INTERNATIONAL JOURNAL OF CLIMATOLOGY
Int. J. Climatol. 25: 99–116 (2005)
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/joc.1116
ON INTERANNUAL CHARACTERISTICS OF CLIMATE PREDICTION
CENTER MERGED ANALYSIS PRECIPITATION OVER THE KOREAN
PENINSULA DURING THE SUMMER MONSOON SEASON
KYUNG-JA HA,* SUNG-KYU PARK and KI-YOUNG KIM
Department of Atmospheric Sciences, Pusan National University, Busan 609-735, South Korea
Received 12 September 2003
Revised 16 August 2004
Accepted 16 August 2004
ABSTRACT
To improve understanding of the characteristics of large-scale environments associated with summer monsoon precipitation
in Korea, we have constructed an index to measure the Changma using area-averaged and period-averaged Climate
Prediction Center merged analysis precipitation over Korea during the years 1979–2001. Strong and weak Changma
years are defined as years in which the precipitation anomaly is over ±0.5σ . The domain and period for this analysis are
selected over the Korean peninsula (31.25–41.25 ° N, 123.75–131.25 ° E) from late June to late July. In strong Changma
years, strong upper level divergence occurs near the Korean peninsula, located at the entrance of the mid-latitude jet core,
leading to low pressure in the lower layer. Between strong and weak Changma years, 850 hPa southwesterly flows are
significant throughout the East China Sea and southern Japan.
Clear differences exist between strong and weak Changma years. Weak Changma is well correlated with a weak Pacific
high, as seen in the changes in sea-level pressure and 500 hPa geopotential height, and wind speed in the lower level jet.
Strong Changma is well correlated with strong migratory highs in northern parts of the Korean peninsula. It is shown
that the long-term changes in summer monsoon rainfall, such as those between the two climate regimes of 1979–92 and
1993–2001, are characterized by a shift of maximum rainfall periods from late July to mid August. Copyright  2005
Royal Meteorological Society.
KEY WORDS:
Changma rainfall index; dynamical monsoon index; interannual characteristics; long-term changes
1. INTRODUCTION
Summer monsoon rainfall prediction is of great importance in the Far East Asian monsoon region. Monsoon
rainfall, defined as the seasonal mean or intraseasonal mean rainfall, plays an important role in affecting water
resources and agricultural affairs. Since ancient times, summer monsoon rainfall has been variously referred to
as ‘Baiu’ (Japanese), ‘Meiyu’ (Chinese), and ‘Changma’ (Korean), all of which possess characteristics of an
annual march and synoptic activities. They are characterized by frontal rain systems between the two major
anticyclonic circulations over the subtropics and the polar regions. Many articles exist on the association
between local summer monsoon rainfall and the Asian monsoon. The climatological intraseasonal oscillation
(CISO) of the Asian summer monsoon has large regional differences. The dominant periodicities of CISO
are about 30–40 days over both the Indian Ocean and the western Pacific (Lau and Chan, 1986; Wang and
Xu, 1997) and about 60 days over the extratropics in East Asia and northern India (Wang and Xu, 1997).
On the other hand, Kang et al. (1999) investigated the principal mode of Changma from the perspective of
climatological variation of the Asian summer monsoon.
* Correspondence to: Kyung-Ja Ha, Department of Atmospheric Sciences, Pusan National University, Busan 609-735, South Korea;
e-mail: [email protected]
Copyright  2005 Royal Meteorological Society
100
K.-J. HA, S.-K. PARK AND K.-Y. KIM
Previous studies on the interannual variability of the East Asian summer monsoon have emphasized the
thermal effect of the Tibetan Plateau (Nitta, 1983; Luo and Yanai, 1984). The thermal state and convective
activities over the western Pacific warm pool are also considered to play an important role in the interannual
variability of the East Asian summer monsoon (Nitta, 1987). Also, the pronounced year-to-year variability is
partially explained by El Niño and the southern oscillation (ENSO; Zhang et al., 1996). However, the linkage
between the Far East Asian monsoon rainfall and ENSO still remains unclear.
Numerous studies also exist on the decadal variability of the East Asian summer monsoon, and many
of these studies have investigated the summer rainfall over central China and southern Japan (Yatagai and
Yasunari, 1994; Nitta and Hu, 1996). Ho et al. (2003) have examined long-term climate change in Korea.
Apparently, the cooling trend of the surface temperature in Mongolia due to deforestation and expansion of
the arid area (Wang and Li, 1990; Xue, 1996) might be associated with the climate change (Ho et al., 2003).
It is well known that Changma in Korea is accompanied by the northward penetration of the subtropical
Pacific high to the south of the Korean peninsula (Lim, 1997). Oh et al. (2000) found that, at the onset and
retreat times of Changma, the geopotential height increases rapidly over the Korean peninsula and southern
Japan, and that the increase in geopotential height in the mid–high latitudes persists for several days prior
to the Changma onset. Lu et al. (2001) have also emphasized that Changma retreats are associated with
planetary-scale teleconnection patterns of circulation in the mid–high latitudes.
Because of the importance of predictability of summer monsoon rainfall, our attention should revert back
to the understanding of year-to-year variability. Many studies (Webster and Yang, 1992; Goswami et al.,
1999; Lau et al., 2000) about the variability of the Asian summer monsoon have mostly targeted South Asia,
East–Southeast Asia, and the Indo-Pacific region. Changma in Korea has been poorly understood on an
interannual time scale. The purpose of the present study is to provide a more comprehensive understanding
of the interannual variability of summer monsoon rainfall over Korea and the causes of this in the large-scale
circulations. To accomplish our objective, we will define the interannual characteristics of mean rainfall during
Changma with the use of climatic analysis data. In particular, we have recently experienced record-breaking
torrential rainfall during the Changma season or before/after Changma. However, it is still uncertain whether
these extreme events are episodic or/and a part of long-term variation in the climatic regime. We attempt
to investigate the characteristics of the long-term changes, such as decadal variations in circulation states
associated with precipitation.
