ArticlePDF Available

Global Analysis of Climate Change Projection Effects on Atmospheric Rivers

Authors:

Abstract and Figures

A uniform, global approach is used to quantify how atmospheric rivers (ARs) change between Coupled Model Intercomparison Project Phase 5 historical simulations and future projections under the Representative Concentration Pathway (RCP) 4.5 and RCP8.5 warming scenarios. The projections indicate that while there will be ~10% fewer ARs in the future, the ARs will be ~25% longer, ~25% wider, and exhibit stronger integrated water vapor transports (IVTs) under RCP8.5. These changes result in pronounced increases in the frequency (IVT strength) of AR conditions under RCP8.5: ~50% (25%) globally, ~50% (20%) in the northern midlatitudes, and ~60% (20%) in the southern midlatitudes. The models exhibit systematic low biases across the midlatitudes in replicating historical AR frequency (~10%), zonal IVT (~15%), and meridional IVT (~25%), with sizable intermodel differences. A more detailed examination of six regions strongly impacted by ARs suggests that the western United States, northwestern Europe, and southwestern South America exhibit considerable intermodel differences in projected changes in ARs.
Content may be subject to copyright.
Global Analysis of Climate Change Projection
Effects on Atmospheric Rivers
Vicky Espinoza
1,2
, Duane E. Waliser
2
, Bin Guan
2,3
, David A. Lavers
4
,
and F. Martin Ralph
5
1
Sonny Astani Civil and Environmental Engineering Department, University of Southern California, Los Angeles, CA, USA,
2
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA,
3
Joint Institute for Regional Earth System
Science and Engineering, University of California, Los Angeles, CA, USA,
4
European Centre for Medium-Range Weather
Forecasts, Reading, UK,
5
Center for Western Weather and Water Extremes, Scripps Institution of Oceanography at University
of California, San Diego, CA, USA
Abstract A uniform, global approach is used to quantify how atmospheric rivers (ARs) change between
Coupled Model Intercomparison Project Phase 5 historical simulations and future projections under the
Representative Concentration Pathway (RCP) 4.5 and RCP8.5 warming scenarios. The projections indicate that
while there will be ~10% fewer ARs in the future, the ARs will be ~25% longer, ~25% wider, and exhibit
stronger integrated water vapor transports (IVTs) under RCP8.5. These changes result in pronounced
increases in the frequency (IVT strength) of AR conditions under RCP8.5: ~50% (25%) globally, ~50% (20%) in
the northern midlatitudes, and ~60% (20%) in the southern midlatitudes. The models exhibit systematic
low biases across the midlatitudes in replicating historical AR frequency (~10%), zonal IVT (~15%), and
meridional IVT (~25%), with sizable intermodel differences. A more detailed examination of six regions
strongly impacted by ARs suggests that the western United States, northwestern Europe, and
southwestern South America exhibit considerable intermodel differences in projected changes in ARs.
Plain Language Summary Atmospheric rivers (ARs) are elongated strands of horizontal water
vapor transport, accounting for over 90% of the poleward water vapor transport across midlatitudes.
These rivers in the skyhave important implications for extreme precipitation when they make landfall,
particularly along the west coasts of many midlatitude continents (e.g., North America, South America, and
West Europe) due to orographic lifting. ARs are important contributors to extreme weather and precipitation
events, and while their presence can contribute to benecial rainfall and snowfall, which can mitigate
droughts, they can also lead to ooding and extreme winds. This study takes a uniform, global approach
that is used to quantify how ARs change between Coupled Model Intercomparison Project Phase 5 historical
simulations and future projections under the Representative Concentration Pathway (RCP) 4.5 and RCP8.5
warming scenarios globally. The projections indicate that while there will be ~10% fewer ARs in the future,
the ARs will be ~25% longer, ~25% wider, and exhibit stronger integrated water vapor transports under
RCP8.5. These changes result in pronounced increases in the frequency (integrated water vapor transport
strength) of AR conditions under RCP8.5: ~50% (25%) globally, ~50% (20%) in the northern midlatitudes,
and ~60% (20%) in the southern midlatitudes.
1. Introduction
Atmospheric rivers (ARs) are elongated strands of horizontal water vapor transport, accounting for over 90%
of the poleward water vapor transport across midlatitudes (Zhu & Newell, 1998). These rivers in the skyhave
important implications for extreme precipitation when they make landfall, particularly along the west coasts
of many midlatitude continents (e.g., North America, South America, and western Europe) and especially
when encountering orographic lifting (e.g., Neiman et al., 2009; Ralph et al., 2004). ARs are important contri-
butors to extreme weather and precipitation events, and while their presence can contribute to benecial
rainfall and snowfall (Dettinger et al., 2011; Guan et al., 2010), which can mitigate droughts (Dettinger,
2013), they can also lead to ooding (e.g., Lavers et al., 2011; Leung & Qian, 2009; Neiman et al., 2011;
Ralph et al., 2006, 2013; Ralph & Dettinger, 2011) and extreme winds (Waliser & Guan, 2017). These important
impacts have motivated a number of climate change studies on ARs, with studies to date focusing mainly
only on the west coasts of North America (Dettinger, 2011; Gao et al., 2015; Hagos et al., 2016; Payne &
ESPINOZA ET AL. 4299
Geophysical Research Letters
RESEARCH LETTER
10.1029/2017GL076968
Key Points:
Globally, atmospheric rivers (ARs) are
~10% fewer, ~25% longer, ~25%
wider, and with stronger moisture
transport under the RCP8.5 scenario
In the midlatitudes where ARs are
most frequent, AR conditions are
~5060% more frequent and AR
transport is ~20% stronger in
the future
Systematic low biases exist in the
midlatitudes in historical AR frequency
(~10%), zonal (~15%), and meridional
(~25%) moisture transport
Supporting Information:
Supporting Information S1
Correspondence to:
D. E. Waliser,
duane.e.waliser@jpl.nasa.gov
Citation:
Espinoza, V., Waliser, D. E., Guan, B.,
Lavers, D. A., & Ralph, F. M. (2018).
Global analysis of climate change
projection effects on atmospheric rivers.
Geophysical Research Letters,45,
42994308. https://doi.org/10.1029/
2017GL076968
Received 26 JUL 2017
Accepted 10 APR 2018
Accepted article online 19 APR 2018
Published online 7 MAY 2018
©2018. American Geophysical Union.
All Rights Reserved.
Magnusdottir, 2015; Pierce et al., 2013; Radićet al., 2015; Shields & Kiehl, 2016a, 2016b; Warner et al., 2015)
and Europe (Gao et al., 2016; Lavers et al., 2013; Ramos et al., 2016; Shields & Kiehl, 2016a).
The rst climate change study on ARs was conducted by Dettinger (2011) and focused on landfalling ARs
in California using seven Coupled Model Intercomparison Project (CMIP) Phase 3 models with the A2
greenhouse-gas emissions scenario. This study found AR frequency increases of about 30% depending
on the model by the end of the 21st century and noted increases in storm temperature, length of AR
season, and peak AR intensity values. Note that climate change studies on ARs have generally dened
AR frequency as the fraction of days a particular grid point has an AR detected over it. Warner et al.
(2015) extended the consideration to a larger area along the west coast of North America using 10
CMIP5 (Taylor et al., 2012) models for a historical period (19701999) and projection period (20702099)
for the Representative Concentration Pathway (RCP) 8.5 warming scenario. This study conducted a multi-
model mean (MMM) analysis with results indicating ~230290% increase in AR days. The analysis also
showed that there would be an increase in extreme values of integrated water vapor transport (IVT)
magnitude and IWV of ~30%. Precipitation values for the MMM showed ~1539% increase. Using more
CMIP5 models (a total of 24), Gao et al. (2015) indicated an ~50600% increase in AR days under
RCP8.5, depending on the season and landfall location along western North America. Another study by
Payne and Magnusdottir (2015) using 28 CMIP5 models found 2335% increases in projected AR landfall
dates in this region under RCP8.5. Hagos et al. (2016) showed increases in projected AR landfall days by
35% under RCP8.5 based on a 29-member ensemble of the National Center for Atmospheric Research
Community Earth System Model. Other studies have also examined projected AR changes in western
North America, as summarized in Table 1.
For the European region, Lavers et al. (2013) used ve CMIP5 models, comparing historical (19802005) and
projection (20742099) periods for the RCP4.5 and RCP8.5 scenarios. This study also used an IVT-based
threshold approach for detecting ARs. This study found that AR frequency approximately doubled in
Britain under RCP8.5 and determined that the change was dominated by the thermodynamic (moistening)
response to warming rather than from the inuence of wind changes. Gao et al. (2016) conducted a study
with a focus on comparing the inuences of thermodynamic and dynamic effects on ARs and the quantica-
tion of the number of AR days across the European sector. By using 24 CMIP5 models, this study found that
AR frequency increased by ~127275% by the end of the century under RCP8.5. Not only did the study nd
that the projected increases in AR frequency were inuenced by thermodynamic processes but found that
variability in wind speed and direction related to shifts in the midlatitude jet stream played a dominant role
in the changes of ARs in the European sector. Other studies have also examined projected AR changes in
Europe, as summarized in Table 1.
