1
|
Ma W, Yuan Z, Lau AKH, Wang L, Liao C, Zhang Y. Optimized neural network for daily-scale ozone prediction based on transfer learning. THE SCIENCE OF THE TOTAL ENVIRONMENT 2022; 827:154279. [PMID: 35248640 DOI: 10.1016/j.scitotenv.2022.154279] [Citation(s) in RCA: 5] [Impact Index Per Article: 2.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 01/03/2022] [Revised: 02/27/2022] [Accepted: 02/28/2022] [Indexed: 06/14/2023]
Abstract
Tropospheric ozone (O3) pollution is worsening in China, and an accurate forecast is a prerequisite to lower the O3 peak level. In recent years, machine learning techniques have attracted increasing attention in O3 prediction owing to their high efficiency and simple operation. However, the accuracy of predicting the daily O3 level is low. This study proposed a novel model by coupling long short-term memory neural network with transfer learning (TL-LSTM), with meteorology and pollutant concentration information as the model input. L2 regularization was applied to reduce the risk of overfitting and to improve the accuracy and generalization ability of the model prediction. Our results indicated that by transferring the knowledge in the model configuration from the hourly LSTM module, TL-LSTM greatly improves the predictability of the daily maximum 8 h average (MDA8) of O3 in Hong Kong. The coefficient of determination (R2) increased from 0.684 to 0.783 and the mean square error (MSE) reduced from 1.36 × 10-2 to 1.05 × 10-2. Furthermore, R2 and MSE were the highest in summer, indicating an under-prediction of peak O3 levels. This was a result of the limited number of high O3 days, which did not provide sufficient knowledge for the model to make an accurate prediction. Sobol analysis indicated that wind speed was the most sensitive factor in O3 prediction, largely due to the development of land-sea breeze circulation which effectively traps pollutants and expedites O3 formation. The results clearly demonstrate the effectiveness of the TL-LSTM in predicting the daily O3 concentration in Hong Kong. Thus, TL-LSTM can be promulgated into other photochemically active regions to assist in O3 pollution forecasting and management.
Collapse
Affiliation(s)
- Wei Ma
- School of Environment and Energy, South China University of Technology, Guangzhou 510006, China
| | - Zibing Yuan
- School of Environment and Energy, South China University of Technology, Guangzhou 510006, China.
| | - Alexis K H Lau
- Division for the Environment, Hong Kong University of Science and Technology, Hong Kong, China
| | - Long Wang
- Guangdong Academy of Environmental Sciences, Guangzhou 510045, China
| | - Chenghao Liao
- Guangdong Academy of Environmental Sciences, Guangzhou 510045, China
| | - Yongbo Zhang
- Guangdong Academy of Environmental Sciences, Guangzhou 510045, China
| |
Collapse
|
2
|
Fang T, Zhu Y, Wang S, Xing J, Zhao B, Fan S, Li M, Yang W, Chen Y, Huang R. Source impact and contribution analysis of ambient ozone using multi-modeling approaches over the Pearl River Delta region, China. ENVIRONMENTAL POLLUTION (BARKING, ESSEX : 1987) 2021; 289:117860. [PMID: 34332168 DOI: 10.1016/j.envpol.2021.117860] [Citation(s) in RCA: 14] [Impact Index Per Article: 4.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 06/10/2021] [Revised: 07/07/2021] [Accepted: 07/26/2021] [Indexed: 06/13/2023]
Abstract
Quantification of source impacts and contributions is a key element for the design of effective air pollution control policies. In this study, O3 source impacts and contributions were comprehensively assessed over the Pearl River Delta (PRD) region of China using brute-force method (BFM), response surface modeling with BFM (RSM-BFM) and differential method (RSM-DM) respectively, high-order decoupled direct method (HDDM), and ozone source apportionment technology (OSAT). The multi-modeling comparison results indicated that under typical nonlinear atmospheric conditions during the O3 formation, BFM, RSM-BFM, and HDDM seemed to be appropriate for assessing the impact of single source emissions; however, the results of HDDM could deviate from those of BFM when the emission reduction ratio was higher than 50 %. Under multi-source control scenarios, the results of source contribution analyses obtained from RSM-DM and OSAT were reasonably well, but the performance of OSAT was limited by its capability in representing the nonlinearity of O3 response to emission reductions of its precursors, particularly NOx. The results of this pilot study in the PRD showed that the RSM-DM appeared to replicate the nonlinearity of O3 chemistry reasonably well (e.g., O3 disbenefits due to local NOx emission reductions in Guangzhou city). Based on the source contribution results, on-road mobile (including both NOx and VOC emissions) and industrial process (mainly VOC emissions) sources were identified as two major contribution sectors by both RSM-DM and OSAT, contributing an average of 31.5 % and 11.4 % (estimated by RSM-DM) and 29.2 % and 13.0 % (estimated by OSAT) respectively to O3 formation in 9 cities of the PRD. Therefore, the reinforced emission reductions on NOx and VOC from on-road mobile and industrial process sources in the central cities (i.e., Guangzhou, Foshan, Dongguan, Shenzhen, and Zhongshan) were suggested to effectively mitigate the ambient O3 levels in the PRD.
