Atmospheric and Climate Sciences, 2011, 1, 1-8
doi:10.4236/acs.2011.11001 Published Online January 2011 (http://www.SciRP.org/journal/acs)
Copyright © 2011 SciRes. ACS
Spatio-Temporal Variations of Precipitation Extremes in
the Yangtze River Basin (1960-2002), China
Qiang Zhang1,2, Xiaohong Chen1,2, Becker Stefan3
1Department of Water Resources and Environment, Sun Yat-sen University, Guangzhou, China
2Key Laboratory of Water Cycle and Water Security in Southern China of Guangdong
High Education Institute, Sun Yat-sen University, Guangzhou, China
3Department of Environmental, Geographic, and Geological Sciences (EGGS), Lehman College,
The City University of New York, 315 Gillet Hall, 250 Bedford Park Blvd. West, Bronx, NY, USA
E-mail: ylian@illinois.edu
Received December 10, 2010; revised January 11, 2010; accepted January 21, 2010.
Abstract
Daily precipitation data during 1960-2002 from 150 stations in the Yangtze River basin were analyzed with
the help of linear trend analysis. Highest 5-day and 10-day precipitation amount (R5D and R10D) and per-
centile daily precipitation maxima (prec95p for 95th percentile and prec99p for 99th percentile) were accepted
as the precipitation extreme index. The frequency of the R5D and R10D was in downward trend, this phe-
nomenon is more obvious in the middle Yangtze River basin; The stations with total precipitation of R5D
and R10D are in significant upward trend (> 95% confidence level) are mostly located in the lower Yangtze
River basin and the south-western part of the Yangtze River basin; 2) the spatial distribution of the frequency
of total precipitation of the percentile daily precipitation maxima is similar to that of R5D and R10D. How-
ever the frequency of prec95p and prec99p is in significant upward trend. The upward trend of total precipi-
tation changes of prec95p and prec99p is more obvious than that of frequency of prec95p and prec99p. The
regions dominated by upward trend of frequency/total precipitation of prec95p and prec99p are also the low-
er Yangtze River basin and south-western part of the Yangtze River basin. Therefore the occurrence prob-
ability of the flash floods in the lower Yangtze River basin and south-western part of the Yangtze River ba-
sin will be greater.
Keywords: Extreme Precipitation, Linear Trend, Spatial Distribution Pattern, Yangtze River Basin
1. Introduction
In recent years, the greenhouse effect has become the
focus of large-scale investigations. Tremendous impacts
of the observed global warming on the global hydrologi-
cal cycle at various spatial and temporal scales are likely
to be more sensitive [1,2]. Some physical and empirical
research results, together with General Circulation Model
(GCM) experiments suggested that global warming may
lead to a more intense hydrological cycle and with an
associated increase in the frequency and/or magnitude of
heavy precipitation [3]. Effects of precipitation changes
on hydrological process rely on the type of changes in
number of rainy days and/or daily precipitation amount.
Increase in daily precipitation amount will be favorable
for increase of watershed runoff and river discharge [4].
The global warming may lead to the increasing inten-
sity and magnitude of extreme climatic events. Extreme
climatic changes may exert more serious negative im-
pacts on human society. Increases of extreme climatic
events are drawing increasing concerns from meteorolo-
gists and hydrologists of the world [5-7]. Some studies
indicate that over the last half century weather patterns
have become more variable, with more frequent and
more intense rainfall events [8,9], changes in the timing
and location of precipitation [5]. The Yangtze River
(Changjiang) (Figure 1), being the longest river in China
and the third longest river in the world, plays a vital role
in the economic development of China. The river origi-
nates in the Qinghai-Tibet Plateau and flows about 6300
km eastwards to the East China Sea. The Yangtze floods
occurred annually, being the most serious natural disas-
ters with tremendous anti-disaster expenditures, being
known as serious trouble of China [10]. Historical records
2 Q. ZHANG ET AL.
Figure 1. Location of the Yangtze River basin.
of the Yangtze floods during recent 200 years are com-
plete. 19th century is the third cold period of the global
little ice age, there are 8 floods occurred to the Yangtze
River basins. 20th century is characterized by global
warming, about 19 floods occurred to the Yangtze River.
The frequency is clearly higher than that of 19th century.
