Scientific African,
Год журнала:
2023,
Номер
20, С. e01670 - e01670
Опубликована: Апрель 13, 2023
Monthly
precipitation
data
that
span
from
1986
–
2019
(34)
years
were
downloaded
the
website
of
Nigeria
Meteorological
Agency,
(NIMET),
Abuja,
and
used
to
assess
rainfall
trend
drought
occurrences
in
Sudan
Savanna
Nigeria.
The
non-parametric
Mann
Kendall
test
was
employed
for
trends
while
Standardized
Precipitation
Index
(SPI)
generate
SPI
values
graphs.
Our
findings
show
is
on
declining
Gombe
(z-value
=
-0.5930)
Kaduna
-2.520)
but
increasing
Sokoto
1.9272).
alone
with
P-value
0.0117
which
less
than
0.05
level
significance
showed
monotonic
trend.
In
Sokoto,
even
though
there
an
increase,
it
not
since
P-
value
0.0540
greater
significant
0.05.
only
two
extreme
droughts
2004
2013
-2.04
-2.08
respectively.
Some
experienced
between
moderate
severe
all
stations
near
normal
condition
predominant
stations.
Apart
incidence
flooding
2.755,
a
few
incidences
wetness
or
other
Water Resources Research,
Год журнала:
2024,
Номер
60(5)
Опубликована: Май 1, 2024
Abstract
Droughts
are
among
the
most
devastating
natural
hazards,
occurring
in
all
regions
with
different
climate
conditions.
The
impacts
of
droughts
result
significant
damages
annually
around
world.
While
drought
is
generally
described
as
a
slow‐developing
hazardous
event,
rapidly
developing
type
drought,
so‐called
flash
has
been
revealed
by
recent
studies.
rapid
onset
and
strong
intensity
require
accurate
real‐time
monitoring.
Addressing
this
issue,
Generative
Adversarial
Network
(GAN)
developed
study
to
monitor
over
Contiguous
United
States
(CONUS).
GAN
contains
two
models:
(a)
discriminator
(b)
generator.
architecture
employs
Markovian
discriminator,
which
emphasizes
spatial
dependencies,
modified
U‐Net
generator,
tuned
for
optimal
performance.
To
determine
best
loss
function
four
networks
functions,
including
Mean
Absolute
Error
(MAE),
adversarial
loss,
combination
Square
(MSE),
MAE.
Utilizing
daily
datasets
collected
from
NLDAS‐2
Standardized
Soil
Moisture
Index
(SSI)
maps,
network
trained
SSI
Comparative
assessments
reveal
proposed
GAN's
superior
ability
replicate
values
Naïve
models.
Evaluation
metrics
further
underscore
that
successfully
identifies
both
fine‐
coarse‐scale
patterns
abrupt
changes
temporal
important
identification.
International Journal of Climatology,
Год журнала:
2024,
Номер
44(3), С. 812 - 830
Опубликована: Янв. 9, 2024
Abstract
Understanding
the
spatiotemporal
historical
drought
pattern
and
their
sensitivity
effect
on
potential
evapotranspiration
(PET)
vegetation
coverage
changes
is
essential
for
efficient
mitigation
policies
under
climate
change.
In
this
study,
we
used
standardized
precipitation
index
(SPEI)
at
multiple
timescales,
such
as
SPEI‐01,
SPEI‐03,
SPEI‐06,
SPEI‐09
SPEI‐12;
explored
regional‐scale
dry
wet
annual
across
seven
sub‐regions
of
South
Asia
from
1902
to
2018.
Results
suggest
that
1981
2018,
extreme
SPEI
has
increased
in
Asia,
which
mostly
affects
summer
winter
growing
seasons,
is,
SPEI‐06
SPEI‐12
Asia.
The
frequency
events
during
had
an
extremely
year
starting
1998
affected
region.
Data
past
18
years
showed
land
changing
detection
forests,
cultivated
land,
arid
savanna
farmland;
by
contrast,
there
been
significantly
reduced
permanent
ice
snow,
mixed
open
shrub,
grasslands,
wetlands,
water
bodies
evergreen
broadleaf
forests.
Seasonal
presented
diverse
characteristics
showing
a
trend
Afghanistan,
India,
Pakistan
Sri
Lanka
autumn
winter.
Afghanistan
Bhutan
are
compared
with
other
occurring
45.3%
44.4%.
Lanka,
India
driest
regions
due
high
frequency,
duration
intensity.
correlation
between
PET
crop
stress
(CWSI),
regional
ET
p
reduction
(Er)
indicated
considerably
negative
correlation,
while
positive
was
found
CWSI
Er,
NDVI
Er.
This
study
provides
comprehensive
assessment
PET,
SPEI,
can
help
formulating
long‐term
adaptive
strategies
reduce
cumulative
impacts
droughts.
Sustainable Development,
Год журнала:
2025,
Номер
unknown
Опубликована: Янв. 23, 2025
ABSTRACT
Sustainable
development
requires
orderly
migratory
flows
that
avoid
the
resulting
socio‐economic
tensions
in
countries
of
origin
and
destination.
Issues
such
as
climate,
armed
conflicts
lack
job
stability
encourage
human
displacement
towards
large
cities,
simultaneously
generating
problems
densification
at
destination
depopulation
abandoned
areas.
In
this
study,
we
analyzed
link
between
determinants
quality
life
immigration,
with
aim
identify
factors
need
to
be
strengthened
“emptied”
territories
and,
way,
achieve
a
reasonable
distribution
population.
We
use
methodological
advance
random
forest
order
correctly
address
complexity
variability
data.
The
study
is
limited
Spain's
19
autonomous
regions,
which
suffer
from
both
overpopulation,
covering
broad
period
guarantees
robustness
results
(2008–2021).
There
evidence
importance
labor
market,
health
education
settlement
migrants.
Decision‐makers
strengthen
these
aspects
under‐inhabited
areas
by
directing
financial
resources
enhancing
their
attractiveness.
This
will
make
it
possible
redirect
exodus
re‐establishing
economic
social
whole.
