Winter wave climate, storms and regional cycles: The SW Spanish Atlantic coast

Abstract

Climatic change-related impacts on coastal areas became an important issue in past decades and nowadays threaten many human settlements and activities. Coastal hazards are linked to flooding and erosion processes associated with sea level rise and the increased strength of hurricanes, cyclones and storms. The main aim of this work is the characterization of coastal storms in Cadiz (SW Spain) and the determination of their recurrence intervals and relationships with several regional cycles. Storm characterization was carried out using the Storm Power Index (Dolan and Davis, 1992) and five classes were obtained, from class I (weak events) to V (extreme events). Storm occurrence probability was 96% for class I (i.e. almost one event per year) to 3% for class V. The return period for class V was 25 years and ranged from 6 to 8 years for classes III and IV storms, e.g. significant and severe events. Classes I and II showed a period of recurrence ranging from 1 to 3 years. Stormy winter seasons were 2009/10 (12 events), 1995/6 and 2002/3 (with 10 events each) and 1993/4 (8 events). Approximately 40% of the change in monthly wave data and storminess indices was related to several teleconnection patterns, the most important drivers of change being the Arctic Oscillation (AO), 21.45%, and the North Atlantic Oscillation (NAO), 19.65%. It is interesting to note that a great number of storms, larger storm duration and higher values of Storm Power Index were only observed when neutral to strong negative NAO and AO phases occurred at the same time (89 storms and 3355 h) and/or when there was an abrupt change of NAO and AO phases, i.e. they moved from a positive to negative phase without passing through a neutral phase. The results obtained in this work have wider applications for ocean and coastal management. It is suggested that methodology used can be easily applied in different areas where wave buoy data are available. In the same way, information obtained with this kind of work constitutes the first step in the development of coastal protection plans to preserve socio-economic activities from the impact of severe storm events.

Full text

INTERNATIONAL JOURNAL OF CLIMATOLOGY
Int. J. Climatol. 33: 2142 –2156 (2013)
Published online 30 August 2012 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/joc.3579
Winter wave climate, storms and regional cycles: the SW
Spanish Atlantic coast
N. Rangel-Buitrago* and G. Anfuso
Departamento de Ciencias de la Tierra, Facultad de Ciencias del Mar y Ambientales, Universidad de C´adiz, C´adiz, Spain
ABSTRACT: Climatic change-related impacts on coastal areas became an important issue in past decades and nowadays
threaten many human settlements and activities. Coastal hazards are linked to flooding and erosion processes associated
with sea level rise and the increased strength of hurricanes, cyclones and storms. The main aim of this work is the
characterization of coastal storms in Cadiz (SW Spain) and the determination of their recurrence intervals and relationships
with several regional cycles. Storm characterization was carried out using the Storm Power Index (Dolan and Davis, 1992)
and five classes were obtained, from class I (weak events) to V (extreme events). Storm occurrence probability was 96%
for class I (i.e. almost one event per year) to 3% for class V. The return period for class V was 25 years and ranged from
6 to 8 years for classes III and IV storms, e.g. significant and severe events. Classes I and II showed a period of recurrence
ranging from 1 to 3 years. Stormy winter seasons were 2009/10 (12 events), 1995/6 and 2002/3 (with 10 events each) and
1993/4 (8 events). Approximately 40% of the change in monthly wave data and storminess indices was related to several
teleconnection patterns, the most important drivers of change being the Arctic Oscillation (AO), 21.45%, and the North
Atlantic Oscillation (NAO), 19.65%. It is interesting to note that a great number of storms, larger storm duration and
higher values of Storm Power Index were only observed when neutral to strong negative NAO and AO phases occurred at
the same time (89 storms and 3355 h) and/or when there was an abrupt change of NAO and AO phases, i.e. they moved
from a positive to negative phase without passing through a neutral phase. The results obtained in this work have wider
applications for ocean and coastal management. It is suggested that methodology used can be easily applied in different
areas where wave buoy data are available. In the same way, information obtained with this kind of work constitutes the
first step in the development of coastal protection plans to preserve socio-economic activities from the impact of severe
storm events.
KEY WORDS waves; storm; power index; teleconnection patterns; Cadiz; Spain
Received 12 April 2012; Revised 30 July 2012; Accepted 30 July 2012
1. Introduction
The recent research on climate change effects on coastal
zone has been almost completely devoted to the impacts
of sea level rise as a result of the global warming (Komar
and Allan, 2008; Phillips and Crisp, 2010). Less attention
has been given to the knowledge and trend of wave
climate and storms. As suggested by Keim et al. (2004),
wave climate, occurrence and distribution of extreme
waves and storms are important issues in the occurrence
and amount of coastal erosion, the deterioration and (at
some places) complete disappearance of ecosystems and
the safety of shipping and offshore platforms.
Different researchers around the World have recog-
nized that the height of extreme waves, the number of
storms and their intensity have been increasing in last
decades (Bacon and Carter, 1991; Allan and Komar,
2000; Dupuis et al., 2006; Komar and Allan, 2008;
Soomere, 2008). In a scenario of rising sea levels and
* Correspondence to: N. Rangel-Buitrago, Departamento de Ciencias
de la Tierra, Facultad de Ciencias del Mar y Ambientales, Universidad
de C´
adiz, Pol´
ıgono R´
ıo San pedro s/n, 11510 Puerto Real, C´
adiz, Spain.
E-mail: nelson.rangelbuitrago@mail.uca.es
increasing wave heights, the coastline will suffer huge
impacts in terms of erosion and flooding especially with
respect to low-lying regions that may partly or entirely
disappear (Hanson and Larson, 2008).
In order to reduce the impacts of climate change, it
is important to provide realistic analyses of the natural
variability and trends associated with climatic events, in
this case storms. The homogeneity and quality of wave
data over long periods (more than 25 years) are essential
in any analysis of extreme wave heights, storms and their
respective probability of recurrence. Many approaches
have been used to document wave climate changes: visual
observations, wave buoys data, instrumented ships, satel-
lite altimeters, numerical and physical models (Dupuis
et al., 2006; Komar and Allan, 2008).
This work deals with the analysis of a 27 year long
wave buoy dataset recorded at Cadiz (SW Spain). The
analysis consisted of characterizing winter wave climate,
extreme waves and storms characteristics and distribution
and their relationships with regional cycles (Telecon-
nection Patterns) such as the North Atlantic Oscillation
(NAO) Index, the East Atlantic (EA) pattern and the Arc-
tic Oscillation (AO), among others. The methodology
2012 Royal Meteorological Society

WINTER WAVE CLIMATE, STORMS AND REGIONAL CYCLES 2143
used can be easily applied in different coastal areas
around the World where wave buoy data are available for
a monitoring period greater than 25 years. Information
obtained in this study constitutes a first step in developing
coastal response plans to storm impacts for the protection
of socio-economic activities, especially tourism. This will
be a major concern for coastal managers in future years
because storm impacts will be magnified by predicted
sea level rise linked to global climate change (Jones and
Phillips, 2011).
1.1. Study area
This study investigates the wave climate and storm char-
acteristics of an Atlantic Ocean coastal area facing the
city of Cadiz, in South-West Spain (Figure 1). The area
corresponds with a northwest–southeast oriented coast-
line which is characterized by a diversity of coastal land-
forms and environments including sand spits, beaches,
dunes, saltmarshes, cliffs and rocky shore platforms.
It has semidiurnal tides and a mesotidal range, with
mean values of neap and spring tides of 1.0 and 3.5 m,
respectively, and it is affected by western and eastern
winds. Western winds are related to Atlantic low pressure
systems that can continue for several days and affect
large portions of the Iberian Peninsula. They blow from
WNW to WSW directions with a mean annual velocity
of 16 km h−1and a frequency of 13%. Winds, blowing
from E to SE directions, with an annual frequency of
20% and a mean velocity of 28 km h−1, are originally
formed in the Mediterranean Sea and greatly increase in
velocity due to channeling through the Gibraltar Strait.
Due to coastline orientation, western winds give rise to
both sea and swell waves, while easterly winds have no
significant fetch and mainly give rise to sea waves. The
main longshore drift flows south-eastward.
During the past 40 years, sea level at the Cadiz coast
did not show a single clear trend (Marcos and Tsimplis,
2008; Marcos et al., 2011) and winter storms represent
the main coastal hazard in the investigated littoral since
they largely affect tourism activities, e.g. the sun, sea
and sand (3S) market. They result in major damage
to recreational and protection structures causing sand
starvation and associated reduction in beach width and
scenic value (Anfuso and Gracia, 2005; Meyer-Arendt,
2011).
The area is characterized by extensive sandy beaches
that, in the past decade, have undergone erosion with
locally recorded values greater than 1 m year−1, essen-
tially associated with the impact of storm events (Mu˜
noz
and Enriquez, 1998; Reyes et al., 1999; Benavente et al.,
2002; Anfuso et al., 2007; Rangel-Buitrago and Anfuso,
2011a). In order to balance coastal retreat, in the past
three decades, more than 600 fills and refills have been
Figure 1. Study area with location of the REDCOS buoy n°1316 operated by Puertos del Estado (Spanish Ministry of Public Works).
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

