Journal of Hydrology 530 (2015) 462–475
Contents lists available at ScienceDirect
Journal of Hydrology
journal homepage: www.elsevier.com/locate/jhydrol
Homogenization of precipitation and flow regimes across
China: Changing properties, causes and implications
Qiang Zhang a,b,c,d,⇑, Xihui Gu a,b, Vijay P. Singh e, Chong-Yu Xu f, Dongdong Kong a,b, Mingzhong Xiao a,b,
Xiaohong Chen a,b,c
a
Department of Water Resources and Environment, Sun Yat-sen University, Guangzhou 510275, China
Key Laboratory of Water Cycle and Water Security in Southern China of Guangdong High Education Institute, Sun Yat-sen University, Guangzhou 510275, China
Guangdong Key Laboratory for Urbanization and Geo-simulation, Sun Yat-sen University, Guangzhou 510275, China
d
School of Civil Engineering and Environment, Suzhou University, Anhui 234000, China
e
Department of Biological & Agricultural Engineering and Zachry Department of Civil Engineering, Texas A&M University, College Station, TX 77843-2117, USA
f
Department of Geosciences and Hydrology, University of Oslo, Norway
b
c
a r t i c l e
i n f o
Article history:
Received 6 March 2015
Received in revised form 6 August 2015
Accepted 15 September 2015
Available online 25 September 2015
This manuscript was handled by Geoff
Syme, Editor-in-Chief, with the assistance of
Bellie Sivakumar, Associate Editor
Keywords:
Homogenization
Gini coefficients
ANOSIM analysis
Precipitation regimes
Streamflow regimes
s u m m a r y
Homogenization and similarities of precipitation and flow regimes across China are thoroughly
investigated using Gini coefficient analysis method and the Analysis Of Similarity (ANOSIM) technique,
respectively based on daily precipitation data from 554 meteorological stations and monthly streamflow
data from 370 hydrological stations covering the period of 1960–2000. The results indicate that: (1)
Homogenization of precipitation regimes is increasing from the northwest to the southeast China.
However, different spatial patterns of homogenization of flow regions are identified. Spatially, lower
homogenization of flow regimes is detected in the northeast China and higher homogenization of flow
regimes in the central and southeast China. Temporally, flow regimes during 1961–2000 are characterized mainly by increasing homogenization, and it is particularly true after 1980; (2) precipitation regimes
during 1961–2000 are characterized by decreasing dissimilarities. Larger areas of China are characterized
by decreasing dissimilarities of precipitation regimes during 1980–2000 when compared to those during
1961–1980, which should be due to increasing precipitation concentration and intensifying precipitation
regimes in recent years; (3) distinctly dissimilar precipitation and flow regimes can be identified between
geographically separate river basins. Interregional similarities of flow regimes are enhancing after 1980
when compared to those before 1980 though interregional similarities of precipitation regimes are not
changed much; and (4) spatial mismatch is evident in terms of spatial range and changing degree of flow
and precipitation regimes. Roughly spatial match can be observed between changes of flow and precipitation indicates and it is particularly the case for precipitation and flow changes in dry season such as
winter in China. However, influences of human activities and precipitation changes on streamflow are
varying as for specific river basins, such as the Yangtze and the Yellow River basins. Damming-induced
fragmentation of river basins is the major cause behind higher homogenization of flow regimes. Thus,
human interferences in hydrological processes via damming and construction of reservoirs greatly alter
streamflow vs. precipitation relations. The results of this study provide theoretical and practical grounds
for management and planning of water resources at basin or interbasin scales under the influences of
human activities and climate changes.
Ó 2015 Elsevier B.V. All rights reserved.
1. Introduction
Investigation of hydrological processes is crucial for basin-scale
water resources management and conservation of fluvial
⇑ Corresponding author at: Department of Water Resources and Environment,
Sun Yat-sen University, Guangzhou 510275, China. Tel./fax: +86 20 84113730.
E-mail address: zhangq68@mail.sysu.edu.cn (Q. Zhang).
http://dx.doi.org/10.1016/j.jhydrol.2015.09.041
0022-1694/Ó 2015 Elsevier B.V. All rights reserved.
ecosystem, and is also the first step into study of influences of
climate changes and human activities on hydrological cycle at
regional and global scale (Zhang et al., 2009a, 2014; Dinpashoh
et al., 2011; Xia et al., 2012; Zhou et al., 2014). Streamflow is
affected by diverse natural factors, such as precipitation, temperature, and fluvial underlying attributes, and by human activities,
e.g. irrigation, building of dam/reservoir, and land use and
land cover changes (Zhang et al., 2013; Tran and O’Neill, 2013;
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Li et al., 2014; McIntyre et al., 2013; Gosling, 2014). Due to increasing population and intensifying human activities and also the
subsequent impact of humans on the hydrologic cycle, growing
interest appears concerning how hydrologic variables are affected
by external forcing such as the human activities (Zhang et al.,
2013; Zhan et al., 2013; Ahn and Merwade, 2014). Human activities such as construction of single and cascade reservoirs and water
diversion tunnels/channels have further altered the hydrology and
water quality of rivers (Li et al., 2014). Moreover, since natural flow
variability is of great importance for maintaining ecological integrity and diversity, any alterations resulting from human impacts
might induce complex responses or the potential degradation of
riverine ecosystems (Petts, 2009; Yin and Yang, 2011). Therefore,
investigation of hydrological alterations in terms of frequency,
duration, and timing of flow regimes as results of changing climate
and human activities has been arousing increasing human
concerns in recent decades (Jiang et al., 2014).
Eco-hydrological effects of reservoirs have been widely investigated due to the fact that hydro-environmental changes due to
human alterations of natural flows may significantly influence
hydrological regimes and thus the ecology of rivers (Li et al.,
2014; Zhang et al., 2014; Iacob et al., 2014). In China, there are a
bunch of researches addressing influences of reservoirs on downstream flow regimes and related eco-hydrological effects (Yin
and Yang, 2011; Li et al., 2014). Besides, impacts of climate
changes, particularly precipitation variations, have been widely
investigated (Ma et al., 2010; Liu et al., 2013; Xu et al., 2013).
Bao et al. (2012) analyzed the attribution of climate variability
and human activities for streamflow decrease in three catchments
located in different parts of the Hai River basin, i.e. Taolinkou,
Zhangjiafen and Guantai catchments. Tang et al. (2011) presented
a geomorphology-based non-point source pollution (GBNP) model
that links the processes of rainfall–runoff, soil erosion, sediment
routing, and pollutant transport of the Miyun reservoir, Beijing,
China. There are also other related studies addressing streamflow
regimes and related impacts from human activities and climate
changes and these researches will not be enumerated here with
details. The above-particularized researches are theoretically and
practically significant in the development of human understanding
of flow regimes and related impacts on ecological environment
under the influences of human activities and climate changes.
However, these researches focused on flow regimes of river basin
at local scale.
Rivers in China are heavily regulated and fragmented by
reservoirs and other hydraulic facilities. There are 98,002 reservoirs or hydraulic infrastructures with storage capacity of >0.1 million m3, and the total storage capacity of the reservoirs is about
932.3 billion m3, accounting for 34.5% of the total streamflow of
the rivers in China (Sun et al., 2013). Parts of the reservoirs are
shown in Fig. 1. Besides, China is characterized by different climate
types with different underlying attributes, and different river
basins are dominated by diverse intensity of human activities. In
this case, it is practically and scientifically important investigate
spatiotemporal variations of flow regimes across whole China.
However, such reports have not been found so far and this is the
major motivation of this study.
This study aims to investigate changes of flow regimes and
precipitation changes based on long-term hydro-meteorological
records across China by clarifying diverse influences of reservoirs
on flow regimes at different river basins over China and also possible impacts on diversity of biota of river basins in China. The
results of this study are of significant relevance in terms of
development of human knowledge concerning spatiotemporal
variations of flow regimes over China under influences of changing
climate and human activities and are also crucial in basin-scale
water resources management and training of river basins.
