Interannual characteristics of the surface hydrological variables over the arid and semi-arid areas of northern China


Ma, Z.G.; Fu, C.B.

Global and Planetary Change 37(3-4): 189-200

2003


The characteristics of the surface humid index (SHI) were analyzed based on 160 station data in China from 1951 to 1998. The surface humid index is defined as SHI = (P)/(P (sub e) ), where P (sub e) is potential evaporation suggested by Thornthwaite's method. The difference between the evolutionary features of the SHI in typical arid regions of north China (Huabei and the northwest) was compared. The results show that the SHI is decreasing (drying trend) in the Huabei region of north China but is increasing in some areas of northwest China (wetting trend). Under regional warming, the drought in the center of north China mainly resulted from the decrease in precipitation and is partly due to the increase in evaporation. A dry period of about 40 years was revealed from the historical data over the area. Increasing evaporation caused by increasing temperature probably intensified the drought in that area, but is not the main reason for the drought. It is the less precipitation that mainly results in the present drought in north China. In addition, the SHI variations in different seasons were also analyzed; the result indicates the notable difference of SHI variation between seasons. Finally, the geographical distribution of annual SHI variation over China was given.

Available
online
at
www.sciencedirect.com
ELSEVIER
SCIENCE ELDIR ECT°
Global
and
Planetary
Change
37
(2003)
189-200
GLOBAL
AND
PLANETARY
CHANGE
www.elsevier.com/locate/gloplacha
Interannual
characteristics
of
the
surface
hydrological
variables
over
the
arid
and
semi
-arid
areas
of
northern
China
Ma
Zhuguo
*
,
Fu
Congbin
START
Regional
Center
for
Temperature
East
Asia,
Institute
of
Atmospheric
Physics,
Chinese
Academy
of
Sciences,
Beijing
100029,
China
Accepted
25
July
2002
Abstract
The
characteristics
of
the
surface
humid
index (SHI)
were
analyzed
based
on
160
station
data
in
China
from
1951
to
1998.
The
surface
humid
index
is
defined
as
SHI=(P)/(P
e
),
where
P
e
is
potential
evaporation
suggested
by
Thornthwaite's
method.
The
difference
between
the
evolutionary
features
of
the
SHI
in
typical
arid
regions
of
north
China
(Huabei
and
the
northwest)
was
compared.
The
results
show
that
the
SHI
is
decreasing
(drying
trend)
in
the
Huabei
region
of
north
China
but
is
increasing
in
some
areas
of
northwest
China
(wetting
trend).
Under
regional
warming,
the
drought
in
the
center
of
north
China
mainly
resulted
from
the
decrease
in
precipitation
and
is
partly
due
to
the
increase
in
evaporation.
A
dry
period
of
about
40
years
was
revealed
from
the
historical
data
over
the
area.
Increasing
evaporation
caused
by
increasing
temperature
probably
intensified
the
drought
in
that
area,
but
is
not
the
main
reason
for
the
drought.
It
is
the
less
precipitation
that
mainly
results
in
the
present
drought
in
north
China.
In
addition,
the
SHI
variations
in
different
seasons
were
also
analyzed;
the
result
indicates
the
notable
difference
of
SHI
variation
between
seasons.
Finally,
the
geographical
distribution
of
annual
SHI
variation
over
China
was
given.
0
2003
Published
by
Elsevier
Science
B.V.
Keywords:
surface
hydrological
variables;
climate
variation;
surface
humid
index
(SHI);
drought
1.
Introduction
As
a
serious
environmental
problem,
drought
is
closely
related
to
the
climate
variation
and
other
surface
hydrological
processes.
In
the
past,
many
studies
discussed
drought
only
through
analyzing
the
climate
variable
such
as
precipitation.
However,
drought
is
not
only
a
climate
issue,
but
is
also
a
factor
affected
by
other
environmental
processes,
such
as
*
Corresponding
author.
Fax:
+86-10-62045230.
E-mail
address:
mazg@tea.ac.cn
(Z.
Ma).
evaporation,
runoff
and
surface
air
temperature,
which
determines
that
it
should
be
studied
from
a
multi-
disciplinary
view.
How
to
express
the
total
character
of
drought
is
the
key
to
correctly
understand
the
cause
of
drought.
In
this
paper,
the
surface
humid
index
(SHI),
a
parameter
suggested
by
Hulme
et
al.
(1992),
was
employed
to
analyze
the
state
of
dry
and
wet
in
the
surface
over
the
arid
region
of
north
China.
This
parameter
can
reasonably
reflect
the
dry/wet
state
of
land
surface
in
that
it
considers
two
major
surface
hydrological
processes
—precipitation
and
potential
0921-8181/03/$
-
see
front
matter
©
2003
Published
by
Elsevier
Science
B.V.
doi:10.1016/50921-8181(02)00203-5
190
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
evaporation
—simultaneously
(Manabe,
1981).
Mean-
while,
other
processes
affecting
the
dry
and
wet
state
of
surface
were
also
discussed
here,
and
the
probable
relationship
between
drought
and
regional
warming
was
investigated.
In
the
arid
and
semi
-arid
regions
of
north
China,
the
climate
variation
have
been
studied
by
many
researchers
(Wei
and
Cao,
1998;
Yan,
1995,
1999).
These
works
analyzed
in
detail
the
temporal
and
spatial
variation
of
precipitation
but
did
not
depict
the
overall
characteristics
of
drought.
