Advertisement

Temperatures across Europe: evidence of time trends

Abstract

This paper deals with the analysis of the temperatures in a group of 29 stations located in twelve European countries by looking at the coefficients in a linear time trend regression model and allowing for long memory patterns in the error term. The results indicate that long memory is present in practically all cases, and the time trend coefficients are statistically significant in the majority of the cases implying evidence of increasing warming trends. This pattern is particularly noticeable in the case of several stations located across Italy and France, which might be related with micro climates affecting these regions.

This is a preview of subscription content, log in to check access.

Access options

Buy single article

Instant unlimited access to the full article PDF.

US$ 39.95

Price includes VAT for USA

Subscribe to journal

Immediate online access to all issues from 2019. Subscription will auto renew annually.

US$ 199

This is the net price. Taxes to be calculated in checkout.

Fig. 1

Notes

  1. 1.

    I(1) denotes integration of order 1, which means that the series needs to be first differenced to render it stationary I(0). This latter concept (I(0)) is considered as a minimal requirement in time series to make statistical inference. Fractional integration or I(d) takes place when the number of differences required in the series is d and d is a fractional value.

  2. 2.

    Long memory is a feature of the I(d, d > 0) models that is characterized by the large degree of association between observations which are far distant in time. Mean reversion means that shocks will be transitory disappearing in the long run and takes place when d < 1.

  3. 3.

    A description of the functional form of this method can be found in Gil-Alana and Robinson (1997).

  4. 4.

    The length of the series varies from 67 to 359 observations. Nevertheless, the method employed in the paper, in spite of its asymptotic nature, it performs well even in small samples as it was shown in Gil-Alana and Robinson (1997) and other papers.

  5. 5.

    In other words, if you do a 30 hypothesis tests at P[Type-1 error] = 10%, you will end up rejecting about 3 of the hypothesis where the null is actually true, simply because the test is set up that way. The same situation holds for confidence intervals: the 95% confidence intervals will fail to contain the true value in 5 of every 100 cases just because such intervals are designed for a single case, not multiple times usage. Thus, the 95% intervals reported across Tables 1 and 2 contain a non-ignorable number of false positives.

  6. 6.

    The spectral density function is basically the frequency domain representation of the autocovariances based on Fourier transforms.

  7. 7.

    Once more note that the results for Navacerrada (NAVA), Salamanca (SALA) and Cadiz (CADI) should be taken with caution.

References

  1. Baillie RT, Chung SK (2002) Modelling and forecasting from trend-stationary long memory models with applications to climatology. Int J Forecast 18(2):215–226

  2. Bloomfield P (1973) An exponential model in the spectrum of a scalar time series. Biometrika 60(2):217–226

  3. Brunet M, Jones PD, Sigró J, Saladié O, Aguilar E, Moberg A, Della-Marta PM, Lister D, Walther A, López D (2007) Temporal and spatial temperature variability and change over Spain during 1850–2005. J Geophys Res-Atmos 112(D12):1–28. https://doi.org/10.1029/2006JD008249

  4. Bunde A, Havlin S, Koscielny-Bunde E, Schellnhuber HJ (2001) Long term persistence in the atmosphere: global laws and tests of climate models. Phys A: Stat Mech Applic 302(1-4):255–267

  5. Bunde A, Eichner JF, Kantelhardt JW, Havlin S (2005) Long term memory: a natural mechanism for the clustering of extreme events and anomalous residual times in climate records. Phys Rev Lett 94(4):048701

  6. Caballero R, Jewson S, Brix A (2002) Long memory in surface air temperature: Detection, modelling and application to weather derivative valuation. Clim Res 21(2):127–140

  7. Caloiero T, Coscarelli R, Ferrari E, Sirangelo B (2017) Trend analysis of monthly mean values and extreme indices of daily temperature in a region of southern Italy. Int J Climatol 37(S1):284–297. https://doi.org/10.1002/joc.5003

  8. Dahlhaus R (1989) Efficient parameter estimation for self-similar processes. Ann Stat 17(4):1749–1766

  9. Delvaux C, Ingels R, Vrábeĺ V, Journée M, Bertrand C (2018) Quality control and homogenization of the Belgian historical temperature data. Int J Climatol. https://doi.org/10.1002/joc.5792

  10. Franzke C (2010) Long range dependence and climate noise characteristics of Antarctic temperature data. J Clim 23(22):6074–6081

  11. Franzke C (2012) Nonlinear trends, long-range dependence, and climate noise properties of surface temperature. J Clim 25(12):4172–4183