2144 N. RANGEL-BUITRAGO AND G. ANFUSO
Table I. Characteristics of the five storm classes (modified from Rangel-Buitrago and Anfuso 2011a): range, frequency (number
of cases and percentages), mean values (X) of significant wave height and period, storm duration and storm power index per
class.
CLASS Range (m2h) Frequency Wave height Period Duration Storm power
N% X (m) X (s) X (h) X (m2h)
I–Weak <515 74 57 3.46 7.02 20 242.4
II – Moderate 516–1225 41 31 4.46 7.52 40.2 785.3
III – Significant 1226–2537 11 8 4.56 7.5 89.6 1850
IV – Severe 2538–5167 2 3 5.06 8.6 139.3 3311
V – Extreme >5167 1 1 7.8 9.7 89 5414
performed along the Spanish coast. Investment in beach
nourishment during the 1990s along the Atlantic coast of
Andalusia was US $37M, representing the cost of inject-
ing circa 13 ×106m3of sediment (Mu˜
noz et al., 2001).
2. Methodology
2.1. Wave Data
Real-time wave measurements along the Spanish lit-
toral are available from the coastal buoy network oper-
ated by Puertos del Estado (Spanish Ministry of Pub-
lic Works). Specifically, this work was based on the
analysis of data obtained from the scalar buoy n°
1316 (36.50 °N; 6.33 °W), a waverider-datawell instru-
ment which is located at a water deep of 21 m, in front
of Cadiz city (Figure 1). Data were logged at 1 h inter-
vals and included a time series of 163 237 wave height
and period records acquired between 1983 and 2010. In
this study, average annual and monthly winter values of
significant wave height (Hs)andthe99
th percentile of
Hs(Hs99) were used for the characterization of wave cli-
mate and definition and description of storms. The Hs99
was used because it is associated with high energy events
(Almeida et al., 2011; BACC Author Team, 2008).
According to previous studies carried out by Anfuso
et al. (2007) and Rangel-Buitrago and Anfuso (2011a),
winter season was defined as the October to March
period. The investigated dataset presented a length of
27 years that, according to Komar and Allan (2008),
represents a sufficient time period to analyse wave height
trend, presence of climate-controlled cycles and wave
height annual variations due to climate events.
2.2. Storm definition and characterization
Storminess is a key issue in coastal erosion and climate
change studies (Li et al., 2011). Coastal engineers nor-
mally use the number of storm events or the amount
of hours during which a certain wave height thresh-
old is exceed (Goda, 1988; Lemm, et al., 1999; Barr,
2004). Meteorologists use the number of storm sys-
tems, characterized by wind speed values grater than
a certain threshold or central air pressure values lower
than 1000 hPa (Schmith et al., 1998; Zielinski, 2002;
Keim et al., 2004). Other researchers use the number of
events or the number of hours during which a specific
water level threshold is exceed (Eliot and Clarke, 1986;
Zhang et al., 2000; Bromirski et al., 2003; Pattiaratchi
and Elliot, 2008; Phillips, 2008).
In this work, a storm is defined as a climatic event
during which the significant wave height (Hs) exceeds a
threshold over a minimum, specific time, so the following
three criteria have to meet:
•Wave height Hs≥2.5 m. This threshold reflects the
deep-water wave height at which erosion affected
Cadiz beaches (Plomaritis et al., 2009, 2010; Rangel-
Buitrago and Anfuso, 2011a).
•The minimum storm duration was set at 12 h, in this
way the storm affected the coast at least during a
complete tidal cycle.
•The inter-storm period was set at 1 d in order to create
de-clustered, independent sets of storms (Morton et al.,
1997; Dorsch et al., 2008).
The energy content for each storm was calculated
according to the Storm Power Index (Dolan and Davis,
1992), which is a valuable indicator of storm strength and
associated shoreline erosion (Li et al., 2011), using the
formula:
Hs2td(1)
Where Hsis the maximum significant wave height
in meters and tdis the storm duration in hours. Once
storms were recognized, they were categorized by means
of the natural breaks function analysis (Jenks and Caspall,
1971), into five classes from class I (weak) to class V
(extreme events, limits presented in Table I).
2.3. Storm recurrence period
Frequency analysis was applied to estimate extreme
waves and storm power recurrence periods by means of
the Generalized Extreme Value (GEV) distribution; it is a
family of continuous probability distributions developed
within extreme value theory to combine the Gumbel,
Fr´
echet and Weibull families also known as types I, II
and III extreme value distributions (Coles, 2001). The
GEV distribution was calculated according to:
F(x) =exp{− exp[−(x −u)/α]1/k}(2)
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

