ArticlePDF Available

Reduced spatial extent of extreme storms at higher temperatures



Extreme precipitation intensity is expected to increase in proportion to the water-holding capacity of the atmosphere. However, increases beyond this expectation have been observed, implying that changes in storm dynamics may be occurring alongside changes in moisture availability. Such changes imply shifts in the spatial organization of storms, and we test this by analyzing present-day sensitivities between storm spatial organization and near-surface atmospheric temperature. We show that both the total precipitation depth and the peak precipitation intensity increases with temperature, while the storm's spatial extent decreases. This suggests that storm cells intensify at warmer temperatures, with a greater total amount of moisture in the storm, as well as a redistribution of moisture toward the storm center. The results have significant implications for the severity of flooding, as precipitation may become both more intense and spatially concentrated in a warming climate.
Reduced spatial extent of extreme storms
at higher temperatures
Conrad Wasko
, Ashish Sharma
, and Seth Westra
School of Civil and Environmental Engineering, University of New South Wales, Sydney, New South Wales, Australia,
School of Civil, Environmental and Mining Engineering, University of Adelaide, Adelaide, South Australia, Australia
Abstract Extreme precipitation intensity is expected to increase in proportion to the water-holding
capacity of the atmosphere. However, increases beyond this expectation have been observed, implying
that changes in storm dynamics may be occurring alongside changes in moisture availability. Such changes
imply shifts in the spatial organization of storms, and we test this by analyzing present-day sensitivities
between storm spatial organization and near-surface atmospheric temperature. We show that both the total
precipitation depth and the peak precipitation intensity increases with temperature, while the storms spatial
extent decreases. This suggests that storm cells intensify at warmer temperatures, with a greater total
amount of moisture in the storm, as well as a redistribution of moisture toward the storm center. The results
have signicant implications for the severity of ooding, as precipitation may become both more intense and
spatially concentrated in a warming climate.
1. Introduction
Short-duration extreme precipitation is predicted to intensify as a result of increases in atmospheric tempera-
ture in most locations globally [Kirtman et al., 2013]. Investigation of the historical sensitivity of precipitation
to temperature is an important source of evidence to understand how extreme precipitation might change in
the future [Collins et al., 2013]. In the absence of changes to circulation patterns and relative humidity, ther-
modynamic factors suggest that extreme precipitation intensity should scale at about 7%/°C as governed by
the Clausius-Clapeyron (C-C) relationship, which describes the capacity of the atmosphere to hold moisture
[Trenberth et al., 2003; Westra et al., 2014]. Using present-day climate data, short-duration precipitation has
been found to scale at rates ranging from C-C to double C-C for temperatures below 24°C [Lenderink and
van Meijgaard, 2008; Hardwick Jones et al., 2010; Lenderink et al., 2011; Mishra et al., 2012; Berg et al., 2013],
with the scaling differing with storm duration [Panthou et al., 2014; Wasko et al., 2015]. Similar rates of scaling
have been observed in long-term historical trends of extreme precipitation, supporting the hypothesis that
these relationships may be indicative of future climatic conditions [Westra and Sisson, 2011; Fujibe, 2013;
Westra et al., 2013, 2014].
A mechanism that has been put forward to explain departures from the thermodynamic (C-C) scaling rate is a
change in storm dynamics. This has been attributed to the release of latent heat, which can result in the storm
increasing the area over which it sources moisture [Trenberth et al., 2003]. Observation of scaling rates above
C-C has led to conjecture that there may be changes in dominant precipitation types as convective events have
greater scaling rates than nonconvective events [Haerter and Berg,2009;Berg et al., 2013; Collins et al., 2013;
Molnar et al., 2015] and also that both the thermodynamic and dynamical mechanisms may operate jointly
[Lenderink and van Meijgaard,2008;Collins et al., 2013; Wasko and Sharma, 2015] enhancing the convergence
of moisture possibly from a larger moisture sourcearea [Trenberth et al., 2003; Westra et al., 2014].
The specic causes of changes in precipitation intensity with temperature can therefore be expected to be
associated with distinct temporal and spatial signatures. For example, if thermodynamic factors dominate
and the precipitation intensies in proportion to the water-holding capacity of the atmosphere, a consistent
increase in precipitation intensity across a storm could be expected. In contrast, if dynamic factors dominate,
the spatial extent of each storm cell and the organization of rainfall within the storm cell should change
[Westra et al., 2014]. Although evidence exists of intensifying temporal patterns at higher temperatures
[Wasko and Sharma, 2015], studies investigating possible changes to the spatial signature of storms with
climate change remain limited having focused on areal reduction factors for engineering applications
[Li et al., 2015] and spatial structures in regional climate models [Guinard et al., 2015].
Geophysical Research Letters
Key Points:
Spatial extent of storms reduces as
temperatures increase
Storm patterns are less uniform at
higher temperatures
Moisture is redistributed from the
storm boundaries to the storm center
Supporting Information:
Supporting Information S1
Correspondence to:
A. Sharma,
Wasko, C., A. Sharma, and S. Westra
(2016), Reduced spatial extent of
extreme storms at higher temperatures,
Geophys. Res. Lett.,43, 40264032,
Received 1 MAR 2016
Accepted 5 APR 2016
Accepted article online 7 APR 2016
Published online 25 APR 2016
©2016. American Geophysical Union.
All Rights Reserved.
Here we investigate the relationship between the spatial organization of moisture within a storm and the
near-surface dry-bulb temperature. We rst examine the hypothesis that the distribution of precipitation with
distance from the storm centers differs when storms are stratied by temperature. Subsequently, we examine
the relationship with temperature of a range of statistics that describe the organization of moisture within the
storm. Finally, using the derived relationships, we develop projections of precipitation distribution with dis-
tance from the storm center for higher temperatures.
2. Data and Methods
Subdaily precipitation and 1.2 m dry-bulb temperature were obtained from the Australian Bureau of
Meteorology weather station data set and have been used extensively in previous studies [Hardwick Jones
et al., 2010; Westra and Sisson, 2011; Westra et al., 2012; Wasko and Sharma, 2015]. Precipitation is measured
using a tipping bucket rain gauge or pluviograph and reported every 6 min. In this study the reported
precipitation was accumulated to resolutions of 1 and 3 h. The temperature data resolution is typically 3 to
12 h, depending on the station. The data set contains over 1300 precipitation stations and approximately
1700 temperature stations across Australia.