Table 1
Comparison of Mean Changes in AR Frequency (Percent of Time Steps) and IVT (kg · m
1
·s
1
) Between the Current Study and Previous Studies for the Western U.S. and
Western Europe
Publication Historical period Projection period Geographic region AR Freq (± %) AR IVT (± %)
Dettinger (2011) 19612000 20462065; 20812100 CA Coast +30 +10
Pierce et al. (2013) 19851994 2060s CA Coast +25100 --
Warner et al. (2015) 19701999 20702099 U.S. West Coast +230290 +30
Payne and Magnusdottir (2015) 19802005 20702100 U.S. West Coast +2335 --
Gao et al. (2015) 19752004 20702099 U.S. West Coast +50600 --
Hagos et al. (2016) 19202005 20062099 U.S. West Coast +35 --
Shields and Kiehl (2016a) 19602005 20552100 U.S. West Coast +8 --
Espinoza et al. (2018, current study) 19792002 20732096 U.S. West Coast +45 +30
Lavers et al. (2013) 19802005 20742099 W. Europe +50100 --
Gao et al. (2016) 19752004 20702099 W. Europe +127275 +2050
Ramos et al. (2016) 19802005 20742099 Europe +100300 +30
Shields and Kiehl (2016a) 19602005 20552100 North Atlantic +4 --
Espinoza et al. (2018, current study) 19792002 20732096 W. Europe +60 +30
Note. The bold (italic) region is previous studies focusing on the U.S. West Coast (western Europe). The studies are ordered from oldest to most recent within each
geographic region (bold and italic). Note that each of the studies mentioned above differ in their methodologies, models used, and their study periods limiting
their comparability.
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4300
While all the studies discussed and cited above have tended toward the same general conclusions (Table 1),
that is, nding an increase in AR frequency and IVT, they have been limited to two regions in the Northern
Hemisphere. A uniform global assessment of climate change impacts on ARs has not been performed despite
the global presence and impacts of ARs (Guan & Waliser, 2015; Waliser et al., 2012; Waliser & Guan, 2017; Zhu
& Newell, 1998). For example, despite the number of studies performed on western North America and
western Europe, the differences in data sets and methodologies used make it challenging to use these stu-
dies to compare impacts of climate change effects on ARs in these two regions. This study addresses this
research gap by analyzing climate change impacts on AR frequencies and IVT using a globally consistent
approach on historical climate simulations and future projections of climate change from CMIP5.
2. Models and Methodology
2.1. CMIP5 Model Data
IVT values were constructed from daily values of 3-D wind and water vapor model outputs at four pressure
levels between 500 and 1,000 hPa inclusive, namely, the data described in Lavers et al. (2015; their
Table S1). We used 21 out of the 22 models examined in that study because the IVT data for one model were
not available at the time of this analysis. The horizontal resolution of the models ranges from 1.125° to 2.813°.
The study periods were 19792002 from the historical simulations and 20732096 from the RCP4.5 and
RCP8.5 scenarios, determined as a time frame that contains the maximum number of overlapping years
among all the models and one that spans the same number of years for the historical and the two RCP runs.
The more stringent requirement on the consistency in data set period resulted in three fewer years included
for analysis relative to Lavers et al. (2015).
2.2. AR Global Detection Algorithm
The AR global detection algorithm introduced in Guan and Waliser (2015) was used. Notable AR criteria used
in the algorithm include IVT magnitude at each grid cell within a contiguous region (object) being above
the 85th percentile for that grid cell and season, the length of the object being greater than 2000 km, and
the length-to-width ratio greater than 2. For a given model, AR detection for the historical simulation and
future projections (RCP4.5 and RCP8.5) are all based on the IVT 85th percentile derived from the historical
simulation. AR detection and subsequent calculation of AR frequency and IVT are based on the respective
horizontal resolution of each individual model. Regridding to a common grid is done only when calculating
the multimodel ensemble mean. It is expected that the inherent differences between different models are
much larger than the sensitivity of the results to the choice of horizontal resolution used for AR detection
and subsequent calculations (e.g., Guan & Waliser, 2017). In that regard, the step at which regridding is intro-
duced into the calculations (e.g., before or after AR detection) is not an important consideration in the current
analysis. AR detection results based on the Guan and Waliser (2015) algorithm was found to be consistent
with regional AR detection methods developed for western North America (Neiman et al., 2008), Britain
(Lavers et al., 2011), and East Antarctica (Gorodetskaya et al., 2014), with over ~90% agreement in detected
AR landfall dates.
2.3. AR Frequency and IVT Analysis
The AR frequency is calculated as the number of AR days detected at each grid cell for the given historical or
future projection period normalized by the total number of days in the given period. Mean AR IVT at each grid
cell is based on averaging the IVT values over the days detected as ARs. Once AR frequency and IVT have been
computed for each individual model, they are (bi-linearly) interpolated to a common 1.5° × 1.5° grid (which
matches the grid of the ERA-Interim reanalysis used as the observational reference) to create MMM maps.
To illustrate the calculation procedures, a set of histograms, shown in Figure 1a, are created for a single loca-
tion in the southeast Pacic Ocean (61°S, 216.25°E) from the Geophysical Fluid Dynamics Laboratory Coupled
Model 3 (GFDL-CM3) model simulations. The red (blue) histogram includes all IVT values at that point for the
historical (RCP8.5 projection) period. The number of IVT values sampled in each histogram is shown in the
plots legend. Additional histograms delineate the IVT values of the detected ARs for the historical (pink)
and RCP8.5 (light blue) simulationswith both using the IVT 85th percentile derived from the historical per-
iod for detecting ARs. For this location and model, the histograms show that the RCP8.5 scenario results in a
signicant increase (~140%) in the number of AR days given the historical IVT threshold. An additional
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4301
histogram (green) shows the top N AR IVT values from the RCP8.5, where N is determined by the number
of AR days in the historical simulation; thus, the pink and green histograms have the same number of AR
IVT values.
The AR frequency results discussed in section 3 (e.g., shadings in Figure 2) are essentially referring to AR fre-
quencies as represented by the light blue and pink histograms. The AR IVT values discussed below (e.g.,
vectors in Figure 2) are based on averages of the IVT values within the pink and green histograms. Thus,
the mean AR IVT to be compared below between the historical and future scenarios is based on averaging
over the same number of the most extreme IVT values from each simulation.
3. Results
3.1. Historical Simulations
AR frequency and IVT for the historical, RCP4.5 and RCP8.5 simulations are shown in Figure 2. Also shown is
the ERA-Interim AR frequency and IVT for 19792002 (Figure 2a; cf. Guan & Waliser, 2015) as the observational
Figure 1. (a) Histograms of IVT values (kg · m
1
·s
1
) at a single grid point (61°S, 216.25°E) in the South Pacic Ocean from
the GFDL-CM3 simulations. The red (dark blue) histogram includes all (i.e., both AR and non-AR grid cells) IVT values for the
historical period, that is, 19792002 (RCP8.5 period, i.e., 20732096). The pink (light blue) histogram includes the IVT
values associated with only AR events in the historical period (AR events in the RCP8.5 scenario based on the AR IVT
threshold from the historical period). The green histogram includes the top N AR IVT values for the RCP8.5 scenario,
where N is determined by number of AR events from the historical simulation (i.e., 713). Thus, the number of IVT values
contributing to the pink and green histograms is the same. Multimodel ensemble histograms of (b) IVT values at each
grid cell within the ARs, (c) AR lengths (km), and (d) AR widths (km) for the historical (blue) and RCP8.5 (orange) simulations,
with vertical bars representing the overall mean. N values in (b) are the total number of AR grid cells; N values in (c) and
(d) are the total number of AR objects.
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4302
reference. Comparing Figures 2a and 2b shows that the MMM is a good representation of the observational
reference. For AR frequency, and zonal and meridional IVT components, the spatial correlations (root mean
square error) between the historical simulation and the reference are 0.98, 0.93, and 0.97 (0.93%,
93.14 kg · m
1
·s
1
, and 51.79 kg · m
1
·s
1
), respectively. The good quantitative agreement between
Figures 2a and 2b mainly stems from the MMM capturing the strong latitudinal dependence of the AR
frequency and IVT values and the dominant zonal asymmetries of the midlatitude patterns (see red and
blue lines in Figures 3c, 3f, and 3i). Direct comparison of the MMM with ERA-Interim shows that the AR
frequencies are generally biased low by ~10% in midlatitude regions, with zonal (meridional) IVT biased
low by ~15% (25%), particularly in the Southern Ocean. The relatively good MMM representations of AR
frequency and IVT provide some condence to now consider the MMM projected changes in AR frequency
and IVT (section 3.2). Subsequent to that will be a discussion of the intermodel differences in historical
simulation biases (section 3.3) and projected changes (section 3.4) in AR frequency and IVT.