WINTER WAVE CLIMATE, STORMS AND REGIONAL CYCLES 2145
Where xis the random variable and u,αand k
are location, scale and shape parameters, respectively,
that should be estimated for each sample. The equation
reduces to type I (or Gumbel, for k=0), type III (or
Weibull, for k>0) or type II (for k<0) distribution.
The equation for the type I (Gumbel) distribution is:
F(x) =exp{− exp[−(x −u)/α]}(3)
The methods of parameter estimation for each distri-
bution are also discussed in detail in Rao and Hamed
(2000). In this study, the method of moments and maxi-
mum likelihood was applied to estimate the distribution
of investigated parameters. The root mean square error
was then used to select the appropriate distribution. To
investigate the recurrence period, wave and storm data
were plotted with the Gringorten (1963) plotting position
formula and the Gumbel distribution was fitted to the
data. The distribution function is given by the formula:
TR =N+0.12/m−0.44 (4)
Where Nis the number of annual maximum obser-
vations and mis the rank of extreme waves or storm
power values from the lowest to the highest observa-
tion. According to the Gumbel distribution, the expected
significant values for a selected return period can be esti-
mated as follows (reduced value):
TRp =−ln[−ln(1−1/TR)](5)
2.4. Northern Hemisphere teleconnection patterns
The term Regional Cycle and Teleconnection Pattern
are terms referred to a recurring and persistent, large
scale pattern of pressure and circulation anomalies that
span vast geographical areas (Hatzaki et al., 2006).
Teleconnection patterns are also referred to as preferred
modes of low-frequency (or long time scale) variability
of the atmospheric circulation with geographically fixed
centres of action.
The NOAA Climate Dynamics Research Centre has
shown the existence of several teleconnection patterns
influencing the European region (including the area
investigated in this study):
•North Atlantic Oscillation – NAO (Wallace and Gut-
zler, 1981; Hurrel, 1995; Barnston and Livezey, 1987).
•East Atlantic Pattern – EA (Esbensen, 1984; Wallace
and Gutzler, 1981).
•East Atlantic/Western Russia – EA/WR (Barnston and
Livezey, 1987)
•Scandinavia – SCAND (Barnston and Livezey, 1987).
•Polar/Eurasia – POL (Barnston and Livezey, 1987).
•East Pacific – North Pacific Pattern – EP–NP (Bell
and Janowiak, 1995).
•Pacific/North American Pattern – PNA (Barnston and
Livezey, 1987; Chen and van den Doo, 2003).
•Arctic Oscillation – AO (Zhou et al., 2001; Higgins
et al., 2002).
Previous teleconnection patterns were recognized as
the most important characterizing parameters of the
European climate system, the NAO apparently being
the most important. A link between these teleconnection
patterns and seasonal, interannual and decadal variations
in weather, waves and storminess for the European
Atlantic coastlines was established by different authors
(Wang et al., 2008; Lozano et al., 2004; Hurrel and
Deser, 2009; Dodet et al., 2010). Likewise, changes in
coastal evolution during the second half of the 20th
century were also connected to modulations of the NAO
Index (Lozano et al., 2004; Esteves et al., 2011; Thomas
et al., 2011; O’Connor et al., 2011).
In this study, distributions of average winter monthly
and annual values of NAO, EA, EATL/WRUS, SCAND,
Polar/Eurasia, EP-NP, PNA and AO from 1983 to 2010
were compared with winter wave height data and storm
classes by means of multiple and independent regression
analysis.
3. Results
3.1. Wave climate
Data showed a clear pattern of cyclic variations of aver-
age monthly values of significant wave height (Hs)and
99th percentile of Hs(Hs99 – Figure 2). Waves were usu-
ally low (Hs<0.8m,Hs99 <1.8 m) in May to August
period (late spring to summer), reaching minimum val-
ues in August (Hs=0.6m,Hs99 =1.22 m). During the
winter season, waves rapidly increased in height, reaching
peak values (Hs=1.2mandHs99 =3.1) in December
to January period.
Trends of average monthly and annual values of Hs
and Hs99 during the investigated period (1983–2010)
and associated linear regression analyses, were plotted
in Figure 2; data showed a small decrease in monthly
(Avg: 1; r2:0.018; P: 0.118; trend: −0.0011) and annual
(Avg: 1; r2:0.085; P: 0.14; trend: −0.008) Hsvalues,
whereas Hs99 remained constant. However, low recorded
values of Pearson coefficient revealed that these trends
are not statistically significant (Pmonthly: 0.118; P
annual: 0.14). Similar results were obtained using the
Mann–Kendall trend test and the Wilcoxon rank-sum
test commonly used in such kind of studies (Carter and
Draper, 1988; Bacon and Carter, 1991; Allan and Komar,
2000).
Considering the 27 investigated winters, a great vari-
ability of average monthly wave height was observed.
Low Hsvalues (<0.8 m) were recorded in winter sea-
sons within the 1990/1–1994/5, 1998/9– 2001/2 and
2004/5–2007/8 periods; high Hsvalues (≈1.3m)were
recorded during the 1995/6, 1996/7, 1997/8, 2002/3,
2003/4 and 2009/10 winter season periods. Concerning
extreme wave conditions, the highest monthly value was
7.8 m (recorded December 1989) and the average value
of extreme wave height was 4.6 m.
Figure 2 evidenced a quasi-periodic 3–4 year beha-
viour in the recurrence of high wave height values. A
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

2146 N. RANGEL-BUITRAGO AND G. ANFUSO
Figure 2. Winter monthly and annual Hsand Hs99 for the 1983– 2009 period.
spectral analysis of time series of extreme waves, based
on the Fourier transformation (Boashash, 2003), indicated
a cyclic trend of 3 years. The probability density function
for the period of recurrence and occurrence probability
of maximum annual Hswas obtained (Figure 3) applying
the GEV proposed by Jenkinson (1955). The probability
of occurrence of 3.0–3.5 m waves ranged from the 86 to
96%, meanwhile the probability of occurrence of waves
higher than 6 m did not reach 6%. The recurrence periods
for 4 and 8 m waves were 5 and 50 years respectively.
3.2. Storm characteristics and trend
One hundred and nineteen storms were recorded during
the investigated 27 year time span. Classes I (weak) and
II (moderate) accounted for 55 and 33%, respectively, of
records. These values were very close to those obtained
by Dolan and Davis (1992), Moritz and Moritz (2006),
Mendoza and Jimenez (2008) and Rangel-Buitrago and
Anfuso (2011a) in United States and Spain. Class III
(significant), constituted 8% of the records and classes
IV (severe) and V (extreme) accounted for 3 and 1%,
respectively (Table I).
Associated average wave height and storm duration
values presented important variations (Table I) and aver-
age wave period ranged from 7.0 (class I) to 9.7 s
(class V). Storm power values were larger than the
ones proposed by Dolan and Davis (1992) because of
the major threshold of storm wave height selected, and
apparent longer duration of investigated storms. Dealing
with monthly distribution, classes I and II events were
observed over the entire winter season, class III from
November to March and classes IV and V only in Decem-
ber and January, January being the stormiest month.
The distribution of the most commonly used storminess
indices over the 27 years is shown in Figure 4. Consid-
ering the number of storms per year (Figure 4), stormy
winter seasons were 2009/10 (12 events), 1995/6 and
2002/3 (with 10 events each) and 1993/4 (8 events); no
storms at all were recorded during 1994/5. Despite some
anomalous peaks, the 1983/4–1986/7, 1990/1– 1992/3,
1998/9–2001/2 and 2003/4– 2008/9 winter periods pre-
sented the lowest average number of storms (Figure 4);
the 3 year moving average did not show a clear trend in
storm occurrence whereas indicated a 6–7 years cyclic
behaviour with storm peaks observed in 1989/90, 1995/6,
2002/3 and 2009/10 winter seasons.
Patterns of variability of storms duration and Storm
Power Index were very similar to that found for the
number of storms (Figure 4). This was because, in
general, stormy winter seasons presented a great number
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

WINTER WAVE CLIMATE, STORMS AND REGIONAL CYCLES 2147
Figure 3. Wave Recurrence and probability. Annual maximum wave height plotted versus the reduced value from the Gumbel distribution using
the Gringoten Plotting position and wave occurrence probability plotted versus the annual maximum wave height value.
of storms which gave many hours under storm conditions
and therefore an elevated storm power.
Linear regressions analysis, Mann–Kendall trend tests
and Wilcoxon rank-sum tests did not show a clear
trend of storminess indices (i.e. number of storms: Avg:
4.85; r2:0.040; trend: 0.082) although they indicated a
6–7 year period (return periods: r2: 0.97 and occurrence
probability: r2: 0.95 – Figures 4 and 5). In order to
further investigate storm periodicity, the return period and
occurrence probability of each storm class were estimated
using the method of the maximum likelihood based
on the Gumbel distribution of annual maximum storm
power values. The probability plot for each storm power
value was represented in Figure 5 and derived using the
Gringoten (1963) plotting position analysis suggested by
Cook (2004) and Goel et al. (2004).
4. Discussion
4.1. Wave climate and storm trend
Results obtained confirmed the clear seasonal nature of
wave height in Cadiz area as previously observed by
Men´
endez et al. (2004), Anfuso and Gracia (2005) and
Rangel-Buitrago and Anfuso (2011a).
The time series of average and extreme monthly values
of wave height presented similar behaviour. The lack
of a clear trend of both features disappointed with
observations carried out by several authors who suggested
an increase in mean significant wave height in many
Oceans around the World and, more closely, on the
Northern part of the North Atlantic Ocean. Nevertheless,
results obtained confirmed observations carried out at
the mid and southern latitudes of the North Atlantic
Ocean, by Dolan et al. (1989) and WASA (1998), and
at regional scale, e.g. Cadiz Gulf, by Men´
endez et al.
(2004), Ferreira et al. (2009) and Almeida et al. (2011).
Concerning the distribution of stormy years, it was
quite close to the one presented at local and regional
scales by previously mentioned authors. Specifically,
Rodriguez et al. (2003) used wind records and individu-
ated eight stormy years in Huelva during the 1962–1999
period; two out of three of the recorded storminess sea-
sons after 1983 (e.g. 1989/90 and 1995/6) corresponded
with the two years that recorded the highest values of
Storm Power Index.
Almeida et al. (2011) used HIPOCAS and offshore
wave buoy data recorded in Faro, in the Algarve coast
(Southern Portugal), and affirmed that the storminess
years were 1987, 1989, 1995, 1996, 1997, 2002, 2003
and 2009. All these except one, i.e. 2003, confirmed
this study’s observations since they corresponded with
years characterized by high Storm Power Index values, a
great number of storms and high values of storm duration
(Figure 4). Similarities to previous studies indicate a
general homogeneity of wave climate in the Cadiz Gulf. It
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