2. METHOD
2.1. Data
In this study, summer monsoon rainfall is measured by the 5 day (pentad) mean of the Climate Prediction
Center merged analysis of precipitation (CMAP) data, for the years 1979–2001. CMAP is known as a
representative data set that combines ground-based rain gauges and satellite-based observations (Xie and
Arkin, 1997). Even though the gridded pentad rainfall has been examined as being useful for diagnostics
(Xie et al. 2003), we used the CMAP 10 day mean rainfall data to obtain the interannual variability, avoiding
the optional selection of ground station data. The data used for the large-scale characteristics also include
the sea-level pressure (SLP), geopotential height at 700, 500 and 200 hPa, wind at 850 and 200 hPa, and
specific humidity at 850 hPa from the National Centers for Environmental Prediction–National Center for
Atmospheric Research (NCEP–NCAR) reanalysis. These NCEP–NCAR reanalysis data and CMAP data have
the same horizontal resolution of 2.5° × 2.5° longitude–latitude.
To examine the characteristics of the interannual variability and long-term variability of the summer
monsoon rainfall over the Korean peninsula, we performed composite analyses for strong versus weak
Changma years, later versus earlier regime, and long-lasting versus short-lasting Changma years. We also
performed a Student’s t-test to obtain the significance for composite differences. To show the decadal variation,
we only analysed rain-gauge information from ground-based stations.
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
101
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
2.2. Changma rainfall index
The development of the monsoon rainband associated with Meiyu in China and Baiu in Japan usually
occurs around mid June, whereas Changma in Korea starts in late June (Ho and Kang, 1988). The transient
evolution of the monsoon rainband in northeast Asia is presented by a maximum rain axis over 8 mm/day, as
shown in Figure 1. In relation to the climatological mean of 1979–2001, the northward progress of a 10 day
maximum rain is obviously seen from early June to late July from the south of Japan to northern Korea.
This northward progress is well matched to the progress or intensity of the North Pacific high (NPH) and the
Okhotsk high. In mean SLP fields during mid July, the maximum rain axis occurs in the trough between the
NPH and the Okhotsk high (refer to the thin line in Figure 1).
In late June, when the maximum rain axis lies in Kyushu in Japan, Jeju island, and the East China Sea,
the rainfall in southern Korea increases. Migrating highs from the interior of China might also be correlated
with the progress of the maximum rain axis after mid July. During the transition from early July to mid
July, there is a distinct phase change when maximum rain axes migrate to the north. Although interannual
anomalies are often highly episodic in time and in space, it is obvious that monsoon flows and Baiu fronts
influence the initiation and characteristics of the Korean Changma rainfall. During early July, the maximum
rain axis is separated into two lines. A broken maximum rain axis during early July may imply a complex
association to either Meiyu or Baiu. After late July, the maximum rain axis of R > 8 mm/day is weakened
and disappears. It is apparent that the onset and retreat of maximum rain axes for CMAP are beneficial to
define the Changma frontal rain index.
From late June to late July, the influence of maximum rain axes appears around Korea. We define this
period as the Changma season and the area influenced as the Changma area (rectangular box in Figure 1),
and then define the Changma rainfall index (CRI) as follows:
CRI = area average of (late June–late July 10 day precipitation rate, over lon
= 123.75–131.25 ° E, lat = 31.25–41.25 ° N)
50N
45N
40N
35N
30N
25N
20N
15N
10N
100E
105E
110E
115E
120E
125E
130E
135E
140E
145E
150E
155E
Figure 1. Schematic diagram showing the maximum rain axes in rainbands shown by CMAP data around the Korean peninsula from
early June to late July. The solid lines represent the axes of the maximum rainbands, R > 8 mm/day for 10 day mean precipitation.
Mean SLP during mid July is shown in the background contours. The rectangular box over the Korean peninsula indicates the area
influenced by the maximum rain axes from late June to late July (E: early; M: middle; L: late)
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
102
K.-J. HA, S.-K. PARK AND K.-Y. KIM
Close scrutiny of the intraseasonal variation of the maximum rain axis and the maximum axis (over 6 m/s) of
wind speed at 850 hPa (solid line in Figure 2) shows that the variations of rain axes in the monsoon season
closely resemble those of low-level winds, indicating moisture transport from a warm temperature pool. The
850 hPa wind maximum axis is observed at the trough axis between the warm NPH and the cold Okhotsk
high. As expected, from the 500 hPa geopotential height, the spatial pattern of 10 day 5850 gpm for the NPH
(shown as a dashed line in Figure 2) is also very similar to that of the intraseasonal variation of the maximum
rain axis. Thus, the characteristic variation of the maximum rain axis for the 10 day mean is well correlated
to the Asian monsoon flow from the climatological point of view.
2.3. Interannual variability
We now investigate the interannual variations of Changma rain. A CRI count for Changma rainfall is shown
in Figure 3. The mean CRI = 7.1 mm/day, and the strong and the weak Changma years are defined as years
V850
585dam500
50N
45N
L7
M7
40N
L7
L7
E7
35N
M7
30N
25N
M7
M6
20N
E7
E7 L6
E6
M6
L6
M6
E6
15N
10N
80E
85E
90E
95E 100E 105E 110E 115E 120E 125E 130E 135E 140E 145E 150E 155E 160E 165E 170E
Figure 2. Time evolution patterns of the maximum wind axis (solid) over |V | > 6 m/s in 10 day mean 850 hPa wind field and 5850 gpm
(dashed) contour of 10 day 500 hPa NPH during the boreal summer season from early June to late July
Rainfall anomaly (mm/day)
3
2
1
0.5σ
0
-1
-0.5σ
-2
-3
-4
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01
Year (1979−2001)
Figure 3. Interannual variation of the area (31.25–41.25 ° N, 123.75–131.25 ° E) and period (late June–late July) mean Changma rainfall
anomaly. Years more than 0.5σ and less than −0.5σ are defined as strong and weak Changma years respectively
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
103
in which the precipitation anomaly is more than ±0.5σ . There are 6 years and 7 years respectively in the
categories of weak and strong years. The CRI shows that weak years are mostly seen in the early period, and
strong years in the late period.
