Ocean Engineering 266 (2022) 112874

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Projected wave storm conditions under the RCP8.5 climate change scenario
in the North Atlantic Ocean
Andressa D’Agostini , Mariana Bernardino , C. Guedes Soares *
Centre for Marine Technology and Ocean Engineering (CENTEC), Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal

A R T I C L E I N F O A B S T R A C T

Keywords: The projected future wave storm conditions in the North Atlantic Ocean are investigated using a Lagrangian
North atlantic ocean methodology applied to a 120 years (1980–2100) dataset resulting from a run of the WAVEWATCH III model
Wave storms forced by EC-Earth winds and ice cover under the RCP8.5 climate change scenario. Several wave storm temporal-
Lagrangian approach
spatial characteristics are analysed, among them the number of storms, storm maximum significant wave height
RCP8.5 climate change scenario
WAVEWATCH III
(HSmax), maximum storm area, storm area associated with HSmax, storm track distribution and duration. When
comparing storm conditions by the end of the 21st century (2070–2100) with the historical period (1980–2010),
there is a decrease in the number of storms, although more intense events occur, i.e., with higher HSmax. The
storm track distribution differs in the mid-latitude regions, with a significant decrease in the number of storm
tracks, and a significant increase in latitudes above 65◦ N by the end of the 21st century. The differences between
the two periods are more expressive in the winter season.

1. Introduction
Wave storms are generated by severe weather conditions and can be
considered as a spatial-temporal phenomenon where the significant
wave height rises above a critical threshold and stands above it for a
time interval greater than 12 h (Boccotti, 2000). Other definitions of
storms are possible using different thresholds (Labeyrie, 1991) but this
definition is adopted in this work.
It is worth mentioning that additional features such as directional
distribution could be considered in the definition of storms (Laface et al.,
2015), for example, extratropical storms (Ponce de Leon and Guedes
Soares, 2014; Gramcianinov et al., 2020) have a smaller directional
variability than the tropical storms such as typhoons (Ou et al., 2002; Liu
et al., 2020).
The storm sea states affect a variety of human activities, as they
represent a risk for ship safety and their severe conditions imply longer
journeys and higher fuel consumption (Vettor and Guedes Soares,
2016), especially for the North Atlantic Ocean which presents one of the
world’s most intensely trafficked ship routes and has varied economic
activity, such as natural resources exploitation (e.g., fishing, crude oil,
natural gas). The operation and safety of these economic activities
depend upon favourable weather conditions as calm sea states. The
understanding of wind-wave climate in historical, present and future
conditions, particularly regarding wave storm events, plays a key role in
the planning and safety of several activities such as offshore operations,
design of ship structures (Bitner-Gregersen et al., 2018), and port ac­
tivities (Izaguirre et al., 2021).
Studies investigating historical periods verified an increase in sig­
nificant wave height (Hs) in the North Atlantic Ocean, for both the mean
(Bertin et al., 2013) and extreme values (Young and Ribal, 2019), and an
increase in the number of wave storms events (Bernardino and Guedes
Soares, 2016). Bertin et al. (2013) forced a wave model with 20th-cen­
tury atmospheric reanalysis for 109 years (1900–2009) and found an
increase for Hs, superimposing the inter-annual variability, over the
whole North Atlantic Ocean. A climatological study of wave storms,
using 35 years (1979–2014) of the Era-interim reanalysis evidenced an
increase of 0.17 storms per year of wave extreme events (Hs above 7 m)
(Bernardino and Guedes Soares, 2016). Young and Ribal (2019) ana­
lysed an extensive satellite database over 33 years (1985–2018) and
showed a positive and statistically significant trend for wave extreme
events (90th percentile) of approximately +0.8 cm/year.
Despite the historical increasing trend, the future projections point in
another direction. Several studies present a consensus on a decrease in
Hs mean and extreme conditions over the North Atlantic Ocean (Hemer
et al., 2013; Aarnes et al., 2017; Morim et al., 2018; Lemos et al., 2020a,
b, 2021; Meucci et al., 2020; O’Grady et al., 2021; Belmadani et al.,

* Corresponding author.
E-mail address: c.guedes.soares@centec.tecnico.ulisboa.pt (C. Guedes Soares).

https://doi.org/10.1016/j.oceaneng.2022.112874
Received 15 April 2022; Received in revised form 25 September 2022; Accepted 9 October 2022
Available online 23 October 2022
0029-8018/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
A. D’Agostini et al. Ocean Engineering 266 (2022) 112874

