorological local emission inventories adopted in the EU, Institute (FMI). FMI has trained other Hydromets for instance, is to require and support local in CA to work with the SILAM model, but the authorities in creating develop AQ plans if model is still run at FMI and experts in CA pollutant concentrations exceed limit values. countries do not work with it independently. Compiling local emission inventories is then 3.4.4. Communication one of the requirements for the local AQ plan so that the action plan, which contains policies and Communication and dissemination of AQ data measures to improve AQ, rests on solid analytical have improved greatly in recent years. The main foundations. way of sharing AQ data in CA countries used to be through bulletins (daily, monthly, seasonal). 3.4.3. Modeling However, currently most Hydromets in the AQ modeling is another important technical region publish AQ data on their websites. In component of AQM. Emission inventories de- addition, Kazakhstan, the Kyrgyz Republic, and scribe the level of pollution emitted by different Uzbekistan have developed apps that report sources, while AQ modeling provides information AQ information in real time and in some cases about the concentrations of pollutants due to the provide tips based on the level of air pollution. estimated source emissions. It is the air pollutant AQ data is also available from the AQ monitoring concentrations that drive the health damages of at US embassies in the region. AQ data from air pollution and define population exposure to monitoring at US embassies are published on a air pollution. Thus, knowing the contribution of dedicated website. 23 21 Advanced AQ forecasts also incorporate data on emissions from different sources. 22 https://silam.fmi.fi/doc/SILAM_v4_5_4_userGuide_general.pdf. 23 https://gispub.epa.gov/airnow/index.html?tab=3. Air Quality Management in Central Asia BACK TO CONTENTS 43 Some CA countries are also considering more di- for the adoption of AQ policies and measures. rect approaches to communicating AQ data—for As discussed above, strategic AQ planning has instance, through public display boards. Commu- mostly followed a top-down process. It is unclear nicating AQ data is only one part of the overall to what extent different stakeholders outside of task of improved awareness about the perils of government have been involved in developing air pollution and what individuals can do to im- the governmental AQ action plans for cities such prove the situation. However, simply sharing data as Astana, Almaty, and Bishkek. Stakeholder without attempts to educate and raise aware- engagement is an important step in any strategic ness about sources of air pollution, measures planning, and hence, it is important to strengthen that can be taken, and why it is necessary to take stakeholder engagement on AQ planning in CA, measures is insufficient. Educating and raising along with increasing awareness and improving awareness are essential for not only improving knowledge on the topic of air pollution. public awareness but also galvanizing support 3.5. Recommendations for Improved AQM The similar history of AQM development in CA certain actions listed in the table. Nevertheless, countries and the common challenges that the strengthening the institutional, legal, and policy countries are facing provide opportunities for framework for AQM; updating AQ standards and establishing a standardized approach to AQM approaches to emission inventories; expanding across the region based on cooperation and the AQ monitoring and modeling capabilities; knowledge exchange. Some CA countries have adopting the airshed approach; increasing advanced on certain aspects of AQM and thus stakeholder awareness and engagement; and can assist the other countries in the region. supporting regional cooperation in AQM are For instance, other countries can learn from shared priorities for CA countries. Kazakhstan’s experience on adopting BAT, To effectively improve AQM in Central Asia, it is expanding its automatic AQ monitoring, and prudent to initially focus on updating the legal and developing AQ modeling. On the other hand, the regulatory AQM framework. It is crucial to clearly Kyrgyz Republic and Uzbekistan have adopted define roles and responsibilities for AQM at the AQ standards for PM 2.5 in line with WHO interim national and local levels, to establish interminis- targets, and this experience can aid other CA terial AQ Coordination Committees for holistic countries to update their AQ standards. planning and decision-making, and to enhance Moreover, given the geography of CA countries strategic AQM planning at both national and local where many large cities are located close to the levels by adopting the airshed approach. Updat- border with other countries and the region’s ing AQ standards in line with WHO recommenda- location in the Dust Belt, transboundary tions and focusing on key pollutants will ensure a air pollution is significant in the region as more targeted approach to AQM. Subsequently, demonstrated in Chapter 2. Hence, the efforts expanding and upgrading AQ monitoring net- of one country alone will prove insufficient to works, updating emission inventory systems, improve AQ to a desirable level approaching and building capacities for AQ modeling are es- WHO guidelines. Regional cooperation is thus sential technical components for improved AQM. needed if CA is to combat its air pollution crisis. Additionally, fostering stakeholder engagement, raising awareness, and promoting regional coop- Table 8 outlines recommendations for priority eration will support a comprehensive and collab- short- to medium-term actions applicable to the orative effort to reduce air pollution and improve region as a whole while recognizing that countries public health across the region. within the region might have advanced more on 44 BACK TO CONTENTS Air Quality Management in Central Asia Table 8: Recommended actions for improving AQM in CA countries Legal and regulatory framework ለ Undertaking a review of the current environmental legislation relating to stationary sources, the setting of ELVs, emission fees and charges, industrial permitting, and emission monitoring is important to bring environmental regulation in line with international best practices. ለ The review of environmental legislation relating to stationary sources could also identify sectors, possibilities, and needs for the promotion of cleaner Review and update industrial production and/or the adoption of BAT. sectoral legislation ለ A review of sectoral legislation could identify areas in sectoral legislation for the key emitting where AQ considerations can be strengthened such as in legislation related to sectors residential heating, transport, and agriculture. ለ The review of legislation could also consider revising pollution fees and taxes to provide clear incentives for enterprises to reduce pollution and improve performance by strengthening incentives for enterprises to invest in cleaner production. In addition, developing a section of green taxonomy for AQ improvement projects could attract funding from additional sources for such projects. Reassess the list of ለ Focus on the key ambient air pollutants from a health perspective as pollutants for which recommended by the WHO and environmental agencies such as in the US AQ standards are and EU. adopted ለ Include PM 2.5 in the list of pollutants for which standards are adopted. ለ Developing AQ standards and targets based on the work of the WHO and best Update AQ standards practice approaches is recommended as a critically important action. in line with WHO ለ CA countries could consider WHO interim targets as incremental steps in a recommendations progressive reduction of air pollution toward meeting WHO guideline values. Committed executive ለ The proper functioning of governmental structures responsible for AQM is Strengthen essential for an effective AQM system. Policy making and technical roles governmental roles, should be clearly distinguished and have sufficient mandate to undertake AQM responsibilities, and activities. structures to support ለ Building the capacity of governmental structures to undertake the different an effective AQM requirements of AQ policy formation such as setting target, designing and system assessing policies and measures, compiling an AQ plan, and tracking progress toward achieving the stated targets is a critically important action. Set up a platform/ ለ A platform/mechanism to share information, knowledge, and experience will mechanism for assist CA countries to accelerate the implementation of best practices in information and AQM—for instance, on adoption of BAT and update of AQ standards. knowledge exchange ለ CA countries could identify sources of transboundary pollution and put forward Agree on actions to an action plan on reducing pollution. Potential sources include industrial reduce transboundary enterprises in border regions as well as SDS. air pollution ለ Encourage CA countries that have not joined CLRTAP to do so. Nested planning Strengthen strategic AQM planning on ለ AQ targets are included in various high-level strategic documents in CA national and local countries. However, there is a lack of targeted AQM strategies that can set levels, adopting the the direction for AQM in the given country in a systematic and holistic way. airshed approach Air Quality Management in Central Asia BACK TO CONTENTS 45 Table 8: Recommended actions for improving AQM in CA countries ለ The mandate and capacity of local authorities to engage in AQM need to be strengthened. In addition, AQM at local level needs to adopt the airshed ap- proach in order to fully account for the sources of air pollution in population cen- ters. A possible approach to strengthen AQM planning at the local level through assistance for the definition of (key) airsheds - e.g. the airsheds where the ma- jority of population lives. Subsequently, local authorities could be required to develop AQ plans for airsheds that show noncompliance with AQ standards, like the approach taken in the EU. Technical assistance and guidance on the development of AQ plans could be provided to local authorities. The AQ plans should as a minimum contain emission AQ trends, inventory compilation, source apportionment, and an action plan with policies and measures to improve AQ. Share the ለ AQ modeling is an example of an AQM activity that can be implemented on a implementation of regional scale. Hence, a CA country with more developed modeling capacities could some AQM activities conduct modeling for the entire region in cooperation with the other CA countries. on a regional scale ለ Sharing of other AQM activities could also be discussed. Horizontal and vertical coordination Establish ለ AQM does not fall only under environmental authorities and has key links with interministerial other sectors such as energy, industry, transport and agriculture. Therefore, AQ Coordination national coordination mechanisms need to be put in place that will ensure that Committees sectoral policies and measures consider AQ impacts. Accountability