hydrometeorological variability in the korean han river basin and its sub-watersheds during...
TRANSCRIPT
-
ORIGINAL PAPER
Hydrometeorological variability in the Korean Han River Basinand its sub-watersheds during different El Nino phases
Sun-Kwon Yoon Jong-Suk Kim Joo-Heon Lee
Young-Il Moon
Published online: 16 January 2013
Springer-Verlag Berlin Heidelberg 2013
Abstract This study investigated the characteristic changes
in precipitation and runoff that occur in the Korean Han
River Basin and its sub-basins in association with the cold-
tongue (CT) and warm-pool (WP) El Nino phases during
spring and summer. During the WP El Nino years, rainfall
in spring and its coefficient of variation were higher than
long-term normal precipitation. During the CT El Nino
years, summers tended to be drier than in climatologically
normal years, although the variability in precipitation
during the summer was relatively lower. The data for
runoff showed wetter springs compared to long-term nor-
mal years during both types of El Nino events and signif-
icant changes in runoff during summer under CT El Nino
conditions. During the WP El Nino years, increased runoff
was seen for 95.8 % of all basins and this increase was
statistically significant for 58.3 % of these basins, but
variability in runoff was small. Overall, the findings con-
firm that water resources in the Han River Basin during the
spring and summer are sensitive to CT/WP El Nino events.
Thus, for basins such as these, where seasonal variability
and the uncertainty of hydrologic data are high, investi-
gation of the relationship between climatic factors and
hydrologic parameters is necessary to maintain the stability
of the water supply system and to allow prediction for
water resources.
Keywords Seasonal rainfall Runoff CT El Nino WP El Nino Han River South Korea
1 Introduction
The seasonal characteristics of hydrometeorological vari-
ability are closely related to global climate phenomena and
climate changes (Kim et al. 2008; Grimm 2011). Therefore,
investigating the correlation between climatic factors and
hydrologic data (such as precipitation and streamflow) is
very important for the accurate prediction and management
of water resources (Horel and Wallace 1981; Pizarro and
Lall 2002; Kim et al. 2011, 2012c). However, regional
hydrologic variability has a complex relationship with cli-
mate, including effects across the hydrologic cycle rather
than being limited to independent phenomena. Changes in
global climate systems have significant consequences for
water resources, which are closely associated with weather
in the short term and climate phenomena in the long term.
Thus, understanding the changes in sea surface tempera-
tures (hereinafter referred to as SST) in the tropical
Pacific Ocean is very important because this can provide
useful information regarding how present and future
hydrological runoff may differ among different regions.
This information is also valuable for predicting the types of
natural disasters that may be caused locally and to take
preparatory measures to face these disasters. However,
despite the many studies (Wang et al. 2000; Wu et al. 2003;
Jin et al. 2005; Weng et al. 2007; Lee and Byun 2009;
Feng et al. 2010; Kim et al. 2012a) that have invest-
igated the relationship between climatic factors and
S.-K. Yoon
Climate Research Department, APEC Climate Center,
Busan 612-020, Republic of Korea
J.-S. Kim (&) Y.-I. MoonDepartment of Civil Engineering, The University of Seoul,
Seoul 130-743, Republic of Korea
e-mail: [email protected]
J.-H. Lee
Department of Civil Engineering, Joongbu Univeristy,
Chung-nam 312-702, Republic of Korea
123
Stoch Environ Res Risk Assess (2013) 27:14651477
DOI 10.1007/s00477-012-0683-9
-
hydrometeorological variables, these variables are still not
sufficiently understood for investigation of the changes in
SST in East Asian regions. In particular, abnormal SST
distributions, known as the El-Nino/Southern Oscillation
(hereinafter referred to as ENSO) and the nonlinear
behavior of climate systems can significantly influence
hydrologic characteristics (Piechota and Dracup 1996;
Piechota et al. 1998; She and Xia 2012). In fact, the ENSO
is known to be a major factor affecting atmospheric circu-
lation patterns, and hydrologic environmental changes have
a close relationship with seasonal anomalies in precipitation
and runoff (Schonher and Nicholson 1989; Gershunov et al.
1999; McPhaden et al. 2006; Kwon et al. 2009; Wang et al.
2011). For example, Weng et al. (2007) and Feng et al.
(2010) showed that El Nino phenomena cause significant
changes in the hydrometeorolgical cycle across many
regions of the world.
Today, the effects of global warming and climate
change must also be taken into consideration, as these have
been shown to alter the typical patterns of large-scale SST
rises and dips. This issue is of particular interest to coun-
tries close to the Pacific region (Weng et al. 2007; Kao and
Yu 2009; Yeh et al. 2009; Na et al. 2011). Specifically,
since the late 1970s, a clear change has been seen in the
frequency and intensity, as well as the location of El Nino
(Weng et al. 2007; Chang et al. 2008; Kao and Yu 2009),
and increases in atmospheric CO2 could have further
aggravated these El Nino changes. Indeed, Kug et al.
(2009), based on the transition mechanism of observed SST
data from the Nino3 and Nino4 regions, identified a new
type of El Nino event that is different from the conven-
tional cold-tongue (hereinafter referred to as CT) event.
This new type of event, called a warm-pool (hereinafter
referred to as WP) El Nino event is characterized by
abnormal increases in SST that occur only in the central
Pacific region (Ashok et al. 2007; Ashok and Yamagata
2009; Feng et al. 2010; Pradhan et al. 2011). In addition,
Ren and Jin (2011) further classified El Nino events into
CT and WP on the basis of the SST data provided by Kug
et al. (2009) and proposed a simple transformation to better
depict the WP El Nino that is almost completely inde-
pendent from the CT El Nino. Although several studies
(Wang et al. 2000; Wu et al. 2003; Jin et al. 2005; Lee and
Byun 2009) have been conducted to investigate if a sys-
tematic relationship between climate variability and SST
around the Korean peninsula and East Asia exists, few
studies (Weng et al. 2007; Feng et al. 2010; Kim et al.
2012a) have explored the consequences of this new El
Nino event on the characteristic changes in precipitation
and runoff in the East Asia region. In addition, little
attention has been given to spatial variations in precipita-
tion and runoff focused on the Korean peninsula during the
new type of El Nino.