The close connections between the warmest years and
high frequency of floods are worthy of attention. Con-
nections between temperature, precipitation and dis-
charge of the Yangtze River basin indicate that discharge
changes are in fine agreement with precipitation. If seen
from annual changes, good agreement also happened to
temperature, precipitation and discharge [11]. Severe
precipitation events are among the most devastating
weather phenomena since they are frequently followed
by flash floods and other sever weather conditions such
as hail [12]. It is widely accepted that variability of fre-
quency and intensity of extreme climatic events are
likely to exert much more impact on nature and human
society than that from the mean climatic value [13].
Changes of extreme climatic events and their possible
influences on human society are receiving increasing
concerns from hydrologists and meteorologists of the
world [14-16]. As for the extreme precipitation changes
in the Yangtze River basin, Zhang et al., [11] analyzed
the spatial distribution and trends in the frequency of
precipitation extremes during 1960-2003 using daily
precipitation data from 147 stations in the Yangtze River
basin with the help of Mann-Kendall trend analysis and
IDW (Inverse Distance Weighted) interpolation tech-
nology with Arcview package. The definition of the ex-
treme precipitation is that the daily precipitation exceed-
ing the 95th percentile in a data set is defined as extreme
precipitation for each station. In the present paper, daily
precipitation exceeding the 95th percentile and 99th per-
centile was defined as extreme precipitation of two mag-
nitudes. At the same time, 5-day extreme daily precipita-
tion and 10-day extreme daily precipitation were also
analyzed for further research.
The main objective of this paper is to demonstrate the
changes of the extreme precipitation in different parts of
the Yangtze River basin, showing the possible regional
responses to global climatic changes.
2. Data and Methods
Considering the influences of topography on climatic
changes, in this paper, the whole Yangtze River basin
was divided into three parts [11] based on longitude
(Table 1). Daily precipitation data of 150 stations in the
Yangtze River basin were from China Meteorological
Administration (CMA). There are 35 stations located in
the upper Yangtze River basin, 73 stations located in the
middle Yangtze River basin and 42 stations located in
Table 1. Three regions of the Yangtze River basin analyzed
in the paper.
Upper Middle Lower
Longitude ~104 E 104-113 E 113 E ~
Mean altitude (m. a.s.l) 2551 627 113
Number of stations 10 27 14
Copyright © 2011 SciRes. ACS
Q. ZHANG ET AL.
3
the lower Yangtze River basin. The time interval is from
January, 1960 to December, 2002. The homogeneity of
the precipitation data was detected by calculating the von
Neumann ratio (Q/n-0.5 and R/n-0.5) with the help of
Bayesian procedures [17]. The precipitation data sets of
all stations are significantly homogeneous. Highest 1 day,
5 day and 10 day precipitation amount (R1D, R5D and
R10D) are accepted as index for extreme precipitation
(www.ncdc.noaa.gov/oa/wmo/ccl) for annual extreme
precipitation. Percentile daily precipitation/temperature
maxima (prec95p for 95th percentile and prec99p for 99th
percentile) was widely accepted as a standard to define
daily extreme precipitation/tempera- ture [12,18,19].
Simple linear regression will be used in this paper for
trend test. The simple linear regression method is a pa-
rametric T-test method, which consists of two steps, fit-
ting a linear simple regression equation with the time t as
independent variable and the precipitation variable, Y as
dependent variable, and testing the statistical significance
of the slope of the regression equation. The parametric
T-test requires the data to be tested is normally distrib-
uted. The normality of the data series is first tested in the
study by applying the Kolmogorov-Smirnov test. The
method first compares the specified theoretical cumula-
tive distribution function (in our case normal distribution)
with the sample cumulative density function based on
observations, then calculates the maximum deviation, D,
of the two. If, for the chosen significance level, the ob-
served value of D is greater than or equal to the critical
tabulated value of the Kolmogorov-Smirnov statistic, the
hypothesis of normal distribution is rejected.
3. Results
3.1. Annual Extreme Precipitation Changes
Figure 2 shows the linear trends of the highest 5-day
precipitation amount, Figure 2A and Figure 2B show
the trends of frequency and total precipitation amount of
highest 5 day precipitation amount respectively. Figure
2A indicates that frequency of R5D of 69% stations of
the Yangtze River basin is in downward trend, (without
considering whether the trend is significant or not). The
stations with frequency of R5D in upward trend are
mostly located in the south-western part and south-east-
ern part of the Yangtze River basin. The Middle Yangtze
River basin is dominated by stations with frequency of
R5D in downward trend. In the Yangtze Delta region, the
downward trend of the frequency of R5D are significant
at > 95% confidence level. In the middle Yangtze River
basin, frequency of R5D of 81% stations is in downward
trend (Figure 3).
