International Journal of Climatology,
Год журнала:
2023,
Номер
44(1), С. 286 - 304
Опубликована: Дек. 12, 2023
Abstract
Extreme
rainfall
estimates
for
ungauged
areas
contribute
to
improved
resilience
flooding.
This
study
applies
a
sub‐daily
scaling
method
annual
daily
maximum
series
from
102
UK
weather
stations.
We
analyse
resultant
parameters
by
season,
homogenous
region,
urban
area
and
geographic
factors
investigate
how
varies
temporally
spatially.
Dummy
regression
models
are
built
using
these
variables
predict
the
parameter
any
location
in
United
Kingdom.
Estimated
intensities
validated
with
observations
yield
Mean
Absolute
Errors
of
3.0,
1.9
0.9
mm/h
1‐,
2‐
6‐h
events,
respectively.
also
demonstrate
intensity‐duration‐frequency
curves
at
site
Oxfordshire
scaled
observed
data
find
that
1‐
6‐h,
20‐year
estimated
be
within
9.4%.
With
such
unified
relationships,
it
is
possible
derive
extreme
specified
durations
return
periods
locations.
According
our
cross‐validation
intensities,
more
than
88%
sites
fall
10%
error
bounds.
offers
means
generating
design
as
input
flood
simulation
evaluate
pluvial
risks
areas.
Remote Sensing,
Год журнала:
2025,
Номер
17(3), С. 530 - 530
Опубликована: Фев. 4, 2025
This
study
introduces
methods
for
generating
fused
precipitation
data
by
applying
radial
basis
function
(RBF)
interpolation,
which
integrates
radar
reflectivity-based
with
ground-based
gauge
measurements.
Rain
gauges
provide
direct
point
rainfall
measurements
near
the
ground,
while
radars
capture
spatial
variability
of
precipitation.
However,
radar-based
estimates,
particularly
extreme
events,
often
lack
accuracy
due
to
their
indirect
derivation
from
reflectivity.
The
aims
produce
high-resolution
gridded
ground
merging
estimates
precise
rain
were
sourced
automated
synoptic
observing
systems
(ASOSs)
and
automatic
weather
(AWSs),
data,
based
on
hybrid
surface
(HSR)
composites,
all
provided
Korea
Meteorological
Administration
(KMA).
Although
RBF
interpolation
is
a
well-established
technique,
its
application
unprecedented.
To
validate
proposed
method,
it
was
compared
traditional
approaches,
including
mean
field
bias
(MFB)
adjustment
kriging-based
such
as
regression
kriging
(RK)
external
drift
(KED).
Leave-one-out
cross-validation
(LOOCV)
performed
assess
errors
analyzing
overall
error
statistics,
errors,
in
intensity
data.
results
showed
that
RBF-based
method
outperformed
others
terms
accuracy.
Atmosphere,
Год журнала:
2023,
Номер
14(9), С. 1415 - 1415
Опубликована: Сен. 8, 2023
Understanding
the
link
between
extreme
precipitation
and
changes
in
runoff
sediment
yield
is
of
great
significance
for
regional
flood
disaster
response
soil
water
conservation
decision-making.
This
study
investigated
spatial
temporal
distribution
(characterized
by
10
indices
recommended
Expert
Team
on
Climate
Change
Detection
Indices)
Toudaoguai–Longmen
section
middle
Yellow
River
from
1960
to
2021
quantified
effects
based
method
partial
least
squares
regression
(PLSR).
The
index
showed
an
obvious
upward
trend
last
20
years,
with
increases
central
northern
regions
(upstream)
being
stronger
than
increase
southern
region
(downstream).
However,
decreased
significantly
due
implementation
large-scale
measures
Loess
Plateau,
average
rates
94.7
million
m3/a
13.3
t/a
during
1960–2021,
respectively.
change
points
occurred
1979.
Compared
those
period
1979,
reductions
years
1980–2021
were
52.7%
70.6%,
Moreover,
contributed
35.3%
6.2%
reduction
1980–1999
2000–2021
periods,
respectively,
84.3%
40.0%
yield,
It
indicated
that
other
factors
(such
as
construction)
played
main
roles
decrease
area
recent
years.
Theoretical and Natural Science,
Год журнала:
2025,
Номер
87(1), С. 52 - 61
Опубликована: Янв. 15, 2025
This
article
obtains
nearly
18
years
of
precipitation
data
for
Guangzhou
from
the
National
Centers
Environmental
Information
(NCEI)
under
Oceanic
and
Atmospheric
Administration
(NOAA)
conducts
visualization
analysis
modeling
predictions
using
R
software.Using
classic
SARIMA
model
(the
ARIMA
with
seasonal
components)
as
well
regression
residual
models
to
build
simulate
series.We
also
used
Basic-bootstrap
prediction,
which
involves
resampling
residuals
fitted
replacement,
historical
values
predict
future
obtain
a
parameter,
Full-bootstrap
prediction
(which
assumes
that
have
uncertainty
estimated
coefficients
themselves
uncertainty).Therefore,
not
only
are
reasonable
obtained
through
bootstrap
method,
but
multiple
parameters
derived
this
method
determine
more
parameters,
then
forecast
next
12
periods
model.
Comparing
results
four
models,
showed
best
predictive
performance.
The
study
found
there
is
no
significant
trend
in
changes
over
time,
short-term
fluctuations
still
exist.
Although
some
extreme
appeared,
they
remain
within
predictable
range.Therefore,
relevant
city
government
departments
prone
rainfall
can
make
adjustments
drainage
facilities
based
on
local
forecasts
Water,
Год журнала:
2025,
Номер
17(4), С. 466 - 466
Опубликована: Фев. 7, 2025
This
study
reports
on
the
impact
of
rainfall
patterns
and
land
surface
changes
flood
dynamics
in
Meijiang
River
Basin,
located
upper
reaches
Ganjiang
River.
We
formulated
a
range
spatial
distribution
scenarios
employed
MIKE
SHE
model
to
evaluate
variations
volume,
peak,
timing
peaks.
found
that
under
comparable
areal
conditions,
volumes
fluctuated
by
up
6.22%
among
different
patterns,
whereas
peaks
exhibited
differences
36.23%.
When
center
moved
from
upstream
downstream,
both
volume
peak
initially
increased
before
decreasing,
with
maximum
values
4.2
billion
m3
4900
m3/s,
respectively.
selected
three
basin
scales
(i.e.,
10,000,
1000,
100
km2)
for
comparative
analysis.
In
period
between
1985
2020,
conditions
resulted
decreases
basins
7.61,
11.53,
15.79%,
respectively;
reduction
6.58,
9.60,
10.48%,
delayed
times
3,
2,
2
h,
The
results
this
show
significant
influence
exerted
location
centers,
processes.
particular,
when
area
was
reduced,
underlying
more
obvious.
These
also
prediction
needs
consider
complex
interaction
multiple
factors.