The aim of the present study was therefore to use the
CT/WP El Nino categorization results of Ren and Jin
(2011) to determine their relationships with precipitation
and runoff in the Korean Han River Basin and its sub-
basins. Specifically this study has been conducted to ana-
lyze the effect of different types of El Nino on seasonal
variability and extreme events at the local scale in Korea.
We expect that this diagnostic study could lead a better
understanding of characteristic changes in precipitation and
runoff within a climatological context. Moreover, it may
provide a first step in developing seasonal hydrologic
estimates conditioned upon large-scale climate state for
stable water supply and flood risk management in a
changing climate.
In this paper, the geographical study area and data are
presented in Sect. 2, the methods are described in Sect. 3,
and the analysis results are discussed in Sect. 4. Section 5
presents the discussion and conclusions.
2 Study area and data
2.1 Study area
The Han River Basin and its sub-basins are located in the
central part of the Korean peninsula and are delineated by
the latitudes 3630038550N and longitudes 126240129020E (Fig. 1). The river is 481.7 km in length and thebasins cover a total area of 26,356 km2, which constitutes
23 % of the land area of South Korea. The average basin
slope is 19 %, with an average elevation of 406 m above
sea level (WAMIS 2012), and most of the eastern part of
the basin region is composed of mountains and valleys with
steep slopes (Chang 2008; Kim et al. 2011, 2012b).
The annual average precipitation in the Han River Basin
from the periods 19662007 was 1,253.3 mm (Fig. 2;
Table 1). The spatial distribution of the annual average
precipitation in the basin is relatively even, although high
annual precipitation occurred for some sub-basins in the
central and western parts (Fig. 2a). Precipitation also
exhibits strong seasonal variability in the Han River Basin.
The lowest rainfall month is December (21 mm) and the
highest (307 mm) is July. Precipitation during spring is
167225 mm, and the average spring precipitation for the
23 sub-basins is 200.5 mm, which accounts for 16.0 % of
the average annual precipitation. The spatial variation of
spring precipitation is smaller to the other seasons
(Fig. 2b). During summer, however, average precipitation
is 760.2 mm. This constitutes 60.6 % of the average annual
precipitation for the entire basin region, reflecting the
concentration of precipitation during summer (Kim and
Jain 2011; Kim et al. 2012c). The average summer pre-
cipitation appears to be relatively high in the northwest and
1466 Stoch Environ Res Risk Assess (2013) 27:14651477
123
-
central regions including Seoul, which is adjacent to the
West Sea (Yellow Sea) (Fig. 2c).
Runoff from the Han River Basin during spring is
10.817.7 % of the total annual runoff, with a large pro-
portion coming from some of the sub-basins in the eastern
and central parts of the basin region. The average seasonal
fractional runoff from the 24 sub-basins is 14.4 %
(Fig. 3a), and the coefficients of variation (CV; the ratio of
standard deviation to mean) for spring runoff in each sub-
basin, compared to the average CV for the total region for
19712000, varies from a maximum of 0.83 to a minimum
of 0.44, with an average of 0.64. This suggests that large
variations in runoff occur during the spring (Fig. 3c). The
northwest region dominated the summer runoff, accounting
for 47.267.4 % of the total, and the average seasonal
fractional runoff from sub-basins was estimated at 60.2 %
(Fig. 3b). In addition, the CV for the summer runoff from
each sub-basin, when compared to the normal average of
the CV value for the same 30 years, varied from a maxi-
mum of 0.33 to a minimum of 0.19, with an average 0.23
(Fig. 3d). Thus, runoff is rather evenly distributed across
the 24 sub-basins, confirming that variation in runoff dur-
ing the summer is particularly high. The spatial distribution
of summer runoff revealed a somewhat large variation in
sub-basins located in the southern parts.
2.2 Data
The hydrometeorological data used in this research were
precipitation and runoff data collected from 1966 to 2007
by the Water Management Information System (WAMIS
2012). The average precipitation data was estimated for
24 sub-basins using the Thiessen polygon method on data
from 43 precipitation gauge stations in the Han River
Basin (Kim and Jain 2011). Runoff data available for the
24 sub-basins were calculated using the widely applied
Precipitation-Runoff Modeling System (PRMS), a physi-
cally based deterministic and distributed-parameter rain-
fall-runoff model developed by the US Geological Survey
(Leavesley et al. 1983; Dressler et al. 2006; Bae et al.
2008; Chang and Jung 2010). These daily runoff data,
which are an assessment of the national watershed
research project in Korea (Kim et al. 2012b; WAMIS
2012), are used for analyzing the long-term effects of
hydrological changes at the regional scale and for setting
up mid-to-long-term water resources planning and man-
agement. For daily runoff computations, the PRMS model
requires daily precipitation and minimum and maximum
air temperatures. Bae et al. (2008) also provides a more
detailed description of PRMS model calibration and
regionalization of hydrologic responses over the Korean
peninsula. Furthermore, it is important to note that the
precipitation dataset for the sub-basin (Kumgangsan Dam)
located in North Korea is not available on the WAMIS
system. Therefore, in this study, we used precipitation
information for 23 sub-basins excluding this sub-basin,
but we recommend that analysis results of the runoff data
for the Kumgangsan Dam basin presented here be used
for reference purposes as it may have a large degree of
uncertainty.
