analysis of an oasis microclimate in china’s hyperarid zone
TRANSCRIPT
ORIGINAL ARTICLE
Analysis of an oasis microclimate in China’s hyperarid zone
Qi Feng Æ Zhuo Macuo Æ Xi Haiyang
Received: 13 January 2008 / Accepted: 24 September 2008 / Published online: 14 October 2008
� Springer-Verlag 2008
Abstract The microclimate of a desert oasis in China’s
hyperarid zone was monitored, analysed and compared to
that of nearby forested lands. Factors associated with dif-
ferences in photosynthetically active radiation (PAR)
between clear, cloudy and dust storm days are discussed.
Desert oases were shown to fulfill ecological functions
such as altering solar radiation, adjusting near-ground and
land surface temperatures, reducing temperature differ-
ences, lowering wind velocity, and increasing soil and
atmospheric humidity. Total solar radiation within the oasis
was roughly half that above the forest canopy. During the
growing season, air temperatures in Populus euphratica
Olivier and Tamarix ramosissima Ledeb. woodlands were,
on average, 1.62 and 0.83�C lower, respectively, than that
in surrounding woodlands. The greater the forest cover, the
greater was the difference in temperature. Air temperature
was higher at the upper storey than that at the lower storey
of the community, i.e., air temperature increased with
increasing height above the soil surface. During the
growing season, relative humidity was higher in woodlands
than in surrounding areas: relative humidity in P. euphra-
tica and T. ramosissima woodlands were, on average, 8.5
and 4.2% higher, respectively, than that in the surrounding
area. Mean wind velocity in the P. euphratica forest land
was 0.33 m/s, 2.31 m/s lower than that in the surrounding
area. On dust storm days PAR and total radiation, Q, were
significantly lower than that on cloudy or clear days. Their
ratio, gQ = PAR/Q, was larger and much more variable on
dust storm days than that on clear or cloudy days.
Keywords Hyperarid zone � Desert oasis �Photosynthetically active radiation (PAR) � Microclimate
Introduction
Microclimate is an important environmental factor in the
growth and development of plants. Conversely, plant
communities create different microclimatic environments,
which support different biotic communities (Zhang et al.
2002). Such communities are continuously exchanging
materials (water) and energy (heat) with the surrounding
environment, which, in turn, dictates both current and
future exchanges. While a forest ecosystem’s micromete-
orological conditions can limit trees’ physiologic activities,
including their growth and development, as well as their
rates of photosynthesis and transpiration, the vegetation, a
reflection of local surface conditions, influences the near-
surface microclimatic features of water and energy through
transpiration and the interception and reflection of incom-
ing light. Thus, the vegetation is responsible for differences
in heat balance and water balance in the system. These
differences are also important to human life and produc-
tivity (Ye and Zhou 2000). Therefore, the study of
microclimate is essential in coordinating the relationships
that exist between living things and their environment.
Since the beginning of the twentieth century, the envi-
ronmental characteristic of various forest communities,
including their microclimate have been the focus of
research in China and abroad (Federer 1968; Garrett 1978;
Kimmins 1977; Hirose 1995). While much fruitful forest
ecosystem research has recently been devoted to China’s
Q. Feng (&) � Z. Macuo � X. Haiyang
Cold and Arid Regions Environmental and Engineering
Research Institute, Chinese Academy of Sciences,
No. 260 West Duong Gang Road, 730000 Lanzhou,
People’s Republic of China
e-mail: [email protected]
123
Environ Geol (2009) 58:963–972
DOI 10.1007/s00254-008-1576-6
temperate broadleaved and coniferous forests, as well as to
its more southerly tropical monsoon evergreen-broadleaved
forests (Tan and Huang 1985; Sun and Chen 1995; Chang
et al. 1999), few reported on the microclimatic character-
istics of plant communities in the hyperarid zone.
Associated with wavelengths in the range of 400–
700 nm, PAR represents the fraction of energy of green
plants can absorb from total solar radiation through the
photosynthetic process. As a basic energy source for pho-
tosynthesis, PAR is clearly an important environmental
factor in plant growth (Zhang and Qin 2002; Britton and
Dodd 1976; Sivakumar and Virmani 1984; Sinclair and
Lemon 1974; Sinclair and Knoerr 1982; Alados and Olmo
2000; Yann and Agnes 2000; Zhang and Zhang 2000).
While many researchers, in China and abroad, have
measured PAR, and no monitoring of PAR has been done
during dust storms.
Desert riparian forests represent an important forest
resource in arid desert environments such as those existing
in arid zone of northwest China. The oases which occur on
both sides of desert rivers house ecosystems mainly veg-
etation populated by Populus euphratica Olivier and
Tamarix ramosissima Ledeb. These oasis ecosystems
interact closely with surrounding desert ecosystems (Fu
2000). The study of the microclimate in desert riparian
forest ecosystems will contribute to our understanding of
the degradation of these ecosystems under the influence of
environmental factors currently prevailing in the arid zone
(Cheng et al. 1991; Feng et al. 1997; Zhang et al. 2004).
This study will provide some basic data on the ecological
functions and benefits of riparian forest ecosystems in
the hyper-arid zone, as well as provide a scientific basis
for the rational management, reasonable utilization of
existing forests, and the establishment of artificial forest
ecosystems.
