Skip to main content

Advertisement

Log in

Frequency analysis of droughts using the Plackett copula and parameter estimation by genetic algorithm

  • Original Paper
  • Published:
Stochastic Environmental Research and Risk Assessment Aims and scope Submit manuscript

Abstract

This study aims to model the joint probability distribution of drought duration, severity and inter-arrival time using a trivariate Plackett copula. The drought duration and inter-arrival time each follow the Weibull distribution and the drought severity follows the gamma distribution. Parameters of these univariate distributions are estimated using the method of moments (MOM), maximum likelihood method (MLM), probability weighted moments (PWM), and a genetic algorithm (GA); whereas parameters of the bivariate and trivariate Plackett copulas are estimated using the log-pseudolikelihood function method (LPLF) and GA. Streamflow data from three gaging stations, Zhuangtou, Taian and Tianyang, located in the Wei River basin, China, are employed to test the trivariate Plackett copula. The results show that the Plackett copula is capable of yielding bivariate and trivariate probability distributions of correlated drought variables.

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

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10

Similar content being viewed by others

References

  • Aas K, Czado C, Frigessi A, Bakken H (2007) Pair-copula constructions of multiple dependence. Insur Math Econ. doi:10.1016/j.jnsmatheco.2007.02.001:1-17

  • Amal SH (2005) Goodness-of-fit for the generalized exponential distribution. http://interstat.statjournals.net/YEAR/2005/articles/0507001.pdf

  • Bárdossy A (2006) Copula-based geostatistical models for groundwater quality parameters. Water Resour Res 42:W11416. doi:10.1029/2005WR004754

    Article  CAS  Google Scholar 

  • Bárdossy A, Li J (2008) Geostatistical interpolation using copulas. Water Resour Res 44:W07412. doi:10.1029/2007WR006115

    Article  Google Scholar 

  • Bonaccorso B, Bordi I, Cancelliere A, Rossi G, Sutera A (2003a) Spatial variability of drought: an analysis of SPI in Sicily. Water Resour Manag 17:273–296

    Article  Google Scholar 

  • Bonaccorso B, Cancelliere A, Rossi G (2003b) An analytical formulation of return period of drought severity. Stoch Env Res Risk Assess 17(3):157–174

    Article  Google Scholar 

  • Cabrera E, García-Serra J (1999) Drought management planning in water supply systems: proceedings from the Uimp International Course Held in Valencia, December 1997. Kluwer, Dordrecht

    Google Scholar 

  • Cancelliere A, Salas JD (2004) Drought length properties for periodic-stochastic hydrological data. Water Resour Res 10(2):1–13

    Google Scholar 

  • Chang LC (2008) Guiding rational reservoir flood operation using penalty-type genetic algorithm. J Hydrol 354(1–4):65–74

    Article  Google Scholar 

  • Cunnane C (1978) Unbiased plotting positions—a review. J Hydrol 37(3–4):205–222

    Article  Google Scholar 

  • De Michele C, Salvadori G (2003) A generalized Pareto intensity-duration model of storm rainfall exploiting 2-copulas. J Geophys Res 108(D2):4067. doi:10.1029/2002JD002534

    Article  Google Scholar 

  • Dobrić J, Schmid F (2007) A goodness of fit test for copulas based on Rosenblatt’s transformation. Comput Stat Data Anal 51:4633–4642

    Article  Google Scholar 

  • Dupuis DJ (2007) Using copulas in hydrology: benefits, cautions, and issues. J Hydrol Eng 12(4):381–393

    Article  Google Scholar 

  • Evin G, Favre AC (2008) A new rainfall model based on the Neyman–Scott process using cubic copulas. Water Resour Res 44(3):W03433. doi:10.1029/2007WR006054

    Article  Google Scholar 

  • Favre AC, El Adlouni S, Perreault L, Thiémonge N, Bobée B (2004) Multivariate hydrological frequency analysis using copulas. Water Resour Res 40(1):W01101. doi:10.1029/2003WR002456

    Article  Google Scholar 

  • Gebremichael M, Krajewski WF (2007) Application of copulas to modeling temporal sampling errors in satellite-derived rainfall estimates. J Hydrol Eng 12(4):404–408

    Article  Google Scholar 

  • Genest C, Favre AC (2007) Everything you always wanted to know about copula modeling but were afraid to ask. J Hydrol Eng 12(4):347–367

    Article  Google Scholar 

  • Genest C, Rémillard B, Beaudoin D (2007a) Goodness-of-fit tests for copulas: a review and a power study. Insur Math Econ. doi:10.1016/j.insmatheco.2007.10.005

  • Genest C, Favre AC, Béliveau J, Jacques C (2007b) Metaelliptical copulas and their use in frequency analysis of multivariate hydrological data. Water Resour Res 43:W09401. doi:10.1029/2006WR005275

    Article  Google Scholar 

  • Goel NK, Seth SM (1998) Multivariate modeling of flood flows. J Hydraul Eng 124(2):146–155

    Article  Google Scholar 

  • González J, Valdés JB (2003) Bivariate drought recurrence analysis using tree ring reconstructions. J Hydrol Eng 8(5):247–258

    Article  Google Scholar 

  • Grimaldi S, Serinaldi F (2006) Asymmetric copula in multivariate flood frequency analysis. Adv Water Resour 29(8):1155–1167

    Article  Google Scholar 

  • Gringorten II (1963) A plotting rule for extreme probability paper. J Geophys Res 68(3):813–814

    Article  Google Scholar 

  • Gumbel EJ (1958) Statistics of extremes. Columbia University Press, New York

    Google Scholar 

  • Kao SC, Govindaraju RS (2008) Trivariate statistical analysis of extreme rainfall events via the Plackett family of copulas. Water Resour Res 44(2):W02415. doi:10.1029/2007WR006261

    Article  Google Scholar 

  • Kao SC, Govindaraju RS (2009) A copula-base joint deficit index for droughts. J Hydrol. doi:10.1016/j.jhydrol.2009.10.029

  • Kim TW, Valdés JB, Yoo C (2006) Nonparametric approach for bivariate drought characterization using Palmer Drought Index. J Hydrol Eng 11(2):134–143

    Article  CAS  Google Scholar 

  • Laux P, Wagner S, Wagner A, Jacobeit J, Bardossy A, Kunstmann H (2009) Modelling daily precipitation features in the Volta Basin of West Africa. Int J Climatol 29(7):937–954

    Article  Google Scholar 

  • Mardia KV (1967) Some contributions to contingency-type bivariate distributions. Biometrika 54(1–2):235–249

    CAS  Google Scholar 

  • McCuen RH (2002) Modeling hydrologic change: statistical methods. Taylor & Francis, Boca Raton

    Book  Google Scholar 

  • Mishra AK, Singh VP, Desai VR (2008) Drought characterization: a probabilistic approach. Stoch Environ Res Risk Assess. doi:10.1007/s00477-007-0194-2