For
example,
the
decrease
in
precipitation
does
not
mean
that
drought
ensues
as
a
result
of
the
uncertainty
of
evaporation.
Also,
the
increase
of
precipitation
cannot
represent
wetting
land
surface.
Recently,
some
studies
(Hulme
et
al.,
1992;
Fu,
1994;
Thomas,
2000)
emphasized
the
expression
of
the
drought
characteristics
on
the
view
of
a
multidisciplinary
study.
In
these
papers,
the
evaporation,
the
dry/wet
state
in
the
global
scale,
and
the
drought
index
were
investigated.
Neverthe-
less,
the
dry/wet
state
in
the
land
surface
of
the
arid
area
of
north
China
and
their
differences
between
seasons
and
regions
are
not
clear.
These
studies
did
not
analyze
the
dry/wet
trend
and
their
relation
with
global
warming.
Manabe's
results
(Manabe,
1981;
Wetherald
and
Manabe,
1999)
indicated
that
the
soil
moisture
in
mid
-latitude
in
summer
would
decrease
under
global
warming.
In
that
case,
the
situation
in
the
arid
and
semi
-arid
region
of
north
China
is
an
impor-
tant
scientific
and
social
problem
because
of
its
influence
on
water
resource
management
and
utiliza-
tion.
Therefore,
temporal
and
spatial
variation,
as
well
as
the
differences
between
two
typical
drought
regions,
will
be
investigated
in
this
paper.
In
this
paper,
the
trend
and
regional
difference
of
interannual
variation
were
investigated,
showing
the
test
of
significance
of
the
trend
in
the
regions
and
the
geographical
distribution
of
SHI.
The
long-term
var-
iation
of
surface
hydrological
budget
in
the
two
typical
areas
has
been
investigated.
2.
Data
and
method
The
data
are
from
the
Chinese
Meteorological
Administration
from
1951
to
1998.
There
are
160
observational
stations
in
China
with
monthly
precip-
itation
and
surface
temperature.
Some
long
serial
data
were
taken
from
the
Carbon
Dioxide
Information
Analysis
Center
(CDIAC)
numeric
data
package
NDP012
and
NDP039,
updated
and
enlarged
version
of
the
original
NDP
data
package
by
Tao
et
al.
(1991),
and
added
the
recent
years
data
of
monthly
precip-
itation
and
surface
temperature
from
Chinese
Climate
Center.
The
same
data
were
used
throughout
this
paper.
According
to
Hulme
et
al.'s
(1992)
suggestion,
the
surface
humid
index (SHI)
can
be
written
as
follows:
SHI
=
P
e
(1)
where
P
is
the
observational
monthly
precipitation,
and
P
e
is
the
monthly
potential
evapotranspiration.
Using
modified
Thornthwaite's
scheme
(Ma,
1999),
potential
evapotranspiration
was
calculated
only
fr
om
the
monthly
mean
surface
temperature.
In
formula
(1),
it
is
easy
to
see
that
the
SHI
includes
two
important
factors
affecting
the
water
budget
of
land
surface
precipitation
and
potential
evaporation.
So
it
is
a
rational
parameter
to
be
used
for
analyzing
wet/dry
state
(WDS)
in
the
land
surface,
but
in
winter
of
north
China,
because
the
surface
temperature
is
always
under
0
°C,
P
e
=
0,
SHI'
co.
Therefore,
the
SHI
is
not
suitable
to
indicate
the
WDS
in
that
season.
In
winter,
the
precipitation
can
be
used
for
representing
the
WDS.
The
difference
between
precipitation
and
evaporation
(P
E,
where
P
is
the
observational
monthly
precipitation
and
E
is
the
monthly
evapora-
tion)
was
used
to
explore
the
historical
characteristics.
Evaporation
(E)
was
calculated
using
Gao's
scheme
(Yang
and
Song,
1999).
The
Mann
—Kendall
method
(simply
named
M
—K
method;
Snyers,
1990;
Fu
and
Wang,
1992)
was
used
for
testing
the
significance
of
the
trend
of
the
WDS.
Two
typical
regions
are
the
Huabei
region
(named
as
HR
as
follows)
and
the
northwest
region
of
north
China
(NR),
respectively.
HR
is
fr
om
35°N
to
42°N
and
is
to
the
east
of
110°E.
Because
of
complicated
geography
and
various
climate
conditions,
we
split
the
northwest
region
into
two
subregions,
one
is
the
No.
1
region
(named
as
NR
1)
extended
to
north
of
35°N
and
to
the
west
of
95°E,
the
other
is
the
No.
2
region
(named
as
NR
2)
covering
from
95°E
to
105°E
and
to
the
north
of
35°N.
There
are
24
observational
stations
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
191
50N
NR.2
40N
30N
20N
NR.1
A
\
o
I
/
•••
80E
90E
100E
110E
120E 130E 140E
Fig.
1.
Distribution
of
the
stations
in
the
three
subregions.
in
HR,
13
stations
in
NR
1
and
6
stations
in
NR
2.
The
extent
of
every
region
was
plotted
in
Fig.
1.
3.
The
interannual
variation
of
two
typical
drought
regions
and
their
difference
In
order
to
conveniently
analyze
the
general
regional
characteristics
of
the
WDS
and
to
compare
the
difference
between
two
regions,
the
area
mean
in
a
subregion
was
calculated.
Some
results
depict
the
whole
evidence
as
follows.
3.1.
Interannual
variation
of
SHI
during
the
fl
ood
season
In
north
China,
precipitation
mostly
occurs
in
the
fl
ood
season
(from
May
to
October),
so
the
WDS
in
the
period
can
represent
the
annual
WDS.