463
2. Data
Adequately considering the climate conditions and geographical
features of river basin, 31 provinces of China were divided into ten
large river basins based on the Ministry of Water Resources of China
at http://www.mwr.gov.cn/zwzc/hygb/szygb/, i.e. Songhua River
basin (SHR), Liao River basin (LR), Hai River basin (HR), Huai River
basin (HuR), Yangtze River basin (YTR), Yellow River basin (YR),
Pearl River basin (PR), Southeast Rivers (SER), Southwest Rivers
(SWR) and Northwest Rivers (NWR) (Fig. 1). Table 1 displays
detailed information of these ten river basins. It can been seen from
Table 1 that precipitation is the main driving factor for streamflow
variations in almost all the river basins. However, human activities
play an important role in interfering precipitation vs. streamflow
relationships, such as surface water withdrawal (Table 1), irrigation,
reservoir storage and land use. In this case, other than influences of
precipitation on streamflow variations, combined impacts of precipitation changes and human activities and also other natural factors such as fluvial topography (Fig. 1b) surface water withdrawal,
agricultural irrigation, reservoir storage and land use. Daily precipitation data covering the period of 1961–2000 from 554 meteorological stations and monthly streamflow covering the period of
1960–2000 from 370 hydrological stations are collected. The hydrological data are from the Ministry of Water Resources of China and
the meteorological data are collected from National Meteorological
Information Center of the China Meteorological Administration. The
quality of precipitation and streamflow data were firmly controlled
and were applied in our previous studies (e.g., Zhang et al., 2011b,
2012). Both precipitation and streamflow datasets contain small
amount of missing values. There are 38 rain gauge stations containing missing daily precipitation data. However, only one station had
1.09% missing values, and most of others had less than 0.1% of total
missing values. The missing values in these series were replaced by
the long-term average of the same days or months of other years
(Xiao et al., 2014). For example, for streamflow data, if there is a
missing value in May in the year 1980, the missing value would
be replaced by the average of the streamflow values of May of the
study time period (excluding 1980). Besides, information of building time, storage capacity and locations of the large reservoirs
whose total capacity is more than 100 million built before 2000
are also collected from Hydrology Bureau of Ministry of Water
Resources of China at http://xxfb.hydroinfo.gov.cn/ssIndex.html?
type=2. The irrigation data from 1978 to 2000 and surface water
withdrawal data of 31 provinces are collected from National Bureau
of statistics of China at http://www.stats.gov.cn/ and China Water
Resources Bulletin in 2000, respectively. The data of land use and
topography are provided by Data Center for Resources and
Environmental Sciences, Chinese Academy of Sciences (RESDC) at
http://www.resdc.cn. The locations of the hydrological stations,
meteorological stations and also reservoirs can be found in Fig. 1a.
3. Methodologies
3.1. Gini coefficient
The Gini coefficient (also known as the Gini index or Gini ratio)
is a measure of statistical dispersion with aim to mirror the income
distribution of a nation’s residents, and is the most commonlyused measure of inequality. In this study, the Gini coefficient is
used to analyze the annual and inter-annual distribution of
streamflow regimes. For the sake of completeness, the computation of Gini coefficient is introduced as follows.
Assume that Y is a positive random variable and denotes the
streamflow in this study. The accumulative probability distribution
of Y is F(x), i.e.
464
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Fig. 1. River basins map and also locations of the precipitation stations, hydrological stations and large-scale water reservoirs across China (a) and topography in China
(b). Ten river basins considered in this study are: Songhua River basin (SHR), Liao River basin (LR), Hai River basin (HR), Huai River basin (HuR), Yangtze River basin (YTR),
Yellow River basin (YR), Pearl River basin (PR), Southeast Rivers (SER), Southwest Rivers (SWR) and Northwest Rivers (NWR).
Table 1
Ten river basins considered in this study (China Water Resources Bulletin in 2000).
River
basin
Basin area
(104 km2)
Amount of precipitation
(108 m3)
Amount of surface runoff
(108 m3)
Streamflow-supplying
source
Surface water withdrawal
(108 m3)
SHR
LR
HR
HuR
YTR
YR
PR
SER
SWR
NWR
55.7
21.9
31.82
3.1
180
75.2
45.3
24
85
336
5415.68
1122.74
347.78
1559.36
3062.29
19561.45
3043.46
8548.94
3723.67
9517.54
5659.95
125.18
877.09
9924.09
456.07
4401.16
2117.04
6122.46
1416.11
Precipitation and snow
Precipitation
Precipitation
Precipitation
Precipitation
Precipitation
Precipitation
Precipitation
Precipitation
Glacier and snow
135.92
373.46
1640.48
256.04
291.71
792.72
304.41
493.69
465
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Table 2
Hydrological indices defined in this study addressing hydrological variations over China and the same as precipitation indices using the same symbols.
Hydrological
indices
Definitions
Calculation method
W
Seasonal coefficient of winter flow
SP
Seasonal coefficient of spring flow
SU
Seasonal coefficient of summer flow
F
Seasonal coefficient of autumn flow
W/SP
SP/SU
SU/F
F/W
MAM
Ratio between winter flow and spring flow
Ratio between spring flow and summer flow
Ratio between summer flow and autumn flow
Ratio between autumn flow and winter flow
Monthly coefficient of the maximum monthly
average flow
Monthly coefficient of the minimum monthly
average flow
Monthly coefficient of immoderation
Coefficient of variation
Timing of maximum monthly average flow
Timing of minimum monthly average flow
The ratio between the average of the total monthly flow during December and February of the next
year and the total annual flow
The ratio between the average of the total monthly flow during March and May of the next year and
the total annual flow
The ratio between the average of the total monthly flow during June and August of the next year and
the total annual flow
The ratio between the average of the total monthly flow during September and November of the next
year and the total annual flow
Quotient of W and SP
Quotient of SP and SU
Quotient of SU and F
Quotient of F and W
Ratio between the maximum monthly average flow and total annual flow (%)
MIM
CI
CV
TMAX
TMIM
Ratio between the minimum monthly average flow and total annual flow (%)
Ratio between the minimum and the maximum monthly average flow
Ratio between the standard deviation and the average monthly flow (%)
Average month of occurrence of the maximum monthly average flow
Average month of occurrence of the minimum monthly average flow
FðxÞ ¼ PðY 6 xÞ
ð1Þ
It can be deduced from Eq. (1) that:
F 1 ðuÞ ¼ inffx : FðxÞ > ug; ð0 6 u < 1Þ
ð2Þ
where F 1 ð1Þ ¼ 1, it can be further deduced from Eq. (2) that:
LðpÞ ¼
1
l
Z
p
F 1 ðuÞdu; ð0 6 p 6 1Þ
ð3Þ
0
A
¼12
D
Z
1
LðpÞdp
n
X
djk ¼
100jyij yik j
i¼1
where L(p) is the Lorenz function of the random variable Y, and l
the expectation of Y. Assume that A denotes the area between the
Lorenz curve and the line by y = x, and D denotes the area circled
by y = x, x = 1 and x axis, then the Gini coefficient, G, can be computed as:
G¼
between sample units is analyzed using the Bray–Curtis dissimilarity coefficient. The dissimilarity, djk , between any two samples, j
and k, can be computed as (Warton et al., 2012):
ð4Þ
0
In this study, x denotes months from January to December and
random variable, Y, denotes monthly streamflow of x, and then
Gini coefficient can reflect the homogenization degree of the
annual distribution of the monthly streamflow. Larger Gini
coefficient indicates larger nonuniform degree and vice versa.
The Gini coefficient can be computed using the R package ineq
from http://cran.r-project.org/web/packages/ineq/ index.html.
3.2. Analysis of similarities (ANOSIM)
In ecohydrology, the natural flow regime can be characterized
by four attributes: magnitude, frequency, duration and timing
(Poff and Zimmerman, 2010) and each individual hydrological
attribute mentioned above can have crucial impacts on biotic
ecosystem (Richter et al., 1996). Richter et al. (1996) formulated
33 hydrological indicators to quantitatively depict hydrological
alteration, and which has been widely used in eco-hydrological
domain (Richter et al., 1997; Chen et al., 2010). With respect to
the monthly streamflow and precipitation processes in this study,
14 hydrological indicators are defined and used in this study
(Table 2) (Matteau et al., 2009).
