Tito Maldonado
Aspects of climate variability
during winter and summer in
Central America
Abstract
Climate variability during winter and summer in Central America is examined. In winter, we
focused on the interaction between El Niño Southern Oscillation (ENSO), Pacific Decadal Os-
cillation (PDO), the Caribbean low-level jet (CLLJ) and precipitation. Using the composite
technique it is found that ENSO events are connected to CLLJ anomalies by modulating the
sea-level pressure (SLP) near the east coast of the United States and the Aleutian Low. The pat-
tern displayed by the SLP anomalies (SLPa) is also associated with the Pacific/North America
pattern (PNA). During warm (cold) ENSO phases, negative (positive) anomalies in the SLP field
over the east coast of North America produce cyclonic (anticyclonic) circulations at low levels.
However, the ENSO signal in the SLPa and the PNA pattern are modulated by the phases of
the PDO. Results indicate that when the ENSO and PDO are in phase (out of phase), the SLPa
signal is enhanced (weakened or cancelled), affecting the CLLJ anomalies in both direction and
intensity, also changing the spatial distribution of precipitation.
During summer, the midsummer drought (MSD) in Central America is characterised in or-
der to create annual indexes representing the timing of its phases (start, minimum and end), and
other features relevant for MSD forecasting such as the intensity and the magnitude. The MSD
intensity is defined as the minimum rainfall detected during the MSD, meanwhile the magni-
tude is the total precipitation divided by the total days between the start and end of the MSD. It
is shown that the MSD extends along the Pacific coast, however, a similar MSD structure was
detected also in two stations in the Caribbean side of Central America, located in Nicaragua.
The MSD intensity and magnitude show a negative relationship with El Niño 3.4 and a positive
relationship with the CLLJ index. However for the Caribbean stations the results were not sta-
tistically significant, which indicates that other processes might be modulating the precipitation
during the MSD over the Caribbean coast. On the other hand, the temporal variables (start,
minimum and end) show low and no significant correlations with the same indexes.
The results from canonical correlation analysis (CCA) show good performance to study the
MSD intensity and magnitude, however, for the temporal indexes the performance is not sat-
isfactory due to the low skill to predict the MSD phases. Moreover, we find that CCA allows
producing forecasts of the MSD intensity and magnitude using sea surface temperatures (SST)
with leading times of up to 3 months. Using CCA as diagnostic tool it is found that during June,
an SST dipole pattern upon the neighbouring waters to Central America is the main variability
mode controlling the inter-annual variability of the MSD features. However, there is also ev-
idence that the regional waters are playing an important role in the annual modulation of the
MSD features. The waters in the PDO vicinity might be also controlling the rainfall during the
MSD, however, exerting an opposite effect at the north and south regions of Central America.
Sammanfattning
I den här studien har klimatvariationer under vintern och sommaren i Centralamerika under-
sökts. Betydelsen av El Niño Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO),
Caribbean low-level jet (CLLJ) för nederbördens distribution och magnitude har undersökts.
Med hjälp av statistiska metoder visas att ENSO modulerar marktrycket (SLP) nära USA:s öst-
kust och därmed ger avvikelser i CLLJ. Sambandet visar att SLP anomalier (SLPa) också är
förknippade med tryckfältet i Stilla Havet/Nordamerika (PNA). Under varma (kalla) ENSO fa-
ser producerar negativa (positiva) anomalier i SLP fältet över Nordamerikas ostkust cykloniska
(anticykloniska) cirkulationer på låga nivåer. Resultaten visar att när ENSO och PDO är i fas (ur
fas), förstärks (försvagas eller avbryts) SLPa-signalen vilket påverkar CLLJ till både riktning
och intensitet, så att den geografiska fördelningen av nederbörd förändras.
Under sommaren studerades den s.k. midsommartorkan (MSD) i Centralamerika genom att
definiera index som representerar tidpunkten för faserna av MSD (start, minimum och slut), och
andra funktioner som är relevanta för MSD prognoser såsom intensitet och omfattning. MSD
sträcker sig längs Stillahavskusten, ettMSD-liknande nederbördsmönster uppträder också i data
från i två stationer på den karibiska sidan av Centralamerika. MSD-intensitet och magnitud
för MSD längs Stilla Havs kusten visar ett negativt samband med El Niño 3.4 och ett positivt
samband med CLLJ, för de karibiska stationerna var resultaten inte statistiskt signifikanta, vilket
indikerar att andra processer kan modulera nederbörden under MSD längs den karibiska kusten.
Däremot visar, visar de temporala variablerna (start, minimum och slut) låg korrelation med
samma index.
Canonical correlation analysis (CCA) har använts och visar sig vara är ett bra instrument
för att studera MSDs intensitet och omfattning, men för den temporala strukturen är resultaten
inte tillfredsställande på grund av att metoden inte kan prediktera MSD faserna. Dessutom kan
CCA användas till att producera prognoser för MSD intensitet och omfattning med hjälp av
vattenytans temperatur (SST) med upp till 3 månader i förväg. Genom att använda CCA visas
att det främst är systematiska SST-mönster (under juni månad) i havsområdens som gränsar till
Centralamerika, som styr de årliga variationerna i MSD. Även regionala havsområden spelar en
viktig roll i moduleringen av MSD. Havsområden som dominerar PDO- är också väsentliga för
nederbörden under MSD.

List of papers
This thesis is based on the following papers, which are referred to in the text
by their Roman numerals.
I Maldonado, T., A. Rutgersson, J. Amador, E. Alfaro, and B. Claremar,
2015: Variability of the Caribbean low-level jet during boreal winter:
large-scale forcings. International Journal of Climatology. DOI:
10.1002/joc.4472
II Maldonado, T., A. Rutgersson, E. Alfaro, J. Amador, and B. Claremar,
2015: Interannual variability of the midsummer drought in Central
America and the connection with sea surface temperatures. Submitted.
Reprints were made with permission from the publishers.
In paper I and II the author had the main responsibility for analysing the
data and writing the main text.

Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.1 About the Centre for Natural Disaster Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2 Aims of study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1 The wind field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Moisture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4 Variability elements in the Intra-Americas Seas . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.1 Large-scale variability modulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.2 Regional-scale modulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4.3 Local-scale modulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1 Area of study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2 Reanalysis data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Precipitation data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.4 Sea surface temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.5 El Niño and PDO indexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.1 Compositing ENSO and PDO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2 Detection of the MSD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.3 Canonical Correlation Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.4 Statistical significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.1 El Niño linearities and non-linearities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.2 Influence of the PDO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.3 Influence of the SST during summer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
6 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
7 Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
8 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

1. Introduction
1.1 About the Centre for Natural Disaster Science
Centre for Natural Disaster Science (CNDS) is a Swedish centre for research
on natural disasters. The investigations conducted at CNDS aims to contribute
to improving the ability to prevent and deal with risks in society by raising
awareness of the dynamics and consequences of natural hazards.
Within the CNDS framework the present thesis examines climate variability
aspects and their impact in the precipitation distribution in Central America.
The present work also provides tools for both studying and forecasting one of
the most interesting climate elements that yearly occurs in the region referred
as the mid-summer drought.
1.2 Aims of study
The general objective of this study is to examine the climate variability ele-
ments in Central America during both winter and summer affecting the pre-
cipitation distribution at monthly scales.
More specific targets are:
• Studying the variability of El Niño and its impact in two relevant climate
processes such as the Caribbean low-level jet and precipitation during
boreal winter.
• Analysing the inter-annual variability of the mid-summer drought and its
relationship with the sea surface temperatures of neighbouring oceans.
Central America and the Caribbean regions are home of about eighty mil-
lion of people. Formed by countries and islands among the poorest of the
Americas and even of the world, many of their inhabitants are living in areas
prone to natural disasters. Alfaro et al. (2010) pointed out that most of the nat-
ural disaster reports have been related to hydro-meteorological phenomena.
Indeed, this region is constantly affected by a wide variety of intense mete-
orological events, like travelling easterly waves, tropical cyclones, convec-
tive systems, cold surges coming from the north hemisphere, the mid-summer
drought (in Spanish canícula or veranillo), the warm pools, the trades, and
an intense low-level jet over the Caribbean Sea. Variations in the intensity of
9
these elements in combination with social factors (i.e. people living in risky
areas), can increase the probability of natural disasters.
The understanding of the annual and seasonal precipitation cycle and its
variation through the year, thereby, becomes essential for planning and prepa-
ration for the rainy season in the area. The annual rainfall cycle in Central
America is known to have two contrasting regimes between the Caribbean and
Pacific slopes, due mainly to the interaction of the trades winds with the com-
plex topography in the region While in the Pacific coast a semi annual feature
is observed trough the year, with the first rainfall maximum occurring during
May-June, and a secondary and more intense peak in August to October, the
Caribbean side exhibits wetter and more humid conditions during the winter
months, concomitant to a precipitation minimum in the Pacific coast. The re-
duction in precipitation during July is due to different processes that those at
the beginning of the year. Therefore, this relative decrease is a relevant ele-
ment to explain the annual rainfall cycle. This reduction in rainfall has been
called the mid-summer drought (MSD). Furthermore, almost simultaneously
with the dry periods and the MSD in the Pacific slope, a strong low-level wind
current known as the Caribbean low-level jet (CLLJ) occurs twice per year.
Both elements remain as a scientific challenge since little is known about their
origins, and interactions with other climate processes in the region.
The precipitation regime in Central America is highly modulated by the in-
teraction of several climate processes that occur every year, that range from
local and regional scales such as those mentioned above, to large scales forc-
ing such as El Niño, the Pacific Decadal Oscillation (PDO), and the North
Atlantic Oscillation (NAO). El Niño Southern Oscillation (ENSO), however,
is the main climate modulator during both winter and summer in the region.
ENSO signal, nevertheless, is known to show inherent variability at high and
mid latitudes, that in certain cases has been associated to non-linearities, and
in other cases to the interaction with the PDO.
At regional scales, previous studies have analysed the influence of the sea
surface temperature (SST) anomalies in the precipitation regime in Central
America. Enfield and Alfaro (1999) have shown that the magnitude of the pre-
cipitation during the rainy season (from May to October) in Central America
is highly modulated by the SST anomalies at both oceans, the Pacific and the
Atlantic. Alfaro (2007) and Fallas-López and Alfaro (2012b) studied the pre-
dictability of the rainfall using canonical correlation analysis (CCA) with SST
as predictor. Both studies found that the SST modify the response on precip-
itation in different ways during each rainfall maximum, the first rainfall peak
being modulated mainly by the SST over the tropical Atlantic, and the sec-
ondary maximum being controlled by a dipole in the SST anomalies over the
neighbour oceans to the Central American coast. Moreover, Fallas-López and
Alfaro (2012a) found that the MSD intensity is dependent on several global
climate indexes such as the Atlantic Multi-Decadal Oscillation (AMO), Niño
3 (N3), and PDO. Maldonado and Alfaro (2011) and Maldonado et al. (2013)
10
utilised CCA to explore the distribution of precipitation extreme events during
the second precipitation maximum finding that during the second rainfall max-
imum, the total precipitation and temporal distribution of precipitation events
are controlled by a dipole in the SSTs between the Pacific and the Caribbean-
North Atlantic waters. Given the above evidence, it would be expected that the
SST is involved in the MSD variability at inter-annual scales. In fact, Alfaro
(2014) found that warm (cool) conditions of Niño 3.4 tend to be associated to
drier (wetter) MSD events in some regions of the North Pacific and Central
Valley of Costa Rica.
As mentioned, the general target of this thesis is to examine the climate
variability during both dry periods in Central America: winter and summer.
Winter is defined as the time spanning from December to February (DJF),
while summer is considered from June to August (JJA). Specifically, it is stud-
ied what are the main climate variability modes that exert at changes in the
temporal and spatial distribution of precipitation during each season at inter-
annual scales.
This study contributes to the current knowledge about the weather and cli-
mate dynamics in the Central American region. In practical terms, the results
of this thesis will contribute to the improvement of the design, planning and
implementation of warning and mitigation strategies oriented to the assess-
ment of sever dry conditions, e.g. the droughts reported in the North Pacific
region of Costa Rica during July 2014 (Chinchilla-Ramírez, 2014). The results
will improve the pool of information currently available to both social and eco-
nomical sectors, such as emergency offices, water resources, agricultural and
hydro-electrical power production parties among others, contributing to bet-
ter response to an eventual emergency situation linked to winter and summer
periods.
11
2. Background
The geographical location of Central America plays a significant role to de-
scribe the climate and the variability of the region. Surrounded by two large
water masses, the Eastern Tropical Pacific (ETPac) ocean at the western side
(Fiedler and Lavín, 2006), the Caribbean Sea and North Atlantic ocean at the
eastern side; along with the Gulf of Mexico, the entire area is known as the
Intra-Americas Seas (IAS, Amador 2008) as shown in Fig. 2.1. The IAS is
also important for the global climate, because it receives large amounts of ra-
diation coming from the Sun onto the Earth surface, and the regional waters
act like an energy reservoir.
Figure 2.1. The Intra-Americas seas region is the domain bounded by 5◦ S – 35◦ N,
35◦ – 105◦ W. The contour colours represent the mean wind speed for February for the
period 1950 – 2012. The quivers are the mean wind vectors. Units for both contours
and vectors are in ms−1. The square shows the utilised area to define the Caribbean
low-level jet index (from Paper I).
The IAS region is sensitive to the effect of both large- and regional-scale
dynamical systems acting in its vicinity, such as the North Atlantic Subtrop-
ical High (NASH), strong easterly winds, a large latent heat belt, warm sea
surface temperature SST and intense precipitation. All of these elements de-
pict the general frame of the regional climate and variability in the region
(Wang et al., 2008). Furthermore, the formation of the Western Hemisphere
Warm Pool (WHWP, Wang and Enfield 2001, 2003; Wang and Fiedler 2006),
12
cyclogenesis over the Caribbean Sea (Goldenberg et al., 2001), the influence
of the Inter Tropical Convergence Zone (ITCZ) along with the MSD (Mag-
aña et al. 1999; Karnauskas et al. 2013; Herrera et al. 2015, Paper II), and
the CLLJ (Amador, 1998, 2008) form part of the vast variety of regional cli-
mate components present in this area. Many of the dynamical and physical
mechanisms and interactions with the climate of the region are still not fully
understood (Amador, 2008).
2.1 The wind field
The trade winds are the low-level tropospheric flow, classically known as part
of the equator-ward branch of the Hadley cell transporting a large amount of
the moisture, resulting in convergence with uplift and destabilised stratifica-
tion at the low latitudes. They are responsible of the convective activity and
associated precipitation distribution that take places near or within the ITCZ.
At local scales, the interaction with the topography helps to explain the tem-
poral and spatial rainfall variability in some areas of Central America (Ama-
dor et al., 2003). Two subtropical high-pressure systems located near 30◦ N,
in both the Pacific Ocean (the North Pacific High) and in the North Atlantic
Ocean (the Bermuda or Azores High). In average, there is a relative strong
meridional pressure gradient between the subtropics and tropics, which accel-
erates the air masses towards the equator, generating the trade winds.
Monsoonal systems such as the North American Monsoon System (NAMS)
and the South American Monsoon System (SAMS) interact in different ways
with the trade winds, and are very important mechanisms to explain the pre-
cipitation during the warm season in the ETPac. Vera et al. (2006) review both
systems with more details. According to Amador et al. (2006), the most rele-
vant for ocean-atmosphere dynamics in the ETPac is the NAMS.
During the northern hemisphere winter, the ITCZ is at its southernmost
position (Srinivasan and Smith, 1996), and SSTs over the adjacent areas of
the Caribbean and the Pacific are relatively uniform, with values usually be-
low 28 ◦C (Amador et al., 2003). Trade winds are intensified and a frequent
southward displacement of air masses occurs. From December – March cold
air masses coming from Canada and the polar regions penetrate deep into the
tropics and produce strong wind events associated with intense periods of rain-
fall (Schultz et al., 1997, 1998). Interaction with topography (i.e. wind is fun-
nelled trough topographical gaps in southern Mexico and Central America),
and a strong gradient in the SLP between the basins, a low in the Pacific and
a high in the Caribbean, due to the intrusion of the cold fronts (CF), produce
strong near-surface wind events that can to extend far in the ETPac. High sea
level pressure (SLP) over the southwester Caribbean generates northerly sur-
face winds across the Isthmus of Panama that extend offshore of the Gulf of
Panama into the eastern tropical Pacific region (Amador et al., 2006). Dur-
13
ing this season these intense winds over the Caribbean are associated with the
winter branch of the CLLJ (Wang and Lee 2007; Amador 2008, Paper I), and
its interaction with mountains forcing rainfall in Central America.
In the boreal summer, more complex circulations than those in winter de-
velop in the IAS region. The NAMS, generated due to the seasonal thermal
contrast that occurs in this season over the ETPac, and the North, Central and
South America masses (e.g. Higgins et al. 2003). The NAMS has been shown
to be associated with the summertime precipitation of the region (Mock, 1996;
Higgins et al., 1997). It starts to develop during May – June, and associated
with heavy rainfall in late May or early June over southern Mexico, moving
north to the south-western United States. The mature phase is reached during
July – August and September, and the precipitation regime over North Amer-
ica is related to NAMS (Higgins et al., 1997). Surges of maritime tropical air
that move northward and that are associated with rainfall over the Gulf of Cal-
ifornia and southwester United States (Douglas and Leal, 2003). The decay
phase (late September – October) can be characterize as broadly the reverse of
the onset phase, though at a slower rate (Higgins et al., 2003).
Other low-level flows occur during the boreal summer, particularly the Gulf
of California low-level jet (Douglas, 1995; Douglas et al., 1998), the Chocó
low-level jet (Poveda and Mesa, 2000), and the Caribbean low-level jet (Ama-
dor, 1998, 2008). The former is parallel to the Gulf axis and has a well defined
diurnal and synoptic scale variability. It is located below 2000 m at the north-
ern end of the Gulf of California with wind velocities of 5− 7ms−1. It is a
relevant mechanism for moisture transport into the interior of the continental
regions during the monsoon (Douglas and Leal, 2003).
The second mechanism, the Chocó jet (CJ) develops in the western coast
of Colombia near 5◦ N. It reaches its maximum by October-November, then
decreases its intensity until being almost absent during the period February-
March. Low-level warm air and moisture convergence associated with the
CJ, low surface pressure and orographic vertical motion on the western An-
des, contribute to deep convective activity, which is organized as meso-scale
convective complexes.
The third low-level flow, the CLLJ, peaks in both winter and summer (Ama-
dor, 1998; Amador et al., 2003; Amador, 2008). The annual mean cycle
of the CLLJ index at 925 hPa is shown in Fig. 2.2. During summer, it is
barotropically unstable, and has potential interaction with transients, such as
easterly waves. From May to July, easterly waves loose energy and momen-
tum strengthening the mean current and causing the low-level jet to peak in
July. The CLLJ is an important element to explain the convective activity dur-
ing July through November, and contributes to understand the climate of the
region. In winter, however, its formation is distinct to the summer compo-
nent (Amador 2008; Wang 2007; Cook and Vizy 2010, Paper I), nevertheless,
it is an important mechanism for moisture transport from the Caribbean sea to
14
Central America and the Gulf of Mexico during both seasons (Durán-Quesada
et al., 2010; Gimeno et al., 2012).
Figure 2.2. Mean annual cycle of the Caribbean low-level jet index in ms−1. Note that
since the wind is easterly, the negative values mean stronger zonal wind (from Paper
I).
According to Amador et al. (2006), the wind stress curl has a strong sea-
sonal cycle with a marked meridional migration of the up-welling (positive
curl) and down-welling (negative curl) areas that is mostly associated with
the southeast and northeast trade patterns and its north-south migration in the
eastern tropical Pacific. The distribution of this variable is almost zonal, and
most of the large-scale features are located north of the equator. In general, the
fluctuations in the magnitude of the wind stress curl in the tropics are mainly
related to seasonal atmospheric circulation and the ITCZ.
2.2 Moisture
Durán-Quesada et al. (2010) identified two main moisture sources for Central
America. The first one and the most important is located in the Caribbean.
The second appears to exist near the equatorial Pacific region. These results
were highlighted also by Gimeno et al. (2012) whose stress that the major
amount of humidity in the Continental part of Central America is originated in
the nearby oceans. A third source of moisture was found by Durán-Quesada
(2012) over Venezuela due to water recycling processes.
The intensity and extent of the moisture sources vary throughout the year.
The Caribbean source (CS) does not vary significantly along the year, except a
slight displacement towards the Gulf of Mexico during winter. In contrast, the
Pacific source (PS) shows significant variation throughout the year, and disap-
pears as a source during winter and spring, mainly due to the influence of the
ITCZ (Durán-Quesada et al., 2010). Seasonally, during the boreal summer the
humidity transport from the CS is more effective and contributes with the pre-
cipitation in Central America, while the moisture that departs from PS is not
15
even able to reach the entire Central America region, contributing only to pre-
cipitation in its southernmost portion, specifically at Costa Rica. Through the
boreal winter, there is a reduction in the moisture transport. Consistent with
the precipitation pattern over this region, exhibiting dry conditions (mainly in
the Pacific basin) during this season. Moisture convergence in Central Amer-
ica during winter is clearly less important than during spring (March, April,
May) and autumn (September, October, November), when the moisture flux
over the continental area of Central America becomes more relevant.