2021; Lobeto et al., 2021). For mean conditions the decrease is uniform,
however, for wave extremes, certain regions show an increase, such as
the east coast of North America (Lobeto et al., 2021). Moreover, the
changes in the projected scenarios vary between seasons, mostly in the
winter months (Hemer et al., 2013; Lemos et al., 2020a,b, 2021).
The projected scenario studies used the Representative Concentra­
tion Pathways (RCP) scenarios, RCP4.5 and RCP 8.5, to assess wind-
wave 21st-century future projections. The RCP represents future sce­
narios with different greenhouse gas concentrations until the year 2100
(Moss et al., 2010) adopted by the Intergovernmental Panel on Climate
Change (IPCC) for its Fifth Assessment Report (IPCC et al., 2014). The
RCP 8.5 combines demographic and economic trends representing the
worst-case scenario with the highest greenhouse gas emissions.
Most works concerning future projections use an Eulerian approach,
which analyses the Hs time series at given locations. An alternative
method is a Lagrangian approach, which enables the investigation of
wave storms from a dynamic and spatial perspective (Bernardino et al.,
Fig. 1. North Atlantic domain used in the present work and wave storms tracks
2008). This approach was used in previous studies to assess wave storm
identified in 1997 winter.
dynamics in historical years (Rusu et al., 2018; Bernardino et al., 2020).
Bernardino and Guedes Soares (2015) characterised extreme storms in
the North Atlantic and compared the Lagrangian and Eulerian ap­ covers the extratropical cyclone genesis area described in Gramcianinov
proaches, whereas, with the first one, the authors were able to evaluate et al. (2020).
distinctive features as the storm trajectory and its size as they developed. Following the wave storm definition of Boccotti, 2000), wave storm
In this context, the present work uses the Lagrangian methodology detection was considered when the significant wave height occurred
proposed by Bernardino et al. (2008), to analyse changes in the wave above a threshold, assumed to be 7 m in the present work, with a min­
storm temporal-spatial characteristics in a future climate scenario (RCP imum duration of 12 h. These same thresholds were used by Bernardino
8.5) over the North Atlantic Ocean. The overall characterization of wave and Guedes Soares (2015, 2016) for the North Atlantic. After the storm’s
storm parameters is produced and a comparison between the historical detection, a spatial parameterization of each storm is performed
period (1980–2010) and the future (2010–2100) is performed to un­ considering its geometrical attributes (e.g., storm area above the
derstand how wave storm characteristics may change. Special attention threshold, storm centre), at each point of storm trajectories regarding its
is given to the end of the 21st century period (2070–2100). spatial relation.
The wave data analysed are from a WW3 version 5.16 (WW3DG,
2016), simulation performed by Bernardino et al. (2021, 2022), from
2. Material and methods
1980 until 2100, forced by the EC-Earth sea-ice cover (1◦ x1◦ ; monthly)
and surface wind (U10; 0.5◦ × 0.5◦ ; 3-hourly). Hereafter this wave
It is possible to study wave storm dynamics from different perspec­
model simulation will be referred to as WW3/EC-Earth. WW3 is a
tives. The usual one is the Eulerian approach which considers the sig­
third-generation spectral wave model developed at the US National
nificant wave height field h(x, y, t) as a function of fixed positions x, y,
Centers for Environmental Prediction (NOAA/NCEP), which solves the
and time t. In contrast, the Lagrangian approach considers the wave
random phase spectral action density balance equation for
storms as dynamic spatial structures, represented by vector fields. The
wavenumber-direction spectra (WW3DG, 2016).
advantage of this method over Eulerian is the analysis of each wave
EC-Earth is developed by scientists and an institutions consortium of
storm’s characteristics over time and space.
over 10 European countries (Hazeleger et al., 2010), based on
The Lagrangian method applied in the present work has the
state-of-the-art models including atmosphere, ocean sea ice, and
following steps: storm detection as spatial outliers above a given
biosphere (Hazeleger et al., 2012). The present study used the EC-Earth
threshold, spatial parameterization of each storm and construction of
simulation delivered by the Swedish Meteorological and Hydrological
the trajectories in time. The wave storm area (Ω) is considered in the
Institute (RCA4; Strandberg et al., 2015), which considers the RCP 8.5
space domain (x, y) for each time (t), as the restricted grid points where
future scenario (high greenhouse gas concentration) (Moss et al., 2010).
Ω(t) = {(x, y) : h(x, y, t) > z} (1) The simulation followed the 5th Coupled Model Intercomparison Project
(CMIP5, Taylor et al., 2012) protocol for long-term simulations with
the h(x, y, t), representing the significant wave height field, is above the implementation for a historical run from 1980 until 2005 and a future
wave storm definition threshold z. The pulse dynamics W(x,y,t) function run from 2006 until 2100. EC-Earth was chosen because this model is
is given by known for being an excellent climate model representing well the at­
{ mospheric conditions and their interannual variability (Hazeleger et al.,
h(x, y, t) − z, (x, y)ε Ω(t)
W(x, y, t) = (2) 2012). For example, Aarnes et al. (2017) used wave ensembles for six
0, [(x, y) ∕
∈ Ω(t)] or [t ∕ ∈ [0, D]]
CMIP5 models and found better historical mean significant waves with
the wave model forced by EC-Earth 10-m wind and sea ice
for each time moment 0 ≤ t ≤ D, where D is the total duration of wave
concentration.
storm above level z, that may be interpreted as a spatial stochastic pulse.
The WW3/EC-Earth run experiment had a 0.5◦ spatial grid and 3 h of
Pulse dynamics W(x, y, t) depend on a group of parameters Ξ(t) which
temporal resolution, for an area covering the North Atlantic Ocean
characterise the storm’s spatial shape described by Bernardino and
(Fig. 1). These simulations were validated by Bernardino et al. (2021),
Guedes Soares (2015).
and used in other previous works (Freitas et al., 2022; Bernardino et al.,
The Lagrangian methodology is applied to a historical and projected
2022). Bernardino et al. (2021) assessed the wave model performance
wave data set (period from 1980 to 2100) to perform wave storm
comparing it with ERA5 reanalysis for the historical period between
analysis in the North Atlantic Ocean (Fig. 1). Here the Lagrangian
1980 and 2009. The authors evaluated the significant wave height (Hs),
method enables the analysis of individual wave storm parameters, such
peak period (Tp), and mean period (Tm) for mean and extreme condi­
as its track distribution (Fig. 1). Due to computational limitations, the
tions (95th percentile). The EC-Earth wind magnitude was also
analysed area is concentrated between 30◦ N - 70◦ N (Fig. 1), where it

