research article evapotranspiration trend and its...

Research ArticleEvapotranspiration Trend and Its Relationshipwith Precipitation over the Loess Plateau duringthe Last Three Decades

Zesu Yang12 Qiang Zhang2 and Xiaocui Hao3

1Plateau Atmospheric and Environment Key Laboratory of Sichuan Province College of Atmospheric SciencesChengdu University of Information Technology Chengdu 610225 China2Institute of Arid Meteorology China Meteorological Administration Key Laboratory of Arid Climatic Change and Reducing Disasterof Gansu Province Key Open Laboratory of Arid Climatic Change and Disaster Reduction of CMA Lanzhou 730020 China3Northwest Regional Climate Center Lanzhou 730020 China

Correspondence should be addressed to Zesu Yang zesuyangoutlookcom

Received 25 December 2015 Accepted 5 June 2016

Academic Editor Gang Liu

Copyright copy 2016 Zesu Yang et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

There have been few studies conducted on the changes in actual ET over the Loess Plateau due to the lack of reliable ET dataBased on ET data simulated by the Community Land Model the present study analyzed the changes in ET over the Loess PlateauThe results showed the domain-average ET to have decreased in the past 31 years at a rate of 078mm yearminus1 ET fluctuated muchmore strongly in the 1990s than in the 1980s and 2000s and apart from in autumn ET decreased in all seasons In particular ETin summer comprised about half of the annual ET trend and had the sharpest trend dominating the interannual decline ET alsodecreasedmore sharply in the semiarid than semihumid regionsThe declining trend of ETwas attributed to declining precipitationand air humidity Locally the ET trend was closely related to local mean annual precipitation in areas with precipitation less than400mm ET showed a decreasing trend in areas with precipitation larger than 600mm ET showed an increasing trend and inareas with precipitation in the range of 400ndash600mm could be classified as a transitional zone

1 Introduction

Evapotranspiration (ET)mdashthe process of transferring waterfrom the land surface to the atmospheremdashis the link betweenthe global water cycle energy cycle and carbon cycle andis of critical importance for agriculture hydrology ecologyand the climate system [1 2] Changes in ET will also changethe energy partitioning between sensible and latent heataltering atmospheric dynamics and influencing weather andclimate [3] A detailed understanding of changes in global andregional ET could in turn help us to understand the globalwater cycle and its role in the climate system

Against the background of increased air temperatureassociated with global warming the atmosphere is able tohold more water leading to increases in the fluxes of waterglobally via the water cycle [4] Usually the evidence putforward to support this conclusion involves increases in

precipitation and runoff but not ET trends because the latterare very hard to observe Potential ETmdashthe upper limit of ETindicative of evaporative abilitymdashis relatively easy to obtainplus there are also some other variables that can representevaporative ability including pan evaporation and referenceETMany studies have investigated trends of pan evaporationpotential ET and reference ET regionally or globally Thesetrends however are not consistent among different regionsof the globe Theoretically global warming would increasepan evaporation but there are only a few regions wherethe evidence supports this inference including Israel [5]northeast Brazil [6] Australia [7] andChina [8] In factmostregions indicate an opposite conclusion that is a decreasein pan evaporation for example America [9 10] Canada[11] Greece [12] India [13] Thailand [14] Japan [15] China[16 17] Australia [18] and New Zealand [19] Likewise twostudies on potential ET have also concluded opposite trends

Hindawi Publishing CorporationAdvances in MeteorologyVolume 2016 Article ID 6809749 10 pageshttpdxdoiorg10115520166809749

2 Advances in Meteorology

55∘N

45∘N

35∘N

25∘N

15∘N75∘E 85∘E 95∘E 105∘E 115∘E 125∘E 135∘E

Humid Arid HyperaridSemiaridSemihumid

(a)

40∘N

38∘N

36∘N

34∘N

102∘E 105∘E 108∘E 111∘E 114∘E

0 200

(km)

N

Provincial boundaryRiversPlainHill

DuneTable landMountain landStation

(b)

Figure 1 The (a) location and climate of the Loess Plateau (black line) and (b) its land cover and the locations of the ET observatories (redstars)

in different regions [20 21] Actual ET is closely related topotential ET or pan evaporation but the implication of thetrend in pan evaporation for the trend in actual ET has longbeen disputed Some scientists argue that the declining trendin pan evaporation indicates a declining trend in actual ETOthers suggest that a complementary relationship betweenpotential and actual ET exists with declining ET meaning anincrease in actual ET Recent studies have shown that thesetwo relationships are correct under certain conditions [17 2223] Apparently the positive relationship is satisfied underconditions of sufficient water while the complementaryrelationship is satisfied under water-limited conditions

Potential ET is decreasing in most regions of Chinahowever pan evaporation and potential ET are increasing inthe Loess Plateau [24 25] Since the trend of pan evaporationand potential ET only provides us with a clue for the possibletrend of actual ET and because potential ET and pan evap-oration have a special trend in the Loess Plateau comparedto other regions of China our goal in the present studywas to probe the characteristics of actual ET by examiningwhether the trend in actual ET in the Loess Plateau alsoshows distinct regional features like potential ET Besidesbecause the Loess Plateau is situated in the boundary zonesinfluenced by the Asian summer monsoon precipitationdemonstrates considerable spatial variation [26] which inturn causes spatial variation in land surface hydrothermalprocesses It is therefore also of interest to investigate therelationship between ET trends and precipitation on the localscale

For a long time research on actual ET has been limitedby a lack of reliable ET data [1] In recent years however theGlobal Land Data Assimilation System (GLDAS) has becomean important dataset for global change and water cycleresearch [27 28] Employing the Community Land Model(CLM) GLDAS produces land surface ET data on the globalscale and its land surface variables are reported to be highly

consistent with observations in the Loess Plateau [27] In thisstudy we first validated the ET products of GLDAS using ETdata observed by an eddy covariance (EC) system at four landsurface observatories in the Loess Plateau Then on the basisof GLDAS ET data we analyzed the trends of ET and theirreasons over the Loess Plateau In addition the relationshipbetween ET trends andmean annual precipitation locally wasalso analyzed

2 Study Area and Data

21 Study Area The Loess Plateaumdashlocated to the north ofthe Yin Mountains south of the Tsinling Mountains east ofthe Tai-hang Mountains and west of the Riyue Mountainsmdashis a highly unique land type with special ecosystems Theaverage altitude is approximately 1000m above sea levelincreasing from southeast to northwest The most commontopography is gully and hilly terrain and the specific geo-morphology is shown in Figure 1The Loess Plateau is locatedin the boundary zone of the Asian summer monsoon It is atransitional region from the humid climate in the southeastunder the influence of the summer monsoon to the aridclimate in the northwest (Figure 1 rectangle) and thus canbe divided into two parts that is the semiarid zone in thenorthwest and the semihumid zone in the southeast [29] (Fig-ure 1) The annual average rainfall is approximately 466mmwith 700mm in the southeastern regions and 120mm in thenorthwestern regions The mean annual temperature rangesfrom4∘C to 12∘CThemain soil type is loess which has aweakability to resist wind and soil erosion

