PIAHSProceedings of the International Association of Hydrological SciencesPIAHSProc. IAHS2199-899XCopernicus GmbHGöttingen, Germany10.5194/piahs-371-173-2015Effects of precipitation and potential evaporation on actual
evapotranspiration over the Laohahe basin, northern ChinaLiuY.RenL.njrll9999@126.comYangX.MaM.YuanF.JiangS.State Key Laboratory of Hydrology-Water Resources and
Hydraulic Engineering, College of Hydrology and Water Resources, Hohai
University, Nanjing 210098, ChinaL. Ren (njrll9999@126.com)12June201537137117317912March201512March2015This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://piahs.copernicus.org/articles/371/173/2015/piahs-371-173-2015.htmlThe full text article is available as a PDF file from https://piahs.copernicus.org/articles/371/173/2015/piahs-371-173-2015.pdf
Problems associated with water scarcity are facing new challenges under the
climate change. As one of main consumptions in water cycle on the Earth,
evapotranspiration plays a crucial role in regional water budget. In this
paper, we employ two methods, i.e. hydrological sensitivity analysis and
hydrological model simulation, to investigate the effect of climate
variability and climatic change on actual evapotranspiration (Ea)
within the Laohahe basin during 1964–2009. Calibrations of the two methods
are firstly conducted during the baseline period (1964–1979), then with the
two benchmarked models, simulations in climatic change duration (1980–2009)
are further conducted and quantitative assessments on climatic change-induced
variation of Ea are analysed accordingly. The results show that
affected by combined impacts of decreased precipitation and potential
evapotranspiration, variation of annual Ea in most sub-catchments
suffer a downward trend during 1980–2009, with a higher descending rate in
northern catchments. At decadal scale, Ea shows significant
oscillation in accordance with precipitation patterns. Northern catchments
generally suffer more decadal Ea changes than southern
catchments, implying the impact of climatic change on decadal Ea
is more intense in semi-arid areas than that in semi-humid regions. For whole
changed durations, a general 0–20 mm reduction of Ea is found
in most parts of studied region. For this water-limited region,
Ea shows higher sensitivity to precipitation than to potential
evaporation, which confirms the significant role of precipitation in
controlling Ea patterns, whereas the impact of potential
evapotranspiration variation would be negligible.
Introduction
The compelling phenomenon of global warming (IPCC, 2007) has attracted much
attention recently. Huntington (2006) pointed out that the most important
consequence due to climatic change should be its intensification of
hydrological cycle both at regional and global scales. As the main
consumption of water resources, evapotranspiration plays an important role in
regional water budget, especially for water-deficit areas. During the past 50
years, a downward trend of pan evaporation and/or reference
evapotranspiration has been found in most regions of the world, covering
countries both in northern and southern hemispheres (Peterson et al., 1995;
Chattopadhyay et al., 1997; Roderick et al., 2004; Liu et al., 2004; Burn et
al., 2007).
Location of the Laohahe basin (LHH), distribution of
hydro-meteorological stations, and watershed divides of 9 sub-catchments
with their names.
The decline of pan and reference evaporation indicates accepted reduction in
total energy supply, but it is also debatable whether it implies a similar
tendency for actual evapotranspiration (Ea). Liu et al. (2010)
analyzed the Ea over the Yellow River basin, and a decreasing
trend was detected during 1961–2006. Similar results were obtained in the
Hai River basin, China (Li et al., 2013). In this study, the Laohahe basin,
located in northern China, is chosen as a case study on Ea
modeling and relevant spatio-temporal variations analyzing. To address this,
two methods are adopted. One is the hydrological sensitivity analysis (HS)
implemented on the basis of Fu's equation (Fu, 1981). The other is the
hydrological model simulation (HM) by the Variable Infiltration Capacity
(VIC) model. Specific objectives of our study are to calculate and compare
annual Ea with two different methods, and to comprehensively
estimate the impacts of climate variability and climate change on
Ea at multiple time scales.