Therefore,
from
the
WDS
of
the
fl
ood
season,
we
can
see
the
annual
WDS.
The
data
of
monthly
precipitation
and
monthly
mean
temperature
are
from
the
Chinese
Meteorological
Administration
fr
om
1951
to
1998.
Fig.
2
is
the
interannual
variation
of
the
area
mean
SHI
in
the
fl
ood
season
in
the
different
subregions,
which
was
calculated
by
the
M
—K
method.
It
is
a
10
-
year
running
mean.
Fig.
2
shows
that
there
is
a
dominant
trend
of
decrease
of
SHI
in
HR
since
the
1950s
spaced
by
a
weak
increase
within
1970s.
So
it
can
be
concluded
that
there
has
been
a
drying
trend
(decreasing
SHI)
in
HR
since
the
1950s.
The
drying
trend
reached
the
significance
of
95%
in
the
region.
In
contrast,
the
increasing
trend
of
surface
temperature
(7)
existed
in
the
period,
and
the
trend
also
reached
the
significance
of
95%.
Meanwhile,
we
can
find
that
the
trend
of
SHI
is
opposite
to
that
of
the
T
in
the
region.
The
analysis
also
indicates
the
decreasing
trend
of
precipitation
after
the
middle
of
1970s,
which
is
consistent
with
the
present
decreasing
WDS
in
HR.
Probably,
the
increasing
T
and
decreasing
P
are
the
main
reasons
that
cause
the
drying
trend,
and
intro-
duce
a
warm
and
dry
situation.
From
Fig.
2,
there
is
a
little
increasing
trend
of
SHI
in
NR
1
in
the
fl
ood
season
after
1980s,
but
the
trend
cannot
reach
the
significance.
In
NR
2,
a
decreasing
SHI
exists
after
the
1980s.
Concerning
the
variation
of
the
temperature,
a
decrease
of
T
in
NR
1
and
an
opposite
trend
in
NR
2
can
be
found
after
the
1980s.
According
to
the
analyses
above,
some
conclusions
can
be
drawn.
In
HR
and
NR
2,
the
increasing
temperature
intensifies
the
present
drying
trend
of
the
land
surface
because
of
decreased
precipitation,
then
the
drying
land
surface
will
further
increase
T
This
fact
agrees
with
the
modeling
results
under
global
warming
(Manabe,
1981).
192
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
Huabei(HR)
1.2
SHI
.20
H
1.0
.16
8
co
.12
.6
.08
_
1
.4
.04
0
0
-2
-2
4
SHI
-6
=
-6
1950
1960
1970 1980 1990
2000
Year
NR.1
SHI
H
I I
SH
T
.7
.6
.4
.3
.2
2
0
NR.2
SHI
H
-2
-4
1950 1960 1970 1980 1990
2000
1950 1960
1970 1980 1990
2000
Year
Year
Fig.
2.
Regional
mean
variation
and
their
test
of
SHI
and
air
temperature
(H=
10
-year
running
mean;
T=
annual
mean
temperature.
The
upper
fi
gure
is
original,
and
the
lower
is
the
test
of
trend
of
SHI
by
the
Mann
—Kendall
method).
3.2.
Interannual
variation
in
different
seasons
Although
SHI
in
the
fl
ood
season
basically
repre-
sents
the
characteristics
of
the
WDS
in
the
whole
year,
it
cannot
depict
the
interannual
variation
in
various
seasons
and
their
differences
between
seasons.
In
this
section,
these
will
be
explored.
Fig.
3
shows
the
anomaly
of
the
9
-year
running
area
mean
in
the
three
subregions.
The
facts
as
follows
can
be
obtained.
The
T
(monthly
mean
temperature)
in
north
China
is
generally
under
0
°C
in
winter,
and
the
potential
evapotranspiration
cannot
occur,
so
the
potential
evapotranspiration
is
zero.
Under
this
condition,
the
precipitation
basically
represents
the
WDS.
From
the
precipitation
pattern
in
the
winter,
we
can
find
that
there
is
an
increasing
trend
of
precipitation
in
the
two
subregions
of
the
northwest
part
of
China,
and
is
of
a
notable
period
variation
in
NR
1
fr
om
1951
to
1998.
In
contrast,
a
remarkable
increasing
temperature
has
occurred
in
the
two
subregions.
There
is
no
obvious
trend
of
precipitation
variation,
but
a
period
of
oscil-
lation
of
about
15
years
exists
in
HR.
The
notable
increasing
trends
of
temperature
can
be
detected
in
the
three
subregions
in
winter.
In
spring,
there
is
a
humid
phase
(increasing
WDS)
in
the
three
subregions,
especially
in
NR
1
where
a
dramatic
increase
of
the
SHI
has
occurred
after
the
1980s.
Comparing
the
variation
of
precipitation
and
temperature,
we
can
see
that
the
humid
trend
resulted
fr
om
a
decreasing
temperature
and
an
increasing
precipitation
in
NR
1.
However,
in
NR
2
and
HR,
the
humid
phase
is
only
related
to
the
increasing
precipitation
because
the
increasing
potential
evapo-
transpiration
that
results
from
increasing
temperature
is
not
enough
to
offset
the
increasing
precipitation.
Therefore,
as
a
representative
parameter
of
WDS,
the
SHI
has
been
increased.
This
means
a
cold
and
humid
state
in
NR
1,
but
a
warm
and
humid
state
in
HR
and
in
NR
2.
