Frontiers in Plant Science,
Journal Year:
2024,
Volume and Issue:
15
Published: April 16, 2024
Amidst
the
backdrop
of
global
climate
change,
it
is
imperative
to
comprehend
intricate
connections
among
surface
water,
vegetation,
and
climatic
shifts
within
watersheds,
especially
in
fragile,
arid
ecosystems.
However,
these
relationships
across
various
timescales
remain
unclear.
We
employed
Ensemble
Empirical
Mode
Decomposition
(EEMD)
method
analyze
multifaceted
dynamics
water
vegetation
Bosten
Lake
Watershed
multiple
temporal
scales.
This
analysis
has
shed
light
on
how
elements
interact
with
revealing
significant
insights.
From
March
October,
approximately
14.9–16.8%
areas
permanent
were
susceptible
receding
drying
up.
Both
annual
monthly
values
Lake’s
level
area
exhibited
a
trend
initial
decline
followed
by
an
increase,
reaching
their
lowest
point
2013
(1,045.0
m
906.6
km2,
respectively).
Approximately
7.7%
vegetated
showed
increase
Normalized
Difference
Vegetation
Index
(NDVI).
NDVI
volatility
was
observed
23.4%
areas,
primarily
concentrated
southern
part
study
near
Bosten.
Regarding
components
(6
<
T
24
months),
temperature,
3-month
cumulative
NDVI,
3-month-leading
precipitation
strongest
correlation
changes
area.
For
interannual
(T≥
precipitation,
temperature
displayed
most
robust
alterations
In
both
components,
had
negative
impact
area,
while
exerted
positive
effects.
Through
comparative
analysis,
this
reveals
importance
periodicity
developing
adaptive
strategies
for
achieving
Sustainable
Development
Goals
dryland
watersheds.
introduces
methodology
dissecting
trends
scale
lake
links
variations
different
The
inherent
correlations
uncovered
research
can
serve
as
valuable
guidance
future
investigations
into
regions.
Scientific Reports,
Journal Year:
2023,
Volume and Issue:
13(1)
Published: Feb. 16, 2023
Abstract
Large-scale
soil
salinity
surveys
are
time-costly
and
labor-intensive,
it
is
also
more
difficult
to
investigate
historical
salinity,
while
in
arid
semi-arid
regions,
the
investigation
of
spatial
temporal
characteristics
can
provide
a
scientific
basis
for
prevention
With
this
objective,
study
uses
multi-source
data
combined
with
ensemble
learning
Google
Earth
Engine
build
monitoring
model
observe
evolution
salinization
Werigan–Kuqa
River
Oasis
from
1996
2021
analyze
driving
factors.
In
experiment,
three
models,
Random
Forest
(RF),
Extreme
Gradient
Boosting
(XGBoost),
Light
Machine
(LightGBM),
were
established
using
collected
field
different
years
some
environmental
variables,
After
accuracy
validation
model,
XGBoost
had
highest
prediction
area,
RMSE
17.62
dS
m
−1
,
R
2
0.73
RPIQ
2.45
test
set.
after
Spearman
correlation
analysis
Electrical
Conductivity
(EC)
we
found
that
near-infrared
band
original
band,
DEM
topographic
factor,
vegetation
index
based
on
remote
sensing,
EC
strong
correlation.
The
distribution
generally
characterized
by
good
west
north
severe
east
south.
Non-salinization,
light
salinization,
moderate
gradually
expanded
southward
eastward
interior
western
oasis
over
25
years.
Severe
very
shifted
northern
edge
eastern
southeastern
desert
areas
during
saline
soils
class
distributed
most
part
area
as
well
smaller
1996,
shrinking
size
discontinuous
2021.
terms
change,
non-salinized
increased
198.25
1682.47
km
level
decreased
5708.77
2246.87
overall,
overall
improved.
