Earth s Future,
Journal Year:
2025,
Volume and Issue:
13(1)
Published: Jan. 1, 2025
Abstract
Effective
wildfire
prevention
includes
actions
to
deliberately
target
different
causes.
However,
the
cause
of
an
increasing
number
wildfires
is
unknown,
hindering
targeted
efforts.
We
developed
a
machine
learning
model
ignition
across
western
United
States
on
basis
physical,
biological,
social,
and
management
attributes
associated
with
wildfires.
Trained
from
1992
2020
12
known
causes,
overall
accuracy
our
exceeded
70%
when
applied
out‐of‐sample
test
data.
Our
more
accurately
separated
ignited
by
natural
versus
human
causes
(93%
accuracy),
discriminated
among
11
classes
human‐ignited
55%
accuracy.
attributed
greatest
percentage
150,247
for
which
source
was
unknown
equipment
vehicle
use
(21%),
lightning
(20%),
arson
incendiarism
(18%).
Applied System Innovation,
Journal Year:
2023,
Volume and Issue:
6(4), P. 61 - 61
Published: June 25, 2023
Automated
classification
of
satellite
images
is
a
challenging
task
that
enables
the
use
remote
sensing
data
for
environmental
modeling
Earth’s
landscapes.
In
this
document,
we
implement
GRASS
GIS-based
framework
discriminating
land
cover
types
to
identify
changes
in
endorheic
basins
ephemeral
salt
lakes
Chott
Melrhir
and
Merouane,
Algeria;
employ
embedded
algorithms
image
processing.
This
study
presents
dataset
nine
Landsat
8–9
OLI/TIRS
obtained
from
USGS
9-year
period,
2014
2022.
The
were
analyzed
detect
water
levels
experience
temporal
fluctuations;
these
are
dry
most
time
fed
with
during
rainy
periods.
unsupervised
was
performed
using
GIS
through
several
modules:
‘i.cluster’
used
generate
classes;
‘i.maxlik’
maximal
likelihood
discriminant
analysis,
auxiliary
modules,
such
as
‘i.group’,
‘r.support’,
‘r.import’,
etc.,
used.
document
includes
technical
descriptions
scripts
processing
detailed
comments
on
functionalities
modules.
results
include
identified
variations
within
Algerian
part
Sahara
over
period
(2014–2022),
series
multispectral
classified
GIS.
main
strengths
high
speed,
accuracy,
effectiveness
programming
codes
monitoring.
presented
GitHub
repository,
which
contains
serves
reference
interpretation
monitoring
arid
semi-arid
areas
Africa.
Remote Sensing,
Journal Year:
2024,
Volume and Issue:
16(5), P. 884 - 884
Published: March 2, 2024
Forest
fires
are
caused
by
various
climatic
and
anthropogenic
factors.
In
Republic
of
Korea,
forest
occur
frequently
during
spring
when
the
humidity
is
low.
During
past
decade,
number
fire
incidents
extent
damaged
area
have
increased.
Satellite
imagery
can
be
applied
to
assess
damage
from
these
unpredictable
fires.
Despite
increasing
threat,
there
a
lack
comprehensive
analysis
effective
strategies
for
addressing
fires,
particularly
considering
diverse
topography
Korea.
Herein,
we
present
an
approach
automated
detection
using
Sentinel-2
images
14
areas
affected
in
Korea
2019–2023.
The
performance
deep
learning
(DL),
machine
learning,
spectral
index
methods
was
analyzed,
optimal
model
detecting
derived.
To
evaluate
independent
models,
two
different
burned
exhibiting
distinct
characteristics
were
selected
as
test
subjects.
increase
classification
accuracy,
tests
conducted
on
combinations
input
channels
DL.
combination
false-color
RNG
(B4,
B8,
B3)
damage.
Consequently,
among
DL
HRNet
achieved
excellent
results
both
regions
with
intersection
over
union
scores
89.40
82.49,
confirming
that
proposed
method
applicable
Korean
landscapes.
Thus,
suitable
mitigation
measures
promptly
designed
based
rapid
areas.