Fig. 1 Location of the studyarea. The Han River Basin
(126240E129020E,36300N38550N) is locatedin the center of the Korean
peninsula. The red line indicates
the demilitarized zone (DMZ)
between North and South
Korea. More information on the
sub-basins is available in
Table 1. (Color figure online)
Stoch Environ Res Risk Assess (2013) 27:14651477 1467
123
-
The monthly ENSO data used in this research were
obtained from the National Oceanic and Atmospheric
Administration (NOAA 2012). Among the data collection
sites, observation data from Nino3 (5S5N, 15090W)and Nino4 (5S5N, 160E150W) sites are known to
have strong correlation with El Nino (Trenberth 1997). The
SST data provided by the Hadley Center, called HadISST,
were used to investigate the correlation between hydrome-
teorological parameters (seasonal precipitation and runoff,
annual maximum flow, 7-day lowflow) and large-scale
NA
13001337mm1250130012001250115012001100115010811100
(a) Annual total precipitation
NA 220225mm210220200210190200180190170180167170
(b) Spring total precipitation (MarchMay)
NA 825852mm800825775800750775725750700725670700
(c) Summer total precipitation (JuneAugust)
Fig. 2 Annual total precipitation and seasonal mean precipitationover the Mid-watershed in the Han River basin. a represents annualtotal precipitation, b and c are seasonal precipitation during the spring
(March to May) season and the summer (June to August) season,
respectively. NA indicates that data are not available
Table 1 Sub-basin information for the Han River Basin, South Korea
ID Han River Basin Catchment
area (km2)
Annual
precipitation
(mm)
Fractional flow
(spring/annual)
(%)
Fractional flow
(summer/annual)
(%)
Sub-basins Name
1001 Upstream of Namhan River 2,447.9 1,240 17.3 47.2
1002 Pyeongchang River 1,773.4 1,294 15.5 56.2
1003 Chungju Dam 2,483.8 1,210 16.2 54.1
1004 Dal Stream 1,614.4 1,174 15.0 54.4
1005 Downstream of Chungju Dam 524.4 1,202 15.8 59.0
1006 Seom River 1,491.0 1,298 11.5 60.7
1007 Downstream of Namhan River 2,072.7 1,297 12.6 61.4
1008 Kumgangsan Dam 2,983.0 NA 12.9 61.7
1009 Pyeongwha Dam 351.3 1,081 14.2 57.3
1010 Chuncheon Dam 1,587.4 1,186 16.4 62.5
1011 Inbook Stream 931.3 1,149 16.5 51.6
1012 Soyang River 1,852.0 1,251 17.7 55.1
1013 Euiam Dam 721.7 1,308 15.4 64.6
1014 Hongcheon River 1,566.0 1,302 16.3 63.0
1015 Cheongpyeong Dam 760.6 1,337 15.5 65.6
1016 Kyeongan Stream 561.1 1,266 16.1 62.5
1017 Paldang Dam 43.9 1,191 14.6 67.4
1018 Han River in Seoul 1,537.2 1,291 14.1 63.4
1019 Han River in Goyang 826.3 1,259 13.4 63.1
1020 Gomitan Stream 2,195.2 1,287 12.6 62.4
1021 Upstream of Imjin River 2,072.7 1,290 11.0 63.4
1022 Hantan River 2,452.2 1,292 11.4 62.9
1023 Downstream of Imjin River 1,419.2 1,316 10.8 64.4
1024 Downstream of Hantan River 146.4 1,306 11.5 60.7
Average 1,358.0 1,253.4 14.4 60.2
The hydrometeorological data (19662007) used here were obtained from the Water Management Information System (WAMIS)
NA represents that data is not available
1468 Stoch Environ Res Risk Assess (2013) 27:14651477
123
-
atmospheric circulation patterns. This dataset, which is
available on the Internet (URL: http://www.cpc.ncep.noaa.
gov/data/indices), has a spatial resolution of 1 9 1 and isupdated on a monthly basis (Rayner et al. 2003).
3 Methodology
The effect of different phases of ENSO events on the
spring and summer precipitation and runoff characteristics
was analyzed by adopting the classification criteria pro-
posed by Ren and Jin (2011) for hydrologic variables from
24 sub-basins. The traditional Nino3 and Nino4, which
were recently used to identify the two types of ENSO
events (Kug et al. 2009; Yeh et al. 2009), are highly cor-
related, which means that these indices are not optimal for
differentiating the two types of El Nino (i.e., WP and CT).
To overcome this, we used two new indices (NCT and NWP)
from a simple nonlinear transformation proposed by Ren
and Jin (2011) to distinguish a CT El Nino from a WP El
Nino. The equation is defined as follows:
NCT N3 aN4NWP N4 aN3
a 2=5;N3N4 [ 00; otherwise:
n1
Here, N3 and N4 indicate indices for Nino3 and Nino4,
respectively, and NCT and NWP are new indices based on a
simple transformation from Nino3 and Nino4 indices. A
comparison of Nino3 and Nino4 anomalies using a scatter
plot revealed a correlation of 0.77 between the two events.
However, a comparison between CT El Nino (NCT) and WP
El Nino (NWP) anomalies yielded a correlation of only 0.19.
We assessed the independence between Nino3 and Nino4
data using CT El Nino and WP El Nino indices that
showed a smaller relative correlation (Fig. 4). In addition,
we performed a correlation analysis using time series data
for Nino3 and CT El Nino indices (NCT) for the period
19502011, which afforded a correlation coefficient (q) of0.98 and standard deviation (r) of 0.6964. However,Nino4 shows a relatively weak correlation with NWPindex (q = 0.86, r = 0.40) compared to the results ofNino3 and NCT. To further examine local hydrologic
responses within the context of a seasonal and episodic
event, the five strongest events of CT El Nino type (1972/
1973, 1982/1983, 1986/1987, 1991/1992 and 1997/1998)
were identified using the NCT index. WP El Nino years
(1968/1969, 1990/1991, 1994/1995, 2002/2003 and
2004/2005) were selected for the five strongest events
using the Nwp index during the period 19502011. For
those five CT/WP El Nino years, we also examined the
evolution patterns of the sea surface temperature (SST)
over the western Pacific Oceans.
The variability in precipitation and runoff was analyzed
based on multiple flow metrics such as seasonal volume,
annual maximum flow (AMF), annual lowflow using the
coefficients of variation (CV), and percentage changes for
CT against WP for the five selected El Nino events. CV is
defined as the ratio of the standard deviation to the mean,
which shows the extent of variability in relation to the
mean. The percentage changes are compared to a standard
30-year climatology (19712000), AMF is defined as the
annual maximum flow, and lowflow was calculated from
the streamflow showing the lowest 7-day average flow
occurring each year.
