Characterizing the capability of public DEMs for mapping global floodplain bathymetry
Yaling Lin,
No information about this author
Chenyu Fan,
No information about this author
Kai Liu
No information about this author
et al.
Journal of Hydrology,
Journal Year:
2025,
Volume and Issue:
unknown, P. 133205 - 133205
Published: March 1, 2025
Language: Английский
Seasonal responses of microbial communities to water quality variations and interaction of eutrophication risk in Gehu Lake
Qiqi Chen,
No information about this author
Yuxia Liu,
No information about this author
Meng Zhang
No information about this author
et al.
The Science of The Total Environment,
Journal Year:
2024,
Volume and Issue:
955, P. 177199 - 177199
Published: Nov. 2, 2024
Language: Английский
Temporal dynamics of soil salinization due to vertical and lateral saltwater intrusion at an onshore aquaculture farm
Agricultural Water Management,
Journal Year:
2024,
Volume and Issue:
306, P. 109179 - 109179
Published: Nov. 16, 2024
Language: Английский
From image-level to pixel-level labeling: A weakly-supervised learning method for identifying aquaculture ponds using iterative anti-adversarial attacks guided by aquaculture features
International Journal of Applied Earth Observation and Geoinformation,
Journal Year:
2024,
Volume and Issue:
132, P. 104023 - 104023
Published: July 15, 2024
Aquaculture
mapping
is
essential
for
monitoring
and
managing
aquaculture
resources.
However,
accurately
geotargeting
individual
ponds
from
medium-resolution
remote
sensing
imagery
remains
challenging,
convolutional
deep
learning
methods
identifying
require
labor-intensive
pixel-level
annotations.
This
paper
presents
a
novel
weakly-supervised
method
to
derive
labels
image-level
annotations
ponds.
Our
approach
uses
iterative
anti-adversarial
attacks
refine
localization
results
multi-scale
class
activation
maps
(CAMs).
The
improved
integrates
two
regularization
guided
by
features
form
joint
loss
function
adversarial
samples:
discriminative
water
region
suppression
non-aquaculture
suppression.
We
also
propose
an
feature
termed
CFNDWI
constrain
the
generate
high-quality
pseudo-labels.
As
result,
pseudo-labels
are
used
train
semantic
segmentation
networks
evaluated
performance
of
our
using
commonly-used
backbones
on
10
m
Sentinel-2
imagery.
achieves
Intersection
over
Union
(IoU)
values
0.618–0.655
pseudo-label
generation,
IoU
0.664–0.708
segmentation,
outperforming
state-of-the-art
public
datasets.
effectiveness
each
module
was
testified
through
ablation
experiments.
leverages
knowledge-driven
guide
data-driven
process,
addressing
lack
datasets
model
training.
code
implementing
will
be
accessible
at
https://github.com/designer1024/WSLM-AQ.
Language: Английский
Dynamic Changes and Driving Factors in the Surface Area of Ebinur Lake over the Past Three Decades
Remote Sensing,
Journal Year:
2024,
Volume and Issue:
16(20), P. 3876 - 3876
Published: Oct. 18, 2024
Dryland
lakes
are
indispensable
to
regional
water
resource
systems.
Ebinur
Lake,
the
largest
saline
lake
in
Xinjiang
Uygur
Autonomous
Region,
is
vital
for
biodiversity
and
environmental
stability
but
has
been
facing
predicament
of
gradual
shrinkage
recent
decades.
In
this
study,
we
proposed
a
new
dual-index
method
Landsat
(-5,
-7,
-8,
-9)
data
extract
with
combinations
normalized
difference
index
(NDWI)
modified
NDWI
turbid
waters
(NDWIturbid).
The
showed
high
overall
accuracy
96.36%
Lake.
series
images
from
1992
2023
were
employed
acquire
areas
results
that,
over
past
three
decades,
area
Lake
exhibited
fluctuating
decreasing
trend,
an
average
568.74
±
152.43
km².
northwest
intermittent
significant
changes,
there
was
close
connection
between
core
areas.
Seasonally,
decreased
spring
autumn.
River
inflow,
driven
by
rainfall
human
activities,
primary
factor
affecting
inter/inner
annual
changes
Furthermore,
due
valley
effects,
wind
found
be
critical
diurnal
This
study
should
deepen
understanding
variations
benefit
local
management.
Language: Английский