Study area and methods
Study area
The study area located in the lower reach of the Heihe
river, in the Ejin region of the Inner Mongolia Autonomous
Region of China. Situated in the interior the study area has
an extremely arid climate and is one of the driest regions in
China. This area is characterized by scanty precipitation,
strong evaporation, frequent wind, abundant sources of
sand, and long durations of sunshine. Based on meteoro-
logical data recorded at the Ejin weather station between
1957 and 2002, mean annual precipitation is 42 mm
(max = 103 mm), while the minimum annual of evapo-
ration is 4,035 mm, nearly 90-fold the precipitation. About
70–80 of total annual precipitation occurs between June
and September. Desert oasis vegetation dominated by
riparian trees (P. euphratica) and shrubs (T. ramosissima).
Both species are tolerant to salinity, water logging, and
wind- and sand-blast damage. Serving to maintain an
ecological balance, protect agricultural and livestock pro-
duction and provide timber and fuel wood, they provide
significant ecological, economic and social benefits.
Microclimatic gradients in the desert oasis were
monitored March 2002, October 2003. Observation plots of
P. euphratica forest were located in a 1,333 ha. P. eu-
phratica reserve at Qiaoqiao in the Ejin region of northwest
China (42�210N and 101�150N), with an elevation of
920.46 m AMSL. The young P. euphratica forest aged 20–
25 years, with a density of 500 plants per hectare, a mean
height of roughly 10 m, and a mean breast-height diameter
of 12 cm. It’s under story includes T. ramosissima and
Sophora alopecuroides L. The soil type in the region is a
P. euphratica forest soil and soil texture of the 0–1.5 m
layer ranged from clayey loam to sand. Organic matter
content was 0.724% (w/w) in the 0–0.30 m soil layer and
0.127% (w/w) in the 0.30–2.00 m soil layer. The water
table depth ranged from 1.5 to 3.5 m in depth.
Microclimatic observation plots of naturally arising
T. ramosissima forest were located on former grasslands
fenced-off in 1988. Located at Erdaoqiao in the Ejin region
of northwest China (41�580N and 101�060E), the site
showed a vegetation cover of 70% or more, a plant density
of 1,300 bunches ha-1, a mean plant height of 2 m, and the
T. ramosissima plants were aged at least 20 years. The soil
type was a T. ramosissima soil developed from the parent
material of river alluvium. Localized surface salt spots are
present, and the water table is 2.0–3.5 m in depth.
The measurement of PAR under dust storm, cloudy and
clear day conditions on 28 April, 30 July, 6 September
2002, and respectively.
Study method
A 6 m tall meteorological station, including an automated
micrometeorological station (ICT Co., Australia) and a
Bowen ratio station (COMPELL Co., USA) was located in
each of the forest plot locations. The main parameters
monitored included effective radiation (total radiation, net
radiation and effective radiation), air temperature and
humidity, wind velocity and direction, soil temperature,
soil heat flux, and CO2 concentration. The radiation probe
was located at a height of over 2 m for beneath-canopy and
unforested land measurements and at least 1 m above the
tallest portion of the canopy for above-canopy measure-
ments, while the air temperature and humidity probes were
placed at heights of 1, 2, and 4 m. Wind velocity probes
were mounted at heights of 1, 2, 3, 4, and 6 m, with the
wind vane at a height of 6 m. Ground temperature probes
964 Environ Geol (2009) 58:963–972
123
were placed at depths of 0, 0.05, 0.10, 0.15, 0.20 and
0.40 m. CO2 concentration was determined at heights
of 1.5 and 3.5 m. Two soil heat flux plates were buried at
0.05 and 0.10 m below the soil surface. All the sensors
were connected by a cable to an indoor data logger (Zeno
3200-A. D). Observations were taken at 10 min intervals
on a 24-h a day basis, under all weather conditions.
Meteorological data recorded by the Ejin Meteorological
Bureau in 2002–2003 were used as the weather data for the
control areas located outside the forests. The net radiation
is determined by the exchange degree of the radiation
between the atmosphere and plant association.
Results
Solar radiation
Owing to the strong weakening effect of forest canopy
(absorption and reflection) on total solar radiation, the mean
monthly total solar radiation ð �QtotRmonthÞ beneath the canopies of
P. euphratica and T. ramosissima forests for the whole
growing season were 341.72 and 345.14 MJ m-2, respec-
tively, compared to 681.26 MJ m-2 for nearby non-forested
lands (Table 1). Thus, for the whole growing season,
ð �QtotRmonthÞ in the T. ramosissima forest exceeded that of
P. euphratica forest by 3.42 MJ m-2. The monthly solar
radiation total, QRmonthtot , was lowest in October, with 132.71
and 140.13 MJ m-2 measured for P. euphratica and
T. ramosissima forests, respectively. Thus, QRmonthtot beneath
the canopies of P. euphratica and T. ramosissima forest was
equivalent to 50.2 and 50.7%, respectively, of that over the
surrounding area. This proportion was at a maximum in June
(68% for both forest types), when QRmonthtot was at a maxi-
mum, and dropped gradually to a minimum in October (28.4
and 30.0%, respectively). For occasions when total daily
radiation QRdaytot was roughly the same, the ratio of daily mean
net radiation (QRdaynet ) to QRday
tot were 37.9 and 25.7% for
P. euphratica and T. ramosissima forests, respectively. For
the P. euphratica forest, this ratio peaked in May, June and
July (43.4, 45.0 and 47.1%, respectively), while for
T. ramosissima peak ratios occurred in September and
October (73.7 and 49.6%, respectively (Table 1). While
in the summer the proportion of QRdaynet to QRday
tot radiation of
P. euphratica forest well exceeded that of T. ramosissima
forest, the contrary was the case in the fall.