Figure 2B and Figure 3 demonstrate that total pre-
cipitation of R5D of the Yangtze River basin is mostly in
upward trend, the stations with total precipitation of R5D
in upward trend account for about 63% of the total sta-
tions. Total precipitation of R5D of 27 stations is in sig-
nificant trend, wherein, stations with total precipitation
of R5D in significant upward trend account for 67% of
the total stations with total precipitation of R5D in sig-
nificant trend. Figure 2B also indicates that stations with
total precipitation of R5D in significant trend are mostly
located in lower Yangtze River basin, and those in sig-
nificant downward trends are mostly located in the mid-
dle Yangtze River basin.
Changes of frequency/total precipitation of R10D
show the similar patterns. Stations with frequency of
R10D in downward trend account for about 53% of the
total stations, wherein, 25 stations are in significant trend.
There are 15 stations with frequency of R10D in signify-
cant upward trend and 12 stations with frequency of
R10D in significant downward trend. It seems that fre-
quency of R10D of stations in significant upward-trend
is mostly located in upper Yangtze River basin (Figure.
4A, Figure 3). As for the total precipitation of R10D
(TR10D), altogether 96 stations have TR10D in upward
(A)
Copyright © 2011 SciRes. ACS
4 Q. ZHANG ET AL.
(B)
Figure 2. Spatial distribution of trends of highest 5-day precipitation amount. A: Frequency of highest 5 day precipitation
amount; B: Total of highest 5-day precipitation amount.
Figure 3. Ratio of stations with total precipitation in up-
ward/downward trend to the total stations (B) and ratio of
stations with frequency of extreme precipitation in signifi-
cant upward/downward trend (> 95% confidence level) to
the total stations with frequency of extreme precipitations
in significant upward/downward trend (> 95% confidence
level) (A) in upper, middle and lower reaches of the Yang-
tze River basin respectively
trend (whether this trend is significant or not) in the
Yangtze River basin. Most stations with TR10D in up-
ward trend are mostly located in the lower Yangtze River
basin (Figure 4B, Figure 3). The middle Yangtze River
basin is dominated by the stations with TR10D in
downward trend.
When it comes to the stations with TR10D in signifi-
cant trend, altogether 37 stations have TR10D in signifi-
cant trends, wherein 20 stations have TR10D in signifi-
cant upward trend (> 95% confidence level), accounting
for 74% of the total stations with TR10D in significant
trends (Figure 3). 11 stations with TR10D in significant
upward trend are located in the middle Yangtze River
basin and 9 stations located in the upper Yangtze River
basin, accounting for 69% and 82% of the total station in
the middle and lower Yangtze River basin respectively.
3.2. Seasonal Extreme Precipitation Changes
Summer (June to September) is the main flooding season
of the Yangtze River basin. In this paper, percentile daily
precipitation maxima (prec95p for 95th percentile and
prec99p for 99th percentile) for summer season were ac-
cepted for detection of summer precipitation extreme in
the Yangtze River basin.
Figure 5 indicates that the upper and lower Yangtze
River basin are dominated by and total precipitation of
prec95p in significant upward trend. The regions domi-
nated by significant upward trend of total precipitation
and frequency of prec95p are mainly the Yangtze Delta
region and south-western part of the Yangtze River basin.
The middle Yangtze River basin is dominated by sig-
nificant downward trend of frequency of prec95p (the
stations with precipitation in significant downward trend
of frequency of prec95p account for about 52% of the
total stations with frequency of prec95p in significant
trend) (Figure 6). The spatial distribution of frequency
and total precipitation of prec99p in the Yangtze River
basin is similar to that of prec95p (Figure 7). South part
of the Yangtze River basin is dominated by significant
upward trend of total precipitation changes of prec99p;
however the north part of the Yangtze River basin is
dominated by downward trend of total precipitation
changes of prec99p (Figure 7B, Figure 6). As for the
Copyright © 2011 SciRes. ACS
Q. ZHANG ET AL.
5
(A)
(B)
Figure 4. Spatial distribution of trends of highest 10-day precipitation amount. A: Frequency of highest 10-day precipitation
amount; B: Total of highest 10-day precipitation amount.