Figure 2. AR frequency (shading; percent of time steps) and IVT (vectors; kg · m
1
·s
1
) for (a) ERA-Interim reanalysis for
the historical period (19792002) with six green boxes depicting regions analyzed in Figures S2 and S3, (b) the MMM for the
21 CMIP5 models analyzed in this study for the historical period (19792002), (c) RCP4.5 warming scenario (20732096),
and (d) RCP8.5 warming scenario (20732096), (e) the difference between (c) and (b) with six green boxes depicting
regions analyzed in Figures S2 and S3, and (f) the difference between (d) and (b). Vector magnitudes are indicated by both
their length and their color based on the blue color bar.
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4303
3.2. Future Projections
Figures 2c and 2d illustrate the MMM AR frequency and IVT values for the RCP4.5 and RCP8.5 simulation. For
the RCP8.5 warming scenario, the global mean AR frequency and IVT increase by 49% and 23%, respectively.
Most evident is the considerable increase in AR frequency in the midlatitudes, with values, for example, along
the Southern Ocean (i.e., 3060°S area average) rising from around 612% for the historical simulations to 12
16% and 1420% for the RCP4.5 and RCP8.5 simulations, respectively. For the North Pacic (i.e., 3060°N,
120240° area average) and North Atlantic (i.e., 3060°N, 270360° area average), the AR frequency changes
from around 1012% for the historical simulations to 1416% and 1618% for the RCP4.5 and RCP8.5
simulations, respectively. Similarly, there are marked increases in IVT magnitude. For example, IVT values in
the North Pacic and North Atlantic, rise from about 350 kg · m
1
·s
1
in the historical simulation to about
420 kg · m
1
·s
1
for RCP8.5 scenario. In the Southern Ocean, the IVT values change from 364 to
434 kg · m
1
·s
1
between the historical and RCP8.5 simulations.
Figures 2e and 2f show the changes between the RCP4.5 and RCP8.5 warming scenarios and the historical
simulation, respectively. Referring to the histograms in Figure 1a (see section 2.3), the AR frequency
differences in Figure 2f refer to the differences between the example pink and light blue histograms. The
IVT differences refer to the differences in the average IVT between the events associated with the green
(i.e., RCP8.5) and pink (i.e., historical) histograms. These difference maps indicate AR frequency increases by
~50% globally, ~60% in the Southern Ocean, and ~50% in the Northern Hemisphere (Figure 2f). Moreover,
in these same regions, the mean magnitude (in terms of IVT of the strongest of events) increases by 25%
globally and by 20% in the northern and southern mid-and-high latitudes for RCP8.5 (Figure 2f). There are
Figure 3. (a) Multimodel mean (MMM) of the individual model biases in AR frequency (percent of time steps) relative to ERA-Interim. (b) Intermodel standard devia-
tion (STD) of the individual model biases around the MMM bias in AR frequency. (c) Zonally averaged AR frequency for individual models (grey), the MMM with
standard error of mean (blue), and ERA-Interim (red). (df) As (a)(c) but for AR zonal IVT (kg · m
1
·s
1
). (gi) As (a)(c) but for AR meridional IVT (kg · m
1
·s
1
). The
errors bars in (c), (f), and (i) represent the standard errors of the MMM.
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4304
also considerable increases in eastward (westward) tropical AR IVT in the eastern Pacic (Indian) Oceans
although the frequency of ARs is very low in the tropics. Comparing these results with earlier studies on
western North America and western Europe (Table 1) shows that these increases are similar, albeit more
moderate in a number of cases, to previous studies. It is noted that the different detection and analysis
methods used across these earlier studies make it difcult to compare them, and the uniform approach
used here not only highlights areas not previously considered but also allows more judicious comparison
across different regions.
Figures 1b to 1d highlight additional aggregate measures of projected changes in future AR characteristics.
Figure 1b shows model ensemble average histograms of the AR IVT values from the historical and RCP8.5
simulations. Consistent with the increases in AR frequency in Figure 2f is the overall increase in the number
of values across the range of IVT, and importantly the approximate doubling of extreme AR IVT events. The
model ensemble average histograms of AR lengths and AR widths in Figures 1c and 1d illustrate that these
increases in the frequency of AR conditions in Figures 1b and 2f arise from an increase of ~25% in both
the lengths and widths of future ARs, despite an ~10% reduction in the number of ARs in the future under
the RCP8.5 scenario (see Figures 1c and 1d legend).
3.3. Intermodel Differences in Historical Simulations
As a means to consider model delity and projection uncertainty, we illustrate and discuss intermodel differ-
ences against the observational reference and among the projections. Figure 3 summarizes comparisons
between the historical simulations and the ERA-Interim observed reference values for AR frequency, zonal
IVT, and meridional IVT (rows 1, 2, and 3, respectively). In each row, the left (middle) map shows the MMM
(standard deviation) of the individual model biases. The right map shows the zonal averages of AR frequency
or IVT for each individual model, the MMM, and the ERA-Interim reanalysis. For AR frequency, the gure
shows little systematic bias across the model ensemble (Figure 3a), although rather signicant disagreement
in the subtropics are found (Figures 3b and 3c) as in Payne and Magnusdottir (2015) for ARs in the northeast-
ern Pacic in CMIP5. The intermodel differences suggest considerable model uncertainty in some areas heav-
ily impacted by ARs (e.g., California, Chile), although natural variability is an important factor to consider when
interpreting these intermodel differences because the analysis period (i.e., 24 years) may not be long enough
to average out decadal to interannual natural variability. For the zonal IVT, most of the areas of considerable
systematic bias and larger standard deviation about the bias occur in the tropics (Figures 3e and 3f), where
the frequency of ARs is considerably smaller and thus will not be elaborated on more here. For the meridional
IVT, there is a clear systematic bias in the simulations; namely, the poleward AR transports are too weak by
~25% (Figures 3g and 3i).
3.4. Intermodel Differences in Future Projections
Figure 4 is similar in layout to Figure 3, although in this case it reects the difference between the RCP8.5 and
historical simulations (i.e., projected changes), as opposed to the difference between the historical simulation
and reanalysis (i.e., biases); thus, it shows elements of model agreement across their projected changes in
ARs. Figure 4a represents similar information as Figure 2f (i.e., AR frequency mean change) with a different
color scale, with Figures 4b and 4c indicating that the greatest intermodel difference in the projected AR
frequency changes is in the midlatitudes (~ ±2% and ±4% for Northern and Southern Hemispheres, respec-
tively) where the highest changes in AR frequency occur (~5% and 10% for Northern and Southern
Hemispheres, respectively). In these regions, the magnitude of the intermodel differences is roughly 40%
of the multimodel ensemble mean. Based on the Community Earth System Model large ensemble, Hagos
et al. (2016) estimated natural variability, represented by one standard deviation of the individual ensemble
members, to contribute 23% uncertainty in the multimember ensemble mean in projected changes in AR fre-
quency over the western North America from years 19801999 to 20802099. The much larger model uncer-
tainty shown here (i.e., 40%), represented by one standard deviation across the different models, suggests
that natural variability may not fully explain the intermodel differences in projected changes in AR frequency.
Figures 4d and 4f show that overall the models uniformly project stronger eastward zonal AR IVT in the mid-
latitudes, with increases of about 20% in the southern and northern midlatitudes (i.e., compare Figures 3f and
4f). There is a modest weakening of zonal AR IVT in the tropics, particularly the westward transports in the
eastern Pacic Oceanalthough it is worth noting that this region has very low AR frequency and large inter-
model spread. Figures 4g and 4i indicate that the meridional AR export of moisture out of the tropics slightly
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4305
weakens but strengthens considerably in the midlatitudes. Specically, the projected changes in the mean
zonal and meridional AR IVT are about 60 and 10 kg · m
1
·s
1
, respectively, in the northern midlatitudes
and 100 and 25 kg · m
1
·s
1
, respectively, in southern midlatitudes. These results indicate that there is
about a 30% projected increase in zonal AR IVT and 510% in meridional AR IVT for the RCP8.5 scenario.
While Figures 4e and 4h show pronounced intermodel variability in the tropics, it is important to note
again that this is a region with very low AR frequency (Figure 3c).