22 Data and Data Processing

221 Simulated ET Data The simulated ET data were down-loaded from GLDAS As the main dynamic framework theland surface model is the key part of GLDAS The forcing

Advances in Meteorology 3

Table 1 Information on the validation sites

Site Location Elevation (m) Land cover Precipitation (mm) Temperature (∘C) ClimateDingxi 35∘351015840N 104∘371015840E 1896 Cropland 386 67 SemiaridSACOL 35∘571015840N 104∘081015840E 1966 Grassland 3818 67 SemiaridQingyang 35∘411015840N 107∘511015840E 1280 Cropland 562 88 SemihumidPingliang 35∘341015840N 106∘351015840E 1480 Cropland 5112 87 Semihumid

data were provided by the Global Data Assimilation Sys-tem of the National Centers for Environmental Predictionand the Data Assimilation System of the Goddard SpaceFlight Center This study used monthly ET data simulatedby the Community Land Model (CLM) with a horizon-tal resolution of 1∘ times 1∘ from January 1982 to Decem-ber 2013 A rectangular region of (34∘ndash41∘N 101∘ndash114∘E)including the Loess Plateau was selected The simulationdata can be obtained directly from the website of GLDAShttpdiscscigsfcnasagovhydrologydata-holdings

CLM was developed from three land models the LandSurface Model of the National Center for AtmosphericResearch the land model of the Institute of AtmosphericPhysics Chinese Academy of Sciences land model and theBiosphere-Atmosphere Transfer Scheme As such it has thebest features of these three land models and offers improve-ments regarding certain parameterization schemes CLMwasdeveloped as the land surface module of the CommunityEarth SystemModel and theCommunityAtmosphereModelLand biogeophysical and hydrological processes are the mainprocesses simulated by CLM including (1) vegetation com-position structure and phenology (2) absorption reflectionand transmittance of solar radiation (3) absorption andemission of longwave radiation (4) momentum sensibleheat (ground and canopy) and latent heat (ground evapora-tion canopy evaporation and transpiration) fluxes (5) heattransfer in the soil and snow including phase changes (6)canopy hydrology (interception throughfall and drip) (7)snow hydrology (snow accumulation and melt compactionand water transfer between snow layers) (8) soil hydrology(surface runoff infiltration subsurface drainage and redistri-bution of water within the column) (9) stomatal physiologyand photosynthesis When simulating latent heat flux theequations used for the ET calculation are [30]

119864 = 119864V + 119864119892

119864V = minus120588atm(119902119904

minus 119902sat119879V)

119903total

119864119892

= minus120588atm(119902119904

minus 119902119892

)

1199031015840

aw

1

119903total=119871sun

119903119887

+ 119903sun119904

+119871sha

119903119887

+ 119903sha119904

119903119887

=1

119862V(119880lowast

119889leaf)

minus12

1199031015840

aw =1

119862119904

119906lowast

1

119903119904

= 119898119860

119888119904

119890119904

119890119894

119875atm + 119887

(1)

where 119902119904

is the specific humidity of the canopy roof 119902119892

is the specific humidity of the land surface 119902sat119879V is the

humidity of the canopy 119864V is the vegetation transpiration119864119892

is the evaporation 119903total is the total resistance to watervapor transfer from the canopy to the canopy air includingthe leaf boundary resistance 119903

119887

and stomatal resistance 119903119904

1199031015840awis the aerodynamic resistance to water vapor transfer betweenthe ground and the canopy air A is the photosynthesism is a parameter associated with atmospheric conditionsand plant type b is an empirical parameter 119906lowast is frictionvelocity 119862

119904

is the turbulent transfer coefficient 119871sun and 119871shaare the sunlit and shaded leaf area indices respectively 119903sun

119904

and 119903sha119904

are the sunlit and shaded stomatal conductancewhich can use the average absorbed photosynthetically activeradiation for sunlit and shaded leaves respectively 119888

119904

is theCO2

concentration at the leaf surface 119890119904

is the vapor pressureat the leaf surface 119890

119894

is the saturation vapor pressure inside theleaf at the vegetation temperature and119875atm is the atmosphericpressure

222 In Situ ET Data ET was measured by an EC (eddycovariance) system at four land surface observatories inthe Loess Plateau (see Figure 1 and Table 1) The EC sys-tem included a three-dimensional ultrasonic anemometer(CSAT-3 Campbell USA) and a quick response infraredanalyzer (Li7500 Li-Cor USA) and was set at a height of25mThe sampling frequencywas 10Hz Latent heat fluxwascalculated by

LE = 12058812058212059610158401199021015840 (2)

where LE is latent heat flux 120588 is air density 120582 is the latentheat of vaporization and1205961015840 and 1199021015840 are the fluctuations aroundthe averages of vertical wind speed and specific humidityrespectively

The data from the EC system were processed as follows[31] first quality control and gap filling (i) half-hourly datafrom periods of sensor malfunction were deleted (ii) half-hourly data within 1 h before or after rainfall were deleted(iii) half-hourly data (119909

119894

) were deleted when 119909119894

ge (+4120575) or119909119894

le (minus4120575) where 120575 is standard deviation On average thedatasets contained 189missing or rejected ETdataMissingdata were filled using a look-up table method [32] For each


Table 2 Correlation statistics between predicted ET and observations at four sites in the Loess Plateau

Site Fitting coefficient Correlation coefficient Root-mean-square error (mm) Relative error ()SACL 0969 094 79 181Dingxi 0834 087 94 215Pingliang 0918 095 84 193Qingyang 0935 096 68 224Correlation coefficients passed the significance test at all sites (119875 lt 001)

study site tables were created based on the four seasons Foreach season ET was classified by 20 119877

119899

-classes times 35 VPD (119877119899

net radiation VPD vapor pressure deficit) 119877

119899

consisted of50Wmminus2 intervals from minus200 to 800Wmminus2 Similarly VPDconsisted of 015 kPa intervals ranging from 0 to 51 witha separate class for VPD = 0 Gaps with no mean assignedin the look-up tables were interpolated linearly Tables ofET means and standard deviation were produced by theprocedureThus we filled the missing data by referring to thesynchronous 119877

119899

and VPD Then half-hourly LE data weretransformed to half-hourly ET Finally half-hourly ET datawere aggregated to monthly ET

Since energy disclosure is an inevitable problem for EC[33] The energy closure ratio of ET across all sites wasevaluated and the results showed that the averaged energyclosure ratio was 862 which is superior to the typicalenergy closure ratio of EC measurements (15ndash30) [34]Thus observed ET at the validation sites were consideredreliable Unaccounting for advection fluxes is one of thereasons for energy disclosure at EC sites Another reason isthat EC towers cannot capture the large eddies (with lowfrequency) associated with stationary secondary circulationsthat generate over tall canopies andheterogeneous landscapes[35ndash39]

