The variability of the Asian summer
monsoon in warmer climate
Ding Yihui*
(National Climate Center, CMA)
*Contributers: Wang Zunya, Sun Ying,Liu Yunyun, Liu Yanju, Song Yafang, Zhang Jin
Outline
1. The variability of Asian summer monsoon
under global warming in last 100-years
2. Future change in Asian summer monsoon
in a warmer world in next 100 years
3. Discussions and conclusions
1. The variability of Asian summer
monsoon under global warming in
last 100-years
Climatology of the Asian summer monsoon
Fig.1. The vertically integrated moisture flux transport (surface to
300hPa )averaged for summer (JJA: June, July and August) of 19482006(Unit：kg.m-1.s-1）
Fig.2. The vertically integrated field of divergence of moisture transport for
JJA of 1948-2006. Negative (positive) regions denotes the convergence
(divergence) of moisture（Unit：10-5 kg.m-2.s-1)
Fig.3. Climatological mean (1979-2006) summer precipitation
pattern ( Unit:mm.d-1)
1.16
94.72
8.79
(+30.61) 116.33
NEC
27.00
(-5.82) NC 91.24
49.63
23.49 (+78.81)YHRB97.58
193.06
162.21
AS
29.19
(+85.23)
BOB
143.47
(-892.19)
SIO
28.6
6
(+341.80)
NWNP
202.29
405.21
267.35
SCS
834.39 (+329.27)455.54 (+201.14)111.64
786.60
120.21
37.71 202.54
(+70.42) SC 32.11
53.39
124.59
(+519.28)
SWNP
544.98
267.87
202.46
79.52
Fig.4. Schematic maps of the climatological mean moisture budgets of the Asian-Pacific summer monsoon
region in summer (JJA) (Units:106kg/s). The shaded area refers to the Tibetan Plateau. The positive and
negative values represent net moisture influx and efflux, respectively. SIO: South Indian Ocean(30S°0°N, 40°E-120°E); AS: Arabian Sea(0°-22.5°N, 40°E-80°E); BOB: Bay of Bengal (0°-22.5°N ,
80°E-100°E); SCS: South China Sea(0°-22.5°N , 100°E-120°E) ; SC: South China (22.5°N-27°N,
100°E-120°E); YHRB: Yangtze–Huaihe River Valleys(27°N-35°N, 100°E-120°E); NC: North
China(35°N-42°N,100°E-120°E); NEC: Northeast China(42°N-54°N ,120°E-135°E); NWNP:
northwestern part of North Pacific (22.5°N-45°N , 120°E-160°E); SWNP: southwestern part of
North Pacific(0°-22.5N°, 120°E-160°E).(Liu, et al.,2009)
Time series of observed mean surface temperature
change for last 100 years
Top panel: NH; bottom panel: China. Red: trends , blue:21-yr running average.
气
温
距
平
(K)
年
气
温
距
平
2
OBSWG
OBSt
OBSm21
1.5
1
(C)
0.5
0
-0.5
-1
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
年
year
观测近百余年北半球（上图，Latif等，2009）和中国（下图，根据龚道溢，王绍武，2009计算绘制）年
平均气温距平变化（图中黑线是年平均距平，红线是线性趋势，兰线是21年滑动平均）
Correlation patterns of Asian-Pacific
summer (JJA) monsoon
亚洲-太平洋季风降水的相关分布
西北太平洋关键区6~8月降水与亚洲-太平洋季风区降水的相关
Teleconnection modes of various components of Asian
summer system: holistic correlation for ISO
+
+
+
+
+
+ +
+
+
-
季风子系统的遥相关（季节内变率的整体性）
 夏季风爆发初期，ISM通过“南支”遥相关型影响长江流域梅雨；
 夏季风盛行期间，ISM通过“北支”遥相关型影响华北地区降水；
 WNPSM主要通过经30~60天滤波的CISO影响中国夏季降水。
Weakening of the Asian summer monsoon and
superposed inter-annual and inter-decadal variability .
