Vol.2, No.7, 681-687 (2010) Natural Science
http://dx.doi.org/10.4236/ns.2010.27084
Impacts of oasis on the atmospheric hydrological cycle
over the nearby desert
Qiang Zhang1,2,3*, Yuhe Nan2, Sheng Wang1,3
1Key Laboratory of Arid Climatic Changes and Disaster Reduction of Gansu Province, Institute of Arid Meteorology, China
Meteorological Administration, Lan Zhou, China; *Corresponding Author:
2College of Atmospheric Science, Lan Zhou University, Lan Zhou, China
3Meteorological Bureau of Gansu Province, Lan Zhou, China
Received 22 February 2010; revised 23 April 2010; accepted 28 April 2010.
ABSTRACT
Using the data of “A field experiment on land-
atmosphere interaction over arid region in
Northwest China” carried out in Dunhuang of
Gansu Province from May to June 2000; Char-
acteristics of the atmospheric humidity over
desert and Gobi near oasis in the Northwest
China Arid Region are analyzed. According to
the difference of the characteristics in different
wind directions, the impacts of oasis on at-
mospheric hydrological cycle over desert and
Gobi near it are revealed. The relation of at-
mosphere inverse humidity and negative water
vapor flux to wind direction and atmospheric
stability is studied. It shows that distribution of
the atmosphere inverse humidity is inconsistent
with that of the negative water vapor flux; some-
times 1-hour-average value demonstrates the
characteristic of counter-gradient transfer. And
the diurnal variation of distribution of the coun-
ter-gradient transfer and the effect of atmos-
pheric stability on the counter-gradient transfer
are also given.
Keywords: Desert or Gobi; Hydrologic Cycle;
Atmospheric Inverse Humidity; Negative Water
Vapor Flux; Counter-Gradient Transfer
1. INTRODUCTION
Oasis is one of the underlying heterogeneous factors. It
directly affects the pattern of the atmospheric energy and
hydrology transfer over grid points in the region in a
model [1-3]. The interaction of the oasis system and the
desert system is relative to the maintenance and degen-
eration processes of oasis ecology system. The special
microclimate characteristics and the hydrological cycle
mechanism formed under the condition of the interaction
are two of the important natural factors (except the fac-
tor of water resource) which maintain the oasis system
[4-6].
Some special interaction processes between oasis and
desert have been found in some experiments and studies
carried out in the Northwest China Arid Region [4-12].
However, how the oasis impacts on the water vapor and
energy transfer over desert or Gobi near it is still un-
known. Most of the former field experiments were car-
ried out in the arid climate region where annual precipi-
tation is about 150 mm, but experiments in arider cli-
mate regions where annual precipitation is blow 50 mm,
such as Dunhuang, are scarce.
Through analysis and study the data of “The Dun-
huang Experiment”, the special physical rule of the land-
surface under the interaction between oasis and desert
near it as well as the ecological maintenance mechanism
of oasis will be discussed in the following passages.
2. OBSERVATION DATA AND
ANALYTICAL METHOD
The data analyzed in this paper is from a 20-day inten-
sive observation experiment in Dunhuang Gobi micro-
meteorological central station from May 25 to June 17,
2000 (briefly, “The Dunhuang Experiment”), which is a
part of “Land-atmosphere Interactive Field Experiment
over Arid Region of Northwest China”. The station is
located at 40°10′ N, 94°31′ E. Its surface elevation is
1150 meters above sea level. It lies in the flat Shuang-
dunzi Gobi west to Dunhuang oasis, about 7 km to the
edge of the oasis. Annual precipitation is about 40 mm
and annual potential evaporation is about 3400 mm. Its
main wind direction is east, generally occurring with the
frequency of more than 50 percent, so the oasis strongly
affects the atmosphere over the observation field. The
data include the gradients of wind, temperature and hu-

Q. Zhang et al. / Natural Science 2 (2010) 681-687
682
midity on a tower, soil heat fluxes, the fluctuations of
wind, temperature, humidity and wind observed by
using a tethered balloon within hundreds of meters high.
The sensors of wind, temperature and humidity on the
tower are installed at 18, 8, 4 and 2 m high respectively.
