ADVANCES IN ATMOSPHERIC SCIENCES, VOL. 23, NO. 4, 2006, 513–520
Decreasing Reference Evapotranspiration in a Warming
Climate—A Case of Changjiang (Yangtze)
River Catchment During 1970–2000
N˜), Lebing GONG (÷WX), JIANG Tong (ñ Õ), and Deliang CHEN (
Chong-yu XU∗1,2 (
1
3,4
4,5
)
Department of Geosciences, University of Oslo, P. O Box 1047 Blindern, NO-0316 Oslo, Norway
2
Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008
3
Department of Earth Sciences, Uppsala University, Villavagen 16, S-75236 Uppsala, Sweden
4
5
2
Department of Earth Sciences, Gothenburg University, Sweden
Laboratory for Climate Studies/National Climate Center, China Meteorological Administration, Beijing 100081
(Received 2 September 2005; revised 14 January 2006)
ABSTRACT
This study deals with temporal trends in the Penman-Monteith reference evapotranspiration estimated
from standard meteorological observations, observed pan evaporation, and four related meteorological variables during 1970–2000 in the Yangtze River catchment. Relative contributions of the four meteorological
variables to changes in the reference evapotranspiration are quantified. The results show that both the
reference evapotranspiration and the pan evaporation have significant decreasing trends in the upper, the
middle as well as in the whole Changjiang (Yangtze) River catchment at the 5% significance level, while
the air temperature shows a significant increasing trend. The decreasing trend detected in the reference
evapotranspiration can be attributed to the significant decreasing trends in the net radiation and the wind
speed.
Key words: reference evapotranspiration, Penman-Monteith method, temporal trend, Changjiang
(Yangtze) River catchment, China
doi: 10.1007/s00376-006-0513-4
1. Introduction
Reference evapotranspiration, Eref , i.e. the potential evapotranspiration of grass (e.g. Allen et al.,
1998), is one of the most important things to consider when scheduling run times for an irrigation system, preparing input data for hydrological models of
a water balance study, or calculating actual evapotranspiration for a region and/or catchment (Blaney
and Criddle, 1950; Dyck, 1983; Hobbins et al., 2001;
Xu and Li, 2003; Xu and Singh, 2005). All else being equal, one expects evapotranspiration to increase
with increases in air temperature, which has been
shown by Ramirez and Finnerty (1996) and Wang
et al. (2003) and many others. However, several
studies have reported decreasing trends in pan evaporation (Epan ) in several countries (e.g. Chattopadhyay and Hulme, 1997; Kaiser, 2000, 2001; Thomas,
2000; Hobbins et al., 2004; Liu et al., 2004; Chen
*E-mail: chongyu.xu@geo.uio.no
et al., 2005). This is at odds with the expectation
that global warming caused by increasing emissions of
greenhouse gases would increase potential evapotranspiration. The study of Chattopadhyay and Hulme
(1997) showed that in spite of the general increase
in temperature in recent decades over the Indian region, both Epan and Eref have decreased and that increases in relative humidity and decreases in radiation were both important correlates with the decreasing trend in Eref . Kaiser (2000, 2001) analyzed the
cloud amount, water vapour pressure and relative humidity over China and concluded that the marked decreasing trend in the all-China mean cloud amount has
continued; annual mean station pressure has increased
dramatically over most of China, and annual mean water vapor pressure has increased significantly in northwest and east-central China (no areas show any significant decreases); relative humidity has decreased
significantly in the northeast (where the largest de-
514
DECREASING REFERENCE EVAPOTRANSPIRATION IN CHANGJIANG CATCHMENT
creases in cloud amount are observed) but it has increased significantly in the northwest (consistent with
water vapor increases). The most likely explanation
for such a paradox is the air pollution over China.
Thomas (2000) analyzed the time series (1954–1993)
of Penman-Monteith evapotranspiration estimates for
65 stations in China, for the country as a whole and
for individual stations. His analysis showed that for
China as a whole, the potential evapotranspiration has
decreased in all seasons especially in northwest and
southeast China. South of 35◦ N, sunshine appears to
be most strongly associated with evapotranspiration
changes while wind, relative humidity and maximum
temperature are the primary factors in northwest, central and northeast China, respectively.
Previous studies have provided solid evidence that
several key meteorological variables have caused the
rich spatial temporal and spatial variations of Epan
and Eref and the role played by each variable varied
with region and season. Previous investigations have
also shown that additional analysis is needed to quantify the causes for the change in evaporation, to understand the mechanisms behind the widely observed
decline in reference evapotranspiration and pan evaporation and its implications for understanding the influence of climate change on the global hydrological cycle.