Fig.5. Long-term variation of the East summer monsoon index for 1870-2003
(based on Guo’s monsoon index). Positive(negative)values denote stronger
(weaker)summer monsoon than normal. (IPCC, 2007)
Inter-decadal variability for all India precipitation
12-18,30-40 and 60-80-yr oscillation
年代际变率：12-18年，30-40年，60与80年
(Goswami,2005)
Long–tem of anomalous precipitation in Asian-African
monsoon regions
30
15
0
a
西非 West Africa
（Trenberth, et al. 2007）
b
东非 East Africa
（Trenberth, et al. 2007）
c
南亚 South Africa
（Trenberth, et al. 2007）
-15
降水距平
百分比/%
-30
30
15
0
-15
-30
30
15
0
-15
-30
80
40
d
0
-40
1900
1920
1940
1960
1980
1880-2008年亚非季风区降水量异常
中国华北（71个站序列）
North China
2000
Ren and Feng (2009)
我国降水分布发生了明显变化
Significant change in summer precipitation pattern in
China (1958-2007) unit:%/10yrs
%/10
（
单
位
：
年
）
（中国气象局气候变化中心）
近50年来，西部地区降水约增加15％—50％；东部地区频繁出现“南涝北旱”，华南地
区降水约增加5％—10％，而华北和东北大部分地区约减少10％—30％。 （图：19582007年我国年降水量变化幅度）
Difference in drought days between 1979-2008 and 1951-1978
50
40
30
20
80
90
100
110
120
130
1979-2008年和1951-1978年年平均干旱日数之差
（单位：天）
（第二次国家气候变化评估报告初稿，2009）
Numbers of days of heavy rainfall
近50年来中国大陆极端强降水日数的变化趋势（实心和虚心圆分别代表增
加和下降趋势，按半径大小分别为每10年变化7.5% 以上，7.5%～2.5%，
小于2.5% ，显著变化的地区标有叉号）
Vulnerability of fresh water resources in the context of
global climate change (high stress region in North China)
气候变化下，全球现代淡水资源的脆弱性和他们的管理
IPCC，2007
(a)
(b)
长江
华北
(c)
(d)
华南
梅雨季
123年（1880－2003）中国东部分区降水的变化
Long –term (123yr)variation of summer precipitation in North
China (a), Yangtze River basin (b), South China (c) and Meiyu
season
(a)
Va
ria
nc
e
(
σ
2)
(b)
华北
Period (years)
Period (years)
(c)
(d)
华南
Va
ria
nc
e
(
σ
2)
Period (years)
长江
Va
ria
nc
e
(
σ
2)
Va
ria
nc
e
(
σ
2)
梅雨季
Period (years)
小波分析功率谱：30-40年与80年周期趋势变化分析：1978和1992是两个突变点
Spectrum power of wavelet analysis:30-40-yr and
60-80-yr oscillations
中国东部不同分区夏季降水的主要周期
Major periods of summer precipitation in different
subregions of East China
subdivision
分区
华南
South China
4, 14*, 30*, 80*
2*, 7, 30*
Yangtze
2*, 7*, 20*, 40*
2*, 7,14, 40*
North China
3, 9, 18*, 40*,
80
3*, 9, 18
2, 7, 12, 40*,
80*
2, 7*, 12, 40*
长江中下游
华北
A时段（123年） B时段（54年）
长江中下游5站 (121年)
Meiyu Season
*代表超过50%信度
中国东部三个地区夏季降水的突变点检验
Detection of abrupt change points for different subregions
in East China
Methods
方法
South
China
华南
Yangte
长江中下
游
North
China
Meiyu
华北
梅雨期长江中下
游5站
Running Test
1980, 1992
1978
1965, 1979
1978
Yamamoto et
al., (1986)
1980, 1992
1979
1964, 1980
1980
Mann-Kendall
(1945; 1975)
1993
1982
1975
1978
所有的突变点都超过95%的信度
15
(a)
(b)
Temporal amplitude
10
5
0
-5
-10
1950
1960
1970
1980
1990
2000
2010
Year
15
空间分布
Temporal amplitude
(c)
时间系数
10
(d)
5
0
-5
-10
-15
1950
1960
1970
1980
1990
2000
Year
中国夏季降水的EOF分析（1951-2004）
EOF modes of summer precipitation (JJA) for 1951-2004
(a)(c): EOF1；(b)(d): EOF2
2010
夏季风
水汽输
送向量
的EOF分
析
下图是
时间系
数
60
Temporal amplitude
40
Leading EOF model of
moisture tranport for
Asian summer monsoon
20
0
-20
-40
1950
1965
1980
Year
1995
2010
a
3
b
2
1
0
1951 1954 1957 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005
-1
-2
-3
图2 1951~2005年夏季850hPa风场的EOF分解的第1特征向量（a）及对应的时间系数（b）
1951-1978
1979-1992
1993-2004
中国不同时段，夏季降水距平百分比分布
的变化。（阴影区是正距平，相对于19712000年平均值）
In-decadal shifts of summer
monsoon patterns in China
shading: positive anomaly
percentage
1951-2004中国东部(107.5-130°E)平均夏季异常降水纬度-时间剖面图。单位: mm
Latitude-time cross-section of summer precipitation
departure(107.5-130°E)
850 hPa平均经向风纬度时间剖面图 （unit: ms-1）。阴影区是异常南风。
Latitude-time cross-section of 850hpa V-component
departures, shading: South wind
(a)
(b)
1955-2004 异常夏季水汽输送(a)和水汽汇 (Q2) (b)纬度-时间剖面图。（地面到
300hPa输送总量）。单位: Kgm-1g-1 (a) 和10-5Kgm-1s-1 (b).
