Uncontrolled
groundwater
exploitation
can
lead
to
aquifer
depletion,
land
subsidence,
and
saltwater
intrusion.
Effective
management
is
challenging
due
the
intricate
nature
of
subsurface
hydrogeology
spatiotemporally
variable
pumping,
especially
in
a
multi-aquifer
system.
To
ensure
sustainable
withdrawal,
multi-objective
optimization
an
effective
tool
for
balancing
goals
drawdown
effects.
However,
running
simulation-optimization
using
detailed
models
computationally
expensive,
pushing
decision-makers
decide
based
on
limited
scenarios.
In
this
study,
hydrogeological
framework
was
constructed
Capital
Area,
Louisiana,
allowing
individual
assessment
each
unit
better
understand
aquifer's
condition.
Moreover,
surrogate-assisted
model
developed
determine
set
optimum
withdrawal
schemes
Baton
Rouge
Industrial
District.
The
District
experiences
significant
extraction,
around
31
million
gallons
per
day,
with
concern
encroachment
subsidence.
Long
short-term
memory
(LSTM)
networks
were
applied
construct
efficient
surrogate
from
model.
Integrating
non-dominated
sorting
genetic
algorithm
(NSGA-II)
model,
maximized
total
potential
wells
minimized
overall
energy
expenses
associated
pumping
head
at
monitoring
wells.
proposed
approach
successfully
produced
solutions
that
align
defined
objectives.
Using
LSTM
proved
constructing
complex
simulation
study
introduced
practical
decision-making
management.
