Environmental and Ecological Statistics

, Volume 25, Issue 4, pp 443–469 | Cite as

Modeling and forecasting daily maximum hourly ozone concentrations using the RegAR model with skewed and heavy-tailed innovations

  • Alessandro José Queiroz SarnagliaEmail author
  • Nátaly Adriana Jiménez Monroy
  • Arthur Gomes da Vitória


This paper considers the modeling and forecasting of daily maximum hourly ozone concentrations in Laranjeiras, Serra, Brazil, through dynamic regression models. In order to take into account the natural skewness and heavy-tailness of the data, a linear regression model with autoregressive errors and innovations following a member of the family of scale mixture of skew-normal distributions was considered. Pollutants and meteorological variables were considered as predictors, along with some deterministic factors, namely week-days and seasons. The Oceanic Niño Index was also considered as a predictor. The estimated model was able to explain satisfactorily well the correlation structure of the ozone time series. An out-of-sample forecast study was also performed. The skew-normal and skew-t models displayed quite competitive point forecasts compared to the similar model with gaussian innovations. On the other hand, in terms of forecast intervals, the skewed models presented much better performance with more accurate prediction intervals. These findings were empirically corroborated by a forecast Monte Carlo experiment.


Adaptative LASSO Air pollution Forecasting Ozone SMSN-RegAR model 



The authors thank to Instituto Estadual de Meio Ambiente e Recursos Hídricos of Espírito Santo state for making the data sets used in this paper available. The authors also thank the extremely helpful comments of the associate editor and reviewers, which have improved substantially the quality of the paper.