  • Molenberghs G, Lesaffre E (1994) Marginal modeling of correlated ordinal data using a multivariate Plackett distribution. J Am Stat Assoc 89:633–644

    Article  Google Scholar 

  • Nadarajah S (2007) A bivariate gamma model for drought. Water Resour Res 43(8):W08501. doi:10.1029/2006WR005641

    Article  Google Scholar 

  • Nelsen RB (1999) An introduction to copulas. Springer, New York

    Google Scholar 

  • O’Reilly UM, Yu T, Riolo R, Worzel B (2005) Genetic programming theory and practice II. Springer, Boston

    Book  Google Scholar 

  • Park WJ, Kim YG (1992) Goodness-of-fit tests for the power-law process. IEEE Trans Reliab 41(1):107–111

    Article  Google Scholar 

  • Plackett RL (1965) A class of bivariate distributions. J Am Stat Assoc 60(310):516–522

    Article  Google Scholar 

  • Poulin A, Huard D, Favre AC, Pugin S (2007) Importance of tail dependence in bivariate frequency analysis. J Hydrol Eng 12(4):394–403

    Article  Google Scholar 

  • Rossi G, Cancelliere A, Pereira L, Oweis T, Shatanawi M, Zairi A (2003) Tools for drought mitigation in Mediterranean regions. Kluwer, Dordrecht

    Google Scholar 

  • Rossi G, Vega T, Bonaccorso B (2007) Methods and tools for drought analysis and management. Springer, Dordrecht

    Book  Google Scholar 

  • Salvadori G, De Michele C (2004) Frequency analysis via copulas: theoretical aspects and applications to hydrological events. Water Resour Res 40(12):W12511. doi:10.1029/2004WR003133

    Article  Google Scholar 

  • Salvadori G, De Michele C (2007) On the use of copulas in hydrology: theory and practice. J Hydrol Eng 12(4):369–380

    Article  Google Scholar 

  • Salvadori G, De Michele C, Kottegoda NT, Rosso R (2007) Extremes in nature: an approach using copulas. Springer, Dordrecht

    Google Scholar 

  • Serinaldi F, Grimaldi S (2007) Fully nested 3-copula: procedure and application on hydrological data. J Hydrol Eng 12(4):420–430

    Article  Google Scholar 

  • Serinaldi F, Bonaccorso B, Cancelliere A, Grimaldi S (2009) Probabilistic characterization of drought properties through copulas. J Phys Chem Earth 34(10–12):596–605

    Google Scholar 

  • Shiau JT (2003) Return period of bivariate distributed hydrological events. Stoch Env Res Risk Assess 17(1–2):42–57

    Article  Google Scholar 

  • Shiau JT (2006) Fitting drought duration and severity with two-dimensional copulas. Water Resour Manag 20:795–815

    Article  Google Scholar 

  • Shiau JT, Modarres R (2009) Copula-based drought severity-duration-frequency analysis in Iran. Meteorol Appl. doi:10.1002/met.145

  • Shiau JT, Shen HW (2001) Recurrence analysis of hydrologic droughts of differing severity. J Water Resour Plan Manag 127(1):30–40

    Article  Google Scholar 

  • Shiau JT, Wang HY, Chang TT (2006) Bivariate frequency analysis of floods using copulas. J AmWater Resour Assoc 42(6):1549–1564

    Article  Google Scholar 

  • Shiau JT, Feng S, Nadarajah S (2007) Assessment of hydrological droughts for the Yellow River, China, using copulas. Hydrol Process 21(16):2157–2163

    Article  Google Scholar 

  • Shimokawa T, Liao M (1999) Goodness-of-fit tests for type-I extreme-value and 2-parameter Weibull distribution. IEEE Trans Reliab 48(1):79–86

    Article  Google Scholar 

  • Singh K, Singh VP (1991) Derivation of bivariate probability density functions with exponential marginals. Stoch Hydrol Hydraul 5(1):55–68

    Article  Google Scholar 

  • Singh VP, Zhang L (2007) IDF curves using the Frank Archimedean copula. J Hydrol Eng 12(6):651–662

    Article  Google Scholar 

  • Sklar A (1959) Fonctions de repartition án dimensions et leurs marges. Publ Inst Stat Univ Paris 8:229–231

    Google Scholar 

  • Song SB, Singh VP (2009) Meta-elliptical copulas for drought frequency analysis of periodic hydrologic data. Stoch Environ Res Risk Assess. doi:10.1007/s00477-009-0331-1

  • Tayfur G, Moramarco T (2008) Predicting hourly-based flow discharge hydrographs from level data using genetic algorithms. J Hydrol 352(1–2):77–93

    Article  Google Scholar 

  • Vose MD (1999) The simple genetic algorithm: foundations and theory. MIT Press, Cambridge

    Google Scholar 

  • Wong G, Lambert MF, Metcalfe AV (2008) Trivariate copulas for characterization of droughts. ANZIAM J 49:306–323

    Google Scholar 

  • Wong G, Lambert MF, Leonard M, Metcalfe AV (2009) Drought analysis using trivariate copulas conditional on climate states. J Hydrol Eng. doi:10.1061/(ASCE)HE.1943-5584.0000169

  • Yue S (1999) Applying bivariate normal distribution to flood frequency analysis. Water Int 24(3):248–254

    Article  Google Scholar 

  • Yue S (2000a) Joint probability distribution of annual maximum storm peaks and amounts as represented by daily rainfalls. Hydrol Sci J 45(2):315–326

    Article  Google Scholar 

  • Yue S (2000b) The Gumbel logistic model for representing a multivariate storm event. Adv Water Resour 24(2):179–185

    Article  Google Scholar 

  • Yue S (2000c) The Gumbel mixed model applied to storm frequency analysis. Water Resour Manag 14(5):377–389

    Article  Google Scholar 

  • Yue S, Rasmussen P (2002) Bivariate frequency analysis: discussion of some useful concepts for hydrological application. Hydrol Process 16(14):811–819

    Article  Google Scholar 

  • Yue S, Ouarda TBMJ, Bobée B, Legendre P, Bruneau P (1999) The Gumbel mixed model for flood frequency analysis. J Hydrol 226:88–100

    Article  Google Scholar 

  • Yue S, Ouarda TBMJ, Bobée B (2001) A review of bivariate gamma distributions for hydrological application. J Hydrol 246:1–18

    Article  Google Scholar 

  • Zhang L, Singh VP (2006) Bivariate flood frequency analysis using the copula method. J Hydrol Eng 11(2):150–164

    Article  Google Scholar 

  • Zhang L, Singh VP (2007a) Gumbel–Hougaard copula for trivariate rainfall frequency analysis. J Hydrol Eng 12(4):409–419

    Article  Google Scholar 

  • Zhang L, Singh VP (2007b) Bivariate rainfall frequency distributions using Archimedean copulas. J Hydrol 332:93–109

    Article  Google Scholar 

  • Zhang L, Singh VP (2007c) Trivariate flood frequency analysis using the Gumbel–Hougaard copula. J Hydrol Eng 12(4):431–439

    Article  Google Scholar 

Download references

Acknowledgment

This work was financially supported by the National Science Council, Republic of China (Grant No. NSC-50879070 and NSC-50579065).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Songbai Song.