(a) Seasonal fractional flow (Spring)
17.017.7%16.017.015.016.014.015.013.014.012.013.011.012.010.811.0
(c) Coefficient of variation (Spring)
0.800.830.750.800.700.750.650.700.600.650.550.600.500.550.440.50
(b) Seasonal fractional flow (Summer)
65.067.4%62.565.060.062.557.560.055.057.552.555.050.052.547.250.0
0.300.330.280.300.250.280.230.250.200.230.190.20
(d) Coefficient of variation (Summer)
Fig. 3 Percentage changes ofseasonal fractional flows and the
coefficients of variation during
a and c the spring season(March to May) and b and d thesummer season (June to August)
in the Han River Basin, Korea.
a and b The seasonal fractionalflows. c and d The coefficientsof variation (CV: standard
deviation/absolute mean)
Stoch Environ Res Risk Assess (2013) 27:14651477 1469
123
-
The empirical probability density function (PDF) was
also shown for the five El Nino events using the nonpara-
metric Kernel approach, which is widely used in theoretical
and applied statistics (Bowman and Azzalini 1997, 2007).
The Kernel density plot shows the optimal distribution for
the actual values of the sample without any assumptions of a
parent distribution. The main objective of using the Kernel
approach is to identify the sensitivity of changes in seasonal
precipitation and runoff anomalies (departures from the
19712000 normals) within the context of the CT/WP El
Nino events. In addition, composite analysis (CA), known as
the composite sampling technique, was applied to analyze
the effect of the different El Nino types on the precipitation
and runoff in the Han River Basin. A significance test using
the Monte Carlo re-sampling technique (Ripley 1987;
Becker et al. 1988) was performed for each seasonal variable
with a 90 % confidence interval.
4 Results
4.1 Characteristics of the SST patterns during CT/WP
El Ninos
Figure 5 shows the composite anomalies of the mean value of
SST from December to February, when El Ninos develop. In
the case of CT El Nino years, positive anomalies occurred in
the central and eastern Pacific areas, and negative anomalies
occurred throughout the western Pacific region (Fig. 5a). As
a CT El Nino evolved, the positive SST pattern extended
towards latitude, and the negative SST pattern extended
towards the east. The warm SST anomaly pattern reached its
maximum amplitude in the fall and winter, whereas it
disappeared in the summer decaying phase, and was replaced
by a SST anomaly pattern with a relatively low temperature
in the eastern Pacific region. However, the WP El Nino
evolution pattern showed a definite difference when com-
pared to the CT El Nino (Fig. 5b). In the WP El Nino years,
small positive SST anomalies first appeared in the western
Pacific region in the spring, and these patterns moved towards
the central Pacific direction. In the same period of time, a
negative anomaly was observed in the eastern and western
Pacific regions. A warm SST anomaly pattern was located in
the central Pacific Ocean, and this anomaly grew stronger and
extended to the east during the fall. In the wintertime, a
positive SST anomaly pattern strengthened, but as spring
arrived, the WP El Nino weakened and disappeared.
The main characteristics of the CT El Nino and the WP
El Nino can be summarized as follows: Firstly, in the CT
El Nino years, a warm SST anomaly pattern is located
mostly in the eastern Pacific region, while the WP El Nino
is in the central Pacific region. Secondly, the SST anomaly
pattern appears stronger in the CT El Nino years than in the
WP El Nino years. Thirdly, in the WP El Nino period, the
SST anomaly pattern tends to decline much more rapidly
than in the years of CT El Nino. Previously, we looked at
the evolving pattern of the CT/WP El Ninos, but the
present study attempted to analyze how hydrological flow
occurs due to different aspects that depend on changes in
the CT/WP El Ninos, rather than investigating the physical
mechanism of the two patterns [for a more detailed
explanation of CT/WP El Nino patterns, refer to Kug et al.
(2009) and Feng et al. (2010)]. These different SST
anomaly patterns may affect air circulation, thereby giving
rise to changes in seasonal precipitation and runoff in the
Pacific Rim countries.
(a) Correlation between N3 and N4 (b) Correlation between NCT and NWP
Fig. 4 Scatter plots a for N3 and N4 indices, and b for NCT and NWP indices. CORs denote correlations between two indices in each panel
1470 Stoch Environ Res Risk Assess (2013) 27:14651477
123
-
The spatial distributions of spring and summer precipi-
tation and runoff were remarkably different during the two
different types of El Nino events in the Han River Basin
(Fig. 6). During spring (Fig. 6a), the empirical PDF
showed a higher peak and lower variation for CT El Nino
events. This indicates that seasonal precipitation and runoff
are not sensitive to CT El Nino events. In contrast, the low
peak and large variation in the empirical PDF for precipi-
tation and runoff during WP El Nino events indicate that it
is becoming increasingly difficult to properly manage
water resources systems properly in this region. However,
in the figure shown (Fig. 6b), relatively lower variation was
illustrated for the WP El Nino periods compared to the CT
periods for the summer season.
In the following sections, we describe a finer-scale
analysis carried out to gain a better understanding of the
potential impacts of CT/WP El Ninos on variability in
seasonal precipitation and runoff over the sub-watersheds
in the Han River Basin.
4.2 Seasonal precipitation and runoff during CT/WP El
Ninos
The statistical estimates for CT/WP El Nino years are
summarized in Tables 2 and 3. Figure 7 shows an overall
increased pattern for the percentile average of spring pre-
cipitation in the range of 6.926.1 % for the 23 sub-basins,
with an average of 17.80 % during WP El Nino periods. As
is shown in Table 3, increases in spring precipitation in the
Han River Basin during WP El Nino events can be
expected, but the variation of spring precipitation is high,
so that a decreasing rainfall pattern can be found in some
sub-basins during the WP El Nino years. However, sig-
nificant increases in spring precipitation were observed in
only 2 sub-basins. It was found that precipitation was rel-
atively low for the summer season during the CT El Nino
years, but quite high for almost all sub-basins during the
WP El Nino years (Fig. 8). The summer precipitation
anomalies during WP El Nino years were -3.5 to 38.4 %,
(a) CT El Nio years
(b) WP El Nio years
Fig. 5 Composite sea surface temperature anomalies (SSTA) based on a standard 30-year climatology (19712000) for peak stages of CT andWP El Ninos during DecemberFebruary. a SSTA in the CT El Nino years. b SSTA in the WP El Nino years. (Color figure online)
Stoch Environ Res Risk Assess (2013) 27:14651477 1471
123
-
with an average of 21.8 %, indicating a strong increasing
pattern. For the WP El Nino years, 10 out of 23 sub-basins
showed a significant increase in summer precipitation
mostly occurring in a central-southern region of the Han
River Basin.