More precise quantitative comparisons of model delity and projection uncertainties across the models for a
number of regions impacted by ARs are given in Figures S2 and S3. These include California, the U.S. east
coast, the UK, and southwestern regions of Africa, Australia, and Chile (see green boxes on Figures 2a and
2e). Notable is the degree that California, the UK, and Chile stand out among these regions in the uncertainty
of simulated AR frequency for the historical period and projected changes in AR frequency and IVT.
4. Conclusions
This study represents the rst global examination of the climate change impacts on ARs associated with
future warming scenarios, based on the application of an AR global detection algorithm (Guan & Waliser,
2015) to outputs from 21 CMIP5 models. AR detection for the historical simulation and future projections
(RCP4.5 and RCP8.5) are all based on the IVT 85th percentile derived from the historical simulation.
Comparisons between the observational reference and multimodel historical values of AR frequency and
IVT show that the MMM represent the reference patterns reasonably well, especially the variation with lati-
tude and dominant zonal asymmetries (Figures 2a, 2b, 3, and S1). The results from the analysis of the projec-
tions indicate that for the most part AR frequency and IVT values will increase globally. More specically, for
Figure 4. As Figure 3 but for the changes in AR frequency (percent of time steps) and IVT (kg · m
1
·s
1
) between historical and RCP8.5 simulations.
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4306
the RCP8.5 warming scenario, AR frequency increases by ~50% globally, ~60% in the Southern Ocean, and
~50% in the Northern Hemisphere (Figure 2f). Moreover, in these same regions, the mean magnitude (in
terms of IVT of the strongest of events) increases by 25% globally and by 20% in the northern and southern
mid-and-high latitudes for RCP8.5 (Figure 2f). Notable are the results that for the RCP8.5 scenario, the number
of ARs are projected to decrease slightly by ~10%, yet the ARs will be ~25% longer and ~25% wider, leading
to an overall increase in the frequency of AR conditions (i.e., Figure 2f), and exhibit more extreme IVT values
(Figure 1b). Moreover, examination of the intermodel difference in projected changes suggests agreement
among the models on the general increase in AR frequency globally, and the stronger enhancements to
AR frequency and IVT in the midlatitudes, particularly the Southern Ocean (Figures 4, S1, S2, and S3).
Previous investigations focused on western North America and Europe, along with a related study on IVT
in general (Lavers et al., 2016), have illustrated that thermodynamic response (i.e., moistening) of the atmo-
sphere to the warming dominates, and dynamical effects (e.g., increases in wind speeds) are small. The results
reported here for climate change impacts on AR frequency and IVT are generally consistent with previous stu-
dies that mainly have focused on western North America and western Europe (a number of which are shown
in Table 1). A virtue of the present study is not only highlighting the climate change impacts on ARs beyond
these two regions but also providing a means to more soundly compare, given the uniform global data sets
and methodology, the projected changes, for example, between two given regions (e.g., western North
America and western Europe).
Apart from the general agreement in AR frequency and IVT patterns and values between the multimodel his-
torical simulations and observation reference, there are considerable intermodel variations in terms of AR
representation and projected changes that suggest caution in terms of the projected changes to ARs.
These warrant additional focused efforts on model evaluation and improvement for AR characteristics.
Also notable is that there are a couple of regions that exhibit differences in the sign of the AR frequency
change, with a few models projecting decreases in AR frequency in the subtropical Pacic regions near
North/South America and the western Pacic (Figures 3, 4, S1, and S2). These are expected to be due to shifts
in storm track or subtropical jet features, such as those diagnosed by Hagos et al. (2016), Shields and Kiehl
(2016a), and Gao et al. (2016) in their studies for western North America and western Europe sectors,
although more in-depth study in these regions is warranted. Overall, the results suggest fairly robust agree-
ment at global scales across the models for the climate change impacts on AR frequency and IVT, relatively
good agreement in terms of latitudinal dependencies, and relatively poor agreement at regional scales
(Figures 3, 4, S2, and S3).
The intermodel differences in projected AR changes, particularly at regional scales, may not be fully explained
by natural variability. Better constraining these models in terms of their AR projections is needed given the
large societal impacts of these storms. Further works on eld experiments, process studies, and model eva-
luation and improvement (e.g., Guan & Waliser, 2017; Hagos et al., 2015; Payne & Magnusdottir, 2015;
Ralph et al., 2016; Wick et al., 2013) need to be undertaken to improve the model delity and reduce the
uncertainty in the projections.
References
Dettinger, M. (2011). Climate change, atmospheric rivers, and oods in CaliforniaA multimodel analysis of storm frequency and
magnitude changes. Journal of the American Water Resources Association,47(3), 514523. https://doi.o rg/10.1111/
j.1752-1688.2011.00546.x
Dettinger, M. D. (2013). Atmospheric rivers as drought busters on the U.S. We st Coast. Journal of Hydrometeorology,14(6), 17211732. https://
doi.org/10.1175/JHM-D-13-02.1
Dettinger, M. D., Ralph, F. M., Das, T., Neiman, P. J., & Cayan, D. R. (2011). Atmospheric rivers, oods, and the water resources of California.
Water,3(2), 445478. https://doi.org/10.3390/w3020445
Gao, Y., Lu, J., Leung, L. R., Yang, Q., Hagos, S., & Qian, Y. (2015). Dynamical and thermodynamical modulations on future changes of
landfalling atmospheric rivers over western North America. Geophysical Research Letters,42, 71797186. https://doi.org/10.1002/
2015GL065435
Gao, Y., Lu, J., & Leung, R. (2016). Uncertainties in projecting future changes in atmospheric rivers and their impacts on heavy precipitation
over Europe. Journal of Climate,29(18), 67116726. https://doi.org/10.1175/JCLI-D-16-0088.1
Gorodetskaya, I. V., Tsukernik, M., Claes, K., Ralph, M. F., Neff, W. D., & Van Lipzig, N. P. M. (2014). The role of atmospheric rivers in anomalous
snow accumulation in East Antarctica. Geophysical Research Letters,41, 61996206. https://doi.org/10.1002/2014GL060881
Guan, B., Molotch, N. P., Waliser, D. E., Fetzer, E. J., & Neiman, P. J. (2010). Extreme snowfall events linked to atmospheric rivers and surface air
temperature via satellite measurements. Geophysical Research Letters,37, L20401. https://doi.org/10.1029/2010GL044696
Guan, B., & Waliser, D. E. (2015). Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. Journal of
Geophysical Research: Atmospheres,120, 12,51412,535. https://doi.org/10.1002/2015JD024257
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4307
Acknowledgments
We acknowledge the World Climate
Research Programmes Working Group
on Coupled Modeling, which is respon-
sible for CMIP, and we thank the climate
modeling groups for producing and
making available their model output.
For CMIP the U.S. Department of
Energys Program for Climate Model
Diagnosis and Intercomparison pro-
vides coordinating support and lead
development of software infrastructure
in partnership with the Global
Organization for Earth System Science
Portals. The ERA-Interim reanalysis data
set is available via http://apps.ecmwf.
int/datasets/data/interim-full-daily/,
and the CMIP5 model output data are
available via https://cmip.llnl.gov/
cmip5/data_portal.html. This research
was supported by the NASA Energy and
Water cycle Study (NEWS) program.
DEWs contribution to this study was
carried out on behalf of the Jet
Propulsion Laboratory, California
Institute of Technology, under a con-
tract with the National Aeronautics and
Space Administration. Vicky Espinozas
contribution to this study was made
possible by NASA Jet Propulsion
Laboratorys Year-Round Internship
Program during her graduate studies at
the University of Southern California.