Basic characteristics of PAR
In the oasis under study, both the total daily PAR radiation,
PARRdaytot , and QRday
tot were significantly lesser on the dust
storm day than that on either a clear day or cloudy day
(Table 2). This is attributable to the greatly incr-
eased quantity of dust particles in the air cutting down
PARRdaytot and QRday
tot values by absorbing and scattering
solar radiation. Although PARRdaytot on the cloudy day was
less than that on the dust storm day, it was itself distinctly
smaller than that on the clear day. This is attributable to the
fact that a portion of the solar radiation was absorbed and
reflected by clouds and vapor in the atmosphere. As a rule,
the weakening effect of dust storms on solar radiation was
much greater than that of cloud and vapor on the cloudy
day.
While PARRdaytot and QRday
tot were the largest on the clear
day, followed by the cloudy day and least on the dust storm
day, their ratio, gRday = PARRdaytot /QRday
tot was the largest for
Table 1 Variations in monthly cumulative solar radiation in P. euphratica and T. ramosissima oasis forests (MJ m-2 month-1)
Stand Radiation type* May June July August September October Mean
P. euphratica QRmonthtot 480.44 539.59 407.24 311.23 179.12 132.71 341.72
PARRmonthtot 353.17 385.52 328.72 212.26 141.22 108.22 254.85
QRmonthnet 208.70 242.56 191.61 58.57 41.99 33.97 129.57
T. ramosissima QRmonthtot 482.19 541.05 410.54 314.22 182.69 140.13 345.14
PARRmonthtot 280.25 312.18 256.53 182.69 106.22 81.47 203.22
QRmonthnet 85.48 91.81 72.55 77.29 134.55 69.50 88.53
Surrounding area QRmonthtot 730.35 797.66 779.68 720.11 593.04 466.73 681.26
*QRmonthtot , QRmonth
net , PARRmonthtot : mean monthly total radiation, mean monthly net radiation, mean monthly photosynthetically active radiation
Table 2 Daily cumulative PAR and solar radiation, their ratio gRday,
mean air temperature and relative humidity on different weather days
April 28
(Dust day)
July 30
(Cloudy day)
September
6 (Clear day)
PARRdaytot (MJ m-2 d-1) 4.09 12.4 21.4
QRdaytot (MJ m-2 d-1) 7.24 23.9 43.4
gRday (%) 56.5 51.9 49.3
Mean air temperature (�C) 2.92 27.8 22.7
Mean relative humidity (%) 80.4 30.8 22.5
PARRdaytot , QRday
tot , gRday: Daily cumulative PAR, daily cumulative solar
radiation, and their ratio
Environ Geol (2009) 58:963–972 965
123
the dust storm day (56.5%) and least on the clear day
(49.3%).
On the dust storm day, the mean air temperature of the
oasis was only 2.92�C and the mean relative humidity
(RH%) was 80.4%. The sudden reduction in air tempera-
ture caused by cold air in the dust storm led to vapor
condensation, results in an increase in RH% and increased
absorption of infrared radiation. Coupled with the weak-
ening of solar radiation by dust particles, was their
especially strong absorption in the near infrared band,
which was reflected by an increase in gRday on the dust
storm day. The gRday value was larger on the cloudy day
than that on the clear day, as a result of the quantity of
clouds and vapor being larger on the cloudy day. The
cloudy day conditions reduced the contribution of air
molecules and aerosol particles to the scattering of radia-
tion and increased the absorption of infrared radiation
(Ji and Ma 1993). The air temperature was greater on the
cloudy day than that on the clear day, but this is a result of
the cloudy day being selected in the summer while the
clear day was selected in the fall (Zhang and Qin 2002).
Dust storm weather clearly had a strong weakening effect
on PAR.
Comparison of daily variations in PAR
Plots of diurnal variations in hourly PAR and total radia-
tion totals (PARRhourtot and QRhour
tot , respectively) on dust
storm, clear and cloudy days (Fig. 1), show that daily
variation in PARRhourtot and QRhour
tot closely parallel each
other. On the clear day both PARRhourtot and QRhour
tot were
very large and closely tied to changes in solar latitude
angle. Their variations were stable, and their maximum
values occurred at noon, local time. Comparatively, daily
variations in PARRhourtot on cloudy and dust storm days were
quite variable, being closely tied to shifts in weather
regime, cloud cover, as well as air vapor and dust content.
On the whole, the PARRhourtot and QRhour
tot values were larger
on the clear day than that on either the cloudy or dust storm
day, with PARRhourtot being least overall on the dust storm
day.
The daily maximum PARRhourtot value on the clear day
was greater than that measured in other regions such as
Zhangye and Linze (Zhang et al. 2004). The range of
PARRhourtot values on the clear day was large, whereas on the
dust storm day it was very small, indicating that dust storm
weather can greatly cut down on the PAR reaching plants.
