Change Detection in Climate Time Series Based on Bounded-Variation Clustering

  • Mohammad Gorji Sefidmazgi
  • Mina Moradi Kordmahalleh
  • Abdollah Homaifar
  • Stefan Liess

Abstract

Climate time series are generally nonstationary which means that their statistical properties change with time. Analysis of nonstationary time series requires detecting of change points between a set of clusters, where model of time series in each cluster has different statistical parameters. Common change detection methods are based on assumptions that may not be valid generally. Bounded-variation clustering can solve the change detection problem with minimum restrictive assumptions. In this paper, this method is employed to detect the pattern of changes in the Pacific Decadal Oscillation and the piecewise linear trend of US temperature. An optimal number of the change points are found with the Bayesian information criterion.

Keywords

Time series Non-stationary Change detection Abrupt climate change Autocorrelation 

References

  1. Alley RB, Marotzke J, Nordhaus WD, Overpeck JT, Peteet DM, Pielke RA, Pierrehumbert RT, Rhines PB, Stocker TF, Talley LD, Wallace JM (2003) Abrupt climate change. Science 299(5615):2005–2010. doi:10.1126/science.1081056 CrossRefGoogle Scholar
  2. Estrada F, Perron P, Martinez-Lopez B (2013) Statistically derived contributions of diverse human influences to twentieth-century temperature changes. Nat Geosci 6(12):1050–1055. doi:10.1038/ngeo1999 CrossRefGoogle Scholar
  3. Gorji Sefidmazgi M, Moradi Kordmahalleh M, Homaifar A, Karimoddini A (2014a) A finite element based method for identification of switched linear systems. In: American Control Conference (ACC). IEEE, Portland, USA, pp 2644–2649. doi:10.1109/ACC.2014.6858898
  4. Gorji Sefidmazgi M, Sayemuzzaman M, Homaifar A (2014b) Non-stationary time series clustering with application to climate systems. In: Jamshidi M, Kreinovich V, Kacprzyk J (eds) Advance trends in soft computing, vol 312. Studies in fuzziness and soft computing. Springer International Publishing, Switzerland, pp 55–63. doi:10.1007/978-3-319-03674-8_6
  5. Gorji Sefidmazgi M, Sayemuzzaman M, Homaifar A, Jha M, Liess S (2014c) Trend analysis using non-stationary time series clustering based on the finite element method. Nonlinear Processes Geophys 21(3):605–615. doi:10.5194/npg-21-605-2014 CrossRefGoogle Scholar
  6. Gurobi (2014) Gurobi optimizer reference manual, Houston, USAGoogle Scholar
  7. Hastie T, Tibshirani R, Friedman JH (2009) The elements of statistical learning: data mining, inference, and prediction. Springer, New YorkCrossRefGoogle Scholar
  8. Horenko I (2010a) On clustering of non-stationary meteorological time series. Dyn Atmos Ocean 49(2–3):164–187. doi:http://dx.doi.org/10.1016/j.dynatmoce.2009.04.003
  9. Horenko I (2010b) On the identification of nonstationary factor models and their application to atmospheric data analysis. J Atmos Sci 67(5):1559–1574. doi:10.1175/2010JAS3271.1 CrossRefGoogle Scholar
  10. Jandhyala V, Fotopoulos S, MacNeill I, Liu P (2013) Inference for single and multiple change-points in time series. J Time Ser Anal. doi:10.1111/jtsa12035 Google Scholar
  11. Kaiser O, Horenko I (2014) On inference of statistical regression models for extreme events based on incomplete observation data. Commun Appl Math Comput Sci 9(1):143–174. doi:10.2140/camcos.2014.9.143 CrossRefGoogle Scholar
  12. Kehagias A, Fortin V (2006) Time series segmentation with shifting means hidden markov models. Nonlin Processes Geophys 13(3):339–352. doi:10.5194/npg-13-339-2006 CrossRefGoogle Scholar
  13. Liu RQ, Jacobi C, Hoffmann P, Stober G, Merzlyakov EG (2010) A piecewise linear model for detecting climatic trends and their structural changes with application to mesosphere/lower thermosphere winds over Collm, Germany. J Geophys Res Atmos 115(D22), D22105. doi:10.1029/2010JD014080 CrossRefGoogle Scholar
  14. Lohmann G (2009) Abrupt climate change modeling. In: Meyers RA (ed) Encyclopedia of complexity and systems science. Springer, New York, pp 1–21. doi:10.1007/978-0-387-30440-3_1 CrossRefGoogle Scholar
  15. Lyubchich V, Gel YR, El-Shaarawi A (2013) On detecting non-monotonic trends in environmental time series: a fusion of local regression and bootstrap. Environmetrics 24(4):209–226. doi:10.1002/env.2212 CrossRefGoogle Scholar
  16. Metzner P, Putzig L, Horenko I (2012) Analysis of persistent nonstationary time series and applications. Commun Appl Math Comput Sci 7(2):175–229. doi:10.2140/camcos.2012.7.175 CrossRefGoogle Scholar
  17. Milly PCD, Betancourt J, Falkenmark M, Hirsch RM, Kundzewicz ZW, Lettenmaier DP, Stouffer RJ (2008) Stationarity is dead: whither water management? Science 319(5863):573–574. doi:10.1126/science.1151915 CrossRefGoogle Scholar
  18. Robbins M, Gallagher C, Lund R, Aue A (2011) Mean shift testing in correlated data. J Time Ser Anal 32(5):498–511. doi:10.1111/j.1467-9892.2010.00707.x CrossRefGoogle Scholar
  19. Rodionov SN (2006) Use of prewhitening in climate regime shift detection. Geophys Res Lett 33(12), L12707. doi:10.1029/2006GL025904 CrossRefGoogle Scholar
  20. Ruggieri E (2013) A Bayesian approach to detecting change points in climatic records. Int J Climatol 33(2):520–528. doi:10.1002/joc.3447 CrossRefGoogle Scholar
  21. Seidel DJ, Lanzante JR (2004) An assessment of three alternatives to linear trends for characterizing global atmospheric temperature changes. J Geophys Res Atmos 109(D14), D14108. doi:10.1029/2003JD004414 CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Mohammad Gorji Sefidmazgi
    • 1
  • Mina Moradi Kordmahalleh
    • 1
  • Abdollah Homaifar
    • 2
  • Stefan Liess
    • 3
  1. 1.Department of Electrical EngineeringNorth Carolina A&T State UniversityGreensboroUSA
  2. 2.Department of Electrical and Computer EngineeringNorth Carolina Agricultural and Technical State UniversityGreensboroUSA
  3. 3.Department of Soil, Water, and ClimateUniversity of MinnesotaMinneapolisUSA

Personalised recommendations