Authorea (Authorea),
Год журнала:
2024,
Номер
unknown
Опубликована: Сен. 2, 2024
In
a
context
where
anticipating
future
trends
and
long-term
variations
in
water
resources
is
crucial,
improving
our
knowledge
about
most
types
of
aquifer
responses
to
climate
variability
change
necessary.
Aquifers
with
dominated
by
seasonal
(marked
annual
cycle)
or
low-frequency
(interannual
decadal
driven
large-scale
dynamics)
may
encounter
different
sensitivities
change.
We
investigated
this
hypothesis
generating
groundwater
level
projections
using
deep
learning
models
for
annual,
inertial
(low-frequency
dominated)
mixed
annual/low-frequency
northern
France
from
16
CMIP6
model
inputs
an
ensemble
approach.
Generated
were
then
analysed
changes
variability.
Generally,
levels
tended
decrease
all
scenarios
across
the
2030-2100.
The
showed
slightly
increasing
but
decreasing
types.
As
severity
scenario
increased,
more
inertial-type
stations
appeared
be
affected
Focusing
on
confirmed
observation:
while
significant
amount
less
severe
SSP
2-4.5
scenario,
eventually
slight
yet
statistically
as
increased.
For
almost
Finally,
seemed,
instances,
higher
than
historical
period,
without
any
differences
between
emission
scenarios.
HydroResearch,
Год журнала:
2024,
Номер
7, С. 285 - 300
Опубликована: Янв. 1, 2024
The
present
study
aims
to
thoroughly
review
GWL
depletion
monitoring
studies
completed
between
2000
and
2023
based
on
data-driven
models
GIS
approaches
from
a
global
perspective.
summarizes
the
details
of
reviewed
papers,
including
location,
period,
time
scale,
key
objective,
input
parameter,
applied
model,
performance
metrics,
research
gaps,
limitations,
rate.
mean
rate
varied
worldwide
2.9
±
1.56
1100
33.76
mm/yr
using
7.6
2.98
2046
45.27
GIS-based
approaches.
This
assesses
strength
relationships
various
keywords
analyzed
co-author
networks
Vos-viewer.
It
proposes
groundwater
development
strategy
evaluated
papers
provide
long-term
solution
water
scarcity
problem.
Overall,
this
highlights
existing
gaps
suggests
potential
future
paths
boost
associated
new
knowledge
increase
accuracy
Earth Science Informatics,
Год журнала:
2024,
Номер
18(1)
Опубликована: Дек. 5, 2024
Abstract
Groundwater
models
are
valuable
tools
to
quantify
the
response
of
groundwater
level
hydrological
stresses
induced
by
climate
variability
and
extraction.
These
strive
for
sustainable
management
balancing
recharge,
discharge,
natural
processes,
with
serving
as
a
critical
variable.
While
traditional
numerical
labour-intensive,
machine
learning
deep
offer
data-driven
alternative,
from
historical
data
predict
variations.
The
in
wells
is
typically
recorded
continuous
time
series
essential
implementing
managed
aquifer
recharge
within
particular
region.
Machine
generate
approach
modeling
an
area,
there
need
understand
if
they
most
suitable
improve
model
prediction.
To
address
this
objective,
study
evaluates
two
algorithms
-
Random
Forest
(RF)
Support
Vector
(SVM);
Simple
Recurrent
Neural
Network
(SimpleRNN)
Long
Short-Term
Memory
(LSTM)
changes
West
Coast
Aquifer
System
South
Africa.
Analysis
regression
error
metrics
on
test
dataset
revealed
that
SVM
outperformed
other
terms
root
mean
square
error,
whereas
random
forest
had
best
performance
MAE.
In
accuracy
analysis
predicted
levels,
achieved
highest
MAE
0.356
m
RMSE
0.372
m.
concludes
effective
improved
prediction
level.
Further
research
can
incorporate
more
detailed
geologic
information
area
enhanced
interpretation.
Journal of Hydroinformatics,
Год журнала:
2024,
Номер
26(11), С. 2962 - 2979
Опубликована: Ноя. 1, 2024
ABSTRACT
The
performance
of
regional
groundwater
level
(GWL)
prediction
model
hinges
on
understanding
intricate
spatiotemporal
correlations
among
monitoring
wells.
In
this
study,
a
graph
convolutional
network
(GCN)
with
long
short-term
memory
(LSTM)
(GCN–LSTM)
is
introduced
for
GWL
utilizing
data
from
16
wells
located
in
the
northeastern
Xiangtan
City,
China.
This
designed
to
account
both
hybrid
temporal
dependencies
and
spatial
autocorrelations
It
consists
two
parts:
part
employs
GCNs
extract
characteristics
self-similarity
weight
matrix
an
attribute
wells;
utilizes
LSTM
module
capture
patterns
sequences,
along
monthly
precipitation
temperature
data.
dynamically
predicts
changes
levels,
achieving
higher
accuracy
average
compared
single-well
predictions
using
LSTM.
By
incorporating
autocorrelations,
GCN–LSTM
demonstrated
improvement
goodness-of-fit
approximately
11.21%
over
LSTM-based
individual
Its
application
holds
significant
reference
value
sustainable
utilization
development
resources
City.
