Abstract.
To
enhance
the
timeliness
and
accuracy
of
spatial
prediction
co-seismic
landslides,
we
propose
an
improved
three-stage
strategy
developed
a
corresponding
hazard
assessment
software
named
Mat.LShazard
V1.0.
Based
on
this
software,
evaluate
applicability
in
six
earthquake
events
that
have
occurred
near
Sichuan
Yunnan
region
including
Wenchuan,
Ludian,
Lushan,
Jiuzhaigou,
Minxian
Yushu
earthquakes.
The
results
indicate
first
stage
(within
half-hour
earthquake),
except
for
2013
earthquake,
AUC
values
modelling
performance
other
five
are
above
0.8.
Among
them,
value
Wenchuan
is
highest,
reaching
0.947.
can
meet
requirements
emergency
rescue
with
immediately
obtaining
overall
predicted
information
possible
coseismic
landslide
locations
quake-affected
area.
In
second
third
stages
(Within
12
hours
quake),
improvement
data
quality,
ability
model
based
entire
database
gradually
improved.
database,
exceeds
0.9,
indicating
very
high
accuracy.
Whether
or
(After
3
days
seismic
event),
area
(Ap)
good
agreement
observed
(Ao).
However,
incomplete
meizoseismal
area,
Ap
much
smaller
than
Ao.
When
complete
built,
nearly
identical
to
This
study
provides
new
application
tool
disaster
prevention
mitigation
different
rescue,
temporary
resettlement,
latereconstruction
after
major
earthquake.
Rock Mechanics Bulletin,
Год журнала:
2023,
Номер
2(2), С. 100038 - 100038
Опубликована: Фев. 11, 2023
The
evaluation
of
engineering
rock
mass
quality
is
fundamental
work
for
the
activities
mass.
increasing
scale
necessitates
higher
intelligence,
timeliness,
and
accuracy
in
evaluation.
As
core
aspects
evaluation,
structural
characteristics,
mechanical
classification
have
been
innovated
recent
years.
non-contact
measurement
technology
structure
rapid
interpretation
information
enables
intelligent
extraction
analysis
parameters.
modular
backpack
laboratory
system
mechanics
provides
an
effective
means
to
acquire
parameters
on-site
conveniently.
theory
statistical
(SMRM)
integrates
various
factors
such
as
properties,
geological
environment,
disturbance,
providing
a
theoretical
basis
accurately
evaluating
weakening
anisotropy
cloud
computing
platform
established
based
on
SMRM
can
provide
technical
support
calculation
instant
quality.
development
method
altering
conventional
state
qualitative
semi-quantitative
quality,
supporting
realization
construction
with
intellectualization
informatization.
Ecological Indicators,
Год журнала:
2023,
Номер
157, С. 111278 - 111278
Опубликована: Ноя. 24, 2023
Geological
disasters
occur
frequently
in
hilly
areas
of
Sichuan
province,
and
the
development
rural
settlement
production
life
may
be
affected
by
geological
disasters.This
study
aims
to
assess
spatial
vulnerability
settlements
Province
through
Back-propagation
neural
network,
find
main
factors
that
really
affect
settlements,
analyze
distribution
characteristics
Province,
methods
reduce
settlements.
Through
research,
it
can
found
three
affecting
are
climate
factors,
soil
erosion
degree,
hazard
hidden
danger
degree.
The
quantitative
factor
data
is
ratio
annual
precipitation
population
settlements,the
higher
precipitation,
vulnerability.
second
largest
influencing
degree
erosion,the
more
severe
erosion,
districts
counties.
third
major
hazards,
Other
natural
environment
also
have
a
certain
impact
on
while
socio-economic
almost
no
Finally,
relationship
between
its
location
obtained
from
relevant
all
sample
counties,
combined
with
Sichuan,
summarized:
closer
counties
fracture
zone
Longmen
Mountain
Longquan
Mountain,
settlements.The
west
than
east,
there
little
difference
north
south.
Then,
based
above
achievements,
this
proposes
strategies
suggestions
providing
reference
materials
support
for
local
planning
departments
work.
Remote Sensing,
Год журнала:
2023,
Номер
15(5), С. 1418 - 1418
Опубликована: Март 2, 2023
Bivariate
data-driven
methods
have
been
widely
used
in
landslide
susceptibility
analysis.
However,
the
names,
principles,
and
correlations
of
bivariate
are
still
confused.
In
this
paper,
first
clarified
based
on
a
comprehensive
in-depth
survey.
A
total
eleven
prevalent
identified,
nominated,
elaborated
general
framework,
constituting
well-structured
method
family.
We
show
that
all
depend
empirical
conditional
probabilities
occurrence
to
calculate
susceptibilities,
either
exclusively
or
inclusively.
It
is
those
eight
“conditional-probability-based”
methods,
which
probabilities,
particularly
strongly
correlated
principle,
therefore
expected
very
close
even
same
performance.
also
suggested
conditional-probability-based
apply
“classification-free”
modification,
factor
classifications
avoided
result
dominated
by
single
parameter,
“bin
width”.
Then,
optimization
framework
for
classification-free
modification
obtaining
optimum
results
optimizing
dominant
parameter
bin
width,
proposed.
The
open
software
Automatic
Landslide
Susceptibility
Analysis
(ALSA)
updated
implement
framework.
Finally,
case
study
presented,
confirms
theoretical
expectation
different
performance,
shows
optimal
perform
better
than
conventional
regarding
both
prediction
rate
ability
reveal
quasi-continuous
varying
pattern
sensibilities
landslides
individual
predisposing
factors.
principles
presented
provide
practical
foundations
applications
explorations