Guan, B., & Waliser, D. E. (2017). Atmospheric rivers in 20 year weather and climate simulations: A multimodel, global evaluation. Journal of
Geophysical Research: Atmospheres,122, 55565581. https://doi.org/10.1002/2016JD026174
Hagos, S., Leung, L. R., Yang, Q., Zhao, C., & Lu, J. (2015). Resolution and dynamical core dependence of atmospheric river frequency in global
model simulations. Journal of Climate,28(7), 27642776. https://doi.org/10.1175/JCLI-D-14-00567.1
Hagos, S. M., Leung, L. R., Yoon, J.-H., Lu, J., & Gao, Y. (2016). A projection of changes in landfalling atmospheric river frequency and extreme
precipitation over western North America from the large ensemble CESM simulations. Geophysic al Research Letters,43, 13571363. https://
doi.org/10.1002/2015GL067392
Lavers, D. A., Allan, R. P., Villarini, G., Lloyd-Hughes, B., Brayshaw, D. J., & Wade, A. J. (2013). Future changes in atmospheric rivers and their
implications for winter ooding in Britain. Environmental Research Letters,8(3), 034010. https://doi.org/10.1088/1748-9326/8/3/034010
Lavers, D. A., Allan, R. P., Wood, E. F., Villarini, G., Brayshaw, D. J., & Wade, A. J. (2011). Winter oods in Britain are connected to atmospheric
rivers. Geophysical Research Letters,38, L23803. https://doi.org/10.1029/2011GL049783
Lavers, D. A., Ralph, F. M., Waliser, D. E., Gershunov, A., & Dettinger, M. D. (2015). Climate change intensication of horizontal water vapor
transport in CMIP5. Geophysical Research Letters,42, 56175625. https://doi.org/10.1002/2015GL064672
Lavers, D. A., Waliser, D. E., Ralph, F. M., & Dettinger, M. D. (2016). Predictability of horizontal water vapor transport relative to precipitation:
Enhancing situational awareness for forecasting western U.S. extreme precipitation and ooding. Geophysical Research Letters,43,
22752282. https://doi.org/10.1002/2016GL067765
Leung, L. R., & Qian, Y. (2009). Atmospheric rivers induced heavy precipitation and ooding in the western U.S. simulated by the WRF regional
climate model. Geophysical Research Letters,36, L03820. https://doi.org/10.1029/2008GL036445
Neiman, P. J., Ralph, F. M., Wick, G. A., Lundquist, J. D., & Dettinger, M. D. (2008). Meteorological characteristics and overland precipitation
impacts of atmospheric rivers affecting the west coast of North America based on eight years of SSM/I satellite observations. Journal of
Hydrometeorology,9(1), 2247. https://doi.org/10.1175/2007jhm855.1
Neiman, P. J., Schick, L. J., Ralph, F. M., Hughes, M., & Wick, G. A. (2011). Flooding in western Washington: The connection to atmospheric
rivers. Journal of Hydrometeorology,12(6), 13371358. https://doi.org/10.1175/2011JHM1358.1
Neiman, P. J., White, A. B., Ralph, F. M., Gottas, D. J., & Gutman, S. I. (2009). A water vapor ux tool for precipitation forecasting.
Water Management,162(2), 8394. https://doi.org/10.1680/wama.2009.162.2.83
Payne, A. E., & Magnusdottir, G. (2015). An evaluation of atmospheric rivers over the North Pacic in CMIP5 and their response to warming
under RCP 8.5. Journal of Geophysical Research: Atmospheres,120, 11,17311,190. https://doi.org/10. 1002/2015JD023586
Pierce, D. W., Cayan, D. R., Das, T., Maurer, E. P., Miller, N. L., Bao, Y., et al. (2013). The key role of heavy precipitation events in climate model
disagreements of future annual precipitation changes in California. Journal of Climate,26(16), 58795896. https://doi.org/10.1175/
JCLI-D-12-00766.1
Radić, V., Cannon, A. J., Menounos, B., & Gi, N. (2015). Future changes in autumn atmospheric river events in British Columbia, Canada, as
projected by CMIP5 global climate models. Journal of Geophysical Research: Atmospheres,120, 92799302. https://doi.org/10.100 2/
2015JD023279
Ralph, F. M., & Dettinger, M. D. (2011). Storms, oods, and the science of atmospheric rivers. Eos, Transactions American Geophysical Union,
92(32), 265. https://doi.org/10.1029/2011EO320001
Ralph, F. M., Neiman, P. J., & Wick, G. A. (2004). Satellite and CALJET aircraft observations of atmospheric rivers over the eastern North Pacic
Ocean during the winter of 1997/98. Monthly Weather Review,132(7), 17211745. https://doi.org/10.1175/1520-0493(2004)132%3C172 1:
SACAOO%3E2.0.CO;2
Ralph, F. M., Neiman, P. J., Wick, G. A., Gutman, S. I., Dettinger, M. D., Cayan, D. R., & White, A. B. (2006). Flooding on Californias Russian River:
Role of atmospheric rivers. Geophysical Research Letters,33, L13801. https://doi.org/10.1029/2006GL026689
Ralph, F. M., Neiman, P. J., Zamora, R. J., & Dettinger, M. D. (2013). Observed impacts of duration and season ality of atmospheric-river landfalls
on soul moisture and runoff in coastal in coastal Northern California. Journal of Hydrometeorology,14(2), 443459. https://doi.org/10.1175/
JHM-D-12-076.1
Ralph, F. M., Prather, K. A., Cayan, D., Spackman, J. R., Demott, P., Dettinger, M., & Intrieri, J. (2016). CalWater eld studies designed to quantify
the roles of atmospheric rivers and aerosols in modulating U.S. West Coast precipitation in a changing climate. Bulletin of the American
Meteorological Society,97(7), 12091228. https://doi.org/10.1175/bams-d-14-00043.1
Ramos, A. M., Tomé, R., Trigo, R. M., Liberato, M. L. R., & Pinto, J. G. (2016). Projected changes in atmospheric rivers affecting Europe in CMIP5
models. Geophysical Research Letters,43, 93159323. https://doi.org/10.1002/2016GL070634
Shields, C. A., & Kiehl, J. T. (2016a). Atmospheric river landfall-latitude changes in future climate simulations. Geophysical Research Letters,43,
87758782. https://doi.org/10.1002/2016GL070470
Shields, C. A., & Kiehl, J. T. (2016b). Simulating the pineapple express in the half degree Community Climate System Model, CCSM4.
Geophysical Research Letters,43, 77677773. https://doi.org/10.1002/2016GL069476
Taylor, K. E., Stouffer, R. J., & Meehl, G. A. (2012). An overview of CMIP5 and the experiment design. Bulletin of the American Meteorological
Society,93(4), 485498. https://doi.org/10.1175/BAMS-D-11-00094.1
Waliser, D. E., & Guan, B. (2017). Extreme winds and precipitation during landfall of atmospheric rivers. Nature Geoscience,10(3), 179183.
https://doi.org/10.1038/NGEO2894
Waliser, D. E., Moncrieff, M., Burrridge, D., Fink, A., Gochis, D., Goswami, B. N., et al. (2012). The Yearof tropical convection (May 2008 to April
2010): Climate variability and weather highlights. Bulletin of the American Meteorological Society,93(8), 11891218. https://doi.org/
10.1175/2011BAMS3095.1
Warner, M., Mass, C. F., & Salathé, E. P. (2015). Changes in winter atmospheric rivers along the North American west coast in CMIP5 climate
models. Journal of Hydrometeorology,16(1), 118128. https://doi.org/10.1175/JHM-D-14-0080.1
Wick, G. A., Neiman, P. J., Ralph, F. M., & Hamill, T. M. (2013). Evaluation of forecasts of the water vapor signature of atmospheric rivers in
operational numerical weather prediction models. Weather and Forecasting,28(6), 13371352. https://doi.org/10.11 75/WAF-D-13-00025.1
Zhu, Y., & Newell, R. E. (1998). A proposed algorithm for moisture uxes from atmospheric rivers. Monthly Weather Review,126(3), 725735.
https://doi.org/10.1175/1520-0493
10.1029/2017GL076968
Geophysical Research Letters
ESPINOZA ET AL. 4308
... Since the MENA region lies between these two areas, there is significant uncertainty about the sign and magnitude of future changes to precipitation in much of the region. However, since atmospheric water vapor will increase with increasing temperatures, confidence is high that precipitation extremes will increase in frequency and intensity in the future throughout the MENA region and the rest of the globe [11,12]. This study investigates the impact of a warmer climate on mean and extreme precipitation in the MENA region where the impact of ARs is starting to get recognized. ...
... Changes in precipitation in a warmer climate are governed by many factors. A primary physical mechanism for increases in precipitation is the enhanced water vapor content in a warmer atmosphere, similar analyses on global scales [11,12], this is the first study to investigate future changes in ARs and precipitation for the MENA region using the CMIP multi-model ensemble. ...
... The detection algorithm was applied to both reanalysis data and the CMIP5 models in their native resolution. For the CMIP5 models, once the AR dynamics are estimated, results are re-gridded using a bi-linear function and ensuring that sharp gradients were not a problem in the interpolating scheme, as was done in Espinoza et al. (2018) [11] and Massoud et al. (2019) [12]. This global algorithm had over 90% agreement in detected AR landfall dates with regional algorithms that focused on regions with different climatologies, such as western North America [63], Britain [28], and East Antarctica [64], which provides us with confidence in using this algorithm for our MENA region application. ...