  1. Babu GJ, Rao C (2004) Goodness-of-fit tests when parameters are estimated. Sankhya 66(1):63–74Google Scholar
  2. Biswas A, Hwang JS, Angers JF (2015) Air pollution effects on clinic visits in small areas of Taiwan revisited. Environ Ecol Stat 22(1):17–32CrossRefGoogle Scholar
  3. Bruno F, Guttorp P, Sampson PD, Cocchi D (2009) A simple non-separable, non-stationary spatiotemporal model for ozone. Environ Ecol Stat 16(4):515–529CrossRefGoogle Scholar
  4. Burnett R, Bartlett S, Krewski D, Robert G, Raad-Young M (1994) Air pollution effects on hospital admissions: a statistical analysis of parallel time series. Environ Ecol Stat 1(4):325–332CrossRefGoogle Scholar
  5. Cancho VG, Lachos VH, Ortega EM (2010) A nonlinear regression model with skew-normal errors. Stat Pap 51(3):547–558CrossRefGoogle Scholar
  6. Cao CZ, Lin JG, Shi JQ (2014) Diagnostics on nonlinear model with scale mixtures of skew-normal and first-order autoregressive errors. Statistics 48(5):1033–1047CrossRefGoogle Scholar
  7. Chen PC, Lai YM, Chan CC, Hwang JS, Yang CY, Wang JD (1999) Short-term effect of ozone on the pulmonary function of children in primary school. Environ Health Perspect 107(11):921CrossRefGoogle Scholar
  8. Chiogna M, Pauli F (2011) Modelling short-term effects of ozone on morbidity: an application to the city of Milano, Italy, 1995–2003. Environ Ecol Stat 18(1):169–184CrossRefGoogle Scholar
  9. Flores BE (1986) A pragmatic view of accuracy measurement in forecasting. Omega 14(2):93–98CrossRefGoogle Scholar
  10. Huang B, Banzon VF, Freeman E, Lawrimore J, Liu W, Peterson TC, Smith TM, Thorne PW, Woodruff SD, Zhang HM (2015) Extended reconstructed sea surface temperature version 4 (ERSST. v4). Part I: Upgrades and intercomparisons. J Clim 28(3):911–930CrossRefGoogle Scholar
  11. Johnson N, Kotz S, Balakrishnan N (1994) Continuous univariate probability distributions, vol 1. Wiley, New YorkGoogle Scholar
  12. Kalbarczyk R, Kalbarczyk E, Niedźwiecka-Filipiak I, Serafin L (2015) Ozone concentration at ground level depending on the content of NOx and meteorological conditions. Ecol Chem Eng S 22(4):527–541Google Scholar
  13. Karnosky DF, Skelly JM, Percy KE, Chappelka AH (2007) Perspectives regarding 50 years of research on effects of tropospheric ozone air pollution on US forests. Environ Pollut 147(3):489–506CrossRefGoogle Scholar
  14. Liu PWG (2007) Establishment of a Box–Jenkins multivariate time-series model to simulate ground-level peak daily one-hour ozone concentrations at Ta-Liao in Taiwan. J Air Waste Manag Assoc 57(9):1078–1090CrossRefGoogle Scholar
  15. Liu PWG, Johnson R (2002) Forecasting peak daily ozone levels-I. A regression with time series errors model having a principal component trigger to fit, (1991) ozone levels. J Air Waste Manag Assoc 52(9):1064–1074CrossRefGoogle Scholar
  16. Liu PWG, Johnson R (2003) Forecasting peak daily ozone levels: Part 2—A regression with time series errors model having a principal component trigger to forecast 1999 and 2002 ozone levels. J Air Waste Manag Assoc 53(12):1472–1489CrossRefGoogle Scholar
  17. Liu PWG, Tsai JH, Lai HC, Tsai DM, Li LW (2013) Establishing multiple regression models for ozone sensitivity analysis to temperature variation in Taiwan. Atmos Environ 79:225–235CrossRefGoogle Scholar
  18. Liu W, Huang B, Thorne PW, Banzon VF, Zhang HM, Freeman E, Lawrimore J, Peterson TC, Smith TM, Woodruff SD (2015) Extended reconstructed sea surface temperature version 4 (ERSST. v4): Part II. Parametric and structural uncertainty estimations. J Clim 28(3):931–951CrossRefGoogle Scholar
  19. Ljung GM, Box GE (1978) On a measure of lack of fit in time series models. Biometrika 65(2):297–303CrossRefGoogle Scholar
  20. Meng XL, Rubin DB (1993) Maximum likelihood estimation via the ecm algorithm: a general framework. Biometrika 80(2):267–278CrossRefGoogle Scholar
  21. Mourot G, Gasso K, Ragot JH (1999) Modelling of ozone concentrations using a Takagi-Sugeno model. Control Eng Pract 7(6):707–715CrossRefGoogle Scholar
  22. Pankratz A (2012) Forecasting with dynamic regression models. Wiley, New YorkGoogle Scholar
  23. Percy K, Legge A, Krupa S (2003) Tropospheric ozone: a continuing threat to global forests? In: Karnosky D, Percy K, Chappelka A, Simpson C, Pikkarainen J (eds) Developments in environmental science: air pollution, global change and forests in the new millennium, Chap. 4, vol 3. Elsevier, Amsterdam, pp 85–117CrossRefGoogle Scholar
  24. Prybutok VR, Yi J, Mitchell D (2000) Comparison of neural network models with ARIMA and regression models for prediction of Houston’s daily maximum ozone concentrations. Eur J Oper Res 122(1):31–40CrossRefGoogle Scholar
  25. R Core Team (2017) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, Accessed 2 Oct 2018
  26. Seinfeld JH, Pandis SN (2016) Atmospheric chemistry and physics: from air pollution to climate change. Wiley, New YorkGoogle Scholar
  27. Spellman G (1999) An application of artificial neural networks to the prediction of surface ozone concentrations in the United Kingdom. Appl Geogr 19(2):123–136CrossRefGoogle Scholar
  28. Tibshirani R (1996) Regression shrinkage and selection via the LASSO. J R Stat Soc Ser B (Methodological) 58:267–288Google Scholar
  29. Tsay RS (2005) Analysis of financial time series. Wiley, New YorkCrossRefGoogle Scholar
  30. Wang H, Li G, Tsai CL (2007) Regression coefficient and autoregressive order shrinkage and selection via the LASSO. J R Stat Soc Ser B (Stat Methodol) 69(1):63–78CrossRefGoogle Scholar
  31. Wang P, Baines A, Lavine M, Smith G (2012) Modelling ozone injury to US forests. Environ Ecol Stat 19(4):461–472CrossRefGoogle Scholar
  32. Xie FC, Lin JG, Wei BC (2009) Diagnostics for skew-normal nonlinear regression models with AR(1) errors. Comput Stat Data Anal 53(12):4403–4416CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  • Alessandro José Queiroz Sarnaglia
    • 1
    Email author
  • Nátaly Adriana Jiménez Monroy
    • 1
  • Arthur Gomes da Vitória
    • 2
  1. 1.LECON - Statistics DepartmentFederal University of Espírito SantoVitóriaBrazil
  2. 2.PIVIC ProgramFederal University of Espírito SantoVitóriaBrazil

Personalised recommendations