2.4. Empirical orthogonal function analysis of the summer monsoon rainfall
The monsoon variations are characterized by several major time scales associated with seasonal marches and
intraseasonal oscillations (Wang and Xu, 1997). To describe the climatological variations, empirical orthogonal
function (EOF) analysis is applied to the climatological pentad CMAP data for summertime (Figure 4). The
results of the EOF analyses are similar to those applied to pentad high cloud fractions obtained by Kang et al.
(1999), except for a more southward shift of the spatial pattern in the eigenvector. The first mode shows an
annual mode. In the second eigenvector (Figure 4(c)), the zonally elongated pattern of positive values from
central China to the North Pacific is very similar to the East Asian rainband associated with Meiyu and Baiu.
The time series in Figure 4(d) shows that the second mode develops during early summer. The second mode
is assumed to be a monsoon mode. The third eigenvector, shown in Figure 4(e), is characterized by positive
TIME SERIES
EIGENVECTOR
(a) 1st mode (47.8%)
(b) 1st mode
0.6
50N
0.4
40N
0.2
30N
0
20N
−0.2
10N
−0.4
−0.6
80E
100E
120E
140E
160E
180
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
(c) 2st mode (17.2%)
(d) 2nd mode
50N
0.6
40N
0.4
0.2
30N
0
20N
−0.2
10N
−0.4
80E
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
(e) 3rd mode (8.3%)
(f) 3rd mode
50N
0.6
0.4
40N
0.2
30N
0
20N
−0.2
10N
−0.4
80E
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
Figure 4. First three leading eigenvectors and associated time series obtained from the EOF analysis of the 5 day mean climatological
variation of CMAP from May to August during 1979–2001. Shaded area represents positive region
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
104
K.-J. HA, S.-K. PARK AND K.-Y. KIM
values over a zonal band crossing the Korean peninsula, and the associated time series (Figure 4(f)) has
positive values from mid June to mid July. Both the spatial pattern and associated time series indicate that
the third eigenmode appears to be related to the Changma in Korea. Therefore, Changma can be represented
by the third mode in this study.
Results of EOF analyses in strong and weak Changma years are shown in Figure 5. In strong years, there
are larger positive values from east China to Japan in the third eigenvector, compared with those in weak
years. From the associated time series of the third eigenvector, the retreat of Changma is late in strong years
compared with weak years. Similar results are shown in the time series for the second eigenvector. The third
mode in weak years explains a greater fraction of the total variance than that in strong years. On the other
hand, Changma rainfall is positively correlated with the rainfall in the Indian monsoon region, especially in
the northern part of the Indian continent, and negatively correlated with the rainfall near the South China Sea
(SCS; 10 ° N, 120 ° E). Compared with the third eigenvector in all years (Figure 4(e)), it is found that there is
a marked contrast in the axis of strong positive values of the third eigenvector over the East China Sea for
the strong and weak Changma years.
3. COMPOSITE DIFFERENCES
To examine the characteristics of synoptic fields associated with interannual variability of the monsoon
rainfall over the Korean peninsula, we performed composite analyses for strong versus weak Changma years
and composite differences with various variables. As a result, the CRI difference is characterized by the SLP,
500 hPa geopotential height, and winds at 200 and 850 hPa.
3.1. SLP and upper level geopotential height
Figure 6(a) and (b) shows the SLP composites for the strong Changma years and the strong minus weak
Changma years. The shaded areas show the regions significant at the 95% confidence level. We used the
Student t-test to assess the significance of the composite difference of mean strong and weak Changma years.
This test has assumptions about the independence and randomness of the data, and good validation is required
when there are fewer than 30 observations.
For the difference in SLP between the strong and weak years, significant positive differences are seen
in eastern Mongolia. A trough is located between these two positive SLP centres due to the NPH and the
migrating high over northeastern China. Rectangular squares in these two higher SLPs are considered for the
Changma index in Section 4 and Table I.
Figure 6(c) and (d) is the same as Figure 6(a) and (b) but shows the 500 hPa geopotential height. There is
a strong and deep trough axis over the Bering Sea and the Sea of Okhotsk, and a trough over the west of
Korea in strong years. From the difference field (Figure 6(d)), the most important thing is the higher NPH
in the northwest Pacific. Compared with the SLP pattern of Figure 6(b), the 500 hPa geopotential height is
dominant over the NPH and that over the north of Korea is shallow.
3.2. Winds
We show the 200 hPa wind (Figure 7) with u and v components in strong Changma years and the composite
differences in u-component wind and v-component wind between strong and weak Changma years. From
Figure 7(a), the u-component wind (shading) and the v-component wind (contour) show that the cyclonic
circulation in the region west of Korea is enhanced by the increasing meandering jet.
It is found that the significant circulation is cyclonic in the region through the west and east of Korea. This
cyclonic circulation is associated with the jet core developing close to the east of Korea. The jet streak in the
downstream region over the east of Korea may induce strong divergence in the elevated layer over the surface
trough. We have chosen this downstream jet as one of the Changma indices. Figure 8 shows the 850 hPa wind
field indicating low-level transport for moisture. Advection of moist and warm air by the low-level winds is
essential for generating convective instability and sustaining the convective activity (Ninomiya, 1980). The
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
EIGENVECTOR
105
TIME SERIES
Strong Changma years
1st mode
1st mode (33.5%)
50N
0.6
40N
0.4
0.2
30N
0
20N
−0.2
10N
−0.4
80E
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
2nd mode (13.0%)
2nd mode
50N
0.6
40N
0.4
0.2
30N
0
20N
−0.2
10N
−0.4
80E
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
3nd mode (8.4%)
3rd mode
50N
0.6
40N
0.4
0.2
30N
0
20N
−0.2
10N
−0.4
80E
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
Weak Changma years
1st mode (28.6%)
1st mode
50N
0.6
40N
0.4
0.2
30N
0
20N
−0.2
10N
−0.4
80E
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
2nd mode (13.6%)
2nd mode
50N
0.6
40N
0.4
0.2
30N
0
20N
−0.2
10N
−0.4
80E
50N
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
3rd mode (11.9%)
3rd mode
0.6
0.4
40N
0.2
30N
0
20N
−0.2
10N
−0.4
80E
100E
120E
140E
160E
180
−0.6
1MAY 16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG
Figure 5. As Figure 4, except for strong and weak Changma years
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
106
K.-J. HA, S.-K. PARK AND K.-Y. KIM
(a) strong
SLP
(c) strong
H500(gpm)
(d) strong - weak
H500(gpm)
70N
60N
50N
40N
30N
20N
10N
E0
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W160W 150W
(b) strong - weak
SLP
70N
60N
50N
40N
30N
20N
10N
E0
60E
70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W150W
Figure 6. (a) Composite field for SLP of strong Changma years. (b) The difference in SLP between weak and strong Changma years.