223 Regional Climate Data Climate data including dailyprecipitation temperature air humidity sunshine durationand wind speed were obtained from the National ClimateCenter China At the time of study there were 106 mete-orological stations across the Loess Plateau These stationsrsquoclimate data were aggregated to the monthly timescale tomatch that of the ET data When analyzing the domain-average trend of these climate variables we averaged them forthe 106 stations

224 Vegetation Index Vegetation indices are radiometricmeasures of photosynthetically active radiation absorbed bychlorophyll in the green leaves of vegetation canopies andare therefore good surrogate measures of the physiologicallyfunctioning surface greenness level of a region The latestversion of the GIMMS NDVI (Global Inventory Model-ing and Mapping Studies Normalized Difference VegetationIndex) dataset spans the period July 1981 to December 2013and is termed NDVI3g (third-generation GIMMS NDVIfromAdvanced Very High Resolution Radiometer (AVHRR)sensors) The NDVI3g time series is an improved 8 kmNDVI dataset produced from AVHRR instruments and canbe downloaded from httpecocastarcnasagovdatapub

gimms This study used gridded NDVI data over the LoessPlateau for the period of January 1982 to December 2013During data preprocessing we spliced the images trans-formed the projection and extracted the study area (the LoessPlateau) The horizontal resolution of NDVI was set to 1∘times 1∘to match the horizontal resolution of the ET data simulatedby CLM with the grid values interpolated by linear averagesof the surrounding area

23 Validation of GLDAS ET Product A mismatch existedbetween the scales of the simulated and measured ET whichcould have introduced uncertainty in the results To datean efficient way to overcome this problem remains elusiveHowever in situ ET data can provide a reference for thesimulated ET [40] Given this the ET produced by GLDASwas validated using the ET observed by the eddy covariancesystem at four sites in the Loess Plateau Figure 2 comparesthe monthly CLM-simulated and measured ET at the foursites It can be seen that the simulated ET was highlyconsistent with the measured ET at the four validation sitesAs shown in Table 2 the fitting coefficient was close to 1the average coefficient of determination (1198772) was about 087the average root-mean-square error was about 812mm andthe relative error was about 2032 The precision of remotesensing methods for estimating regional ET is within therange of 15ndash30 [1] CLM produced ET that achieved anupper-middle level of precision indicating that the CLM-simulated ET was reliable for the Loess Plateau

3 Results and Discussion

31 Trends of Annual Decadal and Seasonal ET Figure 3(a)shows the time series of domain-average annual ET over theLoess Plateau It can be seen that ET decreased from 1982 to2013 over the Loess Plateau at a rate of 078mmyearminus1 Thedecreasing trend was weak before 1998 and then acceleratedafter 1998 The accelerated decrease in ET after 1998 may berelated to severe drought caused by strong El Nino eventsin 1998 [41] and the temperature increase slowing down[27] Figure 3(b) shows the time series of the ratio of ET toprecipitation which also shows an obvious decreasing trendThe decreasing trend implies that ET decreased more sharplythan precipitation during the study period water recyclingfrom the land surface to the atmosphere was becomingweaker as was the regional water cycle

In terms of the interdecadal characteristics of changesin ET over the Loess Plateau (Table 3) ET decreased by41mm from the 1980s to the 1990s and then decreased more


0 20 40 60 80 100Measured ET (mm)

SACOL site

R2 = 08782

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Dingxi site

R2 = 07655

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Qingyang site

R2 = 0933

0

20

40

60

80

100ET

sim

ulat

ed b

y CL

M (m

m)

20 40 60 80 1000Measured ET (mm)

Pingliang site

R2 = 09049

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Figure 2 Comparison of simulated and observed ET at four sites (SACOL Dingxi Pingliang and Qingyang) in the Loess Plateau

200

250

300

350

1982 1986 1990 1994 1998 2002 2006 2010Dom

ain-

aver

age a

nnua

l ET

(mm

)

(a)

03

05

07

09

1982 1986 1990 1994 1998 2002 2006 2010

Evap

orat

ion

ratio

(E

Tpr

ecip

itatio

n)

(b)

Figure 3 Trend of the (a) ET and (b) ET to precipitation ratio over the Loess Plateau during the period 1982ndash2013


Table 3 Interdecadal characteristics of ET over the Loess Plateau

Decade Mean (mm) Standard deviation (mm)1980s 3157 2171990s 3116 2942000s 2948 206

0

40

80

120

160

200

1982 1986 1990 1994 1998 2002 2006 2010

ET in

each

seas

on (m

m)

Winter SpringSummer Autumn

Figure 4 Trend of ET in each season over the Loess Plateau during1982ndash2013

rapidly from the 1990s to the 2000s (by 168mm) Besides thestandard deviation for the 1990s was much larger than that ofthe 1980s and 2000s implying that ET fluctuated abnormallystrongly in the 1990s which could have caused disturbance ofthe water cycle leading to drought [42]

The ET trend in each season and their contribution tothe overall ET trend were also investigated (Figure 4) Foreach year the seasonal ET was calculated by summing themonthly ET within the separate seasons It can be seen thatET decreased in all seasons except autumn Summer hadthe largest magnitude accounting for around half of theannual ET ET also decreased most rapidly during summerat a rate of 043mmyearminus1 In spring ET fluctuated stronglyespecially around 1998 and had a relatively small rate ofdecrease ET had a small mean and rate of decrease in winterand in autumn ET increased slightly but nonsignificantlyOverall the summer trend dominated the declining ET trendover Loess Plateau

32 Local ET Trend over the Loess Plateau Although thedomain-average ET decreased over the Loess Plateau during1982ndash2013 it was possible that local ET trends may havediffered due to the complicated topography and distinctfeatures of local climate Figure 5 shows the local ET trendsover the Loess Plateau during the study period As can beseen ET decreased in most parts of the Loess Plateau withthe rate of decrease ranging from 1 to 3mmyearminus1 An ETincrease was apparent in only a few parts of the Loess Plateauwith the rate of increase being less than 2mmyearminus1 in mostcases This small increasing ET trend was dominated by thelarger decreasing ET trend meaning the domain-averageET showed an overall decreasing trend The areas with anincreasing ET trend were located in Tianshui District and theintersection of the Yellow River Luo River and Wei Riverwhere there is a relatively high level of water availability

425∘N

395∘N

365∘N

335∘N101∘E 105∘E 109∘E 113∘E

minus4 minus3 minus2 minus1 0 1 2 3

Climate tendency of evapotranspiration (mm yearminus1)

Figure 5 Distribution of the climate tendency of ET over the LoessPlateau

100

200

300

400

500

1982 1986 1990 1994 1998 2002 2006 2010

ET (m

m)

SemiaridSemihumid

Figure 6 Trend of ET in the semiarid and semihumid areas of theLoess Plateau

Figure 6 further breaks down the ET trend into that of thesemiarid and that of semihumid region of the Loess Plateau Itcan be seen that ETwasmuch larger in the semihumid regionthan in the semiarid region Furthermore ET decreasedmore rapidly in the semiarid region than in the semihumidregion with their rates of decrease being 0025mmyearminus1and 089mmyearminus1 respectively A larger availability ofwaterslowed the rate of decrease in the semihumid region