The sensor of wind direction is installed at 10 m high.
The sensors of supersonic fluctuation instrument were
installed at 2.9 m with a data processing system that
can give the hourly-average momentum flux, sensible
heat flux, and latent heat fluxes averaged. The details
about the data and the station in Dunhuang experiment
have been described in the paper [5,9]. The accuracy of
the instruments refers to the correlated references [13-
14].
During the Dunhuang experiment, there are 445 valid
sets of data which are hourly averaged and 10-day’s data
to show full diurnal variation. So the amount of samples
of hourly data is 445 and that of diurnal variation is 10.
Among the ten days, one day is affected evidently by the
precipitation. In order to make the results universal, the
data which is evidently affected by the precipitation is
picked out. The turbulent sensible and latent heat flux is
computed by the supersonic fluctuation data.
3. ATMOSPHERIC INVERSE HUMIDITY
STRUCTURE AND WATER VAPOR
FLUX
In Figure 1，the characteristics of 9-day average daily
variation of atmospheric specific humidity at four levels
(a) and latent heat flux at 2.9 m high (b) are given in
May-June over Dunhuang Gobi. As seen in Figure 1,
because of the influence of the oasis, the atmospheric
specific humidity over Gobi is inverse humidity from
0:00 to 6:00. In the daytime, although the specific hu-
midity mainly decreases with height, the inverse humid-
ity appears at 2-8 m high after 14:00 in the afternoon.
Figure 1(b) shows the negative water vapor flux domi-
nates at night and the water vapor transfers up in the
daytime. From its overall characteristics, the diurnal
variation of the specific humidity can be divided into
four stages that are respectively called the wet stage, the
losing-water stage, the dry stage and the attainting-water
stage [15].
The daily integration of latent heat flux is 0.117
MJ/m2 and that of sensible heat flux is 8.692 MJ/m2.
Difference between them is in two orders of magnitude.
The climatic Bowen ratio reachs 74.5 which is an ex-
treme arid climate index.
In fact, the impact of the oasis on desert near it is dy-
namic, which is quite different under different types of
synoptic situation. Especially, the wind direction is es-
sential to the impact. The representative data on two full
days are selected to analyze. One is 16 June on which
(a)
2.5
3
3.5
4
4.5
5
5.5
0 24 6 810121416182022
Local time/hour
atmosphere
humidity /g.kg-1
1m 2m
8m 18
m
(a)
(b)
Figure 1. Daily variation of atmospheric specific humidity at
four levels. (a) and latent heat flux at 2.9 m; (b) average data
from May 27 to June 15 over Dunhuang Gobi.
wind is from the oasis (named oasis wind), another is 14
June on which wind is from desert (named desert wind)
expect 4:00 and 5:00 in which wind direction can not be
clearly distinguished into oasis wind and desert wind.
Figure 2 shows the four stages of the wet stage (a), the
losing-water stage (b), the dry stage (c) and the attaint-
ing-water stage (d) in daily variation of the specific hu-
midity profile in the wind from oasis. It indicates that the
specific humidity profile is mainly inverse humidity all
day. In the wet stage, inverse humidity appears below 18
m high. In the losing-water stage Inverse humidity ap-
pears at about 2 m high and in 8-18 m and the humidity
diminishes obviously with time. In the dry stage, inverse
humidity appears only below 2 m, and inverse humidity
disappears in few stronger heating hours. In the water-
gaining stage, inverse humidity strengthens gradually
and moves up, at last, all is inverse humidity below 18m
and humidity increases obviously.
(b)
-10
-5
0
5
10
15
0246810121416182022
local time/hour
surface latent
heat flux/w.m-2
m-2

Q. Zhang et al. / Natural Science 2 (2010) 681-687
683
683
(a)
0
2
4
6
8
10
12
14
16
18
20
5 5.5 66.5 7 7.5
atmosphere humidity/g.kg
-1
height/m
0h
1h
2h
3h
4h
5h
(a) (b)
(c)
0
2
4
6
8
10
12
14
16
18
20
4.2 4.44.6 4.855.2
atmosphere humidity/g.kg
-1
height/m
12h
13h
14h
15h
16h
17h
18h
(d)
0
2
4
6
8
10
12
14
16
18
20
45678
atmosphere humidity/g.kg
-1
height/m
19h
20h
21h
22h
23h
(c) (d)
Figure 2. The four stages of the wet stage. (a) the losing-water stage; (b) and the dry stage; (c) the attainting-water stage; (d) in
daily variation of the specific humidity profile in the wind from oasis.