2148 N. RANGEL-BUITRAGO AND G. ANFUSO
(a)
(b)
(c)
Figure 4. Distribution of the number of storms, duration and sum of storm power per winter year with their respective linear regression and the
3 year moving average. Average values (Avg), Root mean square (r2) and trends also presented.
is broadly exposed to storms approaching from the third
and fourth quadrants, as observed by Rangel-Buitrago
and Anfuso (2011b), who analysed HIPOCAS datasets
at five locations along Andalusia’s Atlantic littoral.
Comparing obtained data in this study with observa-
tions carried out in Northern Europe, a certain correspon-
dence was evidenced with data presented by O’Connor
et al. (2011). They analysed gale-day frequency in North-
ern Ireland finding a peak in storm generation in the
early 1990s, before an apparently decreasing until 2009.
The most energetic conditions recorded in this work
(1995/6, 1996/7, 2002/3 and 2009/10) matched the
extreme weather conditions recorded over the same peri-
ods in Wales (Phillips, 2008; Phillips and Crisp, 2010;
Thomas et al., 2011), Lithuania (Dailidiene et al., 2011;
Kelpsaite et al., 2011), Estonia and Southern Gulf of Fin-
land (Suursaar, 2010). These authors reported significant
changes in wind direction, high wind speeds, mildest tem-
perature records and an increase in damage to coastal
structures (because of storm impacts). In addition, fair
weather conditions recorded in Cadiz area during 1998
were similar to the observations carried out in Eng-
land and reported by Environmental Scientist (2000) that
observed as 1998 and the 1990s respectively were the
third warmest year and the warmest decade on records in
the UK.
Concerning storm occurrence probability in the inves-
tigated area, it was 96% for class I (i.e. almost one
event per year) to 3% for class V. The return period
for class V was 25 years and ranged from 6 to 8 years
for classes III and IV storms, e.g. significant and severe
events (return periods: r2: 0.97 and occurrence proba-
bility: r2: 0.95 – Figure 5). Such periodicity was simi-
lar to the 6–7 year periodicity proposed for Cadiz and
Huelva respectively by Mu˜
noz and Enriquez (1998) and
Rodriguez et al. (2003) and to the 7–8 year periodic-
ity recorded by Ferreira et al. (2009) and Almeida et al.
(2011) in Faro (Southern Portugal) and in the southern
part of North Atlantic by WASA (1998) and Matulla et al.
(2007).
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

WINTER WAVE CLIMATE, STORMS AND REGIONAL CYCLES 2149
Figure 5. Storm recurrence and probability for the different storm classes in Cadiz littoral. Annual maximum storm power plotted versus the
reduced value from the Gumbel distribution using the Gringoten plotting position and storm occurrence probability plotted versus the annual
maximum storm power.
Lastly, classes I and II showed a period of recurrence
ranging from 1 to 3 years, similar to the 2 –3 year
recurrence period for minor storm events detected in
Cadiz and Huelva respectively by Mu˜
noz and Enriquez
(1998) and Rodriguez et al. (2003).
4.2. Storms and Northern Hemisphere teleconnection
patterns
A series of linear regression analyses and Pearson corre-
lation tests were performed considering the annual and
monthly wave data (Hsand Hs99), storminess indices
and teleconnection patterns. The analysis of annual and
monthly datasets gave very different result, the latter pre-
senting clearer trends (Table II).
Multiple linear regression analysis between wave data,
storminess indices and the teleconnection patterns indi-
cated that approximately 40% of the change in monthly
wave data and all used storminess indices related to
changes in the eight teleconnection patterns used in this
work. Drivers of change being the AO, with 21.45%, and
NAO, with the 19.65% (Table II); both being important
controllers on short- and long-term climatic variability
in North Atlantic Ocean, Europe and the Mediterranean
basin (T¨
urke¸s and Erlat, 2008).
Specifically, the AO is an atmospheric circulation
pattern reflecting the non-seasonal sea level pressure
variations north of 20 °N latitude and varies over time
with no particular periodicity; it is characterized by
pressure anomalies of opposite sign, one located in
the Arctic and the other centred at 37–45 °N latitude
(Thompson and Wallace, 1998). The AO is believed to
be causally related to, and thus partially predictive of,
weather patterns in locations many thousands of miles
away, including many major population centres of Europe
and North America.
In the same way, negative AO values play an important
role in determining extreme conditions such as frozen
precipitations, strong winds and extreme weather events
in general over the Northern Hemisphere and in particular
over the North Eastern United States, the Mediterranean
area and China (Thompson and Wallace, 2000; Higgins
et al., 2002; Wettstein and Mearns, 2002; Xoplaki, 2002;
T¨
urke¸s and Erlat, 2008; Mao et al., 2011).
The NAO is defined as the difference in normalized
sea level pressure computed between a station in the
Azores (Ponta Delgada) or Southern Europe (e.g., Gibral-
tar, Lisbon) and another station in Iceland (Stykkishol-
mur). Negative NAO values give rise to changes in the
surface westerly winds across the North Atlantic that
eventually affects The Azores and the western European
coasts (Andrade et al., 2008). In this respect, severity of
winters in northern and western Europe (WMO, 1995;
Esteves et al., 2011; Thomas et al., 2011), temperature
anomalies (Hurrell, 1995; Trigo et al., 2002), intensity
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