An analysis of runoff anomalies during spring showed
some decreases in 3 sub-basins (1006: Seom River, 1016:
Kyeongan Stream, 1017: Paldang Dam) during CT El Nino
periods, while statistically significant changes in spring
runoff were seen in 4 sub-basins (1010: Chuncheon Dam,
1011: Inbook Stream, 1013: Euiam Dam and 1020: Gom-
itan Stream) with an average increase of 47.64 % (Fig. 9).
However, summer runoff data showed a dry tendency, with
high variation during CT El Nino periods. Runoff increases
in summer were also found for most sub-basins during WP
El Nino periods, with statistically significant results for 14
sub-basins (58.8 % of the total area), but with less runoff
variation (Fig. 10). The summer runoff anomalies were
relatively large, with a coefficient of variation of 0.69
during the CT El Nino years, and with high variations,
particularly in the southern parts of the Han River Basin.
4.3 Episodic extreme events during CT/WP El Ninos
Figure 11 shows the change in percentile obtained by
comparing local distributions of annual maximum flow
(AMF) and annual lowflow occurring in the CT/WP El
Nino years. During the CT/WP El Nino years, the AMF
ranged from -61.8 to 53.2 % and showed a variety of local
changes. In the CT El Nino years (Fig. 11a), 71.5 % of the
Han River Basin appeared lower than the long-term nor-
mal, except for some parts of the western Han River Basin.
In six sub-basins, a statistically significant (p \ 0.10)decreasing pattern was noted in 28.6 % of the overall
basins associated with the Han River Basin. In contrast, in
the WP El Nino years (Fig. 11b), increasing patterns
occurred in the mid- and upper-streams in the Han River.
However, the Chungpyung Dam basin was the only basin
to show statically significant results, and the southern
region of the Han River showed a greater decrease than
was seen in long-term normal years. The changes in the
lowflow rate during CT El Nino years (Fig. 11c) were
the reverse of those of the AMF presented in Fig. 11a. The
positive pattern was identified in 23 sub-basins, and stati-
cally significant increasing values were present in 65.6 %
of the total area of the Han River Basin. However, in the
period of the WP El Nino (Fig. 11d), a general weakening
tendency was seen, except for in 4 sub-basins. A statisti-
cally significant decreasing patter occurred in 20.0 % of
total area of the Han River Basin.
5 Discussion and conclusion
During the WP El Nino years, a warm SST anomaly pattern
was located mostly in the central Pacific region, while CT
El Nino was located in the eastern Pacific region.
In the CT El Nino period, the SST anomaly pattern
tended to decline much more steadily than in the WP El
Nino years. These characteristics of the SST patterns are
consistent with the results from Kug et al. (2009) and Feng
et al. (2010).
The seasonal variability in precipitation and runoff over
the Korean Han River Basin for the CT/WP El Nino years
(a) Spring (MarchMay) (b) Summer (JuneAugust)
Percentage of seasonal composite anomaly to the 19712000 normal
Prob
abilit
y de
nsity
func
tion
300% 200% 100% 0 100% 200% 300%
0.000
0.005
0.010
0.015
0.020
0.025
0.030 Precipitation (CT)Precipitation (WP)Streamflow (CT)Streamflow (WP)
Percentage of seasonal composite anomaly to the 19712000 normal
Prob
abilit
y de
nsity
func
tion
300% 200% 100% 0 100% 200% 300%
0.000
0.005
0.010
0.015
0.020
0.025
0.030 Precipitation (CT)Precipitation (WP)Streamflow (CT)Streamflow (WP)
Fig. 6 Empirical probability density function for the seasonalprecipitation and the seasonal runoff by percentage of seasonal
composite anomalies during CT El Nino and WP El Nino over the
spring and summer season in the Han River Basin, Korea. The solid
blue line indicates the distribution of the seasonal precipitation in the
phase of the CT El Nino years and the solid red line indicates the
distribution in the phase of the WP El Nino years. The dotted blue line
indicates the distribution of the seasonal runoff in the phase of the CT
El Nino years and the dotted red line indicates this distribution in the
phase of the WP El Nino years. (Color figure online)
1472 Stoch Environ Res Risk Assess (2013) 27:14651477
123
-
can be summarized as follows: Firstly, during the WP El
Nino years, precipitation in spring and its CV were higher
than long-term normal precipitation. Secondly, in the
CT El Nino years, summers tended to be relatively
drier than in climatologically normal years, although the
variability in precipitation during the summer was lower.
Lastly, significant increases in summer precipitation and
runoff were seen in the WP El Nino years. These increases
mostly occurred in a central-southern region of the Han
River Basin. However, summer runoff data showed a rel-
atively dry tendency, with high variation during CT El
Nino periods.
