Advertisement

Predicting ground-level ozone concentrations by adaptive Bayesian model averaging of statistical seasonal models

  • K. M. Mok
  • K. V. Yuen
  • K. I. Hoi
  • K. M. Chao
  • D. Lopes
Original Paper
  • 226 Downloads

Abstract

While seasonal time-varying models should generally be used to predict the daily concentration of ground-level ozone given its strong seasonal cycles, the sudden switching of models according to their designated period in an annual operational forecasting system may affect their performance, especially during the season’s transitional period in which the starting date and duration time can vary from year to year. This paper studies the effectiveness of an adaptive Bayesian Model Averaging scheme with the support of a transitional prediction model in solving the problem. The scheme continuously evaluates the probabilities of all the ozone prediction models (ozone season, nonozone season, and the transitional period) in a forecasting system, which are then used to provide a weighted average forecast. The scheme has been adopted in predicting the daily maximum of 8-h averaged ozone concentration in Macau for a period of 2 years (2008 and 2009), with results proved to be satisfactory.

Keywords

Adaptive Bayesian model averaging Kalman filter Model switching Ozone prediction Statistical model 

Notes

Acknowledgements

The supports from the Science and Technology Development Fund of the Macau SAR Government under Grant No. 079/2013/A3 and the university Multi-year Research Grant MYRG2014-00038-FST of the research committee of the University of Macau are acknowledged. The Macau Meteorological and Geophysical Bureau is thanked for supplying the data.