To analyze the spatial distribution of moisture within a storm, point observations of precipitation need to
be transformed to spatial elds. First, at each gauge location, independent precipitation events were
identied. Storm events were dened as independent if they were separated by 5h of zero precipitation
for the hourly rainfall series and 15 h of zero precipitation for the 3-hourly rainfall series [Wasko and
Sharma, 2014]. The precipitation chosen for further analysis was the maximum 1 h or 3 h precipitation burst
with each storm. Each precipitation maximum was matched to the coincident temperature, which was cal-
culated as a 24 h moving average centered on the time of the maximum precipitation. The precipitation-
temperature pairs were extended to a spatial eld by nding the precipitation that occurred at the same
time as the maximum from the central gauge at neighboring gauges up to a radius of 50 km. A spatial eld
was considered for analysis if it had a minimum of ve data points and the maximum precipitation in the
spatial eld occurred at the central gauge. All other spatial elds were discarded. For a station to be
included in the analysis it had to have a minimum of 10 years of record length and a minimum of 100
observed independent spatial elds. In total, 93 stations were considered for the 1 h duration and 78 sta-
tions for the 3 h duration.
To isolate extreme events, a 90th percentile exponential quantile regression was tted to the peak
precipitation-temperature pairs [Wasko and Sharma, 2014] and only spatial eld events that exceeded the
regression line were analyzed for their sensitivity to temperature. The nal data set for each site contains
j= 1..mspatial elds matched to their coincident temperature T
. Each spatial eld jcontains i= 1..nprecipita-
tion observations p
at distance d
from the center gauge; thus, p
is the precipitation observation iof spatial
eld j.
A set of statistics was chosen to represent the spatial organization of moisture within the spatial elds and
calculated on the nal mextreme sets of spatial elds. The statistics are as follows:
1. Peak precipitation (PP), which is the maximum precipitation within the spatial eld (therefore by deni-
tion located at the storm center);
2. Total precipitation (PT), which is the two-dimensional integration of the precipitation plotted against the
radial distance from the storm center;
3. Fraction of zero rainfall observations (PZ) in the spatial eld;
4. Coefcient of variation (CV), which is the variance of the observations within the spatial eld divided by
their mean;
5. Effective radius (RE), dened at the centroid of the precipitation where p
are the nprecipitation observations
within the spatial eld jwith distance d
from the central gauge:
; (1)
Geophysical Research Letters 10.1002/2016GL068509
6. Parameters Aand Bof an expo-
nential curve that describes
how the storm decays with
distance from its center [von
Hardenberg et al., 2003; Rebora
and Ferraris, 2006] where the
parameters to be estimated for
each spatial eld j:
pij dij
Here Aand Bwere tted by mini-
mizing the absolute residuals using
a shufed complex evolution opti-
mization algorithm [Andrews et al.,
2011]. A schematic of PP, PT, and
RE is presented in Figure S1 in the
supporting information.
An exponential regression was
then used to nd the relationship
between each desired statistic set
= {PP
} with
temperature T
(equation (3)), with
the exception of the fraction of
zero statistic where a linear regression was used. For a ΔTchange in the temperature, the tted relationship
is as follows [Hardwick Jones et al., 2010; Utsumi et al., 2011], where α
is the rate at which the statistic scales
per degree temperature change:
The data represent a wide range of climatic zones and precipitation-temperature sensitives [Hardwick
Jones et al., 2010; Utsumi et al., 2011; Wasko and Sharma, 2014]. The Australian climate can be split into
three main zones (Figure 2): tropical in the north, temperate in the east, and arid in the southwest. The
precipitation is summer dominant in the tropical north and winter dominant in the subtropical south
mainland. Temperate zones in the east have slightly winter-dominant precipitation, with precipitation
seasonality becoming uniform toward the south. The scaling of the statistics is aggregated across these
climatic zones.
Finally, the spatial elds were projected for a warmer climate on the basis of tted exponential curves
(equation (2)). To represent the current (base) climate, parameters Aand Bwere calculated at a temperature
of 20°C. The curves were then projected in one degree increments using the calculated scaling above (equa-
tion (3)). It was found that the scaling of the parameters Aand Bwas less stable than the peak precipitation
(PP) and effective radius (RE) scaling. Hence, the latter were used to scale the tted curves. The peak precipi-
tation scaling is equivalent to the parameter Ascaling and the inverse of the effective radius scaling is equiva-
lent to the parameter scaling B.
3. Results
3.1. Spatial Organization as a Function of Temperature
In order to conrm the hypothesis that there is a different spatial organization of moisture within storm cells
at differing temperature, we begin by splitting all the spatial elds across Australia into two groups: the rst
for temperatures above 25°C and the second for temperatures below 18°C. If we are able to detect differences
in the organization of moisture within the largest storms in these groups, it would suggest that temperature
does in fact covary with the spatial signature of storm cells. For both groups the greatest 1000 events by
precipitation depth were chosen.
Figure 1. Fitted exponential curves for the greatest 1000 hourly storm bursts
by precipitation depth below 18°C (blue) and above 25°C (red). The vertical
dashed lines are the corresponding effective radii of the storm data dened
as the centroid of the storm by precipitation volume. Three-dimensional
curves are also presented emphasizing the increase in the storm intensity at
the center of the storm and reduced spatial extent at higher temperatures.
Geophysical Research Letters 10.1002/2016GL068509
Figure 1 presents the calculated
exponential curves and the
effective radii (equation (2)) of the
two groups. The condence inter-
vals for the parameters Aand B
were calculated using 1000 boot-
strapped replicates of the precipi-
tation data. As can be seen, the
intensity at the storm center is
greater at high temperatures and
decays more rapidly with distance
from the center. This indicates that
the moisture becomes more con-
centrated near the storm center
for the warmer storms. This is con-
rmed by considering the effective
storm radius: at the higher tem-
perature, the effective storm radius
is located closer to the center.
Consistent results were found
when the analysis was repeated
with spatial eld aggregated on a
regional basis (not shown).
