Using Costs and Health Benefits to Estimate the Priority of Air Pollution Control Action Plan: A Case Study in Taiwan
Abstract
:1. Introduction
2. Research Methodology: ABaCAS-Taiwan
2.1. RSM
2.2. SMAT
2.3. BenMAP
3. Empirical Data and Technical Parameters
3.1. Cost of Pollution Control and Effectiveness of Pollution Reduction
3.2. Dispersion Simulations: Emissions Data
3.3. Health Benefits: BenMAP Parameter Data
4. Analysis of Simulation Results
4.1. Simulations on the Effectiveness of the Pollution Control Measures
4.2. Reduction in Health Risk
4.3. Health Benefits
5. Summary and Conclusions
Author Contributions
Funding
Acknowledgments
Conflicts of Interest
References
- World Health Organization. Ambient (Outdoor) Air Quality and Health. 2018. Available online: https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health (accessed on 13 August 2020).
- Fann, N.; Lamson, A.D.; Anenberg, S.C.; Wesson, K.; Risley, D.; Hubbell, B. Estimating the National Public Health Burden Associated with Exposure to Ambient PM2.5 and Ozone. Risk Anal. 2011, 32, 81–95. [Google Scholar] [CrossRef] [PubMed]
- Wang, H.; Zhu, Y.; Jang, C.; Lin, C.-J.; Wang, S.; Fu, J.S.; Gao, J.; Deng, S.; Xie, J.; Ding, D.; et al. Design and demonstration of a next-generation air quality attainment assessment system for PM2.5 and O3. J. Environ. Sci. 2015, 29, 178–188. [Google Scholar] [CrossRef] [PubMed]
- Zhu, Y.; Lao, Y.; Jang, C.; Lin, C.-J.; Xing, J.; Wang, S.; Fu, J.S.; Deng, S.; Xie, J.; Long, S. Development and case study of a science-based software platform to support policy making on air quality. J. Environ. Sci. 2015, 27, 97–107. [Google Scholar] [CrossRef] [PubMed]
- Sun, J.; Schreifels, J.; Wang, J.; Fu, J.S.; Wang, S. Cost estimate of multi-pollutant abatement from the power sector in the Yangtze River Delta region of China. Energy Policy 2014, 69, 478–488. [Google Scholar] [CrossRef]
- Wang, S.; Xing, J.; Jang, C.; Zhu, Y.; Fu, J.S.; Hao, J. Impact Assessment of Ammonia Emissions on Inorganic Aerosols in East China Using Response Surface Modeling Technique. Environ. Sci. Technol. 2011, 45, 9293–9300. [Google Scholar] [CrossRef]
- Xing, J.; Wang, S.; Jang, C.; Zhu, Y.; Hao, J.M. Nonlinear response of ozone to precursor emission changes in China: A modeling study using response surface methodology. Atmos. Chem. Phys. 2011, 11, 5027–5044. [Google Scholar] [CrossRef] [Green Version]
- Roy, R.; Braathen, N.A. The Rising Cost of Ambient Air Pollution thus Far in the 21st Century: Results from the Briics and the Oecd Countries; OECD Publishing: Paris, France, 2017. [Google Scholar]
- Bae, H.J.; Park, J. Health Benefits of Improving Air Quality in the Rapidly Aging Korean Society. Sci. Total Environ. 2009, 407, 5971–5977. [Google Scholar] [CrossRef]
- Chae, Y.; Park, J. Quantifying Costs and Benefits of Integrated Environmental Strategies of Air Quality Management and Greenhouse Gas Reduction in the Seoul Metropolitan Area. Energy Policy 2011, 39, 5296–5308. [Google Scholar] [CrossRef]
- Berman, J.D.; Fann, N.; Hollingsworth, J.W.; Pinkertion, K.E.; Rom, W.N.; Szema, A.M.; Breysse, P.N.; White, R.H.; Curriero, F.C. Health Benefits from Large-Scale Ozone Reduction in the Untied States. Environ. Health Perspect. 2012, 10, 1404–1410. [Google Scholar] [CrossRef]
- IQAir. 2019 World Air Quality Report. Available online: https://www.iqair.com/world-most-polluted-cities/world-air-quality-report-2019-en.pdf (accessed on 13 May 2020).