(c) As (a), except for 500 hPa geopotential height. (d) As (b) except for 500 hPa geopotential height. Shaded areas represent significant
areas (95% confidence level). The two rectangular boxes in the significant area are used to construct an index
significant area is for the magnitude of the wind vector in the difference field (Figure 8(b)). Figure 8(b) shows
a north–south pattern consisting of an anomalous anticyclone over the northwestern Pacific and a cyclonic
anomaly around Korea. This meridionally oriented sequence of an alternation of low and high is similar to
the Pacific–Japan pattern shown by Nitta (1986, 1987). In this context, it has been known that above (below)normal pressures and suppressed (enhanced) convection near the Philippine islands and the subtropical western
Pacific Ocean are favourable for enhanced (subdued) rainfall over Japan (Nitta, 1987). In Figure 8(b), the
low-level jet is associated with the merging of the southwest flow and the southeast flow from lower latitudes.
However, the southwest flow is greater than that of the southeast flow. It is obvious to note that the low-level
jet from the southwest flow over the Asian summer monsoon has been associated with a strong Changma.
4. EVALUATION OF SUMMER MONSOON INDICES
Many studies (Webster and Yang, 1992; Goswami et al., 1999; Lau et al., 2000) have derived indices for
measuring the Asian monsoon. In this study, to investigate the Changma index for the interannual variability
of the rainfall characteristics over Korea, the significant difference values in the previous figures (rectangular
squares in Figures 6 and 8) are considered. The Changma indices described in this study are associated with the
interannual variability of the area–period-averaged rainfall. Firstly, we try to compare the Changma indices of
this study and the indices used in the previous studies mentioned above. Table I shows the different correlation
coefficients (CCs) for CRI and other indices. Here, CCs are calculated using the complete 1979–2001 record
length, the years selected as the strong and the weak years, the years except for strong Changma years, and
the years except for weak Changma years. These CCs will reveal the causes of strong and weak Changma.
A 95% confidence level was used.
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
(a) strong
UV200(m/s)
70N
30
60N
25
50N
20
40N
15
30N
10
20N
−10
10N
−15
EQ
60E
80E
107
100E
120E
140E
160E
(b) strong - weak
180
160W
−25
U200(m/s)
70N
60N
50N
40N
30N
20N
10N
EQ
60E
80E
100E
120E
140E
160E
(c) strong - weak
180
160W
V200(m/s)
70N
60N
50N
40N
30N
20N
10N
EQ
60E
80E
100E
120E
140E
160E
180
160W
Figure 7. (a) u (shading) and v (line) components of 200 hPa wind in strong Changma years; (b) u and (c) v component differences
between the mean value of the strong Changma years and that of the weak Changma years
The previous indices, such as RM2 for the regional monsoon in East Asia (Lau et al., 2000) and MHI
(monsoon Hadley index) (Goswami et al., 1999) as well as the NINO3 sea-surface temperature (SST) index
are used to calculate the CC for the CRI. The MHI is a measure of baroclinic instability, which is obtained
from the difference of v-component winds between the upper and lower levels over 10–30 ° N, 70–110 ° E;
and the RM2 is for the 200 hPa u-component wind difference between the mean over 25–35 ° N, 110–150 ° E
and the mean over 40–50 ° N, 110–150 ° E.
In this study, the SLP index is in three parts: the south index for the area over 20–35 ° N, 135–150 ° E;
the north index for the area over 50–57.5 ° N, 115–127.5 ° E; and the total of the south plus the north. The
500 hPa geopotential height (25–35 ° N, 135–152.5 ° E), 200 hPa wind (40–45 ° N, 137.5–155 ° E), and 850 hPa
wind (32.5–37.5 ° N, 127.5–147.5 ° E) are used to measure the interannual variability of Changma from the
significant values mentioned in Section 3.
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
108
K.-J. HA, S.-K. PARK AND K.-Y. KIM
(a) strong
UV850 (m/s)
70N
60N
50N
40N
30N
20N
10N
E0
60E
70E
80E
90E
2
100E 110E 120E 130E 140E 150E 160E 170E
4
6
8
10
12
14
180
16
170W 160W 150W
18
(b) strong−weak
UV850 (m/s)
70N
65N
60N
55N
50N
45N
40N
35N
30N
25N
20N
15N
10N
5N
E0
60E
70E
80E
90E
100E 110E 120E 130E 140E 150E 160E 170E
180
170W 160W 150W
Figure 8. Composite fields for (a) the 850 hPa wind of strong Changma years and (b) the difference in the 850 hPa wind between weak
and strong Changma years. The shaded areas represent significant areas (95% confidence level) and are for wind speed
In relation to the previous indices, we have shown that the CCs are not significant. Therefore, it is unsuitable
to explain the strong and the weak rainfall characteristics over Korea using RM2, MHI, or NINO3 SST.
Among the indices from this study, the SLP area index has the highest CC for all years (1979–2001) and
for the strong and weak Changma years. From the CC for the years except for the strong Changma years
and vice-versa the north and south SLP area indexes are important for explaining rainfall characteristics in
strong and weak years, respectively. It is interesting to note that the migrating high over the north of Korea
may play a significant role in reinforcing the strong Changma over Korea, compared with the NPH over the
south of Korea. The 500 hPa area index is well correlated to the weak Changma over the south of Korea and
the 850 hPa wind area index is also correlated to the weak Changma. On the other hand, the 200 hPa wind
area index does not represent well the strong and the weak Changma years separately.
As a result, there are distinct characteristics for the strong and the weak Changma years in terms of the
SLP area index. The north area index of the SLP area index might be associated with frequent occurrences
of the migrating highs over northeastern China. However, the south area index of the SLP area index and
the 500 hPa geopotential height area index associated with the strength of the Pacific high over the south of
Japan might be correlated to the weak Changma.