and transparency ለ The AQ monitoring network in the different CA countries needs to be expanded to provide monitoring to major cities in the region. Automatic AQ stations are the reference method for AQ monitoring. Incorporating lower-cost sensors in Expand and upgrade the AQ monitoring networks’ expansion could also be considered from a cost- AQ monitoring effectiveness perspective. ለ The existing AQ monitoring sites in CA countries need to be reviewed in terms of needs for equipment and infrastructure upgrades and suitability of monitoring sites’ locations. ለ Upgrade and develop additional laboratory capacities for AQ analyses. ለ Ensure capacities for periodic maintenance and calibration of monitoring Strengthen technical equipment. capabilities of the AQ ለ Ensure sufficient data processing and storage infrastructure for expanded AQ monitoring networks monitoring capacities. ለ Build capacity for expanded AQ monitoring capabilities. ለ Update emission inventory methodologies in line with best international Update and practices. strengthen the ለ Clearly define roles and responsibilities for emission inventory compilation as emission inventory well as for activity data collection and data sharing arrangements. systems to meet ለ Periodically (at least annually) compile national emission inventories for key international best pollutants. practice ለ Support the compilation of local emission inventories in key air pollution hot spots. ለ Build capacity for strengthened emission inventory processes. Strengthen capacities ለ Build capacity for key institutions to undertake AQ modeling. for AQ modeling ለ Some modeling can also be conducted on a regional basis. Strengthen stake- ለ Involve various stakeholders in strategic planning for improving the AQM holder engagement system. Support awareness ለ Design and implement educational and awareness raising campaigns on the raising and education sources of air pollution and potential solutions to improving AQ. Source: Original compilation. 46 BACK TO CONTENTS Air Quality Management in Central Asia 4. Financing Air Quality Improvement 4.1. Cost of inaction Investing in AQ improvements yields numerous tion-related crop damage. 27 Cleaner air contrib- economic benefits, including reduced health- utes to higher quality of life enhancing the liva- care costs, increased productivity, and preven- bility of cities, economic competitiveness, and tion of premature deaths. Analyses of existing stronger local economies, making cities more and emerging data by The Lancet Commission attractive for residents and tourists alike 28 . on Pollution and Health confirm that air, water, These comprehensive benefits underscore the and soil pollution have significant health and significant economic advantages of investing in economic costs 24 and enhanced AQ leads to AQ improvements. Although not all these ben- better cognitive development in children, 25 re- efits could be assessed, reduction of mortali- sulting in long-term productivity gains. Better ty attributed to AQ improvements in CA could AQ improves labor productivity26 and reduces be estimated to range between 3 percent and future healthcare demands. Additionally, it AQ 5 percent of the equivalent of regional GDP in boosts agricultural yields by reducing pollu- 2021. Table 9: Annual premature deaths attributable to PM 2.5 pollution and associated economic costs in CA Premature deaths attributable to PM 2.5 pollution Economic Economic cost, Country cost, range, range, % GDP Average US$ billions equivalent in 2021 Range number Kazakhstan 12,694 8,887–16,911 10.8–12.7 5.5–6 Kyrgyz Republic 2,234 1,135–3,507 0.1–0.3 1.5– 3 Tajikistan 3,278 1,876–5,005 0.1–0.3 1–3 Turkmenistan 3,644 2,196–5,588 1.8–2.9 3.5–6 Uzbekistan 27,099 16,999–36,518 2.4–5.4 3–7 CA region combined 65,384 45,937–83,790 15.2–21.7 3–5 Source: Original analysis based on GBD 2021, World Bank. 24 https://www.thelancet.com/commissions/pollution-and-health. 25 Alter et (2024). https://doi.org/10.1186/s12940-024-01122-x 26 Neidell, M., Pestel, N. Air pollution and worker productivity. IZA World of Labor 2023: 363 doi: 10.15185/izawol.363.v2. 27 There is evidence for the impact of air pollution on agricultural productivity and health in Central Asia (Uzbekistan) – see Akramkhanov.A., S. Strohmeier, Y.A. Yigezu, M. Haddad, T. Smeets, G. Sterk, C. Zucca, A. Zakhadullaev, P. Agostini, E.S. Golub, N. Akhmedkhodjaeva, C.S. Erencin, 2021. The Value of Landscape Restoration in Uzbekistan to Reduce Sand and Dust Storms from the Aral Seabed. © World Bank https://documents1.worldbank.org/curated/en/750031635227796665/pdf/The-Value-of- Landscape-Restoration-in-Uzbekistan-to-Reduce-Sand-and-Dust-Storms-from-the-Aral-Seabed.pdf. 28 https://www.cleanairfund.org/theme/economics/. Air Quality Management in Central Asia BACK TO CONTENTS 47 Air pollution, particularly PM 2.5 pollution, OECD countries, as well as population and GDP leads to economic losses due to air pollution’s parameters for the individual CA countries, 31 adverse health impacts, including links to this study estimates that the health cost of premature death. The latest (2021) GBD global PM 2.5 pollution in the CA region ranges between study 29 estimates that over 65,000 premature US$15.2 and US$21.7 billion per year. Hence, deaths in CA could be attributed to ambient the cost of inaction on air pollution is significant PM 2.5 pollution annually (see Table 9). Using and measures to improve AQ will deliver both Value of Statistical Life (VSL) estimates 30 for health and economic benefits. 4.2. Challenges in financing air quality improvement There are clear economic benefits from investing economies, which might already be grappling in AQ improvements. Air pollution is the leading with high debt levels. One encouraging trend environmental risk to health, causing 7 million is that funding for outdoor AQ projects was premature deaths each year at an estimated still larger than funding for fossil fuel projects global cost of US$8.1 trillion in 2019, equivalent in 2021, due to actions to begin phasing down to 6.1 percent of global GDP (World Bank 2022b). coal-fired power around the world. More than 95 percent of deaths caused by If the economic benefits are evident, why is air pollution occur in low- and middle- income AQ funding not keeping pace with the pollution countries. A recently released report on the state problem? Short-term horizons may have of global AQ funding has found that every US$1 contributed to this phenomenon. A recent study spent on air pollution control yields an estimated on air pollution in cities found no strong patterns US$30 in economic benefits (Strinati et al. between  city competitiveness  and  pollution. 2023). Yet, financing for air pollution measures Despite high levels of pollution, usually has been dismally low. Nevertheless, financing exceeding WHO levels, many cities in Asia, of climate mitigation – one of the main sources particularly in China and India, are considered of financing in the environmental sector – could highly competitive. have significant air quality co-benefits that This is because policy makers may have could be enhanced if the air quality and climate prioritized short-term gains in human and capital mitigation synergies are further incentivized in productivity over the benefits of a healthy the provision of climate mitigation finance. environment and clean air. Such decisions International development funding and public can result from not factoring environmental climate finance specifically targeting air pollution degradation, and hence the decline of natural remain significantly low, with only 1 percent of capital, into the policy calculation. Such is the international development funding (US$2.5 case for Mexico and China as they rose to upper- billion per year) and 2  percent of international middle-income status (World Bank 2020b). But public climate finance (US$1.66 billion per year) this equation needs to be updated if developing committed to air pollution control over the last countries are to make longer-term development six years. gains. Numerous studies have shown that a The report also noted a shift in the type of growth model built on the unchecked decline of financing, with a significant portion of AQ the natural capital stock is not sustainable in the funding being provided in the form of loans long term and that all high-income cities have (92 percent) rather than grants. This trend low pollution levels. In addition, experience from has implications for low- and middle-income rapidly developing countries such as China has 29 Global Burden of Disease Study 2021 (GBD 2021) Data Resources | GHDx. 30 VSL estimates for OECD countries range from US$3.2 million to US$3.8 million. 31 Data for 2021 for the individual CA countries about population and GDP-related parameters were obtained from World Bank Open Data | Data. 48 BACK TO CONTENTS Air Quality Management in Central Asia shown that as economies grow, the emerging emission source in question. For instance, large middle class living in urban areas starts to point sources such as industrial enterprises have demand clean environment, including AQ. established revenue streams and even though, in In general, financing for AQ improvement most cases, emission reduction measures entail requires aligning policy reforms with the design high capital costs, those costs can be passed of tailored financial tools from both the public on to the consumers or internalized. On the and private sectors. Targeted public funding other hand, implementing emission reduction for air quality is mainly focused on air quality measures for multiple point sources with highly monitoring infrastructure and on enforcement decentralized ownership, such as non-district of legislation. The sources of air pollution span residential heating or in other areas where across different economic sectors, and hence, multiple actors are involved such as in passenger the mechanisms to finance emission reduction transport, may require incentives or support measures depend on the characteristics of the mechanisms. 