Study area and data
The Laohahe basin is located at the junction of the Hebei Province, Liaoning
Province, and Inner Mongolia Autonomous Region between 41–42.75∘ N,
117.25–120∘ E (Fig. 1), with 18 112 km2 of area. Its
elevation ranges from 427 to 2054 m, and significantly descends from
southwest to northeast. The mean areal annual temperature, precipitation and
runoff during the period of 1964–2009 are 7.58 ∘C, 418.3 and
28.7 mm respectively, and about 80 % of the annual precipitation falls
from May to September every year. Besides the whole Laohahe basin, we select
other 8 sub-catchments (Fig. 1) including 6 headwater catchments (colored
areas) and 2 midstream catchments (marked with red-line boundaries) for
analysis.
Observed daily records of 52 precipitation-gauged stations and 9
discharge-measured ones are provided by the Water Resources Department of
Inner Mongolia Autonomous Region. Meteorological observations are obtained
from four national standardized meteorological stations. All these
hydro-meteorological variables have continuous records from 1964–2009. The
Inverse Distance Weighting method (IDW, Bartier and Keller, 1996) is adopted
to generate distributed precipitation and areal values. Potential evaporation
(PET) is calculated with the Penman-Monteith equation recommended by the Food
and Agriculture Organization (FAO, Allen et al., 1998). In addition, the soil
and vegetation data needed by VIC are collected. The soil data are derived
from the 5-min the Food and Agriculture Organization dataset, and the land
cover data is provided by the Chinese Academy of Science.
MethodologyHydrological sensitivity analysis
Hydrological sensitivity analysis is based on the assumption that a
variation in mean annual runoff or evapotranspiration can be determined by
the following expression (Koster and Suarez, 1999):
ΔX=βΔP+γΔPET.
where ΔX represents the change in runoff or evapotranspiration in
response to alterations in precipitation and PET; β and γ are
the sensitivity coefficients with respect to precipitation and PET
respectively. The sensitivity coefficients are generally derived from
evaporation equations by computing partial derivatives with respect to
precipitation and PET, respectively, which can be further expressed as:
ΔE=∂E∂P×ΔP+∂E∂PET×ΔPET.
In this study, the Fu's equation (Fu, 1981) proposed on the basis of the
Budyko's hypothesis is employed to estimate average annual
evapotranspiration and relevant sensitivity coefficients, whose analytical
expression is described as:
EaP=1+PETP-[1+PETPϖ]1/ϖ.
where ϖ is a model parameter determined by land surface conditions
including relative infiltration capacity, catchment average slope and
fractional vegetation coverage (Yang et al., 2007). Accordingly, sensitivity
coefficients (partial derivatives) based on Eq. (3) are given as:
∂Ea∂P=1-[1+PETPϖ]1ϖ-1∂Ea∂PET=1-[1+PETPϖ]1ϖ-1PETPϖ-1.
Thus, estimating the impact of climatic change on evapotranspiration can be
realized by substituting Eqs. (4) and (5) into Eq. (2). It should also be
noted that the parameter ϖ is calibrated by comparing average annual
Ea derived from Eq. (3) and from water balance analysis during
the baseline period. Relevant performance is evaluated by three statistical
indices: the root of mean square error (RMSE), relative bias (BIAS) and
correlation coefficient (CC).
Comparison of annual Ea simulated by VIC-3L model and
Fu's equation for (a–i) 9 catchments. The black solid (red hollow) circles
represent results in the baseline (subsequent) periods.
Optimized parameters and performance of the VIC model during the
baseline period.
CatchmentsCalibrated parameters Model performance id1d2d3DsDsmaxWsNSCEBIAS (%)XJD0.250.070.520.0160.980.720.9CTL0.260.070.6220.0190.980.70.4JS0.540.070.6220.0380.980.6-6DZ0.310.070.420.0160.980.85.6CF0.270.070.6220.0180.980.713.9TPZ0.20.070.6420.00660.980.822.8LHH0.30.070.5720.0190.980.812.1
Estimates of ΔE (mm yr-1) for 9 catchments derived from
hydrological model simulation method (HM) and hydrological sensitivity
analysis method (HS) during the changed periods.