Collapse
Affiliation(s)
- Tingting Fang
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| | - Yun Zhu
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China; Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Sun Yat-Sen University, Zhuhai, 519000, China.
| | - Shuxiao Wang
- State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, China
| | - Jia Xing
- State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, China
| | - Bin Zhao
- State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, China
| | - Shaojia Fan
- Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Sun Yat-Sen University, Zhuhai, 519000, China
| | - Minhui Li
- Guangdong Provincial Academy of Environmental Science, Guangzhou, 510006, China
| | - Wenwei Yang
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| | - Ying Chen
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| | - Ruolin Huang
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| |
Collapse
|
3
|
Fang T, Zhu Y, Jang J, Wang S, Xing J, Chiang PC, Fan S, You Z, Li J. Real-time source contribution analysis of ambient ozone using an enhanced meta-modeling approach over the Pearl River Delta Region of China. JOURNAL OF ENVIRONMENTAL MANAGEMENT 2020; 268:110650. [PMID: 32510427 DOI: 10.1016/j.jenvman.2020.110650] [Citation(s) in RCA: 16] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 02/07/2020] [Revised: 04/01/2020] [Accepted: 04/23/2020] [Indexed: 05/17/2023]
Abstract
The nonlinear response of O3 to nitrogen oxides (NOx) and volatile organic compounds (VOC) is not conducive to accurately identify the various source contributions and O3-NOx-VOC relationships. An enhanced meta-modeling approach, polynomial functions based response surface modeling coupled with the sectoral linear fitting technique (pf-ERSM-SL), integrating a new differential method (DM), was proposed to break through the limitation. The pf-ERSM-SL with DM was applied for analysis of O3 formation regime and real-time source contributions in July and October 2015 over the Pearl River Delta Region (PRD) of Mainland China. According to evaluations, the pf-ERSM-SL with DM was proven to be effective in source apportionment when the traditional sensitivity analysis was unsuitable for deriving the source contributions in the nonlinear system. After diagnosing the O3-NOx-VOC relationships, O3 formation in most regions of the PRD was identified as a distinctive NOx-limited regime in July; in October, the initial VOC-limited regime was found at small emission reductions (less than 22-44%), but it will transit to NOx-limited when further reductions were implemented. Investigation of the source contributions suggested that NOx emissions were the dominated contributor when turning-off the anthropogenic emissions, occupying 85.41-94.90% and 52.60-75.37% of the peak O3 responses in July and October respectively in the receptor regions of the PRD; NOx emissions from the on-road mobile source (NOx_ORM) in Guangzhou (GZ), Dongguan&Shenzhen (DG&SZ) and Zhongshan (ZS) were identified as the main contributors. Consequently, the reinforced control of NOx_ORM is highly recommended to lower the ambient O3 in the PRD effectively.
Collapse
Affiliation(s)
- Tingting Fang
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| | - Yun Zhu
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China; Southern Marine Science and Engineering Guangdong Laboratory, Sun Yat-Sen University, Zhuhai, 519000, China.