Cross-sections of anomalous moisture (a) transport and divergence
(Q2) (b) for East China
850hPa wind EOF2
Time coefficient
a
A
3
b
2
1
C
0
1951 1954 1957 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005
-1
-2
-3
c
Interannual Variability:
TBO and 4-7-yr oscillations
1951~2005年夏季850hPa风场的EOF分解的第2特征向量（a）、
时间系数（b）及时间系数的最大熵谱曲线（c）
Sea level pressure
a
EOF fields of (a)EOF1 (c) EOF2
and corresponding time coefficients
(b)EOF 1,(d)EOF2
2
b
1
0
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
-1
-2
-3
d
4
3
d
2
1
c
0
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
-1
-2
-3
1951~2005年夏季海平面气压场EOF分解
的第1和2特征向量（a、c）对应的时间系
数（b、d）
(b)
(a)
40
40
20
20
0
0
-20
-20
-40
-40
1940
1960
1980
2000
1940
1960
1980
2000
Time series of the anomalous vertically integrated (from
surface to 100hPa) apparent heat source (Q1) averaged for
all Tibetan Plateau (75～105°E，27.5~42.5°N ) for
summer (a), and spring (b). Solid lines denote 9-yr running
mean curves. Unit: Wm-2
EOF fields of vertically integrated atmospheric heat
source (Q1) (a) EOF1; (b) EOF2 and corresponding
time coefficients; (b) EOF1；(d)EOF2
a
3
b
2.5
2
1.5
1
0.5
0
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
-0.5
-1
-1.5
-2
c
2.5
d
2
1.5
1
0.5
0
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
-0.5
-1
-1.5
-2
-2.5
整层积分的大气热源分布（阴影为负值区）
e
Composite SSTA (Nino 3.4) patterns for strong
summer monsoon phase of the TBO cycle
a
d
e
b
e
c
季风强年海温的季节演变合成图（a、前一年冬季；b、当年
春季；
c、当年夏季；d、当年秋季；e、当年冬季；虚线方框代表
印度洋偶极子关键区和Nino3.4区）
a: preceding winter (year 0)
b；spring (year 1)
c: summer (year1)
d: autumn (year1)
e: following winter (year2)
Box: IOD and Nino 3.4.
Schematic of the anomalous fields for TBO of the
Asian –Pacific forced by ENSO events .
850hPa风场
SSTA<0
SSTA>0
SSTA<0
亚洲-太平洋夏季风准两年振荡的异常场示意图
绿色阴影区代表亚洲-太平洋季风区夏季大气热源的正值区；黑色箭头代表850hPa环
流异常场；灰色阴影区代表异常水汽辐合场；红（蓝）色曲线代表海温异常场
2. Future change in Asian
summer monsoon in a warmer
world in next 100 years
Validations of IPCC model performance
模式名称
模式和CMAP
1979-99夏季平均
降水的相关
模式和GPCP
1979-99夏季平
均降水的相关
模式和中国站点降水资
料1979-99减去195878降水变化的相关
分类情况
1
CGCM3.1(T47)
0.51
0.39
-0.52
3类
2
CGCM3.1(T63)
0.59
0.49
0.31
3类
3
CNRM-CM3
0.84
0.85
-0.45
2类
4
CSIRO
0.75
0.76
-0.10
2类
5
GFDL-CM2.0
0.83
0.83
0.46
1类
6
GFDL-CM2.1
0.82
0.80
-0.22
2类
7
GISS-EH
0.33
0.36
-0.50
3类
8
GISS-ER
0.40
0.44
-0.60
3类
9
FGOALS-g1.0
0.24
0.28
-0.50
3类
10
INM-CM3.0
0.76
0.70
0.15
2类
11
IPSL-CM4
0.68
0.68
-0.42
3类
12
MIROC3.2(hires)
0.80
0.82
0.38
1类
13
MIROC3.2(medres)
0.81
0.76
0.23
1类
14
ECHAM5
0.71
0.60
-0.43
3类
15
MRI-CGCM2.3.2
0.60
0.61
0.43
3类
16
CCSM3
0.57
0.48
-0.09
3类
17
PCM
0.28
0.11
0.46
3类
18
UKMO-HadCM3
0.89
0.84
-0.76
2类
2010-2099中国东部降水变化百分率EOF分析结果
EOF analysis of summer precipitation in East China for 2010-2099
Fig.10. Future percentage changes(%) in summer precipitation for East Asia and its three sub-regions
(South China, Yangtze River Valley(YRV) and North China), relative to climatological mean of 19801999. Projections are based on 19 IPCC AR4 climate models.(Sun and Ding,2009)
2010-2099中国东部降水时间-纬度剖面
Latitude-time cross-section of East Asian
summer precipitation for 2010-2099
Fig.12. Future change of the East Asian summer index for next 100 years( based on the
definition of monsoon index by Lu and Chan, with estimate of the V-component of wind).
