Critical Reviews in Environmental Science and Technology,
Год журнала:
2023,
Номер
54(12), С. 953 - 975
Опубликована: Ноя. 29, 2023
The
application
of
machine
learning
(ML)
techniques
for
understanding
and
predicting
organic
matter
(OM)
harmful
algal
blooms
(HABs)
in
freshwater
systems
has
increased
significantly
with
the
availability
abundant
data
advanced
monitoring
technologies.
However,
there
is
a
lack
comprehensive
reviews
concentrating
on
practical
applications
delving
into
potential
risks
associated
misrepresentation
or
inflation
constructing
ML
models.
This
review
aims
to
bridge
these
gaps
by
providing
overview
various
aspects
context
OM
HABs
systems.
It
covers
rapid
assessment,
early
warning,
driver
analysis,
highlighting
diverse
range
employed
areas.
Furthermore,
it
discusses
challenges
considerations
handling,
including
using
situ
remote
sensing
importance
appropriate
data-splitting
avoid
leakage.
To
ensure
unbiased
reproducible
results,
this
offers
recommendations
model
improvement,
such
as
utilizing
explainable
gain
insights
behavior
avoiding
overreliance
single
algorithm.
also
emphasizes
significance
deploying
models
through
user-friendly
interfaces,
enabling
non-experts
effectively
utilize
real-world
water
environments.
Environment International,
Год журнала:
2023,
Номер
172, С. 107788 - 107788
Опубликована: Янв. 31, 2023
Sediment
organic
carbon
(SOC)
is
a
precious
archive
that
synthesizes
anthropogenic
processes
remove
geochemical
fluxes
from
watersheds.
However,
the
scarcity
of
inspection
about
dynamic
mechanisms
activities
on
SOC
limits
understanding
into
how
key
human
factors
drive
dynamics.
Here,
four
typical
basins
with
similar
natural
but
significantly
diverse
contexts
(high-moderate-low
disturbance:
XJ-ZS
and
YJ-LS)
were
selected
to
reconstruct
sedimentation
rates
(SR)
dynamics
nearly
century
based
200-cm
corers.
A
partial
least
squares
path
model
(PLS-PM)
was
used
establish
successive
(70
years)
multiple
data
(population,
agriculture,
land
use,
etc.)
quantification
methods
for
SOC.
Intensified
disturbances
shifted
all
SR
pre-stable
post-1960s
fluctuating
increases
(total
coefficient:
high:
0.63
<
low:
0.47
medium:
0.45).
Although
use
change
co-critical
driver
variations,
their
trend
extent
differed
under
dams
other
(SOC
mutated
in
high-moderate
stable
low).
For
high
basin,
changes
increased
(0.12)
reduced
(−0.10)
downstream
Furthermore,
mutation
corresponded
soil
erosion
due
urbanization
both
periods
B.
moderate,
reversed
increase
afforestation
cropland
(−0.19)
forest
excitation
effect
deep
ploughing,
which
drought
phase
B
ecological
project
A.
low,
Great
Leap
Forward
deforestation
period
reed
sweep
A,
suggested
minor
substantially
affected
(0.16)
fragile
environments.
Overall,
revealed
terrestrial
aquatic
ecosystems
near
centenary,
especially
use.
This
constructive
agroforestry
management
reservoir
construction,
consistent
expectations
like
upstream
sequestration
stabilization.