As a part of the ongoing research with the main objective of studying the impact of climate change on floods
in the Changjiang (Yangtze) River basin in China, detailed investigation of the variability of evaporation in
the Changjiang (Yangtze) River basin and quantification of the influence of key factors on evaporation
in the region are needed. Recognizing the above concerns, the main aims of this study are formulated as (1)
examining the temporal trend of the pan evaporation
and the reference evapotranspiration calculated by the
Penman-Monteith method (Allen et al., 1998) for different geographic regions and for the whole catchment
of the Changjiang (Yangtze) River, and (2) analyzing the possible contributing factors for the trend in
evapotranspiration and quantifying the contributions
of individual meteorological variables to that trend.
VOL. 23
the Changjiang catchment is divided into three geographic regions corresponding to the three steps: the
upper, the middle and the lower regions. The classification of the upper, middle and lower regions of the
catchment in this study is different from what is determined by the Changjiang River Water Resources
Commission (CWRC) in China, where flood control is
the main concern of the classification. According to
CWRC, the section above the Yichang station (where
the Three Gorges Dam is located) is called the Upper Reach, 4500 km long, with a controlled catchment area of 1 million km2 accounting for 70.4% of
the Yangtze’s total area. From Yichang to Hukou is
the Middle Reach, 955 km long with a catchment area
of 680 000 km2 . The remaining part from Hukou to
the estuary is called the Lower Reach, 938 km long
with an interval catchment area of 120 000 km2 . The
river basin is located in the subtropical and temperate
climate zone. Monsoon winds dominate the climate
of the region, giving rise to humid summers and dry
winters. The advance and retreat of the summer monsoon determines the timing of the rainy season and the
amount of rainfall throughout the basin. The dataset
of 115 National Meteorological Observatory (NMO)
stations with daily observations of pan evaporation,
maximum, minimum and mean air temperature, wind
speed, relative humidity, sunshine hours, and absolute
vapor pressure for the period of 1970 to 2000 was used
in this study. Data have been provided by the National Climate Center (NCC) of the China Meteorological Administration (CMA). The locations of these
stations are also shown in Fig. 1.
3. Method and analysis procedure
The potential evapotranspiration is a function of
atmospheric forcing and surface type. In order to get
rid of the influence of surface type and to study the
2. Study region and data
The Changjiang (Yangtze) River is about 6380 km
long and has a drainage area of 1.8×106 km2 (Fig. 1).
Originating from the Tibetan Plateau, the terrain of
the basin is shaped like a staircase with 3 steps. Located on the Qingzang Plateau in the west, the highest step has an average elevation of over 3000 meters
above sea-level; the second step—the middle part of
the basin—has an average elevation of 1000 meters;
then, water flows to the east China Plain, which is
the third step with an average elevation of about 100
meters. In the following discussion and comparison,
Fig. 1. Location of the Changjiang (Yangtze) River catchment and the meteorological stations (open circles).
NO. 4
XU ET AL.
515
evaporative demand of the atmosphere independent of
crop type, crop development and management practices, the concept of the reference evapotranspiration,
Eref , was introduced by defining the reference crop as
a hypothetical surface of green grass of uniform height,
actively growing and adequately watered (Allen et al.,
1998). The Food and Agriculture Organization (FAO)
Penman-Monteith (P-M) method of reference evapotranspiration (equation 1) is used in this study.
900
0.408∆(Rn − G) + γ
u2 (es − ea )
Ta + 273
,
Eref =
∆ + γ(1 + 0.34u2 )
(1)
(3) The temporal trends of the reference evapotranspiration and pan evaporation for each region and the
whole catchment were tested for significance by using
the parametric t-test. (4) The cause of the decreasing trend in evapotranspiration was analyzed and the
contributions of individual meteorological variables to
the decreasing trend of evapotranspiration were quantified.
where Eref =reference evapotranspiration mm d−1 ,
Rn =net radiation at the crop surface (MJ m−2 d−1 ),
G=soil heat flux density (MJ m−2 d−1 ), T =mean daily
air temperature at 2 m height (◦ C), u2 =wind speed
at 2 m height (m s−1 ), es =saturation vapour pressure (kPa), ea =actual vapour pressure (kPa), es −
ea =saturation vapour pressure deficit (kPa), ∆=slope
vapour pressure curve (kPa ◦ C−1 ), γ=psychrometric
constant (kPa ◦ C−1 ). The computation of all data
required for the calculation of the reference evapotranspiration in this study follows the method and procedure given in Chapter 3 of Allen et al. (1998).