Based on above-mentioned 14 hydrological indicators, the
ANOSIM technique is used to evaluate the statistical similarity
(Poff et al., 2007). The river systems of the 10 river basins are taken
as the basic sample units and the dissimilarity of flow regimes
,
n
X
jyij þ yik j
ð5Þ
i¼1
where yij is the biotic richness of the ith species of the jth sample,
and n is the number of the species. Based on Eq. (5), the ANOSIM
statistic R value showing dissimilarity between biota community
rB and rw can be computed as:
R ¼ ðr B r w Þ=ðNðN 1Þ=4Þ
ð6Þ
where N is the total number of the sample and the range of R lies
between 1 and 1. A larger positive R value indicates higher
similarity of the flow regimes in a hydrological region, and vice
versa. The flow series of a river basin can be divided as predam
and postdam series by the timing when the water reservoir was
built. The ANOSIM statistic R value for the predam and postdam series is then computed respectively. In recent years, few rivers are not
influenced by human activities. The human-induced influences on
flow regimes are intensifying due to the fact that the intensity
and the extent of human activities are increasing. Thus, it is hard
to find a natural river basin as a reference one to exclude effects
of climate changes on regional similarity of flow regimes. In this
case, precipitation changes at 554 stations are analyzed across
China to reflect impacts of climate changes on flow regimes with
ANOSIM statistic R value. Precipitation changes and regional-scale
similarity are put under consideration to investigate impacts of
reservoirs on flow regimes. The ANOSIM statistic R value is computed using the R package vegan from http://cran.r-project.org/
web/packages/vegan/index.html.
4. Results and discussions
4.1. Annual changes of similarity of flow regimes
Basically, it is difficult to identify the time point when human
activities start to exert influences on flow regimes. Generally, the
time when the reservoir was built is usually taken as the time point
when human activities begin to exert impacts on flow regimes
downstream to the reservoirs. It should be noted here that altogether information of 457 large reservoirs in terms of locations,
466
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Fig. 2. Construction time of water reservoirs in different hydrological regions across China. The two black short lines denote 5% (down) and 95% (up) quantiles, respectively.
The edges of the box denote 25% (down) and 95% (up) quantiles, respectively. The bold black line in the box denotes 50% quantile.
storage capacity and so on, and also streamflow data from 370
stations are collected. Flow regimes at one hydrological station
are usually impacted by more than one reservoir. Besides, spatial
distribution of reservoirs is uneven. Therefore, it is difficult to
decide the above-mentioned time point. Based on suggestions by
Poff et al. (2007), the average of the construction time of the reservoirs within a river basin is regarded as the dividing time point
which separates the entire flow series into predam and postdam
flow series. Fig. 2 depicts the temporal distribution of the construction time of reservoirs in the river basins considered in this study.
A closer look at Fig. 2 indicates that the 50% and 75% percentiles of
the reservoir construction time within most river basins are 1980
or before 1980. Economically speaking, 1980 is the time when
China started the economic reform which follows the subsequent
booming socioeconomic development of China, including industry
and agriculture, water resource supply and navigation. Therefore, it
is practically appropriate to make 1980 as the dividing time for
predam and postdam flow series.
Similar to streamflow and also for the sake of consistency of
analysis, the year of 1980 is also taken as the dividing time for precipitation series. Gini coefficients for streamflow and precipitation
series covering the entire time interval, i.e. 1960–2000 and for
subseries prior to and posterior to 1980 are analyzed (Fig. 3). It
can be seen from Fig. 3a–c that Gini coefficients of precipitation
series during three time intervals of 1961–2000, 1961–1980 and
1981–2000 are decreasing from northwest China to southeast
China. And the spatial patterns of Gini coefficients are in
good agreement with the spatial patterns of precipitation amount,
i.e. precipitation amount is increasing in the direction from
northwest to southeast China. Besides, precipitation events in the
northwest China are dominated by 1-day precipitation event.
Non-precipitation days are pretty dominant in the northwest
China. However, durations of consecutive precipitation regimes
in the southeast China are relatively longer, and 3-, 4- and 5-day
precipitation regimes are dominant in the southeast China
(Zhang et al., 2011a). In this sense, Gini coefficients of precipitation
changes as shown in Fig. 3a–c can well reflect spatial patterns in
intra-annual distribution of precipitation regimes.
Streamflow variations in both space and time are the results
of precipitation changes. In this sense, the spatial patterns of
Gini coefficients of streamflow should be similar to those of
precipitation changes. However, it can be seen from Fig. 3d–f that
spatial patterns of Gini coefficients of streamflow are relatively
complicated when compared to those of precipitation and no confirmative spatial patterns can be identified. Generally, the Gini
coefficients of streamflow in the northeast China are relative
higher when compared to other regions of China. Hydrological
processes in the northeast China are heavily influenced by precipitation changes. Being influenced by fluvial topography and also
annual and seasonal precipitation changes, the hydrological processes of the river basins in the northeast China are annually and
seasonally different with distinctly evident low and high flow seasons. Besides, the Gini coefficients of streamflow in mountainous
areas in northeast China are obviously higher than in plains, and
the spatial distribution of Gini coefficients of streamflow are more
consistent with Gini coefficients of precipitation in mountainous
regions than in plain (Figs. 1b and 3). In case of occurrence of abundant and concentrated precipitation in mountainous regions, magnitudes of streamflow in mountainous areas tend to increase
substantially due to prompt runoff processes as a result of larger
terrain slope. This kind of prompt runoff processes can direct influence uneven temporal and spatial distribution of streamflow variations in mountainous regions. In addition, higher Gini coefficients
of streamflow are in good line with Gini coefficients of precipitation in southwest china and the upper Yangtze River basin where
terrain slopes are all relatively larger (Figs. 1b and 3). All these
hydrological and topography attributes cause higher temporal dissimilarity of flow regimes and hence higher Gini coefficients.
Visual comparison between Fig. 3e and f indicates more fragmentations of river basins are dominated by higher Gini coefficients
during 1960–1980 when compared to those during 1981–2000,
which should be attributed to massive construction and functioning of huge amount of reservoirs. The hydrological regulations of
these reservoirs largely dampen the flow regimes with their
impoundment functions. Taking the Yangtze River basin as an
example, based on Xu (2005), up to the end of 1980s, there are
11,931 reservoirs appeared in the upper Yangtze River basin with
a total storage capacity of 2.05 1010 m3. The construction of the
Three Gorges Dam started in 1993 is 3.93 1010 m3 in the total
storage capacity. The Gezhouba Dam started its construction in
1970 and the operation started in earlier 1980s with the total storage capacity of 1.58 109 m3. The appearance of reservoirs greatly
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
467
Fig. 3. Spatial patterns of Gini coefficients of precipitation and streamflow series across China. The time intervals considered are shown in the panels.
altered the hydrological processes of the Yangtze River basin
(Zhang et al., 2006, 2008a, 2009b). Besides, Gini coefficients of flow
regimes in the Yellow River basin are also relatively low, and which
can also be attributed to hydrological regulations of reservoirs.
There are 238 out of 457 large-scale reservoirs were built in the
Yellow and the Yangtze River basins, and all these large-scale
reservoirs and other smaller-scale reservoirs heavily regulated
hydrological processes of these two river basins. Lower Gini coefficients of flow regimes in the northwest China are due to the fact
that the flow generation of the river basins in the northwest China
is from melting snow or icepack. Precipitation changes have little
influence on flow regimes, and it is particularly true for flow
regimes in the plain regions.
Besides spatial patterns of Gini coefficients of precipitation and
streamflow regimes across China, trends of Gini coefficients are
analyzed using modified Mann–Kendall trend test method
(Hamed and Rao, 1998; Daufresne et al., 2009). It can be observed
from Fig. 4a, c and d that trends in Gini coefficients of precipitation
changes are decreasing in the northwest, northeast and southeast
China and this decreasing trend in parts of these river basins is statistically significant at >95% confidence level. Increasing Gini coefficients are identified in the upper Yellow River basin, the middle
and the lower Yangtze River basin, indicating enhancing dissimilarity of intra-annual precipitation distribution. Besides, a closer look
at Fig. 4c and d indicates expanding river basins being dominated
by increasing Gini coefficients during 1981–2000 when compared
to those during 1961–1980. Researches addressing precipitation
extremes across the Yangtze River basin further corroborated the
observations of this study (Zhang et al., 2008b). Analysis results
by Zhang et al. (2008b) indicated intensifying precipitation
extremes in the middle and the lower Yangtze River basin. Spatial
patterns of trends in Gini coefficients of flow regimes are relatively
complicated without fixed spatial patterns. Generally, decreasing
trends of Gini coefficients can be identified in most parts of China.