- South Coast Air Quality Management District. Final 2012 Air Quality Management Plan; South Coast Air Quality Management District: Diamond Bar, CA, USA, 2016. [Google Scholar]
- Xing, J.; Ding, D.; Wang, S.; Zhao, B.; Jang, C.; Wu, W.; Zhang, F.; Zhu, Y.; Hao, J. Quantification of the enhanced effectiveness of no x control from simultaneous reductions of voc and nh 3 for reducing air pollution in the beijing–tianjin–hebei region, china. Atmos. Chem. Phys. 2018, 18, 7799–7814. [Google Scholar] [CrossRef] [Green Version]
- Xing, J.; Wang, S.; Jang, C.; Zhu, Y.; Zhao, B.; Ding, D.; Wang, J.; Zhao, L.; Xie, H.; Hao, J. Abacas: An overview of the air pollution control cost-benefit and attainment assessment system and its application in China. Available online: http://pubs.awma.org/flip/EM-Apr-2017/xing.pdf (accessed on 10 July 2020).
- ABaCAS Project. Available online: http://www.abacas-dss.com/ (accessed on 10 July 2020).
- The Environmental Protection Agency. Technical Support Document for the Proposed Pm NAAQS Rule: Response Surface Modeling; U.S. Environmental Protection Agency Office of Air Quality Planning and Standards: Research Triangle Park, NC, USA, 2006.
- Taiwan Environmental Protection Administration. Air Pollution Control Strategies. 2017. Available online: https://www.ey.gov.tw/Page/5A8A0CB5B41DA11E/87903aa6-740f-49ad-b665-9db863a61045 (accessed on 24 April 2017).
- European Commission. Extern E, Externalities of Energy: Methodology 2005 Update. Available online: http://www.externe.info/externe_d7/sites/default/files/methup05a.pdf (accessed on 28 November 2005).
- United Nations Environmental Protection Agency. BenMAP, Environmental Benefits Mapping and Analysis Program Community Edition, User’s Manual. 2018. Available online: https://www.epa.gov/sites/production/files/2015-04/documents/benmap-ce_user_manual_march_2015.pdf (accessed on 15 July 2020).
- Huang, J.; Zhu, Y.; Kelly, J.T.; Jang, C.; Wang, S.; Xing, J.; Chiang, P.C.; Fan, S.; Zhao, X.; Yu, L. Large-Scale Optimization of Multi-Pollutant Control Strategies in the Pearl River Delta Region of China Using a Genetic Algorithm in Machine Learning. Sci. Total Environ. 2020, 722, 137701. [Google Scholar] [CrossRef] [PubMed]
- Davidson, K.; Hallberg, A.; McCubbin, D.; Hubbell, B. Analysis of PM2.5 Using the Environmental Benefits Mapping and Analysis Program (BenMAP). J. Toxicol. Environ. Health Part A 2007, 70, 332–346. [Google Scholar] [CrossRef] [PubMed]
- Voorhees, A.S.; Wang, J.; Wang, C.; Zhao, B.; Wang, S.; Kan, H. Public Health Benefits of Reducing Air Pollution in Shanghai: A Proof-of-Concept Methodology with Application to BenMAP. Sci. Total Environ. 2014, 485–486, 396–405. [Google Scholar] [CrossRef] [PubMed]
- Chen, L.; Shi, M.; Gao, S.; Li, S.; Mao, J.; Zhang, H.; Sun, Y.; Bai, Z.; Wang, Z. Assessment of Population Exposure to PM2.5 for Mortality in china and Its Public Health Benefit Based on BenMAP. Environ. Pollut. 2017, 221, 311–317. [Google Scholar] [CrossRef] [PubMed]