Overall, a good linear relationship existed between
PARRhourtot and QRhour
tot (Table 3), though this was stronger on
the clear day, when fewer factors were affecting gRhour
(i.e., slope of plots), than that on the cloudy or dust storm
days, when the relationship was weakened by the effects of
dust particles, clouds and vapor.
Analysis of gRhour
Many factors affected gRhour values, including dust parti-
cles, clouds, vapor and solar radiation intensify, etc.
Previous studies (Ji and Ma 1993; Zhang and Qin 2002)
have shown that daily variation in gRhour shows certain
regularity on a clear day, but that its variation on a dust
storm day is complex. Given the weight of multiple factors
(airborne dust and vapor content) affecting it is constantly
changing due to the influence of the unstable stratification
of strong winds in the atmosphere.
Variations in gRhour during the dust storm were indeed
large and inconsistent (Table 4). Values of gRhour [ 70%,
such as those measured for 7:00, 8:00 and 9:00 in the
morning, are rarely found. Such values are likely the result
6:000123456
0
1
2
3
4
5
0.0
0.3
0.6
0.9
1.2
Acc
umul
ated
PA
R
hour
ly (
MJ.
m-2 )
Time
PAR Q(c)
Acc
umul
ated
PA
R
hour
ly (
MJ.
m-2 ) PAR
Q(b)
Acc
umul
ated
PA
R
hour
ly (
MJ.
m-2 )
PAR Q
(a)
18:0014:0010:00
6:00 18:0014:0010:00
6:00 18:0014:0010:00
Fig. 1 Daily variations in PAR under different weather conditions
[PARRhourtot (y-axis) is in MJ m-2 h-1): a the day of 28th April, b the
day of 30th July, c the day of 6th September
Table 3 Relationships between the hourly cumulative PAR
(PARRhourtot ) and hourly cumulative solar radiation (QRhour
tot ) under
different weather conditions
Date Relationships R2
28 April 2002 (dust) PARRhourtot = 0.5005QRhour
tot ? 0.0391 0.8672
30 July 2002 (cloud) PARRhourtot = 0.5353QRhour
tot - 0.0611 0.8824
6 September 2002
(clear)
PARRhourtot = 0.5709QRhour
tot - 0.0870 0.9738
966 Environ Geol (2009) 58:963–972
123
of the influence of the dust storm, which led to very small
PARRhourtot and QRhour
tot values generating a larger gRhour
value. Under dust storm conditions, total solar radiation
intensity over the dust storm day was very small.
According to Ji and Ma (1993), there exists a roughly
logarithmic relation between PARRhourtot and QRhour
tot in the
Linze region. Britton and Dodd (1976), working in Texas,
USA, showed that gRhour decreased with increasing solar
radiation. Dust particles reduce solar radiation by absorb-
ing strongly in the spectral range of 300–700 nm and have
even greater absorption in the near infrared range of 700–
2,000 nm (or even up to 4,000 nm). Thus, their presence
results in large values of g. On the dust storm day, cold air
caused vapor condensation and an increase in humidity,
which also contributed to increasinggRhour values. As a
whole, gRhour values were significant greater on the dust
storm day than that on the clear day.
The mean gRhour value between 7:00 and 11:00 on the
dust storm day exceeded 60% (Table 4). McCree (1976)
reported on instantaneous g value of 58.0–59.0% on an
overcast day. The present study has shown that high dust
concentrations resulting from a dust storm day have a much
greater impact on gRhour or gRday than that the high vapor
content associated with dense cloud.
Air temperature (2 m height) differences between the
open area surrounding the oasis and the oasis’ P. euphra-
tica and T. ramosissima woodlands were 1.62 and 0.83�C,
averaged over the main growing season of June–September
(Table 5). In the summer P. euphratica trees have dense
branches and leaves and thereby can block more solar
radiation than that T. ramosissima forest, hence its greater
ability to reduce monthly mean air temperatures between
June and September. Furthermore, in May P. euphratica
trees start to grow rapidly and hence have a strong ability to
check wind, whereas at this point T. ramosissima shrubs
are only starting to sprout, so their ability to check wind at
two meters’ height is less. As a result, air temperature was
lesser in P. euphratica forestland than that in T. ramo-
sissima forest. In October, the amplitude of daily air
temperature changes was decreasing. This amplitude was
lesser in the surrounding area than that in either the
P. euphratica or T. ramosissima forest, due to the lack of
shelter and its low heat absorbance of the open lands.
Overall, the oasis’ air temperature was lower than that of
the surrounding area during the growing season, indicating
that the forest served to reduce air temperature and
decrease temperature differences. The fact that air tem-
perature was lower in P. euphratica forest compared to
T. ramosissima forest shows that such forest functions are
tied to forest cover and ground surface conditions; the
greater the forest cover, the greater its temperature-
lowering ability.