Table 2 Percentage anomalies in the Han River basins during CT/WP El Ninos
ID CT El Nino years WP El Nino years
Spring (MarchMay) Summer (JuneAugust) Spring (MarchMay) Summer (JuneAugust)
Precipitation Runoff Precipitation Runoff Precipitation Runoff Precipitation Runoff
1001 -8.70 17.31 -2.00 -3.12 19.80 38.14 9.80 23.93
1002 -3.80 5.17 3.10 1.49 17.00 44.49 21.30 42.72
1003 2.30 25.72 0.00 16.31 16.20 49.72 16.10 25.15
1004 7.50 16.63 8.30 16.36 14.00 34.15 26.40 55.49
1005 7.60 6.94 3.90 4.54 25.00 39.21 24.10 45.53
1006 1.40 -0.46 5.20 4.51 17.70 67.17 23.90 53.79
1007 3.20 4.60 6.80 11.03 21.20 52.53 21.80 39.84
1008 NA 35.48 NA 15.02 NA 15.02 NA 40.23
1009 -3.70 28.63 -0.60 -10.12 24.30 49.78 38.40 66.64
1010 4.50 40.33 2.00 -1.28 20.60 28.18 29.20 44.41
1011 1.70 47.50 1.00 0.38 26.10 41.40 33.40 54.90
1012 -2.10 24.04 1.60 -1.42 12.20 22.44 23.40 35.96
1013 14.90 50.77 10.00 10.06 10.90 22.00 20.50 26.14
1014 0.40 22.59 0.50 -1.60 21.00 41.17 25.10 38.33
1015 0.30 26.14 14.40 18.21 12.20 16.50 23.00 33.61
1016 -6.50 -0.24 6.50 7.50 17.60 40.34 30.30 45.78
1017 -14.80 -7.19 18.00 23.94 6.90 -7.70 36.50 55.51
1018 5.80 16.06 15.00 22.61 12.10 16.78 14.70 18.01
1019 2.90 10.16 7.30 8.39 20.00 43.89 19.50 30.74
1020 20.20 51.97 -3.80 -10.40 21.70 40.79 19.50 33.82
1021 18.10 41.46 -2.20 -7.86 20.70 31.84 17.20 27.66
1022 14.50 29.07 5.00 2.73 20.60 28.94 23.00 34.21
1023 8.00 14.02 12.70 15.96 15.00 40.90 6.70 7.76
1024 6.10 17.28 15.80 21.21 16.70 41.05 -3.50 -6.87
NA represents that data is not available
Statistical estimates shown in boldface are statistically significant at the 10 % level
Table 3 Summary statistics for seasonal variability in the Han River Basin
El Nino type Statistics Spring (MarchMay) Summer (JuneAugust)
Precipitation Runoff Precipitation Runoff
CT El Nino Average change (%) 3.47 21.83 5.59 6.85
Significant stations (p \ 0.10) 0/23 4/23 0/24 0/24CV 0.20 0.30 0.44 0.69
WP El Nino Average change (%) 17.80 36.51 21.75 36.39
Significant stations (p \ 0.10) 2/23 8/23 10/24 14/24CV 0.54 0.89 0.24 0.30
CV the ratio of standard deviation to mean
Stoch Environ Res Risk Assess (2013) 27:14651477 1473
123
-
Yeh et al. (2009) reported an increase in the frequency
of WP El Nino events during recent years, in association
with global warming, and stated that the projected global
warming scenarios indicate that these events will become
more frequent. They suggested that this new type of event
is currently affecting the Korean peninsula as well as other
East Asian regions. However, current observational studies
are highly dependent on the size of the available sample;
hence, it remains to be verified whether the recent trend
toward the WP type of El Nino is a sign of climate change
in the region. During the WP El Nino years, a large-scale
ascending motion was observed over the central Pacific and
a sinking motion was noted over the western and eastern
Pacific region associated with the Walker circulation.
However, during the CT El Nino years, an anomalous
rising motion appears over the eastern pacific and a
descending motion over the western Pacific region (Feng
et al. 2010). Furthermore, the location and intensity of the
anomalous western north Pacific (WNP) cyclone or anti-
cyclone forced by different SST warming associated with
NA NA
(a) CT El Nio years (Spring Rainfall) (b) WP El Nio years (Spring Rainfall)
25% ~ 26.1%20 ~ 2515 ~ 2010 ~ 15 5 ~ 10 0 ~ 55 ~ 010 ~ 514.8 ~ 10
NA NA
(c) CT El Nio (Coefficient of Variation) (d) WP El Nio (Coefficient of Variation)
0.70 ~ 0.780.60 ~ 0.700.50 ~ 0.600.40 ~ 0.500.30 ~ 0.400.20 ~ 0.300.10 ~ 0.200.08 ~ 0.10
Fig. 7 Percentage changes ofspring precipitation and
coefficients of variation in
composite anomalies
(departures from the 19712000
normals) during a and c CT ElNino years and b and d WP ElNino years in the Han River
Basin, Korea. The effects of
both phases of ENSO are shown
with different color schemes
(increases in blues and
decreases in reds). The hatched
polygon indicates statistically
significance in seasonal
precipitation (March to May) at
the 10 % confidence level. NA
indicates that data are not
available. (Color figure online)
NA NA
(a) CT El Nio years (Summer Rainfall) (b) WP El Nio years (Summer Rainfall)
35% ~ 38.4%30 ~ 3525 ~ 3020 ~ 2515 ~ 2010 ~ 15 5 ~ 10 0 ~ 53.8 ~ 0
NA NA
(c) CT El Nio (Coefficient of Variation) (d) WP El Nio (Coefficient of Variation)
0.55 ~ 0.590.50 ~ 0.550.45 ~ 0.500.40 ~ 0.450.35 ~ 0.400.30 ~ 0.350.25 ~ 0.300.20 ~ 0.250.12 ~ 0.20
Fig. 8 Percentage changes ofsummer precipitation and
coefficients of variation in
composite anomalies
(departures from the 19712000
normals) during a and c CT ElNino years and b and d WP ElNino years in the Han River
Basin, Korea. The effects of
both phases of ENSO are shown
with different color schemes
(increases in blues and
decreases in reds). The hatched
polygon indicates statistically
significance in seasonal
precipitation (June to August) at
the 10 % confidence level. NA
indicates that data are not
available. (Color figure online)
1474 Stoch Environ Res Risk Assess (2013) 27:14651477
123
-
the CT/WP El Nino could have potential implications for
the regional moisture transport in many countries bordering
the Pacific rim (Wang et al. 2000; Weng et al. 2007; Lee
et al. 2009; Feng et al. 2010).
The structure and function of stream and riparian eco-
systems is closely associated with the natural hydrologic
regime. Therefore, the effects of episodic, seasonal, or
long-term climate changes on the hydrologic regime must
be clarified in order to gain a better understanding of the
interaction between human and environmental systems. In
this study, we aimed to relate the sensitivity of relevant
flow metrics to climate variability and extremes in the
Korean Han River Basin. We found that water resources
during the spring and summer in the Han River Basin are
very sensitive to WP El Nino events, but not as sensitive to
CT El Nino events. However, it should be noted that this
study is based on relatively short observations and limited
conditions.