33 Attribution of the ET Trend To elucidate the factorscausing the decreasing ET trend we examined the temporalevolution and trends of annual precipitation air humiditytemperature sunshine duration wind speed and vegetationconditions Annual precipitation and air humidity showed anobvious decreasing trend during the study period (Figures7(a) and 7(b)) meaning that water availability was decreasingover the Loess Plateau The trend in sunshine duration wasnot significant (Figure 7(c)) Temperature and wind speedshowed marked increasing trend indicating that the evapo-ration potential was increasing (Figures 7(d) and 7(e)) NDVIalso increased (Figure 7(f)) which may have contributedto an increase in transpiration These results indicate that


1982 1986 1990 1994 1998 2002 2006 2010

y = minus09503x + 44248

R2 = 00204

200

300

400

500

600A

nnua

l pre

cipi

tatio

n (m

m)

(a)1982 1986 1990 1994 1998 2002 2006 2010

y = minus01118x + 65133

R2 = 01508

52

56

60

64

68

72

RH (

)

(b)

1982 1986 1990 1994 1998 2002 2006 20105

55

6

65

7

75

8

Suns

hine

dur

atio

n (h

)

(c)

6

65

7

75

8

85

9

95

10

105

1982 1986 1990 1994 1998 2002 2006 2010

Air

tem

pera

ture

(∘C)

y = 00599x + 76372

R2 = 05211

(d)

1982 1986 1990 1994 1998 2002 2006 2010

y = 00148x + 1555

R2 = 06626

1

12

14

16

18

2

22

24

Win

d sp

eed

(ms

)

(e)1982 1986 1990 1994 1998 2002 2006 2010

y = 00007x + 02938

R2 = 03643

024

027

03

033

036

ND

VI

(f)

Figure 7 Trends of annual (a) precipitation (b) air humidity (c) sunshine duration (d) air temperature (e) wind speed and (f) NDVI overthe Loess Plateau


minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)

Local mean annual precipitation

Figure 8 Relationship between the climate tendency rate of ET andlocal mean annual precipitation

the declining trend in annual precipitation and air humiditycaused the decreasing trend in ET which outweighed theeffects of increasing temperature wind speed and NDVI

Increased potential ET [24 25] and decreased actualET imply that a complementary relationship between thetwo exists over the Loess Plateau Similar research ondifferent regions of China [17] showed that potential andactual ET were both controlled by available energy andboth showed decreasing trends in humid and semihumidregions while in arid and semiarid regions potential ETwas controlled by available energy and showed a decreasingtrend and actual ET was controlled by available water andshowed an increasing trend The Loess Plateau howevershows an opposite ET trend compared to the arid andsemiarid regions of China as a whole The increasing ETtrend in other arid and semiarid regions was attributed toincreased precipitation relative humidity and cloud cover[43] Our results showed that water availability representedby precipitation and air humidity was decreasing during thestudy period which caused the decreasing trend of actualET In the study area potential evapotranspiration (ET)is the upper limit of evapotranspiration when there is nowater limit and represents the evaporative demand Actualevapotranspiration is evapotranspiration mainly limited bythe water supply of the area Potential ET is higher than actualET it means that the ET demand is higher than the actual ETindicating that precipitation cannotmeet the ET demand andall precipitation (except run-off) transforms into ET resultingin a better correlation between actual ET and precipitation

34 Local ET Trends in Response to Mean Annual Precip-itation From the above analysis it is apparent that wateravailability is the dominant factor controlling ET over theLoess Plateau Therefore the different local ET trends maybe related to local water availability Figure 8 shows therelationship between the climate tendency rate of local ETand mean annual precipitation It can be seen that theclimate tendency rate of local ET was closely related tolocal mean annual precipitation The climate tendency rateof local ET changed from negative to positive when localmean annual precipitation increased This implies that ET

changed from being supply-limited to energy-limited due toincreased water availability The climate tendency rate wasnegative in supply-limited areas where local mean annualprecipitationwas less than 400mm becausewater supplywasdecreasing in these areas it could be negative or positive inthe transition zone where local mean annual precipitationwas within the range 400ndash600mm and it was positive inenergy-limited areas where local mean annual precipitationwas greater than 600mm because the ET potential wasincreasing and water availability was sufficient in these areasIn the semiarid region local mean annual precipitation wasless than 400mm so the local climate tendency was negativeand local ET showed a decreasing trend In most of the semi-humid region local mean annual precipitation was withinthe range 400ndash600mm so the local climate tendency waseither negative or positive and local ET showed a decreasingor increasing trend in different areas For the intersectionareas of the Yellow River Luo River and Wei River localmean annual precipitationwas greater than 600mm the localclimate tendency rate was positive and the local ET showedan increasing trend

The strong solar radiation of the Loess Plateau meansrelatively high levels of available energy and therefore theonly limit for ET is the availability of water An increase inwater resources leads to changes in the climate tendency rateof ET from negative to positive and this causes two opposingfeedback mechanisms in the region In areas with little pre-cipitation (less than 400mm) the decreasing ET contributesless moisture to the atmosphere and decreases precipitationlocally further intensifying aridity In areas with large quan-tities of precipitation (greater than 600mm) the increas-ing ET contributes more moisture to the atmosphere andincreases precipitation locally further intensifying humidityThese mechanisms are quite different from those in humidregions where ET increases under low-precipitation condi-tions and decreases under high-precipitation conditions dueto available energy being reduced under high-precipitationconditions [44]

4 Conclusion

Based on CLM-simulated ET the trend of ET was analyzedover the Loess Plateau from 1982 to 2013 The domain-average ET decreased during the study period at a rate of078mmyearminus1 In particular the decreasing trend acceler-ated after 1998 with ET decreasing more from the 1990s to2000s compared to the 1980s to 1990s Stronger variation wasalso apparent for the 1990s leading to an easier facilitation ofdrought Apart from autumn all seasons showed a decreasingtrend with summer dominating the interannual ET trendRegionally most of the Loess Plateau showed a decreasingET trend with only a few small areas showing an increasingtrend (located in Tianshui District and the intersectionareas of the Yellow River Luo River and Wei River) ETdecreased more rapidly in the semiarid region than in thesemihumid region The declining trend of ET was attributedto declining precipitation and air humidity over the LoessPlateau during the study period Local ET trends were closely


related to local mean annual precipitation Areas with meanannual precipitation less than 400mm showed decreasingET trends areas with mean annual precipitation larger than600mm showed increasing ET trends and areas with meanannual precipitation within the range 400ndash600mm weretransitional Overall the decreasing trend of ET indicateslocal water cycling weakened over the Loess Plateau duringthe study period causing an energy partitioning preferencetoward sensible heat rather than latent heat which accel-erated surface warming and intensified land-atmosphereinteractions