2
A. D’Agostini et al.
compared with ERA5 wind. The paper presents discrepancies panels for
all the variables. In general, the WW3/EC-Earth Hs evaluation demon­
strated a good agreement in most North Atlantic Ocean for mean and
extreme conditions, with higher discrepancies for extreme values on the
western coast of Greenland (~1.5 m).
From the Lagrangian methodology, several wave storm parameters
are estimated: number of wave storms over the years, their duration in
hours, storm maximum significant wave height (HSmax) in meters,
maximum storm area (km2), storm area associated with HSmax (km2),
and storm tracks distances (km) and distribution. The analyses included
an overall characterization of storm parameters in the 120 years
(1980–2100), with annual and seasonal analysis of the above-cited pa­
rameters, and analysis of the differences between 30-year slices between
historical (1980–2010) and future (2070–2100) projections, including
the Kruskal-Wallis (KW) (Kruskal and Wallis, 1952), non-parametric test
with 95% confidence interval to identify significant physical changes in
wave storm characteristics between historical periods. To also under­
stand the changes spatially, wave storm track density is computed for
every 30 years from 1980 to 2010 to 2070–2100, and the differences
between periods are estimated. For better visualization of change in
patterns between future and historical periods a Gaussian filter (sigma
value equals 1) is applied to wave storm track density.
3. Results
A Lagrangian method is applied to wave data (Hs) from 120 years
(1980–2100) dataset from the simulation of WW3/EC-Earth under the
RCP8.5 emission scenario. The wave storm parameters oscillate through
the 120 years, presenting a high interannual variability with a visible
negative trend for wave storm occurrence (Fig. 2A), especially at the end
of the 21st century, and a smoother positive trend for wave storm HSmax
(Fig. 2C). Storm duration (Fig. 2B) and track distances (Fig. 2F) do not
present notable trends, and storm areas at HSmax (Fig. 2D) and storm
maximum areas (Fig. 2E) present a slightly negative trend.
To investigate major differences in wave storm conditions between
the historical period (1980–2010) and the future projections
Fig. 2. Temporal evolution of the number of wave storms (A) and annual mean values of wave storms parameters: Wave storms duration (hours) (B), Wave storms
HSmax (meters) (C), Wave storms area at HSmax (km2) (D), Maximum wave storm area (km2) (E), Wave storm tracks distances (km) (F). The gray line indicates the
mean, and the dashed blue line indicates the trend in the 120 years (1980–2100). (For interpretation of the references to colour in this figure legend, the reader is
referred to the Web version of this article.)
3
Ocean Engineering 266 (2022) 112874
(2070–2100) a statistical test is performed (Table 1), and wave storm
frequency distribution is examined (Figs. 3 and 4). The wave storm
occurrence reduction in future projections (2070–2100) presents a mean
13% lower than the 1980–2010 period and this difference is statistically
significant (Table 1). In all months for the analysed 30-year periods, the
wave storm occurrence was reduced by the end of the 21st century
(Fig. 4). The winter months presented 176 fewer occurrences, repre­
senting 42,72% of the total reduction in the number of storms (412). The
storm HSmax also presents significant differences between historical
and projected periods, with an HSmax slightly higher by the end of the
21st century (2070–2100) (Table 1) indicating that despite the storm
occurrence decreasing, the events may be more severe. Storm duration,
area of HSmax, maximum area, and track distance present a greater
standard deviation at the end of 21st indicating greater variability,
however, did not present significant differences between historical and
projected periods (Table 1). The confidence intervals (CI) in Table 1
corroborate with the significance of the KW test, where the CI are
different for the number of wave storms and have just a little intersec­
tion (0.2 m) of CI for the Wave storm HSmax parameter when comparing
both analysed periods (1980–2010 and 2070–2100).
The relative frequency distribution of wave storm parameters be­
tween historical and projected periods shows few differences (Fig. 3).
The HSmax relative frequency present slightly higher occurrences in
HSmax above 9 m for projected wave storms at about 3% higher, which
indicates more intense storms (Fig. 3B), also demonstrated by the higher
HSmax mean (Table 1). After 48 h about 70% of the wave storms
dissipate (Fig. 3A), and about 65% of storm track distances occur until
1500 km (Fig. 3E) for both periods (1980–2010 and 2070–2100). Wave
storm area parameters at HSmax (Fig. 3C) and the maximum area
(Fig. 3D) exhibit similar relative frequency distribution between his­
torical and projected periods, with more than 50% occurring in less than
300 km2 of the area associated with HSmax, and in less than 400 km2 for
storm maximum area in a single track.
Wave storms in the North Atlantic present a high seasonal distribu­
tion, with severe wave conditions in winter months (December, January,
and February - DJF) and calmer conditions in summer months (June,
A. D’Agostini et al. Ocean Engineering 266 (2022) 112874