Compared with Figure 2, the daily variation of the
specific humidity profile over desert near oasis in the
four stages in the wind from desert (given in Figure 3) is
quite different. It shows the specific humidity decreases
progressively with height all day. In the wet stage, weak
inverse humidity appears mainly in 2 m in general, but
in all layer in transient wind direction. In the losing-
water stage, inverse humidity below 2 m diminishes gra-
dually and disappears or moves up to 2-8 m. In the dry
stage, inverse humidity in 2-8 m and air becomes moist
obviously. In the attainting-water stage, the profiles of
specific humidity decreases up and changes gradually
into constant humidity with time. Except 4:00 and 5:00,
there is no water vapor advection from oasis to Gobi, but
the weak inverse humidity still maintains in long time in
dry stage. The reasons are worthy to be discussed.
Figure 4 is the daily variation of the latent heat flux at
2.9 m over Gobi near the oasis in the wind from the oa-
sis (a) and from Gobi or desert (b). In the wind from the
oasis, positive latent heat flux is much smaller than that
in the wind from the desert ,and further, negative water
vapor flux appears mainly in the evening, which are own
to the effect of atmospheric inverse humidity. But in the
wind from the desert, not only is the latent heat flux
bigger, but also the water vapor still transfers generally
up in the evening.
The daily integration of the latent heat flux is –0.051
MJ/m2 in the wind from oasis and is 0.443 MJ/m2 in the
wind from desert. The contributions of the latent heat
flux to the surface heat balance in different kind of wind
(b)
0
2
4
6
8
10
12
14
16
18
20
5 5.56 6.57 7.5
atmosphere humidity/g. kg-1
height/m
6h
7h
8h
9h
10
h
10h
atmosphere humidity/g·kg-1

Q. Zhang et al. / Natural Science 2 (2010) 681-687
684
(a)
0
2
4
6
8
10
12
14
16
18
20
55.566.577.5
atmosphere humidity/g.kg
-1
height/m
0h
1h
2h
3h
4h
5h
(b)
0
2
4
6
8
10
12
14
16
18
20
5 5.5 6 6.577.5
atmosphere humidity/g. kg-1
height/m
6h
7h
8h
9h
10h
11h
atmos phere humidity/g·kg
-1
(a) (b)
(d)
0
2
4
6
8
10
12
14
16
18
20
45678
atmosphere humidity/g.kg
-1
height/m
19h
20h
21h
22h
23h
(c) (d)
Figure 3. The four stages of the wet stage. (a) the losing-water stage; (b) the dry stage; (c) the attainting-water stage; (d) in daily
variation of the specific humidity profile in the wind from desert.
directions are completely different. When the latent heat
flux is converted to the water vapor flux, its daily inte-
gration is –0.0155 mm in the wind from the oasis and is
0.1355 mm in the wind from desert. In the common
synoptic background annual evaporation over desert and
Gobi is bigger than the climatic mean value without the
infect of the oasis. This result means that it is the effect
of the oasis makes the climatic state and the vegtation
ecological type better in the desert region near the oasis
than that in dersert or Gobi. It can be called the clmatic
effect of inverse humidity. Dunhuang Oasis is smaller,
so maybe its climatic effect isn’t more evidient than that
of bigger or more prosperous oasis. The point isn’t found
in former relative studies on arid region.
-15
-10
-5
0
5
10
15
20
0246810 12 14 16 18 20 22
local time/hour
surface latent
flux/w.m
-2
desert wind
oasis wind
Figure 4. (a) Characteristics of daily variation of the latent
heat flux at 2.9 m over Gobi in the wind from oasis; (b) and
from Gobi or desert.
