On the variation of frequency-based rainfall amounts

A case study for evaluating recent extreme rainfalls in Korea

J. H. Ahn, T. W. Kim, Chulsang Yoo, Y. N. Yoon

Research output: Contribution to journalArticle

Abstract

Extreme rainfall events recently occurring in Korea have been shown to change frequency-based rainfall amounts quite significantly. Regardless of the reason for these extremes, the general concern of most hydrologists is how to handle these events for practical applications in Hydrology. Our study aim is to evaluate these extremes with their effect on frequency-based rainfall amounts, especially if they can be assumed to be within normal levels. As there is no commonly accepted methodology to be applied to this kind of study, we follow simplified steps such as: (1) estimation of the climatological variance of frequency-based rainfall amounts, (2) estimation of confidence intervals of frequency-based rainfall amounts (lower and upper bounds for the 5 and 1% significance levels estimated using the climatological variance), and (3) evaluation of the effect of extra rainfall events on the frequency-based rainfall amounts. Twelve stations on the Korean peninsula are selected as they have relatively longer data length. The annual maximum rainfall data collected from 1954 to 1998 are used. From this study we concluded that (1) at least 30 years of data length should be used for the frequency analysis in order to assure the stability of the variance of frequency-based rainfall amounts, (2) the climatological variances estimated all range from 5 to 8% of the frequency-based rainfall amounts, and (3) even though the frequency-based rainfall amount seems to become extreme with seemingly abnormal events, it still remains under its upper bound for the 5 or 1% significance levels estimated using the climatological variance, as well as it decays exponentially to the normal level as extra events are added. Thus, we conclude that we do not need to panic over seemingly abnormal events occurring so far, but just need to consider the variability inherent in frequency-based rainfall amounts.

Original languageEnglish
Pages (from-to)217-227
Number of pages11
JournalStochastic Environmental Research and Risk Assessment
Volume17
Issue number3
DOIs
Publication statusPublished - 2003 Sep 1

Fingerprint

Rainfall
Rain
Extremes
rainfall
Significance level
Hydrology
Frequency Analysis
frequency analysis
confidence interval
Annual
Confidence interval
Upper and Lower Bounds
hydrology
Decay
Upper bound

Keywords

  • Climatological variance
  • Confidence limits
  • Frequency analysis
  • Global warming
  • Rainfall

ASJC Scopus subject areas

  • Environmental Engineering
  • Environmental Science(all)
  • Environmental Chemistry
  • Water Science and Technology
  • Statistics and Probability
  • Civil and Structural Engineering

Cite this

On the variation of frequency-based rainfall amounts : A case study for evaluating recent extreme rainfalls in Korea. / Ahn, J. H.; Kim, T. W.; Yoo, Chulsang; Yoon, Y. N.

In: Stochastic Environmental Research and Risk Assessment, Vol. 17, No. 3, 01.09.2003, p. 217-227.

Research output: Contribution to journalArticle

@article{dbbe51e596564a7bb3ad53a08afaf95f,
title = "On the variation of frequency-based rainfall amounts: A case study for evaluating recent extreme rainfalls in Korea",
abstract = "Extreme rainfall events recently occurring in Korea have been shown to change frequency-based rainfall amounts quite significantly. Regardless of the reason for these extremes, the general concern of most hydrologists is how to handle these events for practical applications in Hydrology. Our study aim is to evaluate these extremes with their effect on frequency-based rainfall amounts, especially if they can be assumed to be within normal levels. As there is no commonly accepted methodology to be applied to this kind of study, we follow simplified steps such as: (1) estimation of the climatological variance of frequency-based rainfall amounts, (2) estimation of confidence intervals of frequency-based rainfall amounts (lower and upper bounds for the 5 and 1{\%} significance levels estimated using the climatological variance), and (3) evaluation of the effect of extra rainfall events on the frequency-based rainfall amounts. Twelve stations on the Korean peninsula are selected as they have relatively longer data length. The annual maximum rainfall data collected from 1954 to 1998 are used. From this study we concluded that (1) at least 30 years of data length should be used for the frequency analysis in order to assure the stability of the variance of frequency-based rainfall amounts, (2) the climatological variances estimated all range from 5 to 8{\%} of the frequency-based rainfall amounts, and (3) even though the frequency-based rainfall amount seems to become extreme with seemingly abnormal events, it still remains under its upper bound for the 5 or 1{\%} significance levels estimated using the climatological variance, as well as it decays exponentially to the normal level as extra events are added. Thus, we conclude that we do not need to panic over seemingly abnormal events occurring so far, but just need to consider the variability inherent in frequency-based rainfall amounts.",
keywords = "Climatological variance, Confidence limits, Frequency analysis, Global warming, Rainfall",
author = "Ahn, {J. H.} and Kim, {T. W.} and Chulsang Yoo and Yoon, {Y. N.}",
year = "2003",
month = "9",
day = "1",
doi = "10.1007/s00477-002-0123-3",
language = "English",
volume = "17",
pages = "217--227",
journal = "Stochastic Environmental Research and Risk Assessment",
issn = "1436-3240",
publisher = "Springer New York",
number = "3",