4.3. Options for public funding and financing, including through concessional finance The main benefit of clean air in the form of residential heating and cooking is one example. improved health outcomes is a public good and In contrast to pollution from industrial facilities, needs to be backed by public finance. However, pollution from decentralized residential heating governments in general dedicate a very small and cooking comes from multiple point sources share of their budgets to air pollution, with that with diverse ownership structures. The source spending focused mainly on monitoring and of co-financing in these situations would likely enforcement. Governments are not the primary come from household savings. Direct subsidies financier of abatement measures. Pollution to make the upfront cost of low-polluting heating abatement is ultimately the responsibility of and cooking equipment more affordable may be the polluter, who must finance the abatement justified, particularly in the case of low-income measure but even so public finance can be used households. By extension of this argument, to incentivize investments in pollution reduction district heating companies providing subsidized measures by the private sector. The government community heating may also be included may introduce a program of state incentives and for direct subsidies to afford low-polluting subsidies to defray part of the cost of meeting an technology. emission standard. A policy of state support can Moreover, fiscal interventions such as eliminat- lower the burden for households and ease the ing or repurposing subsidies are other possibil- way for enterprises to obtain credit. ities for financing air quality improvement ac- From a financial standpoint, public funding tions and influencing the choices and behavior of then becomes the crucial factor in mobilizing businesses and households. The first step in this private funds, whether through enterprises’ or process is to identify the subsidies that could be households’ self-financing or bank financing. eliminated or repurposed to facilitate achieve- This approach is particularly effective for de- ment of air quality and broader environmental risking private investments. Market-based goals (e.g. climate change mitigation and reduc- approaches work well for commercial entities tion of short-lived pollutants that affect both with a revenue stream. The level of state support human health and contribute to climate change). needed for these entities to secure market- After identifying the subsidies to be eliminated/ based financing may be minimal, for example, a repurposed, an analysis needs to be carried out credit guarantee. Other situations may call for that assesses the potential impacts on air pol- more generous state support. Decentralized lution and on vulnerable households from sub- Air Quality Management in Central Asia BACK TO CONTENTS 49 sidies elimination/repurposing. For instance, if waste, this would have a negative impact on air fossil fuel subsidies were eliminated, then such quality) and what will be the impact and the po- an analysis should consider what might fossil fu- tential support needed to cushion the effect of els be substituted with (if fossil fuels are substi- subsidies’ elimination/repurposing on vulnerable tuted with unsustainable biomass or burning of households. Box 5: Public funding for AQ in selected countries United States: The US EPA administers the Clean Air Act grants program, which allocates funds to state and local AQM agencies to support their air pollution control programs. Additionally, the Diesel Emissions Reduction Act (DERA) program provides funding to reduce emissions from older, diesel-powered engines. China: The Chinese government has implemented various funds and financial policies aimed at reducing air pollution. For example, in response to its severe air pollution issues, particularly in cities such as Beijing, the government has set up funds to support clean energy projects and pollution control measures. United Kingdom: The UK supports a global Clean Air Fund, along with various philanthropies, to finance initiatives to improve AQ, including the development of clean air zones and investment in electric vehicle infrastructure. India: India launched the National Clean Air Programme (NCAP) that aims to reduce particulate matter pollution by 20–30 percent relative to 2017 levels by 2024. Various funding mechanisms, including state and central government funds, support the implementation of city-specific AQM plans under this program. Another common form of public sector funding This illustrates the importance of such AQ action is through targeted funds. Many environmental plans not only from an analytical and policy- funds are created to deal with a particular type of making perspective but also from a funding environmental hazard such as oil spills or storage perspective. Regional/city AQ action plans of hazardous waste. On the other hand, the generally lack in CA, which might also impede concept of an Air Quality Fund is for supporting public funding for AQ improvement projects. policies, innovations, and infrastructure China achieved perhaps the most notable im- developments necessary to reduce air pollution. provement in AQ among developing countries While funds can act as a catalyst, the main cost in recent times. Its ambitious National Air Pollu- of pollution abatement and/or clean technology tion Control Plan (2013–2017) cost an estimat- is ultimately borne by the polluters. The main ed Y1.65 trillion, funded mainly by national and source of finance for such funds is likely to be local governments. Still, this was only a fraction from pollution charges or resource fees. This of the roughly Y67 trillion economy32 at that funding source is likely to be small, and other time, and the estimated public health benefit public funds will have to supplement. In addition, outweighed the cost of action 33 by34 , 35 member the examples in Box 4 suggest that generally states, the spending to maintain good AQ can public finance is allocated to support the be as low as less than 0.1 percent of GDP. implementation of regional/city AQ action plans. 32 World Bank data for 2015. 33 https://link.springer.com/article/10.1007/s42524-019-0074-8. 34 The US EPA’s total budget for 2023 was US$11.8 billion, of which just US$1.1 billion aimed at air pollution. The US government budget that year was roughly US$6 trillion. https://www.epa.gov/newsreleases/epa-releases-fy-2023-congressional-budget-justification. 35 https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Government_expenditure_on_environmental_protection. 50 BACK TO CONTENTS Air Quality Management in Central Asia In keeping down the overall economic cost of its contribution to other environmental AQM, the government has the important role and socioeconomic development objec- of prioritizing the most economically efficient tives as well as risks to achieving these measures and targeting its policies and support objectives; to maximize the financial contribution from ለ Policy consistency and coherence, the private sector and households. Effective coordination, and integration, among prioritization and signaling can be supported by others, ensuring that achieving the AQ the following: objectives does not significantly harm a. A strong analytical underpinning for pollution other policy objectives; abatement programs. The analytical work ለ Consideration of other stakeholders , should support a robust prioritization of such as impacts on the public (citizens investments based on and civil society organizations) and busi- ለ Cost-effectiveness and accounting for nesses, and their views; and benefits, including positive co-benefits, ለ Realism — an assessment of the tech- such as climate change mitigation and nical feasibility, financial affordability, negative externalities; political accepta-bility, and overall ease ለ A holistic assessment of the AQ invest- of implementation. Lessons should be ment, including cross-sectoral links and gleaned from other countries. Box 6: Designing good AQ programs The analytical underpinnings of an air pollution program are a crucial ingredient for a cost- effective investment program that meets financing requirements. When cost and pollution benefits are optimized, investment projects are more attractive whether financed by public or private sources. In one of the more successful World Bank projects in air pollution, China - Hebei Air Pollution Prevention and Control Project (World Bank 2020c), one of the critical success factors was the analytical foundation leading to the Hebei action plan. This plan’s success stood out due to its reliance on source apportionment and modeling to identify sources and contributions of various pollutants to AQ in a specific area. Also critical was a methodology for prioritizing each set of measures to maximize the benefit to AQ relative to other criteria such as the cost of measures and ease of implementation. b. Targeted incentives and policies, including year. Other criteria, such as positive impact on establishing clear emission reductions health (primarily due to reduced exposure to across sectors. PM 2.5 and its precursors), and (or) city economy, c. Support for financial structures that mobilize or the economic and financial performance (or additional capital. status) of project proponents, and other posi- tive externalities, as well as cost-effectiveness d. Prioritization of scarce public resources on and implementability (here meaning: an AQ proj- monitoring, enforcement, and institutional ect or measure is politically acceptable or even capacity building as these will not be borne most welcome, and relatively easy to implement by private owners. from technical, economic, and resource per- Applying the principles above, project selec- spectives, over a reasonable time frame) would tion criteria shall not be limited to quantities also be important. Emission quantities would be of emission reduction, measured in tons per adjusted to the relative hazard of the respective Air Quality Management in Central Asia BACK TO CONTENTS 51 pollutant (for example, reducing emission of di- sites, or a CHPP running in coal and mazut) to oxins by just a few kilograms per year, combined neighboring territories (in case of Almaty to with PM 2.5 emission reduction, may outweigh Almaty oblast), thus worsening the environ- quite a few tons of CO 2 , or methane [CH 4 ] emis- mental situation there. This trade-off should sions). be accounted for. An integrated approach is therefore required ለ Large business may wish to comply with to achieve the set social (health), environmen- existing requirements related to emissions tal (including decarbonization), and economic (for example, not exceeding the BAT-ELVs), objectives considering interlinkages, synergies, thus reducing pollution charges and fines. and trade-offs, between priority actions toward In some cases, they also enjoy cost savings each objective. For instance, cost-effective (for example, the cost of operating and strategies for (a) air pollution reduction and (b) maintaining the special landfills for ash and decarbonization often would differently priori- slag waste [ASW] generated by CHPPs) or tize pollutants, emission sources, and interven- higher resource productivity (a typical result tions. Hence, while focusing on maximizing syn- of introducing BAT). ergies, AQ projects would have the overarching ለ SMEs would prioritize measures resulting objectives to reduce health risk and decarbonize in cost savings (for example, reducing the economy. payments for heating and hot water supply Furthermore, different groups of economic or reducing pollution charges and fines, if agents: households, business entities—large applicable) and reducing risks. ones and small and medium enterprises (SMEs), ለ Households would be interested in both (a) — and the government (public authorities at improving the environmental dimension of the national and sub-sovereign levels) may their quality of life (for example, reducing have somewhat different (each time multiple) indoor air pollution