In order to understand the role of the CLLJ for moisture transport, it is worth
to say that the CLLJ acts not only as a moisture belt, but also as a humidity
collector that is capable of modulating surface evaporation as a result of its
moisture content (Wang et al., 2007). Many studies found that the core of
CLLJ wind jet is consistent with the maximum nucleus of moisture gain over
the Caribbean Sea (Wang, 2007; Wang and Lee, 2007; Wang et al., 2007;
Durán-Quesada et al., 2010). The major contribution occurs during boreal
summer, and for the case of the CS region this is in good agreement with
the maximum observed winds in the core of the CLLJ. However, the second
maximum of the CLLJ in boreal winter is not associated with any important
transport, mainly due to the incidence of the dry season, which is characterized
by less intense jet winds than in summer and a minimal amount of precipitable
water.
Durán-Quesada et al. (2010) highlight that the contribution of moisture to
Central America that originates in the PS region is partly determined by the
presence of the CJ, which in turn allows the development of deep convection in
the region. This contribution is more noticeable over northern Colombia when
it appears to be combined with the effect of orographic lifting as described
by Poveda and Mesa (2000). The importance of the PS region is greatest dur-
ing those parts of the boreal summer and autumn that coincide with the max-
imum velocities within the core of the CJ. A significant part of the moisture
transported by this jet is unable to reach Central America completely, mainly
as a result of the loss of moisture in the ITCZ and the presence of a mountain
range in Costa Rica.
2.3 Precipitation
The most relevant synoptic influence present in the Caribbean and Central
America area is the NASH. This subtropical high produces strong easterly
trades winds towards the equator, being in Central America the dominant
wind regime. The interaction between the trades and the topography, plus
the location of the isthmus, imprint upon the region particular climate and
weather features. This interaction produces two regional climates – Pacific
and Caribbean (Taylor and Alfaro, 2005; Amador et al., 2006), nonetheless,
the complex orography is capable to produce very local climate and weather
16
patterns, noted mainly in the high variability observed in precipitation with
respect to the altitude (Fernández et al., 1996; Amador et al., 2003).
The Pacific coast presents a bimodal rainfall distribution around the year
(Fig. 2.3, upper panel). A first maximum occurs in May-June, due to the
migration north of the ITCZ. However, the migration of the ITCZ neither ex-
plains the generalized deep convection during the rainy season over the whole
region since the ITCZ is not found at latitudes 10◦ – 12◦ N (Alfaro, 2000),
nor explain the beginning and the ending of the rainy spell. Additionally to
this migration, the SST of the neighbouring seas have warmed reaching about
29 ◦C, deep convection activity is developed along with a sub-tropical lower-
tropospheric cyclonic circulation anomaly over the subtropics.
15
Santa Cruz
10
5
0
50 100 150 200 250 300 350
25
Bluefields
20
15
10
5
0
50 100 150 200 250 300 350
Days of year
Figure 2.3. Examples of the mean precipitation annual cycle in the Pacific (top) and
Caribbean (bottom) coasts.
A relative minimum of precipitation through July-August upon this slope,
is due to the MSD (Magaña et al. 1999; Karnauskas et al. 2013; Herrera et al.
2015, Paper II). During the occurrence of the MSD in these months, the con-
vective activity diminishes, due to a decrease of about 1◦ C in the SST over
the ETPac, the cyclonic circulation anomaly weakens, corresponding to an
anticyclonic acceleration of the low-level flow and, therefore, to an intensifi-
cation of the trade winds. This leads to a formation of divergence anomalies
that inhibit deep convection activity, and the strengthening of the easterlies,
forcing upward motion and intense precipitation over the Caribbean side, and
subsidence and clear skies upon the Pacific slope (Hidalgo et al., 2015).
The second maximum peaks during August through October due to the
presence of fewer deep clouds, which would produce an increment of incom-
17
Daily precipitation rate Daily precipitation rate
      mm day-1              mm day-1        
ing solar radiation heating the SST above 28◦ C. Then, this warming in the
SST produces an increase of evaporation from the oceans to the atmosphere;
in addition, weakened trade winds and a low-level convergence anomaly lead
to enhanced deep convection. Normally, this season presents the highest fre-
quency of extreme events over the Pacific slope (Alfaro et al., 2010). While
in the Caribbean coast, rainfall decreases during these months, due to a de-
crease in the strength of the trade winds (Taylor and Alfaro, 2005; Amador
et al., 2006). From December to March, the Pacific slope of Central America
presents warm and mostly dry conditions. During these moths, the ITCZ is at
its southernmost position (Srinivasan and Smith, 1996).
The Caribbean side shows (Fig. 2.3, lower panel) a different rainfall mode,
as pointed out by Taylor and Alfaro (2005). Precipitation during the win-
ter months along this coast, is mostly related to humidity convergence, mid-
latitude air intrusions (Schultz et al., 1997, 1998), and less frequent low-level
cloud systems traveling from the east (Velásquez, 2000).
2.4 Variability elements in the Intra-Americas Seas
2.4.1 Large-scale variability modulators
ENSO is the most influential large scale index on the climate of the IAS (Wang
and Fiedler, 2006). Nevertheless, the influence of other variability modes in
the Atlantic Ocean such as the AMO, and the NAO have shown to have an
important contribution in the variability of the precipitation field of Central
America (Enfield and Alfaro 1999; Alfaro 2007; among others). Recently,
the influence of the PDO on rainfall over the Central America (Mantua and
Hare 2002; Fallas-López and Alfaro 2012b, Paper I), the influence of the
Madden-Julian Oscillation (Martin and Schumacher, 2011) on precipitation
in the Caribbean islands have also been studied.
El Niño Southern Oscillation
ENSO is a two-component physical mechanism that describes the coupling
of SST anomalies in the tropical Pacific Ocean and the fluctuation in tropical
SLP gradient in the Western and Eastern Pacific Hemispheres, also known
as the Southern Oscillation (SO). El Niño is associated with unusual strong
warming events that occur every two to seven years in concert with basin-
scale tropical Pacific anomalies (Wang and Fiedler, 2006). ENSO variability
in the eastern tropical Pacific is centred along the equator, but is closely related
to variability of the tropical WHWP. Authors such as Chen and Taylor (2002)
and Bell and Chelliah (2006) suggest that ENSO forces variations in the SST
field and vertical wind shear that trigger the inter-annual variability of the
hurricane season during ENSO events. Moreover, an increase (decrease) of
precipitation has been associated with cold (warm) ENSO phases (Dai and
Wigley 2000; Giannini et al. 2000, Paper I).
18
Variability of the surface winds has been also observed related to ENSO.
During summer, the flow over surface has been found to increase during El
Niño events while a reduction occurs in the opposite phase. Examples of that,
the changes in the CJ (Poveda and Mesa, 2000) and the CLLJ (Wang, 2007;
Amador, 2008). The core intensity of the CLLJ varies with ENSO phases
in such way that during warm (cold) events the jet core is stronger (weaker)
than normal in the boreal summer, surface wind stress and wind stress curl
area expected to be stronger (weaker) than normal in the easternmost portion
of the eastern tropical Pacific. Contrary to what happens in summers, the
jet core is weaker (stronger) than normal during warm (cold) ENSO phases
in winter (Amador 2008, Paper I). These variations in the wind field influence
the vertical wind shear and cyclone-genetic processes (Amador, 2008; Amador
et al., 2010).
It can be followed that ENSO is able to affect precipitation distribution
by modulating two different mechanisms: a) evaporation variability linked to
SST and surface drag variations and b) transport of moisture due to wind flow
modulation.
The Pacific Decadal Oscillation
The PDO is a long-lived El Niño-like pattern of the variability of the Pacific
climate (Zhang et al., 1997). Mantua et al. (1997) highlight the PDO as the
dominant pattern of Pacific Decadal Variability (PDV). Even when the PDO
is referred as an El Niño-like pattern, it differs from ENSO in the time scale
of the persistence of events and its fingerprint is more noticeable in the extrat-
ropics rather than the tropics (Mantua and Hare, 2002). The pattern of a warm
PDO phase is featured by cooler than normal SSTs in the central North Pacific
and warmer than normal SSTs along the west coast of the Americas. Sym-
metry in SST anomalies patterns between northern and southern hemispheres
exhibited by the PDO has been noted by Evans et al. (2001). The mechanisms
of the PDO are complex and still an open issue, however, some studies suggest
the importance of the tropical coupling for the existence of the PDO (Feng
et al., 2010). Schneider and Cornuelle (2005), propose the PDO to evolve
from a composition between the forcing due to El Niño 3.4 (see Trenberth
1997 or for definition of El Niño regions), and the changes of the Aleutian low
(interannual frequencies) and the Kuroshio-Oyashio Extension (decadal time
scales). The importance that the PDO may have for global climate is related to
a correlation between the PDO index (Mantua et al., 1997) and precipitation
anomalies. Warm PDO phases have been found associated with anomalously
dry periods in the eastern coasts of Eurasia, Northwest Pacific of USA, Cen-
tral America and northern South America. While the same phase seems to
be related to wetter than normal conditions in the Gulf of Alaska, South-west
USA and Mexico, South-east Brazil, South central South America and west-
ern Australia. In Paper I is found that during winter the PDO forcing is more
appreciable in combination with the El Niño phases. The PDO intensifies El
19
Niño signal in SLP anomalies, and precipitation if both processes have the
same phase. The PDO, however, modulates precipitation mainly over Mexico
and southern part of USA.
The North Atlantic Oscillation
Greatbatch (2000) made a detailed revision of the NAO. It is the most impor-
tant variability mode in the North Atlantic Ocean. One way to define and index
for NAO is like the difference between normalized mean winter (December to
March) SLP anomalies at Lisbon, Portugal and Stykkisholmur, Iceland (Hur-
rell, 1996). NAO is an element of the Arctic Oscillation (AO) patterns, then
the former is linked with SST anomalies in the East/South-east of Greenland.
NAO has been also associated with the strengthen of the NASH and the
North Eastern trade winds, affecting thus the circulation in the IAS. Thereby,
the impact of the NAO is associated with the climate features of the IAS, SST
in the Tropical North Atlantic, the size of the WHWP and the CLLJ. Seasonal
variations of the NAO may be reflected in its influence on the IAS easterly
winds and precipitation patterns (Wang, 2007). Malmgren et al. (1998) found
that during the boreal summer the NAO index has an inverse relation with the
observed precipitation patterns over Puerto Rico.
The Atlantic Multi-Decadal Oscillation
The AMO is a 70 years period mode in SST (Delworth and Mann, 2000; Kerr,
2000). This signal has also been found in different modeling analysis (Del-
worth et al., 1993; Delworth and Greatbatch, 2000; Latif et al., 2004) and it
was detected as the first rotated EOF with a large response in the North At-
lantic SST (Mestas-Nuñez and Enfield, 2001). The AMO signal modulates
precipitation variations in different regions such as the IAS (Giannini et al.,
2000). Wetter (drier) conditions over Central America (north-east Brazil) dur-
ing JJA (DJF) were found by Zhang and Delworth (2006). AMO has also been
determined to be of importance modulating the impact of ENSO on drought.
Connection between AMO and other regional features such as the Atlantic
Warm Pool (AWP) has been also studied. Wang et al. (2008) show that warm
(cool) phases of the AMO are associated with repeated large (small) AWPs,
suggesting the relationship between the AMO and Atlantic tropical cyclones.
The latter is in agreement with results that indicate the presence of multi-
decadal variations in hurricane activity due to the Atlantic SST (Goldenberg
et al., 2001). AMO is then of importance related to the low frequency vari-
ability of precipitation as it modulates the distribution of moisture and extreme
rainfall events (Maldonado et al., 2013).
The Madden-Julian Oscillation
The MJO has been identified as a 40-50 days oscillation in zonal wind anoma-
lies in the tropical Pacific (Madden and Julian, 1994). Zhang (2005) makes a
review of the basics of the Madden-Julian Oscillation (MJO). Recent studies
20
have analyzed both the ocean-MJO relationship (Webber et al., 2010), and the
interaction between the MJO and ENSO (Tang and YU, 2008; Moon et al.,
2010). Some findings have shown that this oscillation has effects on the trop-
ical troposphere and strong impact on tropical convection. The MJO-related
zonal wind anomalies in the ETPac region might be associated with the in-
crease of rainfall in the North America Monsoon region. Also, the amplifi-
cation of easterly waves as a trigger of gulf surges development that may be
related with the variability of the moisture sources that feed the North America
Monsoon System (Lorenz and Hartmann, 2006). These authors also suggested
that westerly phase of the MJO are associated with the enhancement of favor-
able conditions for Mesoscale Convective System (MCS) development in the
region. Barlow and Salstein (2006) found a positive (negative) phases of the
MJO are associated with the increase (decrease) of precipitation in the re-
gion. Recently, Martin and Schumacher (2011) showed evidence about a link
between the intensity of the CLLJ and the MJO, and it is suggested to lead
changes in precipitation. Both of these last studies find some relation between
the occurrence of extreme events and the MJO phase.
2.4.2 Regional-scale modulators
Regional circulations systems
Two important regional circulation systems modulate the weather and climate
in the region. The first mechanism is the CLLJ over the Caribbean Sea, and
the second the Chocó LLJ.
The CLLJ is important for moisture transport in the region, and also to ex-
plain the convective activity during the summer months. During the winter, it
has a second maximum, nonetheless, it has shown no relation with convection
in during this season (Amador et al. 2006; Amador 2008, Paper I). The Chocó
LLJ peaks in October-November, and contributes to the moisture transport for
the southernmost part of Central America (Durán-Quesada et al., 2010). Fur-
thermore, the Chocó jet is associated with deep convection activity over the
western Andes region in Colombia (Poveda and Mesa, 2000).
SST of neighboring oceans
The influence of the SST anomalies in the precipitation variability field has
been widely studied (Enfield and Alfaro, 1999; Alfaro, 2000, 2007). The be-
ginning and ending of the rain spells are related to fluctuations in the SST
of the Atlantic and Pacific Oceans, and, these anomalies are related to the
magnitude of rainfall and frequency of rainy days (Maldonado and Alfaro,
2010, 2011; Maldonado et al., 2013). Amador et al. (2006) pointed out that
the seasonal cycle of SST is important in defining key climatological features,
especially during summer-autumn, such as the WHWP development (Wang
and Enfield, 2001; Wang and Fiedler, 2006), the appearance of MSD (Magaña
21
et al. 1999; Karnauskas et al. 2013; Herrera et al. 2015, Paper II), and favorable
areas for cyclogenesis (Goldenberg et al., 2001). During the northern winter,
SST isotherms over the Caribbean and the eastern tropical Pacific are mostly
zonally distributed, with values usually below 28−29◦ C, except in the central
eastern tropical Pacific, and to the west of Central America, where there is a
maximum of SST all year. A result of this, and a relatively strong vertical trade
wind shear, and reduced evaporation, no major convective activity occurs in
most of the Pacific coast of Central America during this season. Also, during
boreal winter, the ITCZ is at its southernmost position (Srinivasan and Smith,
1996).
During boreal summer, the WHWP dominates the SST distribution over
most of the eastern tropical Pacific region (Magaña et al., 1999; Wang and
Enfield, 2001, 2003). In the Caribbean warm pool, organized activity is barely
observed, due mainly to a strong vertical wind shear and strong subsidence
associated with regional scale circulations, such as those associated with the
low-level jet described above.
During warm (cold) ENSO phases, the CLLJ shows stronger (weaker) than
normal wind speeds (Amador et al., 2003, 2006; Amador, 2008). This fluc-
tuation is reflected in SST anomalies over the Caribbean Sea, north coast of
Venezuela; a strong (weak) jet results in negative (positive) SST anomalies
over this region due to strong (weak) Ekman transport. In this way, the jet
may have a role in coupling SST anomalies in eastern Pacific during El Niño
or La Niña events with anomalies over some regions of the Caribbean dur-
ing summer. Variations in surface variables (precipitation and temperature) in
different sectors of Mesoamerica, including its west coast, are the result of a
combination of fluctuations in the equatorial tropical Pacific and in the tropi-
cal north Atlantic (Amador et al., 2006). Studies such as Alfaro et al. (1998);
Alfaro and Cid (1999a,b); Enfield and Alfaro (1999) show that the strongest
rainfall signal occurs when tropical north Atlantic and tropical Pacific SST
anomalies are in a configuration of meridional dipole (antisymmetric) across
the ITCZ, that is, when this anomalies have an opposite sign. The rainy season
in south Central America tends to start early and end late in years that begin
with warm SST in the tropical North Atlantic. Ending dates are also delayed
when the eastern tropical Pacific is cool.
2.4.3 Local-scale modulators
Topography is the main local modulator of the variability in the region. In-
teraction between the terrain with the induced flow coming from the Pacific
ITCZ, due to some prevailing synoptic system, produces a type of disturbance
that contributes to precipitation in Central America, which is named the “tem-
porales” (Hanstenrath, 1991; Fernández et al., 1996). These are periods of
weak-moderate nearly continuous rain, lasting several days and affecting a
22
relatively large region. Their definition includes the condition that the wind
must be weak; however, Amador et al. (2003) have shown that in some cases
winds can be intense and long-lasting.
Fernández et al. (1996) identifies at least four synoptic conditions that can
eventually generate conditions for a temporal, such as a deep lower and mid-
dle troposphere troughs in the easterlies (hurricanes excluded), intrusions of
an upper troposphere troughs in mid-latitude westerlies, outbreaks of cold air
fromNorth America, and the direct and indirect effect of hurricanes. Velásquez
(2000) also found that westward- traveling, low-level cloud system over the
Caribbean reaching the Pacific, which are not necessarily associated with mid-
latitude cold air intrusions. The frequency of these events presents a great deal
of interannual and intraseasonal variability, and their relationship to ENSO or
to other large-scale climatic signal is still unclear (Amador et al., 2006).
The sea-breeze circulations are other relevant regional modulators of the
climate, on islands and peninsulas since it favors the development of convec-
tive system. Marked by a diurnal cycle due to the thermal contrast between
the coastline and the sea, it can be related to a diurnal cycle of precipitation.
However, in coastal regions with nearby complex topographical, like Central
America, the induced flow can impact the temporal a spatial distribution of
the meso-scale features in rainfall. Besides that, there is evidence in regions
with similar characteristics to the Central American isthmus (i.e. west coast
of Colombia), that the sea breeze can penetrate over near-coastal mountains
into a valley until approximately 100 km (Warner et al., 2003). Nevertheless,
note that such a feature of the sea breeze circulation has not been reported in
Central America yet, but it might be a local modulator on many regions of this
area.
23
3. Data
3.1 Area of study
Two slightly different domains were used for each paper. In Paper I, a do-
main bounded by 22◦ S – 63◦ N and 15◦ E – 180◦ W is selected, while, for
Paper II the domain was increased to 111◦ to 15◦ E, covering basically all the
Pacific and Atlantic oceans. Both domains, however, capture the most impor-
tant global and regional climate variability modes affecting the climate and
weather of Central America and the Caribbean, such as El Niño, NAO, AMO,
PDO, Pacfic North American (PNA) pattern, WHWP, and TNA.
3.2 Reanalysis data
In Paper I, data from the National Centers for Environmental Prediction-National
Center for Atmospheric Research (NCEP-NCAR) reanalysis project (Kalnay
et al., 1996) are used because this reanalysis has previously been utilised for
climate studies in this region (e.g. Amador 2008; Herrera et al. 2015) and
its time series are long enough to capture long-lived events such as the PDO.
Thus, in Paper I the period 1950 – 2010 is taken for the SLP and horizontal
wind fields.
3.3 Precipitation data
Monthly total precipitation estimates are obtained from the Global Precipita-
tion Climatology Centre (GPCC) (Schneider et al., 2011). The GPCC data
are gridded with a 0.5◦ horizontal resolution. The dataset comprises monitor-
ing products based on quality-controlled gauge station data from 1901 to the
present. Nevertheless, in Paper I, the climatology is estimated for the 1950 –
2010 period.
A total of 25 gauge stations with daily observations of precipitation pro-
vided by the meteorological services in Central America are used in Paper II.
Their location is shown in Fig. 3.1. Note that the majority (23) of the sta-
tions are situated along the Pacific coast with only two stations located along
the Caribbean coast. The two Caribbean stations possess relevant features to
describe the precipitation behaviour in the Caribbean slope, such as their lo-
cation at the CLLJ exit and their contrasting annual precipitation cycle to that
24
in the Pacific. Thus, in order to achieve a better representation of the com-
plete region, the Caribbean stations are considered for the analysis. The gaps
in the time-series are filled using the methodology described in Alfaro and
Soley (2009), which uses autoregressive models of order 1. This method can
filter persistent signals comparable to the length of the filter, and the estimated
values are consistent with the statistical properties of the time series without
external superposition of the data. The precipitation time series length is from
1961 to 2012.
17.5
●
15.0 ● ●
●
● ●
● ●● ●
12.5 ●
● ● ●
●
●●
10.0 ●● ●●●
●
7.5 ●
5.0
−95 −90 −85 −80 −75
Figure 3.1. Spatial distribution of the gauge stations over Central America. From
Paper II.