River basins with increasing Gini coefficients distribute sporadically among those with increasing Gini coefficients (Fig. 4b). In this
case, flow regimes tend to be similar in temporal distribution with
increasing low flow and decreasing high flow events. Specifically,
Fig. 4e and f illustrate that more parts of river basins are dominated
by decreasing Gini coefficients of flow regimes when compared to
those during 1960–1980, which apparently convince significant
human influences on flow regimes, particularly the construction
of large reservoirs. Intensifying human activities play enhancing
roles in alterations of flow regimes when compared to influences
from climate changes on flow regimes, and this result implies crucial attention should be paid to human factors in basin-scale water
resources management and water balance modeling studies.
In general, annual distribution of precipitation and streamflow
matches well in both space and time, however precipitation is only
one of the main factors influencing the streamflow changes in both
space and time. Therefore, human activities such as agricultural
irrigation, building of reservoirs and human withdrawal of
freshwater, are also analyzed (Fig. 5). Fig. 5 clearly illustrates that
irrigated agricultural areas are evidently increasing in most
468
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Fig. 4. Mann–Kendall Trend of Gini coefficients of precipitation and streamflow across China. The orange and red colors denote increasing and significantly increasing trend,
respectively. However, the light green and green color denote decreasing and significantly decreasing trend, respectively. (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)
Fig. 5. Human activities that have potential impacts on relationship between precipitation and streamflow. The human activities considered in this study are mainly irrigated
areas, total reservoir capacity and withdrawals of surface water.
469
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Table 3
ANOSIM-based homogenization analysis results of precipitation regimes in river basins across China.
River basins
R before 1980
before 1980
p
R after 1980
after 1980
p
SHR
LR
HR
HuR
YTR
SWR
SER
PR
NWR
YR
0.212
0.314
0.194
0.303
0.416
0.384
0.213
0.285
NA
0.163
0.010
0.002
0.059
0.025
0.001
0.028
0.093
0.001
NA
0.061
0.197
0.278
0.183
0.292
0.370
0.370
0.274
0.267
NA
0.178
0.013
0.005
0.097
0.023
0.001
0.017
0.047
0.001
NA
0.007
One-sided t test
t9 = 0.36, P > 0.5
Note: R the mean precipitation indices of the predam and postdam precipitation series, respectively. A positive R indicates increased precipitation
is significance computed based on permutation of R.
similarity and vice versa. p
Table 4
ANOSIM-based homogenization analysis results of streamflow regimes in river basins across China.
River basins
before 1980
R
before 1980
p
after 1980
R
after 1980
p
SHR
LR
HR
HuR
YTR
SWR
SER
PR
NWR
YR
0.160
0.160
0.039
0.278
0.247
0.168
0.206
0.419
0.471
0.173
0.019
0.074
0.339
0.102
0.038
0.127
0.172
0.001
0.011
0.035
0.257
0.285
0.182
0.358
0.359
0.254
0.255
0.482
0.629
0.249
0.018
0.009
0.021
0.041
0.005
0.049
0.128
0.001
0.003
0.013
One-sided t test
t10 = 2.33, P < 0.05
Note: R the mean hydrological indices of the predam and postdam streamflow series, respectively. A positive R indicates increased streamflow
is significance computed based on permutation of R.
similarity and vice versa. p
provinces of China and particularly the Hebei, Henan, Shangdong
and Anhui provinces in the Yellow River basin, Hai River basin
and Huai River basin, and the irrigated crop areas in these
provinces reach 4.548 106 ha, 5.081 106 ha, 4.955 106 ha,
3.520 106 ha, respectively, with water volume for irrigation of
1.72 1010 m3, 1.50 1010 m3, 1.94 1010 m3, 1.63 1010 m3
(Fig. 5a, and b). Liu and Zheng (2004) indicated that higher than
60% of the decrease of streamflow can be attributed to human
activities. Irrigation cause evident decrease of fluvial streamflow
with decreased annual variability of streamflow processes in the
Huang–Huai–Hai Plain (e.g. Zhang et al., 2011b). Therefore, annual
precipitation patterns in the Yellow River basin, Huai River basin
and Hai River basin are temporally uneven where precipitation
amount during wet seasons (June–September in the Yellow River
and Huai River basins, and May–October in the Hai River basin)
account for 60–70%, 50–80% and 50–60% of the total annual precipitation amount, and Gini coefficients of precipitation can reach
about 0.7. However, annual distribution of streamflow changes in
these above-mentioned river basins are relatively even with Gini
coefficients of about 0.3 (Fig. 3). Besides, the total reservoir capacities in the middle and lower Yangtze River basin, the Pearl River
basin and the Southeast Rivers are substantially larger than those
in other river basins, and the human withdrawal of freshwater is
also massive. Specifically, the total reservoir capacity in the Hubei,
Hunan, Jiangxi, Guangdong and Guangxi provinces is respectively
9.92 1010 m3, 4.02 1010 m3, 2.94 1010 m3, 4.29 1010 m3,
3.78 1010 m3 with water volume by human withdrawal of
2.59 1010 m3, 2.95 1010 m3, 2.05 1010 m3, 4.21 1010 m3,
2.80 1010 m3 (Fig. 5b, and d). Storage effects of reservoirs can
greatly alter the timing and magnitude of flows and substantially
reduce changing variability of streamflow processes. Thus, the Gini
coefficients of streamflow variations are evidently decreasing
(Fig. 4b) though increased Gini coefficients of precipitation changes
are observed in the middle and lower Yangtze River basin, the Pearl
River basin and the Southeast Rivers (Fig. 4a, c and d).
4.2. Regional homogenization
Regional homogenization of precipitation and flow regimes
within each river basin is investigated using ANOSIM techniques
(Tables 3 and 4). Regional similarity of precipitation regimes before
and after 1980 is analyzed (Table 3). Different spatial similarities of
precipitation events can be detected for the river basins considered
in this study. Spatial similarity of precipitation events before 1980
in the Yangtze River basin (YTR), the Southwest Rivers (SWR) and
the Liao River basin (LR) is statistically significant, implying significant spatial similarity of precipitation processes, and that in the
Yellow River basin (YR) and the Hai River basin (HR) is not statistically significant, implying insignificant spatial similarity, or spatial dissimilarity, of precipitation processes. Comparison between
values indicates that the spatial dissimilarities of precipitation
R
regimes after 1980 tend to be apparent when compared to those
before 1980. This result tends to indicate increasing spatial dissimilarities of precipitation changes after 1980 or enhancing uneven
spatial distribution of precipitation regimes across these river
basins considered in this study. However, hypothesis test results
using the one-sided student t test method indicate no significant
intensification of spatial dissimilarities of precipitation processes
within the river basins studied in this study after 1980 when
compared to that before 1980.