from using coal, wood, or objectives, priorities, and incentives for heating oil for heating, hot water production, investing, or not, their resources in this or that and cooking) and (b) achieving time and cost class of measures. savings (for example, connection to piped ለ The national government may wish (a) to natural gas helps save time previously spent progress toward the green economy model, on identifying suppliers of coal or wood or more resource efficient and less carbon heating oil; transportation and storage costs; intensive; (b) to have a healthier population and time spent to deliver a portion of fuel with longer life expectancy and more years from storage facility to the individual stove or of productive work and life; and (c) to fulfil boiler). Furthermore, as their income may not its international commitments, not least be enough to cover respective investment under the Paris Agreement on climate (NDC costs in one go, access to a multi-year soft pledges). loan, ideally complemented by tax benefits ለ Oblast and City Akimats may wish to im- (for example, deduction of the amount prove the AQ and overall environmental situ- invested from the personal income tax ation on their territory, thus reducing health base), 36 would be a good incentive. risks and making the oblast or city a more at- ለ Investors, financiers, and development tractive place to live, work, and do business partners (DPs) would have their own as well as for tourists. However, some im- priorities and incentives to invest in, or provements could be made by transferring (co-)fund, projects aimed to improving AQ. polluting sources (such as sludge filtration For DPs, compliance with their mandate fields, or municipal waste landfills or dump and priorities set in the country or regional 36 Such practice exists in some EU countries, for example, in France. 52 BACK TO CONTENTS Air Quality Management in Central Asia development assistance strategies might be In conclusion, respective project will have much a key criterion. more chances to be implemented if it corre- ለ Moral incentives (recognition of their sponds to priorities of public authorities, the proj- contribution to improving the state of ect proponent (project owner), and the financier environment in the city [oblast, country], to or investor (and DP, if applicable), and all of them decarbonization, or to transition to greener have sufficient incentives to implement and (or) economy) would be important for all actors. support the project. 4.4. Incentives for private sector investments 4.4.1. Policy instruments that are available Though effective in theory, pollution charge to incentivize private investments: systems in their current form have not been an pollution charges, feebates, regulations, effective tool for CA countries in incentivizing and the pros and cons of various options pollution abatement measures and managing Pollution charges, while a common tool in AQ. Comparative studies have found that ELVs environmental policy, are often insufficient as in CA, particularly for air pollutants from large an incentive for polluters to invest in pollution combustion plants, were more lenient than in control. Moreover, pollution charges alone many OECD countries (OECD 2019). The process cannot generate the full amount of funding for setting ELVs was also less transparent than needed for key investments to improve AQ. the comparators. Moreover, the pollutant of main Pollution charges may be set too low, and concern in CA, which is PM 2.5 , is not among the polluters may prefer to pay the charges — and pollutants subject to the charges in all countries. continue polluting — rather than make significant As pollution levies were revised upward over time, investments in cleaner technologies. In some there was also no clear relationship between the industries where the most cost-effective controls rates and the marginal cost of abatement or the involve fundamental process changes or the use marginal social cost of pollution. As there was of relatively new technologies, the large capital often limited communication from authorities outlay may be prohibitive. This is particularly about the technical bases of rate revisions, true for smaller companies with poor access pollution charges were treated more as a form to financing or those in financial distress. More of revenue collection than an instrument to often, poor enforcement of pollution charges, incentivize pollution control. ambiguous regulations with loopholes, or gaps in jurisdictional oversight are reasons that pollution A more effective system would involve shifting charges exert no pressure on polluters. the focus of the charges from penalizing noncompliance to re-incentivizing compliance Pollution charges in their current form over (for example, through a system of rebates). much of CA have had shortcomings. From This would also avoid overlap with other policy the start of its post-Soviet era, CA countries instruments, such as a carbon-based fuel excise. have implemented pollution charges but not These can improve the efficacy of the system at levels sufficient to incentivize polluters and will position the pollution charge system to to reduce their emissions. Revenue was not play a more effective role in a broad package exclusively dedicated to environmental cleanup of different instruments aimed at incentivizing and protection initiatives, and the principle environmental compliance. of ‘polluter pays’ was a recent adoption. This principle assigns the cost of pollution to the While improvements can be made to the system, polluting party, with the cost set to prevent and pollution charges themselves are not sufficient compensate for damages caused by emissions to shift polluter behavior. Neither are they suited into the natural environment. to raise sufficient funding to substantially finance Air Quality Management in Central Asia BACK TO CONTENTS 53 the pollution abatement or process technology terprises to implement, as a minimum, bench- choices necessary to meet higher environmental mark levels of resource efficiency and clean standards. The revenue collected from pollution industrial production. In the EU, these take the charges in CA is not significant. Given the low form of BAT. BAT establish the minimum avail- base, there is scope to raise air pollution charges able techniques to reduce emissions and pollu- and other environmentally relevant revenue. 37 tion for specific industries. Enterprises operat- Nevertheless, the total revenue that can be ing in each sector are then obliged to apply for mobilized will still be insufficient as a primary IEP based on adoption of BAT. BAT also have an source of air pollution financing. impact on ELVs because emission limit values In addition to or instead of pollution charges, for sectors for which BAT are available are set most advanced AQM systems (for example, in as ranges based on the emission performance EU and the US) adopt legislation that obliges en- prescribed in BAT. Table 10: Air pollution charges in CA, as % of GDP Air pollution-related charges (% of GDP, 2020) Kazakhstan 0.30 Kyrgyz Republic 0.13 Tajikistan 0.03 Uzbekistan n.a. Source: World Bank and OECD data for environmentally related tax revenue. Kazakhstan is currently the only CA country Key barriers to private sector funding for AQ that mandates the adoption of BAT for improvement are the risks associated with the certain enterprises. Newly commissioned investment and at times the uncertain regulatory category I enterprises must obtain IEP based environment. Therefore, clear regulations that on BAT to start operating, whereas existing unequivocally point to a move toward cleaner category I enterprises have deadlines for the production are needed. Public support through implementation of BAT. The pollution fees and de-risking mechanisms such as guarantees, charges for those enterprises are progressively concessional loans, and blended finance will increasing every few years until the specified help private investment in AQ improvement deadline. On the other hand, enterprises that measures. Public-private partnerships (PPPs) have obtained IEP based on BAT are not liable to are another potential instrument that can be pay pollution fees and charges. utilized to lower the financial risk of reducing emissions for private enterprises. 4.4.2. PPPs can help the private sector fund and finance some of the large (risky) Integrating AQ financing opportunities with investments that are needed in transport climate change mitigation funding can also and residential/commercial heating sectors support more active private sector participation Private sector financing is essential for in the funding of emission reduction measures. achieving substantial emission reductions, Air pollution and climate change are intricately especially when there is a need for large-scale linked, given that emitting sources such as coal- technology investments that fund and build fired power plants and diesel-fueled vehicles new infrastructure, like infrastructure for EVs contribute to both problems. Therefore, there is deployment, energy efficient equipment in a strong alignment between the climate change SMEs and homes, and green infrastructure. mitigation agenda and the local AQ agenda. 37 The average OECD level of environmentally relevant tax revenue is approximately 1.6 percent of GDP. 54 BACK TO CONTENTS Air Quality Management in Central Asia Because climate change mitigation strategies novative ways in which concessional finance is will, for the most part, also yield improvements disbursed, implemented, and monitored. This in AQ and health at the local level, every section provides a brief summary of some of opportunity should be taken by those concerned those new financing models and innovations. primarily with local air pollution to leverage on Moreover, development finance institutions financing sources and strategies for climate (DFIs) are increasingly using new financing change mitigation. models to attract private sector participation in AQ financing. The Kyrgyz Republic Air Qual- 4.4.3. Innovative financing schemes to use concessional finance to attract more private ity Improvement Project 38 is a recent example investment from CA of leveraging innovative concessional financing for residential heating, involving the Concessional finance is a conventional option establishment of a revolving mechanism to sup- to leverage financing for AQ improvement proj- port transition to clean heating (see Box 7). ects. However, there are multiple new and in- Box 7: Revolving mechanism to support transition to clean heating in the Kyrgyz Republic The Kyrgyz Republic Air Quality Improvement Project is backed by a US$50 million concessional loan from the World Bank’s IDA, spread over 50 years with a 10-year grace period. The project’s objectives are to strengthen the capacity of the Kyrgyz Republic to manage AQ and to reduce net PM 2.5 and GHG emissions in Bishkek. The project has three components: Component 1: Strengthen the AQM system in Kyrgyz Republic Component 2: Support the adoption of clean heating solutions Component 3: Improve urban greening. An innovative financing mechanism will be utilized in the execution of Component 2; see the figure below. MoF PIU Special account for revolving Credit line Repayment Business loan PFIs Suppliers Repayment Repayment of loan Payment for Consumer lending equipment Clean heating equipment and O&M and services HHS HHs (clients) As a result of implementing a revolving mechanism, it is expected that the number of supported households by the project will almost double. It is estimated that the initial investments will support 13,000 households during the project duration with scale up to 20,000 households within 10 years of implementation of the revolving mechanism. 