Figure 5. Ratio of stations with precipitation in upward/downward trend to the total stations (B) and ratio of stations with
precipitation in significant upward/downward trend (>95% confidence level) to the total stations with precipitations in sig-
nificant upward/downward trend (>95% confidence level) in upper, middle and lower reaches of the Yangtze River basin
respectively (A).
Copyright © 2011 SciRes. ACS
6 Q. ZHANG ET AL.
(A)
(B)
Figure 6. Spatial distribution of trends of 95th percentile daily precipitation maxima. A: Frequency of 95th percentile daily
precipitation maxima; B: Total of 95th percentile daily precipitation maxima.
changes of frequency of prec99p (Figure 7A), mostly
stations with upward trends of precipitation of prec99p
are concentrated in the lower Yangtze River basin and
south-western part of the Yangtze River basin (Figure 6).
4. Summary and Conclusions
Extreme climatic events will exert tremendous influences
on human society. Frequency and total precipitation of
the highest 5 day and 10 day precipitation amount (R5D
and R10D) and percentile daily precipitation maxima
were analyzed with the help of linear trend analysis.
Some interesting results were obtained: 1) the frequency
of the highest 5-day and 10-day precipitation amount
(R5D and R10D) was in downward trend, some stations
are with frequency of R5D and R10D in significant
downward trend, this phenomenon is more obvious in the
middle Yangtze River basin; however the changes of the
total precipitation of R5D and R10D are in upward trend.
The stations with total precipitation of R5D and R10D
are in significant upward trend (> 95% confidence level)
are mostly located in the lower Yangtze River basin and
the south-western part of the Yangtze River basin; 2) the
spatial distribution of the frequency of total precipitation
of the percentile daily precipitation maxima is similar to
that of R5D and R10D. What is different is that the fre-
quency of prec95p and prec99p is in significant upward
trend. However the upward trend of total precipitation
changes of prec95p and prec99p is more obvious than that
of frequency of prec95p and prec99p. The regions domi-
nated by upward trend of frequency/total precipitation of
prec95p and prec99p are also the lower Yangtze River
basin and south-western part of the Yangtze River basin;
3) the global warming greatly influenced the spatial dis-
tribution pattern of the precipitation changes in the
Yangtze River basin of China. More intensified precipi-
tation will occur to the south-western part of the Yangtze
River basin and lower Yangtze River basin. Therefore,
the occurrence probability of the flash floods in the lower
Yangtze River basin and south-western part of the
Yangtze River basin will be greater. Furthermore, the
Yangtze Delta region is geomorphologically low-lying,
Copyright © 2011 SciRes. ACS
Q. ZHANG ET AL.
7
Figure 7. Ratio of stations with summer (June - September) precipitation in upward/downward trend to the total stations (B)
and ratio of stations with summer (June - September) precipitation in significant upward/downward trend (>95% confidence
level) to the total stations with precipitations in significant upward/downward trend (>95% confidence level) in upper, middle
and lower reaches of the Yangtze River basin respectively (A).
(A)
(B)
Figure 8. Spatial distribution of trends of 99th percentile daily precipitation maxima. A: Frequency of 99th percentile daily
precipitation maxima; B: Total of 95th percentile daily precipitation maxima.
Copyright © 2011 SciRes. ACS
8 Q. ZHANG ET AL.
and prone to floods inundation and this region is one of
the economically developed region in China. More good
countermeasures will be necessary for future intensifying
floods threats.
5. Acknowledgements
The work described in this paper was financially sup-
ported by the National Natural Science Foundation of
China (Grant No.: 41071020), Project of the Guangdong
Science and Technology Department (Grant No.: 2010B
050800001), the Program for Outstanding Young Teach-
ers of the Sun Yat-sen University (Grant No.: 2009-37000-
1132381), the Key National Natural Science Foundation
of China (Grant No.: 50839005), a grant from the Re-
search Grants Council of the Hong Kong Special Ad-
ministrative Region, China (CUHK405308).
6. References
[1] B. C. Hewitson, A Methodology for Developing Re-
gional Climate Change Scenarios from General Circula-
tion Models, WRC Report, Vol. 1, No. 594, 1997, pp.
35.
[2] D. Labat, Y. Goddéris, J. L. Probst, J. L. Guyot, “Evi-
dence for Global Runoff Increase Related to Climate
Warming,” Advances in Water Resources, Vol. 27, 2004,
pp. 631-642.