Table 1
Annual mean, minimum and maximum (min - max), confidence interval (CI), standard deviation (STD) of wave storm characteristics and Kruskal-Wallis p-value from
differences for each parameter between the end of the 21st century (2070–2100) and historical period (1980–2010). The values with statistically significant changes at
the 5% level are in bold.
years interval mean min - max CI STD KW p-value

Number of wave storms 1980–2010 105.6 79–126 101.6–109.6 10.5 0.00002

2070–2100 91.9 74–117 87.7–95.9 10.8
Wave storm duration (h) 1980–2010 35.2 12–147 34.5–35.9 1.9 0.42037
2070–2100 35.7 12–141 34.9–36.4 2.1
Wave storm HSmax (m) 1980–2010 9.41 7–19.2 9.34–9.48 0.19 0.01949
2070–2100 9.56 7–19.8 9.46–9.65 0.25
Wave storm area of HSmax (km2) 1980–2010 498.7 11.4–4355.2 476.7–521.0 58.3 0.79014
2070–2100 492.2 11.4–4355.2 462.1–522.1 79.2
Maximum area of wave storms (km2) 1980–2010 591.7 11.4–5094.7 565.6–617.8 68.6 0.95284
2070–2100 592.0 12.2–5506.5 556.1–627.6 94.4
Wave storm tracks distance (km) 1980–2010 1316.7 0–6455 1284.6–1348.8 84.6 0.19835
2070–2100 1353.8 0–6332 1312.5–1395.0 108.6

July, and August) (Woolf et al., 2002; Stopa and Cheung, 2014). This
strong seasonality explains the much lower values in the results of Fig. 5
for summer than for winter season differences in the storm wave track
densities. A wave storm occurrence decrease is observed over the
months for the end of the 21st century (2070–2100), especially in the
winter months (Fig. 4). Storm distribution in the summer months re­
mains similar between historical and future periods (Fig. 4).
To better understand changes spatially, storm track density for
1980–2010 and 2070–2100 years are computed, and their differences
are verified (Figs. 5 and 6). When analysing the seasonality of wave
storm track distribution, it is possible to verify the differences between
wave storm climates in 1980–2010 and 2070–2100 are heterogeneously
distributed over the North Atlantic Ocean (Fig. 5). The winter months
(DJF) exhibited greater differences (order of 4%) with a significant in­
crease of storms in high latitudes (~65◦ N), mainly in the region between
Greenland and Iceland, and a significant reduction in a more extensive
area between 55◦ -65◦ N and 55◦ -10◦ W (Fig. 5, DJF). In the summer
months (JJA) wave storm tracks distribution differences remain lower
than 1% in magnitude and has a significant decrease in the North
Atlantic Ocean among 45◦ -55◦ N and 50◦ -15◦ W (Fig. 5, JJA). Especially
for the winter months, the ice retraction over the 2070–2100 period
seems to influence the differences found in high latitudes where storms
increased (Fig. 5, DJF). Less ice cover represents greater wind fetch.
Considering the wave storm change overview between historical
(1980–2010) and projected (2070–2100) periods in the North Atlantic
Ocean tends to have calmer conditions (Fig. 6), with a maximum value
reduction of 5,45%. Significant differences in wave storm decrease occur
mainly among 50◦ -65◦ N and 50◦ -10◦ W, other minor significant decrease
regions occur between 40◦ -50◦ N and 20◦ -30◦ W, and 35◦ -40◦ N and 55◦ -
45◦ W (Fig. 6), probably associated with wind intensity reduction. Above
latitude 65◦ N, wave storms raised significantly, probably related to sea
ice retraction (Fig. 6), with a maximum increase of 5,58%.
4. Discussion
The projected changes in extreme wave climate are probably asso­
ciated with several factors, including sea ice cover retraction and fewer
storms occurrence. The intense greenhouse gas concentration at the end
of the 21st century will lead to a strong increase in surface air temper­
ature in the Arctic when compared to the extratropical regions, and
consequent sea ice extent loss (Overland et al., 2014). The weaker
temperature gradient between regions drives a reduction in baroclinic
instability and cyclogenesis and the consequent reduction of about 17%
in explosive cyclones over the North Atlantic Ocean at the end of the
21st century as studied by Seiler and Zwiers (2016), which could be
related to the significant reduction in the number of wave storms found
in the present study. Despite the decrease in storm occurrence, Seiler
and Zwiers (2016) also found a slight intensification of the explosive
cyclones on North America’s East Coast associated with the jet stream
speed increase. The eastern coast of Greenland is also a favourable re­
gion for cyclogenesis, with more than 2 cyclones (10− 6 km2)− 1 per
month (Gramcianinov et al., 2020), the projected increase of wave
storms above 65◦ N found in this work could be related to a cyclogenesis
intensification in this region.
Despite not using ocean currents interaction in our wave simulation,
it could represent another possible source of impact on wave climate
change. Wave-current interaction can result in relevant wave field het­
erogeneity, as pointed out by Ponce de Leon and Guedes Soares (2021,
2022). This may be relevant in the North Atlantic Ocean considering
that the Atlantic Meridional Overturning Circulation (AMOC) might
decline between 6 and 8 Sv (34–45%) by 2100 (Weijer et al., 2020).
The North Atlantic Ocean presents a strong seasonal distribution and
variability of wave storms, with higher occurrence in winter months and
high interannual variability (Young, 1999; Woolf et al., 2002; Stopa and
Cheung, 2014; Bernardino and Guedes Soares, 2016). The present sim­
ulations between the years 1980 and 2100 show high interannual
variability (Fig. 2) and a major difference, between historical and the
end of the 21st century, in storm track densities for winter months than
for summer months (Fig. 5). These differences according to season are
probably related to extratropical cyclogenesis in the North Atlantic
Ocean which is more active during the winter months (Gramcianinov
et al., 2020).
In line with the results presented here, other studies also verified a
wave climate change with less intense waves in the North Atlantic Ocean
(Hemer et al., 2013; Morim et al., 2018; Lemos et al., 2020), and more
severe wave conditions along high latitudes (above 65◦ N) (Aarnes et al.,
2017; Bricheno and Wolf, 2018; Meucci et al., 2020; O’Grady et al.,
2021), probably related to sea ice retraction and increased cyclogenesis.
With a similar wave model experiment to the present study, Bricheno
and Wolf (2018) used WW3 forced by EC-Earth model for projected
future wave conditions considering two RCP emission scenarios (4.5 and
8.5) for the European Atlantic coast and verified a decrease in significant
wave height mean of about 5–10%, with an increase in waves above
60◦ N regions associated with sea ice reduction.
Aarnes et al. (2017) analysed wave model ensembles, for RCP4.5 and
RCP8.5 scenarios for Northeast Atlantic. The authors found a mean
significant wave height decrease by the end of the 21st century, espe­
cially under RCP8.5 scenario, except in higher latitudes. Meucci et al.
(2020) also used wave model ensembles although focusing on changes
in extreme wind-wave events by 2100 and noted a significant reduction
(− 5 to − 15%) in extreme Hs for mid and low latitudes in the North
Atlantic basin, and an increase in high latitudes. Using a 10-member
ensemble Lemos et al. (2021) found an 11% decrease in the annual
mean of Hs by the end of the 21st century considering the RCP8.5 sce­
nario, and strong seasonality in projected changes with reduction peaks
of 9% in winter, 13% in spring and 15% in the autumn season. For
tropical cyclone-related wave climate, Belmadani et al. (2021) found a
decrease in average Hs throughout the hurricane season (July to