The procedure of analysis is as follows: (1) Daily
values of the reference evapotranspiration from 1970 to
2000 for the 150 stations were calculated by the P–M
method. (2) The mean annual reference evapotranspiration and pan evaporation for each station were
computed by summing up the daily values; and the
annual values for each region and the whole catchment
were calculated as the arithmetic mean of the regions.
The time series of annual Eref and Epan from 1970
and 2000 averaged over the upper, the middle and the
lower regions as well as the whole catchment are plotted in Fig. 2. The parametric t-test is then used to
test the significance of the temporal trend for each
of these four time series. The normality of the data,
which is required by the t-test, is pre-tested by using
the Kolmogorov-Smirnov method and the result (not
shown) reveals that all the variables shown in Fig. 2
are normally distributed at the 95% confidence level.
The p-values for the detected trends are given in Table
1 (column 2). It is seen that (1) both the pan evaporation and reference evapotranspiration in all the regions
have decreased during the last 3 decades. (2) The decreasing trend is the strongest in the upper region for
both Eref and Epan and becomes weaker in the middle
and lower regions. (3) For the reference evapotranspiration, the decreasing trend is significant in the upper,
the middle and the whole catchment at the 10% significance level. For the pan evaporation, the decreasing
1050
1800
950
L
900
U
W
850
M
1980
1990
Year
2000
Epan (mm yr −1)
)
−1
Eref (mm yr
4.1 Temporal trends in mean annual evapotranspiration, pan evaporation and meteorological variables
(a)
1000
800
1970
4. Results and discussion
(b)
1600
U
1400
L
W
1200
1000
1970
M
1980
1990
Year
2000
Fig. 2. Annual reference evapotranspiration, pan evaporation and their linear trends in the Changjiang
catchment and its sub-regions. Letters U, M, L, and W stand for upper, middle, and lower regions and
the whole catchment respectively.
516
DECREASING REFERENCE EVAPOTRANSPIRATION IN CHANGJIANG CATCHMENT
VOL. 23
Table 1. Significance test p-values of the trends for reference evapotranspiration and relevant meteorological variables
in the three regions and the whole catchment of the Changjiang (Yangtze) River. (The italic numbers mean that the
linear trend is not significant at the 10% level).
Epan
Eref
TA
WD
RH
RN
Upper region
0.03
<0.01
0.02
<0.01
<0.01
<0.01
Middle region
0.71
0.03
0.04
<0.01
0.20
<0.01
Lower region
0.17
0.48
<0.01
<0.01
0.15
<0.01
Whole catchment
0.14
0.05
<0.01
<0.01
0.80
<0.01
14
W
12
U
10
8
1970
(a)
1980
1990
Year
U
1.6 W
1.4
1.2
1
M
0.8 (b)
1980
1990
Year
2000
9
−1
d )
70
U
1970
L
1970
−2
Relative humidity (%)
L
M
W
65
1.8
2000
80
75
−1
16
Wind speed (m s
L
M
(c)
1980
1990
Year
2000
Net radiation (MJ m
Air temperature ( oC)
18
)
Note: Epan =pan evaporation, Eref =reference evapotranspiration, TA=air temperature, WD=wind speed, RH=relative
humidity, RN=net radiation.
8.5
U
L
W
8
M
(d)
7.5
1970
1980
1990
Year
2000
Fig. 3. Annual temperature, wind speed, relative humidity and net radiation and their linear trends in
the Changjiang catchment and its sub-regions. Labels are as in Fig. 2.
trend is significant at the 10% significance level in the
upper region and significant at the 20% significance
level in the lower region and the whole catchment
(when the p-values in the table are smaller than 0.20).
(4) The significance of the decreasing trend in Eref and
Epan differs in different regions. This is logical since it
is well-known that the pan evaporation coefficient, i.e.
the ratio of Etref to Etpan , is not a constant in either
region or season. This result is consistent with that of
other studies (e.g. Liu et al., 2004; Chen et al., 2005).
In order to analyze the main causes of the decreasing trend, the same trend analysis procedure is performed on the major meteorological variables in each
region that determine the evaporation values. The meteorological parameters that are examined include air
temperature, relative humidity, wind speed and net
radiation. The results are shown in Fig. 3 and Table 1
(columns 4–7), which reveal that the air temperature
NO. 4
XU ET AL.