3.4 Sea surface temperature
The extended reconstructed sea surface temperatures (ERSSTv3b, Xue et al.
2003; Smith et al. 2007) are used in Paper II. The SST anomalies are con-
structed using a combination of observed data along with models and histori-
cal sampling grids. This global database has a horizontal resolution of 2.5 by
2.5 degrees. The domain bounded by -22◦ to 63◦ N and 111◦ to 15◦ E. Note
that this domain is a slight modification of the domain used in Paper I. This
change was done in order to have a complete representation of both oceans:
the Pacific and Atlantic. The SST anomalies are used as predictors for the
canonical correlation analysis method.
25
3.5 El Niño and PDO indexes
El Niño and PDO indexes are obtained from the National Oceanic and Atmo-
spheric Administration’s (NOAA) Earth System Research Laboratory, Physi-
cal Sciences Division website1. The ENSO episodes were determined using
the definition of the Oceanic Niño Index (ONI)2 from the NOAA’s Climate
Prediction Centre (CPC). The ONI definition of warm (cold) events consists
of a threshold of ±0.5 ◦C for the 3-month running mean of SST anomalies in
the Niño 3.4 region, based on centred 30-year base periods updated every 5
years. The MJO index as is defined by the CPC3.
1http://www.esrl.noaa.gov/psd/data/climateindices/list/
2http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/
ensoyears.shtml
3http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_mjo_index/
pentad.html
26
4. Methods
4.1 Compositing ENSO and PDO
To explore the teleconnections of El Niño, a composite analysis of El Niño
and La Niña years is used in Paper I to estimate the monthly mean composite
anomalies of SLP, horizontal wind and divergence at 925 hPa and precipi-
tation in February. The composite anomalies were calculated relative to the
climatology of neutral years (i.e. neither warm nor cool ENSO events). Ta-
ble 4.1 shows the ENSO classification of the years using the ONI definition.
Note that, according to this definition, the ENSO events from 1950 to 2010
are almost evenly distributed between 18 El Niño and 16 La Niña events.
In Paper I the linear and non-linear effects of the ENSO events were studied
by addition and subtraction of the ENSO phases. Assuming that El Niño and
La Niña have the opposite signal in the anomalies (linear response) over the
same regions, subtracting the anomalies, the signal of ENSO should be ampli-
fied in such areas. In regions where the subtraction of the anomalies decreases,
this indicates that the ENSO signal in those areas does not have the exact op-
posite sign during the ENSO phases. With the addition of the anomalies it
can identify the regions in which the ENSO signal does not have the opposite
signal during each of the ENSO phases, and so a non-linear behaviour.
Similarly, a composite analysis is performed in Paper I to study the influ-
ence of the PDO on the perturbation of the wind and precipitation. Table
4.1 also shows the classification of the PDO phases according to CPC. The
mean composite anomalies in this case are relative to the 61-year (i.e. 1950 –
2010) climatology for February. Notably, as for the general cases of ENSO,
the cases are almost evenly distributed when one considers the combination of
ENSO and PDO. This examination of ENSO and PDO events clearly differs
from those conducted by, for example, Hoerling et al. (1997); Gershunov and
Barnett (1998a); and DeWeaver and Nigam (2002). However, it does capture
features that are important relative to the distribution of the ENSO and PDO
events, such as the phase shift in La Niña events mentioned by DeWeaver and
Nigam (2002). These authors also pointed out that the method used to calcu-
late the composites does not influence the results of estimating the non-linear
response of the ENSO events, being more important the distribution of the
ENSO events.
27
Table 4.1. Classification of ENSO events during February according to ONI. The
PDO information is from http://jisao.washintong.edu/pdo (from Paper I).
High PDO Low PDO
El Niño 1983, 1987, 1988, 1992 1953, 1954, 1958, 1959
1995, 1998, 2003, 2010 1964, 1966, 1969, 1973
1977, 1978
La Niña 1985, 1989, 1996, 1999 1950, 1951, 1955, 1956
2000, 2001, 2006, 2008 1974, 1975, 1976
2009
Neutral 1981, 1982, 1984, 1986 1952, 1957, 1960, 1961
1990, 1991, 1993, 1994 1962, 1963, 1965, 1967
1997, 2002, 2004, 2005 1968, 1970, 1972, 1979
2007 1980
4.2 Detection of the MSD
In Paper II the daily precipitation times series were filtered using a running
triangular weight average with a window of 31 days, to avoid or minimise
interruptions of the MSD due to weakening of the trades and/or approaching
of the ITCZ, as suggested by Ramírez (1983) and Alfaro (2014).
An algorithm to systematically identify the features of theMSDwas applied
to the filtered daily precipitation time series. This algorithm seeks for the tim-
ing of each phase of the MSD (start, minimum, and end), besides the intensity
and magnitude of the MSD. This approach provides an annual value for each
quantity, which were used as indexes to characterise the MSD. The start of the
MSD is considered as the moment when the decrease in precipitation initiates,
usually after May-Jun. The end of the MSD is when the precipitation stops
increasing, normally taking place around September-October. The minimum
occurs in between the start and end of the MSD. The MSD intensity is defined
as the minimum rainfall detected during the MSD (or the depth of the valley
in Fig. 4.1), meanwhile the magnitude is the total precipitation divided by the
total number of days between the start and end of the MSD.
The description of the whole MSD detecting process is explained as fol-
lows: first, the algorithm scans for the precipitation minimum in the filtered
time series within a reasonable range of days for the existence of the MSD,
determined by the climatologies of each station, i.e. typically from 1 June to
30 August. If the precipitation minimum falls outside this period, it is not con-
sidered a MSD event. Second, the nearest local maxima to the minimum are
sought to determine the start and the end of the MSD. The shortest distance
allowed between the local mxima and the minimum is 5 days. Consequently,
the shortest MSD would have a duration of 10 days. Moreover, the local max-
ima have the restriction that the difference of precipitation between them and
the minimum should be at least 20% per day. It is worth mentioning, that not
28
all the years present the two local maxima and the minimum. For the scope of
Paper II those years missing any of the MSD phases are removed. The causes
for those fails in the detection of the MSD phases are several, however, they
can be attributed to anomalous years with a extended dry season at the begin-
ning of the year, combined with an earlier and severe MSD, or anomalous drier
condition during the secondary precipitation maximum, years with no occur-
rence MSD at all, and in most cases due to missing data in the observations.
Figure 4.1 shows the schematics of this algorithm.
1990
C
●
A
●
B
●
160 180 200 220 240 260 280 300
Days of year
Figure 4.1. Filtered daily time series for the observed precipitation in station La Ar-
gentina during 1990. The dots represent the beginning (A), minimum (B) and end (C)
of the MSD. In this year the MSD started by day 195 (14 July), reached the minimum
at day 239 (27 August) and ended by day 280 (7 Oct). From Paper II.
29
daily precipitation (mm)
0 5 10 15 20
4.3 Canonical Correlation Analysis
Canonical correlation analysis (CCA, Wilks 2011) is applied to investigate
the variability of the MSD. CCA is a statistical technique that searches for
pairs of patterns in two multivariate data sets (fields), and constructs sets of
transformed data variables by projecting the original data onto those patterns.
These new variables maximise the interrelationships between the two fields.
The new variables can be used, analogously to the regression coefficient in the
multiple regression. The new variables or vector weights are also known as
canonical vectors, and their projection with their respective fields as canonical
variates. CCA can be useful in different applications, such as: i) to obtain
diagnostic aspects of the coupled variability of two fields, in the case when
the time series of observations of the two fields are simultaneous, or ii) to
perform statistical forecasts, in the case when the time series of observations
of one field precede the other (e.g. Alfaro 2007; Maldonado and Alfaro 2011;
Fallas-López and Alfaro 2012b; Maldonado et al. 2013).
Therefore in Paper II, CCA is used as a diagnostic tool, however, mod-
els for statistical forecasting can be built for future use. Consequently, CCA
is employed in order to analyse the relationship between the SSTs monthly
anomalies (SSTA, also denoted field X) with each of the indexes (start, min-
imum, end, intensity and magnitude of MSD) estimated to represent each of
the features of the MSD (each index would be the field Y , for the correspond-
ing model), that is, it is sought a statistical relationship between the large and
regional scale features of the SSTA and the characteristics of the MSD in each
of the stations (local scale).
The above CCA methodology is based on Maldonado et al. (2013) and it is
implemented as follows: the fields (SSTA, MSD indexes) are first reduced by
means of principal component analysis (PCA) to assure stability in the CCA
parameters. A maximum of 17 EOFs and CCA modes in the filtering stage are
allowed. This threshold was suggested by Gershunov and Cayan (2003) and
Alfaro (2007) to avoid overparameterisation.
The optimal combination of EOFs and CCAmodes are calculated by means
of the goodness index (R2, Fig. 5.9) . Notice that any set EOFs will produce
unique CCA modes for that especific set, then, once the best combination
of EOFs is determined, that set EOF is capturing the maximum variability
in each field (X and Y ), separately, for each specific CCA model (intensity
and magnitude). The maximum possible number of CCA modes, however, is
determined first by the minimum number of EOFs between both fields. Then
with the goodness index, the maximum number of CCAmodes is found for the
best fit to avoid any overparameterisation in the model Ŷ = bT ·X , where the
elements of b are the ordinary least-squares regression coefficients computed
with CCA, and Ŷ is the predicted value of Y .
The R2 is computed using cross-validation models with 1-month window
from 1961-2012 for each station in all the models. This metric also allows
30
identifying the best month to predict any of the MSD features. It is worth
mentioning that at the end the models would not necessarily have 17 EOF and
CCA modes. Stations having indexes with more than 30% of missing data
(i.e. fails detecting MSD events) are discarded. In total, 21 stations are left for
the CCA models. Precipitation data are transformed to a normal distribution
using percentiles, to achieve better performance in the EOF. The 21 stations
having less than 30% of missing data are filled using the long-term means of
each index. Note that filling the data in this step is not the same procedure
described previously in Section 3.3.
4.4 Statistical significance
In Paper I, the statistical significance for the composite analysis was tested
using bootstrapping estimated in a way similar to that of Gershunov and Bar-
nett (1998a). Using bootstrapping without replacement an artificial time series
with normally distributed statistical noise is generated. The anomalies would
then be considered significant at the 90% level if the absolute values exceed
the 5th or 95th percentiles. A similar relationship between the absolute values
of the anomalies was used to estimate the 95% and 99% confidence levels.
31
5. Results
In the next sections is presented a summary of the results from both paper. In
Sections 5.1 and 5.2 describe the interactions between El Niño, PDO, SLP and
precipitation found in Paper I, while in Section 5.3 summarises the outcomes
from CCA models used in Paper II to study the MSD variability.
5.1 El Niño linearities and non-linearities
The ENSO episode signals in the SLP are shown in Fig. 5.1. During warm
episodes (Fig. 5.1(a)), there are significant negative SLPa over the Aleutian
Peninsula in Alaska and the east coast of North America (at the 95% and
99% confidence levels, respectively). In contrast, during cold events (Fig.
5.1(b)), almost the opposite pattern is observed, with the SLP being higher
than normal in the same regions at the 95% significance level. Hoerling et al.
(1997) noted that, during El Niño, the centre of this SLPa pattern extends
to the Tropical Northern Hemisphere (TNH) teleconnection, while La Niña
extends to the PNA teleconnection pattern. However, La Niña anomalies are
shifted south over the Pacific and east over the Atlantic relative to the El Niño
signal. Zárate-Hernández (2013) found evidence of a similar configuration of
the SLPa associated with the low (high) frequency and penetration of CFs in
the Caribbean region.
The linear component of El Niño teleconnections (Fig. 5.1(c)) is domi-
nant over the Aleutian Low and the east coast of the United States. The Gulf
of Mexico and the Caribbean region SLPa display linear behaviour, with a
higher order of magnitude than the non-linear behaviour for the same rea-
son as explained above. Non-linear anomalies are observed mainly over the
Pacific coastal and central United States (Fig. 5.1(d)) and could be associ-
ated with the southward shift of the SLPa in the North Pacific during La Niña
events. DeWeaver and Nigam (2002) found that the non-linear component of
the geopotential field is not negligible over the tropics. Same composites were
estimated for a geopotential height of 500 hPa (not shown). Here, El Niño
exhibits an asymmetric pattern compared with La Niña pattern. Similar to the
low SLPa levels (Fig. 5.1(a)) during El Niño, two significant troughs are lo-
cated over the Alaskan Peninsula and over the east coast of North America
connected by a band of low pressure. A ridge can be observed centred at 60◦
N, 90◦ W. During La Niña over the Alaskan Peninsula and the eastern part
of North America, the geopotential height anomalies both have an opposite
32
sign to El Niño and are relatively lower in intensity. Two ridges separated by
a relatively strong trough are centred at 50◦ N, 120◦ W. This asymmetry was
previously found by DeWeaver and Nigam (2002), who highlighted that the
asymmetry observed in these cases could be related to the shift observed in La
Niña events in 1976 – 1977, reflecting the inherent variability of ENSO events.
The linear component tends to dominate in the mid-latitudes between 30◦ and
60◦ N, with relatively important effects in the vicinities of the Caribbean Sea
and the Central Pacific. The non-linear component displays more interaction
focused in mid-latitudes over the North Pacific to the Alaskan Peninsula, the
west coast, and northern North America.
Figure 5.1. February composites of El Niño (a) and La Niña (b) used to estimate
the mean SLP anomalies with NCEP-NCAR reanalysis data. The linear component
(c) is estimated as the difference between El Niño and La Niña anomalies, whereas
the non-linear component (d) is the sum of both. Contour levels are spaced 2 hPa
apart. Shaded regions are significant anomalies with confidence levels of 90 – 95%
(horizontal lines), 95 – 99% (vertical lines), and >99% (slant lines). From Paper I.
Composites of the horizontal wind anomalies at 925 hPa during warm and
cold episodes (Fig. 5.2(a) and (b)) confirm previous results reported by Ama-
dor et al. (2003) and Amador (2008). Warm (cold) episodes during boreal
winter and lower (higher) than normal SLPa over the east coast of the United
States and over the Alaskan Peninsula (described above) are associated with
cyclonic (anti-cyclonic) circulation at low levels, resulting in westerly (east-
erly) wind anomalies and weakening (enhancing) the intensity of the CLLJ.
However, the anomalies over the Gulf of Mexico and the Caribbean Sea are
mainly dominated by the linear (Fig. 5.2(c)) rather than the non-linear com-
ponent (Fig. 5.2(d)). This indicates that, at low atmospheric levels, El Niño
signal behaves linearly over this region, as do the jet anomalies during ENSO
phases.
Precipitation composites exhibit a dominant contrasting pattern during El
Niño and La Niña events. Drier conditions are found from the Yucatan Penin-
33
Figure 5.2. February horizontal wind anomalies at 925 hPa for (a) El Niño, (b) La
Niña, (c) El Niño - La Niña (linear anomalies), and (d) El Niño + La Niña (non-linear
anomalies). The reference vector is in ms−1 (Paper I).
sula to northern South America during warm events, while Mexico and Cuba
experience wetter conditions (Fig. 5.3(a)). During cold episodes, the inverse
pattern of rainfall anomalies is observed, with essentially the same spatial dis-
tribution (Fig. 5.3(b)) as during warm events, except for Yucatan Peninsula,
Guatemala, Honduras, and El Salvador, where precipitation anomalies are
mostly negative in both phases. Durán-Quesada (2012) pointed out that these
contrasting precipitation patterns arise due to the relationship between the
CLLJ and the ENSO phases. During warm (cold) events, the CLLJ is weak-
ened (enhanced), the moisture transport efficiency is weakened (enhanced),
and less (more) humidity is transported to Central America. Non-linear anoma-
lies are observed in the rainfall having the same order of magnitude as does
the linear counterpart (Fig. 5.3(d)), however, with no statistical significance.
These non-linear anomalies in the rainfall result in wetter conditions over
Costa Rica and the west coast of southern Mexico, and drier conditions over
the rest of Central America and Mexico. The cause of the non-linear rainfall
anomalies is difficult to determine; however, from these results, it cannot be
related to the non-linear signal in the horizontal wind over the Caribbean Sea
or in the pressure pattern over the Atlantic, as the non-linear signal in those
fields is very low relative to the linear counterpart.
Comparing the mean precipitation anomalies with the divergence compos-
ites (Fig. 5.4), convergence (divergence) is observed over regions with drier
(wetter) conditions during El Niño or La Niña (Fig. 5.4(a) and (b)). The lin-
ear component of the divergence dominates over continental Central America
(Fig. 5.4(c)). The non-linear component (Fig. 5.4(d)) makes a significant
contribution over southern Central America, even though the other fields anal-
ysed previously indicated that the non-linear component can be ignored in the
tropics. However, the non-linearity of precipitation cannot be linked only to
34
Figure 5.3. February anomalies of total monthly precipitation in millimetres (mm)
for (a) El Niño, (b) La Niña, (c) El Niño – La Niña (linear anomalies), and (d) El
Niño + La Niña (non-linear anomalies). Shaded regions are significant anomalies
with confidence levels of 90 – 95% (horizontal lines), 95 – 99% (vertical lines), and
>99% (slant lines). From Paper I.
the divergence. Other regional elements could control the precipitation dis-
tribution, such as the interaction of the mean wind with topography or other
geographical elements present in the region. DeWeaver and Nigam (2002)
pointed out that PDO might be responsible for some observed variability in
ENSO events, and in some cases such variability has been taken erroneously
as the non-linearity of El Niño episodes.
5.2 Influence of the PDO
Since the PDO is a long-lived phenomenon lasting approximately 30 – 40
years, a 30-year running average (not shown) was applied to the PDO index.
This procedure reveals a negative PDO phase dominating from 1950 to 1981
and a positive PDO phase from 1981 to 2010. Table 4.1 lists the classifica-
tion of El Niño and PDO phases. The anomalies are estimated relative to the
climatology of the 61-year period from 1950 to 2010.
Figure 5.5(a) shows the SLPa for the El Niño and positive PDO (PDO+)
sub-composite. Note the similarity between the spatial distribution of the
35
Figure 5.4. As in Fig. 5.2 but showing the wind divergence at 925 hPa. Contour levels
are spaced every 0.4×10−6 s −1 (from Paper I).
SLPa and the SLPa pattern found for the El Niño composites (Fig. 5.1(a)).
Nevertheless, the SLPa are intensified and shifted south relative to the El
Niño case shown in Fig. 5.1(a), while a lower-than-normal pressure band
is located over Central North America, conditions that enhance the Northern
Hemisphere storms that reach low latitudes (Gershunov and Barnett, 1998b).
This configuration of the SLPa spatial distribution is called a constructive pat-
tern. In addition, significant negative SLPa (at 95 – 99% confidence levels)
are located near each coast of North America, along with cyclonic circulation
that weakens the trades and the jet over the Caribbean Sea (Fig. 5.6(a)). Rain-
fall anomalies indicate drier-than-normal conditions over Central America and
the Yucatan Peninsula and wetter-than-normal conditions in southern Mexico
(Fig. 5.7(a)). Significant anomalies in rainfall are found in south-western
Mexico (positive), Nicaragua, and Panama (negative).
The other constructive case is that of La Niña and negative PDO (PDO-)
Fig. 5.5(d)). In this case, the spatial distribution of SLPa is similar to the
SLPa pattern found in La Niña composites (Fig. 5.1(b)). La Niña signal in the
SLPa is modulated by the PDO-, producing more intense and extended SLPa.
This condition can also block storms from the north, preventing them from
reaching southern latitudes (Gershunov and Barnett, 1998b). Significant SLPa
over the eastern North America are accompanied by anticyclonic circulation,
36
Figure 5.5. February mean SLP anomalies for the El Niño and PDO phase sub-
composites: (a) El Niño and PDO+, (b) La Niña and PDO+, (c) El Niño and PDO-,
(d) La Niña and PDO-, (e) neutral and PDO+, and (f) neutral and PDO-. Contours are
spaced every 1 hPa. Shaded regions are significant anomalies with confidence levels
of 90 – 95% (horizontal lines), 95 – 99% (vertical lines), and >99% (slant lines). From
Paper I.
enhancing the trades and the CLLJ (Fig. 5.6(d)). Rainfall anomalies present
conditions that are drier than normal over Mexico and wetter than normal over
Central America and the Yucatan Peninsula (Fig. 5.7(d)).
In the combination of El Niño with PDO-, or of La Niña with PDO+, the
ENSO signal is modulated differently in comparison with the constructive
cases mentioned above. Therefore, the condition in which ENSO and PDO
are out of phase is regarded as a destructive pattern. Note that in Fig. 5.5(b)
the spatial distribution of the SLPa is different from that of the constructive
case (Fig. 5.5(d)), the high pressure anomaly centres being shifted to 160◦
W (over the north Pacific) and 30◦ W (north Atlantic). In the Atlantic, the
trades are enhanced and, in addition, the wind over the Caribbean changes in
direction (becoming south-easterly, Fig. 5.6(b)) compared with the construc-
tive case. Precipitation anomalies display similar distributions but with wetter
conditions in Central America (Fig. 5.7(b)).
