Abstract
The Halphen family of distributions is a flexible and complete system to fit sets of observations independent and identically distributed. Recently, it is shown that this family of distributions represents a potential alternative to the generalized extreme value distributions to model extreme hydrological events. The existence of jointly sufficient statistics for parameter estimation leads to optimality of the method of maximum likelihood (ML). Nevertheless, the ML method requires numerical approximations leading to less accurate values. However, estimators by the method of moments (MM) are explicit and their computation is fast. Even though MM method leads to good results, it is not optimal. In order to combine the advantages of the ML (optimality) and MM (efficiency and fast computations), two new mixed methods were proposed in this paper. One of the two methods is direct and the other is iterative, denoted respectively direct mixed method (MMD) and iterative mixed method (MMI). An overall comparison of the four estimation methods (MM, ML, MMD and MMI) was performed using Monte Carlo simulations regarding the three Halphen distributions. Generally, the MMI method can be considered for the three Halphen distributions since it is recommended for a majority of cases encountered in hydrology. The principal idea of the mixed methods MMD and MMI could be generalized for other distributions with complicated density functions.
Similar content being viewed by others
References
Abramowitz M, Stegun IA (1972) Handbook of mathematical functions. Dover, New York, 1006 pp
Barndorff-Nielsen OE (1978) Hyperbolic distributions and distributions on hyperbolae. Scand J Stat 5:151–157
Barndorff-Nielsen O, Halgreen C (1977) Infinite divisibility of the hyperbolic and generalized inverse Gaussian distributions. Zeitschrift für Wahrscheinlichkeitstheorie und Verwandte Gebiete 38(4):309–311
Blasield P (1978) The shape of the generalized inverse Gaussian and hyperbolic distributions. Research report No. 37. Dept. Theor. Statist., Aarhus University
Bobée B (1979) Comment on: the log–Pearson type 3 distribution: the T-year event and its asymptotic standard error by maximum likelihood theory. Water Resour Res 15:189–190
Bobée B (1999) Extreme flood events valuation using frequency analysis: a critical review. Houille Blanche 54(7–8):100–105
Bobée B, Ashkar F (1988) Generalized method of moments applied to LP3 distribution. J Hydraul Eng (ASCE) 114:899–909
Bobée B, Ashkar F (1991) The Gamma family and derived distributions applied in hydrology. Water Resources Publications, Littleton, Colorado, 203 pp
Bobée B, Ashkar F, Perreault L (1993) Two kinds of moment ratio diagrams and their applications in hydrology. Stoch Hydrol Hydraul 7:41–65
Chaire en Hydrologie Statistique (CHS) (2002) HYFRAN: Software for frequency analysis in hydrology (in French). Technical report. INRS-Eau, Quebec, Canada
Chebana, F, El Adlouni S, Bobée B (2006) Propriétés des estimateurs du maximum de vraisemblance des paramètres et quantiles des lois de Halphen (in French). Research report, I 219. INRS-ETE, Québec, Canada
Chebana F, El Adlouni S, Bobée B (2008a) Method of moments of the Halphen distribution parameters. Stoch Environ Res Risk Assess 22:749–757. doi:10.1007/s00477-007-0184-4
Chebana F, El Adlouni S, Bobée B (2008b) Lois de Halphen: méthodes d’estimation mixtes et comparaison (in French). Research report, R-994. INRS-ETE, Québec, Canada, 72 pp
Chow VT, Maidment DR, Mays LR (1988) Applied hydrology. McGraw-Hill, New York
El Adlouni S, Bobée B (2007) Sampling techniques for Halphen distributions. J Hydrol Eng 12(6):592–604
El Adlouni S, Bobée B, Ouarda TBMJ (2008) On the tails of extreme event distributions in hydrology. J Hydrol 355:16–33
El Adlouni S, Chebana F, Bobée B (2009) Generalized extreme value vs. Halphen system: an explanatory study. J Hydrol Eng ASCE. doi:10.1061/(ASCE)HE.1943-5584.0000152
Fitzgerald DL (2000) Statistical aspects of Tricomi’s function and modified Bessel functions of the second kind. Stoch Environ Res Risk Assess 14:139–158
Guillot P (1964) Une extension des lois A de Halphen comprenant comme cas limite la loi de Galton–Gibrat in (French). Revue de Statistique Appliquée 12:63–73
Halphen E (1941) Sur un nouveau type de couBRe de fréquence. Comptes Rendus de l’Académie des Sciences, 213, 633–635. Published under the name of «Dugué» due to war constraints
Halphen E (1955) Les fonctions factorielles. Publications de l’Institut de Statistique de l’Université de Paris, vol IV, Fascicule I. pp 21–39