International Journal of Applied Earth Observation and Geoinformation,
Journal Year:
2024,
Volume and Issue:
132, P. 104034 - 104034
Published: July 30, 2024
The
fusion
of
remote
sensing
and
artificial
intelligence,
particularly
deep
learning,
offers
substantial
opportunities
for
developing
innovative
methods
in
rapid
disaster
mapping
damage
assessment.
However,
current
models
wildfire
burnt
area
detection
are
mostly
constrained
handling
imbalanced
training
samples
with
non-burnt
areas
oversampled,
boundary
a
mix
unburnt
pixels,
regions
varying
environmental
contexts,
leading
to
poor
model
generalizability.
In
response,
this
paper
proposes
novel
U-Net
based
model,
known
as
BiAU-Net,
which
incorporates
attention
mechanisms
well-designed
loss
function,
enabling
the
focus
on
improve
accuracy
efficiency,
especially
detecting
edges
small
areas.
Unlike
traditional
single-input
image
segmentation,
proposed
BiAU-Net
considers
temporal
changes
two
inputs,
pre-
post-fire
Sentinel-2
imagery,
enhancing
performance
across
diverse
Five
independent
from
different
continents
selected
study
cases,
one
all
five
testing,
demonstrate
generalizability
model.
We
used
Fire
Disturbance
Climate
Change
Initiative
v5.1
product
European
Space
Agency
baseline
evaluation.
experiment
results
indicate
that
BiAU-Net:
(1)
significantly
outperformed
improvements
11.56%
Overall
Accuracy,
29.08%
Precision,
7.06%
Recall,
19.90%
F1-score,
15.44%
Balanced
29.90%
Kappa
Coefficient,
28.29%
Matthews
Correlation
Coefficient
(MCC);
(2)
largely
surpassed
its
variants
most
areas;
(3)
demonstrated
good
testing
continents;
(4)
achieved
highest
overall
compared
state-of-the-art
models,
evidenced
by
F1-score
MCC
values.
GEOMATICA,
Journal Year:
2024,
Volume and Issue:
76(1), P. 100008 - 100008
Published: July 1, 2024
Wildfires
pose
an
increasing
risk
to
expanding
urban
population
centers,
and
critical
habitats
for
plant
animal
species.
Improving
current
wildland
management
strategies
are
vital
mitigating
loss
of
global
biodiversity
preventing
the
displacement
residents.
Accurate
maps
areas
burned
by
wildfires
is
a
primary
source
information
required
developing
strategies.
Advancements
in
underlying
technologies
mapping
comes
from
three
key
areas:
1)
remotely
sensed
data,
2)
cloud
geoprocessing
platforms,
3)
emerging
image
processing
algorithms.
Trends
across
these
were
explored
this
review,
addition
in-depth
discussion
comparison
optimal
usage
scenarios.
This
review
provides
crucial
insights
researchers
practitioners
keen
on
exploring
methods
that
hold
potential
improve
wildfire
area
procedures.
Hydrological Sciences Journal,
Journal Year:
2023,
Volume and Issue:
69(1), P. 139 - 148
Published: Nov. 15, 2023
Addressing
post-fire
impacts
largely
depends
on
burn
"severity."
A
singular
severity
classification
that
encompasses
the
holistic
effects
of
fire
all
ecosystem
processes
does
not
currently
exist.
Lumping
vegetation
and
soil
into
one
metric,
or
using
them
interchangeably,
can
induce
large
inaccuracies
uncertainties
in
intended
response
to
forcing.
Often,
"severity"
reflects
vegetation,
which
be
measured
through
remote
sensing.
Vegetation
is
likely
more
apropos
for
ecological
research,
whereas
relevant
hydrological
analyses.
This
paper
reviews
different
remotely
sensed
products
(mis)used
modeling,
provides
examples
when
may
(not)
match
severity,
summarizes
potential
synergistic
future
sensing
with
situ
metrics.
While
focus
this
western
United
States,
lessons
principles
apply
universally.