4
A. D’Agostini et al. Ocean Engineering 266 (2022) 112874

Fig. 3. Relative frequency distribution of wave storm parameters between the end of the 21st century (2070–2100) and the historical period (1980–2010). A- Wave
storms duration (hours), B- Wave storms HSmax (meters), C- Wave storms area at HSmax (km2), D- Maximum wave storm area (km2), E− Wave storm tracks
distances (km).

Fig. 4. Monthly distribution of wave storm occurrence between the end of the 21st century (2070–2100) and the historical period (1980–2010).

5
A. D’Agostini et al.
Fig. 6. Percentage change of wave storms tracks distribution by the end of the
21st century (2070–2100) relative to the 1980–2010 period. The regions with
statistically significant changes at the 5% level are hatched. The maximum ice
cover is marked in black lines hatched for the 1980–2010 period and dotted for
the 2070–2100 period.
November) driven by a weaker and poleward-displaced anticyclone.
There is a consensus towards a decrease of significant wave height in
the North Atlantic Ocean regarding all seasons and scenarios, which are
generally agreeing with projected wind changes (Morim et al., 2018).
Considering the present historical simulation period (1980–2010), the
work of Bernardino and Guedes Soares (2016) presented related results
of wave storms’ relative frequencies when analysing 35 years
(1979–2014) of ERA-Interim dataset and the authors also used the
Lagrangian method. The authors related the wave storm occurrence over
the years with positive and negative phases of the North Atlantic
Oscillation (NAO), and despite not finding significant differences, wave
storms were longer in duration, with longer tracks, higher HSmax, and
affected a larger maximum area under negative NAO.
It is well known that uncertainties are inherent to climate modelling
and present a cascading behaviour, especially for projected future cli­
mates (Morim et al., 2018). The use of ensemble models enables
assessing robustness and uncertainty and is being applied in recent
studies (Hemer et al., 2013; Aarnes et al., 2017; Morim et al., 2020;
Lemos et al., 2020b; Meucci et al., 2020). Further investigation
considering the Lagrangian methodology over multi-model ensembles
shall provide a better understanding of model uncertainties.
5. Conclusions
In this study, the wave storm conditions are investigated under a
climate change scenario in the North Atlantic Ocean. To achieve this
goal a Lagrangian approach was employed over 120 years of
(1980–2100) wave simulation which used wind and sea ice cover from
EC-Earth under the RCP8.5 scenario. The wave storms were considered
when the significant wave height is higher than 7 m and its duration is
6
Ocean Engineering 266 (2022) 112874
Fig. 5. Percentage change in winter (DJF) and sum­
mer (JJA) of wave storms tracks distribution by the
end of the 21st century (2070–2100) relative to the
1980–2010 period. The regions with statistically sig­
nificant changes at the 5% level are hatched. The
maximum ice cover is marked in black lines hatched
for the 1980–2010 period and dotted for the
2070–2100 period. The colour scales are different
between panels. (For interpretation of the references
to colour in this figure legend, the reader is referred
to the Web version of this article.)
longer than 12 h.
The Lagrangian perspective allowed a temporal-spatial character­
ization of the wave storms, including their frequency over the years,
wave storm duration, storm maximum significant wave height (HSmax),
maximum storm area, storm area associated with HSmax, and storm
track distances. All the cited characteristics experienced a high inter­
annual variability, with highlights for the reduction in the number of
storms and increased HSmax between the projected (2070–2100) and
the historical period (1980–2010). The storm duration, track distance,
area of the storm and associated with HSMax do not present significant
differences at the end of the 21st century. On the other hand, the results
were significant for the wave storm occurrence and maximum HSmax.
By the end of the 21st century wave storms are found to be overall less
frequent in the North Atlantic Ocean by about − 5,5%, although these
events may be more severe (higher HSmax).
