Water,
Год журнала:
2024,
Номер
16(20), С. 2944 - 2944
Опубликована: Окт. 16, 2024
Planning,
managing
and
optimising
surface
water
quality
is
a
complex
multifaceted
process,
influenced
by
the
effects
of
both
climate
uncertainties
anthropogenic
activities.
Developing
an
innovative
robust
decision
support
framework
(DSF)
essential
for
effective
efficient
management,
so
it
can
provide
information
on
assist
policy
makers
resource
managers
to
identify
potential
causes
deterioration.
This
crucial
implementing
actions
such
as
infrastructure
development,
legislative
compliance
environmental
initiatives.
Recent
advancements
in
computational
domains
have
created
opportunities
employing
artificial
intelligence
(AI),
advanced
statistics
mathematical
methods
use
improved
management.
study
proposed
comprehensive
conceptual
DSF
minimise
adverse
extreme
weather
events
change
quality.
The
utilises
machine
learning
(ML),
deep
(DL),
geographical
system
(GIS)
statistical
techniques
foundation
this
outcomes
from
our
three
studies,
where
we
examined
application
ML
DL
models
predicting
index
(WQI)
reservoirs,
utilising
find
seasonal
trend
rainfall
quality,
exploring
connection
between
streamflow,
GIS
show
spatial
temporal
variability
hydrological
parameters
WQI.
Three
potable
supply
reservoirs
Toowoomba
region
Australia
were
taken
area
practical
implementation
DSF.
serve
mechanism
distinct
characteristics
understand
correlations
rainfall,
streamflow
will
enable
enhance
their
making
processes
selecting
management
priorities
safeguard
face
future
variability,
including
prolonged
droughts
flooding.
Marine Pollution Bulletin,
Год журнала:
2025,
Номер
213, С. 117601 - 117601
Опубликована: Янв. 31, 2025
Coastal
ecosystems
in
Pacific
Island
Countries
and
Territories
are
vital
to
local
livelihoods,
yet
increasingly
face
pressures
from
urbanization.
In
Fiji,
the
Greater
Suva
Urban
Area,
where
one-third
of
nation's
population
live,
exemplifies
these
challenges.
This
study
examines
spatial
temporal
water
quality
variations
coastal
zone,
focusing
on
physicochemical,
nutrients,
clarity
parameters.
Using
a
Seabird
Scientific
SBE19
CTD
Thermo
Orion™
AQUAfast™
colorimeter,
coupled
with
hierarchical
clustering
principal
component
analysis,
six
clusters
were
identified,
influenced
by
oceanic
processes,
river
inputs,
anthropogenic
activities.
Key
findings
highlight
nutrient
enrichment
near
urban
centers
particularly
at
Kinoya
Sewage
Treatment
Plant
outfall,
ammonia
exceeded
17.8
mg/L,
significant
variation
observed
nitrate
(up
0.24
±
0.06
mg/L)
nitrite
concentrations
mouths.
Seasonal
runoff
contributed
elevated
turbidity
3.5
NTU)
total
suspended
solids
14.7
levels
during
wet
months.
Salinity,
temperature
exhibited
strong
seasonal
variability,
reflecting
land-ocean
interactions
restricted
exchange.
These
emphasize
need
for
targeted
action
mitigate
pollution,
runoff,
wastewater
impacts.
provides
cost-effective
monitoring
framework
management,
offering
insights
sustainable
resource
management
Fiji
other
regions
amidst
urbanization
climate
change.
Heliyon,
Год журнала:
2024,
Номер
10(13), С. e33695 - e33695
Опубликована: Июнь 28, 2024
The
water
quality
index
(WQI)
is
a
widely
used
tool
for
comprehensive
assessment
of
river
environments.
However,
its
calculation
involves
numerous
parameters,
making
sample
collection
and
laboratory
analysis
time-consuming
costly.
This
study
aimed
to
identify
key
parameters
the
most
reliable
prediction
models
that
could
provide
maximum
accuracy
using
minimal
indicators.
Water
from
2020
2023
were
collected
including
nine
biophysical
chemical
indicators
in
seventeen
rivers
Yancheng
Nantong,
two
coastal
cities
Jiangsu
Province,
China,
adjacent
Yellow
Sea.
Linear
regression
seven
machine
learning
(Artificial
Neural
Network
(ANN),
Self-Organizing
Maps
(SOM),
K-Nearest
Neighbor
(KNN),
Support
Vector
Machines
(SVM),
Random
Forest
(RF),
Extreme
Gradient
Boosting
(XGB)
Stochastic
(SGB))
developed
predict
WQI
different
groups
input
variables
based
on
correlation
analysis.
results
indicated
improved
2022
but
deteriorated
2023,
with
inland
stations
exhibiting
better
conditions
than
ones,
particularly
terms
turbidity
nutrients.
environment
was
comparatively
Nantong
Yancheng,
mean
values
approximately
55.3–72.0
56.4–67.3,
respectively.
classifications
"Good"
"Medium"
accounted
80
%
records,
no
instances
"Excellent"
2
classified
as
"Bad".
performance
all
models,
except
SOM,
addition
variables,
achieving
R2
higher
0.99
such
SVM,
RF,
XGB,
SGB.
RF
XGB
total
phosphorus
(TP),
ammonia
nitrogen
(AN),
dissolved
oxygen
(DO)
(R2
=
0.98
0.91
training
testing
phase)
predicting
values,
TP
AN
(accuracy
85
%)
grades.
"Low"
grades
highest
at
90
%,
followed
by
level
70
%.
model
contribute
efficient
evaluation
identifying
facilitating
effective
management
basins.