- Sacks, J.D.; Lloyd, J.M.; Zhu, Y.; Anderton, J.; Jang, C.J.; Hubbell, B.; Fann, N. The Environmental Benefits Mapping and Analysis Program-Community Edition (BenMAP–CE): A Tool to Estimate the Health and Economic Benefits of Reducing Air Pollution. Environ. Model. Softw. 2018, 104, 118–129. [Google Scholar] [CrossRef]
- Lin, W.-Y.; Hsiao, M.-C.; Wu, P.-C.; Fu, J.S.; Lai, L.-W.; Lai, H.-C. Analysis of Air Quality and Health Co-Benefits Regarding Electric Vehicle Promotion Coupled with Power Plant Emissions. J. Clean. Prod. 2020, 247, 119152. [Google Scholar] [CrossRef]
- Ashok, A.; Lee, I.H.; Arunachalam, S.; Waitz, I.A.; Yim, S.H.L.; Barrett, S.R.H. Development of a Response Surface Model of Aviation’s Air Quality Impacts in the United States. Atmos. Environ. 2013, 77, 445–452. [Google Scholar] [CrossRef]
- Long, S.; Zhu, Y.; Jang, C.; Lin, C.J.; Wang, S.; Zhao, B.; Gao, J.; Deng, S.; Xie, J.; Qiu, X. A Case Study of Development and Application of a Streamlined Control and Response Modeling System for PM2.5 Attainment Assessment in china. J. Environ. Sci. 2016, 41, 69–80. [Google Scholar] [CrossRef]
- Kelly, J.T.; Jang, C.J.; Timin, B.; Gantt, B.; Reff, A.; Zhu, Y.; Long, S.; Hanna, A. A System for Developing and Projecting PM2. 5 Spatial Fields to Correspond to Just Meeting National Ambient Air Quality Standards. Atmos. Environ. X 2019, 2, 100019. [Google Scholar] [CrossRef]
- Ostro, B.; Chestnut, L. Assessing the Health Benefits of Reducing Particulate Matter Air Pollution in the United States. Environ. Res. 1998, 76, 94–106. [Google Scholar] [CrossRef]
- Fann, N.; Baker, K.R.; Fulcher, C.M. Characterizing the PM2.5-Related Health Benefits of Emission Reductions for 17 Industrial, Area and Mobile Emission Sectors Across the U.S. Environ. Int. 2012, 49, 141–151. [Google Scholar] [CrossRef] [PubMed]
- Grabow, M.L.; Spak, S.N.; Holloway, T.; Stone, B.; Mednick, A.C.; Patz, J.A. Air Quality and Exercise-Related Health Benefits from Reduced Car Travel in the Midwestern United States. Environ. Health Perspect. 2012, 120, 68–76. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Boldo, E.; Linares, C.; Lumbreras, J.; Borge, R.; Narros, A.; García-Pérez, J.; Fernández-Navarro, P.; Pérez-Gómez, B.; Aragonés, N.; Ramis, R.; et al. Health Impact Assessment of a Reduction in Ambient PM2.5 Levels in Spain. Environ. Int. 2011, 37, 342–348. [Google Scholar] [CrossRef] [PubMed]
- Broome, R.A.; Fann, N.; Cristina, T.J.; Fulcher, C.; Duc, H.; Morgan, G.G. The Health Benefits of Reducing Air Pollution in Sydney, Australia. Environ. Res. 2015, 143, 19–25. [Google Scholar] [CrossRef]
- Liou, J.L. Estimating the Health Benefit of Particulate Matter Reduction in Taiwan: Application of Benefit per Ton Method; Chung-Hua Institution for Economic Research Publishing: Taipei, Taiwan, 2019. [Google Scholar]
- Environmental Protection Administration of Taiwan. Air Pollution Control Action Plan. Available online: https://www.ey.gov.tw/Page/448DE008087A1971/5638596f-c460-4a12-9e62-d623d34f67d1 (accessed on 21 December 2017).
- Environmental Protection Administration of Taiwan. Taiwan Emission Data System. Available online: https://teds.epa.gov.tw/TEDS.aspx.2017 (accessed on 1 November 2016).