The wave storm track density is estimated to understand the spatial
changes in the projected scenarios. When comparing the historical
(1980–2010) period and the end of the 21st century (2070–2100), the
distribution of wave storm tracks density differs between mid-latitude
regions, with a significant decrease in latitudes below 65◦ N, and a sig­
nificant increase in latitudes above 65◦ N, presenting more expressive
differences in the winter season for both regions.
CRediT authorship contribution statement
Andressa D’Agostini: Methodology, Formal analysis, Visualization,
Writing – original draft. Mariana Bernardino: Methodology, Writing –
original draft. C. Guedes Soares: Writing – review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Data availability
No data was used for the research described in the article.
Acknowledgements
We acknowledge the Swedish Meteorological and Hydrological
Institute for the EC-Earth simulation used in this study. This work was
performed within the Project CLIMENA - “CLimate change Impacts on
the Marine Environment of the North Atlantic”, which is co-funded by
the European Regional Development Fund (Fundo Europeu de Desen­
volvimento Regional - FEDER) and by the Portuguese Foundation for
Science and Technology (Fundação para a Ciência e a Tecnologia - FCT)
under contract number PTDC/EAM-OCE/28561/2017. This work con­
tributes to the Strategic Research Plan of the Centre for Marine Tech­
nology and Ocean Engineering (CENTEC), which is financed by the
Portuguese Foundation for Science and Technology (Fundação para a
A. D’Agostini et al.
Ciência e Tecnologia) under contract UIDB/UIDP/00134/2020.
References
Aarnes, O.J., Reistad, M., Breivik, O., Bitner-Gregersen, E., Eide, L.I., Gramstad, O.,
Magnusson, A.K., Natvig, B., Vanem, E., 2017. Projected changes in significant wave
height toward the end of the 21st century: Northeast Atlantic. J. Geophys. Res.:
Oceans 122, 1–22. https://doi.org/10.1002/2016JC012264.
Belmadani, A., Dalphinet, A., Chauvin, F., Pilon, R., Palany, P., 2021. Projected future
changes in tropical cyclone-related wave climate in the North Atlantic. Clim. Dynam.
56, 3687–3708. https://doi.org/10.1007/s00382-021-05664-5.
Bernardino, M., Boukhanovsky, A., Guedes Soares, C., 2008. Alternative approaches to
storm statistics in the ocean. Proceedings of the International Conference on
Offshore Mechanics and Arctic Engineering - OMAE 2, 1061–1068. https://doi.org/
10.1115/OMAE2008-58053.
Bernardino, M., Guedes Soares, C., 2015. A Lagrangian perspective of the 2013/2014
winter wave storms in the North Atlantic. In: Guedes Soares, C., Santos, T.A. (Eds.),
Maritime Technology and Engineering. Taylor & Francis Group, London, UK,
pp. 1381–1387. https://doi.org/10.1201/b17494-187.
Bernardino, M., Guedes Soares, C., 2016. A climatological analysis of storms in the North
Atlantic. In: Guedes Soares, C., Santos, T.A. (Eds.), Maritime Technology and
Engineering. Taylor & Francis Group, London, UK, pp. 1021–1026. https://doi.org/
10.1201/b21890-137, 3.
Bernardino, M., Rusu, L., Guedes Soares, C., 2020. Evaluation of extreme storm waves in
the Black Sea. Journal of Operational Oceanography 1–15. https://doi.org/10.1080/
1755876X.2020.1736748.
Bernardino, M., Goncalves, M., Guedes Soares, C., 2021. Marine climate projections
towards the end of the 21st century in the North Atlantic. ASME. J. Offshore Mech.
Arct. Eng. 143, 061201 https://doi.org/10.1115/1.4050698.
Bernardino, M., Goncalves, M., Campos, R.M., Guedes Soares, C., 2022. Extremes and
Variability of Wind and Waves across the Oceans until the End of the 21st Century.
Manuscript submitted for publication.
Bertin, X., Prouteau, E., Letetrel, C., 2013. A significant increase in wave height in the
North Atlantic Ocean over the 20th century. Global Planet. Change 106, 77–83.
https://doi.org/10.1016/j.gloplacha.2013.03.009.
Bitner-Gregersen, E.M., Vanem, E., Gramstad, O., Hørte, T., Aarnes, O.J., Reistad, M.,
Breivik, Ø., Magnusson, A.K., Natvig, B., 2018. Climate change and safe design of
ship structures. Ocean Eng. 149, 226–237. https://doi.org/10.1016/j.
oceaneng.2017.12.023.
Boccotti, P., 2000. Wave Mechanics for Ocean Engineering. Elsevier Science, New York.