Vertical variations in air temperature
During the growing season, the air temperature above the
oasis forest canopy was greater than that beneath the can-
opy in the interior of the oasis. For the P. euphratica forest
the temperature difference from top to bottom varied
between 0.97 and 5.74�C, while for the T. ramosissima
forest it varied between 0.61 and 3.49�C (Fig. 2). The air
temperature in the P. euphratica forestland increased rap-
idly from the ground surface to a height of 2 m and then
decreased from the height of 2–4 m. In contrast, for the
T. ramosissima forest the situation was the contrary, the air
temperature decreased rapidly from the ground surface to a
height of 1 m, but then increased from the height of 1–2 m,
with no changes occurring above a height of 2 m. The
vertical distribution of air temperatures in the forests has
the following features: (1) the air temperature over the
oasis’ forest canopy was greater than that under the canopy
and the beneath-canopy air temperature gradually
decreased from top to bottom, (2) in the P. euphratica
forest the largest vertical temperature variations occurred
Table 4 Hourly cumulative PAR (PARRhourtot ) and solar radiation (QRhour
tot ), and their ratio gRhour on a dust storm day
7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00
PARRhourtot (MJ m-2 h-1) 0.22 0.15 0.32 0.53 0.48 0.42 0.43 0.51 0.40 0.34 0.20 0.09
QRhourtot (MJ m-2 h-1) 0.31 0.19 0.43 0.81 0.79 0.73 0.82 0.98 0.75 0.75 0.47 0.21
gRhour (%) 71.8 81.6 75.1 66.1 60.2 56.7 52.0 51.8 53.0 44.7 44.0 44.6
Table 5 Variations in monthly mean air temperature in P. euphratica and T. ramosissima forests (�C)
Stand May June July August September October Mean
P. euphratica forest 24.06 25.36 27.31 25.34 15.12 10.15 21.22
T. ramosissima forest 22.17 25.86 28.32 25.62 16.48 12.67 21.85
Surrounding area 20.30 26.5 28.90 27.30 16.90 8.90 21.47
Environ Geol (2009) 58:963–972 967
123
in the range from the ground surface to a height of 1 m,
while in the T. ramosissima forest the amplitude of varia-
tion above the height of 2 m was small, (3) under
extremely dry weather conditions the vertical variation in
air temperature was very much reduced.
The mean RH% was greater in the woodlands than that
in the surrounding area (Table 6). Averaged over the
growing season RH% was 8.5 and 4.2% (absolute) greater
in the P. euphratica and T. ramosissima woodlands,
respectively, than that in the surrounding area, and similar
trends were apparent for individual months. The forests
maintained a higher humidity due to the shading function
of the forest canopy. The humidity difference between the
woodlands and the surrounding area was smallest between
May and June, i.e. 4.9–6.3% for P. euphratica and 4.0–
4.7% for T. ramosissima. In contrast, the difference for the
P. euphratica woodland was at its highest (9.3–16%) in
September-October. The RH% peaked between August and
October. Its amplitude of variation was lesser in the
woodlands than that in the surrounding area, indicating that
forests play a unique role in maintaining conditions of
lower water requirements for plant survival in the extre-
mely arid zone. Different forest types have different effects
on the relative humidity: P. euphratica forest is tree forest,
so its canopy has a greater shading function than that the
shrubs of T. ramosissima forest. Therefore, it is not sur-
prising that the P. euphratica forest’s RH% was 4.33%
higher (absolute) than that of the T. ramosissima forest.
Vertical distribution of air humidity
In forested lands, the RH% above the forest canopy was
lower than that below (Table 4). The range of variation in
amplitude of RH% in the P. euphratica forest (25.8–
58.6%) was greater than in the T. ramosissima forest (24.5–
38.7%). The largest ranges of the relative humidity in the
P. euphratica forest occurred in May and October and
varied irregularly in between (Fig. 3). This may be related
to the tree species, which is broadleaved. The RH% at a
0
1
2
3
4
5
5
0
1
2
3
4
5
Air temperature
Hei
ght (
m)
May Jun Aug Sep Oct
(a)
Hei
ght
(m)
(b)
3025201510
5 3025201510
Fig. 2 Vertical air temperature variations in a P. euphratica and
b T. ramosissima forests
Table 6 Variation in monthly mean relative humidity in P. euphratica and T. ramosissima forests (%)
Stand May June July August September October Mean
P. euphratica forest 27.50 30.50 39.40 34.90 48.60 44.10 37.48
T. ramosissima forest 25.25 30.34 34.30 34.25 36.41 38.32 33.15
Surrounding area 21.20 25.60 31.50 34.10 32.50 34.80 29.00
100
1
2
3
4
5
0
1
2
3
4
5
Hei
ght
(m)
Relative humidity (%)
(b)
Hei
ght
(m)
May Jun Aug Sep Oct
(a)
6050403020
10 6050403020
Fig. 3 Vertical humidity variations in a P. euphratica and
b T. ramosissima forests
968 Environ Geol (2009) 58:963–972
123
height of 1 m in the T. ramosissima woodland changed
little with time, and the RH% was greater at the height of
2 m than that at 4 m, because the shrub canopy was at 2 m.
Soil temperature
Between May and September, the P. euphratica forest had
dense branches and leaves, and, thus, a stronger ability to
block solar radiation than that T. ramosissima forest
(Table 7). Consequently, during these months monthly mean
soil temperature ð �Tmthsoil Þ in the P. euphratica forest was
lower than that in the T. ramosissima forest. In October,
P. euphratica’s defoliation led to a large decrease in its
ability to block solar radiation, so that the P. euphratica
forest’s ð �Tmthsoil Þ rose above that of the T. ramosissima forest.
