Long term analysis of wet and dry years in Seoul, Korea

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

This study investigated the recurrence characteristics of wet and dry years using the annual precipitation data collected in Seoul, Korea since 1776. More than one half of these 200 years of precipitation data are those collected in the Chosun Age using an old Korean rain gauge called Chukwooki. The recurrence characteristics of wet and dry years were investigated for several sets of thresholds, represented by use of the mean and standard deviation (stdv) of the annual precipitation such as mean±0.5stdv, mean±0.75stdv and mean±1.0stdv. These sets of thresholds have been decided so as to make the occurrence of wet and dry years follow the Poisson distribution. For a given set of thresholds, the wet, dry, and normal years are categorized, then the transition probabilities among those years were derived and compared. The average return periods were also derived using the stationary probabilities of wet and dry years. This analysis has been applied not only to the entire data but also to partial data sets composed of the data before and after the long dry period (lasting about 25 years) around 1900. This was to compare and detect the difference in the recurrence characteristics as well as the difference between the old Chukwooki and the modern flip-bucket style rain gauge data. As results, the overall recurrence patterns of wet and dry years have been found to be very similar. Only one obvious difference may be the return period of extremely dry years (for the threshold of mean-stdv), which after the long dry period was found to be longer than that before the long dry period (8.03 and 6.77 years, respectively). A similar result could also be found in the occurrence probability (or, the inverse of return period) of consecutive dry years estimated by applying the Poisson process. That is, for the lowest threshold of mean-stdv, the occurrence probability of consecutive dry years before the long dry period was higher than those after the long dry period. Thus, we may conclude that the possibility of long dry periods is decreasing recently.

Original languageEnglish
Pages (from-to)24-36
Number of pages13
JournalJournal of Hydrology
Volume318
Issue number1-4
DOIs
Publication statusPublished - 2006 Mar 1

Fingerprint

Korean Peninsula
rain gauges
return period
gauge
buckets
analysis
rain

Keywords

  • Dry year
  • Occurrence probability
  • Poisson process
  • Return period
  • Transition probability
  • Wet year

ASJC Scopus subject areas

  • Soil Science
  • Earth-Surface Processes

Cite this

Long term analysis of wet and dry years in Seoul, Korea. / Yoo, Chulsang.

In: Journal of Hydrology, Vol. 318, No. 1-4, 01.03.2006, p. 24-36.

Research output: Contribution to journalArticle

@article{7e04f96a8f6144d685ca233c6ef559b7,
title = "Long term analysis of wet and dry years in Seoul, Korea",
abstract = "This study investigated the recurrence characteristics of wet and dry years using the annual precipitation data collected in Seoul, Korea since 1776. More than one half of these 200 years of precipitation data are those collected in the Chosun Age using an old Korean rain gauge called Chukwooki. The recurrence characteristics of wet and dry years were investigated for several sets of thresholds, represented by use of the mean and standard deviation (stdv) of the annual precipitation such as mean±0.5stdv, mean±0.75stdv and mean±1.0stdv. These sets of thresholds have been decided so as to make the occurrence of wet and dry years follow the Poisson distribution. For a given set of thresholds, the wet, dry, and normal years are categorized, then the transition probabilities among those years were derived and compared. The average return periods were also derived using the stationary probabilities of wet and dry years. This analysis has been applied not only to the entire data but also to partial data sets composed of the data before and after the long dry period (lasting about 25 years) around 1900. This was to compare and detect the difference in the recurrence characteristics as well as the difference between the old Chukwooki and the modern flip-bucket style rain gauge data. As results, the overall recurrence patterns of wet and dry years have been found to be very similar. Only one obvious difference may be the return period of extremely dry years (for the threshold of mean-stdv), which after the long dry period was found to be longer than that before the long dry period (8.03 and 6.77 years, respectively). A similar result could also be found in the occurrence probability (or, the inverse of return period) of consecutive dry years estimated by applying the Poisson process. That is, for the lowest threshold of mean-stdv, the occurrence probability of consecutive dry years before the long dry period was higher than those after the long dry period. Thus, we may conclude that the possibility of long dry periods is decreasing recently.",
keywords = "Dry year, Occurrence probability, Poisson process, Return period, Transition probability, Wet year",
author = "Chulsang Yoo",
year = "2006",
month = "3",
day = "1",
doi = "10.1016/j.jhydrol.2005.06.002",
language = "English",
volume = "318",
pages = "24--36",
journal = "Journal of Hydrology",
issn = "0022-1694",
publisher = "Elsevier",
number = "1-4",