Competing Interests

The authors declare that they have no competing interests

Acknowledgments

The CLM-simulated ET data were provided by GLDAS Thiswork was jointly supported by the Major National BasicResearch Program of China 2013CB430200 (2013CB430206)the National Science Foundation of China (31300376)and the Postdoctoral Science Foundation of China(2012M512044)

References

[1] K Wang and R E Dickinson ldquoA review of global terrestrialevapotranspiration observation modeling climatology andclimatic variabilityrdquo Reviews of Geophysics vol 50 no 2 ArticleID RG2005 2012

[2] Y Huang D Jiang D Zhuang Y Zhu and J Fu ldquoAn improvedapproach for modeling spatial distribution of water use profitmdasha case study in Tuhai Majia Basin Chinardquo Ecological Indicatorsvol 36 pp 94ndash99 2014

[3] K E Trenberth J T Fasullo and J Kiehl ldquoEarthrsquos global energybudgetrdquo Bulletin of the AmericanMeteorological Society vol 90no 3 pp 311ndash323 2009

[4] P J Durack S E Wijffels and R J Matear ldquoOcean salinitiesreveal strong global water cycle intensification during 1950 to2000rdquo Science vol 336 no 6080 pp 455ndash458 2012

[5] S Cohen A Ianetz and G Stanhill ldquoEvaporative climatechanges at BetDagan Israel 1964ndash1998rdquoAgricultural and ForestMeteorology vol 111 no 2 pp 83ndash91 2002

[6] V D P R Da Silva ldquoOn climate variability in Northeast ofBrazilrdquo Journal of Arid Environments vol 58 no 4 pp 575ndash5962004

[7] D G C Kirono R N Jones and H A Cleugh ldquoPan-evapo-ration measurements and Morton-point potential evaporationestimates in Australia are their trends the samerdquo InternationalJournal of Climatology vol 29 no 5 pp 711ndash718 2009

[8] J Q Xu S Haginoya K Saito and K Motoya ldquoSurface heatbalance and pan evaporation trends in Eastern Asia in theperiod 1971ndash2000rdquo Hydrological Processes vol 19 no 11 pp2161ndash2186 2005

[9] M L Roderick andG D Farquhar ldquoThe cause of decreased panevaporation over the past 50 yearsrdquo Science vol 298 no 5597pp 1410ndash1411 2002

[10] M T Hobbins J A Ramırez and T C Brown ldquoTrends inpan evaporation and actual evapotranspiration across the con-terminous US paradoxical or complementaryrdquo GeophysicalResearch Letters vol 31 no 13 Article ID L13503 2004

[11] D H Burn and N M Hesch ldquoTrends in evaporation for theCanadian Prairiesrdquo Journal of Hydrology vol 336 no 1-2 pp61ndash73 2007

[12] G Papaioannou G Kitsara and N Melanitis ldquoTemporalanalysis of hydrometeorological characteristics in Creterdquo inProceedings of the Symposium onWater Resources ManagementNew Approaches and Technologies (EWRA rsquo07) Chania Greece2007

[13] N Chattopadhyay and M Hulme ldquoEvaporation and potentialevapotranspiration in India under conditions of recent andfuture climate changerdquoAgricultural and Forest Meteorology vol87 no 1 pp 55ndash73 1997

[14] T Tebakari J Yoshitani and C Suvanpimol ldquoTime-space trendanalysis in pan evaporation over Kingdom ofThailandrdquo Journalof Hydrologic Engineering vol 10 no 3 pp 205ndash215 2005

[15] J Asanuma and H Kamimera ldquoLong-term trend of pan evapo-rationmeasurements in Japan and its relevance to the variabilityof the hydro-logical cyclerdquo in Proceedings of the Symposiumon Water Resource and its Variability in Asia MeteorologicalResearch Institute Tsukuba Japan March 2004

[16] H-C Zuo Y Bao C-J Zhang and Y-Q Hu ldquoAn analytic andnumerical study on the physical meaning of pan evaporationand its trend in recent 40 yearsrdquo Chinese Journal of Geophysicsvol 49 no 3 pp 680ndash688 2006

[17] B Liu Z N Xiao and Z G Ma ldquoRelationship between panevaporation and actual evaporation in different humid and aridregions of Chinardquo Plateau Meteorology vol 29 no 3 pp 629ndash636 2010

[18] M L Roderick and G D Farquhar ldquoChanges in Australianpan evaporation from 1970 to 2002rdquo International Journal ofClimatology vol 24 no 9 pp 1077ndash1090 2004

[19] M L Roderick and G D Farquhar ldquoChanges in New Zealandpan evaporation since the 1970srdquo International Journal ofClimatology vol 25 no 15 pp 2031ndash2039 2005

[20] W Abtew J Obeysekera and N Iricanin ldquoPan evaporationand potential evapotranspiration trends in South FloridardquoHydrological Processes vol 25 no 6 pp 958ndash969 2011

[21] X F Qiu Y Zeng and Q L Miu ldquoCalculating land surfaceactual evapotranspiration using Conventional meteorologicaldatardquo Science in China Series D-Earth Sciences vol 33 no 3pp 281ndash288 2004

[22] A J Teuling M Hirschi A Ohmura et al ldquoA regional perspec-tive on trends in continental evaporationrdquoGeophysical ResearchLetters vol 36 no 2 Article ID L02404 2009

[23] Y J Wang B Liu and J Q Zhai ldquoRelationship between poten-tial and actual evaporation in Yangtze River Basinrdquo Advances inClimate Change Research vol 7 no 6 pp 393ndash399 2011

[24] X Q Xie and L Wang ldquoChanges of potential evaporationin Northern China over the past 50 yearsrdquo Journal of NatureResources vol 22 no 5 pp 683ndash691 2007

[25] Z Li ldquoSpatiotemporal variations in the reference crop evap-otranspiration on the Loess Plateau during 1961ndash2009rdquo ActaEcologica Sinica vol 32 no 13 pp 4139ndash4145 2012

[26] S Lin and Y RWang ldquoSpatial-temporal evolution of precipita-tion in China loess plateaurdquo Journal of Desert Research vol 27no 3 pp 502ndash508 2007


[27] Q Zhang J Huang L Zhang and L Y Zhang ldquoWarming anddrying climate over Loess plateau area in China and its effect onland surface energy exchangerdquo Acta Physica Sinica vol 62 no13 Article ID 019202 2013

[28] W Wang X J Wang and P Wang ldquoAssessing the applicabilityof GLDAS monthly precipitation data in Chinardquo Advances inWater Science vol 25 no 6 pp 769ndash778 2014

[29] J Y Zheng Y H Yin and B Y Li ldquoA new scheme for climateregionalization in Chinardquo Acta Geographica Sinica vol 65 no1 pp 3ndash13 2010

[30] K W Oleson Y Dai G Bonan M Bosilovich R Dickson andP Dirmeyer Technical Description of the Community LandModel (CLM) 2004