In the El Niño and PDO- case (Fig. 5.5(c)), the SLPa pattern is also dis-
torted. In the Pacific, the SLPa are located farther north and their extent and
intensity are decreased. In the Atlantic, the SLPa are located farther east and
their intensity is also reduced. The circulation linked to the SLPa in the At-
lantic does not greatly modify the trades over the Caribbean (Fig. 5.6(c)).
37
Figure 5.6. February horizontal wind anomalies at 925 hPa for (a) El Niño and PDO+,
(b) La Niña and PDO+, (c) El Niño and PDO-, (d) La Niña and PDO-, (e) neutral and
PDO+, and (f) neutral and PDO-. The reference vector is in ms −1 (from Paper I).
Rainfall anomalies (Fig. 5.7(c)) display a distribution similar to that of their
constructive counterparts, though south-western Mexico is drier than normal.
Note that the rainfall anomalies are not statistically significant in Mexico or
Central America.
The influence of the PDO on the perturbations of the Central American
climate is measured during neutral ENSO conditions (Fig. 5.5(e) and (f)).
The SLPa spatial distribution in the Atlantic is similar during each phase, but
with opposite sign. Consequently, the magnitude of the SLPa is less than
when combined with ENSO events. On the other hand, negative SLPa are
found over the Pacific during each PDO phase, though negative SLPa are in-
tensified during PDO-. Moreover, in each PDO phase, the observed SLPa
do not produce marked changes in the horizontal wind at 925 hPa over the
Caribbean (Fig. 5.6), so the association between the PDO and the perturbation
in the CLLJ’s intensity and direction is negligible or non-existent. Although
the rainfall anomalies (Fig. 5.7(e) and (f)) tend to be positive during PDO+
(PDO-) phases in most of the study domain, these results are not statistically
significant.
The divergence field is shown in Fig. 5.8. In the constructive cases (Fig.
5.8(a) and (d)), the divergence anomaly pattern represents an intensification
of the pattern found during both the El Niño and La Niña phases, besides
having the opposite sign in each case. During El Niño (La Niña) positive
(negative) anomalies are observed mainly over Costa Rica and Panama, and
38
Figure 5.7. February anomalies of total monthly precipitation in millimetres (mm) for
(a) El Niño and PDO+, (b) La Niña and PDO-, (c) El Niño and PDO-, (d) La Niña and
PDO-, (e) neutral and PDO+, and (f) neutral and PDO-. Shaded regions are significant
anomalies with confidence levels of 90 – 95% (horizontal lines), 95 – 99% (vertical
lines), and >99% (slant lines). From Paper I.
over the north-eastern coast of South America, separated by negative (positive)
anomalies over Colombia and Venezuela. In the destructive cases, however,
each combination behaves differently. During the El Niño and PDO- case (Fig.
5.8(c)), larger changes are observed from 15◦ N towards the North Pole, over
continental North America, and in the equatorial eastern Pacific. Note that
the same region displays similar behaviour in the changes in distribution of
the precipitation anomalies (Fig. 5.7(a) and (c)). The other destructive case,
i.e. La Niña and PDO+ (Figure 5.8(b)), occurs in a different manner: negative
anomalies are intensified over Costa Rica, Panama, and South America, but
significant positive anomalies are centred on 40◦ N and 90◦ Wwhere reduced
precipitation is observed. In the neutral cases, nevertheless, relevant changes
happen only near the equator over continental South America. Regarding the
influence of the PDO on rainfall, it is worth mentioning that the modulated
39
Figure 5.8. February divergence anomalies at 925 hPa for (a) El Niño and PDO+, (b)
La Niña and PDO+, (c) El Niño and PDO-, (d) La Niña and PDO-, (e) neutral and
PDO+, and (f) neutral and PDO-. Contours levels are spaced every 0.4× 10−6 s−1
(from Paper I).
divergence helps to explain some observed changes in the spatial distribution
of precipitation, though there are other mechanisms related to combinations
of the PDO with PNA and/or El Niño. These mechanisms are associated with
changes in the distribution of the latent and sensible heat flux anomalies, mod-
ulating mainly the SST anomalies in the Gulf of Mexico (Muñoz et al., 2010)
that could also be associated with changed precipitation distribution in the mid
latitudes or with the changed storm tracks observed in the Northern Hemi-
sphere by Gershunov and Barnett (1998a) and Zárate-Hernández (2013).
5.3 Influence of the SST during summer
Correlations of the MSD features with El Niño 3.4 and the CLLJ indexes (Pa-
per II) show that theMSD intensity and magnitude have a negative relationship
40
with El Niño 3.4 and a positive relationship with the Caribbean low-level jet
(CLLJ) index, however, for the Caribbean stations the results were not statisti-
cally significant, which is indicating that other processes might be modulating
the precipitation during the MSD over the Caribbean coast. On the other hand,
the temporal variables (start, minimum and end) show low and no significant
correlations with the same indexes.
Such results are also observed in models based on CCA. Using the SST
anomalies as predictor (X field), CCA identifies SST patterns that are related
to the perturbations of the MSD features (Y field, predictant). The goodness
index (R2) is shown in Fig. 5.9. The intensity and magnitude show the best
skill score compared to the other indexes. These results show that the models
to study the timing of the MSD phases have a poor performance using the
CCA technique. Consequently, those models are not considered in the further
analysis. Both CCAmodels for MSD intensity and magnitude show one of the
highest values of R2 in April, which means in operational terms, information
concerning the MSD intensity and magnitude can be retrieve up to 3 months
in advance of an event. This would be valuable for preparation and planning
of the societal, economical and agricultural activities during the MSD period.
Notice that both models also shown the highest results in July, concurrently
with the existence of the MSD, and the CLLJ.
0.35
0.3
0.25
start
0.2 minimum
end
intensity
0.15 magnitude
0.1
0.05
0
-0.05
JAN FEB MAR APR MAY JUN JUL AUG
  SST  
Figure 5.9. Goodness index (R2) estimated as the average of the Pearson correlation
between synthetic time series generated by cross-validation models and the observed
MSD features per station (Paper II).
41
R2
In Paper II, however, the SSTA patterns identified by the model in June is
analysed, since for that month, CCA models show a relative high R2, and there
is also 1-month leading time for forecasting the MSD. Besides, the SST does
not change significantly from June to July, when the CCA models have the
best performance, thus, the SST in July is expected to be close to the SST in
June. The best combination of EOF and CCA modes for the MSD intensity
are X = 14, Y = 4 EOFs and 2 CCA modes, and for the MSD magnitude
X = 11, Y = 6 EOFs and 3 CCA modes, meaning that for each model (for
intensity and magnitude respectively) the best fit is achieved when using 2 and
3 modes (canonical variates) respectively, capturing the maximum influence
of the SST on the precipitation field, and specifically in the modulation of
the MSD. Figure 5.10 shows the Pearson correlation between the predicted
time series generated using cross-validation models and the observations of
both intensity and magnitude in each station. In each case, the station exhibits
relatively high significant correlation, about 0.35 in average.
The X loadings (correlation between the canonical vectors of the SST and
the SSTA) of the first mode controlling the MSD intensity shows a dipolar
pattern in the correlation with the SSTAs surrounding Central America (Fig.
5.11a,b). The highest positive correlations are found over the Pacific in the El
Niño region, whereas the highest negative correlation are found over the Tropi-
cal North Atlantic and near to the Brasil coast. The influence of this variability
mode on the precipitation in Central America has been previously studied by
Enfield and Alfaro (1999); Maldonado and Alfaro (2011) andMaldonado et al.
(2013), but for the secondary peak of precipitation during August-October
(ASO), thus, this results shows that the rainfall during the MSD is governed
by the same variability mode present during the highest precipitation season
in Central America. On the other hand, for the MSDmagnitude, the first mode
exerts more influence of the SSTA over the El Niño region, and the regional
waters close the west Mexican coast. Notice that in both cases also high pos-
itive correlations are found near the western coast of Mexico revealing that
both models are affected by the influence of regional waters, varying with the
same phase that the superficial temperatures in the El Niño region. The Y
loadings in both models are negatively correlated with most of the stations
(Fig. 5.11c,d). That means for a state in which the water over the Pacific is
warmer (colder) than normal, plus the SST over the Atlantic colder (warmer)
than normal, the intensity and magnitude of the MSD drier (wetter). This
results also connects the MSD period with the second rainfall peak, that is,
SSTs conditions for drier MSDs, would persist leading to less rainfall during
the second peak, following the results in Maldonado et al. (2013). The tem-
poral scores (canonical variates) of this mode in both models show that this
mode has mainly inter-annual variations (Fig. 5.11e,f).
The X loadings of the second mode are shown in Fig. 5.12a,b. The MSD in-
tensity is dominated by negative correlations with SST anomalies in the Trop-
ical North Atlantic, the Caribbean and Gulf of Mexico and the eastern Pacific,
42
17.5
15.0 ●
●
12.5
10.0 ● ● sig.level
● 90%
7.5 95%
99%
5.0 r
Magnitude
0.6
17.5
0.4
15.0 0.2
●
0.0
12.5
●
10.0
7.5
5.0
−95 −90 −85 −80 −75
Figure 5.10. Pearson correlation of the CCA models with the time series of intensity
(top) and magnitude (bottom) of MSD. Only significant correlations are plotted (Paper
II).
being more important than the first region. The X loadings for the MSD mag-
nitude on the other hand, reveal a tripolar configuration in the North Pacific.
This tripolar setting is located in the region where the PDO develops. The
PDO has shown more influence in the southern USA and northern Mexico
than in Central America (Muñoz et al. 2010, Paper I), and that could be the
reason that its influence is observed in the second mode. The Y loadings in
both cases (Fig. 5.12c,d) show a non-uniform distribution of the correlation.
For the MSD intensity (Fig. 5.12c), the stations are positively correlated with
this mode, meanwhile, the MSD magnitude (Fig. 5.12d) shows a clear divi-
sion between northern and southern Central America, being the former region
(Guatemala, El Salvador, Honduras and west Nicaragua) negatively correlated
and the latter region (Costa Rica, Panama and east Nicaragua) positively corre-
lated. The temporal scores for the intensity model exhibit mainly inter-annual
variability with a negative trend after 1990 (Fig. 5.12e), while the temporal
43
a) b)
111W 150E 150W 90W 30W 0E 111W 150E 150W 90W 30W 0E
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
c) d)
95W 90W 85W 80W 75W 95W 90W 85W 80W 75W
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
e) f)
Can. Corr. = 0.85 Can. Corr. = 0.85
1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010
year year
Figure 5.11. CCA mode 1 for both MSD intensity and magnitude. In the left col-
umn are the X (upper) and Y(middle) loadings, and the time scores (bottom) for the
intensity. In the right column the same but for the magnitude (Paper II).
scores of the magnitude (Fig. 5.12f) present a decadal variation, that again
possibly relates to the influence of the PDO.
44
SST (black), Intensity (blue) 6 8 10 12 14 16 18 −20 20 40 60
−0.4 −0.2 0.0 0.2 0.4
SST (black), Magnitude (blue) 6 8 10 12 14 16 18 −20 20 40 60
−0.4 −0.2 0.0 0.2 0.4
a) b)
111W 150E 150W 90W 30W 0E 111W 150E 150W 90W 30W 0E
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
c) d)
95W 90W 85W 80W 75W 95W 90W 85W 80W 75W
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
e) f)
Can. Corr. = 0.67 Can. Corr. = 0.66
1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010
year year
Figure 5.12. Same as Fig. 5.11 but for Mode 2 (Paper II).
The MSD magnitude is the only feature with a third mode, the X loadings
(Fig. 5.13a) reveal that this mode is dominated by regional SSTs in the Pacific
and Tropical North Atlantic. The north and south Pacific show a dipole being
positive and negative correlations respectively. The Y loadings (Fig. 5.13b)
45
SST (black), Intensity (blue) 6 8 10 12 14 16 18 −20 20 40 60
−0.4 −0.2 0.0 0.2 0.4
SST (black), Magnitude (blue) 6 8 10 12 14 16 18 −20 20 40 60
−0.4 −0.2 0.0 0.2 0.4
show a north/south division being positively correlated to the north and nega-
tively to the south, while the temporal scores (Fig. 5.13c) reveal an interannual
variation of this mode.
a)
111W 150E 150W 90W 30W 0E
−1.0 −0.5 0.0 0.5 1.0
b)
95W 90W 85W 80W 75W
−1.0 −0.5 0.0 0.5 1.0
c)
Can. Corr. = 0.58
1960 1970 1980 1990 2000 2010
year
Figure 5.13. CCA mode 3 for MSD magnitude. X (a) and Y (b) loadings, and (c) the
time scores (Paper II).
46
SST (black), Magnitude (blue) 6 8 10 12 14 16 18 −20 20 40 60
−0.4 −0.2 0.0 0.2 0.4
6. Concluding remarks
The interaction between El Niño and the PDO was studied in Paper I using
composite analysis. The interaction between these two variability modes dur-
ing boreal winter is summed up as follows:
• During warm (cold) events, the SLPa align with the PNA teleconnection
pattern, and cyclonic (anticyclonic) circulation is observed, resulting in
westerly (easterly) wind anomalies over the Caribbean Sea, weakening
(enhancing) the jet intensity. The PDO enhances or attenuates the ENSO
signal in the SLPa (i.e. during El Niño events concomitant with a posi-
tive PDO, the PNA teleconnection pattern is enhanced), whereas during
El Niño in combination with negative PDO, the PNA pattern is not ob-
served, resulting in no significant modification of the circulation over
the Caribbean Sea.
• Changes in the circulation due to the PDO phases during La Niña episodes
are observed in the direction change of the meridional wind anomalies
in the Caribbean Basin. Cold events plus positive (negative) PDO result
in an anomalous northward (southward) wind component. Due to the
Rossby problem adjustment, the relationship between the SLPa and the
CLLJ anomalies needs further study.
• Despite large changes in the circulation pattern found due to the in-
fluence of the PDO, changes in the monthly cumulative precipitation
anomaly were minimal over the Central American Isthmus. The small
perturbations in rainfall were related to small fluctuations in the diver-
gence in the tropical regions of the American continent.
The outputs of CCA models tailored in Paper II show:
• that the MSD intensity is mainly modulated by a dipolar configuration
in the SST anomalies formed between the Pacific and the Tropical North
Atlantic and Caribbean sea, meanwhile, the first mode controlling the
MSD magnitude shows more influence from El Niño and the regional
waters close to the western coast of Mexico and Central America. This
variability mode in both models has an inter-annual scale, with negative
effects in precipitation, i.e. when this mode is positive for both intensity
and magnitude, a reduction in rainfall is observed in almost all the sta-
tions. Similar SST dipole has been reported to influence the anomalies
47
in rainfall during the months of the highest precipitation events (August-
November), and it has been associated with intensification/reduction
of the trades, leading to a decrease/increase of precipitation in Central
America during the second rainfall period, however, these results show
this mode is present prior to the quarter ASO and is also modulating the
precipitation during boreal summer.
• The second mode in both models presents more complex structures, and
it is difficult to distinguish a general pattern affecting the intensity as
well as magnitude of the MSD. For the intensity model, the SST shows
a negative correlation pattern over the equatorial Pacific and North At-
lantic waters, nevertheless, the latter region dominates; the time series
for this mode show a negative trend after 1990. Meanwhile, for the
magnitude, a tripolar configuration over the vicinity of the PDO devel-
opment region is found. Also negative correlations are found over the
Gulf of Mexico. The influence of the PDO should be taken with cau-
tion and needs more analysis since the PDO has shown more influence
in the southern USA and northern Mexico than in Central America. The
time series exhibited inter-decadal variability. As previously mentioned,
the correlations between both variables (intensity and magnitude) and
this mode, however, do not depict a clear pattern in the stations; for
the MSD intensity the correlations reveal a positive association with the
SSTA, meanwhile, for the MSD magnitude shows a division between
the north and south Central America, being the north (Guatemala, Hon-
duras, El Salvador and west Nicaragua) affected negatively and the south
and Caribbean (Costa Rica, Panama and east Nicaragua) positively. It is
clear that the second mode is not controlling the precipitation in both
models in the same way . The North Atlantic waters become more im-
portant for the MSD intensity, hence, could be associated with the same
controlling mechanism present during the first precipitation maximum
(Alfaro, 2007; Fallas-López and Alfaro, 2012b). While for the MSD
magnitude the tripolar configuration might suggest the influence of the
PDO, also noted in the contrasting effects of this mode between north
and south Central America.
• The MSD magnitude is the only variable with higher CCA modes. The
third mode reveals the influence of regional waters close to the Central
America Pacific and a dipole formed between the north (positive) and
south (negative) Pacific, affecting with opposite sign the eastern (nega-
tive) and western (positive) part of Central America. This mode presents
an inter-annual variability scale.
• The CCA allows identifying patterns of the SST affecting variables de-
scribing the MSD such as the intensity and magnitude, however, it shows
48
a poor performance related to the temporal variables. Another benefit
achieved using CCA was the identification of particular months suitable
for prediction of the intensity and magnitude of the MSD, being capable
of forecasting up to 3 months in advance, which is a reasonable time
in terms of practical matters related to prevention and planing for the
season. Finally, it is worth pointing out that this analysis also provides
a systematic method to study the MSD features, that can be used for
statistical forecasts of such a phenomenon.
49
7. Future work
The project contemplates the elaboration of at least two papers more. The
objectives of those studies are summarised next.
• A further study of MSD extreme events using CCA analysis. This tech-
nique has shown good performance to capture the main variability pro-
cesses affecting the precipitation distribution mainly during the second
half of the year. Moreover, improving the understanding about the MSD
extreme events contributes to the scientific knowledge about the climate
in the region, also provides a systematic method to analyse and predict
the rainfall during the boreal summer. Besides, it is planed to match the
reports from affected sectors such as agricultural, and hydro-electrical
power production with the MSD forecast produced using CCA to evalu-
ate the information obtained with this approach in terms of natural dis-
asters.
• It is planed to study the CLLJ momentum sources through an integrated
model such as an Earth System Model, allowing to integrate all the cli-
mate elements at global and regional scales, since the interaction of the
CLLJ is known to encompass all these scales. Understanding the mech-
anisms involved in the formation, maintenance and dissipation of the
CLLJ will also provide better knowledge of its interactions with other
meteorological processes such as the MSD or the hurricanes occurring
frequently during the second half of the year.
50
8. Acknowledgements
I would like to thank my supervisor team Anna Rutgersson, Bjön Claremar,
Eric Alfaro and Jorge Amador for their full support. Without their guidance,
collaboration and effort this licentiate would never happen. Big thanks to all
PhD students in the meteorology group, present and former for sharing laughs
and burdens.
This licentiate was carried out within the CNDS research school, supported
by the Swedish International Development Cooperation Agency (Sida) through
their contract with the International Science Programme (ISP) at Uppsala Uni-
versity (contract number: 54100006).
51
References
Alfaro, E., 2000: Eventos cálidos y fríos en el Atlántico tropical norte. Atmosfera,
13 (2), 109–119.
Alfaro, E., 2007: Uso del análisis de correlación canónica para la predicción de la
precipitación pluvial en Centroamérica. Ingeniería y Competitividad, 9 (2), 33–48,
URL http:
//bibliotecadigital.univalle.edu.co/xmlui/handle/10893/1622.
Alfaro, E., 2014: Caracterización del “veranillo” en dos cuencas de la vertiente del
Pacífico de Costa Rica, América Central (Characterization of the Mid Summer
Drought in two Pacific slope river basins of Costa Rica, Central America).
Spanish. International Journal of Tropical Biology, 62 (4), 1–15, URL
https://www.academia.edu/9493294/Caracterizaci%C3%B3n_del_
veranillo_en_dos_cuencas_de_la_vertiente_del_Pac%C3%ADfico_de_
Costa_Rica_Am%C3%A9rica_Central_Characterization_of_the_Mid_
Summer_Drought_in_two_Pacific_slope_river_basins_of_Costa_Rica_
Central_America_._Spanish.
Alfaro, E. ., L. Cid, and D. Enfield, 1998: Relaciones entre la precipitación en
Centroamérica y los Océanos Pacífico y Atlántico Tropical. Investigaciones
Marinas, 26, 59–69.
Alfaro, E. and L. Cid, 1999a: Ajuste de un modelo VARMA para los campos de
anomalías de precipitación en Centroamérica y los índices de los océanos Pacífico
y Atlántico Tropical. Atmosfera, 12, 205–222.
Alfaro, E. and L. Cid, 1999b: Análisis de las anomalías en el inicio y el término de la
estación lluviosa en Centroamérica y su relación con los océanos Pacífico y
Atlántico Tropical. Tópicos Meteorológicos Oceanográficos, 6, 1–13.
Alfaro, E. and J. Soley, 2009: Descripción de dos métodos de rellenado de datos
ausentes en series de tiempo meteorológicas. Revista de Matemáticas: Teoría y
Aplicaciones., 16 (1), 59–74, URL http:
//revistas.ucr.ac.cr/index.php/matematica/article/view/1419.