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
109
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
Table I. Correlation coefficients between precipitation variability and variables (indices). The four precipitation time series
are used, which are for all years from 1979 to 2001, strong and weak Changma years, years except for strong Changma,
and years except for weak Changma. The variable indices include ENSO SST index, MHI, RM 2, area-averaged SLP
index, area-averaged 500 hPa geopotential height index, and area-averaged wind index. Areas used in averaging are
different from each variable, such as Figure 6 (SLP, H500), Figure 7 (UV200) and Figure 8 (UV850)
Index
NINO3
MHI
RM2
SLP area
South + north
South
North
500 hPa HGT area
200 hPa wind area
850 hPa wind area
1979–2001
Strong and weak
Changma years
Years except for
strong Changma
Years except for
weak Changma
0.24
−0.07
−0.28
0.26
−0.16
−0.36
0.26
0.02
−0.17
0.02
−0.07
−0.45
0.64
0.57
0.49
0.63
0.46
0.58
0.70
0.62
0.57
0.68
0.51
0.58
0.60
0.62
0.35
0.67
0.42
0.58
0.37
−0.17
0.50
−0.04
0.31
0.24
5. LONG-TERM VARIATION
5.1. A trend
The recent characteristics of rainfall in Korea from the mid 1990s, especially for the summer, show a
significant difference from the climatological features in earlier times. Record-breaking rainfall and flash
floods have been experienced during the month of August since 1998 (Yun et al., 2001; Hwang and Park,
2000). However, it is still uncertain whether these extreme events are episodic or part of a long-term climate
variation. In this section, we attempt to investigate the characteristics of long-term change, such as decadal
variation, for two climate regimes of earlier and later years.
To examine the long-term variability of the summer monsoon rainfall, rainfall data for 13 representative
stations (Table II) over Korea for the period of May–October of 1979–2001 are used. Two peaks of rainfall
(Ib : 25 June–29 July; Ia : 14 August–12 September) are seen from the 23 year time series of pentad rainfall
during 1979–2001 (not shown). More than half of the annual rainfall in Korea is concentrated in these two
Table II. Location information for the surface weather stations used in the
present study
Name
Gangneung
Seoul
Incheon
Ulleung-do
Chupungnyeong
Pohang
Daegu
Jeonju
Ulsan
Gwangju
Busan
Yeosu
Jeju
Copyright  2005 Royal Meteorological Society
Latitude
Longitude
Elevation (m)
37° 45
37° 34
37° 29
37° 29
36° 13
36° 02
35° 53
35° 49
35° 33
35° 08
35° 06
34° 44
33° 31
128° 54
126° 58
126° 38
130° 54
128° 00
129° 23
128° 37
127° 09
129° 19
126° 55
129° 02
127° 44
126° 32
26.0
85.5
68.9
221.1
245.9
5.6
57.8
51.2
31.5
70.9
69.2
67.0
22.0
Int. J. Climatol. 25: 99–116 (2005)
110
K.-J. HA, S.-K. PARK AND K.-Y. KIM
summer rainy periods (Ho and Kang, 1988). The primary rainy period is termed Changma. As the monsoon
trough moves further north to northeastern China, the primary rainy period ends. When the monsoon trough
retreats southward of the Korean peninsula, the secondary rainy period, known as the Autumn Changma,
which is called ‘Shurin’ in Japan (Matsumoto, 1992), starts. During the secondary rainy period, typhoons
account for some of Korea’s precipitation.
Using mean rainfall for the two peaks Ia and Ib and the rainfall during the period between the two peaks,
i.e. II (30 July–13 August), the turning point from the earlier regime to the later regime was investigated.
The turning point for the year is computed as the following two-peak index:
Two-peak index =
Ia + Ib
− II
2
The time series of two-peak indices are shown in Figure 9(a). A negative value implies one-peak characteristics and a positive value represents two-peak characteristics. The two-peak index decreases abruptly in
1993 and since then negative values appear frequently. Therefore, 1993 is selected as the turning point. Years
before and after the turning point are termed the earlier regime (PI, 1979–1992) and later (PII, 1993–2001)
regime respectively.
Figure 9(b) shows the time series of regime-average pentad rainfall for both regimes (five-point moving
average) of later and earlier regime. From late July to mid August the rainfall climate changes from light
rainfall to heavy rain characteristics as the regime changes. This means that the withdrawal time for Changma
has tended to be later in the recent decade, so the last decade has many more years of long Changma than the
earlier decade. Lu et al. (2002) mentioned that the withdrawal of Changma is important for determining the
(a)
(Ib+Ib)/2-II
10
Rain(mm / day)
5
0
PI
−5
−10
PII
Turning point
−15
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
(b)
10
Rain (mm / day)
9
easier(1979~1992)
later(1993~2001)
8
7
6
5
4
3
2
1
16MAY 1JUN 16JUN 1JUL 16JUL 1AUG 16AUG 1SEP 16SEP 1OCT 16OCT
Figure 9. (a) Time series of two-peak index (solid line) and 5 year averaged two-peak index (dashed line). Years before and after the
turning point are termed the earlier (PI) and later (PII) regimes respectively. (b) Time series of regime average pentad rainfall for both
regimes (five-point moving average) of later (dashed line) and earlier (solid line) regimes
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
111
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
rain amount for the summer monsoon season. On the other hand, rainfall decreases in the later regime during
the secondary rainy period. In this study, however, we focus on the period from late July to mid August
associated with Changma withdrawal.
Composite analyses are performed for two climate regimes during June–August. We depict differences of
the mean CMAP between the earlier and later regimes. Figure 10(a) and (b) shows the CMAP (mm/day) in
the later regime and the difference between the two regimes. The shaded areas show the significant regions
at the 95% confidence level. There is significantly more rainfall along the edge of the continent, especially
over the East China Sea, Kyushu and south China in the later regime.
Figure 10(c) and (d) is the same as Figure 10(a) and (b) but shows the 700 hPa geopotential height.