[31] SM Liu ZW Xu Z L Zhu Z Z Jia andM J Zhu ldquoMeasure-ments of evapotranspiration from eddy-covariance systems andlarge aperture scintillometers in the Hai River Basin ChinardquoJournal of Hydrology vol 487 pp 24ndash38 2013

[32] E Falge D Baldocchi R Olson et al ldquoGap filling strategiesfor long term energy flux data setsrdquo Agricultural and ForestMeteorology vol 107 no 1 pp 71ndash77 2001

[33] S M Liu Z W Xu W Z Wang et al ldquoA comparison of eddy-covariance and large aperture scintillometer measurementswith respect to the energy balance closure problemrdquoHydrologyand Earth System Sciences vol 15 no 4 pp 1291ndash1306 2011

[34] T Foken ldquoThe energy balance closure problem an overviewrdquoEcological Applications vol 18 no 6 pp 1351ndash1367 2008

[35] M Mauder T Foken R Clement et al ldquoQuality control ofCarboEurope flux datamdashpart 2 inter-comparison of eddy-covariance softwarerdquo Biogeosciences vol 5 no 2 pp 451ndash4622008

[36] T Foken R Leuning S R Oncley MMauder andM AubinetldquoCorrections and data quality controlrdquo in Eddy Covariance MAubinet T Vesala and D Papale Eds pp 85ndash131 SpringerBerlin Germany 2012

[37] J Cuxart L Conangla and M A Jimenez ldquoEvaluation of thesurface energy budget equation with experimental data andthe ECMWF model in the Ebro Valleyrdquo Journal of GeophysicalResearch D Atmospheres vol 120 no 3 pp 1008ndash1022 2015

[38] J Cuxart B Wrenger D Martınez-Villagrasa et al ldquoEstimationof the advection effects induced by surface heterogeneities inthe surface energy budgetrdquo Atmospheric Chemistry and PhysicsDiscussions vol 2016 pp 1ndash16 2016

[39] Z Gao H Liu E S Russell J Huang T Foken and S P OncleyldquoLarge eddies modulating flux convergence and divergencein a disturbed unstable atmospheric surface layerrdquo Journal ofGeophysical Research Atmospheres vol 121 no 4 pp 1475ndash14922016

[40] Q P Liu Y C Yang H Z Tian B Zhang and L Gu ldquoAssess-ment of human impacts on vegetation in built-up areas inChinabased on AVHRR MODIS and DMSP OLS nighttime lightdata 1992ndash2010rdquo Chinese Geographical Science vol 24 no 2pp 231ndash244 2014

[41] M Jung M Reichstein P Ciais et al ldquoRecent decline in theglobal land evapotranspiration trend due to limited moisturesupplyrdquo Nature vol 467 no 7318 pp 951ndash954 2010

[42] D Li ldquoAssessing the impact of interannual variability of pre-cipitation and potential evaporation on evapotranspirationrdquoAdvances in Water Resources vol 70 pp 1ndash11 2014

[43] B Liu ZMa J Feng andRWei ldquoThe relationship between panevaporation and actual evapotranspiration in Xinjiang since1960rdquo Acta Geographica Sinica vol 63 no 11 pp 1131ndash11392008

[44] Y Song Y Ryu and S Jeon ldquoInterannual variability of regionalevapotranspiration under precipitation extremes a case studyof the Youngsan River basin in Koreardquo Journal of Hydrology vol519 pp 3531ndash3540 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

ClimatologyJournal of

EcologyInternational Journal of


EarthquakesJournal of


Hindawi Publishing Corporationhttpwwwhindawicom

Applied ampEnvironmentalSoil Science

Volume 2014

Mining


Journal of

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal of

Geophysics

OceanographyInternational Journal of


Journal of Computational Environmental SciencesHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering


GeochemistryHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Atmospheric SciencesInternational Journal of


OceanographyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in


MineralogyInternational Journal of


MeteorologyAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Paleontology JournalHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Geological ResearchJournal of


Geology Advances in


55∘N

45∘N

35∘N

25∘N

15∘N75∘E 85∘E 95∘E 105∘E 115∘E 125∘E 135∘E

Humid Arid HyperaridSemiaridSemihumid

(a)

40∘N

38∘N

36∘N

34∘N

102∘E 105∘E 108∘E 111∘E 114∘E

0 200

(km)

N

Provincial boundaryRiversPlainHill

DuneTable landMountain landStation

(b)

Figure 1 The (a) location and climate of the Loess Plateau (black line) and (b) its land cover and the locations of the ET observatories (redstars)

in different regions [20 21] Actual ET is closely related topotential ET or pan evaporation but the implication of thetrend in pan evaporation for the trend in actual ET has longbeen disputed Some scientists argue that the declining trendin pan evaporation indicates a declining trend in actual ETOthers suggest that a complementary relationship betweenpotential and actual ET exists with declining ET meaning anincrease in actual ET Recent studies have shown that thesetwo relationships are correct under certain conditions [17 2223] Apparently the positive relationship is satisfied underconditions of sufficient water while the complementaryrelationship is satisfied under water-limited conditions

Potential ET is decreasing in most regions of Chinahowever pan evaporation and potential ET are increasing inthe Loess Plateau [24 25] Since the trend of pan evaporationand potential ET only provides us with a clue for the possibletrend of actual ET and because potential ET and pan evap-oration have a special trend in the Loess Plateau comparedto other regions of China our goal in the present studywas to probe the characteristics of actual ET by examiningwhether the trend in actual ET in the Loess Plateau alsoshows distinct regional features like potential ET Besidesbecause the Loess Plateau is situated in the boundary zonesinfluenced by the Asian summer monsoon precipitationdemonstrates considerable spatial variation [26] which inturn causes spatial variation in land surface hydrothermalprocesses It is therefore also of interest to investigate therelationship between ET trends and precipitation on the localscale

For a long time research on actual ET has been limitedby a lack of reliable ET data [1] In recent years however theGlobal Land Data Assimilation System (GLDAS) has becomean important dataset for global change and water cycleresearch [27 28] Employing the Community Land Model(CLM) GLDAS produces land surface ET data on the globalscale and its land surface variables are reported to be highly

consistent with observations in the Loess Plateau [27] In thisstudy we first validated the ET products of GLDAS using ETdata observed by an eddy covariance (EC) system at four landsurface observatories in the Loess Plateau Then on the basisof GLDAS ET data we analyzed the trends of ET and theirreasons over the Loess Plateau In addition the relationshipbetween ET trends andmean annual precipitation locally wasalso analyzed

2 Study Area and Data

21 Study Area The Loess Plateaumdashlocated to the north ofthe Yin Mountains south of the Tsinling Mountains east ofthe Tai-hang Mountains and west of the Riyue Mountainsmdashis a highly unique land type with special ecosystems Theaverage altitude is approximately 1000m above sea levelincreasing from southeast to northwest The most commontopography is gully and hilly terrain and the specific geo-morphology is shown in Figure 1The Loess Plateau is locatedin the boundary zone of the Asian summer monsoon It is atransitional region from the humid climate in the southeastunder the influence of the summer monsoon to the aridclimate in the northwest (Figure 1 rectangle) and thus canbe divided into two parts that is the semiarid zone in thenorthwest and the semihumid zone in the southeast [29] (Fig-ure 1) The annual average rainfall is approximately 466mmwith 700mm in the southeastern regions and 120mm in thenorthwestern regions The mean annual temperature rangesfrom4∘C to 12∘CThemain soil type is loess which has aweakability to resist wind and soil erosion