- Pope, C.A., III; Burnett, R.T.; Thun, M.J.; Calle, E.E.; Krewski, D.; Ito, K.; Thurston, G.D. Lung Cancer, Cardiopulmonary Mortality, and Long-Term Exposure to Fine Particulate Air Pollution. JAMA 2002, 287, 1132–1141. [Google Scholar] [CrossRef] [Green Version]
- Crouse, D.L.; Peters, P.A.; Van Donkelaar, A.; Goldberg, M.S.; Villeneuve, P.J.; Brion, O.; Khan, S.; Atari, D.O.; Jerrett, M.; Pope, C.A., III. Risk of Nonaccidental and Cardiovascular Mortality in Relation to Long-Term Exposure to Low Concentrations of Fine Particulate Matter: A Canadian National-Level Cohort Study. Environ. Health Perspect. 2012, 120, 708–714. [Google Scholar] [CrossRef]
- Lepeule, J.; Laden, F.; Dockery, D.; Schwartz, J. Chronic Exposure to Fine Particles and Mortality: An Extended Follow-Up of the Harvard Six Cities Study from 1974 to 2009. Environ. Health Perspect. 2012, 120, 965–970. [Google Scholar] [CrossRef]
- Cesaroni, G.; Badaloni, C.; Gariazzo, C.; Stafoggia, M.; Sozzi, R.; Davoli, M.; Forastiere, F. Long-Term Exposure to Urban Air Pollution and Mortality in a Cohort of More than a Million Adults in Rome. Environ. Health Perspect. 2013, 121, 324–331. [Google Scholar] [CrossRef] [Green Version]
- Jerrett, M.; Burnett, R.T.; Beckerman, B.S.; Turner, M.C.; Krewski, D.; Thurston, G.; Martin, R.V.; Van Donkelaar, A.; Hughes, E.; Shi, Y.; et al. Spatial Analysis of Air Pollution and Mortality in California. Am. J. Respir. Crit. Care Med. 2013, 188, 593–599. [Google Scholar] [CrossRef]
- Beelen, R.; Raaschou-Nielsen, O.; Stafoggia, M.; Andersen, Z.J.; Weinmayr, G.; Hoffmann, B.; Wolf, K.; Samoli, E.; Fischer, P.; Nieuwenhuijsen, M. Effects of Long-Term Exposure to Air Pollution on Natural-Cause Mortality: An Analysis of 22 European Cohorts Within the Multicentre Escape Project. Lancet 2014, 383, 785–795. [Google Scholar] [CrossRef]
- Pope, C.A., III; Turner, M.C.; Burnett, R.T.; Jerrett, M.; Gapstur, S.M.; Diver, W.R.; Krewski, D.; Brook, R.D. Relationships Between Fine Particulate Air Pollution, Cardiometabolic Disorders, and Cardiovascular Mortality. Circ. Res. 2015, 116, 108–115. [Google Scholar] [CrossRef] [PubMed]
- Viscusi, W.K.; Aldy, J.E. The Value of a Statistical Life: A Critical Review of Market Estimates Throughout the World. J. Risk Uncertain 2003, 27, 5–76. [Google Scholar] [CrossRef]
- Viscusi, W.K.; Huber, J.; Bell, J. Assessing Whether There Is a Cancer Premium for the Value of a Statistical Life. Health Econ. 2014, 23, 384–396. [Google Scholar] [CrossRef] [PubMed]
- Liou, J.-L. Effect of Income Heterogeneity on Valuation of Mortality Risk in Taiwan: An Application of Unconditional Quantile Regression Method. Int. J. Environ. Res. Public Health 2019, 16, 1620. [Google Scholar] [CrossRef] [Green Version]
- Directorate-General of Budget. Accounting and Statistics, Executive Yuan, R.O.C. Price Index Database. 2020. Available online: http://statdb.dgbas.gov.tw/pxweb/Dialog/price.asp?mp=4 (accessed on 8 April 2020).
- Directorate-General of Budget. Accounting and Statistics, Executive Yuan, R.O.C. Salary and Productivity Survey Database. Available online: https://earnings.dgbas.gov.tw/query_payroll_C.aspx?mp=4 (accessed on 10 July 2017).
- Environmental Protection Administration of Taiwan, Air Quality Modeling and Simulation Standards. Available online: https://oaout.epa.gov.tw/law/LawContent.aspx?id=GL005316 (accessed on 11 August 2015).