Appendices

Appendix 1

The method for estimating the trivariate cross-product ratio entails the following steps:

  1. (1)

    Compute C UV , C VW , and C UW using Eq. 3.

  2. (2)

    For computing C UVW , Eq. 9 can be rewritten as

    $$ \begin{aligned} \left( {\psi_{UVW} - 1} \right)z^{4} & + \left[ { - \psi_{UVW} \left( {a_{1} + a_{2} + a_{3} + a_{4} } \right) + \left( {b_{1} + b_{2} + b_{3} } \right)} \right]z^{3} \\ & + \left\{ {\psi_{UVW} \left[ {a_{1} a_{2} + \left( {a_{1} + a_{2} } \right)\left( {a_{3} + a_{4} } \right) + a_{3} a_{4} } \right] - \left[ {b_{1} b_{2} + b_{3} \left( {b_{1} + b_{2} } \right)} \right]} \right\}z^{2} \\ & + \left\{ { - \psi_{UVW} \left[ {a_{1} a_{2} \left( {a_{3} + a_{4} } \right) + a_{3} a_{4} \left( {a_{1} + a_{2} } \right)} \right] + b_{1} b_{2} b_{3} } \right\}z + \psi_{UVW} a_{1} a_{2} a_{3} a_{4} = 0 \\ \end{aligned} $$
    (37)

    Let f(z) represent the left side of Eq. 37. Newton’s iterative estimating method of z can be expressed as:

    $$ z_{n + 1} = z_{n} - {\frac{{f(z_{n} )}}{{f^{\prime}(z_{n} )}}} $$
    (38)

    where f (z) is the first derivative of f(z) with respect to z; z n and z n+1 are the nth and (n + 1)th iteratively computed values of z. Then, C UVW can be obtained by Newton’s iteration method.

  3. (3)

    Compute P 000, P 010, P 100, P 011, P 110, P 101, P 011, P 111 using Eq. 8.

  4. (4)

    Compute \( {\frac{{\partial C_{UV} }}{\partial u}},\;{\frac{{\partial C_{UV} }}{\partial v}},\;{\frac{{\partial C_{UW} }}{\partial u}},\;{\frac{{\partial C_{UW} }}{\partial w}},\;{\frac{{\partial C_{VW} }}{\partial v}},\;{\text{and}}\;{\frac{{\partial C_{VW} }}{\partial w}} \) using Eqs. 5 and 6, respectively.

  5. (5)

    The first derivative of P 000, P 010, P 100, P 011, P 110, P 101, P 011, P 111 with respect to u, v, w is expressed, respectively, as:

    $$ \begin{gathered} {\frac{{\partial P_{000} }}{\partial u}} = - {\frac{{\partial P_{100} }}{\partial u}} = {\frac{{\partial C_{UVW} }}{\partial u}} \hfill \\ {\frac{{\partial P_{010} }}{\partial u}} = - {\frac{{\partial P_{110} }}{\partial u}} = - {\frac{{\partial C_{UVW} }}{\partial u}} + {\frac{{\partial C_{UW} }}{\partial u}} \hfill \\ {\frac{{\partial P_{001} }}{\partial u}} = - {\frac{{\partial P_{101} }}{\partial u}} = - {\frac{{\partial C_{UVW} }}{\partial u}} + {\frac{{\partial C_{UV} }}{\partial u}} \hfill \\ {\frac{{\partial P_{011} }}{\partial u}} = - {\frac{{\partial P_{111} }}{\partial u}} = {\frac{{\partial C_{UVW} }}{\partial u}} - {\frac{{\partial C_{UV} }}{\partial u}} - {\frac{{\partial C_{UW} }}{\partial u}} + 1 \hfill \\ \end{gathered} $$
    (39)
    $$ \begin{gathered} {\frac{{\partial P_{000} }}{\partial v}} = - {\frac{{\partial P_{010} }}{\partial v}} = {\frac{{\partial C_{UVW} }}{\partial v}} \hfill \\ {\frac{{\partial P_{100} }}{\partial v}} = - {\frac{{\partial P_{110} }}{\partial v}} = - {\frac{{\partial C_{UVW} }}{\partial v}} + {\frac{{\partial C_{VW} }}{\partial v}} \hfill \\ {\frac{{\partial P_{001} }}{\partial v}} = - {\frac{{\partial P_{011} }}{\partial v}} = - {\frac{{\partial C_{UVW} }}{\partial v}} + {\frac{{\partial C_{UV} }}{\partial v}} \hfill \\ {\frac{{\partial P_{101} }}{\partial v}} = - {\frac{{\partial P_{111} }}{\partial v}} = {\frac{{\partial C_{UVW} }}{\partial v}} - {\frac{{\partial C_{UV} }}{\partial v}} - {\frac{{\partial C_{VW} }}{\partial v}} + 1 \hfill \\ \end{gathered} $$
    (40)
    $$ \begin{gathered} {\frac{{\partial P_{000} }}{\partial w}} = - {\frac{{\partial P_{001} }}{\partial w}} = {\frac{{\partial C_{UVW} }}{\partial w}} \hfill \\ {\frac{{\partial P_{100} }}{\partial w}} = - {\frac{{\partial P_{101} }}{\partial w}} = - {\frac{{\partial C_{UVW} }}{\partial w}} + {\frac{{\partial C_{VW} }}{\partial w}} \hfill \\ {\frac{{\partial P_{010} }}{\partial w}} = - {\frac{{\partial P_{011} }}{\partial w}} = - {\frac{{\partial C_{UVW} }}{\partial w}} + {\frac{{\partial C_{UW} }}{\partial w}} \hfill \\ {\frac{{\partial P_{110} }}{\partial w}} = - {\frac{{\partial P_{111} }}{\partial w}} = {\frac{{\partial C_{UVW} }}{\partial u}} - {\frac{{\partial C_{UW} }}{\partial w}} - {\frac{{\partial C_{VW} }}{\partial w}} + 1 \hfill \\ \end{gathered} $$
    (41)
  6. (6)