}

TY - JOUR

T1 - Long term analysis of wet and dry years in Seoul, Korea

AU - Yoo, Chulsang

PY - 2006/3/1

Y1 - 2006/3/1

N2 - This study investigated the recurrence characteristics of wet and dry years using the annual precipitation data collected in Seoul, Korea since 1776. More than one half of these 200 years of precipitation data are those collected in the Chosun Age using an old Korean rain gauge called Chukwooki. The recurrence characteristics of wet and dry years were investigated for several sets of thresholds, represented by use of the mean and standard deviation (stdv) of the annual precipitation such as mean±0.5stdv, mean±0.75stdv and mean±1.0stdv. These sets of thresholds have been decided so as to make the occurrence of wet and dry years follow the Poisson distribution. For a given set of thresholds, the wet, dry, and normal years are categorized, then the transition probabilities among those years were derived and compared. The average return periods were also derived using the stationary probabilities of wet and dry years. This analysis has been applied not only to the entire data but also to partial data sets composed of the data before and after the long dry period (lasting about 25 years) around 1900. This was to compare and detect the difference in the recurrence characteristics as well as the difference between the old Chukwooki and the modern flip-bucket style rain gauge data. As results, the overall recurrence patterns of wet and dry years have been found to be very similar. Only one obvious difference may be the return period of extremely dry years (for the threshold of mean-stdv), which after the long dry period was found to be longer than that before the long dry period (8.03 and 6.77 years, respectively). A similar result could also be found in the occurrence probability (or, the inverse of return period) of consecutive dry years estimated by applying the Poisson process. That is, for the lowest threshold of mean-stdv, the occurrence probability of consecutive dry years before the long dry period was higher than those after the long dry period. Thus, we may conclude that the possibility of long dry periods is decreasing recently.

AB - This study investigated the recurrence characteristics of wet and dry years using the annual precipitation data collected in Seoul, Korea since 1776. More than one half of these 200 years of precipitation data are those collected in the Chosun Age using an old Korean rain gauge called Chukwooki. The recurrence characteristics of wet and dry years were investigated for several sets of thresholds, represented by use of the mean and standard deviation (stdv) of the annual precipitation such as mean±0.5stdv, mean±0.75stdv and mean±1.0stdv. These sets of thresholds have been decided so as to make the occurrence of wet and dry years follow the Poisson distribution. For a given set of thresholds, the wet, dry, and normal years are categorized, then the transition probabilities among those years were derived and compared. The average return periods were also derived using the stationary probabilities of wet and dry years. This analysis has been applied not only to the entire data but also to partial data sets composed of the data before and after the long dry period (lasting about 25 years) around 1900. This was to compare and detect the difference in the recurrence characteristics as well as the difference between the old Chukwooki and the modern flip-bucket style rain gauge data. As results, the overall recurrence patterns of wet and dry years have been found to be very similar. Only one obvious difference may be the return period of extremely dry years (for the threshold of mean-stdv), which after the long dry period was found to be longer than that before the long dry period (8.03 and 6.77 years, respectively). A similar result could also be found in the occurrence probability (or, the inverse of return period) of consecutive dry years estimated by applying the Poisson process. That is, for the lowest threshold of mean-stdv, the occurrence probability of consecutive dry years before the long dry period was higher than those after the long dry period. Thus, we may conclude that the possibility of long dry periods is decreasing recently.

KW - Dry year

KW - Occurrence probability

KW - Poisson process

KW - Return period

KW - Transition probability

KW - Wet year

UR - http://www.scopus.com/inward/record.url?scp=31044448557&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=31044448557&partnerID=8YFLogxK

U2 - 10.1016/j.jhydrol.2005.06.002

DO - 10.1016/j.jhydrol.2005.06.002

M3 - Article

AN - SCOPUS:31044448557

VL - 318

SP - 24

EP - 36

JO - Journal of Hydrology

JF - Journal of Hydrology

SN - 0022-1694

IS - 1-4

ER -