38 Development Projects : Kyrgyz Republic Air Quality Improvement Project – P177467. Air Quality Management in Central Asia BACK TO CONTENTS 55 Box 7: Revolving mechanism to support transition to clean heating in the Kyrgyz Republic The support for the adoption of clean heating solutions by single-family homes (SFHs) in Bishkek will address the main barriers around high up-front costs and lack of access to financing. Sub- loans in local currency through the participating financial intermediaries (PFIs) will be provided to SFHs interested in switching to clean heating solutions and private entities interested in expanding their businesses to clean heating technologies and services in Bishkek. The innovative approach in the project is that the financing will be implemented though a revolving mechanism at two levels: at the PFIs and the Ministry of Finance (MoF). The loan from the MoF to the PFIs will cover the cost of funds and the foreign exchange risk, and it will be provided on a longer-term maturity than that of sub-loans, so the PFIs may use the funds multiple times until full repayment to the MoF. Similarly, the MoF will manage a special account where the returned funds will be accumulated and channelled to the same or other PFIs, thus maximizing the number of beneficiaries and impact on AQ. It is expected that such revolving mechanisms will continue for around 20 years, with the intention to extend the financing to more households after the project closing. Both state-owned and private commercial banks will have access to financing under this component. The PFIs will provide sub-loans to households with expected five years maturity. The size of the single loan for households will likely not exceed US$5,000 per household. They are fully responsible for repayment of the loan to the MoF. They can reuse the funds during the 10-year period, allocating funds multiple times before repaying the MoF. The PFIs may also provide loans to energy service companies (ESCOs). ESCOs can use the loan to supply and install clean heating systems at SFHs, then collect the costs of investments and operation and maintenance (O&M) according to the agreement with the household. It is also possible to receive a license for heat supply services and provide such services at the defined tariff for such services. Source: Original elaboration by World Bank Group staff. An alternate financing structure involves a The innovation around this type of financing is bond issuance, usually from a local or municipal the option to include a results-based payment authority that incorporates a results-based mechanism with a donor, philanthropic institu- payment mechanism. This could be distinct tion, a DFI, or the issuing body’s national govern- from general municipal fundraising. One ment. These so-called outcome bonds mobilize mechanism developed by the World Bank is so- private capital by linking the bond’s returns to called ‘Emission reduction-linked bond’. For verified achievement of predefined goals (for any kind of ‘environmental’ or ‘air quality’ bond, example, emission reduction goals). These pay- the appropriate analysis should already be in ments lower the effective interest cost when place—such as an emissions inventory and AQ specific air pollution/quality milestones are met. modeling to effectively identify and assess air Concessional funding may also be used to fur- pollution and GHG emissions sources. In the ther lower financing costs. The financial struc- case of authorities new to bonds, additional ture for emissions reduction-linked bond and for technical assistance should be applied, such sustainability-linked bond in Rwanda, described as (a) support for related institutions and below, is presented in Annex IV. stakeholders, (b) project preparation and Sustainability-linked debt is another result- structuring, and (c) bond structuring advisory based, innovative financing instrument that services. The authorities could then on-lend the can be tailored to support AQ improvements. bond proceeds through a call for proposal or a In sustainability-linked debt, interest rates are PPP with a commercial bank to fund pollution tied to the achievement of predefined targets by abatement projects. the borrower. For AQ improvement, predefined 56 BACK TO CONTENTS Air Quality Management in Central Asia targets can include emission reduction and/or as emission reductions or increased energy concentration reduction goals, for instance. efficiency to trigger the interest step-downs. These new financing models are not new and The execution of these new financing have been executed to advance environmental mechanisms requires different actors to and social objectives in different countries. For come together and is best realized through a example, in 2023 the Government of Rwanda facilitator40 capable of convening governments, used US$10 million of its IDA allocation to investors, and donors. Oftentimes these credit enhance a sustainability-linked bond facilitators are incentivized by the option of issued by the Development Bank of Rwanda. concurrently participating directly as an investor Bond proceeds were on-lent to participating or in a credit enhancement role. Financing financial institutions that reported on specific structures of this kind have the added benefit performance indicators. On achieving set levels of leverage, thereby increasing the available of the performance indicators, the Development financing. For a given amount of IDA, multiples Bank was rewarded with step-downs (reduction) of that amount are raised at more favorable in interest payments. The credit enhancement terms from private sources. and other features of the transaction brought Key steps involve development of key the overall funding cost of the financing to performance indicators (KPIs) that reflect 7 percent, which was approximately half the positive AQ outcome that is measurable, normal financing cost. verifiable, and appealing to potential donors. Another example is the Uruguay Sovereign The next step is to explore opportunities on Sustainability-Linked Bond, issued in March emerging environmental markets to trade 2023, which raised US$700 million by linking verifiable environmental benefits collateral investor returns to Uruguay’s performance to AQ improvements, reflected in KPI (GHG on reducing GHG emissions intensity and emission reduction, plastic pollution reduction, maintaining native forest area. 39 In this case, biodiversity preservation, and so on). The final interest rate reductions were designed to kick in step is to explore different designs of innovative when CH4 emissions were reduced in line with financial instruments and marketing them with Uruguay’s NDC. An ‘air quality’ version of such potential donors, along with consultations financing can be structured for IDA countries with governmental agencies and financial in CA, using performance indicators such institutions. 39 https://deuda.mef.gub.uy/innovaportal/file/30968/1/sslb_annual_report.pdf. 40 For instance, the Breath Better Bond Initiative is currently under development by the International Finance Corporation (IFC) and aims to pilot the initiative in one to three developing country cities. Ultimately, the initiative aims to mobilize an estimated US$4 billion in sustainable infrastructure investment to reduce the health impacts of air pollution and mitigate GHG emissions. https://www.climatefinancelab.org/ideas/the-breathe-better-bond-initiative/. Air Quality Management in Central Asia BACK TO CONTENTS 57 5. Conclusion: The Way Forward This report aimed to provide a succinct 50 percent and 80 percent of total exposure in assessment of the key AQ-related issues for CA. cities—and all excess exposure—is caused by The report first delved into analytical work to anthropogenic emissions, which can be controlled establish the key contributors to PM 2.5 population through dedicated policy interventions. exposure in CA countries and in selected cities in Although cities emerge as the major pollution hot the region. The identification of key air pollution spots in CA, only a limited share of PM 2.5 exposure sources then allowed for the assessment of within the city boundaries originates from local potential measures that would improve AQ at the emission sources. In most cities, more than 80 city level. The report then described key features percent of the local contribution originates from of the institutional setup for AQM in CA. Common solid fuel combustion in the residential sector, issues and potential solutions to strengthen the road transport, and industrial sources within AQM system in CA were identified. The report and around the cities. Between 10 percent and also considered a crucial challenge in AQM— 50 percent of the PM 2.5 in ambient air in CA financing. The report outlined the general caused by anthropogenic sources consists of types of financing for AQ improvement and secondary PM 2.5 . formed from the precursor provided recommendations for enabling AQ emissions SO 2 (mainly from coal combustion), financing, including through innovative financing NO x (predominantly from mobile sources), and mechanisms. NH 3 (from livestock manure management and Topographic, meteorological, and socioeconom- fertilizer application) in large upwind areas ic conditions in CA are very different from those (airsheds). in other countries and regions in the world, in This implies that cities in CA can improve their terms of emission densities, the relative impor- AQ only to a limited extent on their own. Efforts tance of the various source sectors that con- to achieve international AQ standards need to tribute to PM 2.5 population exposure, and the involve measures in surrounding regions and relevance of natural and transboundary sources. sometimes even in other countries. This requires Owing to the low population density and high ur- a completely new governance approach to AQM, banization rates in CA, poor AQ in CA is predom- geared toward cooperation between the different inantly a problem in urban areas, where PM 2.5 jurisdictions within the countries and the region concentrations are typically 6 to 12 times above and a clear distribution of responsibilities across the WHO guideline of 5 µg/m³. In major cities (for different governance levels, that is, from the example, Dushanbe, Tashkent, and Almaty), con- national level to provinces and individual cities. centrations surpass even