3.2. Scaling of Spatial
Field Statistics
Having found that temperature
affects the spatial organization of
moisture within storm cells, we
now focus on the statistics
describing the spatial organization
for the most extreme events, that
is, those above the 90th percentile
exponential quantile regression.
The scaling of peak precipitation
and effective storm radius with
temperature for the extreme spa-
tial elds is presented in Figure 2.
The peak precipitation increases
with temperature at 95% of the sites (Figure 2a), and the effective radius decreases with temperature at
82% of the sites (Figure 2b). An example of the scaling calculations is presented in Figure S2. Consistent
increases in the peak precipitation and decreases in the effective radius were found for differing threshold
percentiles; however, the magnitude of the scaling was reduced for less extreme threshold percentiles.
The scaling presented in Figure 2 can be grouped by climatic zone and has been presented as box plot dis-
tributions in Figure 3. The median scaling of the peak precipitation in the arid zone is approximately 5%°C
however, the total storm precipitation scales at a slightly lower rate of approximately 4%°C
. As the effective
radius reduces at higher temperatures this suggests that the precipitation is greater at the storm center and
lesser at the storm extents.
This effect is more pronounced for the temperate zone results. The peak precipitation scaling is approxi-
mately 7%°C
and the effective radius has a negative scaling at a median rate of 3%°C
. Consequently,
the scaling in the total precipitation is less than the peak precipitation. This conrms that moisture is being
redistributed from the storm boundaries to the storm center, resulting in less precipitation at the storm
boundaries and a reduced storm size.
Figure 2. Scaling of the peak precipitation and effective radius for 1 h
duration events. Positive scaling is shown in red and negative scaling in
blue. (a) Circles show the scaling of peak precipitation, while the background
shading denotes the Koppen climate classication [Peel et al., 2007]. General
climatic zones are also shown. All the gauge sites used in the analysis are
shown as grey dots. (b) Circles show the scaling of effective radius.
Geophysical Research Letters 10.1002/2016GL068509
The scaling of the peak precipitation
and storm size is stronger at short
durations as shown by comparing
the 1 h scaling to the 3 h scaling
(Figure S3). For 3 h duration bursts,
97% of the sites have positive peak
precipitation scaling and 74% of the
ing. However, the median scaling of
the peak precipitation is reduced to
for the temperate zone
and 3%°C
for the arid zone
(Figure S4). Although the median
remained similar for the effective
radius scaling for the arid zone, in
the temperate zone it decreased in
magnitude from 3%°C
for the
1 h duration to 2%°C
for the 3 h
duration storms, suggesting that the
redistribution of moisture from the
storm boundaries to the storm cen-
ter is less at longer storm durations.
The scaling of the fraction of obser-
vations with zero precipitation (PZ)
and the coefcient of variation (CV)
in each spatial eld also supports
the observed trends for the 1 h
(Figure S5) and 3 h durations
(Figure S6). Throughout Australia,
there is a positive scaling in the
fraction of zeros and coefcient of
variation at 87% and 92% of the
sites, respectively, for a 1 h duration.
Similar statistics are obtained for a
3 h duration. These results point to
a less uniform spatial distribution of
precipitation occurring over a smal-
3.3. Projection for Higher
Using the scaling relationships
developed above, projections were
developed to illustrate how the
average storm shape for 1 h and 3
durations might change with
temperature increases of up to
3°C (Figure 4). The scaling applied
is the mean scaling for all the
stations within the climatic zone
of interest. The blue curve repre-
sents the storm shape for a base
climate of 20°C and the red curve
Figure 3. Box plots of peak precipitation scaling, effective radius scaling, and
total precipitation scaling for hourly precipitation grouped by arid and
temperate zones.
Figure 4. Predicted precipitation distribution due to temperature increase.
Each panel consists of four curves. The blue curve is tted to a temperature
of 20°C. Each subsequent curve changes from blue to red in 1°C increments, up
to 23°C. The precipitation is standardized by the peak precipitation depth for
the baseline climate of 20°C to ensure the results are independent to the
baseline temperature choice.
Geophysical Research Letters 10.1002/2016GL068509
represents the shape for a projected increase of 3°C (i.e., 23°C); the intermediate curves represent increases in
1° increments. The precipitation in Figure 4 is standardized by the peak precipitation depth for the baseline
climate so the results presented are independent of the baseline temperature choice.
There is minimal change in the storm shape for the arid zone, with an increase in overall precipitation depth
occurring at the same rate as the scaling of the peak precipitation regardless of the storm duration. In
contrast, in the temperate zone, peak precipitation at the storm center increases at a greater rate than the
precipitation at the storm boundaries. In fact there is a decrease in precipitation at the storm boundaries
for both 1 h and 3 h duration. This presents the possibility of a smaller storm extent with redistribution of
moisture to the storm center at higher temperatures. The redistribution of moisture from the storm boundary
to the storm center is less for longer durations (e.g., 3 h), suggesting thatthe convergence of moisture is more
prominent for shorter durations.
4. Conclusions
It is increasingly understood that extreme rainfall scaleswith atmospheric temperature at or above the Clausius-
Clapeyron rate, but how this relates to the spatial organization of events is not well known. The results here
show that within a storm burst, precipitation scaling is not spatially uniform, with moisture being redistributed
from the storm boundaries to the storm center as temperature increases. Whereas the positive scaling in the
total precipitation supports the assertion that the thermodynamic mechanism of increasing moisture capacity
dominates increases in precipitation intensity, the redistribution of moisture from the storm boundaries and
nonconstant precipitation scaling with radial distance from the storm center suggests that storm dynamics
may also be changing. The results therefore suggest that both thermodynamic and dynamic factors result in
a storm with greater precipitation intensity with more moisture concentrated at the center. This is consistent
with trends from regional climate modeling which predict precipitation structures with larger volume and
greater heterogeneity in the future [Guinard et al., 2015]. If the identied historical relationships were to be
maintained as global temperatures increase, more concentrated spatial storm events could be expected with
higher temperatures. This redistribution could have signicant implications for ood severity, as precipitation
may become both more intense and spatially concentrated in a warming climate. Future studies will focus
on replicating this work using observations from precipitation radars and linking synoptic weather patterns
to extreme precipitation patterns to better understand the drivers of precipitation extremes.