Chapter
This study explores historical mean climate and future projected change of atmospheric rivers (ARs) and precipitation in the Middle East and North Africa (MENA) region. A suite of models from the Coupled Model Intercomparison Project Phase 5 (CMIP5, historical and RCP8.5 scenarios) and other observations are used to estimate AR frequency and mean daily precipitation. Despite the arid-to-semi-arid climate in MENA, parts of this region experience frequent and intense ARs, which largely contribute to the total annual precipitation, such as in the mountainous areas of Turkey and Iran. By the end of this century, this study reports that AR frequency is projected to increase by ~20–40% for the North Africa and Mediterranean regions (including any areas with latitudes 35 N and higher). However, in these regions, mean daily precipitation (i.e., regardless of the presence of ARs) is projected to decrease by ~15–30%. For other regions within MENA, such as the Arabian Peninsula and the Horn of Africa, minor changes in AR frequency are expected (±10%), yet mean precipitation for these regions is projected to increase (~50%). Generally, the sign of change in projected AR frequency is opposite to the sign of change in projected mean daily precipitation for most areas within the MENA region.KeywordsClimate changePrecipitationAtmospheric riversMiddle EastNorth Africa
... Since the MENA region lies between these two areas, there is a large uncertainty about the sign and magnitude of future changes in precipitation in much of the region. Yet, since atmospheric water vapor will increase with increasing temperatures, there is confidence that precipitation extremes will increase in frequency and intensity in the future throughout MENA and the rest of the globe (Espinoza et al. 2018;). ...
... The importance of AR storms has motivated an increasing number of studies related to climate change impacts on ARs (as reviewed in (Payne et al. 2020), with many studies focusing on Western North America Pierce et al. 2013;Payne et al. 2015;Warner et al. 2015;Gao et al. 2015;Radić et al. 2015;Hagos et al. 2016;Shields et al. 2016b;Gershunov et al. 2019;Huang et al. 2020) and Europe (Lavers et al. 2013;Gao et al. 2016;Ramos et al. 2016;Shields et al. 2016a). Espinoza et al. (2018) and provided a global view of AR frequency and strength in future climates and reported that increases in both frequency and strength of ARs is expected to occur for many regions of the globe by roughly 25-50%. ...
... Finally, we use the CMIP5 set of models to investigate historical and future changes in ARs and precipitation for the MENA region. Although studies have performed similar analyses on global scales (Espinoza et al. 2018;, this is one of the first to explore future changes in ARs and precipitation for the MENA region using the CMIP multi-model ensemble (c.f. ). ...
Chapter
In this paper, a new hybrid DRASTIC-based fuzzy C-means (FCM) clustering technique is utilized to find the real relation among the affecting parameters of each hydrogeological point resulting in vulnerability and the fuzzy membership degree of each point to the “most-vulnerable class”. This procedure can be done instead of holding a summation through all affecting parameters to form vulnerability index as implemented in the ordinary DRASTIC method. In DRASTIC, any changes in one point’s parameter value may cause that point to move to another vulnerability class or points which have obviously different parameter values may belong to the same vulnerability class. While in fuzzy logic, each point partly belongs to each vulnerability class and does not necessarily belong to a specific one. This is the main motivation to use FCM clustering technique. In this paper, the vulnerability map of Damaneh-Daran aquifer, located in Isfahan province in central Iran, is prepared using DRASTIC and hybridizing DRASTIC and FCM. The analytical-experimental investigations reveal the weighting power of 1.75 is the best value among 1.25, 1.5, 1.75 and 2. In this weighting power, there are approximately 51%, 21% and 1% decreases in the area percentages covered by low, medium and high vulnerability clusters, respectively, while the area percentages covered by very low and very high clusters increases 8 and 5 times than those of the ordinary DRASTIC, respectively, mainly due to partial membership of the hydrogeological points in the fuzzy clusters, making the areas covered much more evenly distributed among different vulnerability classes. To validate the proposed model, the final vulnerability indices were compared with the nitrate concentration of the aquifer assuming four fuzzy intensity levels. The results indicate the FCM-DRASTIC-based vulnerability zoning have more correlation with the nitrate concentration zoning of the aquifer than the ordinary DRASTIC model.KeywordsClusteringDRASTICFuzzy C-meansGroundwaterVulnerability
... As a result, a side effect of SAI exacerbating the increases 355 in high-latitude AR activity and associated mean and extreme precipitation is observed, particularly over northeastern China, the Korean Peninsula and Japan. The presented future changes in AR activity over East Asia, particularly the increase in AR frequency over southern China, agree well with previous AR projection studies (Espinoza et al., 2018;Kamae et al., 2021). The presented increases in AR 360 length and size under the warming scenarios have also been found globally (Espinoza et al., 2018;Zhao, 2020). ...
... The presented future changes in AR activity over East Asia, particularly the increase in AR frequency over southern China, agree well with previous AR projection studies (Espinoza et al., 2018;Kamae et al., 2021). The presented increases in AR 360 length and size under the warming scenarios have also been found globally (Espinoza et al., 2018;Zhao, 2020). These similarities imply additional confidence in the reported future AR projection in this study. ...
Preprint
Full-text available
Atmospheric rivers (ARs) are closely associated with historical extreme precipitation events over East Asia. The projected increase in such weather systems under global warming has been extensively discussed in previous studies, while the role of stratospheric aerosol, particularly for the implementation of stratospheric aerosol intervention (SAI), in such a change remains unknown. Based on an ensemble of the UK Earth System Model (UKESM1) simulations, here we investigate changes in the frequency of ARs and their associated mean and extreme precipitation under a range of climate forcing, including greenhouse gas emission scenarios of high (SSP5-8.5) and medium (SSP2-4.5) levels, the deployment of SAI geoengineering (G6sulfur) and solar dimming (G6solar). The result indicates a significant increase in AR frequency and AR-related precipitation over most of East Asia in a warmer climate and the most pronounced changes are observed in southern China. Comparing to G6solar and both the SSPs scenarios, the G6sulfur simulations indicate that SAI is effective in partly ameliorating the increases in AR activity over the subtropical region; however, it may result in more pronounced increases in ARs and associated precipitation over the mid-high latitude regions, particularly northeastern China and Japan. Such a response is associated with the further weakening of the mid-latitude westerly jet stream under SAI that favours the high-latitude AR activity. This is driven by the decreased meridional gradient of thermal expansion in the mid-high troposphere associated with aerosol cooling across the tropical region, though SAI effectively ameliorates the widespread increase in thermal expansion under climate warming. Such a side effect of SAI over the populated region implies that caution must be taken when considering geoengineering approaches to mitigating hydrological risk under climate change.
... In many tropical regions, moisture appears to be transported in a sort of 'atmospheric river', traveling across borders, channeled narrowly in the vertical, but spread out widely in the horizontal, compared to atmospheric rivers in the extratropics, which are narrow in the horizontal and often associated with extreme precipitation (e.g., Gimeno et al., 2014;Rutz et al., 2019). There are many open questions regarding these expansive moisture transports in the tropics and how they differ from atmospheric rivers in the midlatitudes, which are better characterized (e.g., Espinoza et al., 2018). The sources of water, for instance, remain unclear. ...
Thesis
The response of trade-wind clouds to warming remains uncertain, raising the specter of a large climate sensitivity. Decreases in cloud fraction are thought to relate to interplay among convective mixing, turbulence, radiation, and the large-scale environment. The EUREC4A (Elucidating the role of cloud-circulation coupling in climate) field campaign made extensive measurements that allow for deeper physical understanding and the first process-based constraint on the trade cumulus feedback.I first use EUREC4A observations to improve understanding of the characteristic vertical structure of the trade-wind boundary layer and the processes that produce this structure. This improved physical understanding is then applied to the evaluation of trade cumulus feedbacks. Ideas developed support new conceptual models of the structure of the trade-wind boundary layer and a more active role of clouds in maintaining this structure, and show little evidence for a strong trade cumulus feedback to warming.
... As a large fraction of the precipitation over coastal WA comes from EPEs, the future mass balance changes over WA will be highly sensitive to the future changes in the atmospheric drivers of EPEs identified in this study. A westward (and poleward) shift of ASL (J. S. Hosking et al., 2016;Brown et al., 2020) along with a likely increase in the frequency of atmospheric rivers land falling on Antarctica by the end of the 21st century (Espinoza et al., 2018;Payne et al., 2021) are expected to make a positive contribution to future mass balance changes over WA. However, the projected change in ENSO, and hence the associated teleconnection, remain highly uncertain. ...
Article
Full-text available
We investigate the atmospheric drivers of extreme precipitation over the Amundsen Sea Embayment (ASE) of West Antarctica (WA) using daily output from the RACMO2 model and reanalysis data (1979–2016). Overall, 93.7% of days with extreme precipitation at the two coastal stations of ASE are associated with the four dominant Empirical Orthogonal Function (EOF) modes of geopotential height anomalies (at 850 hPa) over WA. The second EOF mode, associated with a coupled pattern consisting of an Amundsen Sea Low and a blocking high to the east, is the main driver of extreme precipitation over ASE, linked to 44.75% of extreme precipitation days. This is followed by EOF‐3 (associated with El Niño Southern Oscillation/PSA‐1), EOF‐4 (likely associated with more frequent “atmospheric river” events), and EOF‐1 (i.e., Southern Annular mode) with a contribution of 22.16%, 21.1%, and 12%, respectively. Extreme precipitation linked to EOF‐2 and EOF‐4 is more intense (by ∼2 mm/day) than the rest.