517
980
(1)
940
ref
E (mm yr
-1
)
960
(2)
920
(3)
900
(4)
(5)
880
860
1970
1975
1980
1985
year
1990
1995
2000
Fig. 4. Comparison of the original reference evapotranspiration [Line (5)] with recalculated reference evapotranspirations. Line (1): recalculated Eref with detrended data series for all variables
except air temperature; Line (2): recalculated Eref with detrended data series for all variables;
Line (3): recalculated Eref with detrended data series for all variables except wind speed; Line (4):
recalculated Eref with detrended data series for all variables except net radiation.
shows a significant increasing trend in all the regions,
while wind speed and net radiation show a significant
decreasing trend in all regions at the 10% significance
level. No significant temporal trend was detected for
relative humidity except in the upper region where an
increasing trend is detected. The increasing trend in
air temperature over the past 3 decades is consistent
with the global warming as reported across most of
the globe while the decreasing trend in wind speed requires more attention, although both can be due to
a real climate change and/or a change in the environment of the measuring stations. Further study is
needed to determine the extent to which the decrease
in wind speed is due to natural climate changes and
changes in the environment of the measuring stations.
As for the decreasing trend found for net radiation, the decrease in the global radiation (sunshine
duration) is the most likely cause. Kaiser (2000, 2001)
found that the most likely explanation for the decrease
of sunshine duration is the increasing air pollution in
China. The findings of Kaiser (2000, 2001) were cited
and used by Liu et al. (2004) in studying the temporal trend and spatial variation of pan evaporation
in China. By examining solar radiation measurements
in eastern China (Shanghai, Nanjing and Hangzhou),
Zhang et al. (2004) also came to the conclusion that
the global radiation has decreased in recent decades
and the acceleration of air pollution and decrease of
relative sunshine are the possible causes for the decrease of global and direct radiation.
Another study by Liu et al. (2004) also found that
unlike in other parts of the world, the decrease in solar
irradiance in China was not always accompanied by an
increase in cloud cover and precipitation. Therefore,
they speculated that the increased air pollution in the
area may play an important role in the decrease of
solar irradiance in China.
4.2 Contribution of meteorological variables to
the trend of reference evapotranspiration
In order to quantify the contributions of the trends
of meteorological variables to the decreasing trend
of evapotranspiration, the following studies are performed. (1) Removing the increasing/decreasing trend
in the meteorological variables to make them stationary time series. (2) Recalculating the reference evapotranspiration using the detrended meteorological variables. (3) Recalculating the reference evapotranspiration by using, in each time, the original data (with
trend) for one variable and the detrended data for
other variables. (4) Comparing the result with the
original evapotranspiration and considering the difference as the influence of the trend by that variable.
The above analysis is done for each region and the
whole catchment. For illustrative purposes, the original reference evapotranspiration [Line (5)] and the recalculated reference evapotranspiration time series for
the whole catchment are shown in Fig. 4. Line (2)
shows the recalculated evapotranspiration when detrended values of all four meteorological values are
used to calculate the reference evapotranspiration. It
is plotted for comparison. Lines (1), (3) and (4) show
the recalculated Eref with the detrended data series of
all variables except for air temperature, wind speed,
or net radiation, respectively. The difference between
Line (2) and lines (1), (3) and (4) reveals the contribution of each variable to the recalculated evapotranspiration. For example, the difference between Lines
518
DECREASING REFERENCE EVAPOTRANSPIRATION IN CHANGJIANG CATCHMENT
(1) and (2) shows that if the detrended data series are
used for all variables except for air temperature, the
recalculated evapotranspiration will have an increasing trend; in other words, the difference between these
two lines represents the contribution of temperature
to the trend in evapotranspiration. Same explanation
applies to the other lines. The difference between Line
(2) and Line (5) reflects the combined effect of all variables. The line with the influence of relative humidity
is not shown because the trend of relative humidity is
insignificant in the whole catchment. Similar results
are also obtained for the three sub-regions (not shown)
except that in the upper region the effect of relative
humidity is notable.
To better understand the result presented in Fig.
4, a sensitivity analysis of reference evapotranspiration to the meteorological variables is performed. The
procedure of the study is as follows: (1) Determination of scenarios for changes in meteorological variables by adjusting the historic data series by adding
delta changes (∆X = 0, ±10%, ±20%, ±30% of X(t),
where X is the meteorological variable, and t is the
date), (2) recalculation of reference evapotranspiration
for each station by using one scenario at each time, (3)
calculation of the relative changes between the recalculated and the original reference evapotranspiration,
and (4) plotting of the relative changes in reference
evapotranspiration against the relative changes in each
meteorological variable.