Analysis of the spatial similarities of flow regimes tells a
distinctly different story (Table 4). It can be seen from Table 4 that,
470
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
1.0
Precipitation R value
Hydrologic regions
Mean R
YR
NWR
PR
SER
SWR
YTR
HuR
HR
LR
SHR
0.37
0.39
0.07
0.68
0.86
0.2
0.45
0.37
0.17
0.12
NA
0.29
0.19
0.1
0.64
0.83
0.11
0.38
0.23
-0.01
NA
0.06
0.34
0.15
0.08
0.72
0.87
0.32
0.38
0.37
NA
0.15
0.21
0.27
0.28
0.15
0.22
0.58
0.2
0.02
NA
0.62
0.36
0.46
0.22
0.33
0.36
-0.03
-0.06
0.2
NA
-0.08
0.28
0.21
0.32
0.29
0.29
0.14
0.5
0.68
NA
0.15
0.3
0.23
0.19
0.27
0.34
0.4
0.16
0.57
0.38
0.24
0.64
0.26
0.28
0.11
NA
0.23
0.36
-0.01
0.42
0.23
0.4
0.41
0.06
0.12
NA
0.11
0.2
0.14
0.26
0.26
0.13
0.42
0.18
0.04
NA
0.24
0.29
0.29
0.19
0.53
0.13
0.59
0.82
0.03
NA
0.13
0.08
0.26
0.21
0.09
0.42
-0.01
-0.08
0.22
NA
0.08
0.03
0.41
0.59
0.47
0.2
0.65
0.28
0.99
NA
0.14
0.73
0.41
0.39
0.26
SHR
LR
HR
HuR
YTR
SWR
0.58
0.82
0.53
0.16
NA
0.52
-0.09
0.76
0.96
0.79
0.96
0.44
0.65
0.38
NA
0.2
0.46
0
0.18
0.69
0.56
0.67
0.23
0.31
NA
0.33
0.5
0.12
0.21
0.06
0.01
-0.03
0.04
0.38
NA
0.21
0.62
0.82
0.22
0.24
0.48
0.22
0.26
0.32
NA
0.38
0.16
0.41
0.6
0.27
0.14
0.35
0.37
0.28
0.37
0.39
0.45
0.68
0.14
NA
0.99
-0.05
0.52
0.42
0.3
0.38
0.27
0.16
0.46
NA
0.31
0.28
0
0.23
0.22
0.5
0.57
0.41
0.5
NA
0.53
0.67
0.59
0.33
0.46
0.43
0
0.16
0.28
NA
0.51
0.28
0.67
0.16
0.1
0.36
0.2
0.45
0.37
NA
0.35
0.41
0.32
0.47
0.44
0.18
0.32
0.26
0.27
0.33
SER
PR
NWR
YR
Mean R
0.8
0.6
Streamflow R value
Mean R
YR
NWR
PR
SER
SWR
YTR
HuR
HR
LR
SHR
0.4
0.2
0.0
Hydrologic regions
Fig. 6. ANOSIM-based R results of the relations between precipitation and streamflow in the 10 river basins of China. The upper panel shows the relations between
precipitation and streamflow during 1960–1980 and the lower panel during 1980–2000.
values are larger than 0, implywhether before or after 1980, the R
ing different degrees of spatial similarities of streamflow. The R
values of flow regimes in the Pearl River basin, the Northwest
rivers and the Yangtze River basin are large and are statistically
significant at 0.05 significance level, evincing higher spatial simi values of flow regimes
larities of the flow regimes; However, the R
in the Huai River basin, Hai River basin and Southwest rivers are
small and are not statistically significant at 0.05 significance level,
evincing low spatial similarities of the flow regimes. Comparison
values of flow regimes before and after 1980 in the river
between R
basins considered in this study indicates larger R values after 1980
than those before 1980, implying increasing spatial similarities or
homogenization of flow regimes. Besides, hypothesis test results
using the one-sided student t test indicate significant difference
of R values before and after 1980, showing enhancing spatial similarities of flow regimes of the river basins (Table 4). Furthermore,
comparison between Figs. 2 and 3 indicates decreasing spatial similarities of precipitation regimes but increasing spatial similarities
of flow regimes, and which well corroborates remarkable human
influences on spatial and temporal variations of flow regimes.
Fig. 6 illustrates ANOSIM-based R values showing the similarities of precipitation and flow regimes within different river basins.
It can be seen from Fig. 6 that distinctly dissimilar precipitation
and flow regimes can be identified between geographically separate river basins and these findings are in good agreement with
those by Poff et al. (2007) that neighboring regions may be less
likely to be homogenized, given their relatively similar climates
and potentially similar dam management strategies. In this sense,
interregional similarities of precipitation or flow regimes behave in
a very similar way at regional and global scales. This finding is relevant for inter-basin water resources management such as interbasin water transfer. With respect to precipitation regimes, higher
pairwise interregional homogenization or similarities of precipitation regimes before and after 1980 can be detected between SHR,
LR, HR, YR in the north China and SER, PR in the south China. When
it comes to flow regimes however, higher pairwise interregional
similarities can be observed between SHR in northeast China and
YTR, PR in south China. Besides, higher interregional similarities
can also be identified between SHR, LR, YTR, PR. Comparatively,
higher interregional similarities of precipitation regimes are not
identified. Furthermore, interregional similarities of precipitation
regimes are subject to no significant alterations. However, interregional similarities of flow regimes between river basins are
enhancing after 1980 when compared to those before 1980, implying significantly decreasing interregional dissimilarities of flow
regimes though interregional similarities of precipitation regimes
are not changed much. This finding clearly signifies increasingly
significant human influences on flow regimes in both space and
time.
4.3. Changes in attributes of flow and precipitation regimes
Fig. 7 depicts changes of flow and precipitation indices as listed
in Table 2. The precipitation indices of CI, MIM, F/W, SU/F, SP/SU
and W/SP increase during periods after 1980 when compared to
those before 1980. These results indicate increased monthly precipitation in dry seasons, such as winter, autumn, and possibly
decreased monthly precipitation in wet seasons, such as summer.
This kind of seasonal shift of precipitation regimes, i.e. dry seasons
are in wetting tendency and wet seasons drying tendency, has
been well corroborated by published researches (e.g. Zhang et al.,
2011a). Wherein, wetting tendency of winter is evident across
China and which can be reflected by positive W in 80% of the river
basins in China (Fig. 7). However, difference of seasonal precipitation amount is increasing. Changing magnitude of TMAX is smaller
when compared to that of TMIM, indicating not evident shifts of
the timing of the monthly average maximum precipitation but
apparent shifts of the timing of the monthly average minimum
precipitation regimes, and which also further convince the
seasonal shifts of precipitation extremes.
As for flow indices, CI is in significant changes when compared
with other 13 flow regimes. Significant increase of CI and MIM can
be observed after 1980, implying decreasing anomalies between
flow regimes during high and low flow seasons. Besides, CV of flow
regimes of 70% of the river basins is decreasing, also indicating
dampening processes of flow regimes. From the seasonal perspec-
471
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Precipitation and streamflow metrics
Percentage change in precipitation (%)
TMIM
TMAX
CV
CI
MIM
MAM
F/W
SU/F
SP/SU
W/SP
F
SU
SP
W
-9
1.2
-0.3
5.1
6.4
0.4
4.5
4.5
-2
12.3
-1.8
1
-1.7
5.7
9.2
0.1
-0.8
17.6
10.6
-1
-4.3
-4.3
16.2
-17.7
-2.4
-1
13.2
-9.6
-11.9
0.3
-3.4
31.8
18.3
-4
14.7
14.7
34.9
-41.1
1.9
-2.8
22.5
-19
24.9
1.4
-3
22.3
15.7
-3.1
31.6
31.6
-1.7
13.7
2.1
0.6
-1.6
6.5
13.2
3.1
2.3
78.6
61.1
4.3
3.1
3.1
0.6
59.2
0
2.1
-6.2
33.8
13.5
0.3
0.5
87.8
68.3
3.2
5.3
5.3
33.3
74.4
-0.1
-0.6
11.2
58.1
14
-3.3
-0.6
3.5
-2.4
-1.6
5.6
5.6
90.1
-3.5
-4.2
-4.1
19.5
6.6
11.6
-0.3
-7.3
77.7
42
-5.7
-6.8
-6.8
29
20
5.6
-4.8
9.4
30
SHR
LR
HR
HuR
30.9
0.3
-0.8
22.8
19.9
-0.2
18.9
18.9
-11.9
17.8
-3.4
6.9
-5.4
10.3
12
-1.3
-0.8
29.2
24.5
0.9
-7.8
-7.8
16.8
18.2
4.8
-4.9
10.8
21.4
18
-2.6
-1.8
4.3
5.1
-2.1
20.6
20.6
-1.7
5.9
0.2
3.2
-1.9
1.5
6.7
-2.6
-4.2
15.2
12.5
-3.5
-2.2
-2.2
13.1
20.8
-5.1
-4.2
4.6
24.9
-5.6
-0.7
-1.6
90.5
64.9
-1.7
25.4
25.4
3.6
40.2
-7.4
3
3.7
12.9
10.8
-3.5
-0.6
-5.2
-6.4
-2.4
33.3
33.3
5.5
29
-15.8
5.2
8.3
16
40
20
0
Percentage change in streamflow (%)
10.8
-0.4
-6.9
35.3
24.3
-5.2
9.8
9.8
-13.6
26.7
2.1
6.8
-7.7
14.8
7.8
-0.1
-3.9
9.9
5.6
-4.4
-5.3
-5.3
19.6
-9.9
3.7
-4.4
12.6
0.2
7.9
-0.9
-12.4
49.3
43.7
-9.3
-21.5
-21.5
1.1
9.7
11.1
-0.1
-3.1
6.7
12.4
-1.1
-7.7
23.4
14.3
-7.1
-1.5
-1.5
14.6
2.5
-0.6
-4.6
8.2
12.3
34.7
-3.3
3
-1.8
-1.6
1.5
13.5
13.5
-5.9
0.5
-5.8
4
-3.5
-1.8
14.7
-6.5
-5
25.6
16.2
-1.7
32.3
32.3
10.3
21.1
-14.1
6.3
5.2
21.5
YTR
SWR
SER
PR
NWR
YR
TMIM
TMAX
CV
CI
MIM
MAM
F/W
SU/F
SP/SU
W/SP
F
SU
SP
W
-20
-40
Hydrologic regions
Fig. 7. Temporal changes of 14 hydro-meteorological variables in terms of precipitation and streamflow after 1980 when compared to those before 1980.