The ð �Tmthsoil Þ of the 0.20–0.40 m layer changed little over
the season, and was lower in the P. euphratica forest soil
than in that of the T. ramosissima forest (Fig. 4). Vertical
variation in ð �Tmthsoil Þ in the P. euphratica forest soil was not
significant between May and August, but varied greatly in
September and October, when the 0–0.10 m soil layer’s
ð �Tmthsoil Þ changed dramatically. The 0.10–0.40 m soil layer’s
ð �Tmthsoil Þ tended to decrease from May to August, but
increased from September to October. This was tied to the
P. euphratica tree’s growth regime, and it’s defoliation in
September through October, which led to a lessened ability
to block solar radiation. Vertical variation in ð �Tmthsoil Þ in the
T. ramosissima forest tended to be consistent across
months. In the hot summer, P. euphratica forest, unlike
T. ramosissima forest, significantly reduced the tempera-
ture at the ground surface and in the 0–0.20 m soil layer.
Variations in wind speed
The variation of wind force was mainly manifested in the
marked reduction of wind speed due to the forest canopy’s
blocking of the wind. The May–October mean wind speeds
in the P. euphratica and T. ramosissima forests were 0.33
and 0.72 m s-1, respectively, some 2.70 and 2.31 m s-1
lower than that of the surrounding area (Fig. 5). Clearly
P. euphratica forest had a much greater ability to block
wind than that does T. ramosissima forest. For individual
months, wind speed in the surrounding area is [T. ramo-
sissima forest [ P. euphratica forest. From this, it can be
Table 7 Variations in monthly mean soil (1.8 m) temperatures in P. euphratica and T. ramosissima forests (�C)
Stand May June July August September October Mean
P. euphratica forest 15.67 20.29 22.17 19.07 16.43 15.58 18.20
T. ramosissima forest 17.60 21.95 22.67 22.70 17.74 13.90 19.43
050
40
30
20
10
0
50
40
30
20
10
0
Hei
ght
(m)
Soil temperature (°C)
(b)
Hei
ght
(m)
May Jun Aug Sep Oct
(a)
5040302010
Fig. 4 Vertical soil temperature variations in a P. euphratica and
b T. ramosissima forests
0.0
1
2
3
4
5
6
1
2
3
4
5
6
Hei
ght
(m)
Wind speed (m/s)
(b)
Hei
ght
(m)
May Jun Aug Sep Oct
(a)
1.51.00.5
Fig. 5 Vertical wind velocity variations in a P. euphratica and
b T. ramosissima forests
Environ Geol (2009) 58:963–972 969
123
seen that forests in the hyper-arid zone can significantly
reduce wind speed (Table 8).
The common feature in the vertical variation of wind
velocity in the P. euphratica and T. ramosissima forests
was that wind velocity increased with increasing height
(Figure 5). The mean wind velocities at heights of 1, 2, 3,
4, and 6 m in the P. euphratica forest were 0.23, 0.24, 0.25
and 0.50 m s-1, respectively, while in the T. ramosissima
forest they were 0.24, 0.45, 0.72, 0.91 and 1.28 m s-1,
respectively. However, the wind velocity varied less with
height in the P. euphratica forest than that in the T. ra-
mosissima forest. The wind velocity at a height of 1–3 m in
the P. euphratica forest changed little, while the wind
velocity in the T. ramosissima forest continuously
increased from a height of 1 m to one of 6 m.
Comparison of different plants communities
Comparisons were made between a humid climate mon-
soon evergreen-broadleaf forest in the south Asian tropics,
a water conservation forest in the arid zone of a temperate
region, and the presently studied desert riparian forest from
an extremely dry climatic zone (Table 9). The latter forest
has been shown above to have a significant effect on
microclimate: it can diffuse solar radiation, reduce air,
ground surface and soil temperatures, decrease wind speed,
increase air and soil humidity, and, thus, plays even more
significant ecological rule than other forests.
The �QtotRmonth
under the canopy of the desert riparian forest
was about 50% of that of the bare land surrounding it,
while a corresponding proportion of 72.8% was found for
the monsoon evergreen-broadleaf forest. Thus, the desert
riparian forest can trap solar radiation more effectively than
that the monsoon forest. Similarly, the P. euphratica forest
has a stronger temperature lowering function than the
monsoon forests of Betula platyphylla Sukaczev and Pinus
tabulaeformis Carr. (Table 9).
Since the forest canopy prevents the exchange of gas
from inside to outside the forest, water in the interior of the
forest forms a small circulating system or high-humidity
environment. The RH% in the P. euphratica and T. ra-
mosissima forests in the hyper-arid zone were low, but
were nonetheless 8.5 and 4.2% higher, respectively, than
that outside the forest. This compares with values of 9% for
monsoon forest, and 7 and 4% for water conservation
B. platyphylla and Pinus tabulaeformis forests, respec-
tively. This shows that desert riparian forests in a hyper-
arid environment have a significant moisture-holding
function.
Soil temperatures in the monsoon, water-conservation
and desert riparian forests were 4–5, 6–8, and 7–10�C less,
respectively, that in the open areas surrounding them. The
desert riparian forest significantly reduced the temperature
at the ground surface and in the 0–0.20 m soil layer, and
buffered sharp variations in soil temperature. This is
favorable to soil microorganism activity, reduces evapo-
ration from the topsoil and enhances the effectiveness of
precipitation.