References

  1. Abdul-Wahab SA, Al-Alawi SM (2002) Assessment and prediction of tropospheric ozone concentration levels using artificial neural networks. Environ Model Softw 17:219–228. doi: 10.1016/S1364-8152(01)00077-9 CrossRefGoogle Scholar
  2. Ajami NK, Gu C (2010) Complexity in microbial metabolic processes in soil nitrogen modeling: a case for model averaging. Stoch Environ Res Risk Assess 24:831–844. doi: 10.1007/s00477-010-0381-4 CrossRefGoogle Scholar
  3. Balashov NV, Thompson AM, Young GS (2017) Probabilistic forecasting of surface ozone with a novel statistical approach. J Appl Meteorol Climatol 56:297–316. doi: 10.1175/JAMC-D-16-0110.1 CrossRefGoogle Scholar
  4. Barrero MA, Grimalt JO, Cantón L (2006) Prediction of daily ozone concentration maxima in the urban atmosphere. Chemom Intell Lab Syst 80:67–76. doi: 10.1016/j.chemolab.2005.07.003 CrossRefGoogle Scholar
  5. Beck JL, Yuen KV (2004) Model selection using response measurements: Bayesian probabilistic approach. J Eng Mech 130:192–203. doi: 10.1061/(ASCE)0733-9399(2004)130:2(192) CrossRefGoogle Scholar
  6. Chao KM (2013) Development of an efficient and robust air quality prediction system for ground-level ozone in Macau. M.Sc. Thesis, University of MacauGoogle Scholar
  7. Chao KM, Hoi KI, Yuen KV, Mok KM (2012) Adaptive modelling of the daily behavior of the boundary layer ozone in Macau. ISRN Meteorol 2012:1–7. doi: 10.5402/2012/434176 CrossRefGoogle Scholar
  8. Chu HJ, Lin CY, Liau CJ, Kuo YM (2012) Identifying controlling factors of ground-level ozone levels over southwestern Taiwan using a decision tree. Atmos Environ 60:142–152. doi: 10.1016/j.atmosenv.2012.06.032 CrossRefGoogle Scholar
  9. Cobourn WG (2007) Accuracy and reliability of an automated air quality forecast system for ozone in seven Kentucky metropolitan areas. Atmos Environ 41:5863–5875. doi: 10.1016/j.atmosenv.2007.03.024 CrossRefGoogle Scholar
  10. Du X, Wu Y, Fu L, Wang S, Zhang S, Hao J (2012) Intake fraction of PM2.5 and NOx from vehicle emissions in Beijing based on personal exposure data. Atmos Environ 57:233–243. doi: 10.1016/j.atmosenv.2012.04.046 CrossRefGoogle Scholar
  11. Dueñas C, Fernández MC, Cañete S, Carretero J, Liger E (2002) Assessment of ozone variations and meteorological effects in an urban area in the Mediterranean Coast. Sci Total Environ 299:97–113. doi: 10.1016/S0048-9697(02)00251-6 CrossRefGoogle Scholar
  12. Dutot AL, Rynkiewicz J, Steiner FE, Rude J (2007) A 24-h forecast of ozone peaks and exceedance levels using neural classifiers and weather predictions. Environ Model Softw 22:1261–1269. doi: 10.1016/j.envsoft.2006.08.002 CrossRefGoogle Scholar
  13. He HD, Lu WZ (2012) Decomposition of pollution contributors to urban ozone levels concerning regional and local scales. Build Environ 49:97–103. doi: 10.1016/j.buildenv.2011.09.019 CrossRefGoogle Scholar
  14. Hoeting JA, Madigan D, Raftery AE, Volinsky CT (1999) Bayesian model averaging: a tutorial. Stat Sci 14:382–401. doi: 10.1214/ss/1009212519 CrossRefGoogle Scholar
  15. Hoi KI, Yuen KV, Mok KM (2009) Prediction of daily averaged PM10 concentrations by statistical time-varying model. Atmos Environ 43:2579–2581. doi: 10.1016/j.atmosenv.2009.02.020 CrossRefGoogle Scholar
  16. Hoi KI, Yuen KV, Mok KM (2013a) Bayesian model class selection of daily ground-level ozone prediction model. In: Proceedings of the 4th international conference on environmental management, engineering, planning and economics (CEMEPE) and SECOTOX Conference, June 24–28, Mykonos Island, Greece, pp 425–431Google Scholar
  17. Hoi KI, Yuen KV, Mok KM (2013b) Improvement of the multilayer perceptron for air quality modelling through an adaptive learning scheme. Comput Geosci 59:148–155. doi: 10.1016/j.cageo.2013.06.002 CrossRefGoogle Scholar
  18. Hsieh NH, Cheng YH, Liao CM (2014) Changing variance and skewness as leading indicators for detecting ozone exposure-associated lung function decrement. Stoch Environ Res Risk Assess 28:2205–2216. doi: 10.1007/s00477-014-0887-2 CrossRefGoogle Scholar