Andrews, F. T., B. F. W. Croke, and A. J. Jakeman (2011), An open software environment for hydrologi cal model assessment and development,
Environ. Modell. Software,26(10), 11711185, doi:10.1016/j.envsoft.2011.04.006.
Berg, P., C. Moseley, and J. O. Haerter (2013), Strong increase in convective precipitation in response to higher temperatures, Nat. Geosci.,
6(3), 181185, doi:10.1038/ngeo1731.
Collins, M., et al. (2013), Long-term Climate Change: Projections, Commitments and Irreversibility, in Climate Change 2013: The Physical
Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change,edited
by T. Stocker et al., pp. 10291136, Cambridge Univ. Press, Cambridge U. K., and New York.
Fujibe, F. (2013), Clausius-Clapeyron-like relationship in multidecadal changes of extreme short-term precipitation and temperature in
Japan, Atmos. Sci. Lett.,14(3), 127132, doi:10.1002/asl2.428.
Guinard,K.,A.Mailhot,andD.Caya(2015),Projectedchangesincharacteristics of precipitation spatial structures over North America,
Int. J. Climatol.,35(4), 596612, doi:10.1002/joc.4006.
Haerter, J. O., and P. Berg (2009), Unexpected rise in extreme precipitation caused by a shift in rain type?, Nat. Geosci.,2(6), 372373,
Hardwick Jones, R., S. Westra, and A. Sharma (2010), Observed relationships between extreme sub-daily precipitation, surface temperature,
and relative humidity, Geophys. Res. Lett.,37, L22805, doi:10.1029/2010GL045081.
Kirtman, B., et al. (2013), Near-term climate change: Projections and predictability, in Climate Change 2013: The Physical Science Basis.
Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by T. Stocker et al.,
pp. 9531028, Cambridge Univ. Press, Cambridge, U. K., and New York.
Lenderink, G., and E. van Meijgaard (2008), Increase in hourly precipitation extremes beyond expectations from temperature changes,
Nat. Geosci.,1(8), 511514, doi:10.1038/ngeo262.
Lenderink, G., H. Y. Mok, T. C. Lee, and G. J. van Oldenborgh (2011), Scaling and trends of hourly precipitation extremes in two different
climate zonesHong Kong and the Netherlands, Hydrol. Earth Syst. Sci.,15(9), 30333041, doi:10.5194/hess-15-3033-2011.
Li, J., A. Sharma, F. Johnson, and J. Evans (2015), Evaluating the effect of climate change on areal reduction factors using regional climate
model projections, J. Hydrol.,528, 419434, doi:10.1016/j.jhydrol.2015.06.067.
Mishra, V., J. M. Wallace, and D. P. Lettenmaier (2012), Relationship between hourly extreme precipitation and local air temperature in the
United States, Geophys. Res. Lett.,39, L16403, doi:10.1029/2012GL052790.
Molnar, P., S. Fatichi, L. Gaál, J. Szolgay, and P. Burlando (2015), Storm type effects on super ClausiusClapeyron scaling of intense rainstorm
properties with air temperature, Hydrol. Earth Syst. Sci.,19(4), 17531766, doi:10.5194/hess-19-1753-2015.
Geophysical Research Letters 10.1002/2016GL068509
The authors are grateful for the funding
support from the Australian Research
Council for this project. Westra was
supported by ARC Discovery project
DP150100411. The data set can be
obtained from
Panthou, G., A. Mailhot, E. Laurence, and G. Talbot (2014), Relationship between surface temperature and extreme rainfalls: A multi-time-scale
and event-based analysis, J. Hydrometeorol.,15(5), 19992011, doi:10.1175/JHM-D-14-0020.1.
Peel, M., B. Finlayson, and T. McMahon (2007), Updated world map of the Köppen-Geiger climate classication, Hydrol. Earth Syst. Sci.,11,
Rebora, N., and L. Ferraris (2006), The structure of convective rain cells at mid-latitudes, Adv. Geosci.,7,3135.
Trenberth, K. E., A. Dai, R. M. Rasmussen, and D. B. Parsons (2003), The changing character of precipitation, Bull. Am. Meteorol. Soc.,84(9),
12051217, doi:10.1175/BAMS-84-9-1205.
Utsumi, N., S. Seto, S. Kanae, E. E. Maeda, and T. Oki (2011), Does higher surface temperature intensify extreme precipitation?, Geophys. Res.
Lett.,38, L16708, doi:10.1029/2011GL048426.
von Hardenberg, J., L. Ferraris, and A. Provenzale (2003), The shape of convective rain cells, Geophys. Res. Lett.,30(24), 2280, doi:10.1029/
Wasko, C., and A. Sharma (2014), Quantile regression for investigating scaling of extreme precipitati on with temperature, Water Resour. Res.,
50, 36083614, doi:10.1002/2013WR015194.
Wasko, C., and A. Sharma (2015), Steeper temporal distribution of rain intensity at higher temperatures within Australian storms, Nat. Geosci.,
8(7), 527529, doi:10.1038/ngeo2456.
Wasko, C., A. Sharma, and F. Johnson (2015), Does storm duration modulate the extreme precipitation-temperature scaling relationship?,
Geophys. Res. Lett.,42, 87838790, doi:10.1002/2015GL066274.
Westra, S., and S. A. Sisson (2011), Detection of non-stationarity in precipitation extremes using a max-stable process model, J. Hydrol.,
406(12), 119128, doi:10.1016/j.jhydrol.2011.06.014.
Westra, S., R. Mehrotra, A. Sharma, and R. Srikanthan (2012), Continuous rainfall simulation: 1. A regionalized subdaily disaggregation
approach, Water Resour. Res.,48, W01535, doi:10.1029/2011WR010489.
Westra, S., L. Alexander, and F. Zwiers (2013), Global increasing trends in annual maximum daily precipitation, J. Clim.,26, 39043918,
Westra, S., H. J. Fowler, J. P. Evans, L. Alexander, P. Berg, F. Johnson, E. J. Kendon, G. Lenderink, and N. M. Roberts (2014), Future changes to the
intensity and frequency of short-duration extreme rainfall, Rev. Geophys.,52, 522555, doi:10.1002/2014RG000464.