2150 N. RANGEL-BUITRAGO AND G. ANFUSO
Table II. Pearson correlation values for winter annual (a) and monthly (b) data.
a) Annual data correlations
Teleconnection
pattern
Hs
mean
monthly
Hs99
percentile
Number
of
storms
Duration
of
storms
Avg
duration
Max
duration
Sum
SPI
Avg
SPI
Max
SPI
North Atlantic Correlation −0.48 −0.59 −0.51 −0.50 −0.33 −0.52 −0.42 −0.25 −0.23
Oscillation P-value 0.01 0.00 0.01 0.01 0.09 0.01 0.03 0.21 0.25
(NAO) r20.23 0.35 0.26 0.25 0.11 0.27 0.17 0.06 0.05
East Atlantic Correlation 0.33 0.34 0.32 0.31 0.25 0.24 0.32 0.31 0.30
(EA) P-value 0.09 0.08 0.11 0.11 0.20 0.22 0.10 0.12 0.13
r20.11 0.12 0.10 0.10 0.06 0.06 0.10 0.10 0.09
East Correlation 0.13 0.10 −0.13 −0.15 −0.11 −0.20 −0.11 −0.08 −0.11
Atlantic/Western P-value 0.52 0.62 0.53 0.47 0.59 0.33 0.59 0.70 0.59
Russia (EA/WR) r20.02 0.01 0.02 0.02 0.01 0.04 0.01 0.01 0.01
Scandinavia Correlation 0.45 0.48 0.41 0.38 0.35 0.44 0.23 0.13 0.06
(SCAND) P-value 0.02 0.01 0.03 0.05 0.07 0.02 0.26 0.53 0.78
r20.21 0.23 0.17 0.14 0.13 0.20 0.05 0.02 0.00
Polar/Eurasia Correlation 0.11 0.06 −0.08 −0.10 −0.09 −0.06 −0.07 −0.07 −0.09
Pattern P value 0.58 0.76 0.70 0.63 0.67 0.78 0.72 0.72 0.64
(POL) r20.01 0.00 0.01 0.01 0.01 0.00 0.01 0.01 0.01
Arctic Correlation −0.43 −0.47 −0.52 −0.51 −0.34 −0.46 −0.39 −0.22 −0.16
Oscillation P-value 0.03 0.01 0.01 0.01 0.09 0.02 0.04 0.26 0.42
(AO) r20.18 0.22 0.27 0.26 0.11 0.21 0.15 0.05 0.03
Pacific/North Correlation −0.09 −0.13 0.11 −0.03 −0.16 −0.12 −0.06 −0.14 −0.19
American P-value 0.67 0.52 0.57 0.89 0.44 0.54 0.77 0.48 0.33
Pattern (PNA) r20.01 0.02 0.01 0.00 0.02 0.01 0.00 0.02 0.04
East Pacific/ Correlation −0.07 −0.12 0.11 −0.03 −0.16 −0.12 −0.07 −0.16 −0.22
North Pacific P-value 0.72 0.56 0.60 0.88 0.42 0.55 0.72 0.43 0.28
Pattern (EP/NP) r20.01 0.01 0.01 0.00 0.03 0.01 0.01 0.03 0.05
b) Monthly data correlations
North Atlantic Correlation −0.43 −0.46 −0.49 −0.40 −0.36 −0.38 −0.35 −0.33 −0.33
Oscillation P-value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
(NAO) r20.18 0.22 0.24 0.16 0.13 0.14 0.12 0.11 0.11
East Correlation 0.14 0.07 0.06 0.08 0.13 0.11 0.09 0.17 0.16
Atlantic P-value 0.11 0.42 0.49 0.35 0.13 0.21 0.29 0.06 0.07
(EA) r20.02 0.00 0.00 0.01 0.02 0.01 0.01 0.03 0.02
East Atlantic/ Correlation 0.16 0.07 0.07 −0.01 0.00 0.01 −0.02 0.00 0.00
Western Russia P-value 0.06 0.44 0.41 0.94 0.96 0.91 0.82 0.99 0.96
(EA/WR) r20.03 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00
Scandinavia Correlation 0.22 0.20 0.18 0.21 0.17 0.22 0.17 0.17 0.12
(SCAND) P-value 0.01 0.02 0.03 0.02 0.05 0.01 0.04 0.04 0.18
r20.05 0.04 0.03 0.04 0.03 0.05 0.03 0.03 0.01
Polar/Eurasia Correlation 0.00 0.03 −0.04 −0.07 −0.06 −0.03 −0.05 −0.02 −0.03
Pattern P-value 0.96 0.73 0.68 0.42 0.50 0.70 0.53 0.81 0.76
(POL) r20.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Arctic Correlation −0.47 −0.43 −0.51 −0.48 −0.43 −0.44 −0.40 −0.39 −0.37
Oscillation P-value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
(AO) r20.22 0.19 0.26 0.23 0.18 0.20 0.16 0.15 0.14
Pacific/North Correlation 0.07 0.04 0.11 0.05 0.01 0.02 0.05 0.05 0.07
American P-value 0.40 0.63 0.23 0.58 0.94 0.81 0.57 0.59 0.41
Pattern (PNA) r20.01 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.01
East Pacific/ Correlation 0.00 0.00 0.06 −0.01 −0.05 −0.03 0.00 0.00 0.03
North Pacific P-value 0.98 0.98 0.46 0.95 0.60 0.70 0.98 0.96 0.76
Pattern (EP/NP) r20.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
In bold significant correlation values.
and frequency of winter daily precipitation and storms
over the Iberian Peninsula (Mu˜
noz-D´
ıaz and Rodrigo
(2003); Gallego et al., 2005; Almeida et al., 2011) are
partially influenced by this teleconnection pattern.
The relative importance of the NAO index in the devel-
opment of extreme events in the Northern Hemisphere
was confirmed in this study and it was in agreement
with several authors such as: Matulla et al. (2007) that
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

WINTER WAVE CLIMATE, STORMS AND REGIONAL CYCLES 2151
Figure 6. Monthly values of number of storms, storm durations, North Atlantic Oscillation (NAO) and Arctic Oscillation (AO). The dashed
polygons highlight ‘stormy’ years. NAO phases: SN, strong negative; N, negative; NT, neutral; P, positive; SP, strong positive.
showed as the NAO index was not particularly help-
ful in the determination of storm conditions in central
Europe and Andrade et al. (2008) that explained the ‘par-
tial responsibility’ of the NAO index over storminess and
their temporal distribution over the Azores region from
1865 to 2009. Previous assumptions are in accordance
with the observations of Allan et al. (2009), which found
a poor correlation between the ‘Gibraltar – SW Iceland
NAO’ and the distribution of severe winter storms over
the British Isles and Almeida et al. (2011), which indi-
cated that NAO index variations were not exclusively
responsible of winter wave and storm distribution along
the South coast of Portugal.
Considering that the AO and the NAO were the
principal teleconnection patterns involved in changes of
storminess patterns over the study area (Table II), their
relationships with number and duration of storms and
five step moving average were investigated (Figure 6). In
order to facilitate the analysis, NAO data were divided
in five phases (strong negative, negative, neutral, positive
and strong positive) according to Pinto et al. (2009) and
Donat et al. (2010). Negative phases of NAO and AO
presented a certain level of correlation with the different
storminess indices, and as an example the number (C:
−0.49, r2: 0.24) and duration of storms (C: −0.40,
r2: 0.16) were presented in Figure 6 and Table II. It is
interesting to remark that a great number of storms, larger
storm durations and higher values of Storm Power Index
were only observed when neutral to strong negative NAO
and AO phases occurred at the same time (89 storms
and 3355 h – Figure 6) and/or when occurred an abrupt
change of NAO and AO phases, i.e. they moved from
a positive phase to a negative phase without passing
through a neutral phase. The behaviour of NAO and
AO indexes was described by Thompson and Wallace
(1998) who observed as the AO resembles the NAO in
numerous respects. However, the AO’s primary centre
of action covers more of the Arctic, giving it a more
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

2152 N. RANGEL-BUITRAGO AND G. ANFUSO
Figure 7. Pressure and Geopotential height anomalies at 500 mb over the 2009/10, 1995/96 and 1963/64 winter seasons (from NCEP-NCAR
reanalysis). Great storminess and extremely coldest conditions were recorded during the aforementioned winter seasons.
zonal symmetrical appearance. Hence, although the AO
and the NAO patterns resemble each other, there is a clear
distinction that could play a guiding role in determining
the physical mechanisms that control the variability of the
Northern Hemisphere climate (Wallace, 2000). Rogers
and McHugh (2002) examined whether the NAO and
the AO were inseparable spatial modes of atmospheric
circulation in the Northern Hemisphere: the analysis
of the spring, summer and autumn sea level pressure
fields revealed that the NAO and the AO-liken patterns
occurred as separate regional teleconnections forming the
first and second principal components, respectively.
Specifically, in the investigated area, 1995/1996 was
one of the most storminess years over past decades
(Figure 6) and the great number of recorded storms
gave rise to important economic losses and damages in
the Gulf of Cadiz (Ballesta et al., 1998; Reyes et al.,
1999; Anfuso and Gracia, 2005). The 2009/10 winter
season was the stormiest season during the investigated
time span, with 12 storm events equivalent to 614 h
(26 d). The exceptional nature of the 2009/2010 winter
under different climatic aspects was confirmed by several
studies. In this sense, Cattiaux et al. (2010) observed as
it was one of the coldest winter seasons since 1949 and
Cohen et al. (2010) recorded extreme cold temperatures,
snow precipitations and storms in most of the major
population centres of the industrialized countries of the
Northern Hemisphere.
Climatic anomalies were associated with an extreme
persistence of negative phases of AO and NAO indices
and an exceptional Northern Hemisphere mean atmo-
spheric circulation episode (Wang et al. 2010) reflected
by the distribution of the geopotential height anoma-
lies at 500 mb (Figure 7). This parameter exhibited a
strong zonal hemispheric pattern, with anomalously high
pressures over the pole during 2009/2010 (Figure 7).
Such a distribution favoured the development of neg-
ative phases for the AO and NAO at the same time
(Thompson and Wallace, 1998; L’Heureux et al., 2010).
Similar results were recorded by Cohen et al. (2010) and
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