- Zhao, B.; Wang, S.; Donahue, N.M.; Jathar, S.H.; Huang, X.; Wu, W.; Hao, J.; Robinson, A.L. Quantifying the effect of organic aerosol aging and intermediate-volatility emissions on regional-scale aerosol pollution in china. Sci. Rep. 2016, 6, 28815. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Lai, H.-C.; Ma, H.-W.; Chen, C.-R.; Hsiao, M.-C.; Pan, B.-H. Design and application of a hybrid assessment of air quality models for the source apportionment of pm2.5. Atmos. Environ. 2019, 212, 116–127. [Google Scholar] [CrossRef]
Pollution Control Measures | Technology Cost (in 100 Million NTD) | Expected Pollution Reduction (Metric Tons/Year) 1,2 | |||
---|---|---|---|---|---|
PM2.5 | SOx | NOx | VOC | ||
Regulate power facilities (implement stricter power sector regulations and standards) | 20.8 | 143 | 12,092 | 17,163 | - |
Regulate state-owned businesses (Taiwan Steel Group: install pollution control devices on emission sources and utilize optimal feasible technologies; CPC Corporation: implement tail gas recovery and replace heavy-oil fuels) | 268.0 | 174 | 1948 | 1343 | 49 |
Accelerate the decommissioning of 5000 industrial and 1000 commercial boilers) | 540.0 | 175 | 4962 | 2936 | 7 |
Improve control of smoke from 7000 restaurants | 4.1 | 788 | - | - | - |
Change fuel-burning customs and traditions (increase centralized burning to 22,000 metric tons) | 10.0 | 95 | - | 30 | - |
Regulate fugitive dust from construction sites and stockpiles of dust-generating materials | - | 672 | - | - | - |
Control smoke from the burning of agricultural waste (reduce the area of open-air burning by 90%) | 2.5 | 466 | - | - | - |
Control fugitive dust from riverbeds | - | 720 | - | - | - |
Strengthen air pollution controls in port areas (implement reductions in ship speeds, regulate ship fuel usage, promote the use of shore power) | 1863.0 | 801 | 16,385 | 2123 | - |
Retire 80,000 Stage 1 and Stage 2 diesel trucks | 40.9 | 5395 | - | 71,149 | 7584 |
Install exhaust filters in 38,000 Stage 3 diesel trucks | 193.0 | 243 | - | - | - |
Eliminate 1 million 2-stroke motorcycles | - | 457 | 1 | 260 | 7743 |
Promote electric vehicles (up to 2100 vehicles) for the transportation of fresh produce | 10.9 | 7 | - | 34 | 9 |
Strengthen emission standards for automobiles that are 10 years or older, and set up air quality maintenance zones, where the entry of highly polluting vehicles is restricted or forbidden | - | 410 | - | 2587 | 5315 |
Type of Database (Base Year = 2013, Target Year = 2019) | Data Description |
---|---|
Pollutant Monitoring Data | PM2.5 monitoring data across Taiwan in 2013: 24.1 μg/m3 (seasonal average) 1 |
Death Rate | All causes (2013): 0.0165% 2,3 Cardiovascular diseases (2013): 0.0532% 2,4 Respiratory diseases (2013): 0.0026% 2,5 |
Population Data | Import of actual population in the all-age minimum statistical area in 2013: 22,306,759 6 |
Health Impact Function (Cr-Function) | Meta-analysis was applied to health impact function using a random effect approach. 7 |
VSL (Benefit Function) | Salary income—VSL elasticity () 8: 0.2476 Consumer Price Index 9: CPI2014: 98.93; CPI2019: 102.55 Recurring income 10: W2014: 40,189 NTD/month; W2019: 42,851 NTD/month . 8: 357.9 million NTD 11: 364.5 million NTD |
Air pollution Control Measure | Average ∆PM2.5 (μg/m3) 1 | Expected Reduction in Number of Deaths (90% Confidence) 2 |
---|---|---|
Regulate power facilities (stricter power sector regulations and standards) | 0.405 (0.03–1.13) | 827 (213–2022) |
Install pollution control devices in state-owned businesses (e.g., Dragon Steel, China Steel Corporation, CPC Corporation) | 0.050 (0–0.44) | 82 (21–199) |
Decommissioning of boilers (accelerate the retirement of 5000 industrial boilers and 1000 commercial boilers | 0.061 (0.01–0.19) | 114 (29–280) |
Control smoke from 7000 restaurants | 0.114 (0.01–0.21) | 206 (53–504) |
Change fuel-burning customs and traditions (increase centralized burning to 22,000 metric tons) | 0.016 (0–0.03) | 27 (7–67) |
Regulate fugitive dust from construction sites and stockpiles of dust-generating materials, increase conformance to 90% | 0.098 (0.01–0.18) | 176 (45–430) |