Geophysical Research Letters 10.1002/2016GL068509
... Extreme rainfall is expected to change under warming with respect to its duration, frequency, and intensity (Goswami et al., 2006;Swain et al., 2018). Moreover, the increasing temperature may also change precipitation's temporal and spatial patterns (Wasko & Sharma, 2015;Wasko et al., 2016). It was reported that temporal patterns were less uniform and the peak precipitation was more intense as temperatures increase (Wasko and Sharma, 2015;Visser et al., 2021), and the spatial extent of storms had tended to decrease as well (Wasko et al., 2016). ...
... Moreover, the increasing temperature may also change precipitation's temporal and spatial patterns (Wasko & Sharma, 2015;Wasko et al., 2016). It was reported that temporal patterns were less uniform and the peak precipitation was more intense as temperatures increase (Wasko and Sharma, 2015;Visser et al., 2021), and the spatial extent of storms had tended to decrease as well (Wasko et al., 2016). Previous studies have typically used daily rainfall data to detect long-term trends in precipitation and extremes, as daily rainfall data are readily available and usually have long temporal series (Goswami et al., 2006;Miao et al., 2016;Papalexiou and Montanari, 2019). ...
Hourly precipitation data from 1971-2020, collected from 98 stations distributed across the Yellow River basin, were analyzed to detect changes in characteristics on rainfall and rainfall erosivity for all storms and storms with extreme erosivity (greater than 90th percentile). Results showed that over the past 50 years rainfall erosivity at both event and seasonal scale over the whole basin increased significantly (p < 0.05) with rates of 5.46% and 6.86% decade⁻¹, respectively, compared to the 1981-2010 average values. Approximate 80% of 98 stations showed increasing trends and 20% of stations had statistically significant trends (p < 0.1). The increase of rainfall erosivity resulted from the significant increasing trends of average storm precipitation (p < 0.1), duration (p < 0.1) , rainfall energy (p < 0.05) and maximum 1-h intensity (p < 0.05). In addition, the total extreme erosivity showed significant upward trends at a relative rate of 6.05% decade⁻¹ (p < 0.05). Extreme erosivity storms occurred more frequently and with higher rainfall energy during the study period (p < 0.05). Trends for seasonal total and extreme erosivity were also estimated based on daily rainfall data, and the changing magnitudes were similar to those based on hourly rainfall data, which suggested daily rainfall can be applied to detect interannual and long-term variations of rainfall erosivity in the absence of rainfall data with higher resolution. It was suggested that soil and water conservation strategies and vegetation projects conducted within the Yellow River basin should be continued and enhanced in the future.
... The CI values for practically all stations in the study region, indicate an increasing trend, which correlates to a considerable rise in severe precipitation occurrences in recent years. Global warming may be one of the main factors contributing to this phenomenon [75]. Research on the relationship between precipitation concentration and climate change is helpful for dealing with natural disasters caused by climate change. ...
Full-text available
Precipitation, as an important part of the hydrological cycle, is often related to flood and drought. In this study, we collected daily rainfall data from 21 rainfall stations in Shaanxi Province from 1961 to 2017, and calculated eight extreme climate indices. Annual and seasonal concentration indices (CI) were also calculated. The trends in the changes in precipitation were calculated using the M–K test and Sen’s slope. The results show that the precipitation correlation index and CI (concentration index) in Shaanxi Province are higher in the south and lower in the north. For the annual scale, the CI value ranges from 0.6369 to 0.6820, indicating that Shaanxi Province has a high precipitation concentration and an uneven distribution of annual precipitation. The eight extreme precipitation indices of most rainfall stations showed a downward trend during the study period, and more than half of the stations passed the 0.05 confidence interval test. Among them, the Z value of PRCPTOT (annual total precipitation in wet days) at Huashan station reached −6.5270. The lowest slope of PRCPTOT reached −14.3395. This shows that annual rainfall in Shaanxi Province has been decreasing in recent decades. These findings could be used to make decisions about water resources and drought risk management in Shaanxi Province, China.
... By breaking the total precipitation volume into its area and depth components, it can be seen that the change in precipitation in- tensity shows a signal very similar to that of the total precipitation volume. The spatial extent of the precipitation, meanwhile, decreases (-2 %) in the warmer climate, in line with the results of Wasko et al. (2016) and Armon et al. (2022). It is thus clear that it is not changes in the spatial extent of the system, but rather higher local intensities which drive the increase in total precipitation volume. ...
Extreme precipitation is a weather phenomenon with tremendous damaging potential for property and human life. As the intensity and frequency of such events is projected to increase in a warming climate, there is an urgent need to advance the existing knowledge on extreme precipitation processes, statistics and impacts across scales. To this end, a working group within the German-based project ClimXtreme, has been established to carry out multidisciplinary analyses of high-impact events. In this work, we provide a comprehensive assessment of a selected case, affecting the Berlin metropolitan region (Germany) on 29 June 2017, from the meteorological, impacts and climate perspectives, additionally estimating the contribution of climate change to its extremeness. Our analysis shows that this event occurred under the influence of a mid-tropospheric trough over western Europe and two short-wave surface lows over Britain and Poland, inducing relevant low-level wind convergence along the German-Polish border. Several thousand convective cells were triggered in the early morning of 29 June, displacing northwards slowly under the influence of a weak tropospheric flow (10 m s-1 at 500 hPa). A very moist and warm southwesterly flow was present south of the cyclone over Poland, in the presence of moderate Convective Available Potential Energy (CAPE). We identified the soil in the Alpine-Slovenian region as the major moisture source for this case (63 % of identified sources). Maximum precipitation amounted to 196 mm d-1, causing the largest insured losses due to a heavy precipitation event in the period 2002 to 2017 (€60 Mill.) over the area. A household-level survey revealed that the inundation duration was 4 to 12 times larger than other surveyed events in Germany in 2005, 2010 and 2014. The climate analysis showed return periods of over 100 years for daily aggregated precipitation, and up to 100 years and 10 years for 8 h and 1 h aggregations, respectively. The event was the 29th most extreme event in the 1951–2021 climatology in terms of severity and the second with respect to the number of convective cells triggered from 2001 to 2020 over Germany. The conditional attribution demonstrated that warming since the pre-industrial era caused a small, but significant increase of 4 % in total precipitation and 10 % for extreme intensities. The aerosol sensitivity experiments showed that increased anthropogenic aerosols induce larger cloud cover and probability of extreme precipitation (> 150 mm d-1). Our analysis allowed relating interconnected aspects of extreme precipitation. For instance, the link between the unique meteorological conditions of this case and its climate extremeness, or the extent to which this is attributable to already-observed anthropogenic climate change.