(c)
0
2
4
6
8
10
12
14
16
18
20
2 2.5 3 3.5
atmosphere humidity/g.kg-1
height/m
12h
13h
14h
15h
16h
17h
18h
atmosphere humidity/g·kg-1

Q. Zhang et al. / Natural Science 2 (2010) 681-687
685
685
4. DISTRIBUTION OF ATMOSPHERIC
INVERSE HUMIDITY AND NEGATIVE
WATER VAPOR FLUX AS WELL AS
COUNTER-GRADIENT TRANSFER
Figure 5 is a comparison of distribution of the wind
from the oasis and inverse humidity as well as negative
water vapor within 1-2 m over Gobi in Dunhuang. It
indicates that frequency of oasis wind keeps at about 50
percent all day, frequency of the atmosphere inverse
humidity is bigger at night than that in the daytime. Fre-
quency of inverse humidity is bigger than that of oasis
wind at night, but smaller at noon and in the afternoon.
There are two main reasons for this feature. On the one
hand, the turbulent mixing is very strong and demolishes
the inverse humidity structure in the daytime. Thus the
inverse humidity structure is hard to maintain. On the
other hand, surface evaporation in the daytime can partly
counteract the effect of the water vapor advection.
Therefore, if the water vapor advection isn’t strong
enough the inverse humidity will not appear in the day-
time. But at night the turbulent mixing is weaker,
evaporation is less and atmosphere is of generalized con-
servation (It means atmospheric motion and exchange is
inactive), and inertia of the atmosphere inverse humidity
is strong. Thus, the inverse humidity can still keep for a
long time, even if there isn’t the water vapor advection
from the oasis. These conclusions haven’t been drawn
from the observations in desert near Zhangye-Linze
prosperous oasis in Heihe region [2]. There are two main
reasons about it. On the one hand, the wind direction is
more irregular in the observation of Heihe region which
is affected by Qilian Mountain. On the other hand, per-
haps the water vapor transfer of Dunhuang oasis isn’t as
strong as Zhangye-Linze oasis’s.
Generally the direction of the water vapor flux is con-
trolled by the specific humidity gradient. For example, if
the specific humidity decreases with height the water
vapor flux will be positive, namely upwards, vice versa.
0
10
20
30
40
50
60
70
80
90
100
0 2 4 6 810121416182022
local time/h
probablity
inverse humidity
wind from oasis
ne
g
ative moist flux
Figure 5. A comparison of distribution of the wind from the
oasis, inverse humidity and negative water vapor within 1-2 m
over Gobi in Dunhuang.
Figure 5 shows that, the distribution of the negative wa-
ter vapor flux is consistent with that of the inverse spe-
cific humidity at night, but is quite inconsistent with it in
the daytime. Frequency of the negative water vapor flux
is under 20 percent and less than that of the inverse spe-
cific humidity in the daytime. It means that, the latent
heat bulk transfer coefficients must be positive, namely
counter-gradient transfer appears most times. It also ap-
pears over desert near oasis in Heihe region, but is omit-
ted as error in the former studies. The counter-gradient
transfer is partly related to the observational error be-
cause both the humidity and latent heat flux are too
small. But its distribution shows such good rule which
makes us to look for other reasons. Some studies show
that the counter-gradient transfer can result from the
horizontal heterogeneity which is caused by the motion
of convective eddy, complicated underlying and cumulus
convection [6,7]. The strong horizontal heterogeneity of
water vapor caused by distribution of oases in arid re-
gion is a typical mesoscale process. Based on Taylor
hypothesis, temporal averaged values can respond to the
effect of the spatial heterogeneity, and it results in oppo-
site symbol between the hourly averaged gradient and
the turbulent flux.
Figure 6 is a comparison of the characteristics of
daily variation of distribution of counter-gradient trans-
fer (a) and the bulk Richardson number (b). Figure 6(a)
indicates that the frequency of the counter-gradient
transfer over desert and Gobi near oasis in the daytime is
bigger than that at night. Its maximum is over 60 percent
at 9:00 o’ and the minimum is about 20 percent at 18:00.
It basically maintains about 40 percent in a longer time.