Alfaro, E. J., A. Quesada, and F. Solano, 2010: Análisis del impacto en Costa Rica de
los ciclones tropicales ocurridos en el mar Caribe desde 1968 al 2007. Diálogos
Revista Electrónica de Historia, 11 (2), 22–38, URL
http://www.scielo.sa.cr/scielo.php?script=sci_abstract&pid=
S1409-469X2010000200002&lng=en&nrm=iso&tlng=pt.
Amador, J. A., 1998: A Climatic Feature of the Tropical Americas: The Trade Wind
Easterly Jet. Top. Meteor. Oceanogr., 5 (2), 91–102, URL
http://imn.ac.cr/publicaciones/revista/1998/diciembre/
1-Amador,Jorge-dic98.pdf.pdf.
Amador, J. A., 2008: The Intra-Americas Sea Low-level Jet Overview and Future
Research. Annals of the New York Academy of Sciences, 1146 (1), 153–188,
doi:10.1196/annals.1446.012, URL http:
//onlinelibrary.wiley.com/doi/10.1196/annals.1446.012/abstract.
52
Amador, J. A., E. J. Alfaro, O. G. Lizano, and V. O. Magaña, 2006: Atmospheric
forcing of the eastern tropical Pacific: A review. Progress In Oceanography,
69 (2–4), 101–142, doi:10.1016/j.pocean.2006.03.007, URL http:
//www.sciencedirect.com/science/article/pii/S0079661106000292.
Amador, J. A., E. J. Alfaro, E. R. Rivera, and B. Calderón, 2010: Climatic Features
and Their Relationship with Tropical Cyclones Over the Intra-Americas Seas.
Hurricanes and Climate Change, Springer Netherlands, Netherlands, Vol. 2,
149–173.
Amador, J. A., J. R. Chacón, and S. Laporte, 2003: Climate and climate variability in
the Arenal Basin of Costa Rica. Climate, Water and Trans-boundary Challenges in
the Americas, H. Díaz and B. Morehouse, Eds., Academic Publishers, Kluwer,
317–349.
Barlow, M. and D. Salstein, 2006: Summertime influence of the Madden-Julian
Oscillation on daily rainfall over Mexico and Central America. Geophysical
Research Letters, 33 (21), n/a–n/a, doi:10.1029/2006GL027738, URL http:
//onlinelibrary.wiley.com/doi/10.1029/2006GL027738/abstract.
Bell, G. D. and M. Chelliah, 2006: Leading Tropical Modes Associated with
Interannual and Multidecadal Fluctuations in North Atlantic Hurricane Activity.
Journal of Climate, 19 (4), 590–612, doi:10.1175/JCLI3659.1, URL
http://journals.ametsoc.org/doi/abs/10.1175/JCLI3659.1.
Chen, A. A. and M. A. Taylor, 2002: Investigating the link between early season
Caribbean rainfall and the El Niño+ 1 year. Int. J. Climatol., 22 (1), 87–106,
doi:10.1002/joc.711.
Chinchilla-Ramírez, G., 2014: Resumen Meteorológico Julio 2014. Boletín
Meteorológico Mensual. Instituto Meteorológico Nacional, San José, Costa Rica,
URL http:
//www.imn.ac.cr/boletin_meteo/historial/2014/BMET072014.pdf,
1–42 pp.
Cook, K. H. and E. K. Vizy, 2010: Hydrodynamics of the Caribbean Low-Level Jet
and Its Relationship to Precipitation. Journal of Climate, 23 (6), 1477–1494,
doi:10.1175/2009JCLI3210.1, URL
http://journals.ametsoc.org/doi/abs/10.1175/2009JCLI3210.1.
Dai, A. and T. M. L. Wigley, 2000: Global patterns of ENSO-induced precipitation.
Geophysical Research Letters, 27 (9), 1283–1286, doi:10.1029/1999GL011140,
URL http:
//onlinelibrary.wiley.com/doi/10.1029/1999GL011140/abstract.
Delworth, T., S. Manabe, and R. J. Stouffer, 1993: Interdecadal Variations of the
Thermohaline Circulation in a Coupled Ocean-Atmosphere Model. Journal of
Climate, 6 (11), 1993–2011,
doi:10.1175/1520-0442(1993)006<1993:IVOTTC>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(1993)006%
3C1993%3AIVOTTC%3E2.0.CO%3B2.
Delworth, T. L. and R. J. Greatbatch, 2000: Multidecadal Thermohaline Circulation
Variability Driven by Atmospheric Surface Flux Forcing. Journal of Climate,
13 (9), 1481–1495,
doi:10.1175/1520-0442(2000)013<1481:MTCVDB>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(2000)013%
53
3C1481%3AMTCVDB%3E2.0.CO%3B2.
Delworth, T. L. and M. E. Mann, 2000: Observed and simulated multidecadal
variability in the Northern Hemisphere. Climate Dynamics, 16 (9), 661–676,
doi:10.1007/s003820000075, URL
http://link.springer.com/article/10.1007/s003820000075.
DeWeaver, E. and S. Nigam, 2002: Linearity in ENSO’s Atmospheric Response. J.
Climate, 15, 2446 – 2461.
Douglas, M. W., 1995: The Summertime Low-Level Jet over the Gulf of California.
Monthly Weather Review, 123 (8), 2334–2347,
doi:10.1175/1520-0493(1995)123<2334:TSLLJO>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0493(1995)123%
3C2334%3ATSLLJO%3E2.0.CO%3B2.
Douglas, M. W. and J. C. Leal, 2003: Summertime Surges over the Gulf of
California: Aspects of Their Climatology, Mean Structure, and Evolution from
Radiosonde, NCEP Reanalysis, and Rainfall Data.Weather and Forecasting,
18 (1), 55–74, doi:10.1175/1520-0434(2003)018<0055:SSOTGO>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0434(2003)018%
3C0055:SSOTGO%3E2.0.CO%3B2.
Douglas, M. W., A. Valdez-Manzanilla, and R. Garcia Cueto, 1998: Diurnal
Variation and Horizontal Extent of the Low-Level Jet over the Northern Gulf of
California. Monthly Weather Review, 126 (7), 2017–2025,
doi:10.1175/1520-0493(1998)126<2017:DVAHEO>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0493(1998)126%
3C2017%3ADVAHEO%3E2.0.CO%3B2.
Durán-Quesada, A. M., 2012: Sources of moisture for Central America and transport
based on a Lagrangian approach: variability, contributions to precipitation and
transport mechanisms. Ph.D. thesis, Universidade de Vigo.
Durán-Quesada, A. M., L. Gimeno, J. A. Amador, and R. Nieto, 2010: Moisture
sources for Central America: Identification of moisture sources using a Lagrangian
analysis technique. Journal of Geophysical Research: Atmospheres, 115 (D5),
n/a–n/a, doi:10.1029/2009JD012455, URL http:
//onlinelibrary.wiley.com/doi/10.1029/2009JD012455/abstract.
Enfield, D. B. and E. J. Alfaro, 1999: The Dependence of Caribbean Rainfall on the
Interaction of the Tropical Atlantic and Pacific Oceans. Journal of Climate, 12 (7),
2093–2103, doi:10.1175/1520-0442(1999)012<2093:TDOCRO>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(1999)012%
3C2093%3ATDOCRO%3E2.0.CO%3B2.
Evans, M. N., M. A. Cane, D. P. Schrag, A. Kaplan, B. K. Linsley, R. Villalba, and
G. M. Wellington, 2001: Support for tropically-driven Pacific decadal variability
based on paleoproxy evidence. Geophys. Res. Lett, 28 (19), 3689–3692, URL
http://ic.ltrr.arizona.edu/pp/pdv.pdf.
Fallas-López, B. and E. J. Alfaro, 2012a: Uso de herramientas estadísticas para la
predicción estacional del campo de precipitación en América Central como apoyo
a los Foros Climáticos Regionales. 1: Análisis de tablas de contingencia. Revista
de Climatología, 2012, 12: 61-79, URL
http://webs.ono.com/reclim7/reclim12e.pdf.
54
Fallas-López, B. and E. J. Alfaro, 2012b: Uso de herramientas estadísticas para la
predicción estacional del campo de precipitación en América Central como apoyo
a los Foros Climáticos Regionales. 2: Análisis de Correlación Canónica. Revista
de Climatología, 12, URL http://webs.ono.com/reclim8/reclim12g.pdf.
Feng, L., D. Wu, X. Lin, and X. Meng, 2010: The effect of regional
ocean-atmosphere coupling on the long-term variability in the Pacific Ocean.
Advances in Atmospheric Sciences, 27 (2), 393–402,
doi:10.1007/s00376-009-8195-3, URL
http://www.springerlink.com/index/10.1007/s00376-009-8195-3.
Fernández, W., R. Chacón, and Melgarejo, 1996: On the rainfall distribution with
altitude over Costa Rica. Revista Geofísica, 44, 57–72.
Fiedler, P. C. and M. F. Lavín, 2006: Introduction: A review of eastern tropical
Pacific oceanography. Progress in Oceanography, 69 (2-4), 94–100,
doi:10.1016/j.pocean.2006.03.006, URL
http://linkinghub.elsevier.com/retrieve/pii/S0079661106000280.
Gershunov, A. and T. Barnett, 1998a: ENSO influence on intraseasonal extreme
rainfall and temperature frequencies in the contiguous United States: Observations
and model results. J. Climate, 11 (7), 1575–1586,
doi:10.1175/1520-0442(1998)011<1575:EIOIER>2.0.CO;2.
Gershunov, A. and T. Barnett, 1998b: Interdecadal modulation of ENSO
teleconnections. Bull. Am. Meteor. Soc., 79 (12), 2715–2725,
doi:10.1175/1520-0477(1998)079<2715:IMOET>2.0.CO;2.
Gershunov, A. and D. R. Cayan, 2003: Heavy Daily Precipitation Frequency over the
Contiguous United States: Sources of Climatic Variability and Seasonal
Predictability. Journal of Climate, 16 (16), 2752–2765,
doi:10.1175/1520-0442(2003)016<2752:HDPFOT>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442%282003%
29016%3C2752%3AHDPFOT%3E2.0.CO%3B2.
Giannini, A., Y. Kushnir, and M. A. Cane, 2000: Interannual Variability of Caribbean
Rainfall, ENSO, and the Atlantic Ocean*. Journal of Climate, 13 (2), 297–311,
doi:10.1175/1520-0442(2000)013<0297:IVOCRE>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(2000)013%
3C0297%3AIVOCRE%3E2.0.CO%3B2.
Gimeno, L., et al., 2012: Oceanic and terrestrial sources of continental precipitation.
Reviews of Geophysics, 50 (4), n/a–n/a, doi:10.1029/2012RG000389, URL http:
//onlinelibrary.wiley.com/doi/10.1029/2012RG000389/abstract.
Goldenberg, S. B., C. W. Landsea, A. M. Mestas-Nuñez, and W. M. Gray, 2001: The
Recent Increase in Atlantic Hurricane Activity: Causes and Implications. Science,
293 (5529), 474–479, doi:10.1126/science.1060040, URL
http://www.sciencemag.org/content/293/5529/474.
Greatbatch, R. J., 2000: The North Atlantic Oscillation. Stochastic Environmental
Research and Risk Assessment, 14 (4-5), 213–242, doi:10.1007/s004770000047,
URL http://link.springer.com/article/10.1007/s004770000047.
Hanstenrath, S., 1991: Climate Dynamics of the Tropics. Kluwer Academic
Publishers.
Herrera, E., V. Magaña, and E. Caetano, 2015: Air–sea interactions and dynamical
processes associated with the midsummer drought. Int. J. Climatol., 35,
55
1569–1578, doi:10.1002/joc.4077, URL
http://onlinelibrary.wiley.com/doi/10.1002/joc.4077/abstract.
Hidalgo, H. G., A. M. Durán-Quesada, J. A. Amador, and E. J. Alfaro, 2015: The
Caribbean Low-Level Jet, the Inter-Tropical Convergence Zone and Precipitation
Patterns in the Intra-Americas Sea: A Proposed Dynamical Mechanism.
Geografiska Annaler: Series A, Physical Geography, 97 (1), 41–59,
doi:10.1111/geoa.12085, URL
http://onlinelibrary.wiley.com/doi/10.1111/geoa.12085/abstract.
Higgins, R. W., Y. Yao, and X. L. Wang, 1997: Influence of the North American
Monsoon System on the U.S. Summer Precipitation Regime. Journal of Climate,
10 (10), 2600–2622,
doi:10.1175/1520-0442(1997)010<2600:IOTNAM>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(1997)010%
3C2600%3AIOTNAM%3E2.0.CO%3B2.
Higgins, R. W., et al., 2003: Progress in Pan American CLIVAR Research: The
North American Monsoon System. Atmósfera, 16 (1), URL
http://revistas.unam.mx/index.php/atm/article/view/8504.
Hoerling, M., A. Kumar, and M. Zhong, 1997: El Niño, La Niña, and the
nonlinearity of their teleconnections. J. Climate, 10 (8), 1769–1786,
doi:10.1175/1520-0442(1997)010<1769:ENOLNA>2.0.CO;2.
Hurrell, J. W., 1996: Influence of variations in extratropical wintertime
teleconnections on northern hemisphere temperature. Geophys. Res. Lett., 23 (6),
665–668, doi:10.1029/96GL00459, URL
http://onlinelibrary.wiley.com/doi/10.1029/96GL00459/abstract.
Kalnay, E., et al., 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer.
Meteor. Soc., 77 (3), 437–471,
doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2, URL http:
//dx.doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Karnauskas, K. B., R. Seager, A. Giannini, and A. J. Busalacchi, 2013: A simple
mechanism for the climatological midsummer drought along the Pacific coast of
Central America. Atmósfera, 26 (2), 261–281,
doi:10.1016/S0187-6236(13)71075-0, URL http:
//www.sciencedirect.com/science/article/pii/S0187623613710750.
Kerr, R. A., 2000: A North Atlantic Climate Pacemaker for the Centuries. Science,
288 (5473), 1984–1985, doi:DOI:10.1126/science.288.5473.1984.
Latif, M., et al., 2004: Reconstructing, Monitoring, and Predicting
Multidecadal-Scale Changes in the North Atlantic Thermohaline Circulation with
Sea Surface Temperature. Journal of Climate, 17 (7), 1605–1614,
doi:10.1175/1520-0442(2004)017<1605:RMAPMC>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(2004)017%
3C1605%3ARMAPMC%3E2.0.CO%3B2.
Lorenz, D. J. and D. L. Hartmann, 2006: The Effect of the MJO on the North
American Monsoon*. Journal of Climate, 19 (3), 333–343,
doi:10.1175/JCLI3684.1, URL
http://journals.ametsoc.org/doi/abs/10.1175/JCLI3684.1.
Madden, R. A. and P. R. Julian, 1994: Observations of the 40–50-Day Tropical
Oscillation—A Review. Mon. Wea. Rev., 122 (5), 814–837,
56
doi:10.1175/1520-0493(1994)122<0814:OOTDTO>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0493(1994)122%
3C0814:OOTDTO%3E2.0.CO;2.
Magaña, V., J. A. Amador, and S. Medina, 1999: The Midsummer Drought over
Mexico and Central America. Journal of Climate, 12 (6), 1577–1588,
doi:10.1175/1520-0442(1999)012<1577:TMDOMA>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442%281999%
29012%3C1577%3ATMDOMA%3E2.0.CO%3B2.
Maldonado, T. and E. Alfaro, 2010: Comparación de las salidas del modelo MM5v3
con datos observados en la Isla del Coco, Costa Rica. Tecnología en Marcha,
23 (4), 3–28.
Maldonado, T. and E. Alfaro, 2011: Predicción estacional para ASO de eventos
extremos y días con precipitación sobre las vertientes Pacífico y Caribe de
América Central, utilizando análisis de correlación canónica. InterSedes, 12 (24),
78–108, URL http://www.intersedes.ucr.ac.cr/ojs/index.php/
intersedes/article/view/301.
Maldonado, T., E. Alfaro, B. Fallas-López, and L. Alvarado, 2013: Seasonal
prediction of extreme precipitation events and frequency of rainy days over Costa
Rica, Central America, using Canonical Correlation Analysis. Adv. Geosci., 33,
41–52, doi:10.5194/adgeo-33-41-2013, URL
http://www.adv-geosci.net/33/41/2013/.
Malmgren, B. A., A. Winter, and D. Chen, 1998: El Niño–Southern Oscillation and
North Atlantic Oscillation Control of Climate in Puerto Rico. Journal of Climate,
11 (10), 2713–2717,
doi:10.1175/1520-0442(1998)011<2713:ENOSOA>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(1998)011%
3C2713%3AENOSOA%3E2.0.CO%3B2.
Mantua, N. J. and S. R. Hare, 2002: The Pacific Decadal Oscillation. Journal of
Oceanography, 58 (1), 35–44, doi:10.1023/A:1015820616384, URL
http://link.springer.com/article/10.1023/A%3A1015820616384.
Mantua, N. J., S. R. Hare, Y. Zhang, J. M. Wallace, and R. C. Francis, 1997: A
Pacific Interdecadal Climate Oscillation with Impacts on Salmon Production. Bull.
Amer. Meteor. Soc., 78 (6), 1069–1079,
doi:10.1175/1520-0477(1997)078<1069:APICOW>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0477(1997)078%
3C1069:APICOW%3E2.0.CO;2.
Martin, E. R. and C. Schumacher, 2011: Modulation of Caribbean Precipitation by
the Madden–Julian Oscillation. Journal of Climate, 24 (3), 813–824,
doi:10.1175/2010JCLI3773.1, URL
http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3773.1.
Mestas-Nuñez, A. M. and D. B. Enfield, 2001: Eastern equatorial Pacific SST
variability: ENSO and non-ENSO components and their climatic associations.
Journal of Climate, 14 (3), 391–402, URL
http://journals.ametsoc.org/doi/pdf/10.1175/1520-0442(2001)014%
3C0391%3AEEPSVE%3E2.0.CO%3B2.
Mock, C. J., 1996: Climatic controls and spatial variations of precipitation in the
western United States. Journal of climate, 9 (5), 1111–1125, URL
57
http://cat.inist.fr/?aModele=afficheN&cpsidt=3095716.
Moon, J.-Y., B. Wang, and K.-J. Ha, 2010: ENSO regulation of MJO teleconnection.
Climate Dynamics, 37 (5-6), 1133–1149, doi:10.1007/s00382-010-0902-3, URL
http://www.springerlink.com/index/10.1007/s00382-010-0902-3.
Muñoz, E., C. Wang, and D. Enfield, 2010: The Intra-Americas Sea springtime
surface temperature anomaly dipole as fingerprint of remote influence. J. Climate,
23, 43–56, doi:http://dx.doi.org/10.1175/2009JCLI3006.1.
Poveda, G. and O. J. Mesa, 2000: On the existence of Lloró (the rainiest locality on
earth): Enhanced ocean-land-atmosphere interaction by a low-level jet.
Geophysical Research Letters, 27 (11), 1675–1678, doi:10.1029/1999GL006091.
Ramírez, P., 1983: Estudio Meteorológico de los Veranillos en Costa Rica. Nota de
investigación. 5, Instituto Meteorológico Nacional, Ministerio de Agricultura y
Ganadería., San José, Costa Rica.
Schneider, N. and B. D. Cornuelle, 2005: The Forcing of the Pacific Decadal
Oscillation*. Journal of Climate, 18 (21), 4355–4373, doi:10.1175/JCLI3527.1,
URL http://journals.ametsoc.org/doi/abs/10.1175/JCLI3527.1.
Schultz, D. M., W. E. Bracken, and L. F. Bosart, 1998: Planetary- and Synoptic-Scale
Signatures Associated with Central American Cold Surges. Monthly Weather
Review, 126 (1), 5–27,
doi:10.1175/1520-0493(1998)126<0005:PASSSA>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0493(1998)126%
3C0005:PASSSA%3E2.0.CO%3B2.
Schultz, D. M., W. E. Bracken, L. F. Bosart, G. J. Hakim, M. A. Bedrick, M. J.
Dickinson, and K. R. Tyle, 1997: The 1993 Superstorm Cold Surge: Frontal
Structure, Gap Flow, and Tropical Impact. Mon. Wea. Rev., 125 (1), 5–39,
doi:10.1175/1520-0493(1997)125<0005:TSCSFS>2.0.CO;2, URL http:
//dx.doi.org/10.1175/1520-0493(1997)125<0005:TSCSFS>2.0.CO;2.
Smith, T., R. Reynolds, T. C. Peterson, and J. Lawrimore, 2007: Improvements to
NOAA’s Historical Merged Land–Ocean Surface Temperature Analysis
(1880–2006). J. Climate, 21, 2283 – 2296.
Srinivasan, J. and G. L. Smith, 1996: Meridional Migration of Tropical Convergence
Zones. Journal of Applied Meteorology, 35 (8), 1189–1202,
doi:10.1175/1520-0450(1996)035<1189:MMOTCZ>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0450(1996)035%
3C1189:MMOTCZ%3E2.0.CO%3B2.
Tang, Y. and B. YU, 2008: An Analysis of Nonlinear Relationship between the MJO
and ENSO. Journal of the Meteorological Society of Japan. Ser. II, 86 (6),
867–881, URL
http://japanlinkcenter.org/JST.JSTAGE/jmsj/86.867?from=Google.
Taylor, M. A. and E. J. Alfaro, 2005: Climate of Central America and the Caribbean.