The mean wind velocities in the B. platyphylla, Pinus
tabulaeformis, P. euphratica and T. ramosissima forests
were 0.92, 1.01, 2.70, and 2.31 m s-1 lower than that in
their respective surrounding areas. This shows that the
desert riparian forests in the hyper-arid zone could signif-
icantly reduce wind velocity.
Conclusions and discussion
From the comparison of microclimatic characteristics of
the interior of a desert oasis in northwest China and the
surrounding area, we were able to draw the following
conclusions: total solar radiation and air temperature are
lower, but RH% higher, under forested portions of the oasis
than that in the surrounding area. Oasis plant communities
have ecological functions including diffusing solar radia-
tion, adjusting near-ground layer, land surface and soil
temperatures, reducing wind velocity, increasing soil and
atmospheric moisture, and decreasing soil erosion, etc.
Therefore, they can help conserve local water and soil
resources, improve the environment and aid in the main-
tenance of an ecological balance.
The microclimatic differences between P. euphratica
and T. ramosissima forest ecosystems include lower
growing season total solar radiation in P. euphratica than
that T. ramosissima forest, greater summertime net solar
radiation in P. euphratica forest than that in T. ramosissima
forest, and the contrary in the fall. Air temperatures in the
P. euphratica forest, which decreased from ground surface
to canopy, were lower than that in the T. ramosissima
forest, where, by contrast, the air temperature increased
Table 8 Monthly variations in monthly mean wind speed in the P. euphratica and T. ramosissima forests (m s-1)
Stand May June July August September October Mean
P. euphratica forest 0.35 0.35 0.29 0.41 0.25 0.29 0.33
T. ramosissima forest 0.83 0.75 0.84 0.69 0.76 0.45 0.72
Surrounding area 2.90 3.50 3.00 3.20 2.60 3.00 3.03
970 Environ Geol (2009) 58:963–972
123
from the ground surface to the canopy. The P. euphratica
forest’s RH% was, on average, 4.3% (absolute) higher than
that in the T. ramosissima forest. The amplitude of varia-
tion in RH% was greater in the P. euphratica forest than
that in the T. ramosissima forest. Monthly mean soil tem-
peratures were lower in P. euphratica forest than that in
T. ramosissima forest and tended to decrease from the land
surface to the subsurface soil layer. The P. euphratica
forest had a greater ability to check wind and sand, than
that did the T. ramosissima forest. Vertical variations in
wind velocity in the P. euphratica and T. ramosissima
forests increased with height, but largely in the latter case.
On the dust storm day, oasis PARRhourtot was less than that
on either a clear or cloudy day. The dust storm reduced
PAR largely than that clouds or vapor. On the clear day
PARRhourtot and QRhour
tot paralleled each other and showed a
strong linear relationship, whereas on the cloudy and dust
storm days the relationship was weaker. On the dust storm
day, gRhour (slope of relationship) values were very large
and varied in an irregular manner.
Desert riparian forest can diminish solar radiation inside
the forest, reduce temperature, increase humidity and
decrease wind velocity. Such functions conserve water and
soil resources, improve ecosystems, maintain an ecological
balance, protect agricultural and livestock production in
oases, provide better living and production conditions for
local people, and yield greater ecological, economic and
social benefits in a hyper-arid zone characterized by scanty
precipitation and high evaporation. It is urgent that desert
riparian habitats in China be restored and protected. Res-
toration of P. euphratica and T. ramosissima forests and
associated vegetation could employ artificial rejuvenation
and reproduction techniques to enlarge and diversify the
forests in such a manner as to optimize population and
community structures and so give full play to the forests’
primary function of protecting the environment of the arid
zone.
Acknowledgments This research was supported by a grant from
National Natural Sciences Foundation of China (No. 40671010;
40725001, 40501012), the Key Project of the Chinese Academy of
Sciences (KZCX2-XB2-04-02), National Key Technology R & D
Program (No. 2007BAD46B01).