... Hence, most rainfall activities in the tropical region cannot be detected using the existing rainfall monitoring network. It has been reported that rising temperature would enhance convective moisture convergence Schumacher 2015, Shiru et al. 2020), which eventually would cause an increase in the amount of convective rainfall and decrease their spatial extents (Wasko et al., 2016). Therefore, much higher spatial resolution rainfall data will be required for tropical rainfall analysis in the near future. ...
Full-text available
Satellite-based precipitation (SBP) is emerging as a reliable source for high-resolution rainfall estimates over the globe. However, uncertainty in SBP is still significant, limiting their use without evaluation and often without bias correction. The bias correction of SBP remains a challenge for atmospheric scientists. The present study evaluated the performance of six SBPs, namely, SM2RAIN-ASCAT, IMERG, GSMaP, CHIRPS, PERSIANN-CDS and PERSIANN-CSS, in replicating observed daily rainfall at 364 stations over Peninsular Malaysia. The bias of the most suitable SBP was corrected using a novel machine learning (ML)-based bias-correction method. The proposed bias-correction method consists of an ML classifier to correct the bias in estimating rainfall occurrence and an ML regression model to correct the rainfall amount during rainfall events. Besides, the study evaluated the performance of different widely used ML algorithms for classification and regression to select the most suitable algorithms for bias correction. IMERG showed better performance, showing a higher correlation coefficient (R²) of 0.57 and Kling-Gupta Efficiency (KGE) of 0.5 compared to the other products. The performance of random forest (RF) was better than the k-nearest neighbourhood (KNN) for both classification and regression. RF classified the rainfall events with a skill score of 0.38 and estimated the rainfall amount during rainfall events with the modified index of agreement (md) of 0.56. Comparison of IMERG and bias-corrected IMERG (BIMERG) revealed an average reduction in RMSE by 55% in simulating observed rainfall. The proposed bias correction method performed much better when compared with the conventional bias correction methods such as linear scaling and quantile regression. The BIMERG could reliably replicate the spatial distribution of heavy rainfall events, indicating its potential for hydro-climatic studies like flood and drought monitoring in the study area.
... This means that, although the total rain area of HPEs shrinks, their cores are getting larger in future simulations. Similar findings were reported by Peleg et al. (2018) using historic radar observations, and temperature as a proxy, over the eastern Mediterranean and by Wasko et al. (2016) using rain gauges in Australia. Both studies showed that total rain area and the convective core area scale with temperature in opposite directions: total area exhibits a negative scaling, while the area of the convective cores is positively scaled with temperature; this is probably related to an enhanced moisture convergence into the convective cores from the total storm extent. ...
Full-text available
Heavy precipitation events (HPEs) can lead to deadly and costly natural disasters and are critical to the hydrological budget in regions where rainfall variability is high and water resources depend on individual storms. Thus, reliable projections of such events in the future are needed. To provide high‐resolution projections under the RCP8.5 scenario for HPEs at the end of the 21st century, and to understand the changes in sub‐hourly to daily rainfall patterns, weather research and forecasting (WRF) model simulations of 41 historic HPEs in the eastern Mediterranean are compared with “pseudo global warming” simulations of the same events. This paper presents the changes in rainfall patterns in future storms, decomposed into storms' mean conditional rain rate, duration, and area. A major decrease in rainfall accumulation (−30% averaged across events) is found throughout future HPEs. This decrease results from a substantial reduction of the rain area of storms (−40%) and occurs despite an increase in the mean conditional rain intensity (+15%). The duration of the HPEs decreases (−9%) in future simulations. Regionally maximal 10‐min rain rates increase (+22%), whereas over most of the region, long‐duration rain rates decrease. The consistency of results across events, driven by varying synoptic conditions, suggests that these changes have low sensitivity to the specific synoptic evolution during the events. Future HPEs in the eastern Mediterranean will therefore likely be drier and more spatiotemporally concentrated, with substantial implications on hydrological outcomes of storms.
The present study examined extreme rainfall events (EREs) in central India during the summer monsoon season, focusing on their spatial characteristics. A station‐based gridded and a station‐satellite‐blended dataset was used to examine long‐term and recent variations in precipitation characteristics for 50 years (1951–2000) and 38 years (1981–2018), respectively. A precipitation system approach (PSA) was applied to identify the ERE precipitation systems and categorized spatial sizes of ERE systems into three categories: sporadic, intermediate, and massive ERE precipitation systems. Conventionally, the ERE frequency is equal to the total number of ERE grids, whereas PSA counts ERE systems. The sporadic precipitation grid contributes 42% of all the ERE grids, and sporadic EREs frequency increases in the long‐term. The long‐term trend of intermediate and massive EREs does not increase and quite sharply increases, respectively, and these EREs are also intensifying. Recent 38 years have shown a reverse in the trends of ERE characteristics, the frequency and intensity of intermediate and massive EREs have decreased, whereas the massive ERE systems are broadening. In general, different sizes of precipitation systems associated with phenomena such as thunderstorms, meso‐scale systems or synoptic systems occur in the environment. A large difference in the long‐term analysis of extreme rainfall events over central India can be seen between the grid based conventional approach and the precipitation system approach. Precipitation system approach gives more physical meaning of the extreme rainfall events long‐term changes.