Jørgensen B (1982) Statistical properties of the generalized inverse Gaussian distribution. Lecture notes in statistics. Springer, New York
Kirby W (1974) Algebraic boundedness of sample statistics. Water Resour Res 10(2):220–222
Larivaille, P. (1960) Lois de Halphen: Estimation du paramètre d’échelle (in French). Publication Interne de L’E.D.F., Direction de L’Équipement, 3 (194). EDF, Grenoble, France
Le Cam L, Morlat G (1949) Les lois des débits des rivières françaises (in French). La Houille Blanche. 1–7. No spécial B
Morlat G (1951) Note sur l’estimation des débits de crues (in French). La Houille Blanche, No spécial B, pp 663–681
Morlat G (1956) Les lois de probabilité de Halphen in (French). Revue de Statistique Appliquée 3:21–43
Natural Environment Research Council (NERC) (1975) Flood studies report, vol 1, London
Perreault L, Bobée B, Rasmussen PF (1997) Les lois de Halphen (in French). Research report, R-498, INRS-Eau, Université du Québec
Perreault L, Bobée B, Rasmussen PF (1999a) Halphen distribution system. I: Mathematical and statistical properties. J Hydrol Eng 4:189–199
Perreault L, Bobée B, Rasmussen PF (1999b) Halphen distribution system. II: Parameter and quantile estimation. J Hydrol Eng 4:200–208
Puig P (2008) A note on the harmonic law: a two-parameter family of distributions for ratios. Stat Probab Lett 78(3):320–326
Rao AR, Hamed KH (2000) Flood frequency analysis. CRC Press, Boca Raton
Seshadri V (1993) The inverse Gaussian distribution. Clarendon Press, Oxford
Seshadri V (1997) Halphen’s laws. In: Kotz S, Johnson NL (eds) Encyclopedia of statistical sciences, 1, pp 302–306
Sichel HS (1975) On a distribution law for word frequencies. J Am Stat Asses 70:542–547
US Water Resources Council (1981) Guidelines for Determining Flood Flow Frequency. Bull. 17B, Washington DC
Watson GN (1996) A treatise on the theory of bessel functions, 2nd edn. Cambridge University Press, Cambridge, 812 pp
Willeke GE, Hosking JRM, Wallis JR, Guttman NB (1995) The national drought atlas (draft), IWR Rep. 94-NDS-4, US Army Corps of Eng., Fort Belvoir, Va
Acknowledgments
The authors wish to thank the Natural Sciences and Engineering Research Council (NSERC) for financial support. The authors wish to thank the Editor-in-Chief and the two anonymous reviewers for their useful comments which led to the improvement of the paper.
Author information
Authors and Affiliations
Corresponding author
Additional information
Codes related to the presented estimation methods are available upon request from the authors.
Appendices
Appendix A: ML estimation
The object of this appendix is to provide some elements related to the ML estimation. These elements are required to the development of the MMD and MMI methods. Hereafter, we present in more details the steps (a) and (b) given in Sect. 3.2. The main steps of the ML method, based on the previous numerical resolution approach for the three Halphen distributions, are illustrated in Fig. 8.
A.1 ML for HA
-
a.
For a fixed ν, α(ν) and m(ν) are solutions of the system of Eqs. 23 and 24 which can be rewritten as:
$$ D_{A} (\alpha (\nu ),\nu ) = {\frac{A}{H}} $$(38)$$ m(\nu ) \, = {\frac{A}{{R_{A} (\alpha (\nu ), \, \nu )}}} $$(39)where
$$ D_{A} (\alpha ,\nu ) = {\frac{{K_{\nu + 1} (2\alpha )K_{\nu - 1} (2\alpha )}}{{\left[ {K_{\nu } (2\alpha )} \right]^{2} }}} $$(40)and
$$ R_{A} (\alpha ,\nu ) = {\frac{{K_{\nu + 1} (2\alpha )}}{{K_{\nu } (2\alpha )}}} $$(41)
-
b.
In order to determine the variation interval of ν, we find signs of l ′HA (U) and l ′HA (–U) where l ′HA (ν) is the derivative of the partial log-likelihood function log L HA, we have:
$$ l^{\prime}_{\text{HA}} ( - U) = n\left[ {\log \left( {{\frac{G}{H}}{\frac{1}{U}}} \right) + \Uppsi (U)} \right]\;{\text{and}}\;l^{\prime}_{\text{HA}} (U) = n\left[ {\log \left( {{\frac{G}{A}}U} \right) - \Uppsi (U)} \right] $$(42)where \( \Uppsi (z) = \partial \left[ {\log \Upgamma (z)} \right]/\partial z \) is the digamma function and \( U = AH^{ - 1} /\left[ {AH^{ - 1} - 1} \right]. \) Note that U > 1, since H ≤ A. Hence:
-
If l HA ′(U) > 0 and l HA ′(–U) > 0, then the value of ν that maximizes log L HA(ν|α, m) is greater than U, and in that case, the sample would better presented by a Gamma distribution;
-
If l HA ′(U) < 0 and l HA ′(–U) < 0, then the value of ν that maximizes log L HA is less than −U, and in that case, the sample would be better presented by an Inverse Gamma distribution.
