Statistical Science

Nonparametric Analysis of Temporal Trend When Fitting Parametric Models to Extreme­Value Data

Peter Hall and Nader Tajvidi

Full-text: Open access

Abstract

A topic of major current interest in extreme­value analysis is the investigation of temporal trends. For example, the potential influence of “greenhouse” effects may result in severe storms becoming gradually more frequent, or in maximum temperatures gradually increasing, with time. One approach to evaluating these possibilities is to fit, to data, a parametric model for temporal parameter variation, as well as a model describing the marginal distribution of data at any given point in time. However, structural trend models can be difficult to formulate in many circumstances, owing to the complex way in which different factors combine to influence data in the form of extremes. Moreover, it is not advisable to fit trend models without empirical evidence of their suitability. In this paper, motivated by datasets on windstorm severity and maximum temperature, we suggest a nonparametric approach to estimating temporal trends when fitting parametric models to extreme values from a weakly dependent time series. We illustrate the method through applications to time series where the marginal distributions are approximately-Pareto, generalized­Pareto, extreme­value or Gaussian. We introduce time­varying probability plots to assess goodness of fit, we discuss local­likelihood approaches to fitting the marginal model within a window and we propose temporal cross­validation for selecting window width. In cases where both location and scale are estimated together, the Gaussian distribution is shown to have special features that permit it to playa universal role as a “nominal” model for the marginal distribution.

Article information

Source
Statist. Sci., Volume 15, Number 2 (2000), 153-167.

Dates
First available in Project Euclid: 24 December 2001

Permanent link to this document
https://projecteuclid.org/euclid.ss/1009212755

Digital Object Identifier
doi:10.1214/ss/1009212755

Mathematical Reviews number (MathSciNet)
MR1788730

Keywords
Bandwidth, , , , , , , , cross­validation extreme­value distribution kernel location estimate nonparametric regression Pareto distribution probability plot scale estimate

Citation

Hall, Peter; Tajvidi, Nader. Nonparametric Analysis of Temporal Trend When Fitting Parametric Models to Extreme­Value Data. Statist. Sci. 15 (2000), no. 2, 153--167. doi:10.1214/ss/1009212755. https://projecteuclid.org/euclid.ss/1009212755


Export citation

References

  • Carroll, R. J., Ruppert, D. and Welsh, A. H. (1998). Local estimating equations. J. Amer. Statist. Assoc. 93 214-227.
  • Chambers, J. M. and Hastie, T. J. (1992). Statistical Models in S. Wadsworth, Pacific Grove, CA.
  • Cleveland, W. S. and Devlin, S. J. (1988). Locally-weighted regression: an approach to regression analysis by local fitting. J. Amer. Statist. Assoc. 83 597-610.
  • Davison, A. C. and Ramesh, N. I. (2000). Local likelihood smoothing of sample extremes. J. Roy. Statist. Soc. Ser. B 62 191- 208.
  • Fan, J. (1993). Local linear regression smoothers and their minimax efficiencies. Ann. Statist. 21 196-216.
  • Fan, J. and Gijbels, I. (1996). Local Polynomial Modelling and its Applications. Chapman and Hall, London.
  • Fan, J., Heckman, N. E. and Wand, M. P. (1995). Local polynomial kernel regression for generalized linear models and quasi-likelihood functions. J. Amer. Statist. Assoc. 90 141- 150.
  • Fan, J., Yao, Q. and Tong, H. (1996). Estimation of conditional densities and sensitivitymeasures in nonlinear dynamical systems. Biometrika 83 189-206.
  • Gu, C. (1990). Adaptive spline smoothing in non-Gaussian regression models. J. Amer. Statist. Assoc. 85 801-807.
  • Gu, C., Bates, D. M., Chen, Z. and Wahba, G. (1989). The computation of GCV functions through Householder tridiagonalization with application to the fitting of interaction spline models. SIAM J. Matrix Anal. Appl. 10 457-480.
  • Hall, P. and Marron, J. S. (1991). Local minima in crossvalidation functions. J. Roy. Statist. Soc. Ser. B 53 245-252.
  • Hall, P., Wolff, C. L. and Yao, Q. (1999). Methods for estimating a conditional distribution function. J. Amer. Statist. Assoc. 94 154-163.
  • Hart, J. D. and Vieu, P. (1990). Data-driven bandwidth choice for densityestimation based on dependent data. Ann. Statist. 18 873-890.
  • Jones, P. D. (1994). Hemispheric surface air temperature variations: a reanalysis and an update to 1993. J. Climate 7 1794-1802.
  • Leadbetter, M. R., Lindgren, G. and Rootz´en, H. (1983). Extremes and Related Properties of Random Sequences and Processes. Springer, New York.
  • Leadbetter, M. R. and Rootz´en, H. (1988). Extremal theoryfor stochastic processes. Ann. Probab. 16 431-478.
  • Lehmann, E. L. (1983). Theory of Point Estimation. Wiley, New York.
  • Nicholls, N. and Lavery, B. (1992). Australian rainfall trends during the twentieth century. Internat. J. Climatology 12 153-163.
  • Peligrad, M. (1986). Recent advances in the central limit theorem and its weak invariance principle for mixing sequences of random variables. In Dependence in Probability and Statistics (E. Eberlein and M. S. Taqqu, eds) 193-223. Birkh¨auser, Boston.
  • Raftery, A. E. (1989). Discussion of Smith (1989). Statist. Sci. 4 378-381.
  • Rootz´en, H. and Tajvidi, N. (1997). Extreme value statistics and windstorm losses: a case study. Scand. Actuarial J. 70-94.
  • Ruppert, D. (1995). Empirical-bias bandwidths for local polynomial nonparametric regression and densityestimation. Unpublished manuscript.
  • Severini, T. A. and Staniswallis, J. G. (1994). Quasilikelihood estimation in semiparametric models. J. Amer. Statist. Assoc. 89 501-511.
  • Smith, R. L. (1989). Extreme value analysis of environmental time series: an application to trend detection in ground-level ozone (with discussion). Statist. Sci. 4 367-393.
  • Staniswallis, J. G. (1989). The kernel estimates of a regression function in likelihood-based models. J. Amer. Statist. Assoc. 84 276-283.
  • Torok, S. J. and Nicholls, N. (1996). A historical annual temperature dataset for Australia. Austral. Met. Mag. 45 251- 260.
  • Weisberg, S. and Welsh, A. H. (1994). Estimating the missing link function. Ann. Statist. 22 1674-1700.
  • Yoshihara, K. (1976). Limiting behaviour of U-statistics for stationaryabsolutelyregular processes. Z. Wahrsch. Verw. Gebiete 35 237-252.