The daily variation of the counter-gradient transfer
may be related to the atmospheric stability. Comparing
Figure 6(a) with Figure 6(b), the daily variation of the
distribution of counter-gradient transfers and is mainly
correlative to that of the Richardson number. The bigger
the frequency of counter-gradient transfer is, the stronger
the atmospheric instability is, vice versa.
The atmospheric stability indirectly affects the cou-
nter-gradient transfer through its influence on the oasis
effect. Generally, the stronger the heterogeneity is, the
counter-gradient transfer caused by the temporal aver-
aging [16] is more obvious. At night the atmosphere over
oasis is stable, inactive and its impact on desert near it is
smaller, so the frequency of the counter-gradient transfer
is less. But in the daytime the great instability makes the
atmosphere over oasis active, and the influence of at-
mosphere over the oasis on the atmosphere over desert
near oasis is greater, so the frequency of the counter-
gradient transfer is bigger. After the surface atmosphere
exchanges strongly between oasis and desert during a
period in the daytime, horizontal distribution of atmos

Q. Zhang et al. / Natural Science 2 (2010) 681-687
686
(a)
0
10
20
30
40
50
60
70
80
90
100
1357911 13 1517192123
local time/hour
distribution of counter-
gradient moisture transfer/%
observed curve
Polynomial (observed curve)
(a)
(b)
-0.8
-0.4
0
0.4
0.8
1.2
0 246810121416182022
local time/hour
Riichardson number
(b)
Figure 6. (a) A comparison of the characteristics of daily varia-
tion of distribution of counter-gradient transfer; (b) and the
bulk Richard-son number.
pheric humidity tends to even and the impact of the oasis
on desert become weaker. So the counter-gradient trans-
fer of water vapor appears at 9:00, not at noon or in the
afternoon.
5. CONCLUSIONS AND DISCUSSIONS
The daily variation of the specific humidity over desert
near oasis can be divided into four stages which are the
wet stage, the losing-water stage, the dry stage and the
attaining-water stage. This is similar to the general Gobi
or desert. Differences caused by the oasis wind are given
in the following content. Firstly, the atmosphere specific
humidity over Gobi near oasis is basically in inverse
humidity and the water vapor flux is negative at night.
Secondly, the specific humidity decreases progressively
with height and the water vapor flux is positive in the
daytime. Thirdly, the difference between the daily inte-
grated values of the latent and sensible heat flux are in
two orders of magnitude.
The structure of humidity profiles and the water vapor
transfer in desert wind are very different from those in
oasis wind. In the wind from desert, the specific humid-
ity decreases with height all day and the water vapor
basically transfers up. However, in the wind from oasis,
the inverse humidity is dominant and the water vapor
flux is negative all day except few hours. The daily inte-
grated value of water vapor flux in the wind from oasis
is significantly different from that in the wind from de-
sert. Such effect of the oasis indicates that less annual
precipitation can support the climatic state or the vegta-
tion ecological type with greater annual evaporation over
desert region near the oasis. Namly, the effect of the oa-
sis makes the climatic state and the vegtation ecological
type better in the dersert region near the oasis than that
in the desert or Gobi.
The atmosphere over oasis is conservative at night and
active in the daytime. So the frequence of inverse hu-
midity is bigger than that of the oasis wind at night,
while is smaller in the daytime. Because of the influence
of horizontal spatial heterogeneity of water vapor caused
by distribution of oasis, sometimes the direction of water
vapor transfer is inconsistent with specific humidity gra-
dient, namely the counter-gradient transfer of water va-
por appears in a longer time.
The frequency of the counter-gradient transfer in the
daytime is bigger than that at night and it is much rela-
tive to the atmospheric stability. These features are rela-
tive to the counteraction of the turbulent mixing on the
horizontal heterogeneity of water vapor over oasis.
The effects of thermal circulation between oasis and
desert near it, which were found in the former studies
[12] on the Northwest China Arid Region, doesn’t ap-
pear in the Dunhuang experiment. The flowing are the
possible reasons. Firstly, Dunhuang oasis is smaller. Sec-
ondly, the thermal circulation is weaker. At last, its in-
fluence on atmospheic humidity over gobi or desert may
be offset by the strong turbulent mixing resulting from
extreme arid environment.