With increasing focus on large-scale planning and allocation of resources for protection against future flood risk, it is necessary to analyze and improve the deficiencies in the existing flood modeling approach through a better understanding of the interactions among overland flow, subsurface flow, and river hydrodynamics. Recent studies have shown that it is possible to improve flood inundation modeling and mapping using physically based distributed models that incorporate the observable data through assimilation and simulate hydrologic fluxes using the fundamental laws of conservation of mass at multiple spatiotemporal scales. However, despite the significance of distributed modeling in hydrology, it has received relatively less attention within the context of flood hazards. While significant strides have been taken by the research community in estimating streamflow, surface-subsurface volumes, and soil moisture; predicting the flood depths and extents accurately across large scales remains a challenge for distributed models. This chapter addresses the challenges that exist within the flood modeling approach while advocating for a large-scale holistic approach that goes beyond streamflow prediction to provide flood inundation extents and depths. As the way forward, this chapter presents an overview of physically based distributed modeling by providing valuable insights into the essential characteristics of distributed models and highlighting the important factors that need to be considered for large-scale flood simulation. Finally, the chapter provides a prototype modeling framework for large-scale distributed flood modeling using the Interconnected Channel and Pond Routing (ICPR) model for the Wabash River Basin in the United States to illustrate how distributed models can be applied for flood inundation mapping for a large watershed.
Full-text available
Flood modeling and flood forecasting uncertainties are essential to estimate the accuracy of modeling. Uncertainty permits us to evaluate the performance of a model and allows us to account for discharge, flood extent, and flood depth in probabilistic terms. The classical approach used to consist of calibration/validation to reduce the uncertainty of the model with different methods. However, multi-model flood modeling has emerged as a method to estimate uncertainty in flood modeling. It gained awareness from the Bayesian modeling approach using Markov chain/stochastic to estimate the likelihood of parameters derived from classical sensitivity analysis. Furthermore, the development of ensemble methods in the weather forecast/climate field has led to a practical way to estimate the uncertainty of models and is currently used in flood forecast modeling deriving from the strong uncertainty associated with atmospheric initial conditions. This study reviewed the uncertainties associated with flood modeling and forecasting and the various techniques for reducing the uncertainty of flood models
Variability and spatiotemporal changes in precipitation characteristics can have profound socioenvironmental impacts. Several studies have shown that the frequency and/or magnitude of precipitation events have changed over the contiguous United States (CONUS) in the past decades. Most previous studies used only one precipitation dataset and only investigated mean or extreme precipitation. Here, using 6 gridded daily precipitation datasets, we show that there are substantial discrepancies in the changes in characteristics of both extreme and non-extreme precipitation events from 1983 to 2017. Our results highlight that using a single record to study precipitation changes can potentially lead to biased results. Using different datasets enables examining the overall agreements and discrepancies in precipitation characteristics. For example, we show that almost all datasets agree that some areas show statistically significant changes in the annual precipitation maxima; however, the locations and signs of changes are not consistent across datasets. There is a relative agreement between datasets on changes in the total annual precipitation. When examining other percentiles of the precipitation distribution, including non-extreme values, however, we find widespread discrepancies among different precipitation products (e.g., what part of the precipitation distribution is changing). In fact, depending on the source of data, there exist opposing trends and patterns of change in precipitation characteristics. This highlights the need to further investigate non-extreme precipitation events to unravel potential non-extreme but “unexpected” or “unusual” patterns. Finally, we argue that protocols for data selection are needed to address the issue of inter-data variability and to ensure reliability of statistical analysis.
Abstract North‐central Colorado experienced an extreme precipitation event (EPE) in September 2013, during which the equivalent of 80% of the region's annual average precipitation fell in a few days. Widespread flooding occurred above ground, but the short‐ and long‐term subsurface response remains unclear. The objective of the study is to better understand the dynamic subsurface response, namely how the water table and soil water storage responded to a large amount of infiltration in a short period of time and how the hydrologic properties of the subsurface influence the response. Better understanding of subsurface response to EPEs is expected to increase with the advent of more intense and frequent EPEs in the coming decades. A one‐dimensional subsurface flow model using HYDRUS‐1D, was built to simulate and examine infiltration of an EPE at a site in the Boulder Creek Watershed, Colorado. Model calibration was conducted with local field data to fit site observations over a 6‐yr period. A rapid water table depth response in field observations was observed, with the modeled subsurface storing water for 18 mo acting as a hydro‐buffer during recovery. To examine influence on model results, a sensitivity study of soil hydraulic parameters was conducted. The sensitivity study found that changes in n, an empirical parameter related to pore‐size distribution, most significantly affects water table depth. The implications are that one‐dimensional models may provide useful estimates of water table fluctuations and subsurface hydro‐buffer capacities in response to EPEs, which could be of use to regions preparing for EPE effect on water resources.
Full-text available
Climate models have continued to be developed and improved since the AR4, and many models have been extended into Earth System models by including the representation of biogeochemical cycles important to climate change. These models allow for policy-relevant calculations such as the carbon dioxide (CO2) emissions compatible with a specified climate stabilization target. In addition, the range of climate variables and processes that have been evaluated has greatly expanded, and differences between models and observations are increasingly quantified using ‘performance metrics’. In this chapter, model evaluation covers simulation of the mean climate, of historical climate change, of variability on multiple time scales and of regional modes of variability. This evaluation is based on recent internationally coordinated model experiments, including simulations of historic and paleo climate, specialized experiments designed to provide insight into key climate processes and feedbacks and regional climate downscaling. Figure 9.44 provides an overview of model capabilities as assessed in this chapter, including improvements, or lack thereof, relative to models assessed in the AR4. The chapter concludes with an assessment of recent work connecting model performance to the detection and attribution of climate change as well as to future projections.
Full-text available
The mechanisms that cause changes in precipitation, as well as the resulting storm dynamics, under potential future warming remain debated. Measured sensitivities of precipitation to temperature variations in the present climate have been used to constrain model predictions, debate precipitation mechanisms and speculate on future changes to precipitation and flooding. Here, we analyse data sets of precipitation measurements at 6-min resolution from 79 locations throughout Australia, covering a broad range of climate zones, along with sub-daily temperature measurements of varying resolution. We investigate the relationship between temporal patterns of precipitation intensity within storm bursts and temperature variations in the present climate by calculating the scaling of the precipitation fractions within each storm burst. We find that in the present climate, a less uniform temporal pattern of precipitation-more intense peak precipitation and weaker precipitation during less intense times-is found at higher temperatures, regardless of the climatic region and season. We suggest invigorating storm dynamics could be associated with the warming temperatures expected over the course of the twenty-first century, which could lead to increases in the magnitude and frequency of short-duration floods.