(a) W
(b) SP
(c) SU
(d) F
(e) W-SP
(f) SP-SU
(g) SU-F
(h) F-W
(i) MAM
(j) MIM
(k) CI
(l) CV
(m) TMAX
(n) TMIM
-1
-0.8 -0.6
-0.4 -0.2
0
0
0.2
0.6
1
0.4
0.8
>1
Fig. 8. Spatial patterns of precipitation indices (defined in Table 1) after 1980 when compared to those before 1980.
tive, winter flow is increasing in most river basins of China. Six out
of ten river basins are dominated by decreased summer flow, but
increased spring and winter flow after 1980 when compared to
that before 1980. Seasonal difference of flow regimes is enhanced
and it can be observed from changes in SP/SU and W/SP. Timing
of occurrence of dry months is becoming to be late and that of
the wet months earlier. All these findings indicate similarities of
flow regimes seasonally and seasonal shifts of flow regimes in most
river basins of China. Comparisons between changes of flow and
precipitation regimes in terms of precipitation and flow indices
indicate that seasonal shifts and monthly variations of flow
regimes are partly influenced by seasonal behaviors of precipitation regimes. However, influences of enhancing hydrological regulation activities of the reservoirs on flow changes cannot be
ignored. It is a tough job to differentiate influences of human activities and climate changes on flow regimes during specific months
or seasons due to mixed impacts of these two major driving
factors. However, it can be well confirmed that hydrological regulation activities of booming building of reservoirs have increasingly
significant impacts on changes of flow regimes in both space and
time, and it is particularly true for changes of flow regimes after
1980.
472
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
(a) W
(b) SP
(c) SU
(d) F
(e) W-SP
(f) SP-SU
(g) SU-F
(h) F-W
(i) MAM
(j) MIM
(k) CI
(l) CV
(m) TMAX
(n) TMIM
-1
-0.8 -0.6
-0.4 -0.2
0 No Data
0
0.2
0.6
1
0.4
0.8
>1
Fig. 9. Spatial patterns of flow indices (defined in Table 1) after 1980 when compared to those before 1980.
Fig. 8 demonstrates spatial patterns of changes in precipitation
indices defined in Table 2 during period of after 1980 when compared to those before 1980. It can be seen from Fig. 8a–d that
increasing winter precipitation is dominant across China. Besides,
increase of precipitation in spring is also prevalent. Most parts of
China are dominated by decrease of autumn precipitation. Seasonal anomalies of precipitation are large between winter and
spring and also between autumn and winter due to significant
increase of winter precipitation amount. Changes of MAM are
moderate when compared to MIM (Fig. 8i and j), showing that
the fractional contribution of monthly precipitation in dry seasons
to annual precipitation amount is apparently increased and which
can also be mirrored by evident increase of CI (Fig. 8k). Fig. 8l indicates moderate decrease of CV, indicating increasing similarities of
precipitation regimes over China. As for the timing of the maximum/minimum monthly average precipitation (Fig. 8m and n),
the timing of the minimum monthly average precipitation events
is delayed in most parts of China and it is particularly the case in
the central and southeast China. Changes in the timing of the maximum monthly average precipitation are moderate.
As for changes of flow indices in space (Fig. 9), increase of winter flow is evident in most river basins in China such as the Yellow
River basin, the Yangtze River basin and river basins in northeast
China. However, decrease of winter flow is coming to be evident
from spring to autumn (Fig. 9b–d). Increased winter precipitation
should contribute much to the increase of winter flow in China.
Increased difference of flow regimes between winter and spring
is significant and which should be attributed to evident increase
of winter flow but not moderate changes of spring flow
(Fig. 9e, a and b). Difference between summer and autumn streamflow is larger in the Yellow River basin and north parts of the
Yangtze River basin due to increasing summer flow but decrease
autumn flow. However, the pairwise comparison of flow difference
between other seasons shows smaller difference of flow due to
moderate variations of flow regimes. Besides, MAM and TMAX
are decreasing with moderate changing magnitude; however,
MIM and TMIM are in significant decrease in most parts of China
(Fig. 9i, j, m and n). The above-mentioned changes of MAM, MIM,
TMAX and TMIM are partly attributed to seasonal shifts of precipitation regimes and should also partly due to impoundment
functions and hydrological regulation activities of reservoirs.
Hydrological regulations of reservoirs mainly cause dampening of
high flow regimes but inflating of low flow regimes. Fig. 9k and l
indicate increased fractional contribution of low flow events to
annual streamflow. Besides, decreased CV in most river basins of
China also indicates significant dampening of flow regimes and it
is particularly the case for flow regimes of the river basins in central and southeast China. It should be noted here that comparison
between Figs. 8 and 9 clearly indicate mismatch of the precipitation and flow indices in spatial patterns, implying increasingly
remarkable human roles in the changes of flow regimes in both
space and time when compared to influences of precipitation
changes on flow regimes. As for some specific regions such as the
northwest China, the flow processes of the plain regions are the
results of the melting or melted snow and/or icepack in the headwater regions of the river basins.
Catchments and associated processes are governed both by
external factors (e.g., climate inputs) and also by their own internal
factors (e.g., landscape properties) that occur over a wide range of
spatial and temporal scales (e.g. Sivakumar et al., 2015). In this
case, land-use change is taken into account as an important factor
for streamflow variations. In this study, land use changes are analyzed based on the land use data during end-1980 and 2000 with
focus on changes of arable land, forest land, grassland and artificial
land changes. To enhance the visual effects of changes of land use
types, the spatial resolution of the land use types of 1 km 1 km is
resampled as 20 km 20 km (Fig. 10) (e.g. Liu et al., 2014). Then,
changes of land use types such as cultivated lands, forests, grass
lands and artificial surfaces in 2000 relative to end-1980s are analyzed for each pixel with spatial resolutions of 20 km 20 km. The
land use types of each pixel are marked with those dominated by
the largest changing rate. In other words, if the land use type of
a pixel with the largest changing rate is grass land, then the pixel
is marked with grassland, implying that the grass lands are subject
to the largest changing magnitude in this pixel. In this case, taking
cultivated lands as a case study (the upper right panel of Fig. 10),
changes of cultivated lands are the most evident in the northwest
China, the Inner Mongolia and parts of the Northeast China. Closer
look at Fig. 10 indicates increase of 2.83 106 h m2 in arable land
in the northwest China and also in the northeast China (Liu et al.,
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
473
Fig. 10. Temporal variations of land use types from 1980s to 2000 (unit: km2).
2014). Expansion of arable land directly caused decrease of
10.89 105 h m2 and 3.44 106 h m2 in forest land and grassland
(Liu et al., 2014). Specifically, evident decrease of forest land can be
found in northeast China, and grassland in northeast China, north
China and northwest China. Increase of arable land is in close relation with increasing water demand for irrigation and hence
decrease of fluvial streamflow amount. Furthermore, faster expansion can be detected for artificial land. The period of 1980–2000
witnessed increase of 1.76 106 h m2 in artificial land such as
urbanized land and these increase of artificial land is found mainly
in the Huang–Huai–Hai Plain, the Yangtze Delta, the Pearl River
Delta and also Sichuan Plain. Shifts from diverse land use types
to mainly artificial land types lead directly to dull land use types.