| | - Jicheng Jang
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| | - Shuxiao Wang
- State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, China
| | - Jia Xing
- State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, China
| | - Pen-Chi Chiang
- Graduate Institute of Environmental Engineering, National Taiwan University, Taipei, 10673, Taiwan; Carbon Cycle Research Center, National Taiwan University, 10672, Taiwan
| | - Shaojia Fan
- Southern Marine Science and Engineering Guangdong Laboratory, Sun Yat-Sen University, Zhuhai, 519000, China
| | - Zhiqiang You
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| | - Jinying Li
- Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, College of Environment and Energy, South China University of Technology, Guangzhou Higher Education Mega Center, Guangzhou, 510006, China
| |
Collapse
|
4
|
Source Contributions to Ozone Formation in the New South Wales Greater Metropolitan Region, Australia. ATMOSPHERE 2018. [DOI: 10.3390/atmos9110443] [Citation(s) in RCA: 9] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
Abstract
Ozone and fine particles (PM2.5) are the two main air pollutants of concern in the New South Wales Greater Metropolitan Region (NSW GMR) due to their contribution to poor air quality days in the region. This paper focuses on source contributions to ambient ozone concentrations for different parts of the NSW GMR, based on source emissions across the greater Sydney region. The observation-based Integrated Empirical Rate model (IER) was applied to delineate the different regions within the GMR based on the photochemical smog profile of each region. Ozone source contribution was then modelled using the CCAM-CTM (Cubic Conformal Atmospheric model-Chemical Transport model) modelling system and the latest air emission inventory for the greater Sydney region. Source contributions to ozone varied between regions, and also varied depending on the air quality metric applied (e.g., average or maximum ozone). Biogenic volatile organic compound (VOC) emissions were found to contribute significantly to median and maximum ozone concentration in North West Sydney during summer. After commercial and domestic sources, power generation was found to be the next largest anthropogenic source of maximum ozone concentrations in North West Sydney. However, in South West Sydney, beside commercial and domestic sources, on-road vehicles were predicted to be the most significant contributor to maximum ozone levels, followed by biogenic sources and power stations. The results provide information that policy makers can use to devise various options to control ozone levels in different parts of the NSW Greater Metropolitan Region.
Collapse
|
5
|
Collet S, Kidokoro T, Karamchandani P, Jung J, Shah T. Future year ozone source attribution modeling study using CMAQ-ISAM. JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION (1995) 2018; 68:1239-1247. [PMID: 29999477 DOI: 10.1080/10962247.2018.1496954] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 03/27/2018] [Revised: 06/24/2018] [Accepted: 06/29/2018] [Indexed: 06/08/2023]
Abstract
To achieve the current United States National Ambient Air Quality Standards (NAAQS) attainment level for ozone or particulate matter, current photochemical air quality models include tools to determine source apportionment and/or source sensitivity. Previous studies by the authors have used the Ozone and Particulate Matter Source Apportionment Technology and Higher-order Decoupled Direct Method probing tools in CAMx to investigate these source-receptor relationships for ozone. The recently available source apportionment for CMAQ, referred to as the Integrated Source Apportionment Method (ISAM), was used in this study to conduct future year (2030) source attribution modeling. The CMAQ-ISAM ozone source attribution results for selected cities across the U.S. showed boundary conditions were the dominant contributor to the future year highest July maximum daily 8-hour average (MDA8) ozone concentrations. Point sources were generally larger contributors in the eastern U.S. than in the western U.S. The contributions of on-road mobile emissions were around 5 ppb at most of the cities selected for analysis. Off-road mobile source contributions were around 20 ppb or nearly 30%. Since boundary conditions play an important role in future year ozone levels, it is important to characterize future year boundary conditions accurately. The current implementation of ISAM in CMAQ 5.0.2 requires significant computing resources for ozone source attribution, making it difficult to conduct long-term simulations for large domains. The computing requirements for PM source attribution are even more onerous. CMAQ 5.2 was released after this study was completed, and does not include ISAM. If an efficient version of ISAM becomes available, it could be used in long-term ozone and PM2.5 studies. Implications: Ozone source attribution results provide useful information on important emission source contribution categories and provide some initial guidance on future emission reduction strategies. This study explains a new source apportionment technique, CMAQ-ISAM, and compares it to CAMx OSAT. The techniques have similar results: ozone's highest source contributor is boundary conditions, followed by point sources, then off-road mobile sources. The current version of ISAM in CMAQ 5.0.2 requires significant computing resources for ozone source attribution, while the computing requirements for PM source attribution are even more onerous. CMAQ 5.2 was released after this study was completed, and does not include ISAM.