(Sun and Ding,2009)
2010-2099东亚850hPa水汽输送变化时间-纬度剖面
Latitude –time cross-section of 850hpa moisture
transport in East Asia for 2010-2099
Projection of South Asian summer monsoon
The land-sea thermal contrast for June-July-AugustSeptember (JJAS) between the TP region(20o-40°N, 60o100°E) and the tropical Indian Ocean (TIO, 10°S-10°N,
60o-100°E) (see the black boxes in Fig. 1a) was computed
for both the upper and lower troposphere using TCupper=
Thickness(200-500hPa, TP) – Thickness(200-500hPa, TIO)
and TClower = Temperature (near-surface [2m], TP) –
Temperature (500-850hPa, TIO). The 500-850hPa
temperature over the TIO is used as a proxy of nearsurface temperature after height adjustment for
comparison with 2m air temperature over the TP, which is
2-5km above the mean sea level.
(Sun, Ding and Dai,2010)
Long-term (1979-2000) mean of June-July-August (JJAS) 850hPa
winds (arrows, in m s-1) and 200-500hPa thickness anomaly (colors, in
geopotential meter [gpm]) relative to the mean of the domain (10°S45°N, 30°-140°E).
Temporal evolution of (a) JJAS
MI (black, in m s-1), TCupper
(red, gpm), and TClower(blue,
gpm), and (b) JJAS -U200
(black, m s-1) and U850 (green,
m s-1) anomalies (relative
to1980-1999 mean) averaged
over 0°-20°N and 40°110°E, and TCupper (red,
gpm), and TClower (blue, gpm)
from 1951 to 2099 based on
IPCC AR4 7-model arithmetic
mean under observation-based
forcing during 1951-2000 and
the A1B scenario for 2001-2099.
U200 anomalies in (b) were
multiplied by -1 to show the
weakening of 200hPa easterly
winds.
IPCC 7-model
averaged (a) timeheight cross-section of
JJAS temperature
departures (K, from
1980-1999 mean)
averaged over the TIO
during 1951-2099. (b)
Latitude-height crosssection of JJAS
temperature change
(K) from 1980-1999 to
2080-2099averaged
between 60°E and
100°E. The
topography along
90°E is shown by the
black areas.
(c) Longitude-height cross-section of change (K, from 1980-1999 mean) in JJAS meridional
temperature gradient (temperature changes averaged from 20o-40°N minus that from
10°S-10°N) for 2080-2099. The topography along 35°N is shown by the black areas. (d)
Time series of JJAS 200-500hPa thickness anomalies (gpm, from 1980-1999 mean) over the
TP (solid line) and TIO (dashed line) during 1951-2099.
3. Discussions and conclusions
(1) In recent three decades, North and Northeast China have
suffered from severe and persistent droughts while the Yangtze
River basin and South China have undergone much more significant
heavy rainfall/floods events. This long-term change in the summer
precipitation and associated large-scale monsoon circulation features
have been examined by using about 123-yr (1880–2002) records of
precipitation in East Asia. One dominating mode of the inter-decadal
variability of the summer precipitation in China is the near-80-yr
oscillation. Then, on this basis, a possible explanation of this longterm change in relation to significant weakening of the Asian
summer monsoon, possibly due to the abrupt increase in the
preceding winter and spring snow over the Tibetan Plateau and
warming of the sea surface temperature in tropical central and
eastern Pacific since about 1978,has been set forward.
But we cannot answer whether the
anthropogenic forcing has caused the
changes of patterns of rainfall,
floods/droughts in China, and East Asian
monsoon. Natural fluctuation of climate
change can also play an important role. It
needs the further studies.
(2) It seems that the major rainfall belts
would move northward by about 2040, but
unstable. Afterwards, the summer
precipitation in North China would increase
considerably and stably. Furthermore, this
anthropogenically-driven precipitation shift
would appear to be consistent with the
occurrence of rainfall peak period caused by
the natural near-80-yr cycle. But this
coincidence will be reliable?
(3)The above analyses show that the differential increases
of the upper-tropospheric temperature over the TIO and
TP lead to the changed relationship between the SASM
intensity and tropospheric thermal contrasts over the
SASM regions in a GHG-induced warmer climate. The
weakening of the SASM circulation is directly related to
the decrease of upper-tropospheric TP-TIO thermal
contrast, which in turn is caused by the larger uppertropospheric warming over the TIO than over the TP. The
fact that the SASM weakens as the lower-tropospheric
thermal contrast increases in the 21st century implies a
smaller role of this thermal contrast in determining the
SASM intensity than suggested by previousstudies for the
20th century.
(4) Uncertainty of modeling the
enhancement of water vapour in upper
troposphere in tropics and subtropics