6. ACKNOWLEDGEMENTS
We thank Mr. Zeyong Hu, Xuhong Hou, Ping Hou and Yanjiang Nei et al.
for their assistance in the observation and data procession. We also
thank for the comments of the reviewer. This work was supported
by the National Natural Science Foundation of China (Grant No.
40830957) and the National Key Program for Basic Sciences – Re-
search on Formation Mechanism and Prediction Theory of Heavy
Climatic Disasters in China (Grant No.G1999040904-2).
REFERENCES
[1] Zhang, Q. and Huang, R.H. (2004) Water vapor exchange

Q. Zhang et al. / Natural Science 2 (2010) 681-687
687
687
between soil and atmosphere over a gobi surface near an
oasis in summer. Journal of Applied Meteorology, 43(12),
1917-1928.
[2] Zhang, Q. and Huang, R.H. (2004) Parameters of land
surface processes on Gobi in northwest China. Bound-
ary-Layer Meterology, 110(3), 471-478.
[3] Zeng, X. and Pielke, R.A. (1995) Landscape-induced
atmospheric flow and its parameterization in large-scale
numerical models. Journal of Climate, 8(5), 1156-1177.
[4] Zhang, Q., Hu, Y.Q., Cao, X.Y. and Liu, W.M. (2000)
About some problems of arid climate system of North-
west China. Journal of Desert Research (in Chinese),
20(4), 357-362.
[5] Zhang, Q. and Cao, X.Y. (2003) A study of surface heat
and radiation budget energy in Dunhuang region. Chi-
nese Journal of Atmospheric Sciences, 27(2), 165-176.
[6] Zhang, Q., Wang, S. and Sun, Z.X. (2009) A study of
Atmospheric Boundary layer structure during a clear day
in the Arid Region of Northwest China. ACTA Meteor
Sinica, 23(3), 327-337.
[7] Zhang, Q. and Hu, Y.Q. (1992) Characteristics of micro-
meteorology over the farmland in Oasis Heihe region.
Plateau Meteorite (in Chinese), 11(4 ), 361-370.
[8] Zhang, Q. and Zhao, M. (1999) Field experiment and
numerical simulation of inverse humidity of atmosphere
over desert near oasis. ACTA Meteor Sinica (in Chinese),
57(6), 729-740.
[9] Zhang, Q., Song, L.C. and Wei, G.A. (2003) Characteris-
tics of hydrologic transfer between soil and Atmosphere
over Gobi near oasis at the end of summer. Advance in
Atmospheric Science, 20(3), 442-452.
[10] Zhang, Q., Wang, S., Michael, B., Tian, W.S. and Huang,
R.H. (2010) The characteristics of sensible heat and mo-
mentum transfer coefficients over the Gobi in the northwest
China. International Journal of Climatology, 2, 586-589.
[11] Wang, J.M. and Mitsuta, Y. (1991) Turbulence structure
and transfer characteristics in the surface layer of the
HEIFE Gobi area. Journal of the Meteorology Society of
Japan, 69(5), 587-593.
[12] Hu, Y.Q., Yang, X.L., Zhang, Q. and Zuo, H.C. (1992)
The characters of energy budget on the Gobi and desert
surface in Hexi region. ACTA Meteor Sinica, 6(1), 82-91.
[13] Zhang, Q. and Hu, Y.Q. (1992) The instrumental accu-
racy and observational error about micrometeorological
mast of Chinese side in “HEIFE”. Plateau Meteorite (in
Chinese), 11(4), 460-469.
[14] Wang, J.M., et al. (1992) A real-time, low cost turbulence
data acquisition and processing system. Plateau Meteor-
ite (in Chinese), 11(4), 451-490.
[15] Zhang, Q., Wei, G.A. and Huang, R.H. (2002) The bulk
transfer coefficients of the atmospheric momentum and
sensible heat over desert and Gobi in arid climate region
of Northwest China. Science in China (Series D), 45(5),
468-480.
[16] Deardorf, J.W. (1972) Theoretical expression for the
counter gradient vertical heat flux. Journal of Geophysi-
cal Research, 77(30), 5900-5904.