the highest WHO inter- Such a new governance approach requires im target of 35 µg/m³ by up to a factor of two. substantial strengthening of the current In CA, soil and desert dust storms cause episodes AQM practices. Governmental roles and with extremely high PM 2.5 concentrations, responsibilities for AQM could be better clarified originating from the large deserts in the west. and mechanisms for inter- and intra-institutional On an annual basis these episodes account for coordination could be established. There is an 20–50 percent of total exposure in the cities. urgent need to update AQM-related legislation At the same time, this also means that between and practices—namely, updating AQ standards 58 BACK TO CONTENTS Air Quality Management in Central Asia in line with the latest scientific evidence and and employing customized financial mechanisms improving methodologies and capacities for that address specific pollution sources to attract structuring emission inventories. The technical private investment and manage risks. Key steps infrastructure for AQM such as AQ monitoring to design innovative financial instruments are stations, laboratories, and equipment also needs based on measurable results and exploration of to be upgraded and strengthened. In addition, environmental markets. the significant share of transboundary air There have been concerted efforts by CA pollution in the region calls for improved regional countries to address the growing challenge of cooperation, coordination, and resource sharing. air pollution. However, achieving sustainable Achieving lasting improvements in AQ requires improvement in AQ requires actions on all the investments in technical infrastructure but more items discussed in this report—from technical importantly in emission reduction measures capabilities for AQ assessment through that also contribute to energy security of the institutional strengthening and innovative CA countries. Therefore, a strategic approach approaches to financing AQ improvement. to financing that combines policy alignment, Given the scale of the challenge and the shared tailored financial tools, and private sector experience of CA countries, it is prudent for engagement is needed. Essential components countries to join resources. This collaboration of sustainable financing for AQ improvement would lead to, among others, accelerated take- include integrating AQM investments with up of best AQM practices and increased funding climate action to allow access to broader funding opportunities for AQ improvement from public, opportunities, creating supportive environments private, and DPs’ sources due to potential through policy reforms and robust institutions, economies of scale. Air Quality Management in Central Asia BACK TO CONTENTS 59 References Akramkhanov A., S. Strohmeier, Y.A. Yigezu, M. Haddad, T. Smeets, G. Sterk, C. Zucca, A. Zakhadullaev, P. Agostini, E.S. Golub, N. Akhmedkhodjaeva, C.S. Erencin, 2021. The Value of Landscape Restoration in Uzbekistan to Reduce Sand and Dust Storms from the Aral Seabed. © World Bank https://documents1.worldbank.org/curated/en/ 750031635227796665/pdf/ The-Value-of- Landscape-Restoration-in-Uzbekistan-to-Reduce-Sand-and-Dust-Storms-from-the-Aral- Seabed.pdf Alter, N.C., Whitman, E.M., Bellinger, D.C. et al. Quantifying the association between PM 2.5 air pollution and IQ loss in children: a systematic review and meta-analysis . Environ Health 23, 101 (2024). https://doi.org/10.1186/s12940-024-01122-x Almeida, S. M., M. Manousakas, E. Diapouli, Z. Kertesz, L. Samek, E. Hristova, K. Šega, R. Padilla Alvarez, C. A. Belis, and K. Eleftheriadis. 2020. “Ambient Particulate Matter Source Apportionment Using Receptor Modelling in European and Central Asia Urban Areas.” Environmental Pollution 266 (November): 115199. https://doi.org/10.1016/j.envpol.2020.115199. Amann, Markus, Imrich Bertok, Jens Borken-Kleefeld, Janusz Cofala, Chris Heyes, Lena Höglund- Isaksson, Zbigniew Klimont, et al. 2011. “Cost-Effective Control of Air Quality and Greenhouse Gases in Europe: Modeling and Policy Applications.” Environmental Modelling & Software 26 (12): 1489–1501. https://doi.org/10.1016/j.envsoft.2011.07.012. Burns, Andrew, Charl Campagne, and David Jooste. 2019. The World Bank Macro-Fiscal Model Technical Description. Washington, DC: World Bank. https://documents1.worldbank.org/curated/ en/294311565103938951/pdf/The-World-Bank-Macro-Fiscal-Model-Technical-Description.pdf. Clean Air Fund. https://www.cleanairfund.org/theme/economics/ De Bruyn, Sander. 2020. Health Costs of Air Pollution in European Cities and the Linkage with Transport. 20.190272.134. Delft, NL: CE Delft. https://epha.org/wp-content/uploads/2020/10/ final-health-costs-of-air-pollution-in-european-cities-and-the-linkage-with-transport.pdf. FAO (Food and Agriculture Organization). 2023. World Food and Agriculture - Statistical Yearbook 2023. Rome, Italy: Food and Agriculture Organization (FAO). https://doi.org/10.4060/cc8166en. Ge, Yao, Mathew R. Heal, David S. Stevenson, Peter Wind, and Massimo Vieno. 2021. “Evaluation of Global EMEP MSC-W (Rv4.34) WRF (v3.9.1.1) Model Surface Concentrations and Wet Deposition of Reactive N and S with Measurements.” Geoscientific Model Development 14 (11): 7021–46. https://doi.org/10.5194/gmd-14-7021-2021. IEA (International Energy Agency). 2021. World Energy Outlook 2021 . Paris: International Energy Agency. IEA. 2022. Tajkistan Energy Sector Review 2022 . Paris: International Energy Agency. IEA. 2023. World Energy Statistics and Balances . Paris: International Energy Agency. https://doi.org/10.1787/enestats-data-en. IHME, GBD Global Burden of Disease Collaborative Network. 2019. Global Burden of Disease Study 2019 (GBD 2019). Seattle, United States: Institute for Health Metrics and Evaluation (IHME). https://vizhub.healthdata.org/gbd-results/. 60 BACK TO CONTENTS Air Quality Management in Central Asia Kiesewetter, G., J. Borken-Kleefeld, W. Schöpp, C. Heyes, P. Thunis, B. Bessagnet, E. Terrenoire, H. Fagerli, A. Nyiri, and M. Amann. 2015. “Modelling Street Level PM 10 Concentrations across Europe: Source Apportionment and Possible Futures.” Atmos. Chem. Phys. 15 (3): 1539–53. https://doi.org/10.5194/acp-15-1539-2015. Lancet commission on Pollution and Health. https://www.thelancet.com/commissions/pollution-and-health McDuffie, Erin E., Randall V. Martin, Joseph V. Spadaro, Richard Burnett, Steven J. Smith, Patrick O’Rourke, Melanie S. Hammer, et al. 2021. “Source Sector and Fuel Contributions to Ambient PM 2.5 and Attributable Mortality across Multiple Spatial Scales.” Nature Communications 12 (1): 3594. https://doi.org/10.1038/s41467-021-23853-y. Middleton, N. J. 2017. “Desert Dust Hazards: A Global Review.” Aeolian Research 24 (February): 53–63. https://doi.org/10.1016/j.aeolia.2016.12.001. Neidell, M., Pestel, N. Air pollution and worker productivity. IZA World of Labor 2023: 363 doi: 10.15185/izawol.363.v2 Nobakht, Mohamad, Maria Shahgedanova, and Kevin White. 2021. “New Inventory of Dust Emission Sources in Central Asia and Northwestern China Derived From MODIS Imagery Using Dust Enhancement Technique.” Journal of Geophysical Research: Atmospheres 126 (4): e2020JD033382. https://doi.org/10.1029/2020JD033382. OECD (Organisation for Economic Co-operation and Development). 2019. “Addressing Industrial Air Pollution in Kazakhstan: Reforming Environmental Payments Policy Guidelines.” OECD Green Growth Studies, OECD Publishing, Paris, https://doi.org/10.1787/0e04ea86-en Shen, S., C. Li, A. van Donkelaar, N. Jacobs, C. Wang, and R. V. Martin, R. V. 2024. “Enhancing Global Estimation of Fine Particulate Matter Concentrations by Including Geophysical a Priori Information in Deep Learning”. https://pubs.acs.org/doi/full/10.1021/acsestair.3c00054 Simpson, D., A. Benedictow, H. Berge, R. Bergström, L. D. Emberson, H. Fagerli, C. R. Flechard, G. D. Hayman, M. Gauss, and J. E. Jonson. 2012. “The EMEP MSC-W Chemical Transport Model–Technical Description.” Atmos. Chem. Phys 12:7825–65. https://doi.org/10.5194/acp-12-7825-2012. Strinati, Costanza, Sean Stout, Jake Connolly, Léa Faucheux, and George Tsitati. 2023. “The State of Global Air Quality Funding.” Climate Policy Initiative. UNEP (United Nations Environment Programme). 2017. Frontiers 2017: Emerging Issues of Environmental Concern. Chapter 4: Sand and Dust Storms: Subduing a Global Phenomenon. Nairobi: United Nations Environment Programme. https://catalogue.unccd.int/765_Frontiers_2017_CH4_EN.pdf UNICEF (United Nations Children’s Fund). 2023. Health and Social Impacts of Air Pollution on Women and Children in Bishkek, Kyrgyz Republic . UNICEF. WHO (World Health Organization). 2021. WHO Global Air Quality Guidelines.Geneva, Switzerland: WHO. World Bank. 2019. “Pollution and City Competitiveness: A Descriptive Analysis.” https://hdl.handle.net/10986/31278. World Bank. 2020a. Fueling Kyrgyz Republic’s Transition to Clean Household Heating Solutions. Washington, DC: World Bank. World Bank. 2020b. Clearing the Air: A Tale of Three Cities Air Quality Management in Central Asia BACK TO CONTENTS 61 World Bank. 2020c. China - Hebei Air Pollution Prevention and Control Project. Implementation Completion Report. https://www.worldbank.org/en/results/2020/05/07/breathing-easier- supporting-chinas-ambitious-air-pollution-control-targets. World Bank. 2021. Clean Air and Cool Planet: Cost-Effective Air Quality Management in Kazakhstan and Its Impact on Greenhouse Gas Emissions. Washington, DC: World Bank. http://hdl.handle.net/10986/36791. World Bank. 2022a. Clean Air and Cool Planet - Volume II. Integrated Air Quality Management and Greenhouse Gas Reduction for Almaty and Nur-Sultan. Washington, DC: World Bank. World Bank. 2022b. The Global Health Cost of PM 2.5 Air Pollution: A Case for Action Beyond 2021. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1816-5. World Bank. 2022c. Kazakhstan Country Climate and Development Report. CCDR Series. Washington, DC: World Bank. https://openknowledge.worldbank.org/entities/publication/dabff214-772e- 50b4-89d9-a172e99accc3 World Bank. 2023a. Air Quality Analysis for Bishkek: PM 2.5 Source Apportionment and Emission Reduction Measures. Washington, DC: World Bank. World Bank. 2023b. Air Quality Assessment for Tashkent. Washington, DC: World Bank. World Bank. 2023c. Uzbekistan Country Climate and Development Report. CCDR Series. Washington, DC: World Bank. https://openknowledge.worldbank.org/entities/publication/686d0297-b311- 4296-ab33-51d1859c6b32 World Bank 2023d. Air Quality Management in Tajikistan. Washington, DC: World Bank. https: //documents1.worldbank.org/curated/en/099081723150539839/pdf/P18001404c- 70c408509f4e0291eb72b882e.pdf World Bank. 2024a. Report on Heating Demand and Supply, and Options to Reduce Demand - Tajikistan. Washington, DC: World Bank. World Bank. 2024b.  Tajikistan Country Climate and Development Report.  