... Recent studies point to the more significant role of ARs in future climates (e.g., Espinoza et al., 2018;Payne et al., 2020). Considering the Clausius-Clapeyron scale, global warming is likely to bring about a wetter atmosphere and therefore a greater availability of water vapor in near-saturation transport events, inducing a greater advection of moisture by ARs. ...
Article
Full-text available
The effect of increased populations concentrated in urban areas, coupled with the ongoing threat of climate change, means that society is becoming increasingly vulnerable to the effects of extreme precipitation. The study of these events is therefore a key topic in climate research, in their physical basis, in the study of their impacts, and in our adaptation to them. From a meteorological perspective, the main questions are related to the definition of extreme events, changes in their distribution and intensity both globally and regionally, the dependence on large‐scale phenomena including the role of moisture transport, and changes in their behavior due to anthropogenic pressures. In this review article, we address all these points and propose a set of challenges for future research. This article is categorized under: Science of Water > Water Extremes Science of Water > Hydrological Processes Main oceanic and terrestrial moisture sources and their area of higher moisture contribution are associated with extreme precipitation events. The rounded areas represent the regions where the source of higher contribution changed compared with climatological mean precipitation. The sources defined are North and South Atlantic Ocean (NATL and SATL), North and South Pacific Ocean (NPAC and SPAC), Mediterranean and Red Seas (MED and REDS), Gulf of Mexico (MEXCAR), Indian Ocean and Zanzibar Current and Arabian Sea (IND and ZANAR), Agulhas Current (AGU), South America (SAM), Sahel Region (SAHEL), and South Africa (SAFR).
... In particular, a positive role has also been demonstrated in that they may act as drought busters for example on the US West Coast (Dettinger, 2013). Additionally, there is a large consensus that the importance of ARs will 40 increase for all the roles they play in the hydrological cycle in the near future, in the context of the atmosphere becoming hotter and wetter under a changing climate than the current one (e.g., Espinoza et al., 2018;Massoud et al., 2019;Lavers et al., 2015;Ramos et al., 2016b). Figure 1a shows an example of a well-defined AR, observable in the field of integrated water vapour (IWV, eq.1 where q is the specific humidity and Ω denotes a vertical integration over the whole tropospheric column) reaching the Irish shore. ...
Article
This study makes use of the new Total Column Vater Vapour Data Record (CDR‐2 (v2)) —developed by the European Space Agency (ESA) in coordination with the Satellite Application Facility on Climate Monitoring (CM SAF)— to analyse the adequacy of the integrated vertical water vapour column (IWV) data provided by the European Centre for Medium‐Range Weather Forecasts (ECMWF) ERA5 and ERA‐Interim reanalyses in regions of critical interest for moisture transport mechanisms. This information is critical for the initialization of moisture transport models —both Eulerian and Lagrangian— used to study the main mechanisms and predict the future evolution of moisture transport events. Particularly, almost 40000 atmospheric river (AR) and nocturnal low‐level jet (NLLJ) events identified on a global scale between 2002 and 2017, have been used to study the variability between the cited reanalyses and the CDR‐2, both in terms of bias in the observed values of IWV during each particular event, and in terms of daily temporal correlation fields. Although some notable discrepancies are reported in the main tropical rainforest regions, it is observed that in regions of high interest for both ARs and NLLJs, the degree of agreement between the reanalyses and CDR‐2 is high. The bias observed in the regions of interest is generally low, and the temporal correlation in the IWV fields is above 0.8 in most areas. ERA5 appears to show slightly better performance than ERA‐Interim when resolving the moisture column, and both show greater similarity to CDR‐2 in mid‐latitudes compared to tropical regions. The probability density functions constructed on a event‐to‐event basis reinforce these ideas. We conclude that the here presented evaluations using CDR‐2 serve to strengthen avaliable evidence that the ECMWF reanalyses can be safely used in the initializations of Lagrangian dispersion models and Eulerian moisture tracer simulations —commonly used for the analysis of main advection mechanisms— in the vast majority of regions critical to the study of ARs and LLJs. They can be also safely used for the detection of moisture source‐sink regions in the study of the global hydrological cycle in these regions.
Article
Full-text available
Atmospheric rivers (ARs) are intensive poleward moisture transport events that are essential to the global hydrological cycle and are often linked to extreme weather events. We categorize the winter North Pacific ARs into two “flavors”: wind‐dominated (windy ARs) and moisture‐dominated (wet ARs) using 40 years of hourly data from fifth generation of the European Centre for Medium‐Range Weather Forecasts Interim Reanalysis. We compare the differences between windy ARs and wet ARs including the lifecycle characteristics (such as genesis locations and changes of meteorological elements through the lifecycle), overall AR frequency, landfall impacts, and variability. The windy ARs are more likely to occur in the midlatitudes, while wet ARs are more active in the subtropics. Windy ARs are associated with intensive surface pressure lows, where the strong pressure gradient can support the strong wind within ARs. Due to larger size and longer lifetime, wet ARs are more likely to produce more precipitation over a lifecycle. By scaling the landfalling ARs, we show that wet ARs dominate the high‐category ARs (Category 4 and 5) with higher spatial frequency and more precipitation, and windy ARs have higher contributions in the lower AR categories especially over British Columbia. Windy ARs are modulated by El Niño Southern Oscillation (ENSO) teleconnections via the anomalous geopotential height and extended subtropical jet. Wet ARs are affected by the anomalous sea surface temperature over the midlatitudes related to ENSO. Sensitivity analysis with an alternate AR detection algorithm shows consistent results on AR flavors but with disagreement on the amplitude.
Article
Full-text available
Tasman Glacier, a temperate maritime glacier in the New Zealand Southern Alps, is rapidly receding. Climate warming is resulting in lengthening of the ablation season, meaning crevasses in the accumulation area are becoming exposed at the surface for longer. We combine measurements of air temperature and wind speed from inside crevasses with surface meteorological data, finding that during summer, in-crevasse air temperature is frequently positive, and can at times exceed surface air temperature. Greatest warming occurred in the widest crevasses during clear-sky conditions, but full depth warming of crevasses also occurred at night. Net shortwave radiation contributes to heating of air in the upper regions of crevasses, but turbulent sensible heat transfer was responsible for driving warm air deeper into crevasses. Crevasses orientated to maximise radiation retrieval, and running parallel to wind flow, have the greatest potential for warming and heat storage. We hypothesise a positive feedback loop in the surface energy-balance system, where crevasses entrain and trap heat, which enhances melting, that in turn enlarges the crevasses, enabling greater heat storage and further melting. Energy-balance models that treat accumulation areas of alpine glaciers as homogeneous surfaces will therefore underestimate snow melt and overestimate mass balance.
Article
Full-text available
Atmospheric rivers (ARs) generate most of the economic losses associated with flooding in the western United States and are projected to increase in intensity with climate change. This is of concern as flood damages have been shown to increase exponentially with AR intensity. To assess how AR-related flood damages are likely to respond to climate change, we constructed county-level damage models for the western 11 conterminous states using 40 years of flood insurance data linked to characteristics of ARs at landfall. Damage functions were applied to 14 CMIP5 global climate models under the RCP4.5 “intermediate emissions” and RCP8.5 “high emissions” scenarios, under the assumption that spatial patterns of exposure, vulnerability, and flood protection remain constant at present day levels. The models predict that annual expected AR-related flood damages in the western United States could increase from $1 billion in the historical period to $2.3 billion in the 2090s under the RCP4.5 scenario or to $3.2 billion under the RCP8.5 scenario. County-level projections were developed to identify counties at greatest risk, allowing policymakers to target efforts to increase resilience to climate change.
Article
Full-text available
Landfalling atmospheric rivers over the North Pacific are evaluated in historical (1980-2005) simulations from 28 models participating in the fifth phase of the Coupled Model Intercomparison Project (CMIP5) and compared to the Modern-Era Retrospective Analysis for Research and Applications and ERA-Interim reanalyses. Landfalling dates are identified as having spatially elongated high wind and moisture features in physical proximity to the coastline. Model performance relative to reanalysis is found to be quite variable. The majority can resolve the spatial structure of landfalling events, but few correctly resolve the frequency distribution, interannual variability in number and amplitude of moisture flux, and median landfalling latitude. The response of a subset of high-performing models to projected warming at the end of the 21st century is investigated using Representative Concentration Pathway (RCP) 8.5 (2070-2100) projections. Selected models show a broadening of the frequency distribution, with the largest increase in frequency occurring equatorward of peak historical frequency. While there is a robust increase in AR-related moisture transport compared to the historical period, this increase is not linked to more extreme AR events within the context of the projected future climate. The occurrence of years with many AR dates is greater at the end of the 21st century than during the historical period. The equatorward increase in peak historical frequency is colocated with increases in the 850 and 250 hPa zonal winds. The response in moisture flux to warming is mostly thermodynamic, but equatorward of its peak distribution, it is dominated by a dynamic response.