WINTER WAVE CLIMATE, STORMS AND REGIONAL CYCLES 2153
1980-1981
1981-1982
1982-1983
1983-1984
1984-1985
1985-1986
1986-1987
1987-1988
1988-1989
1989-1990
1990-1991
1991-1992
1992-1993
1993-1994
1994-1995
1995-1996
1996-1997
1997-1998
1998-1999
1999-2000
2000-2001
2001-2002
2002-2003
2003-2004
2004-2005
2005-2006
2006-2007
2007-2008
2008-2009
2009-2010
YEAR
Avg: 4.37 R2:0.040 Trend: 0.010
Avg: 156.1 R2:0.028 Trend: 6.70 Avg: 185.2 R2:0.006 Trend: 4.5 Avg: 176 R2:0.36 Trend: 41.2
Avg: 4.6 R2:0.30 Trend: 0.06 Avg: 5.5 R2:0.10 Trend: 0.51
0
4
8
12
Number of Storms
0
0
1
−1
2
S.P
P
N.T
N
S.N
−2
200
400
600
Storm Duration (hr)
North Atlantic
Oscillation (NAO)
0
1
−1
2
−2
Artic Oscillation (AO)
Linear Regression
3 Step-Moving Average
Figure 8. Yearly values of number of storms, storm duration, North Atlantic Oscillation (NAO) and Arctic Oscillation (AO) phases. Average
(Avg), Root mean square (r2), lineal regression and the three step moving average also presented. NAO phases: SN, strong negative; N, negative;
NT, neutral; P, positive; SP, strong positive.
L’Heureux et al. (2010) which highlighted that the nega-
tive AO phase of December 2009 was a record result of
an unusual occurrence of two troposphere–stratosphere
coupling events (NAO–AO) that occurred more rapidly
than usual and in quick succession. Similar patterns
of circulation and AO–NAO values (Figures 6 and 7)
were recorded during the 2005/06, 1995/96, 1987/88
and 1963/64 winter seasons. These years were character-
ized by a high number of storms, great storm durations
and a great amount of reported coastal damage (Rangel-
Buitrago and Anfuso, 2011a).
Lastly, considering the decadal behaviour of annual
values of AO and NAO indices, positive values were
recorded during the past three decades (Figure 8). The
1980s were characterized by the general predominance
of positive NAO phases, which specifically ranged from
neutral to strong positive phases. The AO presented clear
negative phases from 1980 to 1987 and positive ones
from 1987 to 1994. Considering storminess, the 1980s
can be classified as a period of ‘calm’ due to the low
recorded condition in wave heights and storminess (35
storms).
In the 1990s, NAO and AO indices presented the
predominance of positive phases, 1995 being the only
year that recorded strong negative phases. During the
1990s, the teleconnection behaviour patterns and gen-
erally low storm activity, coincided with analyses per-
formed by WASA (1998) in the northeast Atlantic coast,
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

2154 N. RANGEL-BUITRAGO AND G. ANFUSO
Ferreira et al. (2009) and Almeida et al. (2011) in Portu-
gal and Rangel-Buitrago and Anfuso (2011a) in the Gulf
of Cadiz. Finally, 2000–2009 was characterized by cyclic
behaviour of both NAO and AO indices, as well as wave
climates and storminess that showed extreme conditions
between 2002 and 2003 and low energy conditions from
2004 to 2007.
5. Conclusions
Coastal erosion and flooding, increased by climatic
change-related processes, represent a great threat to
human activities and settlements developed along the
World’s coastlines especially because human pressure
and occupation in such environments have been grow-
ing considerably in recent decades. The Cadiz coastal
area includes different human settlements and natural
environments, essentially large sandy beaches of great
economic interest because of tourism-related activities.
In past decades, beach reduction and loss of attractive-
ness were related to elevated coastal retreat rates basically
linked to the impact of winter storms since sea level varia-
tions did not show a clear trend. Large amounts of money
were spent counteracting coastal retreat by using beach
nourishment works and hard engineering structures.
An important step towards reducing and preventing
beach erosion and infrastructure damage is to understand
and characterize wave climate and coastal processes, e.g.
past and future trends. Understanding seasonal and annual
wave climate behaviour is crucial in erosion prevention,
as well as assessing beach response, cyclic variance and
evolution. These are key issues in distinguishing real
erosive trends from natural alternating erosion/accretion
phases. Consequently, the applied methodology enables
the characterization of wave climate, e.g. seasonal and
annual periodicity, and storm events according to energy.
This makes it simple and objective, while data can be
easily compared with similar datasets from elsewhere.
In particular, wave climate and storm classes’ distri-
bution evidenced a clear seasonal trend in Cadiz area.
Classes I and II (weak and moderate events) were
observed along the entire winter season which covered
the October to March period, class III events character-
ized the November to March period and classes IV and
V (severe and extreme) events were recorded only in
December and January, with January being the stormiest
month. Wave climate and storminess during the 27 year
period did not show a clear trend. This contradicted the
forecast general increase in the number and intensity of
storms in many oceans around the World and specifi-
cally the Northern Hemisphere, due to climate change.
Storminess presented a cyclic behaviour which occur-
rence probability varied from 96% for class I to 3% for
class V. The return period for class V was 25 years and
ranged from 6 to 8 year for storms belonging to classes III
and IV, e.g. significant and severe events. Observations
broadly confirmed results recorded at mid and low lati-
tudes of the Northern Hemisphere and at other locations
of the Cadiz Gulf, in this sense pointing to a broad homo-
geneity of regional wave climate.
Efforts in the direction of storm characterization and
prediction should also point to the determination of
the general climatic conditions for their formation, e.g.
the comprehension of general atmospheric circulation
systems linked to the teleconnection patterns. In the
Cadiz area, storm distribution appeared linked to negative
phases of AO and NAO indices. Even if further investiga-
tion is needed, it was highlighted that a great number of
storms, larger storm duration and higher values of Storm
Power Index were only observed when neutral to strong
negative AO and NAO phases there were at the same
time and/or when occurred an abrupt change of AO and
NAO phases, i.e. they moved from a positive phase to a
negative phase without passing through a neutral phase.
Hence, accurate short-term predictions of AO and NAO
indices might represent valuable predictive techniques to
forecast storm events, e.g. establishing early and warn-
ing systems in order to reduce the impacts of energetic
erosive events.
Acknowledgements
This work is a contribution to the RESISTE Research
Project (CGL2008-00458/BTE, supported by the Spanish
Ministry of Science and Technology and by European
Funds for Regional Development – F.E.D.E.R.) and to
the Andalusia P.A.I. Research Group no. RNM-328.
Thanks to Puerto del Estado (Spanish Ministry of Public
Works) for offshore wave data records. This work has
been partially developed at the Centro Andaluz de
Ciencia y Tecnolog´
ıa Marinas (CACYTMAR), Puerto
Real (Cadiz, Spain).
References
Allan JC, Komar PD. 2000. Are ocean wave heights increasing in
the eastern North Pacific?. EOS, Transactions of the America
Geophysical Union 47: 561 –567.
Allan R, Tett S, Alexander L. 2009. Fluctuations in autumn – winter
severe storms over the British Isles: 1920 to present. International
Journal of Climatology 29: 357 –371.
Almeida LP. Ferreira O Vousdouskas MI. Dodet G. 2011. Historical
variation and trends in storminess along the Portuguese Southcoast.
Natural Hazards Earth Systems. Science 11: 2407–2417.
Andrade C, Trigo RM, Freitas MC, Gallego MC, Borges P, Ramos
AM. 2008. Comparing historic records of storm frequency and the
North Atlantic Oscillation (NAO) chronology for the Azores region.
The Holocene 15(5): 745 –754.
Anfuso G, Gracia FJ. 2005. Morphodynamic characteristics and short
term evolution of a coastal sector in SW Spain: implications for
coastal erosion management. Journal of Coastal Research 21(6):
1139– 1153.
Anfuso G, Dominguez L, Gracia FJ. 2007. Short and medium-term
evolution of a coastal sector in Cadiz, SW Spain. Catena 70:
229– 242.
BACC Author Team. 2008. Assessment of Climate Change for the
Baltic Sea Basin. Springer-Verlag: Berlin.
Bacon S, Carter DJT. 1991. Wave climate changes in the North
Atlantic and North Sea. International Journal of Climatology 11:
545– 558.
Ballesta M, Morales J, Acosta EA. 1998. Efecto erosivo de los
temporales del invierno 1995– 1996 sobre la playa de Mazagon
(Huelva, SO Espa˜
na): influencia de las construcciones costeras.
Revista de la Sociedad Geol´ogica de Espa ˜na 11(3– 4): 285–296.
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

