Multi-criteria decision-making analysis in selecting suitable plotting positions for IDF curves of storm events

Research output: Contribution to journalArticle

Abstract

Selecting suitable distributions for rainfall data is usually subjective and complex since it requires decision-makers to consider results from various measures of goodness-of-fit indices. In this study, the VIKOR method in multi-criteria decision-making analysis is modified to select the most suitable plotting positions to represent extreme storm intensities in order to build the intensity–duration–frequency (IDF) curves of storm events. This is done by considering the rankings provided by all goodness-of-fit indices used to obtain a compromise solution. Nine plotting positions are considered: Weibull (W), Adamowski (A), Gringorten (G), Hazen (H) and Gumbel (EV I) and two known plotting positions for generalized extreme value (GEV) distribution using Pearson’s skewness and another two using L-skewness. The IDF curves obtained are compared to a reference IDF curves which was found using the GEV distribution. The mean and median for three goodness-of-fit indices, the coefficient of variation of root mean square error, CVRMSE, the mean percentage of difference, Δ, and the coefficient of determination, R2, are taken as the criteria for selection process. The results show that six plotting positions, A, H, W, G and the two plotting positions with L-skewness, are equally superior compared to the other three plotting positions.

Original languageEnglish
Article number625
JournalEnvironmental Earth Sciences
Volume76
Issue number17
DOIs
Publication statusPublished - 1 Sep 2017

Fingerprint

multi-criteria decision making
skewness
Decision making
decision making
selection criteria
meteorological data
Mean square error
Rain
ranking
rainfall
distribution
analysis
index
methodology

Keywords

  • Intensity–duration–frequency (IDF)
  • Plotting positions
  • Storm events
  • VIKOR method

ASJC Scopus subject areas

  • Global and Planetary Change
  • Environmental Chemistry
  • Water Science and Technology
  • Soil Science
  • Pollution
  • Geology
  • Earth-Surface Processes

Cite this

@article{3c93d1d01f73460db93408b87b7bd5d5,
title = "Multi-criteria decision-making analysis in selecting suitable plotting positions for IDF curves of storm events",
abstract = "Selecting suitable distributions for rainfall data is usually subjective and complex since it requires decision-makers to consider results from various measures of goodness-of-fit indices. In this study, the VIKOR method in multi-criteria decision-making analysis is modified to select the most suitable plotting positions to represent extreme storm intensities in order to build the intensity–duration–frequency (IDF) curves of storm events. This is done by considering the rankings provided by all goodness-of-fit indices used to obtain a compromise solution. Nine plotting positions are considered: Weibull (W), Adamowski (A), Gringorten (G), Hazen (H) and Gumbel (EV I) and two known plotting positions for generalized extreme value (GEV) distribution using Pearson’s skewness and another two using L-skewness. The IDF curves obtained are compared to a reference IDF curves which was found using the GEV distribution. The mean and median for three goodness-of-fit indices, the coefficient of variation of root mean square error, CVRMSE, the mean percentage of difference, Δ, and the coefficient of determination, R2, are taken as the criteria for selection process. The results show that six plotting positions, A, H, W, G and the two plotting positions with L-skewness, are equally superior compared to the other three plotting positions.",
keywords = "Intensity–duration–frequency (IDF), Plotting positions, Storm events, VIKOR method",
author = "{Mohd Ariff}, Noratiqah and Jemain, {Abdul Aziz} and {Abu Bakar}, {Mohd Aftar}",
year = "2017",
month = "9",
day = "1",
doi = "10.1007/s12665-017-6952-7",
language = "English",
volume = "76",
journal = "Environmental Earth Sciences",
issn = "1866-6280",
publisher = "Springer Verlag",
number = "17",

}

TY - JOUR

T1 - Multi-criteria decision-making analysis in selecting suitable plotting positions for IDF curves of storm events

AU - Mohd Ariff, Noratiqah

AU - Jemain, Abdul Aziz

AU - Abu Bakar, Mohd Aftar

PY - 2017/9/1

Y1 - 2017/9/1

N2 - Selecting suitable distributions for rainfall data is usually subjective and complex since it requires decision-makers to consider results from various measures of goodness-of-fit indices. In this study, the VIKOR method in multi-criteria decision-making analysis is modified to select the most suitable plotting positions to represent extreme storm intensities in order to build the intensity–duration–frequency (IDF) curves of storm events. This is done by considering the rankings provided by all goodness-of-fit indices used to obtain a compromise solution. Nine plotting positions are considered: Weibull (W), Adamowski (A), Gringorten (G), Hazen (H) and Gumbel (EV I) and two known plotting positions for generalized extreme value (GEV) distribution using Pearson’s skewness and another two using L-skewness. The IDF curves obtained are compared to a reference IDF curves which was found using the GEV distribution. The mean and median for three goodness-of-fit indices, the coefficient of variation of root mean square error, CVRMSE, the mean percentage of difference, Δ, and the coefficient of determination, R2, are taken as the criteria for selection process. The results show that six plotting positions, A, H, W, G and the two plotting positions with L-skewness, are equally superior compared to the other three plotting positions.

AB - Selecting suitable distributions for rainfall data is usually subjective and complex since it requires decision-makers to consider results from various measures of goodness-of-fit indices. In this study, the VIKOR method in multi-criteria decision-making analysis is modified to select the most suitable plotting positions to represent extreme storm intensities in order to build the intensity–duration–frequency (IDF) curves of storm events. This is done by considering the rankings provided by all goodness-of-fit indices used to obtain a compromise solution. Nine plotting positions are considered: Weibull (W), Adamowski (A), Gringorten (G), Hazen (H) and Gumbel (EV I) and two known plotting positions for generalized extreme value (GEV) distribution using Pearson’s skewness and another two using L-skewness. The IDF curves obtained are compared to a reference IDF curves which was found using the GEV distribution. The mean and median for three goodness-of-fit indices, the coefficient of variation of root mean square error, CVRMSE, the mean percentage of difference, Δ, and the coefficient of determination, R2, are taken as the criteria for selection process. The results show that six plotting positions, A, H, W, G and the two plotting positions with L-skewness, are equally superior compared to the other three plotting positions.

KW - Intensity–duration–frequency (IDF)

KW - Plotting positions

KW - Storm events

KW - VIKOR method

UR - http://www.scopus.com/inward/record.url?scp=85029183352&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029183352&partnerID=8YFLogxK

U2 - 10.1007/s12665-017-6952-7

DO - 10.1007/s12665-017-6952-7

M3 - Article

AN - SCOPUS:85029183352

VL - 76

JO - Environmental Earth Sciences

JF - Environmental Earth Sciences

SN - 1866-6280

IS - 17

M1 - 625

ER -