    Compute \( {\frac{{\partial C_{UVW} }}{\partial u}},\;{\frac{{\partial C_{UVW} }}{\partial v}} \) and \( {\frac{{\partial C_{UVW} }}{\partial w}} \) as:

    $$ \begin{aligned} & {\frac{{\partial P_{000} }}{\partial u}}P_{011} P_{101} P_{110} + P_{000}\, {\frac{{\partial P_{011} }}{\partial u}}P_{101} P_{110} + P_{000} P_{011}\, {\frac{{\partial P_{101} }}{\partial u}}P_{110} + P_{000} P_{011} P_{101} \,{\frac{{\partial P_{110} }}{\partial u}} \\ & - \psi_{UVW} \left( {{\frac{{\partial P_{111} }}{\partial u}}P_{100} P_{010} P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial u}}P_{010} P_{001} + P_{111} P_{100}\, {\frac{{\partial P_{010} }}{\partial u}}P_{001} + P_{111} P_{100} P_{010}\, {\frac{{\partial P_{001} }}{\partial u}}} \right) = 0 \\ \end{aligned} $$
    (42)
    $$ \begin{aligned} & {\frac{{\partial P_{000} }}{\partial v}}P_{011} P_{101} P_{110} + P_{000}\, {\frac{{\partial P_{011} }}{\partial v}}P_{101} P_{110} + P_{000} P_{011} \,{\frac{{\partial P_{101} }}{\partial v}}P_{110} + P_{000} P_{011} P_{101} \,{\frac{{\partial P_{110} }}{\partial v}} \\ & - \psi_{UVW} \left( {{\frac{{\partial P_{111} }}{\partial v}}P_{100} P_{010} P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial v}}P_{010} P_{001} + P_{111} P_{100} \,{\frac{{\partial P_{010} }}{\partial v}}P_{001} + P_{111} P_{100} P_{010} \,{\frac{{\partial P_{001} }}{\partial v}}} \right) = 0 \\ \end{aligned} $$
    (43)
    $$ \begin{aligned} & {\frac{{\partial P_{000} }}{\partial w}}P_{011} P_{101} P_{110} + P_{000}\, {\frac{{\partial P_{011} }}{\partial w}}P_{101} P_{110} + P_{000} P_{011} \,{\frac{{\partial P_{101} }}{\partial w}}P_{110} + P_{000} P_{011} P_{101} \,{\frac{{\partial P_{110} }}{\partial w}} \\ & - \psi_{UVW} \left( {{\frac{{\partial P_{111} }}{\partial w}}P_{100} P_{010} P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial w}}P_{010} P_{001} + P_{111} P_{100} \,{\frac{{\partial P_{010} }}{\partial w}}P_{001} + P_{111} P_{100} P_{010} \,{\frac{{\partial P_{001} }}{\partial w}}} \right) = 0 \\ \end{aligned} $$
    (44)
  7. (7)

    Compute the first derivative of P 000, P 010, P 100, P 011, P 110 P 101, P 011, P 111 with respect to u, v, w using Eqs. 39, 40 and 41, respectively.

  8. (8)

    Compute the probability density of a bivariate Plackett copula \( {\frac{{\partial^{2} C_{UV} }}{\partial u\partial v}},\;{\frac{{\partial^{2} C_{VW} }}{\partial v\partial w}} \) and \( {\frac{{\partial^{2} C_{UW} }}{\partial u\partial w}} \) using Eq. 4, respectively.

  9. (9)

    Compute second-order derivatives of P 000, P 010, P 100, P 011, P 110 P 101, P 011, P 111 with respect to u, v, w as:

    $$ \begin{gathered} {\frac{{\partial^{2} P_{000} }}{\partial u\partial v}} = - {\frac{{\partial^{2} P_{100} }}{\partial u\partial v}} = - {\frac{{\partial^{2} P_{010} }}{\partial u\partial v}} = {\frac{{\partial^{2} P_{110} }}{\partial u\partial v}} = {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial v}} \hfill \\ {\frac{{\partial^{2} P_{001} }}{\partial u\partial v}} = - {\frac{{\partial^{2} P_{101} }}{\partial u\partial v}} = - {\frac{{\partial^{2} P_{011} }}{\partial u\partial v}} = {\frac{{\partial^{2} P_{111} }}{\partial u\partial v}} = - {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial v}} + {\frac{{\partial^{2} C_{UV} }}{\partial u\partial v}} \hfill \\ \end{gathered} $$
    (45)
    $$ \begin{gathered} {\frac{{\partial^{2} P_{000} }}{\partial u\partial w}} = - {\frac{{\partial^{2} P_{100} }}{\partial u\partial w}} = - {\frac{{\partial^{2} P_{001} }}{\partial u\partial w}} = {\frac{{\partial^{2} P_{101} }}{\partial u\partial w}} = {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial w}} \hfill \\ {\frac{{\partial^{2} P_{010} }}{\partial u\partial w}} = - {\frac{{\partial^{2} P_{110} }}{\partial u\partial w}} = - {\frac{{\partial^{2} P_{011} }}{\partial u\partial w}} = {\frac{{\partial^{2} P_{111} }}{\partial u\partial w}} = - {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial w}} + {\frac{{\partial^{2} C_{UV} }}{\partial u\partial w}} \hfill \\ \end{gathered} $$
    (46)
    $$ \begin{gathered} {\frac{{\partial^{2} P_{000} }}{\partial v\partial w}} = - {\frac{{\partial^{2} P_{010} }}{\partial v\partial w}} = - {\frac{{\partial^{2} P_{001} }}{\partial v\partial w}} = {\frac{{\partial^{2} P_{011} }}{\partial v\partial w}} = {\frac{{\partial^{2} C_{UVW} }}{\partial v\partial w}} \hfill \\ {\frac{{\partial^{2} P_{100} }}{\partial v\partial w}} = - {\frac{{\partial^{2} P_{110} }}{\partial v\partial w}} = - {\frac{{\partial^{2} P_{101} }}{\partial v\partial w}} = {\frac{{\partial^{2} P_{111} }}{\partial v\partial w}} = - {\frac{{\partial^{2} C_{UVW} }}{\partial v\partial w}} + {\frac{{\partial^{2} C_{UV} }}{\partial v\partial w}} \hfill \\ \end{gathered} $$
    (47)
  10. (8)

    Compute \( {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial v}},\;{\frac{{\partial^{2} C_{UVW} }}{\partial v\partial w}} \) and \( {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial w}}. \)

    \( {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial v}} \) can be obtained as:

    $$ \begin{aligned} {\frac{{\partial^{2} P_{000} }}{\partial u\partial v}}P_{011} P_{101} P_{110} & + {\frac{{\partial P_{000} }}{\partial u}}\,{\frac{{\partial P_{011} }}{\partial v}}P_{101} P_{110} + {\frac{{\partial P_{000} }}{\partial u}}P_{011} \,{\frac{{\partial P_{101} }}{\partial v}}P_{110} + {\frac{{\partial P_{000} }}{\partial u}}P_{011} P_{101}\, {\frac{{\partial P_{110} }}{\partial v}} \\ & + {\frac{{\partial P_{000} }}{\partial v}}\,{\frac{{\partial P_{011} }}{\partial u}}P_{101} P_{110} + P_{000} \,{\frac{{\partial^{2} P_{011} }}{\partial u\partial v}}P_{101} P_{110} + P_{000}\, {\frac{{\partial P_{011} }}{\partial u}}\,{\frac{{\partial P_{101} }}{\partial v}}P_{110} + P_{000} \,{\frac{{\partial P_{011} }}{\partial u}}P_{101}\, {\frac{{\partial P_{110} }}{\partial v}} \\ & + {\frac{{\partial P_{000} }}{\partial v}}P_{011} \,{\frac{{\partial P_{101} }}{\partial u}}P_{110} + P_{000} \,{\frac{{\partial P_{011} }}{\partial v}}\,{\frac{{\partial P_{101} }}{\partial u}}P_{110} + P_{000} P_{011} \,{\frac{{\partial^{2} P_{101} }}{\partial u\partial v}}P_{110} + P_{000} P_{011} \,{\frac{{\partial P_{101} }}{\partial u}}\,{\frac{{\partial P_{110} }}{\partial v}} \\ & + {\frac{{\partial P_{000} }}{\partial v}}P_{011} P_{101} \,{\frac{{\partial P_{110} }}{\partial u}} + P_{000} \,{\frac{{\partial P_{011} }}{\partial v}}P_{101} \,{\frac{{\partial P_{110} }}{\partial u}} + P_{000} P_{011} \,{\frac{{\partial P_{101} }}{\partial v}}\,{\frac{{\partial P_{110} }}{\partial u}} + P_{000} P_{011} P_{101} \,{\frac{{\partial^{2} P_{110} }}{\partial u\partial v}} \\ & - \psi_{UVW} \left( {{\frac{{\partial^{2} P_{111} }}{\partial u\partial v}}P_{100} P_{010} P_{001} + {\frac{{\partial P_{111} }}{\partial u}}\,{\frac{{\partial P_{100} }}{\partial v}}P_{010} P_{001} + {\frac{{\partial P_{111} }}{\partial u}}P_{100} \,{\frac{{\partial P_{010} }}{\partial v}}P_{001} + {\frac{{\partial P_{111} }}{\partial u}}P_{100} P_{010} \,{\frac{{\partial P_{001} }}{\partial v}}} \right. \\ & + {\frac{{\partial P_{111} }}{\partial v}}\,{\frac{{\partial P_{100} }}{\partial u}}P_{010} P_{001} + P_{111} \,{\frac{{\partial^{2} P_{100} }}{\partial u\partial v}}P_{010} P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial u}}\,{\frac{{\partial P_{010} }}{\partial v}}P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial u}}P_{010} \,{\frac{{\partial P_{001} }}{\partial v}} \\ & + {\frac{{\partial P_{111} }}{\partial v}}P_{100} \,{\frac{{\partial P_{010} }}{\partial u}}P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial v}}\,{\frac{{\partial P_{010} }}{\partial u}}P_{001} + P_{111} P_{100} \,{\frac{{\partial^{2} P_{010} }}{\partial u\partial v}}P_{001} + P_{111} P_{100} \,{\frac{{\partial P_{010} }}{\partial u}}\,{\frac{{\partial P_{001} }}{\partial v}} \\ & \left. { + {\frac{{\partial P_{111} }}{\partial v}}P_{100} P_{010} \,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial v}}P_{010} \,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} P_{100} \,{\frac{{\partial P_{010} }}{\partial v}}\,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} P_{100} P_{010} \,{\frac{{\partial^{2} P_{001} }}{\partial u\partial v}}} \right) = 0 \\ \end{aligned} $$
    (48)

    Similarly, applying the \( {\frac{\partial }{\partial w}} \) operation to Eqs. 42 and 43, respectively, we can obtain \( {\frac{{\partial^{2} C_{UVW} }}{\partial v\partial w}} \) and \( {\frac{{\partial^{2} C_{UVW} }}{\partial u\partial w}} \).

  11. (9)

    Compute the probability density \( {\frac{{\partial^{3} C_{UVW} }}{\partial u\partial v\partial w}} \) of the trivariate Plackett copula.

    Applying the \( {\frac{\partial }{\partial w}} \) operation to Eq. 45, we can obtain the third-order derivatives as:

    $$ \begin{gathered} {\frac{{\partial^{3} P_{000} }}{\partial u\partial v\partial w}} = - {\frac{{\partial^{3} P_{100} }}{\partial u\partial v\partial w}} = - {\frac{{\partial^{3} P_{010} }}{\partial u\partial v\partial w}} = {\frac{{\partial^{3} P_{110} }}{\partial u\partial v\partial w}} = {\frac{{\partial^{3} C_{UVW} }}{\partial u\partial v\partial w}} \hfill \\ {\frac{{\partial^{3} P_{001} }}{\partial u\partial v\partial w}} = - {\frac{{\partial^{3} P_{101} }}{\partial u\partial v\partial w}} = - {\frac{{\partial^{3} P_{011} }}{\partial u\partial v\partial w}} = {\frac{{\partial^{3} P_{111} }}{\partial u\partial v\partial w}} = - {\frac{{\partial^{3} C_{UVW} }}{\partial u\partial v\partial w}} + {\frac{{\partial^{3} C_{UV} }}{\partial u\partial v\partial w}} \hfill \\ \end{gathered} $$
    (49)

    Applying \( {\frac{\partial }{\partial w}} \) operation to Eq. 48, a new equation is obtained. Because the new equation is too long which contains 128 terms (see Appendix 2), it is omitted here. Substituting Eq. 49 into the new equation, \( {\frac{{\partial^{3} C_{UVW} }}{\partial u\partial v\partial w}} \) is the only unknown and one can then obtain the final formula for computation.

Appendix 2: Third order derivative

The final formula of \( {\frac{{\partial^{3} C_{UVW} }}{\partial u\partial v\partial w}} \) is expressed as