22 Data and Data Processing

221 Simulated ET Data The simulated ET data were down-loaded from GLDAS As the main dynamic framework theland surface model is the key part of GLDAS The forcing






119864 = 119864V + 119864119892

119864V = minus120588atm(119902119904

minus 119902sat119879V)

119903total

119864119892

= minus120588atm(119902119904

minus 119902119892

)

1199031015840

aw

1

119903total=119871sun

119903119887

+ 119903sun119904

+119871sha

119903119887

+ 119903sha119904

119903119887

=1

119862V(119880lowast

119889leaf)

minus12

1199031015840

aw =1

119862119904

119906lowast

1

119903119904

= 119898119860

119888119904

119890119904

119890119894

119875atm + 119887

(1)

where 119902119904





119887



119904


119904

and 119903sha119904


119904

is theCO2



119894



LE = 12058812058212059610158401199021015840 (2)



119894


ge (+4120575) or119909119894






119899



119899


119899











0 20 40 60 80 100Measured ET (mm)

SACOL site

R2 = 08782

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Dingxi site

R2 = 07655

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Qingyang site

R2 = 0933

0

20

40

60

80

100ET

sim

ulat

ed b

y CL

M (m

m)

20 40 60 80 1000Measured ET (mm)

Pingliang site

R2 = 09049

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)


200

250

300

350

1982 1986 1990 1994 1998 2002 2006 2010Dom

ain-

aver

age a

nnua

l ET

(mm

)

(a)

03

05

07

09

1982 1986 1990 1994 1998 2002 2006 2010

Evap

orat

ion

ratio

(E

Tpr

ecip

itatio

n)

(b)





0

40

80

120

160

200

1982 1986 1990 1994 1998 2002 2006 2010

ET in

each

seas

on (m

m)






425∘N

395∘N

365∘N

335∘N101∘E 105∘E 109∘E 113∘E




100

200

300

400

500

1982 1986 1990 1994 1998 2002 2006 2010

ET (m

m)

SemiaridSemihumid





1982 1986 1990 1994 1998 2002 2006 2010

y = minus09503x + 44248

R2 = 00204

200

300

400

500

600A

nnua

l pre

cipi

tatio

n (m

m)

(a)1982 1986 1990 1994 1998 2002 2006 2010

y = minus01118x + 65133

R2 = 01508

52

56

60

64

68

72

RH (

)

(b)

1982 1986 1990 1994 1998 2002 2006 20105

55

6

65

7

75

8

Suns

hine

dur

atio

n (h

)

(c)

6

65

7

75

8

85

9

95

10

105

1982 1986 1990 1994 1998 2002 2006 2010

Air

tem

pera

ture

(∘C)

y = 00599x + 76372

R2 = 05211

(d)

1982 1986 1990 1994 1998 2002 2006 2010

y = 00148x + 1555

R2 = 06626

1

12

14

16

18

2

22

24

Win

d sp

eed

(ms

)

(e)1982 1986 1990 1994 1998 2002 2006 2010

y = 00007x + 02938

R2 = 03643

024

027

03

033

036

ND

VI

(f)



minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)








4 Conclusion




Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in






119864 = 119864V + 119864119892

119864V = minus120588atm(119902119904

minus 119902sat119879V)

119903total

119864119892

= minus120588atm(119902119904

minus 119902119892

)

1199031015840

aw

1

119903total=119871sun

119903119887

+ 119903sun119904

+119871sha

119903119887

+ 119903sha119904

119903119887

=1

119862V(119880lowast

119889leaf)

minus12

1199031015840

aw =1

119862119904

119906lowast

1

119903119904

= 119898119860

119888119904

119890119904

119890119894

119875atm + 119887

(1)

where 119902119904





119887



119904


119904

and 119903sha119904


119904

is theCO2



119894



LE = 12058812058212059610158401199021015840 (2)



119894


ge (+4120575) or119909119894






119899



119899


119899











0 20 40 60 80 100Measured ET (mm)

SACOL site

R2 = 08782

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Dingxi site

R2 = 07655

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Qingyang site

R2 = 0933

0

20

40

60

80

100ET

sim

ulat

ed b

y CL

M (m

m)

20 40 60 80 1000Measured ET (mm)

Pingliang site

R2 = 09049

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)


200

250

300

350

1982 1986 1990 1994 1998 2002 2006 2010Dom

ain-

aver

age a

nnua

l ET

(mm

)

(a)

03

05

07

09

1982 1986 1990 1994 1998 2002 2006 2010

Evap

orat

ion

ratio

(E

Tpr

ecip

itatio

n)

(b)





0

40

80

120

160

200

1982 1986 1990 1994 1998 2002 2006 2010

ET in

each

seas

on (m

m)






425∘N

395∘N

365∘N

335∘N101∘E 105∘E 109∘E 113∘E




100

200

300

400

500

1982 1986 1990 1994 1998 2002 2006 2010

ET (m

m)

SemiaridSemihumid





1982 1986 1990 1994 1998 2002 2006 2010

y = minus09503x + 44248

R2 = 00204

200

300

400

500

600A

nnua

l pre

cipi

tatio

n (m

m)

(a)1982 1986 1990 1994 1998 2002 2006 2010

y = minus01118x + 65133

R2 = 01508

52

56

60

64

68

72

RH (

)

(b)

1982 1986 1990 1994 1998 2002 2006 20105

55

6

65

7

75

8

Suns

hine

dur

atio

n (h

)

(c)

6

65

7

75

8

85

9

95

10

105

1982 1986 1990 1994 1998 2002 2006 2010

Air

tem

pera

ture

(∘C)

y = 00599x + 76372

R2 = 05211

(d)

1982 1986 1990 1994 1998 2002 2006 2010

y = 00148x + 1555

R2 = 06626

1

12

14

16

18

2

22

24

Win

d sp

eed

(ms

)

(e)1982 1986 1990 1994 1998 2002 2006 2010

y = 00007x + 02938

R2 = 03643

024

027

03

033

036

ND

VI

(f)



minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)








4 Conclusion




Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in





119899



119899


119899











0 20 40 60 80 100Measured ET (mm)

SACOL site

R2 = 08782

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Dingxi site

R2 = 07655

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Qingyang site

R2 = 0933

0

20

40

60

80

100ET

sim

ulat

ed b

y CL

M (m

m)

20 40 60 80 1000Measured ET (mm)

Pingliang site

R2 = 09049

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)


200

250

300

350

1982 1986 1990 1994 1998 2002 2006 2010Dom

ain-

aver

age a

nnua

l ET

(mm

)