Around 1980 witnessed not evident variations of precipitation
regimes in space (One-sided t test, p > 0.5, Table 3). However, the
heterogeneity of streamflow processes in space after 1980 substantially decreases (One-sided t test, p > 0.05, Table 4). Therefore, the
homogenization of topographical and underlying features should
be one of the principle factors behind spatial homogenization of
streamflow changes.
In recent years, human-induced global warming has altered and
will continue altering hydrological cycle at regional and global
scales (Milly et al., 2005). Meanwhile, human activities particularly
the building of dams or reservoirs for water supply, hydropower
generation, agricultural irrigation and so forth greatly altered
hydrological cycle at regional scale. Influences from human activities on flow regimes are even more remarkable than those from
warming climate in Asian regions and parts of USA (e.g.
Haddeland et al., 2014). Damming and construction of reservoirs
have been continuing, and up to 2011, 756 large-scale reservoirs
with total capacity of 749.99 billion m3, 3938 moderate-scale reservoirs with total capacity of 111.98 billion m3, and 93,308 smallscale reservoirs with total capacity of 70.35 billion m3 have been
built (Sun et al., 2013). Furthermore, the South-to-North Water
Transfer Project further alters water resources in time and space.
All these human activities further enhance human influences on
spatiotemporal patterns of flow regimes and water resources. Thus,
relations between precipitation and streamflow are coming to be
complicated in both space and time with the intensification of
human activities. This point should arouse considerable concerns
from academic communities in terms of hydrological modeling
and also inter-basin water resources management.
5. Conclusions
Water resources management and also inter-basin water transfer projects are theoretically and practically important in terms of
stability of human society and also sustainable development of
socio-economy. In this study, homogenization and similarities of
precipitation and flow regimes across China are thoroughly investigated based on daily precipitation data covering the period of
1961–2000 from 554 meteorological stations and monthly streamflow data covering the period of 1960–2000 from 370 hydrological
stations. Implications and possible causes behind spatiotemporal
patterns of flow and precipitation regimes are discussed. The
above-mentioned analysis helps to obtain some interesting and
relevant conclusions as follows:
(1) Homogenization of precipitation regimes reflected by the
Gini coefficients is increasing in the direction of from the
northwest to the southeast China. Different spatial patterns
of Gini coefficients of flow regimes are observed across
China when compared to those of precipitation regimes.
Lower homogenization of flow regimes is detected in the
northeast China. River basins in the central, the south and
the southeast China are dominated by higher homogenization of flow regimes. River fragmentations being characterized by lower homogenization of flow regions distribute
sporadically among the river parts with higher homogenization. Damming-induced fragmentation of river basins is the
major cause behind higher homogenization of flow regimes.
Besides, homogenization of flow regimes is greatly enhanced
after 1980 due to booming building of reservoirs.
(2) Precipitation regimes during 1961–2000 are characterized
by decreasing dissimilarities. Increasing dissimilarities are
observed mainly in the Yellow River and parts of the Yangtze
474
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
River basin. Besides, larger areas of China are dominated by
increasing dissimilarities of precipitation regimes temporally during 1961–1980 when compared to those during
1980–2000. This phenomenon should be attributed to
increasing precipitation concentration and intensifying precipitation regimes in recent years. Contrarily, flow regimes
during 1961–2000 are characterized mainly by increasing
similarities or homogenization in both space and time. Particularly, after 1980 large parts of the river basins in China
witness a dominant homogenization of flow regimes. In this
sense, it can be said that flow regimes are influenced by precipitation changes, however, human activities particularly
damming and construction of reservoirs significantly alter
flow regimes at regional scale.
(3) Distinctly dissimilar precipitation and flow regimes can be
identified between geographically separate river basins
and this finding is in good agreement with those based on
studies of flow regimes in USA. Therefore, interregional similarities of precipitation or flow regimes behave in a very
similar way at regional and global scales. Interregional similarities of flow regimes are enhancing after 1980 when compared to those before 1980 though interregional similarities
of precipitation regimes are not changed much. In this case,
human activities tend to have more influences on interregional similarities of flow regimes when compared to those
from precipitation changes.
(4) Analysis of 14 precipitation and flow indices indicates
partial match in spatial patterns of precipitation and flow
regimes. However, spatial mismatch is evident in terms of
spatial range and changing degree of flow and precipitation
regimes. Seasonal shifts of precipitation changes trigger
increasing precipitation in winter. Correspondingly,
streamflow in low flow season such as winter is also increasing. Besides, fractional contribution of precipitation and
streamflow in dry seasons to the annual precipitation and
streamflow increases. Variability of precipitation and flow
changes is dominantly decreasing across China, which seems
to indicate considerable influences of precipitation regimes
on streamflow changes in dry season. However, such
influences are varying from one river basin to another.
Specifically, increase of streamflow is of larger increasing
magnitude when compared to that of precipitation in the
Yangtze River basin. However, the changing directions of
precipitation and streamflow are adverse in the Yellow River
basin. And it is particularly the case when it comes to the
South-to-North water transfer project in China, which significantly alters spatiotemporal distribution of water resources
across China.
Acknowledgements
This work is financially supported by the National Science
Foundation for Distinguished Young Scholars of China (Grant No.:
51425903), the Xinjiang Science and Technology Planning Project
(Grant No.: 201331104), the Natural Science Foundation of Anhui
Province, China, and is fully supported by a grant from the
Research Grants Council of the Hong Kong Special Administrative
Region, China (Project No. CUHK441313). Detailed information
such as data can be obtained by writing to the corresponding
author at zhangq68@mail.sysu.edu.cn. The last but not the least,
our cordial gratitude should be extended to the editor, Prof. Dr.
Geoff Syme, the associate editor, Prof. Dr. Bellie Sivakumar, and
two anonymous reviewers for their professional and pertinent
comments and revision suggestions which are greatly helpful for
further improvement of the quality of this paper.
References
Ahn, K.-H., Merwade, V., 2014. Quantifying the relative impact of climate and
human activities on streamflow. J. Hydrol. 515, 257–266.
Bao, Z., Zhang, J.Y., Wang, G., Fu, G., He, R., Yan, X., Jin, J., Liu, Y., Zhang, A., 2012.
Attribution for decreasing streamflow of the Haihe River basin, northern China:
climate variability or human activities? J. Hydrol. 460–461, 117–129.
Chen, Y.Q., Yang, T., Xu, C.Y., Zhang, Q., Chen, X., Hao, Z.C., 2010. Hydrologic
alteration along the Middle and Upper East River (Dongjiang) basin, South
China: a visually enhanced mining on the results of RVA method. Stoch.
Environ. Res. Risk Assess. 24, 9–18.
Daufresne, M., Lengfellner, K., Sommer, U., 2009. Global warming benefits the small
in aquatic ecosystems. Proc. Natl. Acad. Sci. USA 106 (31), 12788–12793.
Dinpashoh, Y., Jhajharia, D., Fakheri-Fard, A., Singh, V.P., Kahya, E., 2011. Trends in
reference evapotranspiration over Iran. J. Hydrol. 399, 422–433.
Gosling, R., 2014. Assessing the impact of projected climate change on drought
vulnerability in Scotland. Hydrol. Res. 45 (6), 806–816.
Haddeland, I., Heinke, J., Biemans, H., Eisner, S., Flörke, M., Hanasaki, N., Konzmann,
M., Ludwig, F., Masaki, Y., Schewe, J., Stacke, T., Tessler, Z.D., Wada, Y., Wisser, D.,
2014. Global water resources affected by human interventions and climate
change. Proc. Natl. Acad. Sci. 111 (9), 6251–6256.
Hamed, K.H., Rao, A.R., 1998. A modified Mann–Kendall trend test for
autocorrelated data. J. Hydrol. 204 (1–4), 182–196.
Iacob, O., Rowan, J.S., Brown, I., Ellis, C., 2014. Evaluating wider benefits of natural
flood management strategies: an ecosystem-based adaptation perspective.
Hydrol. Res. 45 (6), 774–787.
Jiang, C., Xiong, L.H., Xu, C.-Y., Guo, S.L., 2014. Bivariate frequency analysis of
nonstationary low-flow series based on the time-varying copula. Hydrol.
Process. http://dx.doi.org/10.1002/hyp.10288.