$$ \begin{aligned} {\frac{{\partial^{3} P_{000} }}{\partial u\partial v\partial w}}P_{011} P_{101} P_{110} & + {\frac{{\partial^{2} P_{000} }}{\partial u\partial v}}\,{\frac{{\partial P_{011} }}{\partial w}}P_{101} P_{110} + {\frac{{\partial^{2} P_{000} }}{\partial u\partial v}}P_{011} \,{\frac{{\partial P_{101} }}{\partial w}}P_{110} + {\frac{{\partial^{2} P_{000} }}{\partial u\partial v}}P_{011} P_{101} \,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial^{2} P_{000} }}{\partial u\partial w}}\,{\frac{{\partial P_{011} }}{\partial v}}P_{101} P_{110} + {\frac{{\partial P_{000} }}{\partial u}}\,{\frac{{\partial^{2} P_{011} }}{\partial v\partial w}}P_{101} P_{110} + {\frac{{\partial P_{000} }}{\partial u}}\,{\frac{{\partial P_{011} }}{\partial v}}\,{\frac{{\partial P_{101} }}{\partial w}}P_{110} + {\frac{{\partial P_{000} }}{\partial u}}\,{\frac{{\partial P_{011} }}{\partial v}}P_{101} \,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial^{2} P_{000} }}{\partial u\partial w}}P_{011} \,{\frac{{\partial P_{101} }}{\partial v}}P_{110} + {\frac{{\partial P_{000} }}{\partial u}}\,{\frac{{\partial P_{011} }}{\partial w}}\,{\frac{{\partial P_{101} }}{\partial v}}P_{110} + {\frac{{\partial P_{000} }}{\partial u}}P_{011} \,{\frac{{\partial^{2} P_{101} }}{\partial v\partial w}}P_{110} + {\frac{{\partial P_{000} }}{\partial u}}P_{011} \,{\frac{{\partial P_{101} }}{\partial v}}\,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial^{2} P_{000} }}{\partial u\partial w}}P_{011} P_{101} \,{\frac{{\partial P_{110} }}{\partial v}} + {\frac{{\partial P_{000} }}{\partial u}}\,{\frac{{\partial P_{011} }}{\partial w}}P_{101} \,{\frac{{\partial P_{110} }}{\partial v}} + {\frac{{\partial P_{000} }}{\partial u}}P_{011} \,{\frac{{\partial P_{101} }}{\partial w}}\,{\frac{{\partial P_{110} }}{\partial v}} + {\frac{{\partial P_{000} }}{\partial u}}P_{011} P_{101}\,{\frac{{\partial^{2} P_{110} }}{\partial v\partial w}} \\ & + {\frac{{\partial^{2} P_{000} }}{\partial v\partial w}}\,{\frac{{\partial P_{011} }}{\partial u}}P_{101} P_{110} + {\frac{{\partial P_{000} }}{\partial v}}\,{\frac{{\partial^{2} P_{011} }}{\partial u\partial w}}P_{101} P_{110} + {\frac{{\partial P_{000} }}{\partial v}}\,{\frac{{\partial P_{011} }}{\partial u}}\,{\frac{{\partial P_{101} }}{\partial w}}P_{110} + {\frac{{\partial P_{000} }}{\partial v}}\,{\frac{{\partial P_{011} }}{\partial u}}P_{101}\,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}\,{\frac{{\partial^{2} P_{011} }}{\partial u\partial v}}P_{101} P_{110} + P_{000}\,{\frac{{\partial^{3} P_{011} }}{\partial u\partial v\partial w}}P_{101} P_{110} + P_{000}\,{\frac{{\partial^{2} P_{011} }}{\partial u\partial v}}\,{\frac{{\partial P_{101} }}{\partial w}}P_{110} + P_{000}\,{\frac{{\partial^{2} P_{011} }}{\partial u\partial v}}P_{101}\,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}\,{\frac{{\partial P_{011} }}{\partial u}}\,{\frac{{\partial P_{101} }}{\partial v}}P_{110} + P_{000}\,{\frac{{\partial^{2} P_{011} }}{\partial u\partial w}}\,{\frac{{\partial P_{101} }}{\partial v}}P_{110} + P_{000}\,{\frac{{\partial P_{011} }}{\partial u}}\,{\frac{{\partial^{2} P_{101} }}{\partial v\partial w}}P_{110} + P_{000}\,{\frac{{\partial P_{011} }}{\partial u}}\,{\frac{{\partial P_{101} }}{\partial v}}\,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}\,{\frac{{\partial P_{011} }}{\partial u}}P_{101}\,{\frac{{\partial P_{110} }}{\partial v}} + P_{000}\,{\frac{{\partial^{2} P_{011} }}{\partial u\partial w}}P_{101}\,{\frac{{\partial P_{110} }}{\partial v}} + P_{000}\,{\frac{{\partial P_{011} }}{\partial u}}\,{\frac{{\partial P_{101} }}{\partial w}}\,{\frac{{\partial P_{110} }}{\partial v}} + P_{000}\,{\frac{{\partial P_{011} }}{\partial u}}P_{101}\,{\frac{{\partial^{2} P_{110} }}{\partial v\partial w}} \\ & + {\frac{{\partial^{2} P_{000} }}{\partial v\partial w}}P_{011}\,{\frac{{\partial P_{101} }}{\partial u}}P_{110} + {\frac{{\partial P_{000} }}{\partial v}}\,{\frac{{\partial P_{011} }}{\partial w}}\,{\frac{{\partial P_{101} }}{\partial u}}P_{110} + {\frac{{\partial P_{000} }}{\partial v}}P_{011}\,{\frac{{\partial^{2} P_{101} }}{\partial u\partial w}}P_{110} + {\frac{{\partial P_{000} }}{\partial v}}P_{011}\,{\frac{{\partial P_{101} }}{\partial u}}\,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}\,{\frac{{\partial P_{011} }}{\partial v}}\,{\frac{{\partial P_{101} }}{\partial u}}P_{110} + P_{000}\,{\frac{{\partial^{2} P_{011} }}{\partial v\partial w}}\,{\frac{{\partial P_{101} }}{\partial u}}P_{110} + P_{000}\,{\frac{{\partial P_{011} }}{\partial v}}\,{\frac{{\partial^{2} P_{101} }}{\partial u\partial v}}P_{110} + P_{000}\,{\frac{{\partial P_{011} }}{\partial v}}\,{\frac{{\partial P_{101} }}{\partial u}}\,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}P_{011}\,{\frac{{\partial^{2} P_{101} }}{\partial u\partial v}}P_{110} + P_{000}\,{\frac{{\partial P_{011} }}{\partial w}}\,{\frac{{\partial^{2} P_{101} }}{\partial u\partial w}}P_{110} + P_{000} P_{011}\,{\frac{{\partial^{3} P_{101} }}{\partial u\partial v\partial w}}P_{110} + P_{000} P_{011}\,{\frac{{\partial^{2} P_{101} }}{\partial u\partial w}}\,{\frac{{\partial P_{110} }}{\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}P_{011}\,{\frac{{\partial P_{101} }}{\partial u}}\,{\frac{{\partial P_{110} }}{\partial v}} + P_{000}\,{\frac{{\partial P_{011} }}{\partial w}}\,{\frac{{\partial P_{101} }}{\partial u}}\,{\frac{{\partial P_{110} }}{\partial v}} + P_{000} P_{011}\,{\frac{{\partial^{2} P_{101} }}{\partial u\partial w}}\,{\frac{{\partial P_{110} }}{\partial v}} + P_{000} P_{011}\,{\frac{{\partial P_{101} }}{\partial u}}\,{\frac{{\partial^{2} P_{110} }}{\partial