Figure 5 shows the result of the sensitivity study. It
is seen that the sensitivity of reference evapotranspira-
tion to the meteorological variables varies in different
regions and the two most sensitive variables are relative humidity and net radiation, followed by air temperature and wind speed. The combination of Fig. 5
and Fig. 3 explains the result shown in Fig. 4. For example, on one hand, although relative humidity is one
of the most sensitive variables in almost all regions, it
has no significant contribution to the decreasing trend
in reference evapotranspiration except in the upper region, because the relative humidity in other regions
and over the whole catchment has no significant decreasing or increasing trend. The temperature increase
compensates to some extent the decreasing trend in
evapotranspiration, but the contributions from net radiation and wind speed are much larger, which results
in the significant decreasing trend in evapotranspiration in the Changjiang catchment.
5. Summary and conclusions
The temporal trends in the time series of the reference evapotranspiration, observed pan evaporation,
and meteorological variables during 1970–2000 in the
Yangtze River catchment and its sub-areas were analyzed. The possible causes and quantitative contributions of the major meteorological variables to the
evaporation trends were investigated. The following
conclusions may be drawn from the study. (1) For the
upper and middle regions as well as the whole catchment, there exist significant decreasing trends in reference evapotranspiration and pan evaporation from
20
RN
15
R elative variation in E ref (%)
VOL. 23
10
TA
5
WD
0
Upper
Middle
Lower
0
5
10
15
R elative variation in meteorological variables (% )
RH
20
Fig. 5. The sensitivity of the reference evapotranspiration to
the four major meteorological variables in the Changjiang catchment and its sub-regions. TA=air temperature, WD=wind speed,
RH=relative humidity, RN=net radiation.
NO. 4
XU ET AL.
1970 to 2000. (2) The sensitivity analysis shows that
the most sensitive variables for the reference evapotranspiration are relative humidity and net radiation
followed by air temperature and wind speed. (3) The
decreasing trend detected in reference evapotranspiration is a combined effect of all four meteorological
decreasing trend detected in reference evapotranspiration is a combined effect of all four meteorological
variables. The actual contribution of each variable
to the decreasing trend in reference evapotranspiration depends on the sensitivity of evapotranspiration
to the variable and the magnitude of the trend of the
variable. In the Changjiang (Yangtze) catchment, the
decreasing trends in reference evapotranspiration are
mainly caused by the significant decreases in net radiation and wind speed. Temperature increases result in
an increase in evapotranspiration in all 3 regions, but
this cannot compensate for all the contributions from
the decrease in net radiation and wind speed.
The increasing trend in air temperature over the
past 3 decades is consistent with the global warming
as reported across most of the globe. Further study
is needed to determine the extent to which the decrease in wind speed is due to natural climate changes
and changes in the environment of the measuring stations. Previous studies (Kaiser, 2000, 2001; Zhang et
al. 2004) showed that the most likely explanation for
the decrease in the global radiation (sunshine duration) over China is the increase in air pollution. This
paper presents progress results from on-going research
that studies the impact of climate change on floods
in the Changjiang (Yangtze) River basin in China.
The ongoing research and planned studies include investigation and quantification of natural and human
effects on the changing trend of meteorological variables, calculation and regional mapping of actual evapotranspiration in the catchment by using complementary relationship evaporation models and water balance models, investigation of the effect of the changes
in evapotranspiration on flooding and the hydrological cycle in the region, etc. Reference evapotranspiration and pan evaporation provide a measurement of
the integrated effect of radiation, wind, temperature
and humidity on evaporation. In a humid climate,
reference evapotranspiration provides an upper limit
for actual evapotranspiration and, in an arid climate,
it provides a total available energy value for actual
evapotranspiration. For a given precipitation event, a
decrease in reference evapotranspiration will decrease
actual evapotranspiration in both humid and arid climates, which in turn increases the magnitude and frequency of floods. Quantitative estimation of the effect
of different meteorological variables on evaporation, as
was done in this study, is also an important step in
studying the impact of climate change on evapotran-
519
spiration and water balance components.