  12. Gibelin A-L, Dubuisson B, Corre L, Deaux N, Jourdain S, Laval L, Piquemal J-M, Mestre O, Dennetière D, Desmidt S, Tamburini A (2014) Evolution de la température en France depuis les années 1950: Constitution d’un nouveau jeu de séries homogénéisées de référence. La Météorologie 87:45–53. https://doi.org/10.4267/2042/54336

  13. Gil-Alana LA (2003) An application of fractional integration to a long temperature time series. Int J Climatol 23(14):1699–1710

  14. Gil-Alana LA (2004) The use of the Bloomfield model as an approximation to ARMA processes in the context of fractional integration. Math Comput Model 39(4-5):429.436

  15. Gil-Alana LA (2005) Statistical model for the temperatures in the Northern hemisphere using fractional integration techniques. J Clim 18(24):5537–5369

  16. Gil-Alana LA (2008) Time trend estimation with breaks in temperature time series. Clim Chang 89(3-4):325–337

  17. Gil-Alana LA (2010) Climate change in the Iberian Peninsula. Evidence based on fractional integration techniques. Seguridad y Medio Ambiente 117(1):50–62

  18. Gil-Alana LA (2018) Maximum and minimum temperatures in the United States: time trends and persistence. Atmos Sci Lett 19(4):e810, 1−3. https://doi.org/10.1002/asl.810

  19. Gil-Alana LA, Robinson PM (1997) Testing of unit roots and other nonstationary hypotheses in macroeconomic time series. J Econ 80:241–268

  20. IPCC (2014) Climate Change 2014: synthesis report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. IPCC, Geneva, Switzerland, 151 pp

  21. Klingbjer P, Moberg A (2003) A composite monthly temperature record from Tornedalen in northern Sweden, 1802−2002. Int J Climatol 23(12):1465–1494. https://doi.org/10.1002/joc.946

  22. Mariani, L., and Zavatti, F. (2017), A dataset of the European average temperature anomalies for the period 1655-2016, Climatemonitor.

  23. Matiu M, Ankerst DP, Menzel A (2016) Asymmetric trends in seasonal temperature variability in instrumental records from ten stations in Switzerland, Germany and the UK from 1864 to 2012. Int J Climatol 36(1):13–27. https://doi.org/10.1002/joc.4326

  24. Nemec J, Gruber C, Chimani B, Auer I (2013) Trends in extreme temperature indices in Austria based on a new homogenised dataset. Int J Climatol 33(6):1538–1550. https://doi.org/10.1002/joc.3532

  25. Rea W, Reale M, Brown J (2011) Long memory in temperature reconstructions. Climate Change 107(3-4):247–265

  26. Ribes A, Corre L, Gibelin A-L, Dubuisson B (2016) Issues in estimating observed change at the local scale – a case study: the recent warming over France. Int J Climatol 36(11):3794–3806. https://doi.org/10.1002/joc.4593

  27. Robinson PM (1994) Efficient Tests of Nonstationary Hypotheses. J Am Stat Assoc 89(428):1420–1437

  28. Scorzini AR, Di Bacco M, Leopardi M (2018) Recent trends in daily temperature extremes over the central Adriatic region of Italy in a Mediterranean climatic context. Int J Climatol 38(S1):e741–e757. https://doi.org/10.1002/joc.5403

  29. Shen Z, Shi J, Lei Y (2017) Comparison of the long range climate memory in outgoing lognwage radiation over the Tibetan plateau and the Indian monsoon region. Adv Meteorol 7637351:1–7

  30. Whittle P (1951) Hypothesis testing in time series analysis. Uppsala, Almqvist & Wiksells Boktrycheri AB

  31. Yuan N, Fu Z, Liu S (2013) Long-term memory in climate variability: a new look based on fractional integral techniques. J Geophys Res-Atmos 118(23):12962–12969

  32. Yuan N, Ding M, Huang Y (2015) On the long term climate memory in the surface air temperature records over Antarctica: a non-negligible factor for trend evaluation. J Clim 28(15):5922–5934

Download references

Acknowledgments

Comments from the Editor and three anonymous reviewers are gratefully acknowledged.

Funding information

Prof. Luis A. Gil-Alana gratefully acknowledges financial support from the Ministerio de Economía y Competitividad (ECO2017-85503-R).

Author information

Correspondence to Luis A. Gil-Alana.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gil-Alana, L.A., Sauci, L. Temperatures across Europe: evidence of time trends. Climatic Change 157, 355–364 (2019). https://doi.org/10.1007/s10584-019-02568-6

Download citation

Keywords

  • European temperatures
  • Long memory
  • Time trends