Agronomy,
Journal Year:
2025,
Volume and Issue:
15(4), P. 954 - 954
Published: April 14, 2025
As
global
warming
progresses,
quantifying
drought
thresholds
for
crop
yield
losses
is
crucial
food
security
and
sustainable
agriculture.
Based
on
the
CNN-LSTM
model
Copula
function,
this
study
constructs
a
conditional
probability
framework
under
future
climate
change.
It
analyzes
relationship
between
Standardized
Precipitation–Evapotranspiration
Index
(SPEI)
winter
wheat
yield,
assesses
vulnerability
of
in
various
regions
to
stress,
quantifies
The
results
showed
that
(1)
SPEI
Zhoukou,
Sanmenxia,
Nanyang
was
significantly
correlated
with
yield;
(2)
southern
eastern
higher
than
center,
western,
northern
past
(2000–2023)
(2024–2047);
(3)
there
were
significant
differences
thresholds.
loss
below
30,
50,
70
percentiles
(past/future)
−1.86/−2.47,
−0.85/−1.39,
0.60/0.35
(Xinyang);
−1.45/−2.16,
−0.75/−1.34,
−0.17/−0.43
(Nanyang);
−1.47/−2.24,
−0.97/−1.61,
0.69/0.28
(Zhoukou);
−2.18/−2.86,
−1.80/−2.36,
−0.75/−1.08
(Kaifeng),
indicating
threshold
will
reduce
future.
This
mainly
due
different
soil
conditions
Henan
Province.
In
context
change,
droughts
be
more
frequent.
Hence,
research
provide
valuable
reference
efficient
utilization
agricultural
water
resources
prevention
control
risk
change
Journal of Hydroinformatics,
Journal Year:
2024,
Volume and Issue:
26(8), P. 1852 - 1882
Published: July 30, 2024
ABSTRACT
The
objective
of
this
study
was
to
develop
a
theoretical
framework
based
on
machine
learning,
the
hydrodynamic
model,
and
analytic
hierarchy
process
(AHP)
assess
risk
flooding
downstream
Ba
River
in
Phu
Yen.
made
up
three
main
factors:
flood
risk,
exposure,
vulnerability.
Hazard
calculated
from
depth,
velocity,
susceptibility,
which
depth
velocity
were
using
susceptibility
built
namely,
support
vector
machines,
decision
trees,
AdaBoost,
CatBoost.
Flood
exposure
constructed
by
combining
population
density,
distance
river,
land
use/land
cover.
vulnerability
poverty
level
road
density.
indices
each
factor
integrated
AHP.
results
showed
that
hydraulic
model
successful
simulating
events
1993
2020,
with
Nash–Sutcliffe
efficiency
values
0.95
0.79,
respectively.
All
learning
models
performed
well,
area
under
curve
(AUC)
more
than
0.90;
among
them,
AdaBoost
most
accurate,
an
AUC
value
0.99.