CCDR Series. Washington, DC: World Bank. https://openknowledge.worldbank.org/entities/publication/cc01a571-2254- 4a07-acdd-3bcff0021e23 World Bank. 2025a. Accelerating Access to Clean Air. Washington, DC: World Bank. World Bank. 2025b. Air Quality Management in EU Member States – Governance and Institutional Arrangements: International Experience and Implications . Washington, DC: World Bank. 62 BACK TO CONTENTS Air Quality Management in Central Asia Annex I: The GAINS Model Tool To provide airshed management with a good The GAINS model quantifies the spatial distribu- understanding of where pollution is currently tion of observed pollution concentrations in am- coming from and how pollution can be bient air by bringing together information on the reduced most effectively in the future, the sources of emissions and their socioeconomic analysis employed a well-established model drivers with advanced modeling of atmospher- framework that has enabled decision-makers ic chemistry and transport of pollution (the blue throughout the world to achieve substantial AQ boxes in Figure A1). Based on given projections improvements. Developed by the International of future economic, energy, and agricultural de- Institute for Applied Systems Analysis (IIASA), velopment, GAINS then determines the future the GAINS model (Amann et al. 2011) quantifies improvements in AQ and population exposure for the pathways of atmospheric pollution from each source, offered by about 1,100 proven emis- the socioeconomic driving forces to the most sion control options and the associated costs to relevant health and environmental impacts the overall economy. To inform decision-making (Figure A1). Building on robust scientific about the cost-effectiveness of alternative policy understanding and quality-controlled local intervention options, GAINS explores cooperative data that have been elaborated with strong multi-sectoral portfolios of measures that achieve involvement of local stakeholders, GAINS given AQ and/or climate policy targets at least analyses have informed decision-makers and cost to the economy (the red boxes in Figure A1). stakeholders, among others, in China, South A host of local statistics, measurement data, and Africa, Viet Nam, the EU, and the parties of the policy documents are used to compile the input CLRTAP in Eastern and Western Europe and data to the GAINS model that enable a reliable lo- North America. calized application (the orange boxes in Figure A1). Figure A1: Information flow in the GAINS model analysis National O cial Meteorological Ambient PM2.5 Population statistics Public health statistics emission inventories Data monitoring and projections statistics Socio-economic Emission Atmospheric projections generating CTM models, e.g., EMEP, CMAQ, WRF-chem, to 2030 activities AERMOD, CHIMERE, etc. National and A complete Atmospheric Emission PM2.5 Population Health international emission chemistry and factors concentrations exposure impacts and studies inventory transport model Emission National policy Cost-e ectiveness analysis Policy control documents targets measures National and international evidence Cost data The GAINS framework Source: IIASA. Note: Blue boxes indicate the analysis of the pollution chain from the sources to its impacts. Red boxes refer to the cost-effectiveness analysis, and orange boxes provide the sources of local input data used for the calculations. Air Quality Management in Central Asia BACK TO CONTENTS 63 Emission estimates The GAINS model 41 distinguishes about 400 CH 4 , nitrous oxide [N 2 O], hydrofluorocarbons emission source categories, for which it [HFCs], perfluorocarbons [PFCs], and sulfur estimates the annual emissions of primary PM 2.5 hexafluoride [SF6 ]). and the precursor emissions that generate For each source category i and year, annual secondary PM 2.5 in ambient air (that is, SO 2 , emissions of pollutant p are estimated NO x , NH 3 , and VOCs); other relevant substances considering activity levels (act ), ( uncontrolled ) such as total suspended particulate (TSP), PM 10 , emission factors (emfact ), as well as the removal black carbon, organic carbon (OC), and carbon efficiencies (eff m,p ) and application shares app m monoxide (CO); and the six Kyoto GHGs (CO 2 , of a control measure m . Equation 1 Activity rates act i (that is, the quantities of cific diurnal and seasonal time profiles as well as emission-generating activities) are derived characteristic release heights are considered for from relevant local statistics or, if unavailable, the emissions of each source category. estimated based on experience from other For the future, the evolution of emission- countries/states with comparable conditions. generating activities is extracted as external Emission factors emfact i,p are primarily input to GAINS from other scenario and forecast derived from local measurements deemed studies. Future application rates are derived representative of the specific sources in the considering existing emission control legislation region and local emission inventories to the in the region as well as the natural turnover of extent they are available. The plausibility of local capital stock. data is validated with international literature. If no local data are available, emission factors In CA, application of this approach is described from other countries/states with comparable above provided emissions estimations based on situations are applied. spatial emission densities. In total, GAINS considers about 1,100 proven While comparing emissions on a per capita basis emission control options m (but not all options is instructive for the understanding of the emis- are available to all emission sources), for which sion intensities of the economies of different the emission removal efficiencies are derived countries, actual pollution concentrations in am- from worldwide literature considering the local bient air are more closely related to the spatial conditions. Application rates app m reflect the emission densities (Figures A2 and A3). Overall, share of total activities to which a given measure CA is characterized by its very low population m is applied at a given time. density, which causes much lower spatial emis- Total emissions of a given source category in an sion intensities compared to other countries administrative unit are spatially distributed based and regions in the world (comparisons with the on statistics for large point sources and using EU are shown in Figure 3 and Figure 4). However, appropriate surrogate data for distributed sourc- there are significant exceptions for PM 2.5 in Tajik- es (for example, maps of population distribution, istan and NO x and NH 3 in Uzbekistan, which need road networks, and land use data). Region-spe- further analysis. 41 http://gains.iiasa.ac.at. 64 BACK TO CONTENTS Air Quality Management in Central Asia Figure A2: Spatial densities of primary PM 2.5 emissions in 2020, by economic sector (kg/km2) 0.4 0.3 kg/km2 0.2 0.1 0.0 Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 Power sector Residential Mobile sources Waste Agriculture Industrial processes Industrial combustion Source: Original GAINS analysis. Figure A3: Spatial densities of PM 2.5 precursor emissions in 2020, by economic sector (kg/km2) 1.6 SO2 NOx NH3 1.4 1.2 1.0 kg/km2 0.8 0.6 0.4 0.2 0.0 Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Turkmenistan EU-27 Power sector Residential Mobile sources Waste Agriculture Industrial processes Industrial combustion Source: Original GAINS analysis. The emission figures presented above have been comparisons with local emission inventories are derived from the databases of the GAINS model, hampered, among others, by a limited availability applying the methodology adopted by the LRTAP of complete national inventories. Only Kazakh- Convention to a comprehensive set of emission stan and the Kyrgyz Republic have ratified the sources based on internationally available activi- LRTAP Convention with obligations for submis- ty statistics and representative emission factors sion of complete emission inventories for the key that reflect applied emission controls. However, pollutants according to international standards. Air Quality Management in Central Asia BACK TO CONTENTS 65 PM2.5 concentrations and population exposure With the resulting emission fields of all PM 2.5 evidence indicates long-term exposure to PM 2.5 precursor emissions, PM 2.5 concentrations in as the most powerful predictor for adverse ambient air are then computed over the entire health impacts. With a focus on public health, model domain using reduced-form source- hourly results are aggregated to annual mean receptor relationships that have been derived concentrations as the most relevant metric from the EMEP atmospheric chemistry-transport associated with public health impacts. Also, model (Simpson et al. 2012). The underlying to facilitate AQM at the airshed level, ambient computations of the full EMEP model have been concentrations occurring in the target region performed at hourly time steps for the full year, are aggregated to a population exposure metric, employing the meteorological conditions of 2018 computed as a sum of the products of grid and considering the characteristic seasonal and average PM 2.5 concentrations and population in diurnal time patterns for all emission sources. the grid cell. It should, however, be noted that mean population exposure is by definition lower Annual mean concentrations of PM 2.5 in ambient than the highest concentrations measured at air are computed for primary emissions of hot spots, which are relevant for establishing PM 2.5 at a 10  km × 10  km spatial resolution, compliance with ambient AQ standards. distinguishing the release heights of the different emission sources. The chemical formation and The geophysical approach of the atmospheric atmospheric transport of secondary PM 2.5 in dispersion model employed by GAINS makes it ambient air from the emissions of the relevant possible to track the fate of emissions emerging precursor emissions (that is, SO 2 , NO x , NH 3 , and from specific sources and thereby to quantify VOCs) are modeled with a 0.5° × 0.5° longitude- their contributions to total PM 2.5 concentrations latitude resolution (Kiesewetter et al. 2015). in ambient air in the defined area. This facilitates the development of emission-inventory-based Although public attention and legislative AQM source apportionments of PM 2.5 concentrations focus on episodic concentration peaks at or exposure for specific locations or regions. pollution hot spots, worldwide epidemiological 66 BACK TO CONTENTS Air Quality Management in Central Asia Annex II: Emission Estimates for Central Asian Countries for 2020 Table A1: Emissions of primary PM 2.5 in 2020 (kilotons) Kazakhstan Kyrgyz Republic Tajikistan Uzbekistan Turkmenistan Power 51.7 0.7 0.2 3.7 0.0 Industrial combustion 15.6 0.5 0.4 1.8 2.0 Industrial processes 30.1 1.2 2.6 6.7 0.8 Residential 96.2 17.6 41.5 20.8 0.7 Mobile exhaust 12.6 1.6 1.8 9.5 9.3 Non-exhaust 0.0 0.0 0.0 0.0 0.0 Agriculture 24.0 3.0 5.1 18.3 3.4 Waste 10.9 2.7 3.0 10.9 2.7 Sum 241.1 27.3 54.6 71.7 18.9 Table A2: SO 2 emissions in 2020 (kilotons) Kazakhstan Kyrgyz Republic Tajikistan Uzbekistan Turkmenistan Power 472.1 17.5 18.8 43.5 0.0 Industrial combustion 182.2 5.1 13.1 13.8 0.6 Industrial processes 