Acknowledgments. The authors thank the National Climate Center of China for providing the meteorological data. The overseas Chinese Scholar Fund of the
Chinese Academy of Sciences and the Key Project of the
Chinese Academy of Sciences (KZCX3-SW-331) are jointly
funded to do field work and data gathering in the Yangtze
River and to exchange visitors for this topic in China. The
Swedish Research Council (VR), the Swedish Foundation
for International Cooperation in Research and Higher Education (STINT), the Chinese Ministry of Science and Technology, the Swedish International Development Cooperation Agency (Sida) and the China Meteorological Administration partly supported the research. The authors thank
the two anonymous reviewers for their detailed and constructive comments.
REFERENCES
Allen, R. G., L. S. Pereira, D. Raes, and M. Smith, 1998:
Crop evapotranspiration—Guidelines for computing
crop water requirements. AO Irrigation & Drainage
Paper 56, Food and Agriculture Organization, 1998,
ISBN 92-5-104219-5.
Blaney, H. F., and W. D. Criddle, 1950: Determining water requirements in irrigated area from climatological
irrigation data. U. S. Department of Agriculture, Soil
Conservation Service Tech. Paper. No. 96, 48pp.
Chattopadhyay, N., and M. Hulme, 1997: Evaporation and
potential evapotranspiration in India under conditions
of recent and future climatic change. Agricultural and
Forest Meteorology, 87(1), 55–74.
Chen, D., G. Gao, C.-Y. Xu, J. Guo, and G. Ren, 2005:
Comparison of Thorthwaite method and Pan data
with the standard Penman-Monteith estimates of potential evapotranspiration for China. Climate Res.,
28, 123–132.
Dyck, S., 1983: Overview on the present status of the
concepts of water balance models. New Approaches
in Water Balance Computations (Proceedings of the
Hamburg Workshop), A. Van der Beken, and A. Herrmann, Eds, IAHS Publ. No. 148, 3–19.
Hobbins, M. T., J. A. Ramirez, and T. C. Brown, 2001a:
The complementary relationship in estimation of regional evapotranspiration: An enhanced advectionaridity model. Water Resour. Res., 37(5), 1389–1403.
Hobbins, M. T., J. A. Ramirez, T. C. Brown, and L. H.
J. M. Claessens, 2001b: The complementary relationship in estimation of regional evapotranspiration: The
complementary relationship areal evapotranspiration
and advection-aridity models. Water Resour. Res.,
37(5), 1367–1388.
Hobbins, M. T., J. A. Ramirez, and T. C. Brown, 2004:
Trends in pan evaporation and actual evapotranspiration across the conterminous U. S.: Paradoxical or
complementary? Geophys. Res. Lett., 31, L13503,
doi:10.1029/2004GL019846.
Kaiser, D. P., 2000: Decreasing cloudiness over China: An
updated analysis examining additional variables. Geophys. Res. Lett., 27(15), 2193–2196.
520
DECREASING REFERENCE EVAPOTRANSPIRATION IN CHANGJIANG CATCHMENT
Kaiser, D. P., 2001: Assessing observed temperature and
cloud amount trends for China over the last half of
the twentieth century: What can the sunshine duration tell us? Report No. POO-109147, 12th Symposium on Global Change Studies, sponsored by U. S.
Department of Energy.
Liu, B., M. Xu, M. Henderson, and W. Gong,
2004: A spatial analysis of pan evaporation trends
in China, 1955–2000. J. Geophys. Res., 109,
doi:10.1029/2004JD004511.
Ramirez, J. A., and B. Finnerty, 1996: CO2 and temperature effect on evapotranspiration and irrigated agriculture. Journal of Irrigation and Drainage Engineering, ASCE, 122, 155–163.
Thomas, A., 2000: Spatial and temporal characteristics of
potential evapotranspiration trends over China. International Journal of Climatology, 20, 381–396.
Wang, J., Z. Hao, T. Jiang, J. Shi, and T. Zeng, 2003:
VOL. 23
Study on reference evapotranspiration in the Yangtze
River under increasing temperature. Journal of Lake
Sciences, 15, 277–288. (in Chinese)
Xu, C.-Y., and V. P. Singh, 2005: Evaluation of three
complementary relationship evapotranspiration models by water balance approach to estimate actual regional evapotranspiration in different climatic regions.
J. Hydrol., 308, 105–121.
Xu, Z. X., and J. Y. Li, 2003: A distributed approach
for estimating catchment evapotranspiration: Comparison of the combination equation and the complementary relationship approaches. Hydrol. Process.,
17, 1509–1523.
Zhang, Y. L., B. Q. Qin, and W. M. Chen, 2004: Analysis
of 40 year records of solar radiation data in Shanghai, Nanjing and Hangzhou in Eastern China. Theor.
Appl. Climatol., 78, 217–227.