  19. Kalman RE (1960) A new approach to linear filtering and prediction problems. J Basic Eng 82:35–45. doi: 10.1115/1.3662552 CrossRefGoogle Scholar
  20. Kalman RE, Bucy RS (1961) New results in linear filtering and prediction theory. J Basic Eng 83:95–108. doi: 10.1115/1.3658902 CrossRefGoogle Scholar
  21. Khaniabadi YO, Hopke PK, Goudarzi G, Daryanoosh SM, Jourvand M, Basiri H (2017) Cardiopulmonary mortality and COPD attributed to ambient ozone. Environ Res 152:336–341. doi: 10.1016/j.envres.2016.10.008 CrossRefGoogle Scholar
  22. Khatibi R, Naghipour L, Ghorbani MA, Smith MS, Karimi V, Farhoudi R, Delafrouz H, Arvanaghi H (2013) Developing a predictive tropospheric ozone model for Tabriz. Atmos Environ 68:286–294. doi: 10.1016/j.atmosenv.2012.11.020 CrossRefGoogle Scholar
  23. Kim SE (2010) Tree-based threshold modeling for short-term forecast of daily maximum ozone level. Stoch Environ Res Risk Assess 24:19–28. doi: 10.1007/s00477-008-0295-6 CrossRefGoogle Scholar
  24. Kovač-Andrić E, Brana J, Gvozdić V (2009) Impact of meteorological factors on ozone concentrations modelled by time series analysis and multivariate statistical methods. Ecol Inform 4:117–122. doi: 10.1016/j.ecoinf.2009.01.002 CrossRefGoogle Scholar
  25. Lam KS, Wang TJ, Wu CL, Li YS (2005) Study on an ozone episode in hot season in Hong Kong and transboundary air pollution over Pearl River Delta region of China. Atmos Environ 39:1967–1977. doi: 10.1016/j.atmosenv.2004.11.023 CrossRefGoogle Scholar
  26. Li J, Yang W, Wang Z, Chen H, Hu B, Li J, Sun Y, Fu P, Zhang Y (2015) Modeling study of surface ozone source-receptor relationships in East Asia. Atmos Res 167:77–88. doi: 10.1016/j.atmosres.2015.07.010 CrossRefGoogle Scholar
  27. Lin Y, Cobourn WG (2007) Fuzzy system models combined with nonlinear regression for daily ground-level ozone predictions. Atmos Environ 41:3502–3513. doi: 10.1016/j.atmosenv.2006.11.060 CrossRefGoogle Scholar
  28. Lu WZ, Wang D (2014) Learning machines: rationale and application in ground-level ozone prediction. Appl Soft Comput 24:135–141. doi: 10.1016/j.asoc.2014.07.008 CrossRefGoogle Scholar
  29. Mahapatra A (2010) Prediction of daily ground-level ozone concentration maxima over New Delhi. Environ Monit Assess 170:159–170. doi: 10.1007/s10661-009-1223-z CrossRefGoogle Scholar
  30. Mansfield ML, Hall CF (2013) Statistical analysis of winter ozone events. Air Qual Atmos Heal 6:687–699. doi: 10.1007/s11869-013-0204-0 CrossRefGoogle Scholar
  31. Mikkonen S, Korhonen H, Romakkaniemi S, Smith JN, Joutsensaari J, Lehtinen KEJ, Hamed A, Breider TJ, Birmili W, Spindler G, Plass-Duelmer C, Facchini MC, Laaksonen A (2011) Meteorological and trace gas factors affecting the number concentration of atmospheric Aitken (D p = 50 nm) particles in the continental boundary layer: parameterization using a multivariate mixed effects model. Geosci Model Dev 4:1–13. doi: 10.5194/gmd-4-1-2011 CrossRefGoogle Scholar
  32. Mikkonen S, Laine M, Mäkelä HM, Gregow H, Tuomenvirta H, Lahtinen M, Laaksonen A (2015) Trends in the average temperature in Finland, 1847–2013. Stoch Environ Res Risk Assess 29:1521–1529. doi: 10.1007/s00477-014-0992-2 CrossRefGoogle Scholar
  33. Morales-Casique E, Neuman SP, Vesselinov VV (2010) Maximum likelihood Bayesian averaging of airflow models in unsaturated fractured tuff using Occam and variance windows. Stoch Environ Res Risk Assess 24:863–880. doi: 10.1007/s00477-010-0383-2 CrossRefGoogle Scholar
  34. Özbay B, Keskin GA, Doğruparmak ŞÇ, Ayberk S (2011) Multivariate methods for ground-level ozone modeling. Atmos Res 102:57–65. doi: 10.1016/j.atmosres.2011.06.005 CrossRefGoogle Scholar
  35. Paschalidou AK, Kassomenos PA, Bartzokas A (2009) A comparative study on various statistical techniques predicting ozone concentrations: implications to environmental management. Environ Monit Assess 148:277–289. doi: 10.1007/s10661-008-0158-0 CrossRefGoogle Scholar