Our results indicate that studies on the efficient predic-
tion and management of water resources are urgently
(a) CT El Nio years (Spring Streamflow) (b) WP El Nio years (Spring Streamflow)
60% ~ 67.2%50 ~ 6040 ~ 5030 ~ 4020 ~ 3010 ~ 20 0 ~ 107.7 ~ 0
(c) CT El Nio (Coefficient of Variation) (d) WP El Nio (Coefficient of Variation)
1.40 ~ 1.561.20 ~ 1.401.00 ~ 1.200.80 ~ 1.000.60 ~ 0.800.40 ~ 0.600.20 ~ 0.400.12 ~ 0.20
Fig. 9 Percentage changes ofspring runoff and coefficients of
variation in composite
anomalies (departures from the
19712000 normals) during
a and c CT El Nino years andb and d WP El Nino years in theHan River Basin, Korea. The
effects of both phases of ENSO
are shown with different color
schemes (increases in blues and
decreases in reds). The hatched
polygon indicates statistically
significance in seasonal runoff
(March to May) at the 10 %
confidence level. Note NA
indicates that data are not
available. (Color figure online)
(a) CT El Nio years (Summer Streamflow) (b) WP El Nio years (Summer Streamflow)
35% ~ 38.4%30 ~ 3525 ~ 3020 ~ 2515 ~ 2010 ~ 15 5 ~ 10 0 ~ 53.8 ~ 0
(c) CT El Nio (Coefficient of Variation) (d) WP El Nio (Coefficient of Variation)0.90 ~ 1.040.80 ~ 0.900.70 ~ 0.800.60 ~ 0.700.50 ~ 0.600.40 ~ 0.500.30 ~ 0.400.20 ~ 0.300.16 ~ 0.20
Fig. 10 Percentage changes ofsummer runoff and coefficients
of variation in composite
anomalies (departures from the
19712000 normals) during
a and c CT El Nino years andb and d WP El Nino years in theHan River Basin, Korea. The
effects of both phases of ENSO
are shown with different color
schemes (increases in blues and
decreases in reds). The hatched
polygon indicates statistically
significance in seasonal runoff
(June to August) at the 10 %
confidence level. Note NA
indicates that data are not
available. (Color figure online)
Stoch Environ Res Risk Assess (2013) 27:14651477 1475
123
-
required. This includes the study of natural disasters such
as drought and flood in basins like the Han River Basin,
where seasonal variations and uncertainties are high.
Investigation of the correlations between climatic factors
and hydrologic variables are also required to establish
stable water supplies and realistically adaptable strategies
for resilience against extreme events in a changing climate.
Further examination and understanding of the synergistic
impacts of large-scale atmospheric circulations and the
evolution of the emerging El Nino conditions on local
climate and hydrology will require additional studies using
physical mechanism-based numerical simulations to reduce
the uncertainties.
Acknowledgments This CRI work was supported by the NationalResearch Foundation of Korea (NRF) grant funded by the Korea
government (2012R1A1A2005304) and also funded by The Korea
Meteorological Administration Research and Development Program
under grant CATER 2012-3100 (Development of Drought Outlook &
Response Techniques on Korea and East Asia Region). The authors
thank the two anonymous reviewers for their comments and valuable
suggestions.
References
Ashok K, Yamagata T (2009) Climate change: the El Nino with a
difference. Nature. doi:10.1038/461481a
Ashok K, Behera SK, Rao SA, Weng H, Yamagata T (2007) El Nino
Modoki and its possible teleconnection. J Geophys Res
112:C11007. doi:10.1029/2006JC003798
Bae DH, Jung IW, Chang HJ (2008) Long-term trend of precipitation
and runoff in Korean river basins. Hydrol Process 22:26442656
Becker RA, Chambers JM, Wilks AR (1988) The new S language.
Wadsworth & Brooks/Cole, USA
Bowman AW, Azzalini A (1997) Applied smoothing techniques for
data analysis: the Kernel approach with S-plus illustrations.
Oxford University Press, Oxford
Bowman AW, Azzalini A (2007) R Package Sm: nonparametric
smoothing methods (version 2.2). University of Glasgow, UK
and Universita di padova, Italia
Chang HJ (2008) Spatial analysis of water quality trends in the Han
River basin, South Korea. Water Res 42:32853304
Chang HJ, Jung IW (2010) Spatial and temporal changes in runoff
caused by climate change in a complex large river basin in
Oregon. J Hydrol 388:186207
Chang CWJ, Hsu HH, Wu CR, Sheu WJ (2008) Interannual mode of sea
level in the South China Sea and the roles of El Nino and El Nino
Modoki. Geophys Res Lett 35:L03601. doi:10.1029/2007GL032562
(a) AMF (CT El Nio years) (b) AMF (WP El Nio years)
(c) Lowflow (CT El Nio years) (d) Lowflow (WP El Nio years)
Fig. 11 Percentage changes in composite anomalies (departures fromthe 19712000 normals) of annual maximum flow (AMF) and annual
7-day lowflow in the Han River Basin during the CT/WP Nino years.
a AMF (CT El Nino years). b AMF (WP El Nino years). c Lowflow
(CT El Nino years). d Lowflow (WP El Nino years). The hatchedpolygons indicate statistically significance at the 10 % confidence
level
1476 Stoch Environ Res Risk Assess (2013) 27:14651477
123
-
Dressler KA, Leavesley GH, Bales RC, Fassnacht SR (2006)
Evaluation of gridded snow water equivalent and satellite snow
cover products for mountain basins in a hydrologic model.
Hydrol Process 20:673688
Feng J, Chen W, Tam CY, Zhou W (2010) Different impacts of El
Nino and El Nino Modoki on China rainfall in the decaying
phases. Int J Climatol. doi:10.1002/joc.2217
Gershunov A, Barnett TB, Cayan DR (1999) North Pacific interdec-
adal oscillations seen as factor in ENSO-related North American
climate anomalies. Eos Trans Am Geophys Union. doi:10.1029/
99EO00019
Grimm AM (2011) Interannual climate variability in South America:
impacts on seasonal precipitation, extreme events, and possible
effects of climate change. Stoch Environ Res Risk Assess
25:537554
Horel JD, Wallace JM (1981) Planetary-scale atmospheric phenom-
ena associated with the Southern Oscillation. Mon Weather Rev
109:813829
Jin YH, Kawamura A, Jinno K, Berndtsson R (2005) Detection of
ENSO-influence on the monthly precipitation in South Korea.