From
Fig.
3,
we
can
also
find
that
there
are
phase
differences
in
the
relation
between
precipitation
and
temperature
during
different
periods
in
the
three
subregions.
In
NR
1,
the
trend
of
precipitation
had
been
agreeable
with
that
of
temperature
fr
om
the
middle
1950s
to
the
early
1970s.
But
after
the
1970s,
the
phase
between
them
is
obviously
reversed.
These
phase
differences
between
different
periods
also
exist
in
other
subregions.
There
are
two
positive
anomaly
periods
of
the
SHI
(humid
period)
in
HR
one
is
in
the
middle
of
the
1950s;
the
other
has
started
since
the
early
1970s.
The
latter
was
weaker
than
the
former.
In
NR
1
and
NR
2,
the
latest
humid
period
is
stronger than
others.
In
summer,
the
warming
and drying
trend
existed
before
the
middle
1980s,
and
then
a
remarkable
cold
and
humid
stage
in
NR
1,
as
a
result
of
the
increasing
precipitation
and
decreasing
temperature,
occurs.
In
NR
2,
the
distinguished
humid
period
had
occurred
after
the
1970s,
and
the
humid
state
had
weakened
during
the
middle
1980s,
then
it
has
been
intensified
once
again.
In
this
period,
the
temperature
varied
fr
om
the
notable
decreasing
temperature
to
increasing
tem-
Tem.(°C)
Humid
index
E
E
3
2
0
1
-1
-2
2
1
0
-1
-2
HR
1950
1955
1960
1965
1970
1975
Winter
0.3
0.2
-
0.0
-
0.0
-0.1
-
-0
.2
-
-0
.3
0.5
-
0.0
-0.5
-
0.15
0.10
0.05
at
0.00
E
-0.05
:p
-o.io
-0.15
0.2
0.0
00
E
-0.1
-0.2
-0.3
0.4
0.3
0.2
0.0
-0
.1
F
-0
.2
-0.3
E
1
980
1985
1990
1995
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Spring
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Summer
0.4
0.2
0.0
-0.2
-0.4
-0.6
1950
1955
1980
1965
1970
1975
1980
1985
1990
1995
Autum
Tem.(°C)
Humid
Index
6
4
2
0
-2
-6
0
1
2
1950
NR.
1
'nor
111••
-11
0.8
0.6
-
0.4
-
m
0.2
-
5
ao
-0.2
-0
.4
0.4
-
S4
0.0
E
-0.2
-
~
-0.4
-
-0
.6
1950
1955
1960
1965
1970
1975
1980
Winter
1985
1990
1995
Llilltr
rn
Th
ill1
1
11
1
J
-1-141-1-
E
E
2
p
1955
1960
1965
1970
1975
1980
1985
1990
1995
Spring
0.03
-
0.02
-
0.01
-
0.00
r
I I
I
111
III
0.01
-
0.02
-
003
0.2
-
0.0
-
o
ITT1
11_
U
1 1 1 1 1
-
1.]
-0.3
ne.
0.5
0.4
0.3
0.2
-8:1
-0.2
-0.3
-0.4
0.3
0.2
0.1
0.0
0.0
-0.2
-0.3
-0.4
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Summer
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Autum
Tem.("C)
Humid
index
Tem.("C)
Humid
index
Humid
index
E
2
0
1
-1
2
0
1
-1
-2
NR.2
1950
1955
1960
1985
1970
1975
1980
1985
1990
1995
Winter
0.2
0.1
0.0
-0.1
0C214
0.15
0.10
0.05
000
0.05
0.10
0.15
Ri°
1950
1955
1960
1965
1970
1975
1980
Spring
0.04
0.02
0.00
-0.02
-0.04
-0.06
0.08
0.
0.5
4
0.3
:
0.4
0.3
0.2
0.0
0.0
-0.1
-0.2
-0.3
0.4
0.4
0.2
0.0
-0.2
-0.4
-0.6
1985
1990
1995
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Summer
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Autum
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
Fig.
3.
Anomalies
of
surface
humid
index
in
the
three
subregions.
0.0
-0.1
-0.2
-0.3
194
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
perature
trend.
In
other
words,
the
state
of
climate
had
varied
from
cold
and
humid
(CH)
to
warm
and
humid
(WH).
In
HR,
the
SHI
has
successively
decreased
since
1950s,
relate
little
change
had
occurred
fr
om
the
middle
1960s
to
the
middle
1970s.
Then,
the
SHI
has
sharply
decreased,
and
a
sustained
drying
trend
of
the
SHI
has
been
formed
in
HR
of
summer.
Meanwhile,
we
found
that
although
there
was
a
little
increasing
precipitation,
the
SHI
has
decreased
along
with
the
successively
increased
temperature
in
that
area.
There-
fore,
the
surface
hydrological
variables
(such
as
SHI)
are
greatly
affected
by
surface
climate
variables
like
precipitation
and
surface
temperature.
In
fall,
there
has
been
a
rapid
increase
of
SHI
in
NR
1
after
the
1970s.
The
temperature
and
precip-
itation
have
rapidly
increased
in
the
same
period.
This
indicated
that
though
increasing
temperature
can
decrease
the
moisture
of
land
surface,
the
SHI
in
the
area
has
increased
mainly
due
to
rich
precipitation.
In
NR
2,
the
SHI
has
sustained
small
values
for
a
long
period
after
the
1980s,
and
reached
its
historical
minimum.
The
decrease
of
precipitation
and
increase
of
temperature
may
be
the
main
reason.