  36. Sahu SK, Bakar KS (2012) A comparison of Bayesian models for daily ozone concentration levels. Stat Methodol 9:144–157. doi: 10.1016/j.stamet.2011.04.009 CrossRefGoogle Scholar
  37. Schlink U, Herbarth O, Richter M, Dorling S, Nunnari G, Cawley G, Pelikan E (2006) Statistical models to assess the health effects and to forecast ground-level ozone. Environ Model Softw 21:547–558. doi: 10.1016/j.envsoft.2004.12.002 CrossRefGoogle Scholar
  38. Schwarz G (1978) Estimating the dimension of a model. Ann Stat 6:461–464. doi: 10.1214/aos/1176344136 CrossRefGoogle Scholar
  39. Sharma S, Sharma P, Khare M (2017) Photo-chemical transport modelling of tropospheric ozone: a review. Atmos Environ 159:34–54. doi: 10.1016/j.atmosenv.2017.03.047 CrossRefGoogle Scholar
  40. Shen JC, Chang CH, Wu SJ, Hsu CT, Lien HC (2015) Real-time correction of water stage forecast using combination of forecasted errors by time series models and Kalman filter method. Stoch Environ Res Risk Assess 29:1903–1920. doi: 10.1007/s00477-015-1074-9 CrossRefGoogle Scholar
  41. Sousa SIV, Martins FG, Pereira MC, Alvim-Ferraz MCM (2006) Prediction of ozone concentrations in Oporto city with statistical approaches. Chemosphere 64:1141–1149. doi: 10.1016/j.chemosphere.2005.11.051 CrossRefGoogle Scholar
  42. Sun W, Zhang H, Palazoglu A (2013) Prediction of 8 h-average ozone concentration using a supervised hidden Markov model combined with generalized linear models. Atmos Environ 81:199–208. doi: 10.1016/j.atmosenv.2013.09.014 CrossRefGoogle Scholar
  43. Symeonidis P, Ziomas I, Proyou A (2004) Development of an emission inventory system from transport in Greece. Environ Model Softw 19:413–421. doi: 10.1016/S1364-8152(03)00140-3 CrossRefGoogle Scholar
  44. Thompson ML, Reynolds J, Cox LH, Guttorp P, Sampson PD (2001) A review of statistical methods for the meteorological adjustment of tropospheric ozone. Atmos Environ 35:617–630. doi: 10.1016/S1352-2310(00)00261-2 CrossRefGoogle Scholar
  45. Vingarzan R (2004) A review of surface ozone background levels and trends. Atmos Environ 38:3431–3442. doi: 10.1016/j.atmosenv.2004.03.030 CrossRefGoogle Scholar
  46. Wang D, Lu WZ (2006) Ground-level ozone prediction using multilayer perceptron trained with an innovative hybrid approach. Ecol Model 198:332–340. doi: 10.1016/j.ecolmodel.2006.05.031 CrossRefGoogle Scholar
  47. World Health Organization (2013) Review of evidence on health aspects of air pollution—REVIHAAP: final technical report. Geneva, Switzerland: World Health Organization. http://www.euro.who.int/__data/assets/pdf_file/0004/193108/REVIHAAP-Final-technical-report-final-version.pdf?ua=1. Accessed 25 Jan 2017
  48. Xia L, Shao Y (2005) Modelling of traffic flow and air pollution emission with application to Hong Kong Island. Environ Model Softw 20:1175–1188. doi: 10.1016/j.envsoft.2004.08.003 CrossRefGoogle Scholar
  49. Yuen KV (2010) Bayesian methods for structural dynamics and civil engineering. Wiley, ChichesterCrossRefGoogle Scholar
  50. Yuen KV, Liang PF, Kuok SC (2013) Online estimation of noise parameters for Kalman filter. Struct Eng Mech 47:361–381. doi: 10.12989/sem.2013.47.3.361 CrossRefGoogle Scholar
  51. Zheng J, Swall JL, Cox WM, Davis JM (2007) Interannual variation in meteorologically adjusted ozone levels in the eastern United States: a comparison of two approaches. Atmos Environ 41:705–716. doi: 10.1016/j.atmosenv.2006.09.010 CrossRefGoogle Scholar
  52. Zheng J, Zhong L, Wang T, Louie PKK, Li Z (2010) Ground-level ozone in the Pearl River Delta region: analysis of data from a recently established regional air quality monitoring network. Atmos Environ 44:814–823. doi: 10.1016/j.atmosenv.2009.11.032 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  • K. M. Mok
    • 1
  • K. V. Yuen
    • 1
  • K. I. Hoi
    • 1
  • K. M. Chao
    • 1
  • D. Lopes
    • 1
  1. 1.Department of Civil and Environmental EngineeringUniversity of MacauTaipaChina

Personalised recommendations