Encyclopedia of World Climatology, J. E. Oliver, Ed., Springer, Netherlands,
183–186.
Trenberth, K. E., 1997: The Definition of El Niño. Bulletin of the American
Meteorological Society, 78 (12), 2771–2777,
doi:10.1175/1520-0477(1997)078<2771:TDOENO>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0477(1997)078%
3C2771%3ATDOENO%3E2.0.CO%3B2.
58
Velásquez, R., 2000: Mecanismos físicos de variabilidad climática y eventos
extremos en Venezuela. Lic. thesis, Departamento de Física Atmosférica, Oceánica
y Planetaria, Escuela de Física, Universidad de Costa Rica.
Vera, C., et al., 2006: Toward a unified view of the American monsoon systems.
Journal of Climate, 19 (20), 4977–5000, URL
http://journals.ametsoc.org/doi/pdf/10.1175/JCLI3896.1.
Wang, C., 2007: Variability of the Caribbean Low-Level Jet and its relations to
climate. Clim Dyn, 29 (4), 411–422, doi:10.1007/s00382-007-0243-z, URL
http://link.springer.com/article/10.1007/s00382-007-0243-z.
Wang, C. and D. B. Enfield, 2001: The Tropical Western Hemisphere Warm Pool.
Geophysical Research Letters, 28 (8), 1635–1638, doi:10.1029/2000GL011763,
URL http:
//onlinelibrary.wiley.com/doi/10.1029/2000GL011763/abstract.
Wang, C. and D. B. Enfield, 2003: A Further Study of the Tropical Western
Hemisphere Warm Pool. J. Climate, 16 (10), 1476–1493,
doi:10.1175/1520-0442(2003)016<1476:AFSOTT>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442%282003%
29016%3C1476%3AAFSOTT%3E2.0.CO%3B2.
Wang, C. and P. C. Fiedler, 2006: ENSO variability and the eastern tropical Pacific:
A review. Prog. Oceanogr., 69 (2–4), 239–266, doi:10.1016/j.pocean.2006.03.004,
URL http:
//www.sciencedirect.com/science/article/pii/S0079661106000334.
Wang, C. and S.-k. Lee, 2007: Atlantic warm pool, Caribbean low-level jet, and their
potential impact on Atlantic hurricanes. Geophysical Research Letters, 34 (2), 1–5,
doi:10.1029/2006GL028579, URL http:
//onlinelibrary.wiley.com/doi/10.1029/2006GL028579/abstract.
Wang, C., S.-k. Lee, and D. B. Enfield, 2007: Impact of the Atlantic Warm Pool on
the Summer Climate of the Western Hemisphere. Journal of Climate, 20 (20),
5021–5040, doi:10.1175/JCLI4304.1, URL
http://journals.ametsoc.org/doi/abs/10.1175/JCLI4304.1.
Wang, C., S.-K. Lee, and D. B. Enfield, 2008: Climate Response to Anomalously
Large and Small Atlantic Warm Pools during the Summer. Journal of Climate,
21 (11), 2437–2450, doi:10.1175/2007JCLI2029.1, URL
http://journals.ametsoc.org/doi/abs/10.1175/2007JCLI2029.1.
Warner, T. T., B. E. Mapes, and M. Xu, 2003: Diurnal Patterns of Rainfall in
Northwestern South America. Part II: Model Simulations. Monthly Weather
Review, 131 (5), 813–829,
doi:10.1175/1520-0493(2003)131<0813:DPORIN>2.0.CO;2, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0493%282003%
29131%3C0813%3ADPORIN%3E2.0.CO%3B2.
Webber, B. G. M., A. J. Matthews, and K. J. Heywood, 2010: A dynamical ocean
feedback mechanism for the Madden–Julian Oscillation. Quarterly Journal of the
Royal Meteorological Society, 136 (648), 740–754, doi:10.1002/qj.604, URL
http://onlinelibrary.wiley.com/doi/10.1002/qj.604/abstract.
Wilks, D. S., 2011: Statistical Methods in the Atmospheric Sciences, Volume 100,
Third Edition. 3d ed., Academic Press, Amsterdam ; Boston.
59
Xue, Y., T. Smith, and R. Reynolds, 2003: Interdecadal changes of 30-yr SST
normals during 1871-2000. JOURNAL OF CLIMATE, 16 (10), 1601–1612,
doi:10.1175/1520-0442-16.10.1601.
Zárate-Hernández, E., 2013: Climatología de masas invernales de aire frío que
alcanzan Centroamérica y el Caribe y su relación con algunos índices árticos. Top.
Meteor. Oceanogr., 12 (1), 35 – 55.
Zhang, C., 2005: Madden-Julian Oscillation. Reviews of Geophysics, 43 (2), n/a–n/a,
doi:10.1029/2004RG000158, URL http:
//onlinelibrary.wiley.com/doi/10.1029/2004RG000158/abstract.
Zhang, R. and T. L. Delworth, 2006: Impact of Atlantic multidecadal oscillations on
India/Sahel rainfall and Atlantic hurricanes. Geophysical Research Letters,
33 (17), n/a–n/a, doi:10.1029/2006GL026267, URL http:
//onlinelibrary.wiley.com/doi/10.1029/2006GL026267/abstract.
Zhang, Y., J. M. Wallace, and D. S. Battisti, 1997: ENSO-like interdecadal
variability: 1900-93. Journal of Climate, 10 (5), 1004–1020, URL
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(1997)010%
3C1004%3AELIV%3E2.0.CO%3B2.
60
Paper I
INTERNATIONAL JOURNAL OF CLIMATOLOGY
Int. J. Climatol. (2015)
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/joc.4472
Variability of the Caribbean low-level jet during boreal
winter: large-scale forcings
Tito Maldonado,* Anna Rutgersson,† Jorge Amador,‡ Eric Alfaro§ and Björn Claremar†
Department of Earth Sciences, Uppsala University, Sweden
ABSTRACT: An index capturing the anomalies of the zonal wind at 925 hPa from 1950 to 2010 was defined to explore the
relationship between the fluctuations of the Caribbean low-level jet (CLLJ) and the main climate variability modes affecting
the Intra-Americas Sea Region. El Niño Southern Oscillation (ENSO) events, here defined using the Niño 3.4 index, are found
to be the most important variability modes for the jet anomalies, in agreement with previous studies. However, the Pacific
Decadal Oscillation (PDO) and the Pacific/North American (PNA) teleconnection pattern also show significant correlations
with the CLLJ anomaly index during February. The North Atlantic Oscillation (NAO) and the Arctic Oscillation (AO) reveal
a possible interaction with the jet anomalies that could be connected with the cold fronts and cold air surges arriving to the
Caribbean basin from the Northern Hemisphere during winter. A composite technique is used to explain the correlations with
the Pacific indexes. We found that ENSO events are connected to CLLJ anomalies by modulating the sea-level pressure (SLP)
near the east coast of the United States and the Aleutian Low. The pattern displayed by the SLP anomalies (SLPa) is also
associated with the PNA. During warm (cold) ENSO phases, negative (positive) anomalies in the SLP field over the east coast
of North America produce cyclonic (anticyclonic) circulations at low levels. However, the ENSO signal in the SLPa and the
PNA pattern are modulated by the phases of the PDO. Results indicate that when the ENSO and PDO are in phase (out of
phase), the SLPa signal is enhanced (weakened or cancelled), affecting the CLLJ anomalies in both direction and intensity,
also changing the spatial distribution of precipitation.
KEY WORDS El Niño Southern Oscillation; Pacific Decadal Oscillation; Caribbean low-level jet; climate variability; ENSO
non-linearities
Received 25 March 2015; Revised 29 June 2015; Accepted 15 July 2015
1. Introduction coast, precipitation displays a bimodal annual cycle,
peaking in May–June and August–October, the second
The Central American climate is known to be controlled
precipitation maximum being more intense than the first.
by the trade winds (trades), the dominant wind regime
A relative reduction of rainfall for several weeks in July
year-round (e.g. Alfaro, 2002; Taylor and Alfaro, 2005;
and August is known as the mid-summer drought (MSD;
Amador et al., 2006). The trades are generated by the Magaña et al., 1999; Herrera et al., 2015). On the west-
North Atlantic subtropical high (NASH) and by the pres- ern Caribbean (Greater Antilles), annual rainfall also
sure gradient between the NASH and the low-pressure exhibits a bimodal structure (Martin and Schumacher,
belt formed over the intertropical convergence zone 2011b), with an initial maximum in May, a minimum
(ITCZ). The interaction of the trades with the complex in approximately July–August, and a second maximum
topography of Central America leads to a marked contrast in September–October (Jury et al., 2007, Gamble et al.,
in annual precipitation distribution between the Pacific 2008). The minimum that separates the two rainfall peaks
and Caribbean coasts of Central America. On the Pacific has also been termed the MSD based on its similarity to
the rainfall minimum that occurs on the Pacific coast of
*Correspondence to: T. Maldonado, Department of Earth Sciences, Central America (Magaña et al., 1999).
Uppsala University, Villav. 16, 752 36 Uppsala, Sweden; Centre In the Caribbean Sea, a strong low-level wind flow is
for Geophysical Research, University of Costa Rica, San Pedro present year-round. This wind current displays a bimodal
de Montes de Oca, 11501-2060 San Jose, Costa Rica. E-mail:
tito.maldonado@geo.uu.se annual cycle that peaks in February (winter) and July
† (summer) at approximately 925 hPa (Amador, 1998, 2008;Present address: Department of Earth Sciences, Uppsala University,
Villav. 16, 752 36 Uppsala, Sweden. Amador et al., 2010). This strong wind current is usu-
‡Present address: Centre for Geophysical Research and School of ally called the Intra-Americas Sea low-level jet (IALLJ;
Physics, University of Costa Rica, San Pedro de Montes de Oca, Amador, 2008) or, for concision, the Caribbean low-level
11501-2060 San Jose, Costa Rica.
§ jet (CLLJ). The CLLJ occurrence coincides with reducedPresent address: Centre for Geophysical Research, Centre for Research
in Marine Sciences and Limnology and School of Physics, University of precipitation on the Pacific coast and relatively increased
Costa Rica, San Pedro de Montes de Oca, 11501-2060 San Jose, Costa rainfall on the Caribbean coast of Central America. In
Rica. winter, the CLLJ transports moisture from the Caribbean
© 2015 Royal Meteorological Society
T. MALDONADO et al.
Sea to the eastern Central American continent and, due to impact of wind fluctuation on precipitation distribution;
ascending motion at the jet exit, precipitation occurs over nonetheless, non-linearity might appear because of the
relatively flat areas of Costa Rica and Nicaragua (Amador, inherent non-linearity of ENSO events.
2008). Cold fronts (CFs) arriving from the Northern Hemi- The variability of the CLLJ during each season is report-
sphere and terrain concavity are also elements producing edly associated with different forcings, the fluctuations of
precipitation during the winter months on the Caribbean the jet during winter being related to large-scale features
coast (Amador et al., 2006; Zárate-Hernández, 2013). On such as SST anomalies over the Niño 3.4 region (−5∘ to
the other hand, in summer, the CLLJ plays a significant 5∘N, 120∘ to 170∘W), whereas the CLLJ perturbations in
role in Central American convective activity. The CLLJ is summer are associated with regional-scale forcings such
the main element generating the MSD in July and August as the difference in the SSTs of the tropical North Atlantic
(Herrera et al., 2015). The CLLJ also modulates the ver- (TNA) and El Niño regions (Amador et al., 2010). Jury and
tical wind shear in the Caribbean Sea, one of the mecha- Malmgren (2012) found evidence that the variability of
nisms controlling the formation of tropical cyclones in the the CLLJ is associated with SLP anomalies over the Ama-
region (Amador, 2008; Amador et al., 2010). zons and the Atlantic ITCZ, rather than the North Atlantic
Previous studies (e.g. Amador, 2008; Amador et al., anticyclone. Recently, Hidalgo et al. (2015) presented evi-
2010) have noted that these two peaks in the CLLJ wind dence that the CLLJ during summer is modulated by a
speed might differ in origin, with distinct dynamic and pattern of divergence between lower and higher levels of
physical mechanisms in each season. Many studies (e.g. the troposphere over the Intra-Americas Sea region.
Wang, 2007; Wang and Lee, 2007; Muñoz et al., 2008; Besides the CLLJ variability associated with ENSO
Cook and Vizy, 2010) have attributed the origin of the events, other climate variability modes have been asso-
CLLJ to geostrophic balance produced by the meridional ciated with the CLLJ anomalies in winter. Wang (2007)
pressure gradient (due to the NASH) in the neighbour- mentioned that the North Atlantic Oscillation (NAO)
hood of the jet. This reasoning, however, could be debat- might modulate the jet anomalies through controlling the
able because of the mass field adjustment (or Rossby pressure difference between the NASH and the Icelandic
adjustment) problem (Gill, 1982; Holton, 2004). Her- Low, although he presented no supporting evidence.
rera et al. (2015) point out that the length scale of the Moreover, Muñoz and Enfield (2011) found that the NAO
CLLJ (∼1500 km) suggests that geostrophic equilibrium is not correlated with the jet’s anomalies during spring,
is achieved through the adjustment of the mass field to the although the NAO decays during that season. In addition,
wind; in other words, the meridional pressure gradient in Muñoz and Enfield (2011) documented the influence
this region adapts to the circulation, and not the reverse as of other important climate variability indexes, such as
most studies suggest. Ranjha et al. (2013), nevertheless, the Pacific Decadal Oscillation (PDO) and Pacific/North
found that the CLLJ did not follow their classification of American Pattern (PNA), related to the wind fluctuations
coastal low-level jets, thus, the sources of CLLJ momen- in spring over the Caribbean Sea and to the jet region.
tum in winter and summer remain unknown, and explo- Gershunov and Barnett (1998b) explored the influence
ration of them is beyond the scope of this paper. of the PDO in modulating El Niño signal during winter
Amador (2008) has identified El Niño Southern Oscilla- in North America. Their findings indicate that the PDO
tion (ENSO) events as the main climate variability mode intensifies (or weakens) El Niño signal in the sea-level
responsible for the inter-annual variability of the jet in pressure (SLP) field. Durán-Quesada (2012) pointed
each season. It should be mentioned that when the impact out that the CLLJ variability in winter associated with
of ENSO phases on the climate over Central America is large-scale indexes, such as the ENSO, PDO, NAO, and
analysed, it is usually assumed that the signals of both Madden–Julian Oscillation (MJO), could alter the mois-
episodes (i.e. El Niño and La Niña) have opposite effects, ture transport from the Caribbean to Central America.
that is, their teleconnections have the same spatial distri- Under certain circumstances, this transport could provoke
bution but with the inverse sign of the atmospheric field intensified dry conditions leading to severe drought events.
anomalies. Hoerling et al. (1997) analysed the effects of Notably, although most of these studies describe the asso-
the linear and non-linear teleconnections of El Niño in ciation between climate variability indexes and CLLJ
North America during boreal winter. Their findings indi- anomalies, the connection between CLLJ fluctuations
cated that precipitation in the tropics displays non-linear and climate variability modes is still an open scientific
behaviour, evidenced by the shift in the maximum anoma- question. CFs or cold surges from the Northern Hemi-
lies of tropical rainfall. They also found a large non-linear sphere (Amador et al., 2006, Zárate-Hernández, 2013)
component in the surface climate anomalies in North are additional sources of variability in winter; however,
America, consistent with a phase shift in teleconnections. their interaction with the CLLJ has not previously been
Later, DeWeaver and Nigam (2002) found that ENSO sea explored.
surface temperatures (SSTs) directly affect the large-scale This study examines the CLLJ variability in February
atmospheric circulation through their impact on diabatic (winter) to identify and establish a relationship between
heating and subsequent upper-level divergence over the the CLLJ fluctuations in this season and the main cli-
equatorial Pacific. In previous studies (e.g. Wang, 2007; mate variability modes affecting the Intra-Americas
Amador, 2008), CLLJ anomalies have been assumed to Sea. Section 2 describes the data, while the analytical
behave linearly in response to El Niño events, as has the methodology and techniques are described in Section 3.
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
CARIBBEAN LOW-LEVEL JET AND ITS VARIABILITY
Results are presented in Section 4 and the summary and (AO), WHWP, and AMO, were obtained from the
conclusions in Section 5. National Oceanic and Atmospheric Administration’s
(NOAA’s) Earth System Research Laboratory, Physical
Sciences Division website (http://www.esrl.noaa.gov/
2. Data
psd/data/climateindices/list/, last viewed 23 February
To study the influence of the large- and regional-scale 2015). The ENSO episodes were determined using the def-
modulators of the CLLJ, we selected the domain bounded inition of the Oceanic Niño Index (ONI: http://www.cpc.
by 22∘S–63∘N and 15∘E–180∘W. This domain captures ncep.noaa.gov/products/analysis_monitoring/ensostuff/
the most important global and regional climate variability ensoyears.shtml, last viewed 23 February 2015) from
modes affecting the climate and weather of Central Amer- the NOAA’s Climate Prediction Centre (CPC). The ONI
ica and the Caribbean, such as El Niño as well as the NAO, definition of warm (cold) events consists of a threshold of
Atlantic Multidecadal Oscillation (AMO), PDO, PNA, ±0.5 ∘C for the 3-month running mean of SST anomalies
Western Hemisphere Warm Pool (WHWP), and TNA. in the Niño 3.4 region, based on centred 30-year base peri-
ods updated every 5 years. The MJO index as is defined
2.1. Reanalysis data by the CPC (http://www.cpc.ncep.noaa.gov/products/
In this study, data from the National Centers for Envi- precip/CWlink/daily_mjo_index/pentad.html, last viewed
ronmental Prediction-National Center for Atmospheric 23 February 2015). In this study the MJO index 7 (40∘W)
Research (NCEP-NCAR) reanalysis project (Kalnay et al., is used as presented by Durán-Quesada (2012). We define
1996) are used because this reanalysis has been extensively an index to represent the number of CFs arriving in
utilised for climate studies in this region (e.g. Amador, the Caribbean Sea and northern South America, based
2008; Herrera et al., 2015) and its time series are long in Zárate-Hernández (2013), who numbered the CFs
enough to capture long-lived events such as the PDO. reaching this region. The count is divided into monthly
The NCEP-NCAR reanalysis has previously displayed scores from November to February covering the period
deficiencies in its representation of precipitation in the 1978–2012. Hence, the February score for 1979–2010 is
tropics, mainly at regional scales, although the large-scale selected for the CFs index.
and inter-annual variability features are rendered in good
agreement with observations (Janowiak et al., 1998). On
the other hand, Janowiak et al. (1998) also pointed out that 3. Methods
reanalysis fields such as wind, geopotential height, and
temperature contained proper estimates of the observed 3.1. CLLJ anomaly index
values. Herrera et al. (2015) mentioned that this reanalysis As the aim of this paper is to study the variability of the
can capture important features of the climate in Central maximum wind speed in winter, the February monthly
America, such as the MSD. Therefore, the NCEP-NCAR means were chosen. The maximum wind speed in Jan-
reanalysis is good enough for exploring the large-scale and uary or February has been reported as statistically not
inter-annual variability of the CLLJ for two reasons: (1) it different (Wang, 2007), while using the monthly means
is similar to other reanalyses, such as ERA-Interim (Dee the results do not show a significant difference with the
et al., 2011), in capturing the variability in dynamic fields seasonal means (not shown). The CLLJ anomaly (CLLJa)
(e.g. horizontal wind, geopotential height, and SLP) and index was defined as the areal average of zonal wind
(2) its time coverage exceeds 60 years, which is sufficient anomalies in the region bounded by 12.5∘–17.5∘N and
for studying long-lived phenomena such as the PDO. 70∘–80∘W at 925 hPa (square in Figure 1(a)), in order to
Consequently, the horizontal wind at 925 hPa and SLP describe the monthly fluctuations of zonal wind velocity
data are taken from this reanalysis database. The monthly in February, relative to the climatology over the 61 years
means for the 1950–2010 period were used to estimate from 1950 to 2010. Using this definition of the CLLJ
the monthly climatology of both fields for February, to index allows us to have a good and simple represen-
enable examination of the mean atmospheric conditions tation of the mean features of the jet core, whereas it
when the CLLJ peaks. can be applied to measure other features of the wind
such as the wind shear in the jet area as is described
2.2. Precipitation data below. Thus, having a positive (negative) anomaly means
Monthly total precipitation estimates are obtained from the a weaker (stronger) CLLJ. The annual climatology of
Global Precipitation Climatology Centre (GPCC) (Schnei- monthly means for the index is shown in Figure 1(b).