Significant positive values are represented over northeast China and Mongolia in the composite difference
(Figure 10(d)). Similar results appear at 500 hPa (not shown). Ho et al. (2003) examined the long-term
changes in Korean precipitation between the earlier (1954–77) and later (1978–2001) regimes, showing that
increased geopotential height over the Asian continent leads to a strengthening of the anticyclonic circulation
anomaly in the later regime. Therefore, a northerly wind is enhanced in East Asia along the eastern boundary
of the anticyclone, which transports relatively cool and dry air to East Asia. The northerly wind later interacts
with the prevailing southerly wind that is warm and moist. Hence, moisture convergence and convective
activity intensify, increasing the chance of heavy rainfall. In the composite difference for 850 hPa wind (not
shown), an anticyclonic circulation anomaly associated with positive values in Figure 10(d) emerges over
northeast China and Mongolia. As a result, a northerly wind anomaly appears in Korea, and, as mentioned
above, convergence occurs between cold air associated with the anticyclonic circulation anomaly and warm
air originating from the SCS to the south of the Korean peninsula. This may result in strong baroclinicity to
the south of the Korean peninsula, which may contribute to more rainfall over that region in the later regime,
as shown in Figure 10(b). On the other hand, the increased rainfall over south China in the later regime may
be associated with the stronger southwesterly over the south of that region.
(a) later regime
PRCP
(c) later regime
70N
70N
60N
60N
50N
50N
40N
40N
30N
30N
20N
20N
10N
10N
E0
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W 150W
(b)DIFF .(later−earlier)
E0
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W 150W
PRCP
(d)DIFF .(later−earlier)
70N
70N
60N
60N
50N
50N
40N
40N
30N
30N
20N
20N
10N
10N
E0
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W 150W
H700
H700
E0
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W 150W
Figure 10. (a) Precipitation (mm/day) composite during June–August in the later regime. Contour interval is 2 mm/day. (b) Composite
difference in precipitation between the later and earlier regimes. Contour interval is 0.5 mm/day, and zero line is not shown. (c) As
(a), except for 700 hPa geopotential height. Contour interval is 20 m. (d) As (b), except for 700 hPa geopotential height. Contour
interval is 3 m. Shaded areas represent significant areas with a 95% confidence level
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
112
K.-J. HA, S.-K. PARK AND K.-Y. KIM
5.2. Long and short Changma
As mentioned in Section 5.1, the last decade has more years of long Changma than the early decade.
Comparisons between the long and short Changma with rainfall amounts during late July to mid August
(30 July–13 August) are important for examining the influence of the late withdrawals of Changma. This
period is defined as a late peak period in this study. Figure 11(a) shows the time series of mean precipitation
anomalies over Korea during the late peak period. Long and short Changma years are defined as years in
which precipitation anomalies are more than ±1σ in the later and earlier regimes respectively. Long Changma
years are selected as 1993, 1997, 1998, and 1999 in the later regime, and short years as 1984, 1988, and 1990
in the earlier regime.
(a)Rain anomaly(30Jul−13Aug)
14
12
10
8
6
4
2
0
−2
−4
−6
−8
−10
σ
−σ
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
Mean: 6.7 mm/day
(b) long Changma years
(d) July
70N
70N
60N
60N
50N
50N
40N
40N
30N
30N
20N
20N
10N
10N
EQ
60E
80E 100E 120E 140E 160E 180 160W
EQ
60E
(c) short Changma years
(e) August
70N
70N
60N
60N
50N
50N
40N
40N
30N
30N
20N
20N
10N
10N
EQ
60E
80E 100E 120E 140E 160E 180 160W
80E 100E 120E 140E 160E 180 160W
EQ
60E
80E 100E 120E 140E 160E 180 160W
Figure 11. (a) Time series of mean precipitation anomaly over Korea during the late peak period (30 July–13 August). The data used are
5 day averaged (pentad) precipitation data of 13 weather stations in the Korean peninsula. The years when rainfall is less than −σ in the
earlier regime and more than σ in the later regime are denoted by diamonds (short Changma) and circles (long Changma) respectively.
SLP composites during the late peak period (b) in the long Changma years and (c) in short Changma years. The climatology of SLP in
(d) July and (e) August from 1979 to 2001
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
113
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
Composite analyses for long and short Changma years are performed. Figure 11(b) and (c) shows SLP
composites in long Changma years and in short Changma years. In the long Changma years (Figure 11(b)),
ridges associated with the NPH extend toward the Sea of Okhotsk and south of Japan and a trough develops
between these ridges, with an inclined NPH northeastward from the SCS. The direction of the NPH boundary
might be associated with a continuous moisture supply from the warm pool area. On the other hand, the ridge
extends toward Korea and the Aleutian low develops in short Changma years (Figure 11(c)). From these
results, positive values in composite difference appear over the northwest Pacific, north of the Okhotsk Sea
and the Bering Sea, and negative values around Korea. Similar features are evident in the 700 hPa geopotential
height composite difference (not shown).
Figure 11(d) and (e) indicates the climatology of the SLP in July and August respectively. In July, the
NPH extends toward the south of Japan and the Sea of Okhotsk. In August, on the other hand, the NPH
extends toward Korea and Japan. When compared in Figure 11(b) and (c), the climatology of the SLP in
July and August is similar to the composites for long and short Changma years respectively. As shown in
the 200 hPa geopotential height, the Tibetan High extends more to the northeast in the short Changma years
compared with the long Changma years, which is similar to the climatology of the 200 hPa geopotential
height in August. In addition, the 200 hPa geopotential height field in the long Changma years resembles the
climatology of July. These results mean that, in the long Changma years, the pressure patterns in July are
maintained longer at both the surface and upper level synoptic fields during the late peak period.
In the 850 hPa wind composites (Figure 12(a) and (b)), a southwesterly is extended from the SCS toward the
Korean peninsula in the long Changma years. This southwesterly provides moisture to the Korean peninsula
and causes heavy rain. In the short Changma years, however, the Korean peninsula is not on the path of
the southwesterly because of the circulation associated with the expansion of the NPH toward the Korean
peninsula. In the 200 hPa wind composites (Figure 12(c) and (d)), the westerly jet stream moves northward
so that it is weakened around the Korean peninsula in the short Changma years. Lu (2002) also found that,
around Korea, the upper level jet is weakened and moves poleward for early withdrawals.