However,
the
cold
and
humid
(CH)
state
was
an
obvious
characteristic
of
this
area
before
the
1980s
because
of
increasing
precipitation
and
decreasing
temper-
ature.
In
HR,
the
negative
anomaly
of
the
SHI
since
the
middle
1960s
changed
into
positive
anomaly;
in
other
words,
the
present
WDS
is humid.
The
temper-
ature
variation
gives
the
remarkable
increasing
trend,
which
results
in
a
warm
and
humid
surface
state
in
HR
at
present.
Based
on
the
analyses
above,
some
conclusions
can
be
reached.
In
NR
2,
the
drought
(negative
anomaly
of
the
SHI)
of
land
surface
has
occurred
in
fall,
but
in
HR,
it
has
occurred
in
summer.
These
facts
were
distinguished
because
of
the
decrease
of
precip-
itation
and
the
increase
of
surface
temperature
in
the
region.
In
other
words,
the
decrease
of
precipitation
resulted
in
the
forming
of
the
drought,
the
regional
warming
further
intensifies
the
surface
drought,
and
then
the
drying
land
surface
has
positively
affected
the
atmosphere.
The
complicated
interaction
between
the
surface
hydrological
processes
and
atmosphere
in
the
arid
and
semi
-arid
area
of
north
China
needs
to
be
investigated
deeply
in
the
future.
In
HR,
the
drought
of
land
surface
mainly
occurred
in
summer
because
of
the
decrease
of
precipitation
and
the
increase
of
sur-
face
temperature.
It
should
be
indicated
that
the
anomaly
of
the
SHI
in
fall
changed
from
a
negative
to
a
positive
value
after
the
middle
1980s,
which
means
changing
the
drought
state
into
the
wet
state
in
HR.
Manabe
(1981)
showed
that
the
soil
moisture
in
middle
latitude
would
decrease
under
2
x
CO
2
be-
cause
of
the
decrease
in
precipitation
and
increase
in
surface
air
temperature
in
the
zone.
Therefore,
we
see
that
the
drought
in
HR
and
in
the
east
part
of
north-
west
China
is
oppositely
correlated
with
the
regional
warming.
In
that
case,
whether
the
regional
warming
in
the
subregions
resulted
fr
om
under
2
x
CO
2
con-
dition
needs
further
discussion.
4.
The
test
of
the
seasonal
SHI
trend
in
the
three
subregions
The
results
above
show
notable
trends
of
the
SHI
in
different
seasons
and
different
subregions.
How-
ever,
the
credibility
of
the
results
needs
to
be
tested.
In
this
section,
the
trends
of
different
seasons
in
the
three
subregions
were
tested
by
the
M
—K
method
(Mann
Kendall
method).
Fig.
4
gives
the
test
curves
of
the
SHI
and
the
temperature
in
various
seasons
in
the
three
subre-
gions.
As
shown
in
Fig.
4,
there
is
a
notable
increase
in
the
temperature
trend
in
all
three
sub-
regions
in
winter,
and
they
could
pass
the
signifi-
cance
of
95%.
However,
there
is
the
increasing
trend
in
precipitation
in
NR
2.
Although
there
is
the
positive
anomaly
of
the
SHI
in
HR
during
spring,
this
wet
trend
did
not
pass
the
significance
of
95%
(the
absolute
value
is
more
than
2),
and
the
increas-
ing
temperature
passed
the
significance
of
95%.
Meanwhile,
both
the
wet
and
the
warm
trends
of
temperature
during
spring
passed
the
significance
of
95%
in
NR
1
and
NR
2.
In
summer,
the
drying
trend
of
land
surface
in
HR
can
pass
the
test
of
significance,
and
there
was
no
obvious
trend
of
temperature
variation.
The
WDS
in
NR
2
succes-
sively
maintained
a
cold
and
wet
surface
state.
In
NR
2,
the
decreasing
trend
of
temperature
passed
the
significance
of
95%,
but
the
wet
trend
of
the
WDS
did
not.
In
fall,
the
warming
and drying
trends
of
WDS
in
HR
and
NR
2
passed
the
significance
of
95%.
The
warm
and
wet
states
are
the
main
char-
acteristics
of
NR
1.
6
4
2
0
-2
-4
-6
1950
Winter
.--'"
.-"
6
4
2
0
-2
1950
1960 1970 1980 1990
.•
Spri
4
2
0
-2
-4
-6
-8
1950
4
2
0
-2
-4
-6
-8
1950
1960 1970 1980 1990
-'‘
Summer
=
. .
1960 1970 1980 1990
Autumn
=
. . . .
1960 1970
HR
1980 1990
6
4
2
0
-2
-4
-6
1950
Winter
4
2
0
-2
1950
1960 1970 1980 1990
E
Spring
0
-2
-4
-6
8
1950
6
4
2
0
-2
-4
1950
1960 1970 1980 1990
E
--
Summer
1960
I I
1970 1980 1990
Autumn
=
/A
1960 1970
NR.1
I I
1980 1990
6
4
2
0
-2
-4
-6
1950
Winter
6
4
2
0
-2
-4
1960 1970 1980 1990
Spring
1950
2
0
-2
-4
-6
1950
1960 1970 1980 1990
Summer
••
......
4
2
0
-2
-4
6
1950
1960 1970 1980 1990
=
Y
.
-
:
'-•
.
.
...
.
Autumn
'
-
-*
1960 1970
NR.2
1980
Fig.
4.
