A hidden Markov model for earthquake prediction

  • Cheuk Fung Yip
  • Wai Leong Ng
  • Chun Yip Yau
Original Paper


Earthquake occurrence is well-known to be associated with structural changes in underground dynamics, such as stress level and strength of electromagnetic signals. While the causation between earthquake occurrence and underground dynamics remains elusive, the modeling of changes in underground dynamics can provide insights on earthquake occurrence. However, underground dynamics are usually difficult to measure accurately or even unobservable. In order to model and examine the effect of the changes in unobservable underground dynamics on earthquake occurrence, we propose a novel model for earthquake prediction by introducing a latent Markov process to describe the underground dynamics. In particular, the model is capable of predicting the change-in-state of the hidden Markov chain, and thus can predict the time and magnitude of future earthquake occurrences simultaneously. Simulation studies and applications on a real earthquake dataset indicate that the proposed model successfully predicts future earthquake occurrences. Theoretical results, including the stationarity and ergodicity of the proposed model, as well as consistency and asymptotic normality of model parameter estimation, are provided.


Change-point ETAS model EM algorithm Structural break 

Supplementary material

477_2017_1457_MOESM1_ESM.pdf (351 kb)
Supplementary material 1 (pdf 351 KB)


  1. Adelfio G, Chiodi M (2015) Alternated estimation in semi-parametric space-time branching-type point processes with application to seismic catalogs. Stoch Environ Res Risk Assess 29:443450Google Scholar
  2. Bartolucci F, Farcomeni A (2009) A multivariate extension of the dynamic logit model for longitudinal data based on a latent Markov heterogeneity structure. J Am Stat Assoc 104:816–831CrossRefGoogle Scholar
  3. Bartolucci F, Farcomeni A (2015) A discrete time event-history approach to informative drop-out in mixed latent Markov models with covariates. Biometrics 71:80–89CrossRefGoogle Scholar
  4. Bartolucci F, Farcomeni A (2015a) Information matrix for hidden Markov models with covariates. Stat Comput 25:515–526CrossRefGoogle Scholar
  5. Bartolucci F, Farcomeni A, Pennoni F (2013) Latent Markov models for longitudinal data. Chapman & Hall, CRC Press, Boca RatonGoogle Scholar
  6. Baum LE, Petrie T, Soules G, Weiss N (1970) A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains. Ann Math Stat 41(1):164–171CrossRefGoogle Scholar
  7. Baum LE (1972) An equality and associated maximization technique in statistical estimation for probabilistic functions of Markov processes. Inequalities 3:1–8Google Scholar
  8. Bulla J, Berzel A (2008) Computational issues in parameter estimation for stationary hidden Markov models. Comput Stat 23(1):1–18CrossRefGoogle Scholar
  9. Cameletti M, De Rubies V, Ferrari C, Sbarra P, Tosi P (2016) An ordered probit model for seismic intensity data. Stoch Environ Res Risk Assess 1–10. doi: 10.1007/s00477-016-1260-4
  10. Cosentino P, Ficarra V, Luzio D (1977) Truncated exponential frequency-magnitude relationship in earthquake statistics. Bull Seismol Soc Am 67(6):1615–1623Google Scholar
  11. Davis PM, Jackson DD, Kagan YY (1989) The longer it has been since the last earthquake, the longer the expected time till the next? Bull Seismol Soc Am 79(5):1439–1456Google Scholar
  12. Ebel JE, Chambers DW, Kafka AL, Baglivo JA (2007) Non-Poissonian earthquake clustering and the hidden Markov model as bases for earthquake forecasting in California. Seismol Res Lett 78(1):57–65CrossRefGoogle Scholar
  13. Fierro R, Leiva V (2017) A stochastic methodology for risk assessment of a large earthquake when a long time has elapsed. Stoch Environ Res Risk Assess 1–10. doi: 10.1007/s00477-016-1288-5
  14. Finazzi F, Fassò A (2016) A statistical approach to crowdsourced smartphone-based earthquake early. Stoch Environ Res Risk Assess 1–10. doi: 10.1007/s00477-016-1240-8