Full-text available
Extreme precipitation is thought to increase with warming at rates similar to or greater than the water vapour holding capacity of the air at ~ 7% °C−1, the so-called Clausius–Clapeyron (CC) rate. We present an empirical study of the variability in the rates of increase in precipitation intensity with air temperature using 30 years of 10 min and 1 h data from 59 stations in Switzerland. The analysis is conducted on storm events rather than fixed interval data, and divided into storm type subsets based on the presence of lightning which is expected to indicate convection. The average rates of increase in extremes (95th percentile) of mean event intensity computed from 10 min data are 6.5% °C−1 (no-lightning events), 8.9% °C−1 (lightning events) and 10.7% °C−1 (all events combined). For peak 10 min intensities during an event the rates are 6.9% °C−1 (no-lightning events), 9.3% °C−1 (lightning events) and 13.0% °C−1 (all events combined). Mixing of the two storm types exaggerates the relations to air temperature. Doubled CC rates reported by other studies are an exception in our data set, even in convective rain. The large spatial variability in scaling rates across Switzerland suggests that both local (orographic) and regional effects limit moisture supply and availability in Alpine environments, especially in mountain valleys. The estimated number of convective events has increased across Switzerland in the last 30 years, with 30% of the stations showing statistically significant changes. The changes in intense convective storms with higher temperatures may be relevant for hydrological risk connected with those events in the future.
Full-text available
Evidence that extreme rainfall intensity is increasing at the global scale has strengthened considerably in recent years. Research now indicates that the greatest increases are likely to occur in short-duration storms lasting less than a day, potentially leading to an increase in the magnitude and frequency of flash floods. This review examines the evidence for sub-daily extreme rainfall intensification due to anthropogenic climate change, and describes our current physical understanding of the association between sub-daily extreme rainfall intensity and atmospheric temperature. We also examine the nature, quality and quantity of information needed to allow society to adapt successfully to predicted future changes, and discuss the roles of observational and modelling studies in helping us to better understand the physical processes that can influence sub-daily extreme rainfall characteristics. We conclude by describing the types of research required to produce a more thorough understanding of the relationships between local scale thermodynamic effects, large-scale atmospheric circulation and sub-daily extreme rainfall intensity.
Predicting future precipitation extremes is difficult and therefore many studies have used the historical relationship between precipitation intensity and temperature to consider what might occur in a future warmer climate. In general extreme precipitation intensity is expected to increase as temperatures increase. However, in tropical areas it has been observed that, for higher temperatures, lower precipitation intensities occur, contradicting the expected relationship. This has been thought to be due to limits in moisture availability. In this work we show that the negative scaling found in previous studies may be a result of the analysis methods. By conditioning the precipitation intensity and temperature relationship on storm duration we demonstrate that positive scaling of precipitation intensity with temperature in tropical regions of Australia is possible. We argue that methods for estimating scaling relationships should be modified to include storm duration.
Recent studies have examined the relationship between the intensity of extreme rainfall and temperature. Two main reasons justify this interest. First, the moisture-holding capacity of the atmosphere is governed by the Clausius-Clapeyron (CC) equation. Second, the temperature dependence of extreme-intensity rainfalls should follow a similar relationship assuming relative humidity remains constant and extreme rainfalls are driven by the actual water content of the atmosphere. The relationship between extreme rainfall intensity and air temperature (P-extr-T-a) was assessed by analyzing maximum daily rainfall intensities for durations ranging from 5 min to 12 h for more than 100 meteorological stations across Canada. Different factors that could influence this relationship have been analyzed. It appears that the duration and the climatic region have a strong influence on this relationship. For short durations, the P-extr-T-a relationship is close to the CC scaling for coastal regions while a super-CC scaling followed by an upper limit is observed for inland regions. As the duration increases, the slope of the relationship P-extr-T-a decreases for all regions. The shape of the P-extr-T-a curve is not sensitive to the percentile or season. Complementary analyses have been carried out to understand the departures from the expected Clausius Clapeyron scaling. The relationship between dewpoint temperature and extreme rainfall intensity shows that the relative humidity is a limiting factor for inland regions, but not for coastal regions. Using hourly rainfall series, an event-based analysis is proposed in order to understand other deviations (super-CC, sub-CC, and monotonic decrease). The analyses suggest that the observed scaling is primarily due to the rainfall event dynamic.
Spatial structures of hourly precipitation fields were studied from three simulations of the Canadian Regional Climate Model (CRCM) and observations of the National Centers for Environmental Prediction (NCEP) Stage IV analysis. Each precipitation structure, defined as a contiguous area of precipitation above a given threshold, was analysed through geometric characteristics (position, area, major and minor axes, eccentricity, orientation) and intensity characteristics (volume, mean and maximum intensities, precipitation distribution within the structure) for 16 climatic regions covering North America. While providing new insights on the spatial facet of precipitation, this study aimed to: (1) assess the performance of the CRCM to reproduce observed precipitation structures and (2) analyse the changes in precipitation structures between historical (1961–1990) and future (2071–2100) periods. In addition, the effect of internal variability was investigated using two CGCM-driven CRCM simulations. In order to assess the CRCM performance, a reanalysis-driven CRCM simulation was first compared with observations and then with CGCM-driven CRCM simulations. Results suggest that reanalysis-driven CRCM precipitation structures displayed intensities spatially more homogeneous than observed ones for the central and eastern United States and showed significantly lower precipitation volumes, intensities and areas. However, annual cycles of characteristic values were well reproduced. In addition, CGCM-driven CRCM showed significantly lower precipitation volumes and intensities during summer months for southeastern regions when compared to reanalysis-driven CRCM. Precipitation structures were also larger and shifted further north. Boundary conditions seemed to influence mainly central and eastern regions of North America. In future climate, results suggest more convective summer precipitations for central and eastern regions (increases in volumes, intensities and heterogeneity of structures), drier spring and summer conditions for southwestern regions (decreases in numbers of structures), wetter winter and spring conditions for northern regions (increases in numbers of structures) and wetter autumn conditions for southeastern regions (increases in volumes and intensities).