Improve control smoke from agricultural waste-burning (reduce the area of open-air burning by 90%) | 0.069 (0–0.12) | 123 (31–300) |
Control fugitive dust from riverbeds | 0.105 (0.01–0.19) | 188 (48–460) |
Retire 80,000 Stage 1 and Stage 2 diesel trucks | 1.959 (0.12–4.18) | 3846 (994–9415) |
Install exhaust filters in 38,000 Stage 3 diesel trucks | 0.046 (0.01–0.08) | 87 (22–213) |
Eliminate 1 million 2-stroke motorcycles | 0.088 (0.01–0.16) | 169 (43–413) |
Promote electric vehicles (up to 2100 vehicles) for the transportation of fresh produce | 0.003 (0–0.01) | 5 (1–13) |
Strengthen air pollution controls in port areas (reduce ship speeds, regulate ship fuel usage, promote the use of shore power) | 0.529 (0.04–1.48) | 994 (255–2430) |
Set stricter emission standards for automobiles that are 10 years or older, and set up air quality maintenance zones, where the entry of highly polluting vehicles is restricted or forbidden | 0.107 (0.01–0.21) | 208 (53–508) |
Air Pollution Control Measure | Health Benefit 1 (100 Million NTD) | Percentage of Total Benefit (%) | Technical Cost (100 Million NTD) | Benefit/Cost Ratio | |
---|---|---|---|---|---|
1. Technical Measures | |||||
T1 | Retire 80,000 Stage 1 and Stage 2 diesel trucks | 14,019 | 54.5 | 1863 | 7.5 |
T2 | Strengthen air pollution controls in port areas (ship speed reductions, regulations on ship fuel usage, promoting the use of shore power) | 3623 | 14.1 | 11 | 332.4 |
T3 | Power facility regulations (stricter power sector regulations and standards) | 3014 | 11.7 | 21 | 144.9 |
T4 | Improve control of smoke from 7000 restaurants | 751 | 2.9 | 4 | 183.1 |
T5 | Eliminate 1 million 2-stroke motorcycles | 616 | 2.4 | 193 | 3.2 |
T6 | Improve control of smoke from agricultural waste-burning (reduce the area of open-air burning by 90%) | 448 | 1.7 | 3 | 179.3 |
T7 | Regulate boilers (accelerate the retirement of 5000 industrial boilers and 1000 commercial boilers) | 416 | 1.6 | 540 | 0.8 |
T8 | Install exhaust filters in 38,000 Stage 3 diesel trucks | 317 | 1.2 | 41 | 7.8 |
T9 | Install pollution control devices in state-owned businesses (e.g., Dragon Steel, China Steel Corporation, CPC Corporation) | 299 | 1.2 | 268 | 1.1 |
T10 | Change fuel-burning customs and traditions (increase centralized burning to 22,000 metric tons) | 98 | 0.4 | 10 | 9.8 |
Subtotal | 23,601 | 91.7 | 2954 | 86.9 | |
2. Administrative Measures 1 | |||||
M1 | Tighten emission standards for automobiles that are 10 years or older, and set up air quality maintenance zones, where the entry of highly polluting vehicles is restricted or forbidden | 758 | 2.9 | - | - |
M2 | Control fugitive dust from riverbeds | 685 | 2.7 | - | - |
M3 | Regulations for fugitive dust from construction sites and stockpiles of dust-generating materials, increase conformance to 90% | 642 | 2.5 | - | - |
M4 | Promotion of electric vehicles (up to 2100 vehicles) for the transportation of fresh produce | 18 | 0.1 | - | - |
Subtotal | 2103 | 8.2 | - | - |
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Lai, H.-C.; Hsiao, M.-C.; Liou, J.-L.; Lai, L.-W.; Wu, P.-C.; Fu, J.S. Using Costs and Health Benefits to Estimate the Priority of Air Pollution Control Action Plan: A Case Study in Taiwan. Appl. Sci. 2020, 10, 5970. https://doi.org/10.3390/app10175970
Lai H-C, Hsiao M-C, Liou J-L, Lai L-W, Wu P-C, Fu JS. Using Costs and Health Benefits to Estimate the Priority of Air Pollution Control Action Plan: A Case Study in Taiwan. Applied Sciences. 2020; 10(17):5970. https://doi.org/10.3390/app10175970
Chicago/Turabian StyleLai, Hsin-Chih, Min-Chuan Hsiao, Je-Liang Liou, Li-Wei Lai, Pei-Chih Wu, and Joshua S. Fu. 2020. "Using Costs and Health Benefits to Estimate the Priority of Air Pollution Control Action Plan: A Case Study in Taiwan" Applied Sciences 10, no. 17: 5970. https://doi.org/10.3390/app10175970
APA StyleLai, H.-C., Hsiao, M.-C., Liou, J.-L., Lai, L.-W., Wu, P.-C., & Fu, J. S. (2020). Using Costs and Health Benefits to Estimate the Priority of Air Pollution Control Action Plan: A Case Study in Taiwan. Applied Sciences, 10(17), 5970. https://doi.org/10.3390/app10175970