Catchments1980–1989 1990–1999 2000–2009 HMHSHMHSHMHSXJ-15.3-23.139.525.4-50.5-49.7CTL-14.9-0.948.649.2-42.1-36.8XJD-51.8-64.219.018.0-39.4-29.6YSWZ-63.0-49.8-0.112.3-67.7-48.3JS-33.5-24.927.226.2-35.5-28.4DZ-52.7-50.111.0-0.6-44.8-56.7CF-25.7-24.240.838.0-40.3-36.5TPZ-45.5-51.127.215.6-47.3-56.3LHH-38.6-36.321.224.3-49.6-51.2Hydrological model simulation
Hydrological model is also a widely used tool for assessing hydrological
responses to climate change (Bao et al., 2011; Lan et al., 2013). Different
from hydrological sensitivity analysis, it elaborately depicts hydrological
processes with physically based framework at flexible time step. In this
study, the semi-distributed Variable Infiltration Capacity model (VIC, Liang
et al., 2004) is selected, which plays multiple roles, as both a
hydrological model and land surface model (LSM). VIC has been extensively
used in studies on topics ranging from water resources management to
land-atmosphere interactions. It balances both the water and surface energy
budgets within the grid cell, and accounts for the effects of sub-grid-scale
variability in soil, vegetation, precipitation, and topography on grid-scale
fluxes. Specifically for its evaporation section, the total actual
evapotranspiration over a grid cell is computed as a weighted sum of three
types of evaporation including evaporation from the canopy layer of each
vegetation tile, transpiration from each of the vegetation tile and
evaporation from the bare soil.
Actual evapotranspiration calculated by VIC is mainly controlled by soil and
vegetation parameters, and most of them can be obtained from remote sensing
data and relevant organizations (Hansen et al., 2000; FAO, 1998) without
calibration, except seven parameters (i, d1, d2, d3,
Ds, Dsmax and Ws) which are subject to
calibration based on the agreement between simulated and observed hydrographs
(Xie et al., 2006). For model implementation, the VIC model is run at a daily
time step with 0.0625∘× 0.0625∘ of spatial
resolution. Calibration is conducted on streamflow by optimizing seven
parameters with two criteria: Nash-Sutcliffe Coefficient of Efficiency (NSCE)
and BIAS. Table 1 lists the seven optimized parameters and the calibration
results for main sub-catchments. It is seen that the NSCEs vary between 0.6
and 0.82 and absolute values of the BIAS range from 0.4 to 6 %,
indicating that the calibrated VIC model generally capture the natural
variability of observed hydrographs.
Results and discussionActual evapotranspiration modeling
Fig. 2 shows the comparison of annual Ea derived from the two
models in the baseline period (1964–1979) and change periods (1980–2009),
respectively. The best fitted results occur in XJ, CF and LHH catchments with
all calculated samples close to the 1:1 line (Fig. 2a, g, i). As regards DZ
and TPZ catchments, the Ea series derived from Fu's equation are
generally smaller than that from the VIC model (Fig. 2f, h), and opposite
patterns are exhibited in YSWZ, JS and CTL catchments (Fig. 2b, d, e).
Generally, for all 9 catchments points in the scatter plots are concentrated
with the values of RMSE varying between 20.1 and 50.3 mm and those of CC
almost above 0.8.
Annual series of climate-induced ΔE over the whole changed
periods. Black solid (red dash) lines denote the ΔE calculated using
the hydrological model simulation method (hydrological sensitivity analysis
method). The values in the lower right hand corner of each plot give the
correlation coefficient between the two derived ΔE series.
Spatial distribution of ΔE derived from (a) hydrological
sensitivity analysis method (HS) and (b) hydrological model simulation method
(HM) during the whole changed period of 1980–2009.
