An integrated workflow combining machine learning and wavelet transform for automated characterization of heterogeneous groundwater systems
Scientific Reports,
Journal Year:
2025,
Volume and Issue:
15(1)
Published: Feb. 10, 2025
Groundwater
aquifers
are
complex
systems
that
require
accurate
lithological
and
hydrogeological
characterization
for
effective
development
management.
Traditional
methods,
such
as
core
analysis
pumping
tests
provide
precise
results
but
expensive,
time-consuming,
impractical
large-scale
investigations.
Geophysical
well
logging
data
offers
an
efficient
continuous
alternative,
though
manual
interpretation
of
logs
can
be
challenging
may
result
in
ambiguous
outcomes.
This
research
introduces
automated
approach
using
machine
learning
signal
processing
techniques
to
enhance
the
aquifer
characterization,
focusing
on
Quaternary
system
Debrecen
area,
Eastern
Hungary.
The
proposed
methodology
is
initiated
with
imputation
missing
deep
resistivity
from
spontaneous
potential,
natural
gamma
ray,
medium
utilizing
a
gated
recurrent
unit
(GRU)
neural
network.
preprocessing
step
significantly
improved
quality
subsequent
analyses.
Self-organizing
maps
(SOMs)
then
applied
preprocessed
map
distribution
units
across
groundwater
system.
Considering
mathematical
geological
aspects,
SOMs
delineated
three
primary
units:
shale,
shaly
sand,
sand
gravel
which
aligned
closely
drilling
data.
Continuous
wavelet
transform
further
refined
mapping
hydrostratigraphical
boundaries.
integrated
methods
effectively
mapped
subsurface
generating
3D
model
simplifies
into
four
major
zones.
lithology
deterministically
estimated
shale
volume
permeability,
revealing
higher
permeability
lower
sandy
gravelly
layers.
provides
robust
foundation
flow
contaminant
transport
modeling
extended
other
regions
management
development.
Language: Английский
Long-term assessment of groundwater contamination and associated health risks: A study of heavy metals in the Quaternary aquifer system in Pannonian Basin
Physics and Chemistry of the Earth Parts A/B/C,
Journal Year:
2025,
Volume and Issue:
unknown, P. 103935 - 103935
Published: April 1, 2025
Language: Английский