der et al., 2011). The GPCC data are gridded with a 0.5∘ To explore the variability of the vertical structure of the
horizontal resolution. The dataset comprises monitoring jet, another index is defined as the difference between
products based on quality-controlled gauge station data the monthly anomalies of the U-component at 925 and
from 1901 to the present. Nevertheless, in this study, the 1000 hPa in February. Amador (2008) described the main
climatology is estimated for the 1950–2010 period. features of the jet in each season. The jet is weaker in
winter than in summer (Figure 1(b)). The jet’s vertical
2.3. Global and regional climate variability indexes structure also differs between winter and summer, being
Global climate variability indexes such as the Caribbean less strong but more confined to 925 hPa in February. In
index (CAR), PNA, NAO, TNA, PDO, Arctic Oscillation addition, in February, the CLLJ has a single well-defined
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
T. MALDONADO et al.
(a)
30°N
20°N
10°N
10
Reference vector
0
100°W 80°W 60°W
(b) –6
–7
–8
–9
–10
–11
–12
–13
–14
ne st er
Ju gu b
Au pte
m
Se
Figure 1. (a) February mean horizontal wind at 925 hPa for the 1950–2010 period using the NCEP-NCAR reanalysis. Contour lines are spaced
1m s−1 apart and represent the wind magnitude. Arrows show the mean wind direction and magnitude. The CLLJ index is defined as the February
monthly anomalies of the zonal wind in the area enclosed by the black square. Latitude and longitude coordinates are in degrees. (b) Annual cycle
of the CLLJ monthly means index.
southward branch, whereas in July, it presents two the ONI definition. Note that, according to this definition,
ramifications, one extending towards Central America and the ENSO events from 1950 to 2010 are almost evenly dis-
the other extending to the Gulf of Mexico. This branching tributed between 18 El Niño and 16 La Niña events.
of the CLLJ suggests a connection between the jet and the The linear and non-linear effects of the ENSO events
Americas Monsoon System (AMS), as the jet is the main were studied by addition and subtraction of the ENSO
moisture transport from the Caribbean Sea to the AMS. phases. Assuming that El Niño and La Niña have the
Figure 2 shows the time series of the CLLJa index. It is to opposite signal in the anomalies (linear response) over
be noted from this figure that the variability of the wind the same regions, subtracting the anomalies, the signal of
speed, ranges from approximately −3 to 3m s−1. This ENSO should be amplified in such areas. In regions where
time series also indicates that some observed anomalies in the subtraction of the anomalies decreases, this indicates
the CLLJ can be related to ENSO episodes. that the ENSO signal in those areas does not have the exact
opposite sign during the ENSO phases. With the addition
3.2. Compositing ENSO and PDO of the anomalies we can identify the regions in which the
The winter variability of the CLLJa is found to be mainly ENSO signal does not have the opposite signal during each
associated with El Niño events (Figure 2 and Table 1), of the ENSO phases, and so a non-linear behaviour.
in agreement with previous studies (e.g. Amador, 2008). Similarly, a composite analysis is performed to study
To explore how the teleconnections of El Niño influence the influence of the PDO on the perturbation of the wind
the CLLJ anomalies, a composite analysis of El Niño and and precipitation. Table 2 also shows the classification of
La Niña years is used to estimate the monthly mean com- the PDO phases according to CPC. The mean compos-
posite anomalies of SLP, horizontal wind and divergence ite anomalies in this case are relative to the 61-year (i.e.
at 925 hPa and precipitation in February. The composite 1950–2010) climatology for February. Notably, as for the
anomalies were calculated relative to the climatology of general cases of ENSO, the cases are almost evenly dis-
neutral years (i.e. neither warm nor cool ENSO events). tributed when one considers the combination of ENSO
Table 2 shows the ENSO classification of the years using and PDO. This examination of ENSO and PDO events
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
CLLJ (m s–1)
January
February
March
April
May
July
Octobe
N rovemb
D ee rcember
CARIBBEAN LOW-LEVEL JET AND ITS VARIABILITY
CLLJa index absolute values of the anomalies was used to estimate the
4 95% and 99% confidence levels.
Cold
Warm
2 4. Results
Figure 2 shows the time series of the CLLJ anomaly index.
It is to be noted from this figure that there is a relation-
0 ship between the jet fluctuations and ENSO episodes, i.e.
during warm (cold) events, the jet is weakened (enhanced)
due to an increase in westerly (easterly) wind anomalies
−2 over the jet core region. This association between ENSO
episodes and CLLJ perturbations was previously reported
by, for example, Amador et al. (2003), Wang (2007), and
Amador (2008). However, not all CLLJ anomalies can
−4 be explained by ENSO events, as shown in Figure 2.
1950 1970 1990 2010 This reveals an inherent ENSO variability due to either
Year non-linearities or the presence of additional external forc-
ings that affect the CLLJa.
Figure 2. CLLJa index defined as the areal average of the anomalies The Pearson correlation of the CLLJ index with the
for February 1950–2010, in the region bounded by 12.5∘ –17.5∘N and
∘ ∘ global and regional indexeswas estimated in order to deter-70 –80 W. The x-axis shows the temporal coordinates and the y-axis
the anomalies of the CLLJ in m s−1. The warm (cold) ENSO episodes mine and study what variability modes (besides ENSO)
are represented by dots (triangles). are important in explaining the CLLJ variability. The same
was done with the vertical structure index. In addition, to
explore the lead time between the time series, and so to
Table 1. February Pearson correlation coefficients (r) between
the CLLJa index and global and regional climate modula- determine which indexes are suitable for seasonal fore-
tors. Significance levels are 90% (italic), 95% (bold), and 99% casting of the wind over the Caribbean Sea, correlations
(bold italic). considering up to 5months in advance, from September to
January were calculated, and with the simultaneous cor-
Index r relation in February (perfect prognosis) to study the rela-
Niño 3.4 0.36 tionship with the anomalies of the CLLJ. The results for
PDO 0.22 both the intensity and vertical structure indexes are shown
PNA 0.35 in Tables 3 and 4.
NAO −0.21 Tables 3 and 4 present similar results. The Niño 3.4 SST
AMO −0.18 and the WHWP anomalies have the highest correlation
CAR 0.13
TNA 0.22 values, peaking in December. This result suggests that,
WHWP 0.31 by knowing the state of the SST in the Niño 3.4 region
MJO −0.40 by December or January, one could forecast the expected
CS (CF) 0.44 intensity and vertical structure of the jet in February.
AO −0.31 Other variability modes, such as PDO, PNA, NAO and
AO, display relatively high correlations in winter months,
specifically in February, concurrent with the jet maxi-
clearly differs from those conducted by, for example, Hoer- mum, making them suitable for diagnostic studies. The
ling et al. (1997), Gershunov and Barnett (1998a), and simultaneous correlations of the CLLJ anomaly index and
DeWeaver and Nigam (2002). However, it does capture the majority of the climate modulators considered in this
features that are important relative to the distribution of the study reinforce the use of the February monthly means for
ENSO and PDO events, such as the phase shift in La Niña the composite analysis.
events mentioned by DeWeaver and Nigam (2002). These Gershunov and Barnett (1998b) have documented that
authors also pointed out that the method used to calculate the influence of the PDO in the ENSO signal intensifies
the composites does not influence the results of estimating (reduces) the response to ENSO in winter, if both modes
the non-linear response of the ENSO events, being more are in phase (out of phase).Moreover, the PNA teleconnec-
important the distribution of the ENSO events. tion pattern is known to reach its mature phase in winter
Statistical significance was estimated in a way similar to and is highly correlated with ENSO events. The relation-
that of Gershunov and Barnett (1998a), using bootstrap- ship between the ENSO, PDO, and PNA is explored in the
ping without replacement. In this approach, an artificial next section using composites.
time series with normally distributed statistical noise is Wang (2007) has suggested that the NAO might be
generated. The anomalies would then be considered sig- associatedwith the fluctuations and formation of the CLLJ,
nificant at the 90% level if the absolute values exceed the because the NAO is the main mechanism associated with
5th or 95th percentiles. A similar relationship between the the pressure difference between the Icelandic Low and
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
Anomaly (m s–1)
T. MALDONADO et al.
Table 2. ENSO events in February classified according to ONI. The PDO information is from http://jisao.washington.edu/pdo (last
viewed 23 February 2015).
High PDO Low PDO
El Niño 1983, 1987, 1988, 1992, 1995, 1998, 2003, 2010 1953, 1954, 1958, 1959, 1964, 1966, 1969,
1973, 1977, 1978
La Niña 1985, 1989, 1996, 1999, 2000, 2001, 2006, 1950, 1951, 1955, 1956, 1974, 1975, 1976
2008, 2009
Neutral 1981, 1982, 1984, 1986, 1990, 1991, 1993, 1952, 1957, 1960, 1961, 1962, 1963, 1965,
1994, 1997, 2002, 2004, 2005, 2007 1967, 1968, 1970, 1972, 1979, 1980
Table 3. Pearson correlation between the CLLJ anomalies index and the most important global and regional climate modulators for
the Caribbean Basin. Significance levels are 90% (italic), 95% (bold), and 99% (bold italic).
Niño 3.4 PDO PNA NAO AMO CAR WHWP TNA AO
September 0.33 0.20 0.11 0.02 −0.20 0.07 0.17 0.01 0.06
October 0.36 0.18 0.12 0.00 −0.20 0.07 0.24 0.13 −0.07
November 0.39 0.07 0.12 0.09 −0.18 0.16 0.26 0.10 0.10
December 0.41 −0.08 0.12 −0.02 −0.18 0.16 0.33 0.09 0.21
January 0.39 −0.01 −0.05 0.15 −0.18 0.19 0.31 0.08 0.02
February 0.36 0.22 0.35 −0.21 −0.18 0.13 0.31 0.22 −0.31
the NASH. Table 1 presents estimated Pearson correlation America remains an open question that is not addressed
coefficients. The NAO and the AO show simultaneous here due to the low temporal resolution of the data series
significant negative correlation with the jet anomalies; used. Adequately resolving MJO would require weekly or
however, this result is expected since the close relationship daily data.
of these two indexes (Ambaum et al., 2001). A scatter plot The frequency and variability of CFs and cold surges
of these modes and the CLLJ index is shown in Figure 3 in arriving in the Gulf of Mexico, Central America, and the
order to help with the interpretation of their correlations. Caribbean region have been studied by Zárate-Hernández
The high dispersion of the dots reflects the relatively low (2013). He found that CF frequency and penetration into
correlations of the ENSO, PDO, PNA, NAO, and AO with northern South America are modulated by temperatures
the CLLJ index in February. in the Arctic Belt between 70∘ and 90∘N and by the AO
Regional climate indexes such as the CAR exhibit low phases. Using the results of Zárate-Hernández (2013), a
and no significant correlation with the CLLJa, while the CF index is defined to account for the number of CFs
TNA indexes exhibit a significant positive correlation only arriving in the Caribbean region in February from 1979
in February. On the other hand, theWHWPSSTs show sig- to 2010. The correlation between the CLLJa and the CFs
nificant positive correlation with the wind fluctuation dur- is significant and positive, meaning that if the number
ing the months considered, which reveals the importance of CFs increases, the jet intensity decreases. This find-
of the regional waters in the interaction with the wind; ing agrees with the fact that when a CF arrives in the
however, this interaction is not completely clear and needs Caribbean region, low SLPaweakens the trades, increasing
further analysis (Hidalgo et al., 2015). It should be noted the westerly anomalies and thus reducing the jet intensity.
that the WHWP is also strongly modulated by the influ- This correlation is also corroborated by the AO, which
ence of the ENSO SSTs (Wang and Enfield, 2003; Wang exhibits a significant negative correlation. According to
and Fiedler, 2006) and by the phases of the PNA and PDO Zárate-Hernández (2013), the number of CFs increases if
(Muñoz et al., 2010). the AO is in its negative phase. The latter shows a possible
The interaction of the MJO and CLLJ with the precipita- connection of NAO, AO, and CF with the CLLJ anomalies
tion over the Caribbean has already been studied byMartin that needs further study.
and Schumacher (2011a). They found that precipitation
anomalies are above (below) average in phases 1 and 2 4.1. El Niño composites
(5 and 6), and that the wind decreases (increases) in the The ENSO episode signals in the SLP are shown in
region of the low-level jet. In contrast, we found that the Figure 4. SLPa as shown in Figure 4(a) and (b) is in agree-
MJO index 7 is significantly negatively correlated with the ment with previous results, such as those of Hoerling et al.
CLLJa, meaning that a positive MJO phase 7 in Febru- (1997) and Gershunov and Barnett (1998b). During warm
ary would result in increased wind intensity and decreased episodes (Figure 4(a)), there are significant negative SLPa
precipitation if the mechanisms Martin and Schumacher over the Aleutian Peninsula in Alaska and the east coast
(2011a) examined are valid for the MJO phase 7 in winter. of North America (at the 95% and 99% confidence levels,
Note, however, that how this interaction affects the tem- respectively). In contrast, during cold events (Figure 4(b)),
poral and spatial distribution of precipitation over Central almost the opposite pattern is observed, with the SLP being
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
CARIBBEAN LOW-LEVEL JET AND ITS VARIABILITY
Table 4. Pearson correlation between the vertical structure anomalies index and the most important global and regional climate
modulators for the Caribbean Basin. Significance levels are 90% (italic), 95% (bold), and 99% (bold italic).
Niño 3.4 PDO PNA NAO AMO CAR WHWP TNA AO
September 0.41 0.27 0.11 0.03 −0.04 0.12 0.23 0.01 0.07
October 0.45 0.22 0.12 −0.02 −0.04 0.10 0.28 0.15 −0.10
November 0.48 0.11 0.12 0.11 −0.02 0.19 0.29 0.10 0.05
December 0.50 −0.01 0.13 0.01 −0.01 0.22 0.46 0.09 0.20
January 0.50 0.07 0.02 0.17 −0.01 0.25 0.46 0.09 0.00
February 0.49 0.31 0.38 −0.24 −0.01 0.18 0.48 0.27 −0.34
2 0 2
0 AO
−2
−4 −2
−4
2
1 0 1 2
0
NAO −1
−3 −2 −1 −2
−3
2 0 1 2
1
0 PNA
−1
−3 −2 −1 −2
−3
1 −1 0 1
0
−1 PDO −1
−3 −2 −1 −2
−3
2
1 2
1
Nino 0
−1 0 −1
2
0 1 2
1
0 U
−1
−3 −2 −1 −2
−3
Figure 3. Scatter plot matrix showing the relationship of the CLLJ anomalies with Niño 3.4, PDO, PNA, NAO, and AO. This matrix also shows the
association among each of these indexes.
higher than normal in the same regions at the 95% signif- east coast of the United States. The Gulf of Mexico and
icance level. Hoerling et al. (1997) noted that, during El the Caribbean region SLPa display linear behaviour, with
Niño, the centre of this SLPa pattern extends to the Tropi- a higher order of magnitude than the non-linear behaviour
cal Northern Hemisphere (TNH) teleconnection, while La for the same reason as explained above. Non-linear
Niña extends to the PNA teleconnection pattern. However anomalies are observed mainly over the Pacific coastal
the La Niña anomalies are shifted south over the Pacific and central United States (Figure 4(d)) and could be
and east over the Atlantic relative to the El Niño sig- associated with the southward shift of the SLPa in the
nal. Zárate-Hernández (2013) found evidence of a similar North Pacific during La Niña events. DeWeaver and
configuration of the SLPa associated with the low (high) Nigam (2002) found that the non-linear component of the
frequency and penetration of CFs in the Caribbean region. geopotential field is not negligible over the tropics. Same
The linear component of El Niño teleconnections composites were estimated for a geopotential height of 500
(Figure 4(c)) is dominant over the Aleutian Low and the hPa (not shown). Here, El Niño exhibits an asymmetric
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
T. MALDONADO et al.
60°N (a) 60°N (b)
30°N 30°N
8
0 0 6
4
2
0
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0 –2
60°N 60°N –4
(c) (d) –6
–8
30°N 30°N hPa
0 0
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
Figure 4. February composites of El Niño (a) and La Niña (b) used to estimate the mean SLP anomalies with NCEP-NCAR reanalysis data. The
linear component (c) is estimated as the difference between El Niño and La Niña anomalies, whereas the non-linear component (d) is the sum of
both. Contour levels are spaced 2 hPa apart. Shaded regions are significant anomalies with confidence levels of 90–95% (horizontal lines), 95–99%
(vertical lines), and >99% (slant lines).
60°N 60°N
(a) (b)
30°N 30°N
0 0
2 2
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
60°N 60°N
(c) (d)
30°N 30°N
0 0
2 2
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
Figure 5. February horizontal wind anomalies at 925 hPa for (a) El Niño, (b) La Niña, (c) El Niño – La Niña (linear anomalies), and (d) El Niño +
La Niña (non-linear anomalies). The reference vector is in m s−1.
pattern compared with La Niña pattern. Similar to the low observed in La Niña events in 1976–1977, reflecting the
SLPa levels (Figure 4(a)) during El Niño, two significant inherent variability of ENSO events. The linear compo-
troughs are located over the Alaskan Peninsula and over nent tends to dominate in the mid-latitudes between 30∘
the east coast of North America connected by a band of and 60∘N, with relatively important effects in the vicini-
low pressure. A ridge can be observed centred at 60∘N, ties of the Caribbean Sea and the Central Pacific. The
90∘W. During La Niña over the Alaskan Peninsula and non-linear component displays more interaction focused
the eastern part of North America, the geopotential height in mid-latitudes over the North Pacific to the Alaskan
anomalies both have an opposite sign to El Niño and Peninsula, the west coast, and northern North America.
are relatively lower in intensity. Two ridges separated by Composites of the horizontal wind anomalies at 925 hPa
a relatively strong trough are centred at 50∘N, 120∘W. during warm and cold episodes (Figure 5(a) and (b)) con-
This asymmetry was previously found by DeWeaver firm previous results reported by Amador et al. (2003) and
and Nigam (2002), who highlighted that the asymmetry Amador (2008). Warm (cold) episodes during boreal win-
observed in these cases could be related to the shift ter and lower (higher) than normal SLPa over the east
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
CARIBBEAN LOW-LEVEL JET AND ITS VARIABILITY
30°N 30°N
(a) (b)
20°N 20°N
10°N 10°N
0 0
110°W 100°W 90°W 80°W 70°W 110°W 100°W 90°W 80°W 70°W
30°N 30°N
(c) (d)
20°N 20°N
10°N 10°N
0 0
110°W 100°W 90°W 80°W 70°W 110°W 100°W 90°W 80°W 70°W
–120 –80 –40 0 40 80 120
mm month–1
Figure 6. February anomalies of cumulative monthly precipitation in millimetres (mm) for (a) El Niño, (b) La Niña, (c) El Niño – La Niña (linear
anomalies), and (d) El Niño+LaNiña (non-linear anomalies). Shaded regions are significant anomalies with confidence levels of 90–95% (horizontal
lines), 95–99% (vertical lines), and >99% (slant lines).
coast of the United States and over the Alaskan Penin- El Salvador, where precipitation anomalies are mostly
sula (described above) are associated with cyclonic (anti- negative in both phases. Durán-Quesada (2012) pointed
cyclonic) circulation at low levels, resulting in westerly out that these contrasting precipitation patterns arise due
(easterly) wind anomalies and weakening (enhancing) the to the relationship between the CLLJ and the ENSO
intensity of the CLLJ. However the anomalies over the phases. During warm (cold) events, the CLLJ is weakened
Gulf of Mexico and the Caribbean Sea are mainly dom- (enhanced), the moisture transport efficiency is weakened
inated by the linear (Figure 5(c)) rather than the non-linear (enhanced), and less (more) humidity is transported to
component (Figure 5(d)). This indicates that, at low atmo- Central America. Non-linear anomalies are observed in
spheric levels, El Niño signal behaves linearly over this the rainfall having the same order of magnitude as does
region, as do the jet anomalies during ENSO phases. the linear counterpart (Figure 6(d)), however, with no
The implication of these results for the precipita- statistical significance. These non-linear anomalies in the
tion distribution in Central America is as follows. rainfall result in wetter conditions over Costa Rica and the
Precipitation composites exhibit a dominant contrast- west coast of southern Mexico, and drier conditions over
ing pattern during El Niño and La Niña events. Drier the rest of Central America and Mexico. The cause of
conditions are found from the Yucatan Peninsula to north- the non-linear rainfall anomalies is difficult to determine;
ern South America during warm events, while Mexico however, from these results, it cannot be related to the
and Cuba experience wetter conditions (Figure 6(a)). non-linear signal in the horizontal wind over the Caribbean
During cold episodes, the inverse pattern of rainfall Sea or in the pressure pattern over the Atlantic, as the
anomalies is observed, with essentially the same spa- non-linear signal in those fields is very low relative to
tial distribution (Figure 6(b)) as during warm events, the linear counterpart. Nevertheless, comparing the mean
except for Yucatan Peninsula, Guatemala, Honduras, and precipitation anomalies with the divergence composites
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
T. MALDONADO et al.
40°N 40°N
(a) (b)
20°N 20°N
0 0
120°W 90°W 60°W 120°W 90°W 60°W
40°N 40°N
(c) (d)
20°N 20°N
0 0
120°W 90°W 60°W 120°W 90°W 60°W
–2 –1.6 –1.2 –0.8 –0.4 0 0.4 0.8 1.2 1.6 2
×10–6 s–1
Figure 7. As in Figure 5 but showing the divergence at 925 hPa. Contour levels are spaced every 0.4× 10−6 s−1.