Annual
trends
of
SHI
in
the
three
subregions
in
different
seasons
by
the
M
—K
method
(solid
line
=
SHI;
dotted
line
=
temperature).
1990
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
196
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
5.
The
geographical
variation
of
SHI
The
geographical
variation
of
the
SHI
will
help
us
to
understand
the
spatial
variation
of
the
interaction
between
the
surface
hydrological
processes
and
the
atmosphere.
In
this
section,
the
geographical
variation
of
the
SHI
was
given
in
Fig.
5
by
the
M
—K
method.
In
order
to
highlight
the
interannual
variation
trend
in
the
arid
and
semiarid
areas
of
north
China
and
the
difference
between
them,
and
that
in
the
humid
area
of
south
China,
the
geographical
variation
in
south
of
China
have
also
been
analyzed,
and
the
original
data
was
processed
by
a
9
-year
running
mean.
From
Fig.
5,
some
evidences
show
the
large
geographical
differ-
ences
and
the
seasonal
differences
of
the
SHI;
the
SHI
is
decreasing
in
most
areas
in
the
east
of
100°E
and
in
the
north
of
35
°N.
This
means
a
drying
trend
of
land
surface
in
these
areas.
All
drought
trends
passed
the
significance
of
95%
(the
absolute
value
of
the
number
of
the
contour
in
the
pattern
exceeds
2).
An
interesting
fact
is
that
there
is
an
opposite
variation
trend
in
the
two
sides
of
100°E
in
the
north
part
of
35°N
—the
drought
in
the
east
and
the
wet
state
in
the
west.
Further,
the
geographical
varia-
tion
in
the
fl
ood
season
was
investigated
in
Fig.
5.
The
50\
40\
30\
20\
remarkable
drought
of
land
surface
appeared
in
the
east
part
of
northwest
and
northeast,
and
in
HR,
the
wet
trend
in
the
central
and
western
part
of
northwest.
Meanwhile,
the
seasonal
difference
of
geographical
variation
has
been
analyzed.
As
mentioned
above,
precipitation
in
winter
basically
depicts
the
WDS.
Because
many
analyses
of
the
characteristics
of
the
precipitation
in
the
season
in
China
have
been
per-
formed,
we
will
not
give
a
result
for
winter.
Fig.
6
shows
the
trends
of
WDS
in
three
seasons
by
the
M
—K
method.
The
results
indicate
that
a
smaller
extent
of
the
drought
in
spring
than
in
other
seasons
in
north
China.
Its
position
is
more
northward,
and
an
obvious
drought
trend
appeared
in
most
areas
of
the
northern
part
of
the
northeast
part
of
China
and
in
the
east
of
HR.
The
annual
geographical
variation
of
the
SHI
is
roughly
consistent
with
that
in
the
fl
ood
season.
The
notable
drought
trends
of
land
surface
occurred
in
HR,
in
the
east
part
of
northwest,
in
the
south
part
of
northeast
and
in
the
north
part
of
Xingjiang.
Generally
speaking,
there
are
wet
trends
of
land
surface
in
the
central
northwest
and
the
north
part
of
northeast
China
in
the
summer.
In
fall,
except
for
the
wet
trends
ofland
surface
in
the
west
part
(NR
1)
and
the
plateau
region,
there
are
drought
trends
in
the
rest
regions
of
north
China.
The
".,
------
S
.,
I
0
'‘.
0
0'
0
0
0
/
I I I I
80E
90E
100E
1
10E
120E 130E 140E
Fig.
5.
Annual
trend
of
SHI
in
China
from
1951
to
1998
(solid
line
=
drying;
dashed
line
=wetting).
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
197
50N
40N
30N
20N
50N
40N
30N
20N
Spring
0
r-'i
,
,
-
_
II/
a
80E
90E
100E
110E
120E 130E
140E
/
—•
—•
Summer
O
C12
'
80E
90E
100E
110E 120E 130E
140E
50N
-
Autumn
40N
30N
20N
-
/'s
11
l
I/
/
'V
I
0
0
80E
90E
100E 110E
120E
130E 140E
Fig.
6.
Annual
trend
of
SHI
in
different
seasons
in
China
from
1951
to
1998
(the
order
is
the
same
as
in
Fig.
5).
198
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
Table
1
Wet
and
dry
phases
in
different
stations
Beijing
Huhehaote
Dry
Phase
(<300
mm)
1845-1885
1897
—1949
1964
—present
Wet
phase
(>
300
mm)
1886-1896
1950-1963
Dry
Phase
(<150
mm)
1919-1954
1964
—present
Wet
phase
(
150
imn)
1955-1963
Xian
Wulumuqi
Dry
Phase
Wet
phase
Dry
Phase
Wet
phase
(<150
mm)
(>
150
mm)
(<40
mm)
(>
40
mm)
1927
—1945
1946-1967 1912-1937
1938-1962
1968
—1978
1979-1987
1963-1978
1979
—present
1988
—present
center
of
the
drought
was
located
in
the
Weihe
valley
and
the
lower
reaches the
Yellow
River
valley.
6.
Long-term
variation
of
surface
hydrological
budget
in
two
typical
areas
In
the
analysis,
the
differences
between
the
obser-
vational
precipitation
and
evaporation
calculated
by
Gao's
method
can
be
used
for
representing
the
WDS:
V
w
=P
E,
where
P
is
observational
monthly
pre-
cipitation
and
E
evaporation.
The
characteristics
of
the
WDS
were
discussed
through
analyzing
V
w
varia-
tion.