-
If l ′HA (U) < 0 and l HA ′(–U) > 0, then the value of ν that maximizes log L HA is in the interval ]–U;U[ ; and the ML estimators are solutions of the system composed by equations 38 and 39. In that case, the partial log-likelihood function is given by:
$$ \log L_{\rm HA} (\nu |\alpha ,m) = n\left\{ {\log \left[ {{\frac{{G^{\nu - 1} }}{{2m^{\nu } K_{\nu } (2\alpha )}}}} \right] - {\frac{{\alpha \left[ {K_{\nu + 1} (2\alpha ) + K_{\nu - 1} (2\alpha )} \right]}}{{K_{\nu } (2\alpha )}}}} \right\} $$(43)
-
A.2 ML for HB
where
and
-
b.
In order to establish the variation interval of ν, we determine the sign of l ′HB (ν), the derivative of the partial log-likelihood function log L HB, evaluated at \( V = 1/\left[ {2\left( {Q/A^{2} - 1} \right)} \right], \) we obtain then:
$$ l^{\prime}_{\rm HB} (V) = 2n\left[ {\log \left( {{\frac{G}{A}}2V} \right) - \Uppsi (2V)} \right] $$(48)Hence,
-
If l ′HB (V) ≥ 0, then the argument of the maximum of the partial log-likelihood function log L HB with respect to ν is greater than V. In that case, the Gamma distribution is more appropriate to represent the sample;
-
If l ′HB (V) < 0, then the maximum of the partial log-likelihood function log L HB with respect to ν is reached in the interval ]0;V[ . In that case, the partial log-likelihood function is given by:
$$ \log L_{\rm HB} (\nu |\alpha ,m) = n\left\{ {\log \left[ {{\frac{{2G^{2\nu - 1} }}{{m^{2\nu } ef_{\nu } (\alpha )}}}} \right] - {\frac{{ef_{\nu + 1} (\alpha )}}{{ef_{\nu } (\alpha )}}} + \hat{\alpha }{\frac{{ef_{{\nu + {1 \mathord{\left/ {\vphantom {1 2}} \right. \kern-\nulldelimiterspace} 2}}} (\alpha )}}{{ef_{\nu } (\alpha )}}}} \right\} $$(49)
-
A.3 ML for HIB
-
a.
For a fixed ν, α(ν) and m(ν) are solutions of the following system obtained from Eq. 29 and 30:
$$ D_{B} (\alpha (\nu ),\nu ) = {\frac{{H^{2} }}{IQ}} $$(50)$$ m(\nu ) \, = \, HR_{B} (\alpha (\nu ),\nu ) $$(51)
where the functions D B and R B are given respectively by expressions 46 and 47.
-
b.
To establish the interval in which belongs ν, we determine the sign of l ′HIB (W)which is the derivative of the partial log-likelihood function log L HIB evaluated at \( W = 1/\left[ {2\left( {H^{2} IQ^{ - 1} - 1} \right)} \right], \) given by:
$$ l^{\prime}_{\rm HIB} (W) = 2n\left[ {\log \left( {{\frac{H}{G}}2W} \right) - \Uppsi (2W)} \right] $$(52)Therefore,
-
If l ′HIB (W) ≥ 0, then the value of ν that maximizes the log-likelihood function log L HIB is greater than W. In this case, the Inverse Gamma distribution is more appropriate to represent the sample;
-
If l ′HIB (W) < 0, then the maximum of function log L HIB with respect to ν is reached in the interval]0;W[. In this case, the log-likelihood function is given by:
$$ \log L_{\rm HIB} (\nu |\alpha ,m) = n\left\{ {\log \left[ {{\frac{{2m^{2\nu } }}{{G^{2\nu + 1} ef_{\nu } (\alpha )}}}} \right] - {\frac{{ef_{\nu + 1} (\alpha )}}{{ef_{\nu } (\alpha )}}} + \hat{\alpha }{\frac{{ef_{{\nu + {1 \mathord{\left/ {\vphantom {1 2}} \right. \kern-\nulldelimiterspace} 2}}} (\alpha )}}{{ef_{\nu } (\alpha )}}}} \right\} $$(53)
-
Rights and permissions
About this article
Cite this article
Chebana, F., Adlouni, S.E. & Bobée, B. Mixed estimation methods for Halphen distributions with applications in extreme hydrologic events. Stoch Environ Res Risk Assess 24, 359–376 (2010). https://doi.org/10.1007/s00477-009-0325-z
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00477-009-0325-z