Bricheno, L.M., Wolf, J., 2018. Future wave conditions of Europe, in response to high-end
climate change scenarios. J. Geophys. Res.: Oceans 123 (12), 8762–8791. https://
doi.org/10.1029/2018JC013866.
Freitas, A., Bernardino, M., Guedes Soares, C., 2022. The influence of the Arctic
Oscillation on North Atlantic wind and wave climate by the end of the 21st century.
Ocean Eng. 246, 110634 https://doi.org/10.1016/j.oceaneng.2022.110634.
Gramcianinov, C.B., Campos, R.M., de Camargo, R., Hodges, K.I., Guedes Soares, C., da
Silva Dias, P.L., 2020. Analysis of Atlantic extratropical storm tracks characteristics
in 41 years of ERA5 and CFSR/CFSv2 databases. Ocean Eng. 216, 108111 https://
doi.org/10.1016/j.oceaneng.2020.108111.
Hazeleger, W., Severijns, C., Semmler, T., Ştefǎnescu, S., Yang, S., Wang, X., Wyser, K.,
Dutra, E., Baldasano, J.M., Bintanja, R., Bougeault, P., Caballero, R., Ekman, A.M.L.,
Christensen, J.H., Van Den Hurk, B., Jimenez, P., Jones, C., Kållberg, P., Koenigk, T.,
et al., 2010. EC-Earth: a seamless Earth-system prediction approach in action. Bull.
Am. Meteorol. Soc. 91 (10), 1357–1363. https://doi.org/10.1175/
2010BAMS2877.1.
Hazeleger, W., Wang, X., Severijns, C., Ştefǎnescu, S., Bintanja, R., Sterl, A., Wyser, K.,
Semmler, T., Yang, S., van den Hurk, B., van Noije, T., van der Linden, E., van der
Wiel, K., 2012. EC-Earth V2.2: description and validation of a new seamless earth
system prediction model. Clim. Dynam. 39 (11), 2611–2629. https://doi.org/
10.1007/s00382-011-1228-5.
Hemer, M.A., Fan, Y., Mori, N., Semedo, A., Wang, X.L., 2013. Projected changes in wave
climate from a multi-model ensemble. Nat. Clim. Change 3 (5), 471–476. https://
doi.org/10.1038/nclimate1791.
IPCC, 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups I,
II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change [Core Writing Team. In: Pachauri, R.K., Meyer, L.A. (Eds.). IPCC, Geneva,
Switzerland, p. 151.
Izaguirre, C., Losada, I.J., Camus, P., Vigh, J.L., Stenek, V., 2021. Climate change risk to
global port operations. Nat. Clim. Change 11 (1), 14–20. https://doi.org/10.1038/
s41558-020-00937-z.
Kruskal, W., Wallis, W., 1952. Use of ranks in one-criterion variance analysis. J. Am. Stat.
Assoc. 47 (260), 583–621. https://doi.org/10.2307/2280779.
Labeyrie, J., 1991. Time scales and statistical uncertainties in the prediction of extreme
environmental conditions. Reliab. Eng. Syst. Saf. 32 (3), 243–266. https://doi.org/
10.1016/0951-8320(91)90002-O.
Laface, V., Arena, F., Guedes Soares, C., 2015. Directional analysis of sea storms. Ocean
Eng. 107, 45–53. https://doi.org/10.1016/j.oceaneng.2015.07.027.
Lemos, G., Semedo, A., Dobrynin, M., Menendez, M., Miranda, P.M.A., 2020a. Bias-
corrected cmip5-derived single-forcing future wind-wave climate projections toward
the end of the twenty-first century. J. Appl. Meteorol. Climatol. 59 (9), 1393–1414.
https://doi.org/10.1175/JAMC-D-19-0297.1.
Ocean Engineering 266 (2022) 112874
Lemos, G., Menendez, M., Semedo, A., Camus, P., Hemer, M., Dobrynin, M., Miranda, P.
M.A., 2020b. On the need of bias correction methods for wave climate projections.
Global Planet. Change 186 (December 2019), 103109. https://doi.org/10.1016/j.
gloplacha.2019.103109.
Lemos, G., Menendez, M., Semedo, A., Miranda, P.M.A., Hemer, M., 2021. On the
decreases in North Atlantic significant wave heights from climate projections. Clim.
Dynam., 0123456789 https://doi.org/10.1007/s00382-021-05807-8.
Liu, G.L., Li, X., Wang, J.H., Kou, Y., Wang, X.P., 2020. Research on the statistical
characteristics of typhoon frequency. Ocean Eng. 209, 107489 https://doi.org/
10.1016/j.oceaneng.2020.107489.
Lobeto, H., Menendez, M., Losada, I.J., 2021. Future behavior of wind wave extremes
due to climate change. Sci. Rep. 11 (1), 1–13. https://doi.org/10.1038/s41598-021-
86524-4.
Meucci, A., Young, I.R., Hemer, M., Kirezci, E., Ranasinghe, R., 2020. Projected 21st
century changes in extreme wind-wave events. Sci. Adv. 6 (24) https://doi.org/
10.1126/sciadv.aaz7295.
Morim, J., Hemer, M., Cartwright, N., Strauss, D., Andutta, F., 2018. On the concordance
of 21st century wind-wave climate projections. Global Planet. Change 167
(February), 160–171. https://doi.org/10.1016/j.gloplacha.2018.05.005.