References
Alados I, Olmo FJ (2000) Estimation of photosynthetically active
radiation under cloudy conditions. Agric For Meteorol 102:39–50
Britton CM, Dodd JD (1976) Relationships of photosynthetically
active radiation and shortwave irradiance. Agric Meteorol
17:1–7
Chang J, Pan XD, Ge Y, Chen ZH, Liu K, Chen QC (1999)
Microclimatic characteristics of evergreen Cyclobalanopsisglauca forest. Acta Ecol Sin 19(1):68–75 (in Chinese)T
ab
le9
Com
par
iso
no
fm
icro
clim
ate
ind
iffe
ren
tfo
rest
lan
ds
Item
sM
on
soo
nev
erg
reen
bro
adle
aved
fore
stW
ater
con
serv
atio
nfo
rest
inD
aqin
gM
oun
tain
Des
ert
rip
aria
nfo
rest
Cli
mat
iczo
ne
Su
btr
opic
alm
on
soo
nh
um
idcl
imat
eT
emp
erat
esu
bar
idzo
ne
Mid
-tem
per
ate
hy
per
-ari
dzo
ne
Stu
dy
area
Zh
aoq
in,
Gu
ang
do
ng
Bai
shit
ou
go
ub
asin
inD
aqin
gM
oun
tain
,In
ner
Mo
ng
oli
aE
jin
,In
ner
Mo
ng
oli
a
Lat
itu
de
and
lon
git
ude
23
�100 N
–4
0�2
00 –
42�3
00 N
11
2�3
40 E
–9
9�3
00 –
10
2�0
00 E
To
tal
ann
ual
sola
rra
dia
tio
n
(MJ
m-
2p
ery
ear)
3,4
88
.8–4
,792
.31
5,8
61
.52
6,5
10
–6
,72
0
Mo
nso
on
fore
stS
urr
oun
din
gar
eaB
etula
pla
typhy
lla
Pin
us
tabula
eform
isS
urr
oun
din
gar
eaP
.eu
ph
rati
caT
.ra
mo
siss
ima
Su
rro
un
din
gar
ea
Air
tem
per
atu
re(�
C)
19
.92
1.5
14
.28
14
.87
15
.52
7.8
8.5
21
.47
Rel
.h
um
idit
y(%
)8
77
86
86
56
13
7.5
33
.22
9
So
ilte
mp
erat
ure
(�C
)
0m
19
.62
4.7
13
.19
13
.79
19
.91
19
.97
20
.66
30
.12
0.0
5m
19
.82
4.8
10
.81
12
.66
18
.89
20
.34
18
.26
29
.21
0.1
0m
19
.72
4.2
10
.88
12
.63
18
.45
18
.61
16
.60
27
.99
0.1
5m
20
.12
4.1
10
.90
12
.38
18
.55
17
.94
16
.39
24
.00
0.2
0m
20
.02
4.2
10
.78
12
.21
18
.33
16
.80
15
.24
20
.03
Win
dv
elo
city
(ms-
1)
––
0.2
50
.16
1.1
70
.33
0.7
23
.03
Environ Geol (2009) 58:963–972 971
123
Cheng HS, Kang YH, Feng JC (1991) Preliminary study of plant
growth and water balance in Shapotou region of Tengger Desert.
J Desert Res 11(2):1–10 (in Chinese)
Federer CA (1968) Spatial variation of net radiation, Albedo and
surface temperature of forests. J Appl Meteorol 7:789–795
Feng JZ, Huang ZC, Zhang CL (1997) Several advances in the
quantitative study of environmental plant physiology. J Desert
Res 17(1):89–94
Fu XF (2000) Study of oasis development and environmental
coordination in arid zone. J Desert Res 20(2):197–200 (in
Chinese)
Garrett HE (1978) Spatial and temporal variation in carbon dioxide in
an oak-hickory forest ravine. For Sci 24(2):180–190
Hirose T (1995) Canopy structure and photon flux partitioning among
species in herbaceous plant community. Ecology 76(2):466–474
Ji GL, Ma XY (1993) Characteristics of photosynthetically active
radiation in Zhangye region. Plateau Meteorol 12(2):141–146 (in
Chinese)
Kimmins JP (1977) Forest Ecology, 2nd edn. Prentice Hall, Engle-
wood Cliffs, pp 596
McCree KJ (1976) A solarimeter for measuring photosynthetically
active radiation. Agric Meteorol 3:353–366
Sinclair TR, Lemon ER (1974) Penetration of photosynthetically
active radiation in corn canopies. Agron J 66: 201–205
Sinclair TR, Knoerr KR (1982) Distribution of photosynthetically
active radiation in the canopy of a loblolly pine plantation J Appl
Ecol 19:183–191
Sivakumar MVK, Virmani SM (1984) Crop productivity in relation to
interception of photosynthetically active radiation. Agric For
Meteorol 31:131–141
Sun XF, Chen LZ (1995) Preliminary study of radiation energy
environment of deciduous broadleaved forest in warm temper-
ature. Acta Ecol Sin 15(3):278–286 (in Chinese)
Tan SM, Huang JL (1985) Preliminary approach to the microclimatic
characteristics of mixed Eucalyptus tuoliensis forest. Acta Ecol
Sin 5(3):241–248 (in Chinese)
Yann NV, Agnes B (2000) PAR extinction in shortgrass ecosystems:
effects of clumping, sky conditions and soil albedo. Agric For
Meteorol 105:21–41
Ye DZ, Zhou JF (2000) Origin and control measures of dust storm
weather in North China. Acta Geogr Sin 55(5):513–521 (in
Chinese)
Zhang XZ, Zhang YG (2000) Measuring and modeling photosyn-
thetically active radiation in Tibet Plateau during April–October.
Agric For Meteorol (102): 207–212
Zhang YL, Qin BQ (2002) Basic characteristics of phytosynthetically
active radiation (PAR) and its climatological calculation in
Taihu Lake region, Agricultural Meteorology. J Solar Energy
23(1):118–123 (in Chinese)
Zhang YP, Liu YH, Ma YX, Wang JX, Dou JX, Guo P (2002)
Microclimatic characteristics in different growing stages of
tropic forest. J Nanjing For Univ (Nat Sci) 26(1):83–87 (in
Chinese)
Zhang L, Dong ZC, Huang XL (2004) Study of the model of the
relation between typical xerophyte growth and groundwater.
J Desert Res 1:110–113 (in Chinese)
972 Environ Geol (2009) 58:963–972
123