  15. Huang Q, Ikeya M (1998) Seismic electromagnetic signals (SEMS) explained by a simulation experiment using electromagnetic waves. Phys Earth Planet Inter 109(3):107–114CrossRefGoogle Scholar
  16. Hughes JP, Guttorp P, Charles SP (1999) A non-homogeneous hidden Markov model for precipitation occurrence. J R Stat Soc Ser C (Appl Stat) 48(1):15–30CrossRefGoogle Scholar
  17. Hutton K, Woessner J, Hauksson E (2010) Earthquake monitoring in southern California for seventy-seven years (1932–2008). Bull Seismol Soc Am 100(2):423–446CrossRefGoogle Scholar
  18. Leroux BG (1992) Maximum-likelihood estimation for hidden Markov models. Stoch Process Appl 40(1):127–143CrossRefGoogle Scholar
  19. Leroux BG, Puterman ML (1992) Maximum-penalized-likelihood estimation for independent and Markov-dependent mixture models. Biometrics 48(2):545–558CrossRefGoogle Scholar
  20. Levinson SE, Rabiner LR, Sondhi MM (1983) An introduction to the application of the theory of probabilistic functions of a Markov process to automatic speech recognition. Bell Syst Tech J 62(4):1035–1074CrossRefGoogle Scholar
  21. Ogata Y (1981) On Lewis’ simulation method for point processes. IEEE Trans Inf Theory 27(1):23–31CrossRefGoogle Scholar
  22. Ogata Y (1988) Statistical models for earthquake occurrences and residual analysis for point processes. J Am Stat Assoc 83(401):9–27CrossRefGoogle Scholar
  23. Ogata Y (1998) Space-time point-process models for earthquake occurrences. Ann Inst Stat Math 50(2):379–402CrossRefGoogle Scholar
  24. Orfanogiannaki K, Karlis D, Papadopoulos GA (2010) Identifying seismicity levels via Poisson hidden Markov models. Pure Appl Geophys 167(8–9):919–931CrossRefGoogle Scholar
  25. Park SK, Johnston MJ, Madden TR, Morgan FD, Morrison HF (1993) Electromagnetic precursors to earthquakes in the ULF band: a review of observations and mechanisms. Rev Geophys 31(2):117–132CrossRefGoogle Scholar
  26. Reid HF (1910) The mechanics of the earthquake, vol. II of Lawson, A.C., chairman, The California earthquake of April 18, 1906: report of the State Earthquake Investigation Commission: Carnegie Institution of Washington Publication, p 192 (reprinted in 1969)Google Scholar
  27. Seeber L, Armbruster JG (2000) Earthquakes as beacons of stress change. Nature 407(6800):69CrossRefGoogle Scholar
  28. Sornette D, Knopoff L (1997) The paradox of the expected time until the next earthquake. Bull Seismol Soc Am 87(4):789–798Google Scholar
  29. Team RC (2016) R: language and environment for statistical computing. R Foundation for Statistical Computing, 2005; Vienna, AustriaGoogle Scholar
  30. Telesca L, Giocoli A, Lapenna V, Stabile TA (2015) Robust identification of periodic behavior in the time dynamics of short seismic series: the case of seismicity induced by Pertusillo Lake, southern Italy. Stoch Environ Res Risk Assess 29:1437. doi: 10.1007/s00477-014-0980-6 CrossRefGoogle Scholar
  31. Utsu T (1961) A statistical study on the occurrence of aftershocks. Geophys Mag 30(4):521–605Google Scholar
  32. Viterbi AJ (1967) Error bounds for convolutional codes and an asymptotically optimum decoding algorithm. IEEE Trans Inf Theory 13(2):260–269CrossRefGoogle Scholar
  33. Welch LR (2003) Hidden Markov models and the Baum–Welch algorithm. IEEE Inf Theory Soc Newsl 53(4):10–13Google Scholar
  34. Zucchini W, MacDonald IL (2009) Hidden Markov models for time series: an introduction using R. CRC Press, Boca RatonCrossRefGoogle Scholar
  35. Zucchini W, Raubenheimer D, MacDonald IL (2008) Modeling time series of animal behavior by means of a latent-state model with feedback. Biometrics 64(3):807–815CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  1. 1.Department of Medicine and TherapeuticsChinese University of Hong KongShatinChina
  2. 2.Department of StatisticsChinese University of Hong KongShatinChina

Personalised recommendations