Water,
Год журнала:
2022,
Номер
14(22), С. 3771 - 3771
Опубликована: Ноя. 20, 2022
The
Upper
Krishna
Basin
in
Maharashtra
(India)
is
highly
vulnerable
to
floods.
This
study
aimed
generate
a
flood
susceptibility
map
for
the
basin
using
Frequency
Ratio
and
Statistical
Index
models
of
analysis.
hazard
inventory
was
created
by
370
locations
plotted
ArcGIS
10.1
software.
259
(70%)
were
selected
randomly
as
training
samples
analysis
models,
validation
purposes,
remaining
111
(30%)
used.
Flood
analyses
performed
based
on
12
conditioning
factors.
These
elevation,
slope,
aspect,
curvature,
Topographic
Wetness
Index,
Stream
Power
rainfall,
distance
from
river,
stream
density,
soil
types,
land
use,
road.
model
revealed
that
38%
area
high-
very-high-flood-susceptibility
class.
precision
confirmed
receiver
operating
characteristic
under
curve
value
method.
showed
66.89%
success
rate
68%
prediction
model.
However,
provided
an
82.85%
83.23%
rate.
comparative
most
suitable
mapping
flood-prone
areas
Basin.
results
obtained
this
research
can
be
helpful
disaster
mitigation
preparedness
Ecological Informatics,
Год журнала:
2023,
Номер
79, С. 102427 - 102427
Опубликована: Дек. 11, 2023
Flooding
is
the
most
frequent
and
damaging
threat
to
The
United
Nations
Educational,
Scientific,
Cultural
Organization
(UNESCO)
World
Heritage
Sites,
has
been
exacerbated
by
climate
change.
Hoi
An
Ancient
Town,
one
of
world's
cultural
heritage
sites
in
Vietnam,
facing
inundation
risks
from
increasing
extreme
flood
events
due
anthropogenic
natural
processes.
This
study
combines
Geospatial
Information
System
(GIS),
remote
sensing
data,
landscape
ecology
metrics
assess
risk
City.
analysis
includes
(1)
classification
land-use/land-cover
(LULC)
using
Sentinel-2
data;
(2)
computation
a
index
hazard
characteristics,
physical
demographic
exposures,
socioeconomic
vulnerabilities,
together
with
adaptive
capacity
spatial
metrics,
Fuzzy-Analytic
Hierarchy
Process
(AHP)
method
determine
weight
each
ranking;
(3)
determination
for
UNESCO
Sites
City
surrounding
areas.
By
comparing
model
outputs
historical
locations
(N
=
330),
finds
that
>75%
past
floods
occurred
at
high
or
very
high-risk
areas
forecasted
model.
Results
verified
true
incidents
show
hotspots
are
concentrated
city's
sites,
where
human
changes
have
impacted
structure.
mapping
obtained
synthesis
analysis,
encompasses
hazard,
exposure,
vulnerability,
offering
holistic
knowledge
help
reduce
manage
Sites.
can
be
applicable
other
similar
characteristics.
Journal of Water Resources Planning and Management,
Год журнала:
2023,
Номер
149(10)
Опубликована: Авг. 4, 2023
Floods
have
claimed
the
lives
of
countless
people
and
caused
significant
property
damage
in
many
countries,
putting
their
livelihoods
jeopardy.
The
Vembanad
lake
system
(VLS)
Kerala,
India,
has
faced
adverse
mishappening
during
2018,
2019,
2021
floods
state
due
to
torrential
rainfall.
goal
this
research
is
construct
effective
decision
tree–based
machine
learning
models
such
as
adaptive
boosting
(AdaBoost),
random
forest
(RF),
gradient
machines
(GBMs),
extreme
(XGBoost)
for
integrating
data,
processing,
generating
flood
susceptibility
maps.
There
are
18
conditioning
parameters
considered,
which
include
seven
categories
11
numerical
data.
These
categorical
data
were
converted
bringing
total
amount
input
61.
recursive
feature
elimination
(RFE)
was
utilized
selection
technique,
a
22
layers
chosen
feed
into
generate
efficiencies
evaluated
using
receiver
operating
characteristic
(ROC)–area
under
ROC
curve
(AUC),
F1
score,
accuracy,
kappa.
According
results,
performance
all
four
demonstrated
practical
application;
however,
XGBoost
fared
well
terms
model's
metrics.
For
testing
set,
ROC-AUC
values
XGBoost,
GBM,
AdaBoost
0.90,
whereas
it
0.89
RF.
accuracy
varied
significantly
among
models,
with
scoring
0.92,
followed
by
GBM
(0.88),
RF
(0.87),
(0.87).
As
result,
map
may
be
early
mitigation
actions
future
floods,
land-use
planners
emergency
managers,
assisting
reduction
risk
regions
prone
hazard.
Water,
Год журнала:
2022,
Номер
14(22), С. 3771 - 3771
Опубликована: Ноя. 20, 2022
The
Upper
Krishna
Basin
in
Maharashtra
(India)
is
highly
vulnerable
to
floods.
This
study
aimed
generate
a
flood
susceptibility
map
for
the
basin
using
Frequency
Ratio
and
Statistical
Index
models
of
analysis.
hazard
inventory
was
created
by
370
locations
plotted
ArcGIS
10.1
software.
259
(70%)
were
selected
randomly
as
training
samples
analysis
models,
validation
purposes,
remaining
111
(30%)
used.
Flood
analyses
performed
based
on
12
conditioning
factors.
These
elevation,
slope,
aspect,
curvature,
Topographic
Wetness
Index,
Stream
Power
rainfall,
distance
from
river,
stream
density,
soil
types,
land
use,
road.
model
revealed
that
38%
area
high-
very-high-flood-susceptibility
class.
precision
confirmed
receiver
operating
characteristic
under
curve
value
method.
showed
66.89%
success
rate
68%
prediction
model.
However,
provided
an
82.85%
83.23%
rate.
comparative
most
suitable
mapping
flood-prone
areas
Basin.
results
obtained
this
research
can
be
helpful
disaster
mitigation
preparedness