Article
Full-text available
This study investigates the North Atlantic atmospheric rivers (ARs) making landfall over western Europe in the present and future climate from the multimodel ensemble of phase 5 of the Coupled Model Intercomparison Project (CMIP5). Overall, CMIP5 captures the seasonal and spatial variations of historical landfalling AR days, with the large intermodel variability strongly correlated with the intermodel spread of historical near-surface westerly jet position. Under representative concentration pathway 8.5 (RCP8.5), AR frequency is projected to increase significantly by the end of this century, with 127%–275% increase at peak AR frequency regions (458–558N). While thermodynamics plays a dominant role in the future increase of ARs, wind changes associated with the midlatitude jet shifts also significantly contribute to AR changes, resulting in dipole change patterns in all seasons. In the North Atlantic, the model-projected jet shifts are strongly correlated with the simulated historical jet position. As models exhibit predominantly equatorward biases in the historical jet position, the large poleward jet shifts reduce AR days south of the historical mean jet position through the dynamical connections between the jet positions and AR days. Using the observed historical jet position as an emergent constraint, dynamical effects further increase future AR days over the equatorward flank above the increases from thermodynamical effects. Compared to the present, both total and extreme precipitation induced by ARs in the future contribute more to the seasonal mean and extreme precipitation, primarily because of the increase in AR frequency.
Article
Full-text available
This study examines future changes of landfalling atmospheric rivers (ARs) over western North America using outputs from the Coupled Model Intercomparison Project Phase 5 (CMIP5). The result reveals a strikingly large increase of AR days by the end of the 21st century in the RCP8.5 scenario, with fractional increases between 50% and 600%, depending on the seasons and landfall locations. These increases are predominantly controlled by the super-Clausius-Clapeyron rate of increase of atmospheric water vapor with warming, while changes of winds that transport moisture in the ARs, or dynamical effect, mostly counter the thermodynamical effect of increasing water vapor, limiting the increase of AR events in the future. The consistent negative effect of wind changes on AR days during spring and fall can be linked to the robust poleward shift of the subtropical jet in the North Pacific basin.
Article
Full-text available
Simulations from the Community Earth System Model Large Ensemble project are analyzed to investigate the impact of global warming on atmospheric rivers (ARs) making landfall in western North America. The model has notable biases in simulating the subtropical jet position and the relationship between extreme precipitation and moisture transport. After accounting for these biases, the model projects an ensemble mean increase of 35% in the number of landfalling AR days between the last twenty years of the 20th and 21st centuries under RCP8.5. However, the associated extreme precipitation days increase only by 28% because moisture transport to produce extreme precipitation also increases with warming. Internal variability introduces an uncertainty of ±8% and ±7% in the changes in AR days and associated extreme precipitation days compared to only about 1% difference from accountings for model biases. The significantly larger mean changes compared to internal variability and effects of model biases highlight the robust AR responses to global warming.
Article
Atmospheric rivers (ARs) are narrow, elongated, synoptic jets of water vapor that play important roles in the global water cycle and meteorological/hydrological extremes. Increasing evidence shows ARs have signatures and impacts in many regions across different continents. However, global-scale characterizations of AR representations in weather and climate models have been very limited. Using a recently developed AR detection algorithm oriented for global applications, the representation of AR activities in multi-decade weather/climate simulations is evaluated. The algorithm is applied to 6-hourly (daily) integrated water vapor transport from 22 (2) global weather/climate models that participated in the GASS-YoTC Multi-model Experiment, including four models with ocean-atmosphere coupling and two models with super-parameterization. Multiple reanalysis products are used as references to help quantify model errors in the context of reanalysis uncertainty. Model performance is examined for key aspects of ARs (frequency, intensity, geometry, seasonality), with the focus on identifying and understanding systematic errors in simulated ARs. The results highlight the range of model performances relative to reanalysis uncertainty in representing the most basic features of ARs. Among the 17 metrics considered, AR frequency, the zonal component of the integrated water vapor transport (IVT), fractional zonal circumference, fractional total meridional IVT, and three seasonality metrics have consistently large errors across all models. Possible connections between AR simulation qualities and aspects of model configurations are discussed. Despite the lack of a monotonic relationship, the importance of model horizontal resolution to the overall quality of AR simulation is suggested by the evaluation results.
Article
Atmospheric rivers—long, narrow filaments of large integrated water vapour transport—are associated with weather and water extremes, such as precipitation extremes and flooding in western North America and northern Europe. Here we apply a global detection algorithm for atmospheric rivers to reanalysis data during 1997–2014 to investigate the impact of atmospheric rivers on wind extremes as well as precipitation extremes. We find that atmospheric rivers are associated with up to half of the extreme events in the top 2% of the precipitation and wind distribution, across most mid-latitude regions globally. Landfalling atmospheric rivers are associated with about 40–75% of extreme wind and precipitation events over 40% of the world’s coastlines. Atmospheric rivers are associated with a doubling or more of the typical wind speed compared to all storm conditions, and a 50–100% increase in the wind and precipitation values for extreme events. We also find that the majority of extreme wind events catalogued between 1997 and 2013 over Europe with billion US dollar losses were associated with atmospheric rivers. We conclude that landfalling atmospheric rivers can represent a significant hazard around the globe, because of their association with not only extreme precipitation, but also extreme winds.
Article
Atmospheric Rivers (ARs) are elongated bands of high water vapor concentration extending to the mid-latitudes, which can be associated with intense precipitation and floods over continental areas. We analyze ARs reaching Europe in simulations from six CMIP5 global climate models (CGMs) to quantify possible changes during the current century, with emphasis in five western European prone coastal areas. ARs are represented reasonably well in GCMs for recent climate conditions (1980-2005). Increased vertically integrated horizontal water transport is found for 2074- 2099 (RCP4.5 and RCP8.5) compared to 1980-2005, while the number of ARs is projected to double on average for the same period. These changes are robust between models and are associated with higher air temperatures and thus enhanced atmospheric moisture content, together with higher precipitation associated with extra-tropical cyclones. This suggests an increased risk of intense precipitation and floods along the Atlantic European Coasts from the Iberian Peninsula to Scandinavia.
Article
The latitude of landfall for atmospheric rivers (ARs) is examined in the fully coupled half-degree version of the Community Climate System Model, version 4 (CCSM4) for warm future climate simulations. Two regions are examined: U.S. West Coast/North Pacific ARs, and United Kingdom/North Atlantic ARs. Changes in AR landfall-latitude reflect changes in the atmospheric steering flow. West coast U.S. ARs are projected to push equatorward in response to the subtropical jet climate change. UK AR response is dominated by eddy-driven jets and is seasonally dependent. UK simulated AR response is modest in the winter with the largest relative changes occurring in the seasonal transition months. Precipitation associated with ARs is also projected to increase in intensity under global warming. CCSM4 projects a marked shift to higher rainfall rates for Southern California. Small to modest rainfall rates may increase for all UK latitudes, for the Pacific Northwest, and central and northern California.
Article
Atmospheric rivers are recognized as major contributors to the poleward transport of water vapor. Upon reaching land, these phenomena also play a critical role in extreme precipitation and flooding events. The Pineapple Express (PE) is defined as an atmospheric river extending out of the deep tropics and reaching the west coast of North America. Community Climate System Model (CCSM4) high resolution ensemble simulations for the 20th and 21st centuries are diagnosed to identify the PE. Analysis of the 20th century simulations indicated that the CCSM4 accurately captures the spatial and temporal climatology of the PE. Analysis of the end 21st century simulations indicate a significant increase in storm duration and intensity of precipitation associated with landfall of the PE. Only a modest increase in the number of atmospheric rivers of a few percent is projected for the end of 21st century.
Article
The Western United States is vulnerable to socioeconomic disruption due to extreme winter precipitation and floods. Traditionally, forecasts of precipitation and river discharge provide the basis for preparations. Herein we show that earlier event awareness may be possible through use of horizontal water vapor transport (IVT) forecasts. Applying the potential predictability concept to the NCEP global ensemble reforecasts, across 31 winters, IVT is found to be more predictable than precipitation. IVT ensemble forecasts with the smallest spreads (least forecast uncertainty) are associated with initiation states with anomalously high geopotential heights south of Alaska, a set-up conducive for anticyclonic conditions and weak IVT into the Western United States. IVT ensemble forecasts with the greatest spreads (most forecast uncertainty) have initiation states with anomalously low geopotential heights south of Alaska and correspond to atmospheric rivers. The greater IVT predictability could provide warnings of impending storminess with additional lead times for hydrometeorological applications.