WINTER WAVE CLIMATE, STORMS AND REGIONAL CYCLES 2155
Barnston AG. Livezey RE. 1987. Classification, seasonality and
persistence of low-frequency atmospheric circulation patterns.
Monthly Weather Review 15: 1083 –1126.
Barr S. 2004. Port beach Coastal Erosion Study, Technical Report 427.
Department for planning and infrastructure: Fremantle.
Bell GD, Janowiak JE. 1995. Atmospheric circulation associated with
the Midwest floods of 1993. Bulletin of American Meteorological
Society 5: 681– 695.
Benavente J, Del R´
ıo L, Anfuso G, Gracia, FJ, Reyes JL. 2002. Utility
of morphodynamic characterization in the prediction of beach
damage by storms. Journal of Coastal Research SI36: 56–64.
Boashash B. 2003. Time-Frequency signal Analysis and Processing: A
Comprehensive Reference. Elsevier Science: Oxford.
Bromirski PD, Flinck RE, Cayan DR. 2003. Storminess variability
along the California coast 1858–2000. Journal of Climate 16(6):
982– 993.
Carter DJT, Draper L. 1988. Has the north-east Atlantic become
rougher? Nature 332: 494.
Cattiaux J, Vautard R, Cassou C, Yiou P, Masson-Delmotte V,
Codron F. 2010. Winter 2010 in Europe: a cold extreme in a
warming climate. Geophysical Research Letters 37: L20704, DOI:
10.1029/2010GL044613.
Chen WY, Van den Dool H. 2003. Sensitivity of teleconnection
patterns to the sign of their primary action center. Monthly Weather
Review 131: 2885– 2899.
Cohen J, Foster J, Barlow M, Saito K, Jones J. 2010. Winter
2009– 2010: a case study of an extreme Arctic Oscillation
event. Geophysical Research Letters 37: L17707, DOI: 10.1029/
2010GL044256.
Coles S. 2001. An Introduction to Statistical Modeling of Extreme
Value s . Springer: London.
Cook NJ. 2004. Confidence limits for extreme wind speeds in mixed
climates. Journal of Wind Engineering and Industrial Aerodynamics
92: 41– 51.
Dailidiene I, Davuliene L, Kelpsaite L, Razinkovas A. 2011. Analysis
of the climate change in Lithuanian coastal areas of the
Baltic Sea. Journal of Coastal Research 28(3): 557 –569, DOI:
10.2112/JCOASTRES-D-10-00077.1.
Dodet G, Bertin X, Taborda R. 2010. Wave climate variability in the
North-East Atlantic Ocean over the last six decades. Ocean Modeling
31: 120– 131.
Dolan R, Davis RE. 1992. An intensity scale for Atlantic coast
northeast storms. Journal of Coastal Research 8: 352–364.
Dolan R, Lins H, Hayden B. 1989. Mid-Atlantic coastal storms.
Journal of Coastal Research 4: 417– 433.
Donat M, Leckebusch GC, Pinto JG, Ulbrich U. 2010. European
storminess and associated circulation weather types: future changes
deduced from a multi-model ensemble of GCM simulations. Climate
Research 42: 27– 43.
Dorsch W, Newland T, Tassone D, Tymons S, Walker D. 2008. A
statistical approach to modeling the temporal patterns of ocean
storms. Journal of Coastal Research 24(6): 1430– 1438.
Dupuis H, Michel D, Sottolichio A. 2006. Wave climate evolution in
the Bay of Biscay over two decades. Journal of Marine Systems 63:
105– 114.
Eliot IG, Clarke DJ. 1986. Minor storm impact on the beach face of a
sheltered sandy beach. Marine Geology 73(1–2): 61–83.
Esbensen SK. 1984. A comparison of intermonthly and interannually
teleconnection in the 700 mb geopotential height field during
the Northern Hemisphere winter. Monthly Weather Review 112:
2016– 2032.
Esteves LS, Williams JJ, Brown JM. 2011. Looking for evidence of
climate change impacts in the eastern Irish Sea. Natural Hazards
and Earth System Sciences 11: 1641 –1656.
Ferreira O, Vousdoukas MV, Ciavola P. 2009. MICORE Review of
Climate Change Impacts on Storm Occurrence. MICORE: Florence.
Gallego MC, Garcia JA, Vaquero JM. 2005. The NAO signal in daily
rainfall series over the Iberian Peninsula. Climate Research 29(2):
103– 109.
Goda Y. 1988. On the methodology of selecting design wave height.
In Proceedings of the twenty-first coastal engineering conference.
Malaga, Spain.
Goel NK, Burn DH, Pandey MD, An Y. 2004. Wind quartile
estimation using a pooled frequency analysis approach. Journal of
Wind Engineering and Industrial Aerodynamics 92: 509– 528.
Gringoten LL. 1963. A plotting rule for extreme probability paper.
Journal of Geophysical Research 68(3): 813–814.
Hanson H, Larson M. 2008. Implications of extreme waves and water
levels in the southern Baltic sea. Journal of Hydraulic Research
46(2): 292– 302.
Hatzaki M, Flocas HA, Maheras P, Asimakopoulos DN, Giannakopou-
los C. 2006. Study of future climatic variations of a teleconnection
pattern affecting Eastern Mediterranean. Global NEST Journal 8(3):
195– 203.
Higgins RW, Leetma A, Kousky VE: 2002. Relationships between
climate variability and winter temperature extremes in the Unites
States. Journal of Climate 15: 1555 –1572.
Hurrell JW. 1995. Decadal trends in the North Atlantic Oscillation:
regional temperatures and precipitation. Science 269: 676–679.
Hurrell, JW, Deser C. 2009. North Atlantic climate variability: the role
of the North Atlantic Oscillation. Journal of Marine Systems 78(1):
28– 41.
Jenkinson AF. 1955.. The frequency distribution of the annual
maximum (or minimum) values of meteorological elements.
Quarterly Journal of the Royal Meteorological Society 81: 158– 171.
Jenks GF, Caspal FC. 1971. Error on choropletic maps: definition,
measurement, reduction. Annals of the Association of American
Geographers 61(2): 217– 244.
Jones A, Phillips M. 2011. Disappearing Destinations. CABI:
Wallingford.
Keim BD, Muller RA, Stone GW. 2004. Spatial and temporal
variability of coastal storms in the North Atlantic Basin. Marine
Geology 210: 7– 15.
Kelpsaite L, Dailidiene I, Soomere T. 2011. Changes in wave
dynamics at the south-eastern coast of the Baltic proper during the
1993– 2008. Boreal Environmental Research 16: 220– 232.
Komar P, Allan JC. 2008. Increasing hurricane-generated wave heights
along the U.S. East Coast and their climate controls. Journal of
Coastal Research 24(2): 479–488.
L’Heureux M, Butler A, Jha A, Kumar A, Wang W. 2010. Unusual
extremes in the negative phase of the Arctic Oscillation
during 2009. Geophysical Research Letters 37: L10704, DOI:
10.1029/2010GL043338.
Lemm AJ, Hegge BJ, Masselink G. 1999. Offshore wave climate,
Perth (Western Australia): 1994–1996. Marine and Freshwater
Research 50: 95– 102.
Li F, Roncevich L, Bicknell C, Lowry R, Ilich K, 2011. Interannual
variability and trends of storminess Perth, 1994– 2008. Journal of
Coastal Research 27(4): 738–745.
Lozano I, Devoy RJN, May W, Andersen U. 2004. Storminess and
vulnerability along the Atlantic coastlines of Europe: analysis of
storm records and of a greenhouse gases induced climate scenario.
Marine Geology 210: 205 –225.
Mao R, Gong D, Bao J, Fan Y. 2011. Possible influence of Arctic
Oscillation on dist storm frequency in North China. Journal of
Geographical Sciences 21(2): 207 –218.
Marcos M, Tsimplis M. 2008. Coastal sea level trends in Southern
Europe. Geophysical Journal International 175: 70 –82.
Marcos M, Puyol B, Woppelmann G, Herrero C, Garcia-Fernandez MJ.
2011. The long sea level record at cadiz (southern Spain) from 1880
to 2009. Journal of Geophysical Research 116: C1200, 10 pp, DOI:
10.1029/2011JC007558.
Matulla C, Schoner W, Alexandersson H. 2007. European storminess:
late nineteenth century to present. Climate Dynamics 31: 125 –130.
Mendoza ET, Jimenez JA. 2008. Coastal storm classification on the
Catalan littoral (NW Mediterranean). Ingenier´
ıa Hidr´aulica en
M´exico 23(2): 23– 34.
Men´
endez M, M ´
endez F, Losada I, Medina R, Abascal A. 2004.
Variaciones del r´
egimen extremal del clima mar´
ıtimo en el litoral
espa˜
nol en el periodo 1958–2001. In El Clima entre el Mar y la
Monta˜na, Concha D, Garcia J, Alvarez D. (eds). Universidad de
Cantabria: Santander.
Meyer-Arendt K. 2011. Grand Isle, Louisiana: a historic US Gulf
Coast Resort Adapts to Hurricanes, Subsidence and Sea Level Rise.
In Disappearing Destinations, Jones A, Phillips M. (eds). CABI:
Wallingford, 203–217.
Moritz H, Moritz H. 2006. Evaluating extreme storm power and
potential implications to coastal infrastructure damage, Oregon Coast
USA. In 9th International Workshop on Wave Hindcasting and
Forecasting. Victoria, Canada.
Morton I, Bowers J, Mould G. 1997. Estimating return period wave
heights and winds speeds using a seasonal point process model.
Coastal Engineering 26: 251 –270.
Mu˜
noz JJ, Enriquez J. 1998. Din´
amica litoral de una unidad fisiogr´
afica
completa: Sanl´
ucar– Rota. Revista de Obras Publicas 3375: 34–44.
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)