Hydrol Process 19:40814092
Kao HY, Yu JY (2009) Contrasting Eastern-Pacific and Central-
Pacific Types of ENSO. J Clim 22:615632. doi:10.1175/2008
JCLI2309.1
Kim JS, Jain S (2011) Precipitation trends over the Korean peninsula:
typhoon-induced changes and a typology to characterize climate-
related risk. Environ Res Lett 6:034033
Kim TW, Yoo CS, Ahn JH (2008) Influence of climate variation on
seasonal precipitation in the Colorado River Basin. Stoch
Environ Res Risk Assess 22:411420
Kim JS, Jain S, Moon YL (2011) Atmospheric teleconnection-based
conditional streamflow distributions for the Han River and its
sub-watersheds in Korea. Int J Climatol. doi:10.1002/joc.2374
Kim DW, Choi KS, Byun HR (2012a) Effects of El Nino Modoki on
winter precipitation in Korea. Clim Dyn 38:13131324
Kim JS, Jain S, Yoon SK (2012b) Warm season streamflow
variability in the Korean Han River Basin: links with atmo-
spheric teleconnections. Int J Climatol 32(4):635640
Kim JS, Li RCY, Zhou W (2012c) Effects of the Pacific-Japan
teleconnection pattern on tropical cyclone activity and extreme
events over the Korean peninsula. J Geophys Res 117:D18109
Kug JS, Jin FF, An SI (2009) Two types of El Nino events: cold
tongue El Nino and warm pool El Nino. J Clim 22:14991515.
doi:10.1175/2008JCLI2624.1
Kwon HH, Lall U, Obeysekera J (2009) Simulation of daily rainfall
scenarios with interannual and multidecadal climate cycles for
South Florida. Stoch Environ Res Risk Assess 7:879896
Leavesley GH, Lichty RW, Troutman BM, Saindon LG (1983)
Precipitation-Runoff modeling system. Users manual. Water
Resources Investigations: 83-4238 US Geological Survey,
Reston
Lee SM, Byun HR (2009) Some causes of the May drought over
Korea. Asia-Pacific J Atmos Sci 45(3):247264
Lee SK, Wang C, Mapes B (2009) A simple atmospheric model of the
local and teleconnection response to heating anomalies. J Clim
22:272284
McPhaden MJ, Zebiak SE, Glantz MH (2006) ENSO as an integrating
concept in earth science. Science 314:17401745
Na H, Jang BG, Choi WM, Kim KY (2011) Statistical simulations of
the future 50-year statistics of cold-tongue El Nino and Warm-
Pool El Nino. Asia-Pacific J Atmos Sci 47(3):223233
NOAA (national weather service climate prediction center) (2012)
http://www.cpc.ncep.noaa.gov/data/indices/. Accessed Jan 2012
Piechota TC, Dracup JA (1996) Drought and regional hydrologic
variation in the United States: associations with the El Nino-
Southern oscillation. Water Resour Res 32(5):13591373
Piechota TC, Chiew Francis HS, Dracup JA, McMachon TA (1998)
Seasonal streamflow forecasting in eastern Australia and the El
Nino-Southern Oscillation. Water Resour Res 34(11):30353044
Pizarro G, Lall U (2002) El Nino-induced flooding in the US West:
what can we expect? EOS Trans Am Geophys Union 83:
349352
Pradhan PK, Preethi B, Ashok K, Krishna R, Sahai AK (2011)
Modoki, Indian Ocean Dipole, and western North Pacific
typhoons: possible implications for extreme events. J Geophys
Res 116:D18108. doi:10.1029/2011JD015666
Rayner NA, Parker DE, Horton EB, Folland CK, Alexander LV,
Rowell DP, Kent EC, Kaplan A (2003) Global analyses of sea
surface temperature, sea ice, and night marine air temperature
since the late nineteenth century. J Geophys Res. doi:10.1029/
2002JD002670
Ren HL, Jin FF (2011) Nino indices for two types of ENSO. Geophys
Res Lett 38:L04704. doi:10.1029/2010GL046031
Ripley BD (1987) Stochastic simulation. Wiley
Schonher T, Nicholson SE (1989) The relationship between Califor-
nia rainfall and ENSO events. J Clim 2:12581269
She D, Xia J (2012) The spatial and temporal analysis of dry spells in
the Yellow River basin, China. Stoch Environ Res Risk Assess.
doi:10.1007/s00477-011-0553-x
Trenberth KE (1997) The definition of El Nino. National Center for
Atmospheric Research. Bulletin of the American Meteorological
Society, pp 27712777
WAMIS (water management information system) (2012) http://
wamis.go.kr/eng/. Accessed Jan 2012
Wang B, Wu R, Fu X (2000) PacificEast Asian teleconnection: how
does ENSO affect East Asian climate? J Clim 13:15171536
Wang X, Wang D, Zhou W, Li CY (2011) Interdecadal modulation of
the influence of La Nina events on mei-yu rainfall over the
Yangtze River Valley. Adv Atmos Sci 29(1):157168. doi:
10.1007/s00376-011-1021-8
Weng H, Ashok K, Behera S, Rao S, Yamagata T (2007) Impacts of
recent El Nino Modoki on dry/wet conditions in the Pacific rim
during boreal summer. Clim Dyn 29:113129
Wu R, Hu ZZ, Kirtman BP (2003) Evolution of ENSO-related rainfall
anomalies in East Asia. J Clim 16:37423758
Yeh SW, Kug JS, Dewitee B, Kwon MH, Kirtman BP, Jin FF (2009)
El Nino in a changing climate. Nature 461:511514. doi:
10.1038/nature08316
Stoch Environ Res Risk Assess (2013) 27:14651477 1477
123
Hydrometeorological variability in the Korean Han River Basin and its sub-watersheds during different El Nio phasesAbstractIntroductionStudy area and dataStudy areaData
MethodologyResultsCharacteristics of the SST patterns during CT/WP El NiosSeasonal precipitation and runoff during CT/WP El NiosEpisodic extreme events during CT/WP El Nios
Discussion and conclusionAcknowledgmentsReferences