The
data
is
from
the
long-term
instrumental
databases
in
China
(Tao
et
al.,
1991).
Because
there
are
no
enough
stations
in
the
arid
area,
several
stations
were
added
to
represent
the
three
subregions
(see
Table
1).
600
13.5
Temperature
500
P
-E
_
13.0
E
400
Moist
phase
Moist
Vase
-
12.5
.11
IA ‘1,*
.
E
Lu
d.
300
NO%
11
,
4 0
W
l• r%
/
to/
4
12.0
1
2
N
4
4I
4
51
•••
11.5
200
11.0
1840
1860
1880
1900
1920 1940
1960
1980
2000
Beijing
80
70
60
E
50
a
40
Lu
30
P
-E
20
1840 1860 1880
%.*
r
a
Moiqt
Moist
9
8
-
7
6
5
1900
1920 1940 1960 1980
2000
Wuloumuqi
Fig.
7.
Annual
variation
of
w(=P
E)
in
Beijing
and
Wuloumuqi.
O
a)
a)
E
a)
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
199
Fig.
7
shows
the
V
w
variation
of
the
9
-year
running
mean
in
the
Beijing
station
and
Wulumuqi.
From
Fig.
7,
there
is
a
periodic
variation
of
dry
and
wet
in
the
Beijing
station,
the
representative
of
HR.
But
the
duration
of
dry
phase
and
wet
phase
is
different.
In
general,
the
duration
of
dry
phase
is
about
40
years,
which
is
longer
than
the
duration
of
wet
phase
by
about
29
years.
Also,
we
can
find
that
there
are
three
dry
periods
and
two
wet
periods
for
150
years;
the
present
state
of
HR
is
in
the
latest
dry
phase
since
the
middle
1960s.
This
characteristic
can
be
found
in
other
observational
sites
of
HR.
But
in
the
Xian
station
and
in
Wulumuqi,
the
representatives
of
NR
1
and
NR
2,
respectively,
the
duration
of
dry
phase
and
wet
phase
is
about
20
years.
Comparing
with
HR,
the
duration
is
shorter
and
the
fr
equency
is
higher.
It
is
noted
that
the
present
state
of
WDS
in
NR
1
is
located
in
the
wet
phase,
which
is
opposite
with
the
phase
in
NR
2.
The
above
facts
can
be
found
in
Table
1.
From
the
analyses
of
historical
data
above,
there
are
remarkable
periodic
and
regional
variations
in
the
arid
and
semi
-arid
areas.
What
have
caused
the
present
drought
in
north
China?
Some
ideas
suggested
that
the
drought
is
mainly
due
to
increasing
human
activities
such
as
440
420
400
380
io
a
360
w
340
Precipitation(mm)
increasing
emission
of
CO
2
and
sulphur
(Manabe,
1981;
Wetherald
and
Manabe,
1999).
However,
from
the
analyses,
the
state
of
WDS
in
most
parts
of
north
China
were
located
in
the
natural
dry
phase;
the
main
reason
for
this
should
be
the
decreasing
precipitation
in
the
area.
Nevertheless,
the
increasing
temperature
intensified
the
drought.
The
relationship
between
pre-
cipitation,
temperature
and
evaporation,
drawn
from
the
temperature
and
precipitation
in
Beijing,
was
given
in
Fig.
8.
The
results
show
less
precipitation
but
higher
evaporation,
caused
by
high
surface
air
temperature
in
the
last
30
years.
Also,
we
found
that
the
present
evaporation
reaches
its
highest
level
since
1840,
and
the
precipitation
is
still
in
the
low
phase.
Recently,
the
small
increase
in
precipitation
is
about
100
mm
in
Beijing,
while
evaporation
has
increased
about
55
mm;
the
difference
between
them
is
45
mm.
Generally
specking,
the
P
—E
in
the
wet
phase
should
exceed
400
mm.
The
difference
between
recent
precipitation
and
evaporation
is
the
one-fourth
of
that
value,
so
little
increase
of
precipitation
is
not
enough
to
make the
land
surface
wet.
Meanwhile,
from
the
left
pattern
of
Fig.
8,
a
positive
correlation
between
the
surface
air
temper-
ature
and
evaporation
has
clearly
existed
since
the
middle
1960s,
which
is
a
good
evidence
that
increas-
-
Evaporation
"
Temperature
-
14.0
13.5
13.0
12.5
12.0
11.5
320
11.0
184018601880190019201940196019802000
1000
900
800
700
600
500
184018601880190019201940196019802000
420
o
E
400
2
0
O
IP
a.
ci,
a
380
E
as
>
Lu
360
660
640
E
620
600
.2-
580
0
2
_
560
540
14.0
-
13.5
-
13.0
-
12.5
-
12.0
-
11.5
1980
11.0
2000
1980
2000
Fig.
8.
Annual
variation
of
evaporation,
temperature
and
precipitation
in
Beijing.
200
Z.
Ma,
C.
Fu
/
Global
and
Planetary
Change
37
(2003)
189-200
ing
temperature
intensifies
the
drought.
However,
since
evaporation
is
controlled
by
many
variables
such
as
net
surface
radiation
and
soil
moisture,
etc.,
these
will
be
discussed
in
another
paper.
7.
Summary
The
above
results
indicate
that
the
SHI
is
a
reason-
able
parameter
to
depict
the
wet
and/or
dry
state
on
the
land
surface
based
on
its
solid
physical
basis
consider-
ing
two
fundamental
component
of
the
hydrological
budget
in
the
land
surface.