Li, R., Chen, Q., Tonina, D., Cai, D., 2014. Effects of upstream reservoir regulation on
the hydrological regime and fish habitats of the Lijiang River, China. Ecol. Eng.
http://dx.doi.org/10.1016/j.ecoleng.2014.04.021.
Liu, C.M., Zheng, H.X., 2004. Changes in components of the hydrological cycle in the
Yellow River basin during the second half of the 20th century. Hydrol. Process
18, 2337–2345.
Liu, J.Y., Kuang, W., Zhang, Z., Xu, X., Qin, Y., Ning, J., Zhou, W., Zhang, S., Li, R., Yan, C.,
Wu, S., Shi, X., Jiang, N., Yu, D., Pan, X., Chi, W., 2014. Spatiotemporal
characteristics, patterns and causes of land use changes in China since the late
1980s. Acta Geograph. Sin. 64 (12), 1412–1420 (in Chinese with English abstract).
Liu, Z.F., Xu, Z.X., Fu, G.B., Yao, Z.J., 2013. Assessing the hydrological impacts of
climate change in the headwater catchment of the Tarim River basin, China.
Hydrol. Res. 44 (5), 834–849.
Ma, H., Yang, D.W., Tan, S.K., Gao, B., Hu, Q., 2010. Impact of climate variability and
human activity on streamflow decrease in the Miyun Reservoir catchment. J.
Hydrol. 389, 317–324.
Matteau, M., Assani, A.A., Mesfioui, M., 2009. Application of multivariate statistical
analysis methods to the dam hydrologic impact studies. J. Hydrol. 371, 120–
128.
McIntyre, N., Ballard, B., Bruen, M., Bulygina, N., Buytaert, W., Cluckie, I., Dunn, S.,
Ehret, U., Ewen, J., Gelfan, A., Hess, T., Hughes, D., Jackson, B., Kjeldsen, T., Merz,
R., Park, J.-S., O’Connell, E., O’Donnell, G., Oudin, L., Todini, E., Wagener, T.,
Wheater, H., 2013. Modelling the hydrological impacts of rural land use change.
Hydrol. Res. 45 (6), 737–754.
Milly, P.C.D., Dunne, K.A., Vecchia, A.V., 2005. Global pattern of trends in streamflow
and water availability in a changing climate. Nature 438, 347–350.
Petts, G.E., 2009. Instream flow science for sustainable river management. J. Am.
Water Resour. Asssoc. 45, 1071–1076.
Poff, N.L., Olden, J.D., Merritt, D.M., Pepin, D.M., 2007. Homogenization of regional
river dynamics by dams and global biodiversity implications. Proc. Natl. Acad.
Sci. 104, 5732–5737.
Poff, N.L., Zimmerman, J.K.H., 2010. Ecological responses to altered flow regimes: a
literature review to inform the science and management of environmental
flows. Freshw. Biol. 55, 194–205.
Richter, B.D., Baumgartner, J.V., Powell, J., Braun, D.P., 1996. A method for assessing
hydrologic alteration within ecosystem. Conserv. Biol. 10, 1163–1174.
Richter, B.D., Baumgartner, J.V., Wigington, R., Braun, D., 1997. How much water
does a river need? Freshw. Biol. 37, 231–249.
Sivakumar, Bellie, Singh, Vijay P., Berndtsson, Ronny, Khan, Shakera K., 2015.
Catchment classification framework in hydrology: challenges and directions. J.
Hydrol. Eng. http://dx.doi.org/10.1061/(ASCE).
Sun, Z.G., Zhang, L., Duan, Z., 2013. Volume and spatial distribution of water
reservoirs in China. China Water Resour. 7, 10–11 (in Chinese).
Tang, L., Yang, D., Hu, H., Gao, B., 2011. Detecting the effect of land-use change on
streamflow, sediment and nutrient losses by distributed hydrological
simulation. J. Hydrol. 409, 172–182.
Tran, L.T., O’Neill, R.V., 2013. Detecting the effects of land use/land cover on mean
annual streamflow in the Upper Mississippi River Basin, USA. J. Hydrol. 499,
82–90.
Warton, D., Wright, S.T., Wang, Y., 2012. Distance-based multivariate analyses
confound location and dispersion effects. Methods Ecol. Evol. 3 (1), 89–101.
Xia, J., Qiu, B., Li, Y., 2012. Water resources vulnerability and adaptive management
in the Huang, Huai and Hai River basins of China. Water Int. 37 (5), 509–511.
Xiao, M., Zhang, Q., Singh, V.P., 2014. Influences of ENSO, NAO, IOD and PDO on
seasonal precipitation regimes in the Yangtze River basin, China. Int. J. Climatol.
http://dx.doi.org/10.1002/joc.4228.
Q. Zhang et al. / Journal of Hydrology 530 (2015) 462–475
Xu, J.X., 2005. Variation in grain size of suspended load in upper Changjiang River
and its tributaries by human activities. J. Sediment Res. 3, 8–16 (in Chinese).
Xu, X.Y., Yang, H.B., Yang, D.W., Ma, H., 2013. Assessing the impacts of climate
variability and human activities on annual runoff in the Luan River basin, China.
Hydrol. Res. 44 (5), 940–952.
Yin, X.A., Yang, Z.F., 2011. Development of a coupled reservoir operation and water
diversion model: balancing human and environmental flow requirements. Ecol.
Model. 222, 224–231.
Zhan, C.S., Niu, C.W., Song, X.F., Xu, C.-Y., 2013. The impacts of climate variability
and human activities on streamflow in Bai River basin, northern China. Hydrol.
Res. 44 (5), 875–885.
Zhou, Y., Zhang, Q., Singh, V.P., 2014. Fractal-based evaluation of the effect of water
reservoirs on hydrological processes: the dams in the Yangtze River as a case
study. Stoch. Environ. Res. Risk Assess. 28, 263–279.
Zhang, Q., Liu, C.L., Xu, C.-Y., Xu, Y.P., Jiang, T., 2006. Observed trends of annual
maximum water level and streamflow during past 130 years in the Yangtze
River basin, China. J. Hydrol. 324, 255–265.
Zhang, Q., Chen, G.Y., Su, B.D., Disse, M., Jiang, T., Xu, C.-Y., 2008a. Periodicity of
sediment load and runoff in the Yangtze River basin and possible impacts of
climatic changes and human activities. Hydrol. Sci. J. 53 (2), 457–465.
Zhang, Q., Xu, C.-Y., Zhang, Z., Chen, Y.D., Liu, C.-L., 2008b. Spatial and temporal
variability of precipitation maxima during 1960–2005 in the Yangtze River
475
basin and possible association with large-scale circulation. J. Hydrol. 353, 215–
227.
Zhang, Q., Xu, C.-Y., Chen, Y.D., Jiang, J., 2009a. Abrupt behaviors of the streamflow
of the Pearl River basin and implications for hydrological alterations across the
Pearl River Delta, China. J. Hydrol. 377, 274–283.
Zhang, Q., Xu, C.-Y., Singh, V.P., Yang, T., 2009b. Multiscale variability of sediment
load and streamflow of the Lower Yangtze River basin: possible causes and
implications. J. Hydrol. 368, 96–104.
Zhang, Q., Singh, V.P., Li, J., Chen, X., 2011a. Analysis of the periods of maximum
consecutive wet days in China. J. Geophys. Res. 116, D23106.
Zhang, Q., Singh, V.P., Sun, P., Xi, Chen, Zhang, Z., Li, J., 2011b. Precipitation and
streamflow changes in China: changing patterns, causes and implications. J.
Hydrol. 410, 204–216.
Zhang, Q., Sun, P., Singh, V.P., Chen, X., 2012. Spatial–temporal precipitation changes
(1956–2000) and their implications for agriculture in China. Glob. Planet.
Change 82–83, 86–95.
Zhang, Q., Singh, V.P., Xu, C.-Y., Chen, X., 2013. Abrupt behaviors of streamflow and
sediment load variations of the Yangtze River basin, China. Hydrol. Process. 27
(3), 444–452.
Zhang, Q., Xiao, M., Liu, C.-L., Singh, V.P., 2014. Reservoir-induced hydrological
alterations and ecological in stream flow in the East River, the Pearl River basin,
China. Stoch. Environ. Res. Risk Assess. 28, 2119–2131.