Collapse
Affiliation(s)
- Susan Collet
- a Product Regulatory Affairs , Toyota Motor North America , Ann Arbor , MI , USA
| | - Toru Kidokoro
- b Advanced Powertrain Management System Development Division , Toyota Motor Corporation , Shizuoka , Japan
| | | | - Jaegun Jung
- c Air Sciences Division, Ramboll US Corporation , Novato , CA , USA
| | - Tejas Shah
- c Air Sciences Division, Ramboll US Corporation , Novato , CA , USA
| |
Collapse
|
6
|
Future-Year Ozone Isopleths for South Coast, San Joaquin Valley, and Maryland. ATMOSPHERE 2018. [DOI: 10.3390/atmos9090354] [Citation(s) in RCA: 6] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
Abstract
Many areas of the United States are working toward achieving the 2015 ozone National Ambient Air Quality Standard (NAAQS) attainment level. The objective of this study was to develop future-year (2030) volatile organic compounds and nitrogen oxides (VOC-NOx) isopleth diagrams of the 4th highest maximum daily 8-h average ozone design value concentrations at monitors of interest in the South Coast Air Basin (SoCAB) and San Joaquin Valley (SJV) in California, and in Maryland. The simulation results showed there would be attainment of the 2015 ozone NAAQS in 2030 without further controls at the selected monitors: 27% in SoCAB, 57% in SJV, and 100% in Maryland. The SoCAB ozone isopleths developed in this study were compared with those reported in the South Coast Air Quality Management District 2016 Air Quality Management Plan. There are several differences between the two modeling studies, the results are qualitatively similar for most of the monitors in the relative amounts of additional emission reductions needed to achieve the ozone NAAQS. The results of this study provide insight into designing potential control strategies for ozone attainment in future years for areas currently in non-attainment. Additional photochemical modeling using these strategies can then provide confirmation of the effectiveness of the controls.
Collapse
|
7
|
Zawacki M, Baker KR, Phillips S, Davidson K, Wolfe P. Mobile Source Contributions to Ambient Ozone and Particulate Matter in 2025. ATMOSPHERIC ENVIRONMENT (OXFORD, ENGLAND : 1994) 2018; 188:129-141. [PMID: 30344445 PMCID: PMC6192431 DOI: 10.1016/j.atmosenv.2018.04.057] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/30/2023]
Abstract
The contribution of precursor emissions from 17 mobile source sectors to ambient ozone and fine particulate matter levels across the U.S. were evaluated, using the CAMx photochemical model, to identify which mobile source sectors are projected to have the largest impacts on air pollution in 2025. Both onroad and nonroad sectors contribute considerably to projected air pollution across much of the country. Summer ozone season ozone contributions between 2 and 5 ppb, which are among the highest levels presented on the maps of mobile source sectors, are largely found in the southeast United States from the onroad sectors, most notably light-duty and heavy-duty vehicles, and along the coastline from the Category 3 (C3) marine sector. Annual average PM2.5 contributions between 0.5 to 0.9 μg/m3, which are among the highest levels presented on the maps of mobile source sectors, are found throughout the Midwest and along portions of the east and west coast from onroad sectors as well as nonroad diesel and rail sectors. Additionally, contributions of precursor emissions to ambient ozone and PM2.5 levels were evaluated to understand the range of impacts from precursors in the various mobile source sectors. For most mobile source sectors, in most locations, NOX emissions contributed more to ozone than VOC emissions, and secondary PM2.5 contributed more to ambient PM2.5 than primary PM2.5. The largest ozone levels on the maps showing contributions from mobile source NOX emissions tended to be between 2 and 5 ppb, while the largest ozone levels on the maps showing contributions from mobile source VOC emissions tended to be between 0.9 and 2 ppb, except for southern California where ozone contributions from VOC emissions from onroad light duty vehicles were between 2 and 5 ppb. The largest contributions to ambient PM2.5 on the maps showing primary and secondary contributions from mobile source sectors tended to be between 0.1 and 0.5 μg/m3. The contribution from primary PM2.5 extended over localized areas (urban-scale) and the contribution from secondary PM2.5 extended over more regional (multi-state) areas.
Collapse
Affiliation(s)
- Margaret Zawacki
- US EPA, Office of Transportation and Air Quality, 2000 Traverwood Drive, Ann Arbor, MI 48105 telephone: 1-734-214-4472, fax: 1-734-214-4939
| | - Kirk R Baker
- US EPA, Office of Air Quality, Planning, and Standards, Research Triangle Park, NC 27711 (, )
| | - Sharon Phillips
- US EPA, Office of Air Quality, Planning, and Standards, Research Triangle Park, NC 27711 (, )
| | - Ken Davidson
- US EPA, Office of Transportation and Air Quality, San Francisco, CA
| | - Philip Wolfe
- ORISE participant hosted by the US EPA, Ann Arbor, MI 48105
| |
Collapse
|