Some
conclusions
were
given
as
follows:
In
the
three
subregions,
there
is
the
notable
inter
-
annual
variation
of
the
SHI
and
the
evident
regional
differences.
Opposite
phases
of
interannual
SHI
variation
exist
between
HR
and
NR
1
and
NR
2.
There
was
a
wet
trend
of
land
surface
in
NR
1
in
the
past
20
years,
but
there
were
dramatic
dry
trends
of
land
surface
in
NR
2
and
HR.
There
is
a
large
difference
of
interannual
SHI
variation
trend
between
seasons.
In
HR,
the
drought
trend
occurred
mainly
in
summer;
in
NR
2,
it
occurred
mainly
in
fall.
There
has
been
the
wet
trend
of
land
surface
in
NR
1
in
spring
and
fall.
The
phases
related
to
the
surface
temperature
trend
are
opposite
between
two
seasons
—cold
and
wet
in
spring;
warm
and
wet
in
fall.
The
cold
and
wet
trend
of
land
surface
is
the
interannual
character
in
NR
2
in
spring
and
summer.
The
drought
trends
of
land
surface
occurred
in
most
parts
of
north
China
Except
in
NR
1.
These
facts
can
be
validated
by
the
interannual
trends
of
fl
ood
season.
The
intensity
of
the
drought
of
land
surface
varies
in
the
three
subregions.
Analyses
of
the
long-term
data
show
that
there
are
droughts
in
HR
and
NR
2
at
present
that
are
due
to
the
decreasing
precipitation
rather than
increasing
temperature,
and
that
there
is
a
wet
phase
in
NR
1
due
to
increasing
precipitation.
The
influence
of
the
global
warming
on
the
region
WDS
is
an
important
problem
for
future
study
on
the
interaction
between
the
surface
hydrological
processes
and
climate
variation.
The
above
analyses
show
a
notable
positive
correlation
between
the
surface
air
temperature
and
evaporation.
Because
evaporation
is
controlled
by
several
processes,
the
relationship
between
evaporation
and
other
factors
needs
further
study
in
order
to
understand
the
influence
of
regional
warming
on
surface
hydrological
processes.
Acknowledgements
We
acknowledge
the
comments
from
reviewers.
This
research
was
supported
by
the
National
key
Planning
Development
for
Basic
Research
(Grant
No.
G1999043400),
the National
Nature
Science
Founda-
tion
of
China
(Grant
No.
40145022),
and
the
Key
Project
of
Knowledge
Innovation
Engineering
of
Chinese
Academy
of
Sciences (Grant
No.
KZCX1-10-07).
References
Fu,
C.,
1994.
An
aridity
trend
in
China
in
association
with
global
warming.
In:
Zepp,
R.G.
(Ed.),
Climate
Biosphere
Interaction:
Biogenic
Emission
and
Environmental
Effects
of
Climate
Change.
John
Wiley
&
Sons,
New
York,
pp.
1-17.
Fu,
C.,
Wang,
Q.,
1992.
The
definition
and
detection
of
the
abrupt
climatic
change.
Chin.
J.
Atmos.
Sci.
16
(4),
482-493.
Hulme,
M.,
Marsh,
R.,
Jones,
P.D.,
1992.
Global
changes
in
a
humid-
ity
index
between
1931-60
and
1961-90.
Clim.
Res.
2,
1-22.
Ma,
Z.,
1999.
The
relationship
between
the
soil
moisture
and
cli-
matic
variability
over
East
China
and
a
model
used
for
retrieving
soil
moisture,
PhD
dissertation,
Institute
of
Atmospheric
Physics
in
Chinese
Academy
of
Sciences,
p.
121.
Manabe,
S.,
1981.
Summer
dryness
due
to
an
increase
of
atmos-
pheric
CO
2
concentration.
Clim.
Change
3,
347-386.
Snyers,
R.,
1990.
On
the
statistical
analysis
of
series
of
observa-
tions,
Technical
Note,
143,
WMO,
Geneva,
11,
1990.
Tao,
S.,
Fu,
C.,
Zeng,
Z.,
et
al.,
1991.
Two
long-term
instrumental
climatic
data
bases
of
the
People's
Republic
of
China.
In:
Kai-
ser,
D.P.
(Ed.),
Carbon
Dioxide
Information
Analysis
Center.
Oak
Ridge
National
Laboratory,
Oak
Ridge,
TN.
Thomas,
A.,
2000.
Spatial
and
temporal
characteristics
of
potential
evapotranspiration
trends
over
China.
Int.
J.
Climatol.
20,
381-396.
Wei,
F.,
Cao,
H.,
1998.
Regional
characteristics
of
drought
anoma-
li
es
in
North
China.
Chin.
Q.
Appl.
Meteorol.
9
(2),
205-211.
Wetherald,
R.T.,
Manabe,
S.,
1999.
Detectability
of
summer
dryness
caused
by
greenhouse
warming.
Clim.
Change
43,
495-511.
Yan,
Z.,
1995.
Some
chaotic
features
of
the
wet/dry
fluctuations
in
North
China.
Acta
Meteorol.
Sin.
53
(2),
232-237.
Yan,
Z.,
1999.
Interdecadal
oscillations
of
Precipitation
in
North
China
and
its
relation
with
global
temperature
change.
Q.
J.
Appl.
Meteorol.
10
(special
issue),
16-21.
Yang,
H.,
Song,
Z.,
1999.
Multiple
time
scales
analysis
of
water
resources
in
North
China.
Plateau
Meteorol.
18
(4),
496-508.