v\partial w}} \\ & + {\frac{{\partial^{2} P_{000} }}{\partial v\partial w}}P_{011} P_{101}\,{\frac{{\partial P_{110} }}{\partial u}} + {\frac{{\partial P_{000} }}{\partial v}}\,{\frac{{\partial P_{011} }}{\partial w}}P_{101}\,{\frac{{\partial P_{110} }}{\partial u}} + {\frac{{\partial P_{000} }}{\partial v}}P_{011}\,{\frac{{\partial P_{101} }}{\partial w}}\,{\frac{{\partial P_{110} }}{\partial u}} + {\frac{{\partial P_{000} }}{\partial v}}P_{011} P_{101}\,{\frac{{\partial^{2} P_{110} }}{\partial u\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}\,{\frac{{\partial P_{011} }}{\partial v}}P_{101}\,{\frac{{\partial P_{110} }}{\partial u}} + P_{000}\,{\frac{{\partial^{2} P_{011} }}{\partial v\partial w}}P_{101}\,{\frac{{\partial P_{110} }}{\partial u}} + P_{000}\,{\frac{{\partial P_{011} }}{\partial v}}\,{\frac{{\partial P_{101} }}{\partial w}}\,{\frac{{\partial P_{110} }}{\partial u}} + P_{000}\,{\frac{{\partial P_{011} }}{\partial v}}P_{101}\,{\frac{{\partial^{2} P_{110} }}{\partial u\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}P_{011}\,{\frac{{\partial P_{101} }}{\partial v}}\,{\frac{{\partial P_{110} }}{\partial u}} + P_{000}\,{\frac{{\partial P_{011} }}{\partial w}}\,{\frac{{\partial P_{101} }}{\partial v}}\,{\frac{{\partial P_{110} }}{\partial u}} + P_{000} P_{011}\,{\frac{{\partial^{2} P_{101} }}{\partial v\partial w}}\,{\frac{{\partial P_{110} }}{\partial u}} + P_{000} P_{011}\,{\frac{{\partial P_{101} }}{\partial v}}\,{\frac{{\partial^{2} P_{110} }}{\partial u\partial w}} \\ & + {\frac{{\partial P_{000} }}{\partial w}}P_{011} P_{101}\,{\frac{{\partial^{2} P_{110} }}{\partial u\partial v}} + P_{000}\,{\frac{{\partial P_{011} }}{\partial w}}P_{101}\,{\frac{{\partial^{2} P_{110} }}{\partial u\partial v}} + P_{000} P_{011}\,{\frac{{\partial P_{101} }}{\partial w}}\,{\frac{{\partial^{2} P_{110} }}{\partial u\partial v}} + P_{000} P_{011} P_{101}\,{\frac{{\partial^{3} P_{110} }}{\partial u\partial v\partial w}} \\ & - \psi_{UVW} \left( {{\frac{{\partial^{3} P_{111} }}{\partial u\partial v\partial w}}P_{100} P_{010} P_{001} + {\frac{{\partial^{2} P_{111} }}{\partial u\partial v}}\,{\frac{{\partial P_{100} }}{\partial w}}P_{010} P_{001} + {\frac{{\partial^{2} P_{111} }}{\partial u\partial v}}P_{100}\,{\frac{{\partial P_{010} }}{\partial w}}P_{001} + {\frac{{\partial^{2} P_{111} }}{\partial u\partial v}}P_{100} P_{010}\,{\frac{{\partial P_{001} }}{\partial w}}} \right. \\ & + {\frac{{\partial^{2} P_{111} }}{\partial u\partial w}}\,{\frac{{\partial P_{100} }}{\partial v}}P_{010} P_{001} + {\frac{{\partial P_{111} }}{\partial u}}\,{\frac{{\partial^{2} P_{100} }}{\partial v\partial w}}P_{010} P_{001} + {\frac{{\partial P_{111} }}{\partial u}}\,{\frac{{\partial P_{100} }}{\partial v}}\,{\frac{{\partial P_{010} }}{\partial w}}P_{001} + {\frac{{\partial P_{111} }}{\partial u}}\,{\frac{{\partial P_{100} }}{\partial v}}P_{010}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + {\frac{{\partial^{2} P_{111} }}{\partial u\partial w}}P_{100}\,{\frac{{\partial P_{010} }}{\partial v}}P_{001} + {\frac{{\partial P_{111} }}{\partial u}}\,{\frac{{\partial P_{100} }}{\partial w}}\,{\frac{{\partial P_{010} }}{\partial v}}P_{001} + {\frac{{\partial P_{111} }}{\partial u}}P_{100}\,{\frac{{\partial^{2} P_{010} }}{\partial v\partial w}}P_{001} + {\frac{{\partial P_{111} }}{\partial u}}P_{100}\,{\frac{{\partial P_{010} }}{\partial v}}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + {\frac{{\partial^{2} P_{111} }}{\partial u\partial w}}P_{100} P_{010}\,{\frac{{\partial P_{001} }}{\partial v}} + {\frac{{\partial P_{111} }}{\partial u}}\,{\frac{{\partial P_{100} }}{\partial w}}P_{010}\,{\frac{{\partial P_{001} }}{\partial v}} + {\frac{{\partial P_{111} }}{\partial u}}P_{100}\,{\frac{{\partial P_{010} }}{\partial w}}\,{\frac{{\partial P_{001} }}{\partial v}} + {\frac{{\partial P_{111} }}{\partial u}}P_{100} P_{010}\,{\frac{{\partial^{2} P_{001} }}{\partial v\partial w}} \\ & + {\frac{{\partial^{2} P_{111} }}{\partial v\partial w}}\,{\frac{{\partial P_{100} }}{\partial u}}P_{010} P_{001} + {\frac{{\partial P_{111} }}{\partial v}}\,{\frac{{\partial^{2} P_{100} }}{\partial u\partial w}}P_{010} P_{001} + {\frac{{\partial P_{111} }}{\partial v}}\,{\frac{{\partial P_{100} }}{\partial u}}\,{\frac{{\partial P_{010} }}{\partial w}}P_{001} + {\frac{{\partial P_{111} }}{\partial v}}\,{\frac{{\partial P_{100} }}{\partial u}}P_{010}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + {\frac{{\partial P_{111} }}{\partial w}}\,{\frac{{\partial^{2} P_{100} }}{\partial u\partial v}}P_{010} P_{001} + P_{111} \,{\frac{{\partial^{3} P_{100} }}{\partial u\partial v\partial w}}P_{010} P_{001} + P_{111}\,{\frac{{\partial^{2} P_{100} }}{\partial u\partial v}}\,{\frac{{\partial P_{010} }}{\partial w}}P_{001} + P_{111}\,{\frac{{\partial^{2} P_{100} }}{\partial u\partial v}}P_{010}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + {\frac{{\partial P_{111} }}{\partial w}}\,{\frac{{\partial P_{100} }}{\partial u}}\,{\frac{{\partial P_{010} }}{\partial v}}P_{001} + P_{111}\,{\frac{{\partial^{2} P_{100} }}{\partial u\partial w}}\,{\frac{{\partial P_{010} }}{\partial v}}P_{001} + P_{111}\,{\frac{{\partial P_{100} }}{\partial u}}\,{\frac{{\partial^{2} P_{010} }}{\partial v\partial w}}P_{001} + P_{111}\,{\frac{{\partial P_{100} }}{\partial u}}\,{\frac{{\partial P_{010} }}{\partial v}}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + \,{\frac{{\partial P_{111} }}{\partial w}}\,{\frac{{\partial P_{100} }}{\partial u}}P_{010} \,{\frac{{\partial P_{001} }}{\partial v}} + P_{111} \,{\frac{{\partial^{2} P_{100} }}{\partial u\partial w}}P_{010} \,{\frac{{\partial