(a) long Changma years
(c) long Changma years
70N
70N
60N
60N
50N
50N
40N
40N
30N
30N
20N
20N
10N
10N
EQ
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W 150W
EQ
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W150W
3
20
(b) shoart Changma years
(d) short Changma years
70N
70N
60N
60N
50N
50N
40N
40N
30N
30N
20N
20N
10N
10N
EQ
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W 160W150W
3
EQ
60E 70E 80E 90E 100E 110E 120E 130E 140E 150E 160E 170E 180 170W160W150W
20
Figure 12. The 850 hPa wind composites in (a) the long Changma years and (b) the short Changma years. 200 hPa wind composites
in (c) the long Changma years and (d) the short Changma years. Contour (interval 5 m/s) of right panel represents wind speed
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
114
K.-J. HA, S.-K. PARK AND K.-Y. KIM
Figure 13 shows the 850 hPa water vapour flux divergence composites. In the composite difference (not
shown), significant negative values are found to the east (the East Sea of Korea and Japan) and southwest
(eastern south China) of Korea, so the flux convergence of water vapour over these regions is stronger in long
Changma years. On the other hand, water vapour flux convergence is seen from the southwest to northeast
around Korea in the long Changma years. Therefore, the supply of moisture to the Korean peninsula due to
the southwesterly and water vapour flux convergence in the lower troposphere leads to increased rainfall over
Korea during the late peak period.
6. SUMMARY AND CONCLUSIONS
We have used the CMAP 10 day mean field to depict the intraseasonal variability of the summer monsoon
rainband in the western Pacific. Maximum rain axes over 8 mm/day are used to investigate the northward
evolution of the summer frontal rain characteristics. The local CRI over Korea is defined as an area-mean
CMAP during the Changma period from late June to late July from the time variation of the maximum rain
axis. From the results of the EOF analyses applied to the CMAP data, the second mode depicts the East
Asian rainband, referred to as Meiyu and Baiu, and the third mode appears to be related to the Changma in
Figure 13. The 850 hPa water vapour flux divergence (×109 ) during late peak period (a) in the long Changma years and (b) in the short
Changma years. Negative (positive) values represent flux convergence (divergence). Dark grey areas are mountain areas above 1500 m
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
SUMMER MONSOON RAINFALL CHARACTERISTICS OVER KOREA
115
Korea. Also, the retreat of the Changma is late in strong Changma years, and Changma rainfall is positively
correlated with the rainfall in the Indian monsoon region, especially that in the northern part of the Indian
continent, and negatively correlated with the rainfall near the SCS (10 ° N, 120 ° E). In order to construct
an index associated with the interannual variability of CRI, composite fields for SLP, 500 hPa geopotential
height, 200 and 850 hPa winds are investigated.
From the composite analyses, a strong Changma is characterized by the strengthening of the NPH (in SLP
and 500 hPa geopotential height) over the northwest Pacific and by the frequent occurrence of the migrating
high (in SLP) over northeastern China. These are well represented in the Changma indices. For the 200 hPa
wind, cyclonic circulation over northwestern Korea is seen in strong Changma years. This cyclonic circulation
is associated with the jet core developing to the east of Korea. This jet streak in the downstream region may
induce strong divergence over the surface trough. At 850 hPa, the southwesterly is strongly extended from
the Bay of Bengal through the East China Sea to Japan in the strong Changma years. The downstream jet
over the east of Korea and the southwesterly around southern Japan are also represented in the Changma
indices.
The CRI is more significantly correlated to the indices in SLP, 500 hPa geopotential height, 200 hPa wind
and 850 hPa wind than the previously defined indices of MHI, RM2 and NINO3. An important approach
in this study is to separate the indices in terms of the strong and the weak Changma years. The migrating
high in northeastern China is associated with the Changma, which is well represented in the SLP field. On
the other hand, both the NPH over the northwest Pacific and the 850 hPa wind index are associated with the
weak Changma.
To examine the long-term variability of the summer monsoon rainfall, we have analysed the rainfall data
for 13 representative stations over Korea for the period of May–October of 1979–2001. Pentad rainfall shows
two peaks in the earlier regime (1979–92), but only one peak in the later regime (1993–2001). From late July
to mid August, the rainfall climate changes abruptly from a dry and light rainfall to heavy rain characteristics.
This means that the withdrawal time for Changma has tended to be late in the recent decade.
In the composite analyses for CMAP in the earlier and later regime during June–August, there is
significantly more rainfall along the edge of the continent, especially over the East China Sea, Kyushu
and south China in the later regime. On the other hand, there is little difference in rainfall between the earlier
and later regimes over the Korean peninsula.
The later decade has many more years of long Changma than the early decade. By using the time series
of mean precipitation anomalies over 13 Korean stations during the late peak period (from late July to mid
August), long and short Changma years are selected for the later and earlier regimes respectively. In the
long Changma years, SLP ridges associated with the NPH extend toward the Sea of Okhotsk and south of
Japan, and a trough develops between these ridges. In the short Changma years, however, the ridge extends
toward Korea. The climatology of SLP in July is similar to the composite for long Changma years, and that
in August is similar to the composite for short Changma years. A similar feature is also found in the 200 hPa
geopotential height climatology; the same result is seen (not shown). These results mean that, in the long
Changma years, pressure patterns in July are maintained longer on both the surface and upper level synoptic
fields during the late peak period.
At 850 hPa, the southwesterly is extended from the SCS toward the Korean peninsula in the long Changma
years. In the short Changma years, however, the Korean peninsula is not on the path of the southwesterly
because of the circulation associated with the expansion of the NPH toward the Korean peninsula. In the
short Changma years, the upper level jet moves northward and is weakened around the Korean peninsula.
As a result, there is little change of total rainfall during summer (June–August) over Korea. However, from
late July to mid August there is more rainfall in the later regime because the July pattern is maintained longer
on both the surface and upper level synoptic fields in that regime. On the other hand, there is no significant
difference in SST between the earlier and later regimes (not shown). It is not clear whether the variation of
Korea rainfall during the late peak period is a climatological variation. It is important to understand better
the causes of change in summertime precipitation patterns and their predictability, and further investigations
of these phenomena should be undertaken.