[3] A. M. Fowler, and K. J. Hennessy, “Potential Impacts of
Global Warming on the Frequency and Magnitude of
Heavy Precipitation,” Natural Hazards, Vol. 11, 1995, pp.
283-303.
[4] A. O. Lucero, D. Rozas, “Characteristics of Aggregation
of Daily Rainfall in a Middle-Latitudes Region during a
Climate Variability in Annual Rainfall Amount,” Atmos-
pheric Research, Vol. 61, 2002, pp. 35-48.
[5] E. B. Horton, C. K. Folland, and D. E. Parker, “The
Changing Incidence of Extremes in Worldwide and Cen-
tral England Temperatures to the End of the Twentieth
Century,” Climate change, Vol. 50, 2001, pp. 67-295.
[6] N. Plummer, M. James, S. Neville, N. Ramasamy,
Changes in Climate Extremes over the Australian Re-
gion and New Zealand during the Twentieth Century,”
Climate Change, Vol. 42, 1999, pp. 183-202.
[7] P. Y. Groisman, T. Karl, D. Easterling, R. W. Knight,
“Changes in the Probability of the Heavy Precipitation:
Important Indicators of the Climate Change,” Climate
Change, Vol. 42, 1999, pp. 243-283.
[8] T. R. Karl, R. W. Knight, and N. Plummer, “Trends in
High-Frequency Climate Variability in the Twentieth
Century,” Nature, Vol. 377, 1995, pp. 217-220.
[9] D. R. Easterling, G. A. Meehl, C. Parmesan, S. A.
Changnon, T. R. Karl, and L. O. Mearns, “Climate Ex-
tremes Observations, Modelling, and Impacts,” Science,
Vol. 289, 2000, pp. 2068-2074.
[10] Changjiang Water Resources Commission, Ministry of
Water Resources(CWRC), “Waterlogging and Aridity
Disasters in the Yangtze River Basins,” Chinese Hydrau-
lic Power and Electric Force Press, Beijing, 2002.
[11] Q. Zhang, T. Jiang, M. Germmer, S. Becker, “Precipita-
tion, Temperature and Discharge Analysis from 1951 to
2002 in the Yangtze River Basin, China,” Hydrological
Sciences Journal, Vol. 50, No. 1, 2005, pp. 65-80.
[12] C. Jones, E. D. Waliser, M. K. Lau, W. Stern, “Global
Occurrence of Extreme Precipitation and the Mad-
den-Julian Oscillation: Observations and Predictability,”
American Meteorological Society, Vol. 17, 2004, pp.
4575-4589.
[13] R. W. Katz, and B. G. Brown, “Extreme Events in a
Changing Climate: Variability is More Important than
Average,” Climatic Change, 21, 1992, pp. 289-302.
[14] R. Suppiah, and K. Hennessy, “Trends in the Intensity
and Frequency of Heavy Rain in Tropical Australia and
Links with the Southern Oscillation,” Australian Mete-
orological Magazine, Vol. 45, 1996, pp. 1-17.
[15] R. D. Mora, C. Bouvier, L. Neppel, H. Niel, “Regional
Approach for the Estimation of Low-Frequency of
Low-Frequency Distribution of Daily Rainfall in the
Languedoc-Roussillon Region, France,” Hydrol. Sci. J.,
Vol. 50, No. 1, 2005, pp. 17-29.
[16] P. M. Zhai, A. J. Sun, F. M. Ren, “Changes of Climate
Extremes in China,” Climatic Change, Vol. 42, No. 1,
1999, pp. 203-218.
[17] T. A. Buishand, “Some Methods for Testing the Homo-
geneity of Rainfall Records,” J. Hydrol., Vol. 58, 1982,
pp. 11-27.
[18] J. L. Bell, L. C. Sloan, and M. A. Snyder, “Regional
Changes in Extreme Climatic Events,” Journal of Cli-
mate, Vol. 17, No. 1, 2004, pp. 81-87.
[19] S. N. Diffenbaugh, S. J. Pal, J. R. Trapp, F. Giorgi,
“Fine-Scale Processes Regulate the Response of Extreme
Events to Global Climate Change,” Geophysics, Vol. 102,
No. 44, 2005, pp. 15774-15778.
Copyright © 2011 SciRes. ACS