Multi-scale analysis of Ea changes
The difference between simulated Ea in the changed periods and
baseline periods are further analyzed at multiple time scales (annual,
decadal and whole changed durations) to reflect heterogeneous characteristics
of Ea changes (ΔE). At annual scale, simple linear
regression is introduced to reflect ΔE trend during 1980–2009. As
can be seen in Fig. 3, most sub-catchments suffer negative trends, and the
results are similar to Li et al. (2013) which demonstrated the actual
evapotranspiration in Haihe River basin (close to our study area with similar
climatic conditions) significantly decreased. On the whole, northern
catchments (CTL, XJ and CF) within the Laohahe basin have suffered higher
descending rates than the southern catchments (YSWZ, JS and TPZ). It implies
that tendency of climatic change in semi-arid regions is more significant
than in semi-humid areas.
At decadal scale, a decrease-increase-decrease pattern of ΔE is found
in those 9 catchments (Table 2). For the whole studied basin, ΔE
computed by the hydrological model are -38.6, 21.2 and -49.6 mm in
1980s, 1990s and 2000s, respectively, while corresponding values given by the
hydrological sensitivity analysis method are -35.8, 25 and -50 mm,
respectively. It is obvious that dry climatic circumstances (1980s and 2000s)
have imposed more effect on Ea than the wet status (1990s). For
the two midstream catchments, TPZ catchment shows more reduction of
Ea than CF catchment during dry decade, whereas more increments
of Ea are found in CF catchment in the wet decade. It implies
that the impact of climatic change on Ea is more intense in
semi-arid areas than that in semi-humid regions. Considering the six
headwater catchments, DZ, YSWZ, XJD and JS catchments located in southern
parts generally show similar patterns of Ea variations with TPZ
catchment, while that of CTL and XJ catchments situate in northern parts are
analogous to CF catchment.
During the whole changed duration (1980–2009), a general 0–20 mm reduction
of Ea is found in most parts of the Laohahe basin with larger
reduction in southwest part of the Laohahe basin, in which more than 40 mm
of reduction in Ea is observed (Fig. 4). Additionally, a
0–80 mm increment of Ea is detected in northern and eastern
parts of CF catchment. Some differences in derived spatial distribution of
ΔE from the two methods are also found, especially in TPZ catchment,
where hydrological model simulation method seems to present more reduction of
Ea than hydrological sensitivity analysis method.
Conclusions
In this study, we adopt two different methods to evaluate the impact of
climate change on Ea within the Laohahe basin. Like other
statistical methods, the hydrological sensitivity method treats the studied
basin as a black box, and provides the relationship between climate change
and hydrological responses without consideration of subsistent hydrological
processes. This hypothesis guarantees its effectiveness in general reflecting
Ea changes induced by quantitative change of annual precipitation
and PET, whereas other effects such as seasonal properties of climatic
variables which also influence hydrological responses are not captured. In
contrast, the hydrological model simulation method takes its advantages for
the elaborate depiction of hydrological processes at a daily step. Overall,
at annual scale, the Ea simulated by the two methods is
comparable, meanwhile the spatiotemporal variation of Ea derived
by each method is similar.
Affected by combined impacts of decreased precipitation and PET, most
sub-catchments of this region have suffered a downward trend of annual
Ea with a higher descending rate in northern catchments. At
decadal scale, Ea presents significant decadal oscillation, and
northern catchments generally suffer more changes of Ea than
southern catchments, implying that the impact of climatic change on
Ea is more intense in semi-arid areas than that in semi-humid
regions. For whole changed durations, a general 0–20 mm reduction of
Ea is found in most parts of this region, which is in good
agreement with the pattern of precipitation variability.
Acknowledgements
This work was supported by the Special Basic Research Fund for Methodology in
Hydrology (Grant no. 2011IM011000) from the Ministry of Sciences and
Technology, China, the National Natural Science Foundation of China (Grant
no. 41323001), the National Key Technology R&D Program by Ministry of
Sciences and Technology, China (Grant no. 2013BAC10B02), the 111 Project
(Grant no. B08048) from the Ministry of Education and State Administration of
Foreign Experts Affairs, China, the National Natural Science Foundation of
China (Grant no. 41201031) and the Fundamental Research Funds for the Central
Universities of China (Grant no. 2014B35814, Grant no. 2014B35914).
