Revista Brasileira de Geografia Física,
Journal Year:
2025,
Volume and Issue:
18(4), P. 2551 - 2572
Published: April 22, 2025
In
this
study,
we
identified
flood
risk
areas
using
maximum
streamflow
estimates
derived
from
regional
functions
and
1D
hydraulic
modelling
in
an
urban
ungauged
watershed
Lavras,
Minas
Gerais,
Brazil.
Our
approach
focused
on
employing
simple
techniques
based
secondary
data
to
support
management
regions
lacking
hydrological
monitoring
technical
expertise.
The
study
evaluated
two
developed
for
the
Grande
River
Basin
identify
most
suitable
mapping.
Flood
maps
5,
10,
50
100-year
return
periods
were
created
software
HEC-RAS
assessed
population
vulnerability
under
current
regulated
land
use
scenarios.
By
data,
work
provides
a
practical
decision-making
cities
with
limited
resources,
highlighting
need
improved
drainage
management.
Climate,
Journal Year:
2025,
Volume and Issue:
13(5), P. 93 - 93
Published: May 2, 2025
Accurate
simulation
of
extreme
precipitation
events
is
crucial
for
managing
climate-vulnerable
sectors
in
Southern
Africa,
as
such
directly
impact
agriculture,
water
resources,
and
disaster
preparedness.
However,
global
climate
models
frequently
struggle
to
capture
these
phenomena,
which
limits
their
practical
applicability.
This
study
investigates
the
effectiveness
three
bias
correction
techniques—scaled
distribution
mapping
(SDM),
quantile
(QDM),
QDM
with
a
focus
on
above
below
95th
percentile
(QDM95)—and
daily
outputs
from
11
Coupled
Model
Intercomparison
Project
Phase
6
(CMIP6)
models.
The
Climate
Hazards
Group
Infrared
Precipitation
Stations
(CHIRPS)
dataset
was
served
reference.
bias-corrected
native
were
evaluated
against
observational
datasets—the
CHIRPS,
Multi-Source
Weighted
Ensemble
(MSWEP),
Global
Climatology
Center
(GPCC)
datasets—for
period
1982–2014,
focusing
December-January-February
season.
ability
generate
eight
indices
developed
by
Expert
Team
Change
Detection
Indices
(ETCCDI)
evaluated.
results
show
that
captured
similar
spatial
patterns
precipitation,
but
there
significant
changes
amount
episodes.
While
generally
improved
representation
its
varied
depending
reference
used,
particularly
maximum
one-day
(Rx1day),
consecutive
wet
days
(CWD),
dry
(CDD),
extremely
(R95p),
simple
intensity
index
(SDII).
In
contrast,
total
rain
(RR1),
heavy
(R10mm),
(R20mm)
showed
consistent
improvement
across
all
observations.
All
techniques
enhanced
accuracy
indices,
demonstrated
higher
pattern
correlation
coefficients,
Taylor
skill
scores
(TSSs),
reduced
root
mean
square
errors,
fewer
biases.
ranking
using
comprehensive
rating
(CRI)
indicates
no
single
model
consistently
outperformed
others
relative
GPCC,
MSWEP
datasets.
Among
methods,
SDM
QDM95
variety
criteria.
strategies,
best-performing
EC-Earth3-Veg,
EC-Earth3,
MRI-ESM2,
multi-model
ensemble
(MME).
These
findings
demonstrate
efficiency
improving
modeling
extremes
ultimately
boosting
assessments.
Hydrology,
Journal Year:
2025,
Volume and Issue:
12(5), P. 115 - 115
Published: May 8, 2025
This
study
fosters
tropical
hydroclimatology
research
by
implementing
a
computational
modeling
framework
based
on
artificial
neural
networks
and
machine
learning
techniques.
We
evaluated
two
models,
Multilayer
Perceptron
(MLP)
Support
Vector
Machine
(SVM),
in
their
ability
to
simulate
20-year
monthly
time
series
(2001–2021)
of
minimum
maximum
river
stage
the
Itacaiúnas
River
Basin
(BHRI),
located
eastern
Brazilian
Amazon.
The
models
were
configured
using
explanatory
variables
spanning
meteorological,
climatological,
environmental
dimensions,
ensuring
representation
key
local
regional
hydrological
drivers.
Both
exhibited
robust
performance
capturing
fluviometric
variability,
with
comprehensive
multimetric
statistical
evaluation
indicating
MLP’s
superior
accuracy
over
SVM.
Notably,
MLP
model
reproduced
level
during
sequence
extreme
events
linked
natural
disasters
(floods)
across
BHRI
municipalities.
These
findings
underscore
model’s
potential
for
refining
hydrometeorological
products,
thus
supporting
water
resource
management
decision-making
processes
Amazon
region.
International Journal of Climatology,
Journal Year:
2025,
Volume and Issue:
unknown
Published: May 9, 2025
ABSTRACT
Dynamical
downscaling
(DDS)
datasets
play
a
crucial
role
in
understanding
regional
climate
patterns
and
extreme
weather
events.
This
study
evaluates
the
reproducibility
of
indices
Japan
using
two
DDS
based
on
JRA‐55
reanalysis
for
period
1979–2012.
A
total
48
were
analysed
to
assess
biases,
interannual
variability,
trends
precipitation
temperature
by
comparing
with
AMeDAS
observations,
high‐resolution
automated
meteorological
observation
network
Japan.
Both
reasonably
captured
correlation
coefficients
exceeding
0.6
many
indices.
However,
systematic
biases
underestimations
trend
magnitudes
observed.
For
indices,
DS‐run
(DDS
Non‐Hydrostatic
Model,
NHM)
generally
exhibited
consistent
tendency
toward
negative
across
most
areas,
while
RC‐run
Regional
Climate
NHRCM)
showed
relatively
smaller
some
regions
but
larger
Southwest
Islands
(area
7).
both
runs
successfully
reproduced
variability.
pronounced
TX‐related
particularly
TXm
TXn,
slightly
TXx.
Positive
more
common
TN‐related
especially
area
1.
Trend
analyses
revealed
regionally
varying
patterns.
direction
observed
all
regions,
high
agreement
sign.
magnitude
statistical
significance
varied
depending
index
type
region.
Although
each
run
distinct
characteristics,
shared
highlighting
need
further
improvements
model
performance.
These
findings
suggest
importance
careful
evaluation
when
outputs
impact
assessments
offer
useful
insights
future
improvement
development
strategies