P_{001} }}{\partial v}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial u}}\,{\frac{{\partial P_{010} }}{\partial w}}\,{\frac{{\partial P_{001} }}{\partial v}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial u}}P_{010} \,{\frac{{\partial^{2} P_{001} }}{\partial v\partial w}} \\ & + \,{\frac{{\partial^{2} P_{111} }}{\partial v\partial w}}P_{100} \,{\frac{{\partial P_{010} }}{\partial u}}P_{001} + \,{\frac{{\partial P_{111} }}{\partial v}}\,{\frac{{\partial P_{100} }}{\partial w}}\,{\frac{{\partial P_{010} }}{\partial u}}P_{001} + \,{\frac{{\partial P_{111} }}{\partial v}}P_{100} \,{\frac{{\partial^{2} P_{010} }}{\partial u\partial w}}P_{001} + \,{\frac{{\partial P_{111} }}{\partial v}}P_{100} \,{\frac{{\partial P_{010} }}{\partial u}}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + \,{\frac{{\partial P_{111} }}{\partial w}}\,{\frac{{\partial P_{100} }}{\partial v}}\,{\frac{{\partial P_{010} }}{\partial u}}P_{001} + P_{111} \,{\frac{{\partial^{2} P_{100} }}{\partial v\partial w}}\,{\frac{{\partial P_{010} }}{\partial u}}P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial v}}\,{\frac{{\partial^{2} P_{010} }}{\partial u\partial w}}P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial v}}\,{\frac{{\partial P_{010} }}{\partial u}}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + \,{\frac{{\partial P_{111} }}{\partial w}}P_{100} \,{\frac{{\partial^{2} P_{010} }}{\partial u\partial v}}P_{001} + P_{111} \,{\frac{{\partial P_{100} }}{\partial w}}\,{\frac{{\partial^{2} P_{010} }}{\partial u\partial v}}P_{001} + P_{111} P_{100} \,{\frac{{\partial^{3} P_{010} }}{\partial u\partial v\partial w}}P_{001} + P_{111} P_{100} \,{\frac{{\partial^{2} P_{010} }}{\partial u\partial v}}\,{\frac{{\partial P_{001} }}{\partial w}} \\ & + \,{\frac{{\partial P_{111} }}{\partial w}}P_{100} \,{\frac{{\partial P_{010} }}{\partial u}}\,{\frac{{\partial P_{001} }}{\partial v}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial w}}\,{\frac{{\partial P_{010} }}{\partial u}}\,{\frac{{\partial P_{001} }}{\partial v}} + P_{111} P_{100} \,{\frac{{\partial^{2} P_{010} }}{\partial u\partial w}}\,{\frac{{\partial P_{001} }}{\partial v}} + P_{111} P_{100} \,{\frac{{\partial P_{010} }}{\partial u}}\,{\frac{{\partial^{2} P_{001} }}{\partial v\partial w}} \\ & + \,{\frac{{\partial^{2} P_{111} }}{\partial v\partial w}}P_{100} P_{010} \,{\frac{{\partial P_{001} }}{\partial u}} + \,{\frac{{\partial P_{111} }}{\partial v}}\,{\frac{{\partial P_{100} }}{\partial w}}P_{010} \,{\frac{{\partial P_{001} }}{\partial u}} + \,{\frac{{\partial P_{111} }}{\partial v}}P_{100} \,{\frac{{\partial P_{010} }}{\partial w}}\,{\frac{{\partial P_{001} }}{\partial u}} + \,{\frac{{\partial P_{111} }}{\partial v}}P_{100} P_{010} \,{\frac{{\partial^{2} P_{001} }}{\partial u\partial w}} \\ & + \,{\frac{{\partial P_{111} }}{\partial w}}\,{\frac{{\partial P_{100} }}{\partial v}}P_{010} \,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} \,{\frac{{\partial^{2} P_{100} }}{\partial v\partial w}}P_{010} \,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial v}}\,{\frac{{\partial P_{010} }}{\partial w}}\,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial v}}P_{010} \,{\frac{{\partial^{2} P_{001} }}{\partial u\partial w}} \\ & + \,{\frac{{\partial P_{111} }}{\partial w}}P_{100} \,{\frac{{\partial P_{010} }}{\partial v}}\,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial w}}\,{\frac{{\partial P_{010} }}{\partial v}}\,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} P_{100} \,{\frac{{\partial^{2} P_{010} }}{\partial v\partial w}}\,{\frac{{\partial P_{001} }}{\partial u}} + P_{111} P_{100} \,{\frac{{\partial P_{010} }}{\partial v}}\,{\frac{{\partial^{2} P_{001} }}{\partial u\partial w}} \\ & \left. { + \,{\frac{{\partial P_{111} }}{\partial w}}P_{100} P_{010} \,{\frac{{\partial^{2} P_{001} }}{\partial u\partial v}} + P_{111} \,{\frac{{\partial P_{100} }}{\partial w}}P_{010} \,{\frac{{\partial^{2} P_{001} }}{\partial u\partial v}} + P_{111} P_{100} \,{\frac{{\partial P_{010} }}{\partial w}}\,{\frac{{\partial^{2} P_{001} }}{\partial u\partial v}} + P_{111} P_{100} P_{010} \,{\frac{{\partial^{3} P_{001} }}{\partial u\partial v\partial w}}} \right) = 0 \\ \end{aligned} $$
(50)

Appendix 3: Estimation of trivariate cross product ratio

  1. (1)

    Compute C UV , C VW , and C UW using Eq. 3.

  2. (2)

    Compute the empirical trivariate probability distribution P 0(i), i = 1, 2,…, n.

  3. (3)

    Set the parameters of GA, including population size, probability of selection and crossover and mutation, calculation terminal condition.

  4. (4)

    Generate randomly the cross-product ratio of trivariate Plackett copula and form an initial population.

  5. (5)

    Compute the individual ψ UVW using Eqs. 710 and evaluate the fitness of each individual in the population using Eq. 17.

  6. (6)

    Repeat the following calculation steps until termination condition: ① Select best-ranking individuals to reproduce. ② Breed new generation through genetic operations (crossover and mutation) and give birth to their offsprings. ③ Compute the individual ψ UVW using Eqs. 710 and evaluate the individual fitness of offsprings using Eq. 17. ④ Replace the worst ranked part of population with offspring.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Song, S., Singh, V.P. Frequency analysis of droughts using the Plackett copula and parameter estimation by genetic algorithm. Stoch Environ Res Risk Assess 24, 783–805 (2010). https://doi.org/10.1007/s00477-010-0364-5

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00477-010-0364-5

Keywords

Navigation