38.0 3.3 2.4 42.6 1.0 Residential 151.1 14.0 11.8 36.5 65.1 Mobile exhaust 2.5 3.2 3.5 16.0 13.3 Non-exhaust 0.0 0.0 0.0 0.0 0.0 Agriculture 1.8 0.2 0.4 1.5 0.3 Waste 0.2 0.1 0.1 0.2 0.1 Sum 847.9 43.4 50.1 154.1 80.4 Table A3: NO x emissions in 2020 (kilotons) Kazakhstan Kyrgyz Republic Tajikistan Uzbekistan Turkmenistan Power 298.0 3.7 3.6 29.3 9.4 Industrial combustion 110.1 4.1 5.1 47.0 12.2 Industrial processes 22.4 16.4 12.4 38.3 6.1 Residential 35.8 4.4 5.0 38.0 24.8 Mobile exhaust 157.2 18.3 21.6 188.2 156.6 Non-exhaust 0.0 0.0 0.0 0.0 0.0 Agriculture 42.5 9.6 11.6 73.2 32.5 Waste 1.1 0.2 0.2 0.8 0.0 Sum 667.1 56.7 59.5 414.8 241.6 Air Quality Management in Central Asia BACK TO CONTENTS 67 Table A4: NH 3 emissions in 2020 (kilotons) Kazakhstan Kyrgyz Republic Tajikistan Uzbekistan Turkmenistan Power 0.0 0.0 0.0 0.0 0.0 Industrial combustion 0.1 0.0 0.0 0.1 0.0 Industrial processes 0.2 0.0 0.8 2.8 0.5 Residential 1.7 0.5 0.5 0.5 0.1 Mobile exhaust 1.1 0.0 0.0 0.1 0.1 Non-exhaust 0.0 0.0 0.0 0.0 0.0 Agriculture 143.4 38.4 51.6 294.6 112.6 Waste 5.3 1.8 2.6 9.1 1.7 Sum 151.8 40.7 55.5 307.2 115 68 BACK TO CONTENTS Air Quality Management in Central Asia Annex III: Hourly PM2.5 Concentrations Measured in Various Locations throughout Central Asia While soil and desert dust consists primarily of storms cause episodes of extremely high PM 2.5 coarse particles (PM 10 or larger), it also includes concentrations over large areas, especially in a certain share of PM 2.5 which can be transported the summer. Multi-year measurements in CA by wind over large distances. Even 1,000–2,000 cities show hourly peak concentrations of up to km away from these source regions, dust 900 µg/m³ (Figure A4 to Figure A6). Figure A4: Hourly PM 2.5 concentrations measured at the US embassy in Dushanbe 1000 800 1000 600 µg/m 3 800 400 1000 600 µg/m 3 200 800 400 0 600 200 2020 2021 2022 Hours of the year µg/m 3 400 Source:0Original elaboration. 1000 200 A5: Hourly2020 2021 2022 Hours of the year Figure 800 PM 2.5 concentrations measured at the US embassy in Bishkek ) (ug/m 3) 0 1000 600 2020 2021 2022 Hours of the year 800 400 PM32.5 ) (ug/m 1000 600 200 800 400 2.5 0 3 PM PM 2.5 (ug/m 600 200 2020 2021 2022 Hours of the year 4000 1000 200 2020 2021 2022 Hours of the year 800 ) (ug/m 3) Original elaboration. Source:0 1000 600 2020 2021 2022 Hours of the year Figure 800 400 A6: Hourly PM 2.5 concentrations measured at the US embassy in Tashkent PM32.5 ) (ug/m 1000 600 200 800 400 0 2.5 3 PM PM 2.5 (ug/m 600 200 2020 2021 2022 Hours of the year 400 0 200 2020 2021 2022 Hours of the year 0 2020 2021 2022 Hours of the year Source: Original elaboration. Air Quality Management in Central Asia BACK TO CONTENTS 69 Annex IV: Cost-effective Air Pollution Control Measures for the Various Emission Source Sectors Table A5: Cost-effective air pollution control measures for the various emission source sectors Overlap with potential Source Cost-effective air pollution decarbonization sectors control measures measures ለ Application of high-efficiency flue gas cleaning Power technology, including high-efficiency dust removal generation (electrostatic precipitators), flue gas desulfurization, and selective catalytic reduction ለ Stringent emission standards for industrial facilities above 50 MWth using solid fuels Industrial ለ High-efficiency flue gas cleaning in industrial boilers combustion (especially for solid fuels) including high-efficiency dust removal (electrostatic precipitators), low-NO x burners, flue gas desulfurization, and selective catalytic reduction ለ Improvements in process technology, as well as more Industrial efficient capture and removal of process and fugitive processes emissions from industrial processes ለ Clean alternatives for traditional cooking with solid fuels, such as switch to LPG or piped natural gas stoves, electric induction cookers ለ Accelerated replacement of traditional solid fuel cook Household stoves with new more efficient stoves (including fan- solid fuel use assisted stoves) ለ Accelerated introduction of new heating stoves and boilers with higher combustion efficiencies and basic pollution controls ለ Switch to cleaner heating methods ለ Accelerated introduction of modern boilers with higher Other combustion efficiency and use of low-sulphur fuel residential ለ Switching to cleaner heating methods ለ Strict legislation requiring more frequent and enforced vehicle inspection and maintenance of vehicles with mandatory elimination or repair of high-emitting vehicles Heavy-duty ለ More stringent ELVs for vehicles (the further potential is vehicles estimated assuming an immediate introduction of Euro VI/6 equivalent emission standards for new vehicles) ለ Electric buses 70 BACK TO CONTENTS Air Quality Management in Central Asia Overlap with potential Source Cost-effective air pollution decarbonization sectors control measures measures ለ Accelerated shift to electric cars, vans, and two- wheelers Other ለ Strict legislation requiring more frequent and enforced transport vehicle inspection and maintenance of vehicles with (especially mandatory elimination or repair of high-emitting vehicles light-duty vehicles) and ለ More stringent ELVs for vehicles (the further potential is road dust estimated assuming an immediate introduction of Euro VI/6 equivalent emission standards for new vehicles) ለ Paving and regular cleaning of roads ለ For large industrial farms: Control of NH 3 emissions from livestock production for large industrial farms (covered storage of manure, efficient manure application on land, new animal houses built according to low-emission housing standards) Agriculture ለ Efficient application of mineral nitrogen fertilizers, including improved application of urea (proper timing and dose), and/or potentially the use of urease inhibitors, triggered through proper incentive programs ለ No measures are assumed for small subsistence farms ለ Introduction and efficient enforcement of bans on open Crop residue burning of agricultural residues burning ለ Energetic use of crop residue or conversion of crop residue to pellets ለ Policies to support circular economies through increased recycling of aluminum, steel, paper and plastics, and Municipal solid materials efficiency strategies waste ለ Solid municipal waste management reducing amounts of trash burning and introducing efficient waste collection and recycling schemes ለ Reduced routine flaring during fuel extraction and Other processing ለ Reduced fugitive dust emissions from mining industry Source: Original compilation. Note: The table provides the full portfolio of measures that are critical at the global level. Not all measures are relevant in CA, and the list does not necessarily include measures that might be critical for CA cities but less relevant at the global scale. Air Quality Management in Central Asia BACK TO CONTENTS 71 Annex V: Air Quality and Ozone Standards in Central Asia, EU, and WHO Guidelines Table A6: AQ standards in CA, EU, and WHO guidelines MAC (CA countries)/Limit values (EU)/Guidelines (WHO), in µg/m 3 Averaging period Kyrgyz EU (2024 WHO Kazakhstan Tajikistan Turkmenistan Uzbekistan Republic update)* (2021) Pollutant: PM 2.5 One-time 160 160 160 — — — — 24-hour 35 35 35 — 60 25 (a) 15 (b) 1 year — 25 25 — 35 10 5 Pollutant: PM 10 One-time 300 300 300 — 500 — — 24-hour 60 60 60 — 300 45 (a) 45 (b) 1 year — 40 40 — 50 20 15 Pollutant: NO 2 One-time 200 85 85 85 85 — — 1-hour — — — — — 200 (c) — 24-hour 40 40 40 40 60 50 (a) 25 (b) 1 year — — — — 40 20 10 Pollutant: SO 2 One-time 500 500 500 500 500 — — 1-hour — — — — — 350 (c) — 24-hour 50 50 50 50 50 50 (a) 40 (b) Pollutant: CO One-time 5,000 5,000 5,000 5,000 5,000 — — Maximum — — — — — 10,000 — 8-hour mean 24-hour 3,000 3,000 3,000 3,000 3,000 4,000 (a) 4,000 (b) Sources: Kazakhstan: https://adilet.zan.kz/rus/docs/V2200029011#z10. Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/11957?cl=ru-ru. Tajikistan: Постановление Правительства Республики Таджикистан от 26 июня 2023 года № 286 «О Государственном стандарте высшего профессионального образования для учреждений высшего профессионального образования с международным статусом в Республике Таджикистан». Turkmenistan: UNECE Turkmenistan Environmental Performance Review 2012. Uzbekistan: https://monitoring.meteo.uz/ru/menu/kriterii-kachestva-atmosfernogo-vozduha. EU: Annex I, https://data.consilium.europa.eu/doc/document/PE-88-2024-INIT/en/pdf. WHO: https://www.who.int/news-room/feature-stories/detail/what-are-the-who-air-quality-guidelines. Note: *The October 2024 update of the EU AQ limit values mandates achieving the limit values by 2030. (a) Not to be exceeded more than 18 times per year. (b) Not to be exceeded more than 4 times per year. (c) Not to be exceeded more than 3 times per year. 72 BACK TO CONTENTS Air Quality Management in Central Asia Table A7: Ozone standards in CA, EU, and WHO guidelines MAC (CA countries)/Limit values (EU)/Guidelines (WHO), in µg/m 3 Averaging period Kyrgyz EU (2024 Kazakhstan Tajikistan Turkmenistan Uzbekistan WHO Republic update)* Pollutant: O3 One-time 160 160 160 160 — — — 24-hour 30 30 30 100 — — — Maximum 8-hour daily — — — — 120 (a) 100 100 mean 1 year — — — 30 — — — Maximum 8-hour — — — — 100 (b) — — annual mean Peak — — — — — 60 60 season (c) Sources: Kazakhstan: https://adilet.zan.kz/rus/docs/V2200029011#z10. Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/11957?cl=ru-ru. Tajikistan: Постановление Правительства Республики Таджикистан от 26 июня 2023 года № 286 «О Государственном стандарте высшего профессионального образования для учреждений высшего профессионального образования с международным статусом в Республике Таджикистан». Uzbekistan: https://monitoring.meteo.uz/ru/menu/kriterii-kachestva-atmosfernogo-vozduha. EU: Annex I, https://data.consilium.europa.eu/doc/document/PE-88-2024-INIT/en/pdf. WHO: https://www.who.int/news-room/feature-stories/detail/what-are-the-who-air-quality-guidelines. Note: *The October 2024 update of the EU AQ limit values mandates achieving the limit values by 2030. (a) Not to be exceeded more than 18 times per year averaged over 3 years. (b) Not to be exceeded more than 3 times per year. (c) Average of daily maximum 8-hour mean O 3 concentration in the six consecutive months with the highest six-month running average O 3 concentration. Air Quality Management in Central Asia BACK TO CONTENTS 73 Annex VI: Financial Structure for Emissions Reduction-Linked Bond and for Sustainability- Linked Bond in Rwanda Figure A7: A World Bank financial structure for incentivizing emission reductions Source: Elaboration based on World Bank Case Study Emission Reduction-Linked Notes Mobilize Private Capital for Climate Friendly Project. 74 BACK TO CONTENTS Air Quality Management in Central Asia Figure A8: Rwanda - World Bank Sustainability-Linked Bond structure for a national Development Bank Source: Elaboration based on World Bank Case Study Rwanda Issues First-Ever Sustainability-Linked Bond (SLB) Backed by World Bank’s Innovative Financial Structure. Air Quality Management in Central Asia Summary Report April 2025 AIR QUALITY MANAGEMENT IN CENTRAL ASIA