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)
116
K.-J. HA, S.-K. PARK AND K.-Y. KIM
ACKNOWLEDGEMENTS
We gratefully acknowledge Song Yang for helpful discussions on an earlier version of this paper. This
material is based upon work on ‘Monsoon and Changma’ supported by a Meteorological & Earthquake R&D
Grant from the Korean Meteorological Administration. We would like to acknowledge the support from the
KISTI (Korea Institute of Science and Technology Information) under ‘The Fifth Strategic Supercomputing
Support Program’ with Dr Min-Su Joh as the technical supporter. The use of the computing system of the
Supercomputing Center is also greatly appreciated.
REFERENCES
Goswami BN, Krishnamurthy V, Annamalai H. 1999. A broad scale circulation index for the interannual variability of the Indian
summer monsoon. Quarterly Journal of the Royal Meteorological Society 125: 611–634.
Ho CH, Kang IS. 1988. The variability of precipitation in Korea. Journal of the Korean Meteorological Society 24(1): 38–48 (in Korean).
Ho CH, Lee JY, Ahn MH, Lee HS. 2003. A sudden change in summer rainfall characteristics in Korea during the late 1970s.
International Journal of Climatology 23: 117–128.
Hwang JD, Park CK. 2000. Characteristics of circulation pattern over East Asia associated heavy rainfall events in Korea during the
summer 1999. Journal of the Korean Meteorological Society 36(5): 573–582 (in Korean).
Kang IS, Ho CH, Lim YK, Lau KM. 1999. Principal modes of climatological seasonal and intraseasonal variations of the Asian summer
monsoon. Monthly Weather Review 127: 322–340.
Lau KM, Chan PH. 1986. Aspects of the 40–50 day oscillation during the northern summer as inferred from outgoing longwave
radiation. Monthly Weather Review 114: 1354–1367.
Lau KM, Kim KM, Yang S. 2000. Dynamical and boundary forcing characteristics of regional components of the Asian summer
monsoon. Journal of Climate. 13: 2461–2482.
Lim KH. 1997. The linear relationship between monthly precipitation amounts of Korea and time variation of the geopotential height
in East Asia and the Pacific during the summer season. Journal of the Korean Meteorological Society 33(1): 63–74 (in Korean).
Lu RY. 2002. Association between the surface temperature in Eurasian and the Changma withdrawals. Korean Journal of the Atmospheric
Sciences 5(S): 63–70.
Lu RY, Oh JH, Kim BJ, Baek HJ, Huang RH. 2001. Associations with the interannual variations of onset and withdrawal of the
Changma. Advances in Atmospheric Sciences 18: 1066–1080.
Luo HB, Yanai M. 1984. The large-scale circulation and heat sources over the Tibetan Plateau and surrounding areas during the early
summer of 1979. Part II: heat and moisture budgets. Monthly Weather Review 112: 966–989.
Matsumoto J. 1992. The seasonal changes in Asian and Australian monsoon regions. Journal of the Meteorological Society of Japan
70: 257–273.
Ninomiya K. 1980. Enhancement of Asian subtropical front due to thermodynamic effect of cumulus convection. Journal of the
Meteorological Society of Japan 58: 1–15.
Nitta T. 1983. Observational study of heat sources over the eastern Tibetan Plateau during the summer monsoon. Journal of the
Meteorological Society of Japan 61: 590–605.
Nitta T. 1986. Long-term variations of cloud amounts in the western Pacific region. Journal of the Meteorological Society of Japan 64:
373–390.
Nitta T. 1987. Convective activities in the tropical western Pacific and their impact on the Northern Hemisphere summer circulation.
Journal of the Meteorological Society of Japan 65: 373–390.
Nitta T, Hu ZZ. 1996. Summer climate variability in China and its association with 500hPa height and tropical convection. Journal of
the Meteorological Society of Japan 74: 425–445.
Oh JH, Lu RY, Kwon WT, Ryoo SB. 2000. The 500hPa geopotential fields in the mid–high latitudes associated with Changma. Korean
Journal of the Atmospheric Sciences 3(2): 83–94.
Wang B, Xu X. 1997. Northern Hemisphere summer monsoon singularities and climatological intraseasonal oscillation. Journal of
Climate 10: 1071–1085.
Wang WC, Li K. 1990. Precipitation fluctuation over semiarid region in northern China and the relationship with El Nino/southern
oscillation. Journal of Climate 3: 769–783.
Webster PJ, Yang S. 1992. Monsoon and ENSO: selectively interactive systems. Quarterly Journal of the Royal Meteorological Society
118: 877–926.
Xie P, Arkin PA. 1997. Global precipitation: a 17-year monthly analysis based on gauge observations, satellite estimates and numerical
model outputs. Bulletin of the American Meteorological Society 78: 2539–2558.
Xie P, Janowiak JE, Arkin PA, Adler R, Gruber A, Ferraro R, Huffman GJ, Curtis S. 2003. GPCP pentad precipitation analyses: an
experimental dataset based on gauge observation and satellite estimates. Journal Climate 16: 2197–2214.
Xue Y. 1996. The impact of desertification in the Mongolian and the inner Mongolian grassland on the regional climate. Journal of
Climate 9: 2173–2189.
Yatagai A, Yasunari T. 1994. Trends and decadal-scale fluctuations of surface air temperature and precipitation over China and Mongolia
during the recent 40 year period (1951–1990). Journal of the Meteorological Society of Japan 72: 937–957.
Yun WT, Park CK, Lee JW, Lee HS, Min SK. 2001. Analysis of the Korean heavy rainfall features in summer. Journal of the Korean
Meteorological Society 37(2): 181–194 (in Korean).
Zhang RH, Sumi A, Kimoto M. 1996. Impact of El Niño on the East Asian monsoon: a diagnostic study of the 86/87 and 91/92 events.
Journal of the Meteorological Society of Japan 74: 49–62.
Copyright  2005 Royal Meteorological Society
Int. J. Climatol. 25: 99–116 (2005)