2156 N. RANGEL-BUITRAGO AND G. ANFUSO
Mu˜
noz JJ, Lopez B, Gutierrez JM, Moreno L, Cuenca G. 2001. Cost
of beach maintenance in the Gulf of Cadiz (SW Spain). Coastal
Engineering 42: 143– 153.
Mu˜
noz-Diaz D, Rodrigo FS. 2003. Effects of the North Atlantic
oscillation on the probability for climatic categories of local monthly
rainfall in southern Spain. International Journal of climatology 23(4):
381– 397.
O’Connor M, Cooper AJW, Jackson DWT. 2011. Decadal behavior of
tidal inlet-associated beach systems, northwest Ireland, in relation to
climate forcing. Journal of Sedimentary Research 81: 38–51.
Pattiaratchi C, Elliot M. 2008. Sea level variability in south-west
Australia: from hours to decades. In Proceedings of the 31st
ASCE International Conference on Coastal Engineering. Hamburg,
Germany.
Phillips M. 2008. Consequences of short-term changes in coastal
processes: a case study. Earth Surface Processes and Landforms 33:
2094– 2107.
Phillips M, Crisp S. 2010. Sea level trends and NAO influences: the
Bristol Chanel/Seven Estuary. Global and Planetary Change 73:
211– 218.
Pinto J, Zacharias S, Fink AH, Leckebush GC, Ulbrich U. 2009.
Factors contributing to the development of extreme North Atlantic
cyclones and their relationship with the NAO. Climate Dynamics 32:
711– 737.
Plomaritis T, Anfuso G, Rangel-Buitrago N, Benavente J, Del Rio L.
2009. Erosion and accretion patterns in different morphodynamic
beach states in Cadiz (SW Spain). In 7th International conference on
geomorphology. Melbourne, Australia.
Plomaritis T, Del Rio L, Benavente J, Valladares M. 2010. Storm
thresholds for the Spanish Gulf of Cadiz coast. Geophysical Research
Abstracts 12: 10781.
Rangel-Buitrago N, Anfuso G. 2011a. An application of Dolan and
Davis (1992) classification to coastal storms in SW Spanish littoral.
Journal of Coastal Research SI64: 1891– 1895.
Rangel-Buitrago N, Anfuso G. 2011b. Coastal storm characterization
and morphological impacts on sandy coasts. Earth Surface Process
and Landforms 36(15): 1997 –2010.
Rao AR, Hamed KH. 2000. Flood Frequency Analysis. CRC press:
Boca Raton.
Reyes JL, Martins JT, Benavente J, Ferreira O, Gracia FJ,
Alveri˜
no–Dias JM, L´
opez-Aguayo F. 1999. Gulf of Cadiz beaches:
a comparative response to storm events. Boletin del Instituto Espa˜nol
de Oceanograf´
ıa 15(1– 2): 221 –228.
Rodriguez A, Ruiz F, Caceres LM, Rodriguez J, Pino R, Mu˜
noz JM.
2003. Analysis of the recent storm record in the south western
Spanish coast: implications for littoral management. The Science of
the Total Environment 303: 189–201.
Rogers JC, McHugh MJ. 2002. On the separability of the North
Atlantic Oscillation and Arctic Oscillation. Climate Dynamics 19:
599– 608.
Schmith T, Kass E, Li TS. 1998. Northeast Atlantic winter storminess
1875– 1995 re-analyzed. Climate Dynamics 14: 529– 536.
Soomere T. 2008. Extremes and decadal variations of the northern
Baltic seawave conditions. In Extreme Ocean Waves, Pelinovsky E,
Kharif C. (eds). Springer-Verlag: Berlin, 193– 157.
Suursaar U. 2010. Waves, currents and sea level variations along the
Letipea– Sillamae coastal section of the southern Gulf of Finland.
Oceanologia 52(3): 391– 416.
Thomas T, Phillips MR, Williams AT, Jenkis RE. 2011. Short-term
beach rotation, wave climate and the North Atlantic Oscillation
(NAO). Progress in Physical Geography 35(3): 333– 352.
Thompson DWJ, Wallace JM. 1998. The Artic Oscillation signature
in the wintertime geopotential height and temperature fields.
Geophysical Research Letters 25: 1297– 1300.
Thompson DWJ, Wallace JM, Hegerl GC. 2000. Annular modes in
the extratropical circulation. Part II: trends. Journal of Climate 13:
1018– 1036.
Trigo RM, Osborn TJ, Corte-Real J. 2002. The North Atlantic
Oscillation influence on Europe: climate impacts and associated
physical mechanism. Climate Research 20: 9–17.
T¨
urke¸s M, Erlat E. 2008. Influence of the Arctic Oscillation on the
variability of winter mean temperatures in Turkey. Theoretical and
Applied Climatology 92: 75 –85.
Wallace JM. 2000. North Atlantic Oscillation/annular mode: two
paradigms – one phenomenon. Quarterly Journal of the Royal
Meteorological Society 126: 791– 805.
Wallace JM, Gutzler DS. 1981. Teleconnections in the geopotential
height field during the Northern Hemisphere winter. Monthly
Weather Review 109: 784 –812.
Wang S, McGrath R, Hanafin JA, Lynch P, Semmler T, Nolan P.
2008. The impact of climate change on storm surges over Irish
waters. Ocean Modeling 25: 83 –94.
Wang C, Liu H, Lee S. 2010. The record-breaking cold temperatures
during the winter of 2009/2010 in the Northern Hemisphere.
Atmospheric Science Letters 11: 161 –168.
WASA 1998. Changing waves and storms in the northeast Atlantic.
Bulletin of the American Meteorological Society 79: 741–760.
Wettstein JJ, Mearns LO. 2002. The influence of the North Atlantic-
Arctic Oscillation on mean, variance and extremes of temperature in
the northeastern United States and Canada. Journal of Climate 15:
3586– 3600.
WMO (World Meteorological Organization). 1995. The Global Climate
System Review. Climate System Monitoring1991-1993.WMO:
Geneva.
Xoplaki E. 2002. Climate Variability Over the Mediterranean,PhD
thesis, University of Bern, Switzerland.
Zhang K, Douglas BC, Leatherman SP, 2000. Twentieth-century storm
activity along the U.S. East Coast. Journal of Climate 13:
1748– 1761.
Zhou S, Miller AJ, Wang J, Angell JK. 2001. Trends of NAO and
AO and their associations with stratospheric processes. Geophysical
Research Letters 28: 4107– 4110.
Zielinski G. 2002. A classification scheme for winter storms in the
Easter and central United States with an emphasis on Nor’easters.
Bulletin of the American Meteorological Society 83: 37–51.
2012 Royal Meteorological Society Int. J. Climatol. 33: 2142 –2156 (2013)