Abstract.
Rainfall
is
intrinsically
connected
to
the
incidence
of
landslide
catastrophes.
Exploring
ideal
rainfall
threshold
model
(RTM)
for
an
area
in
order
determine
warning
level
(RWL)
region
daily
hazard
(LHW)
critical
precise
prevention
and
management
local
landslides.
In
this
paper,
a
method
calculating
thresholds
using
multilayer
perceptron
(MLP)
regression
proposed
453
rainfall-induced
First,
study
was
divided
into
subareas
based
on
topography
climate
conditions.
Then,
two
methods,
MLP
ordinary
least
squares
(OLS),
were
utilized
explore
optimal
RTM
each
subregion.
Subsequently,
11
factors
along
with
three
models
selected
predict
susceptibility
(LS).
Finally,
obtain
LHW
result
area,
superposition
matrix
employed
overlay
RWL
LS
prediction
results.
The
following
are
study's
findings:
(1)
RTMs
calculation
methods
different
subregions.
(2)
Three-dimensional
convolutional
neural
network
produces
more
accurate
(3)
validated
anticipated
data
July
19,
2020,
validation
results
proved
correctness
RTM.
Heliyon,
Journal Year:
2024,
Volume and Issue:
10(2), P. e24660 - e24660
Published: Jan. 1, 2024
Many
landslides
can
cause
significant
damage
to
infrastructure,
property,
and
human
life.
To
study
landslide
structure
processes,
geophysical
techniques
are
most
productive
when
employed
in
combination
with
other
survey
monitoring
tools,
such
as
intrusive
sampling.
Here,
the
integration
of
electrical
resistivity
tomography
(ERT)
seismic
refraction
(SRT)
methods
is
used
assess
Thungsong
district,
Nakhon
Si
Thammarat,
south
Thailand,
where
a
hilly
seasons
prolonged
rainfall
region.
The
2D
cross-plot
analysis
P-wave
velocity
values
obtained
by
these
two
introduced
identify
potential
landslide-prone
zones
this
results
model
reveal
detailed
image
subsurface
conditions,
highlighting
areas
low
(lower
than
600
m/s)
Ωm).
These
indicative
weak
zone
be
sliding
materials.
Moreover,
an
sampling
data
from
boreholes
also
for
calibration
validation
geological
data.
This
improve
accuracy
assessment
develop
effective
mitigation
strategies
reduce
risk
area.
In
addition
cross-plot,
volume
material
determined
difference
surface
slipping
plane
elevations.
calculation
roughly
33447.76
m
Geomatics Natural Hazards and Risk,
Journal Year:
2022,
Volume and Issue:
13(1), P. 2425 - 2441
Published: Sept. 9, 2022
Northern
Thailand
is
a
hotspot
for
landslides.
Rainfall-triggered
landslides
in
this
region
have
caused
much
suffering
and
many
fatalities.
In
work,
landslide-triggering
rainfall
threshold
proposed
based
on
data
relating
to
48
triggering
events
that
59
the
study
area.
To
account
different
mechanism
of
landslide
formation,
was
portioned
into
two
parts
duration
events.
A
split
point
3
days
chosen
as
separator
portioning
be
1)
no
longer
than
2)
days.
The
also
required
suitable
variable
antecedent
rainfalls
which
found
cumulative
over
25-day
period
(CR25)
140
mm.
Therefore,
thresholds
combining
with
event
-
(CED)
were
established
by
incorporating
CR25
mm
traditional
ED
threshold.
This
first
attempt
incorporate
difference
formation
dividing
CED
portions
duration.
introduced
shows
positive
sign
prediction,
particularly
term
false
alarm
rate,
ratio,
critical
success
index.
will
useful
warning
system
Open Geosciences,
Journal Year:
2025,
Volume and Issue:
17(1)
Published: Jan. 1, 2025
Abstract
A
recently
introduced
rainfall
threshold
for
landslide
early
warning
systems
combined
a
cumulative
with
event–duration
(ED)
threshold.
Cumulative
event–duration,
known
as
the
CED
threshold,
was
reported
to
perform
better
than
conventional
ED
However,
establishment
of
based
on
frequentist
approach
which
required
an
adequate
number
landslide-triggering
data.
An
alternative
use
data
is
non-triggering
These
events
supply
much
bigger
amount
susceptibility
model.
Although
seldom
considered,
previous
scholars
that
this
has
produced
results
events.
This
study
investigates
reliability
and
prediction
performance
The
designated
negative-CED
(CED
N
)
compared
positive-CED
P
North
Thailand,
hot
spot,
chosen
area.
proposed
assessed
from
three
skill
scores,
including
(1)
true
positive
fraction
(TPF),
(2)
false
(FPF),
(3)
predictive
value
(PPV),
their
variations
over
range
uncertainties.
Rather
possessing
lower
uncertainties
parameters,
negative
provided
compromise
predictions
TPF
FPF
scores
thresholds.
Integrating
event-based
resulted
in
significant
improvement
hence
enhanced
scores.
Keeping
mind
thresholds
were
not
established
data,
care
must
be
taken
when
using
these
it
recommended
they
should
applied
only
areas
where
are
limited.