Morim, J., Trenham, C., Hemer, M., Wang, X.L., Mori, N., Casas-Prat, M., Semedo, A.,
Shimura, T., Timmermans, B., Camus, P., Bricheno, L., Mentaschi, L., Dobrynin, M.,
Feng, Y., Erikson, L., 2020. A global ensemble of ocean wave climate projections
from CMIP5-driven models. Sci. Data 7 (1), 1–11. https://doi.org/10.1038/s41597-
020-0446-2.
Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., Van Vuuren, D.P.,
Carter, T.R., Emori, S., Kainuma, M., Kram, T., Meehl, G.A., Mitchell, J.F.B.,
Nakicenovic, N., Riahi, K., Smith, S.J., Stouffer, R.J., Thomson, A.M., Weyant, J.P.,
Wilbanks, T.J., 2010. The next generation of scenarios for climate change research
and assessment. Nature 463 (7282), 747–756. https://doi.org/10.1038/
nature08823.
O’Grady, J.G., Hemer, M.A., McInnes, K.L., Trenham, C.E., Stephenson, A.G., 2021.
Projected incremental changes to extreme wind-driven wave heights for the twenty-
first century. Sci. Rep. 11 (1), 1–9. https://doi.org/10.1038/s41598-021-87358-w.
Ou, S.H., Liau, J.M., Hsu, T.W., Tzang, S.Y., 2002. Simulating typhoon waves by SWAN
wave model in coastal waters of Taiwan. Ocean Eng. 29 (8), 947–971. https://doi.
org/10.1016/S0029-8018(01)00049-X.
Overland, J.E., Wang, M., Walsh, J.E., Stroeve, J.C., 2014. Future Arctic climate changes:
adaptation and mitigation time scales. Earth’s Future 2 (2), 68–74. https://doi.org/
10.1002/2013ef000162.
Ponce de Leon, S., Guedes Soares, C., 2014. Extreme wave parameters under North
Atlantic extratropical cyclones. Ocean Model. 8178–8188. https://doi.org/10.1016/
j.ocemod.2014.07.005.
Ponce de Leon, S., Guedes Soares, C., 2021. Numerical modelling of the effects of the Gulf
stream on the wave characteristics. J. Mar. Sci. Eng. 9, 42. https://doi.org/10.3390/
jmse9010042.
Ponce de Leon, S., Guedes Soares, C., 2022. Numerical study of the effect of current on
waves in the Agulhas Current Retroflection. Ocean Eng. 264, 112333 https://doi.
org/10.1016/j.oceaneng.2022.112333.
Rusu, L., Bernardino, M., Guedes Soares, C., 2018. Analysis of extreme storms in the
Black Sea. In: Guedes Soares, C., Santos, T.A. (Eds.), Progress in Maritime
Technology and Engineering. Taylor and Francis Group, London, UK, pp. 699–704.
Seiler, C., Zwiers, F.W., 2016. How will climate change affect explosive cyclones in the
extratropics of the Northern Hemisphere? Clim. Dynam. 46, 3633–3644. https://doi.
org/10.1007/s00382-015-2791-y.
Stopa, J.E., Cheung, K.F., 2014. Periodicity and patterns of ocean wind and wave
climate. J. Geophys. Res.: Oceans 119, 5563–5584. https://doi.org/10.1002/
2013JC009729.
Strandberg, G., Bärring, L., Hansson, U., Jansson, C., Jones, C., Kjellström, E.,
Kupiainen, M., Nikulin, G., Samuelsson, P., Ullerstig, A., 2015. CORDEX Scenarios
for Europe from the Rossby Centre Regional Climate Model RCA4; Technical Report
116. Climate research–Rossby Centre: Norrkoping, Sweden.
Taylor, K.E., Stouffer, R.J., Meehl, G.A., 2012. An overview of CMIP5 and the experiment
design. Bull. Am. Meteorol. Soc. 93 (4), 485–498. https://doi.org/10.1175/BAMS-D-
11-00094.1.
Vettor, R., Guedes Soares, C., 2016. Development of a ship weather routing system.
Ocean Eng. 123, 1–14. https://doi.org/10.1016/j.oceaneng.2016.06.035.
WW3DG (WAVEWATCH III R Development Group), 2016. User Manual and System
Documentation of WAVEWATCH III R Version 5.16. Tech. Note 329. NOAA/NWS/
NCEP/MMAB, College Park, MD, USA, p. 326. + Appendices.
Weijer, W., Cheng, W., Garuba, O.A., Hu, A., Nadiga, B.T., 2020. CMIP6 models predict
significant 21st century decline of the atlantic meridional overturning circulation.
Geophys. Res. Lett. 47 (12) https://doi.org/10.1029/2019GL086075.
Woolf, D.K., Challenor, P.G., Cotton, P.D., 2002. Variability and predictability of the
North Atlantic wave climate. J. Geophys. Res.: Oceans 107 (10). https://doi.org/
10.1029/2001jc001124.
Young, I.R., 1999. Seasonal variability of the global ocean wind and wave climate. Int. J.
Climatol. 19 (9), 931–950. https://doi.org/10.1002/(SICI)1097-0088(199907)19:
9<931::AID-JOC412>3.0.CO;2-O.
Young, I.R., Ribal, A., 2019. Multiplatform evaluation of global trends in wind speed and
wave height. Science 364 (6440), 548–552. https://doi.org/10.1126/science.
aav9527.