}

TY - JOUR

T1 - On the variation of frequency-based rainfall amounts

T2 - A case study for evaluating recent extreme rainfalls in Korea

AU - Ahn, J. H.

AU - Kim, T. W.

AU - Yoo, Chulsang

AU - Yoon, Y. N.

PY - 2003/9/1

Y1 - 2003/9/1

N2 - Extreme rainfall events recently occurring in Korea have been shown to change frequency-based rainfall amounts quite significantly. Regardless of the reason for these extremes, the general concern of most hydrologists is how to handle these events for practical applications in Hydrology. Our study aim is to evaluate these extremes with their effect on frequency-based rainfall amounts, especially if they can be assumed to be within normal levels. As there is no commonly accepted methodology to be applied to this kind of study, we follow simplified steps such as: (1) estimation of the climatological variance of frequency-based rainfall amounts, (2) estimation of confidence intervals of frequency-based rainfall amounts (lower and upper bounds for the 5 and 1% significance levels estimated using the climatological variance), and (3) evaluation of the effect of extra rainfall events on the frequency-based rainfall amounts. Twelve stations on the Korean peninsula are selected as they have relatively longer data length. The annual maximum rainfall data collected from 1954 to 1998 are used. From this study we concluded that (1) at least 30 years of data length should be used for the frequency analysis in order to assure the stability of the variance of frequency-based rainfall amounts, (2) the climatological variances estimated all range from 5 to 8% of the frequency-based rainfall amounts, and (3) even though the frequency-based rainfall amount seems to become extreme with seemingly abnormal events, it still remains under its upper bound for the 5 or 1% significance levels estimated using the climatological variance, as well as it decays exponentially to the normal level as extra events are added. Thus, we conclude that we do not need to panic over seemingly abnormal events occurring so far, but just need to consider the variability inherent in frequency-based rainfall amounts.

AB - Extreme rainfall events recently occurring in Korea have been shown to change frequency-based rainfall amounts quite significantly. Regardless of the reason for these extremes, the general concern of most hydrologists is how to handle these events for practical applications in Hydrology. Our study aim is to evaluate these extremes with their effect on frequency-based rainfall amounts, especially if they can be assumed to be within normal levels. As there is no commonly accepted methodology to be applied to this kind of study, we follow simplified steps such as: (1) estimation of the climatological variance of frequency-based rainfall amounts, (2) estimation of confidence intervals of frequency-based rainfall amounts (lower and upper bounds for the 5 and 1% significance levels estimated using the climatological variance), and (3) evaluation of the effect of extra rainfall events on the frequency-based rainfall amounts. Twelve stations on the Korean peninsula are selected as they have relatively longer data length. The annual maximum rainfall data collected from 1954 to 1998 are used. From this study we concluded that (1) at least 30 years of data length should be used for the frequency analysis in order to assure the stability of the variance of frequency-based rainfall amounts, (2) the climatological variances estimated all range from 5 to 8% of the frequency-based rainfall amounts, and (3) even though the frequency-based rainfall amount seems to become extreme with seemingly abnormal events, it still remains under its upper bound for the 5 or 1% significance levels estimated using the climatological variance, as well as it decays exponentially to the normal level as extra events are added. Thus, we conclude that we do not need to panic over seemingly abnormal events occurring so far, but just need to consider the variability inherent in frequency-based rainfall amounts.

KW - Climatological variance

KW - Confidence limits

KW - Frequency analysis

KW - Global warming

KW - Rainfall

UR - http://www.scopus.com/inward/record.url?scp=0344961254&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0344961254&partnerID=8YFLogxK

U2 - 10.1007/s00477-002-0123-3

DO - 10.1007/s00477-002-0123-3

M3 - Article

VL - 17

SP - 217

EP - 227

JO - Stochastic Environmental Research and Risk Assessment

JF - Stochastic Environmental Research and Risk Assessment

SN - 1436-3240

IS - 3

ER -