(a)

03

05

07

09

1982 1986 1990 1994 1998 2002 2006 2010

Evap

orat

ion

ratio

(E

Tpr

ecip

itatio

n)

(b)





0

40

80

120

160

200

1982 1986 1990 1994 1998 2002 2006 2010

ET in

each

seas

on (m

m)






425∘N

395∘N

365∘N

335∘N101∘E 105∘E 109∘E 113∘E




100

200

300

400

500

1982 1986 1990 1994 1998 2002 2006 2010

ET (m

m)

SemiaridSemihumid





1982 1986 1990 1994 1998 2002 2006 2010

y = minus09503x + 44248

R2 = 00204

200

300

400

500

600A

nnua

l pre

cipi

tatio

n (m

m)

(a)1982 1986 1990 1994 1998 2002 2006 2010

y = minus01118x + 65133

R2 = 01508

52

56

60

64

68

72

RH (

)

(b)

1982 1986 1990 1994 1998 2002 2006 20105

55

6

65

7

75

8

Suns

hine

dur

atio

n (h

)

(c)

6

65

7

75

8

85

9

95

10

105

1982 1986 1990 1994 1998 2002 2006 2010

Air

tem

pera

ture

(∘C)

y = 00599x + 76372

R2 = 05211

(d)

1982 1986 1990 1994 1998 2002 2006 2010

y = 00148x + 1555

R2 = 06626

1

12

14

16

18

2

22

24

Win

d sp

eed

(ms

)

(e)1982 1986 1990 1994 1998 2002 2006 2010

y = 00007x + 02938

R2 = 03643

024

027

03

033

036

ND

VI

(f)



minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)








4 Conclusion




Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in


0 20 40 60 80 100Measured ET (mm)

SACOL site

R2 = 08782

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Dingxi site

R2 = 07655

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)

Qingyang site

R2 = 0933

0

20

40

60

80

100ET

sim

ulat

ed b

y CL

M (m

m)

20 40 60 80 1000Measured ET (mm)

Pingliang site

R2 = 09049

20 40 60 80 1000Measured ET (mm)

0

20

40

60

80

100

ET si

mul

ated

by

CLM

(mm

)


200

250

300

350

1982 1986 1990 1994 1998 2002 2006 2010Dom

ain-

aver

age a

nnua

l ET

(mm

)

(a)

03

05

07

09

1982 1986 1990 1994 1998 2002 2006 2010

Evap

orat

ion

ratio

(E

Tpr

ecip

itatio

n)

(b)





0

40

80

120

160

200

1982 1986 1990 1994 1998 2002 2006 2010

ET in

each

seas

on (m

m)






425∘N

395∘N

365∘N

335∘N101∘E 105∘E 109∘E 113∘E




100

200

300

400

500

1982 1986 1990 1994 1998 2002 2006 2010

ET (m

m)

SemiaridSemihumid





1982 1986 1990 1994 1998 2002 2006 2010

y = minus09503x + 44248

R2 = 00204

200

300

400

500

600A

nnua

l pre

cipi

tatio

n (m

m)

(a)1982 1986 1990 1994 1998 2002 2006 2010

y = minus01118x + 65133

R2 = 01508

52

56

60

64

68

72

RH (

)

(b)

1982 1986 1990 1994 1998 2002 2006 20105

55

6

65

7

75

8

Suns

hine

dur

atio

n (h

)

(c)

6

65

7

75

8

85

9

95

10

105

1982 1986 1990 1994 1998 2002 2006 2010

Air

tem

pera

ture

(∘C)

y = 00599x + 76372

R2 = 05211

(d)

1982 1986 1990 1994 1998 2002 2006 2010

y = 00148x + 1555

R2 = 06626

1

12

14

16

18

2

22

24

Win

d sp

eed

(ms

)

(e)1982 1986 1990 1994 1998 2002 2006 2010

y = 00007x + 02938

R2 = 03643

024

027

03

033

036

ND

VI

(f)



minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)








4 Conclusion




Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in




0

40

80

120

160

200

1982 1986 1990 1994 1998 2002 2006 2010

ET in

each

seas

on (m

m)






425∘N

395∘N

365∘N

335∘N101∘E 105∘E 109∘E 113∘E




100

200

300

400

500

1982 1986 1990 1994 1998 2002 2006 2010

ET (m

m)

SemiaridSemihumid





1982 1986 1990 1994 1998 2002 2006 2010

y = minus09503x + 44248

R2 = 00204

200

300

400

500

600A

nnua

l pre

cipi

tatio

n (m

m)

(a)1982 1986 1990 1994 1998 2002 2006 2010

y = minus01118x + 65133

R2 = 01508

52

56

60

64

68

72

RH (

)

(b)

1982 1986 1990 1994 1998 2002 2006 20105

55

6

65

7

75

8

Suns

hine

dur

atio

n (h

)

(c)

6

65

7

75

8

85

9

95

10

105

1982 1986 1990 1994 1998 2002 2006 2010

Air

tem

pera

ture

(∘C)

y = 00599x + 76372

R2 = 05211

(d)

1982 1986 1990 1994 1998 2002 2006 2010

y = 00148x + 1555

R2 = 06626

1

12

14

16

18

2

22

24

Win

d sp

eed

(ms

)

(e)1982 1986 1990 1994 1998 2002 2006 2010

y = 00007x + 02938

R2 = 03643

024

027

03

033

036

ND

VI

(f)



minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)








4 Conclusion




Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in


1982 1986 1990 1994 1998 2002 2006 2010

y = minus09503x + 44248

R2 = 00204

200

300

400

500

600A

nnua

l pre

cipi

tatio

n (m

m)

(a)1982 1986 1990 1994 1998 2002 2006 2010

y = minus01118x + 65133

R2 = 01508

52

56

60

64

68

72

RH (

)

(b)

1982 1986 1990 1994 1998 2002 2006 20105

55

6

65

7

75

8

Suns

hine

dur

atio

n (h

)

(c)

6

65

7

75

8

85

9

95

10

105

1982 1986 1990 1994 1998 2002 2006 2010

Air

tem

pera

ture

(∘C)

y = 00599x + 76372

R2 = 05211

(d)

1982 1986 1990 1994 1998 2002 2006 2010

y = 00148x + 1555

R2 = 06626

1

12

14

16

18

2

22

24

Win

d sp

eed

(ms

)

(e)1982 1986 1990 1994 1998 2002 2006 2010

y = 00007x + 02938

R2 = 03643

024

027

03

033

036

ND

VI

(f)



minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)








4 Conclusion




Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in


minus4

minus3

minus2

minus1

0

1

2

3

4

0 100 200 300 400 500 600 700 800Clim

atic

tend

ency

rate

of E

T (m

my

ear)








4 Conclusion




Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in



Competing Interests


Acknowledgments


References























































Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in





























Volume 2014

Mining


Journal of



Geophysics







Journal of




Advances in











Geology Advances in

research article evapotranspiration trend and its...

Documents