References
Allen, R. G., Pereira, L. S., Raes, D., and Smith, M.: Crop
Evapotranspiration: Guidelines for Computing Crop Water Requirements. United
Nations Food and Agriculture Organization, Irrigation and Drainage Paper,
56, 1–15, 1998.
Bao, Z., Zhang, J., Liu, J., Wang, G., Yan, X., Wan, X., and Zhang, L.:
Sensitivity of hydrological variables to climate change in the Haihe River
basin, China, Hydrol. Process., 26, 2294–2306, 2012.
Bartier, P. M. and Keller, C. P.: Multivariate interpolation to incorporate
thematic surface data using Inverse Distance Weighting (IDW), Comput.
Geosci-uk., 22, 795–799, 1996.
Burn, D. H. and Hesch, N. M.: Trends in evaporation for the Canadian
Prairies, J. Hydrol., 336, 61–73, 2007.
Chattopadhyay, N. and Hulme, M.: Evaporation and potential evapotranspiration
in India under conditions of recent and future climate change, Agr. For.
Meteorol., 87, 55–73, 1997.
Fu, B.: On the calculation of the evaporation from land surface, Sci. Atmos.
Sin., 5, 23–31, 1981 (in Chinese).
Huntington, T. G.: Evidence for intensification of the global water cycle:
review and synthesis, J. Hydrol., 319, 83–95, 2006.
IPCC: Climate change 2007, the physical science basis. Contribution of
working group I to the fourth assessment report of the intergovernmental
panel on climate change. Cambridge University Press, Cambridge, 2007.
Koster, R. D. and Suarez, M. J.: A simple framework for examining the
interannual variability of land surface moisture fluxes, J. Climate, 12,
1911–1917, 1999.
Lan, C., Zhang, Y., Gao, Y., Hao, Z., and Cairang, L.: The impacts of
climate change and land cover/use transition on the hydrology in the upper
Yellow River Basin, China, J. Hydrol., 502, 37–52, 2013.
Li, X., Marco, G., Zhai, J., Liu, X., Su, B., and Wang, Y.: Spatio-temporal
variation of actual evapotranspiration in the Haihe River Basin of the past
50 years, Quatern. Int., 304, 133–141, 2013.Liu, B., Xu, M., Henderson, M., and Gong, W.: A spatial analysis of pan
evaporation trends in China, 1955–2000, J. Geophys. Res., 109,
D15102, 10.1029/2004JD004511, 2004.
Liu, Q. and Yang, Z.: Quantitative estimation of the impact of climate change
on actual evapotranspiration in the Yellow River Basin, China, J. Hydrol.,
395, 226–234, 2010.
Liang, X., Guo, J., and Leung, L. R.: Assessment of the effects of spatial
resolutions on daily water flux simulation, J. Hydrol., 298, 287–310, 2004.
Peterson, T. C., Golubev, V. S., and Groisman, P. Y.: Evaporation losing its
strength, Nature, 377, 687–688, 1995.
Roderick, M. L. and Farquhar, G. D.: Changes in Australian pan evaporation
from 1970 to 2002, Int. J. Climatol., 24, 1077–1090, 2004.
Xie, Z., Yuan, F., Duan, Q., Zheng, J., Liang, M., and Chen, F.: Regional
parameter estimation of the VIC land surface model: methodology and
application to river basins in China, J. Hydrometeorol., 8, 447–468, 2007.
Hansen, M. C., Defries, R. S., Yownshend, J. R. G., and Sohlberg, R.: Global
land cover classification at 1 km spatial resolution using a classification
tree approach, Int. J. Remote. Sens., 21, 1331–1364, 2000.Yang, D., Sun, F., Liu, Z., Cong, Z., Ni, G., and Lei, Z.: Analyzing spatial
and temporal variability of annual water-energy balance in non-humid regions
of China using the Budyko hypothesis, Water Resour. Res., 43,
10.1029/2006WR005224, 2007.