Building virtual forest landscapes to support forest management: the challenge of parameterization
Published: Feb. 28, 2025
Simulation
models
are
important
tools
to
study
the
impacts
of
climate
change
and
natural
disturbances
on
forest
ecosystems.
Being
able
track
tree
demographic
processes
in
a
spatially
explicit
manner,
process-based
landscape
considered
most
suitable
provide
robust
projections
that
can
aid
decision-making
management.
However,
challenging
parameterize
setting
up
new
areas
for
application
studies
largely
depends
data
availability.
The
aim
this
is
demonstrate
parameterization
process,
including
model
testing
evaluation,
area
Italian
Alps
using
available
data.
We
processed
soil,
climate,
carbon
pools,
vegetation,
management
data,
ran
iterative
spin-up
simulations
generate
virtual
best
resembling
current
conditions.
Our
results
demonstrated
feasibility
initializing
with
typically
from
plans
national
inventories,
as
well
openly
mapping
products.
Evaluation
tests
proved
ability
capture
environmental
constraints
driving
regeneration
dynamics
inter-specific
competition
forests
Alps,
simulate
dynamics.
subsequently
be
applied
investigate
development
under
suite
future
scenarios
recommendations
adapting
decisions.
Language: Английский
Improving the Accuracy of Tree Species Mapping by Sentinel-2 Images Using Auxiliary Data—A Case Study of Slyudyanskoye Forestry Area near Lake Baikal
Forests,
Journal Year:
2025,
Volume and Issue:
16(3), P. 487 - 487
Published: March 10, 2025
Timely
and
accurate
information
on
forest
composition
is
crucial
for
ecosystem
conservation
management
tasks.
Information
regarding
the
distribution
extent
of
forested
areas
can
be
derived
through
classification
satellite
imagery.
However,
optical
data
alone
are
often
insufficient
to
achieve
required
accuracy
due
similarity
in
spectral
characteristics
among
tree
species,
particularly
mountainous
regions.
One
approach
improving
integration
auxiliary
environmental
data.
This
paper
presents
results
research
conducted
Slyudyanskoye
Forestry
area
Irkutsk
Region.
A
dataset
comprising
101
variables
was
collected,
including
Sentinel-2
bands,
vegetation
indices,
climatic,
soil,
topographic
data,
as
well
canopy
height.
The
performed
using
Random
Forest
machine
learning
method.
demonstrated
that
significantly
improved
performance
species
model,
with
overall
increasing
from
49.59%
(using
only
bands)
80.69%
(combining
variables).
most
significant
improvement
achieved
incorporation
climatic
soil
features.
important
were
shortwave
infrared
band
B11,
height,
length
growing
season,
number
days
snow
cover.
Language: Английский
Geospatial database generation of forest growth using the QGIS software package
IOP Conference Series Earth and Environmental Science,
Journal Year:
2024,
Volume and Issue:
1415(1), P. 012049 - 012049
Published: Dec. 1, 2024
Abstract
This
article
discusses
the
use
of
machine
learning
methods
in
QGIS
software
package
to
create
a
geographic
information
database
forest
growth.
The
authors
explore
possibilities
applying
algorithms
analyse
satellite
images
different
resolutions
and
geospatial
data
determine
growth
resources.
consider
classification
forecasting
their
advantages
context
creating
for
tracking
managing
research
results
confirm
effectiveness
using
accurate
automated
analysis
dynamics
changes
cover,
which
can
serve
as
basis
decision-making
field
forestry
protection
natural
resource
potential.
Language: Английский