(Figure 7), convergence (divergence) is observed over determine the effects of changes in the ENSO signal on
regions with drier (wetter) conditions during El Niño or SLP, horizontal wind at 925 hPa, and precipitation due
La Niña (Figure 7(a) and (b)). The linear component of the to the influence of the PDO in February. Because the
divergence dominates over continental Central America PDO is a long-lived phenomenon lasting approximately
(Figure 7(c)). The non-linear component (Figure 7(d)) 30–40 years, a 30-year running average (not shown)
makes a significant contribution over southern Central was applied to the PDO index described in Section 2.
America, even though the other fields analysed previ- This procedure reveals a negative PDO phase domi-
ously indicated that the non-linear component can be nating from 1950 to 1981 and a positive PDO phase
ignored in the tropics. However, the non-linearity of from 1981 to 2010. Table 2 lists the classification of
precipitation cannot be linked only to the divergence. El Niño and PDO phases. The anomalies are estimated
Other regional elements could control the precipitation relative to the climatology of the 61-year period from
distribution, such as the interaction of the mean wind with 1950 to 2010.
topography or other geographical elements present in the Figure 8(a) shows the SLPa for the El Niño and pos-
region. DeWeaver and Nigam (2002) pointed out that itive PDO (PDO+) sub-composite. Note the similarity
PDO might be responsible for some observed variability between the spatial distribution of the SLPa and the SLPa
in ENSO events, and in some cases such variability has pattern found for the El Niño composites (Figure 4(a)).
been taken erroneously as the non-linearity of El Niño Nevertheless, the SLPa are intensified and shifted south
episodes. In the following section, we explore the influ- relative to the El Niño case shown in Figure 4(a), while a
ence of the PDO on the ENSO teleconnection in Central lower-than-normal pressure band is located over Central
America. North America, conditions that enhance the Northern
Hemisphere storms that reach low latitudes (Gershunov
4.2. Influence of the PDO and Barnett, 1998b). This configuration of the SLPa
In the previous subsection, the ENSO signal in the spatial distribution was called a constructive pattern by
anomalies of fields such as SLP, horizontal wind at 925 Gershunov and Barnett (1998b). In addition, significant
hPa, and precipitation in February were examined using negative SLPa (at 95–99% confidence levels) are located
a composite analysis of ENSO events. Here, the effects near each coast of North America, along with cyclonic
of the PDO are examined by constructing sub-composites circulation that weakens the trades and the jet over the
of ENSO events and PDO phases. The objective is to Caribbean Sea (Figure 9(a)). Rainfall anomalies indicate
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
CARIBBEAN LOW-LEVEL JET AND ITS VARIABILITY
60°N 60°N
(a) (b)
30°N 30°N
0 0 10
8
6
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
60°N 60°N 4
(c) (d)
2
0
30°N 30°N
–2
–4
0 0
–6
–8
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0 –10
60°N 60°N
(e) (f)
hPa
30°N 30°N
0 0
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
Figure 8. February mean SLP anomalies for the El Niño and PDO phase sub-composites: (a) El Niño and PDO+, (b) La Niña and PDO+, (c) El
Niño and PDO−, (d) La Niña and PDO−, (e) neutral and PDO+, and (f) neutral and PDO−. Contours are spaced every 1 hPa. Shaded regions are
significant anomalies with confidence levels of 90–95% (horizontal lines), 95–99% (vertical lines), and >99% (slant lines).
drier-than-normal conditions over Central America and Atlantic). In the Atlantic, the trades are enhanced and, in
the Yucatan Peninsula and wetter-than-normal conditions addition, the wind over the Caribbean changes in direction
in southern Mexico (Figure 10(a)). Significant anomalies (becoming south-easterly, Figure 9(b)) compared with
in rainfall are found in south-western Mexico (positive), the constructive case. Precipitation anomalies display
Nicaragua, and Panama (negative). similar distributions but with wetter conditions in Central
The other constructive case is that of La Niña and neg- America (Figure 10(b)).
ative PDO (PDO−; Figure 8(d)). In this case, the spatial In the El Niño and PDO− case (Figure 8(c)), the SLPa
distribution of SLPa is similar to the SLPa pattern found pattern is also distorted. In the Pacific, the SLPa are located
in La Niña composites (Figure 4(b)). La Niña signal in the farther north and their extent and intensity are decreased.
SLPa is modulated by the PDO−, producing more intense In the Atlantic, the SLPa are located farther east and
and extended SLPa. This condition can also block storms their intensity is also reduced. The circulation linked to
from the north, preventing them from reaching southern the SLPa in the Atlantic does not greatly modify the
latitudes (Gershunov and Barnett, 1998b). Significant trades over the Caribbean (Figure 9(c)). Rainfall anoma-
SLPa over the eastern North America are accompanied lies (Figure 10(c)) display a distribution similar to that
by anticyclonic circulation, enhancing the trades and the of their constructive counterparts, though south-western
CLLJ (Figure 9(d)). Rainfall anomalies present conditions Mexico is drier than normal. Note that the rainfall anoma-
that are drier than normal over Mexico and wetter than lies are not statistically significant in Mexico or Central
normal over Central America and the Yucatan Peninsula America.
(Figure 10(d)). The influence of the PDO on the perturbations of the
In the combination of El Niño with PDO−, or of La Niña Central American climate is measured during neutral
with PDO+, the ENSO signal is modulated differently ENSO conditions (Figure 8(e) and (f)). The SLPa spa-
in comparison with the constructive cases mentioned tial distribution in the Atlantic is similar during each
above. Therefore, the condition in which ENSO and phase, but with opposite sign. Consequently, the mag-
PDO are out of phase is regarded as a destructive pat- nitude of the SLPa is less than when combined with
tern. Note that in Figure 8(b) the spatial distribution of ENSO events. On the other hand, negative SLPa are
the SLPa is different from that of the constructive case found over the Pacific during each PDO phase, though
(Figure 8(d)), the high-pressure anomaly centres being negative SLPa are intensified during PDO−. Moreover,
shifted to 160∘W (over the north Pacific) and 30∘W (north in each PDO phase, the observed SLPa do not produce
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
T. MALDONADO et al.
60°N 60°N
(a) (b)
30°N 30°N
0 0
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
60°N 60°N
(c) (d)
30°N 30°N
0 0
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
60°N 60°N
(e) (f)
30°N 30°N
0 0
180 150°W 120°W 90°W 60°W 30°W 0 180 150°W 120°W 90°W 60°W 30°W 0
2
Figure 9. February horizontal wind anomalies at 925 hPa for (a) El Niño and PDO+, (b) La Niña and PDO+, (c) El Niño and PDO−, (d) La Niña
and PDO−, (e) neutral and PDO+, and (f) neutral and PDO−. The reference vector is in m s−1.
marked changes in the horizontal wind at 925 hPa over different manner: negative anomalies are intensified over
the Caribbean, so the association between the PDO and Costa Rica, Panama, and South America, but significant
the perturbation in the CLLJ’s intensity and direction is positive anomalies are centred on 40∘N and 90∘W where
negligible or non-existent. Although the rainfall anomalies reduced precipitation is observed. In the neutral cases,
(Figure 10(e) and (f)) tend to be positive during PDO+ nevertheless, relevant changes happen only near the
(PDO−) phases in most of the study domain, these results equator over continental South America. Regarding the
are not statistically significant. influence of the PDO on rainfall, it is worth mentioning
The divergence field is shown in Figure 11. In the that the modulated divergence helps to explain some
constructive cases (Figure 11(a) and (d)), the divergence observed changes in the spatial distribution of precip-
anomaly pattern represents an intensification of the pattern itation, though there are other mechanisms related to
found during both the El Niño and La Niña phases, besides combinations of the PDO with PNA and/or El Niño.
having the opposite sign in each case. During El Niño (La These mechanisms are associated with changes in the
Niña) positive (negative) anomalies are observed mainly distribution of the latent and sensible heat flux anomalies,
over Costa Rica and Panama, and over the north-eastern modulating mainly the SST anomalies in the Gulf of
coast of South America, separated by negative (positive) Mexico (Muñoz et al., 2010) that could also be associ-
anomalies over Colombia and Venezuela. In the destruc- ated with changed precipitation distribution in the mid
tive cases, however, each combination behaves differently. latitudes or with the changed storm tracks observed in the
During the El Niño and PDO− case (Figure 11(c)), larger Northern Hemisphere by Gershunov and Barnett (1998a)
changes are observed from 15∘N towards the North Pole, and Zárate-Hernández (2013).
over continental North America, and in the equatorial
eastern Pacific. Note that the same region displays similar
behaviour in the changes in distribution of the precipitation 5. Summary and conclusions
anomalies (Figure 10(a) and (c)). The other destructive This paper presents a study of current knowledge about
case, i.e. La Niña and PDO+ (Figure 11(b)), occurs in a the variability of the CLLJ during winter (February)
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
CARIBBEAN LOW-LEVEL JET AND ITS VARIABILITY
30°N 30°N
(a) (b)
20°N 20°N
10°N 10°N
80
0 0 60
110°W 100°W 90°W 80°W 70°W 110°W 100°W 90°W 80°W 70°W
30°N 30°N
(c) (d) 40
20
20°N 20°N
0
10°N 10°N
–20
0 0 –40
110°W 100°W 90°W 80°W 70°W 110°W 100°W 90°W 80°W 70°W
30°N 30°N
–60
(e) (f)
–80
20°N 20°N
10°N 10°N
0 0
110°W 100°W 90°W 80°W 70°W 110°W 100°W 90°W 80°W 70°W
Figure 10. February anomalies of cumulative monthly precipitation in millimetres (mm) for (a) El Niño and PDO+, (b) La Niña and PDO−, (c) El
Niño and PDO−, (d) La Niña and PDO−, (e) neutral and PDO+, and (f) neutral and PDO−. Shaded regions are significant anomalies with confidence
levels of 90–95% (horizontal lines), 95–99% (vertical lines), and >99% (slant lines).
and about its impact on precipitation anomalies over the and AO. Results also indicate that, in winter, the CLLJ
Central American region. Using reanalysis data products interacts with regional-scale elements such as theMJO and
for horizontal wind at 925 hPa and SLP for the 61-year with cold surges originating in the Northern Hemisphere
period from 1950 to 2010, we study the large-scale that arrive in the Intra-Americas Sea Region. However,
environment of the CLLJ by calculating monthly mean the MJO reportedly exerts more influence on the weather
anomalies. GPCC precipitation data were used to estimate of North America during neutral years (Xue et al., 2002).
the anomalies of the monthly cumulative rainfall. A CLLJ The interaction between El Niño and the PDO was
anomaly index at 925 hPa was defined as the areal average studied using composite analysis. During warm (cold)
of the February monthly means for the region bounded events, the SLPa align with the PNA teleconnection pat-
by 12.5–17.5∘N and 70–80∘W. We found that the CLLJa tern, and cyclonic (anticyclonic) circulation is observed,
is significantly correlated with El Niño, the PDO, and the resulting in westerly (easterly) wind over the Caribbean
PNA, and significantly negative correlated with the NAO Sea, weakening (enhancing) the jet intensity. The PDO
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
T. MALDONADO et al.
40°N 40°N
(a) (b)
20°N 20°N
0 0
120°W 90°W 60°W 120°W 90°W 60°W
40°N 40°N
(c) (d)
20°N 20°N
0 0
120°W 90°W 60°W 120°W 90°W 60°W
40°N 40°N
(e) (f)
20°N 20°N
0 0
120°W 90°W 60°W 120°W 90°W 60°W
–1.6 –0.8 0 0.8 1.6
×10–6 s–1
Figure 11. February divergence anomalies at 925 hPa for (a) El Niño and PDO+, (b) La Niña and PDO+, (c) El Niño and PDO−, (d) La Niña and
PDO−, (e) neutral and PDO+, and (f) neutral and PDO−. Contours levels are spaced every 0.4× 10−6 s−1.
enhances or attenuates the ENSO signal in the SLPa (i.e. needs further study. Despite large changes in the
during El Niño events concomitant with a positive PDO, circulation pattern found due to the influence of the PDO,
the PNA teleconnection pattern is enhanced), whereas changes in the monthly cumulative precipitation anomaly
during El Niño in combination with negative PDO, the were minimal over the Central American Isthmus.
PNA pattern is not observed, resulting in no significant The small perturbations in rainfall were related to small
modification of the circulation over the Caribbean Sea. fluctuations in the divergence in the tropical regions of the
Moreover, changes in the circulation due to the PDO American continent.
phases during La Niña episodes are observed in the The interactions found among the different climate
direction change of the meridional wind anomalies in elements studied here emphasise the current need for
the Caribbean Basin. Cold events plus positive (negative) integrated analysis of the Central American climate.
PDO result in an anomalous northward (southward) wind However, such analysis presents many difficulties, as
component. Due to the Rossby problem adjustment, the meteorological measurements are unfortunately scarce in
relationship between the SLPa and the CLLJ anomalies this region. Reanalysis, on the other hand, is not sufficient
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
CARIBBEAN LOW-LEVEL JET AND ITS VARIABILITY
for several reasons: (1) the temporal and spatial resolution high circulation. Int. J. Climatol. 28(3): 343–350, doi: 10.1002/
of the data are not necessarily the same or sufficient to joc.1600.
Gershunov A, Barnett T. 1998a. ENSO influence on intraseasonal
capture the main characteristics of the climate elements; extreme rainfall and temperature frequencies in the contiguous United
(2) the time series in the databases are limited in duration; States: observations and model results. J. Clim. 11(7): 1575–1586,
and (3) removing the noise from the climatology data is doi: 10.1175/1520-0442(1998)011<1575:EIOIER>2.0.CO;2.
Gershunov A, Barnett T. 1998b. Interdecadal modulation of ENSO
difficult. This study, however, contributes significantly teleconnections. Bull. Am. Meteorol. Soc. 79(12): 2715–2725, doi:
to our understanding of the atmospheric dynamics of 10.1175/1520-0477(1998)079<2715:IMOET>2.0.CO;2.
the Intra-Americas Sea Region, regarding identifying Gill AE. 1982.Atmosphere–OceanDynamics. Academic Press: London.
Herrera E,MagañaV, Caetano E. 2015. Air–sea interactions and dynam-
and establishing the relationships between the CLLJ ical processes associated with the midsummer drought. Int. J. Clima-
fluctuations in winter and the main climate variability tol. 35: 1569–1578, doi: 10.1002/joc.4077.
modes. Hidalgo HG, Durán-Quesada AM, Amador JA, Alfaro EJ. 2015.
The Caribbean low-level jet, the inter-tropical convergence zone
and precipitation patterns in the Intra-Americas Sea: a proposed
dynamical mechanism. Geogr. Ann. Ser. A: Phys. Geogr. 97(1):
Acknowledgements 41–59.
Hoerling M, Kumar A, Zhong M. 1997. El Niño, La Niña, and the
This study was financed by the Swedish International nonlinearity of their teleconnections. J. Clim. 10(8): 1769–1786, doi:
Development Agency (SIDA) and the International Sci- 10.1175/1520-0442(1997)010<1769:554 ENOLNA>2.0.CO;2.Holton JR. 2004. An Introduction to Dynamic Meteorology. Academic
ence Programme (ISP). The authors would like to thank the Press: San Diego, CA.
Centre for Natural Disaster Science (CNDS) in Uppsala Janowiak J, Gruber A, Kondragunta C, Livezey R, Huffman G.
University, the National Centers for Environmental Pre- 1998. A comparison of the NCEP-NCAR reanalysis precipitation
and the GPCP rain gauge-satellite combined dataset with obser-
diction (NCEP), and the National Center for Atmospheric vational error considerations. J. Clim. 11(11): 2960–2979, doi:
Research (NCAR) for the reanalysis data. We would also 10.1175/1520-0442(1998)011<2960:ACOTNN>2.0.CO;2.
like to acknowledge support via project VI-805-A9-532, Jury M, Malmgren BA. 2012. Joint modes of climate variability
across the inter-Americas. Int. J. Climatol. 32: 1033–1046, doi:
CIGEFI-UCR, provided via a SIDA-CSUCA agreement. 10.1002/joc.2324.
Jury M, Malmgren BA, Winter A. 2007. Subregional precipita-
tion climate of the Caribbean and relationships with ENSO and
NAO. J. Geophys. Res. Atmos. 112: D16107, doi: 10.1029/2006
References JD007541.
Alfaro E. 2002. Some characteristics of the annual precipitation cycle in Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L,
Central America and their relationships with its surrounding tropical Iredell M, Saha S, White G, Woollen J, Zhu Y, Leetmaa A, Reynolds
oceans. Top. Meteorol. Oceanogr. 9: 109–119. R, Chelliah M, Ebisuzaki W, Higgins W, Janowiak J, Mo KC,
Amador JA. 1998. A climatic feature of the tropical Americas: the trade Ropelewski C, Wang J, Jenne R, Joseph D. 1996. The NCEP/NCAR
wind easterly jet. Top. Meteorol. Oceanogr. 5: 91–102. 40-year reanalysis project. Bull. Am. Meteorol. Soc. 77: 437–471, doi:
Amador JA. 2008. The Intra-Americas sea low-level jet. Ann. N. Y. Acad. 10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Sci. 1146: 153–188, doi: 10.1196/annals.1446.012. Magaña V, Amador JA, Medina S. 1999. The midsummer drought
Amador JA, Chacón JR, Laporte S. 2003. Climate and climate vari- over Mexico and Central America. J. Clim. 12: 1577–1588, doi:
ability in the Arenal basin of Costa Rica. In Climate, Water and 10.1175/1520-0442(1999)012<1577:TMDOMA>2.0.CO;2.
Trans-Boundary Challenges in the Americas, Díaz H, Morehouse Martin ER, Schumacher C. 2011a. Modulation of Caribbean precipita-
B (eds). Kluwer Academic Publishers: Holland, The Netherlands, tion by the Madden-Julian Oscillation. J. Clim. 24(3): 813–824, doi:
317–349. 10.1175/2010JCLI3773.1.
Amador JA, Alfaro EJ, Lizano OG, Magaña VO. 2006. Atmospheric Martin ER, Schumacher C. 2011b. The Caribbean low-level jet and its
forcing of the eastern tropical Pacific: a review. Prog. Oceanogr. relationship with precipitation in IPCC AR4 models. J. Clim. 24(22):
69(2–4): 101–142, doi: 10.1016/j.pocean.2006.03.007. 5935–5950, doi: 10.1175/JCLI-D-11-00134.1.
Amador JA, Alfaro EJ, Rivera ER, Calderón B. 2010. Climatic features Muñoz E, Enfield D. 2011. The boreal spring variability of the
and their relationship with tropical cyclones over the Intra-Americas Intra-Americas low-level jet and its relation with precipitation and
seas. In Hurricanes and Climate Change, Elsner JB, Hodges RE, tornadoes in the eastern United States. Clim. Dyn. 36: 247–259, doi:
Malmstadt JC, Scheitlin KN (eds). Springer: The Netherlands, 10.1007/s00382-009-0688-3.
149–173. Muñoz E, Busalacchi AJ, Nigam S, Ruiz-Barradas A. 2008. Winter
Ambaum MHP, Hoskins BJ, Stephenson DB. 2001. Arctic Oscilla- and summer structure of the Caribbean low-level jet. J. Clim. 21(6):
tion or North Atlantic Oscillation? J. Clim. 14: 3495–3507, doi: 1260–1276, doi: 10.1175/2007JCLI1855.1.
10.1175/1520-0442(2001)014<3495:AOONAO>2.0.CO;2. Muñoz E, Wang C, Enfield D. 2010. The Intra-Americas sea springtime
CookKH, Vizy EK. 2010. Hydrodynamics of the Caribbean low-level jet surface temperature anomaly dipole as fingerprint of remote influence.
and its relationship to precipitation. J. Clim. 23(6): 1477–1494, doi: J. Clim. 23: 43–56, doi: 10.1175/2009JCLI3006.1.
10.1175/2009JCLI3210.1. Ranjha R, Svensson G, Tjerström M, Semedo A. 2013. Global dis-
Dee DP, Uppala SM, Simmons AJ, Berrisford P, Poli P, Kobayashi S, tribution and seasonal variability of coastal low-level jets derived
Andrae U, Balmaseda MA, Balsamo G, Bauer P, Bechtold P, Bel- from ERA-Interim reanalysis. Tellus A 65: 20412, doi: 10.3402/tel-
jaars ACM, van de Berg L, Bidlot J, Bormann N, Delsol C, Dragani lusa.v65i0.20412.
R, Fuentes M, Geer AJ, Haimberger L, Healy SB, Hersbach H, Hólm Schneider U, Becker A, Finger P, Meyer-Christoffer A, Rudolf B, Ziese
EV, Isaksen L, Kållberg P, Köhler M, Matricardi M, McNally AP, M. 2011. GPCC full data reanalysis version 6.0 at 0:5: monthly
Monge-Sanz BM, Morcrette J-J, Park B-K, Peubey C, de Rosnay P, land-surface precipitation from rain-gauges built on GTS-based and
Tavolato C, Thépaut J-N, Vitart F. 2011. The ERA-Interim reanaly- historic data, doi: 10.5676/DWD_GPCC/FD_M_V6_050.
sis: configuration and performance of the data assimilation system. Taylor MA, Alfaro EJ. 2005. Climate of Central America and the
Q. J. R. Meteorol. Soc. 137: 553–597, doi: 10.1002/qj.828. Caribbean. In Encyclopedia of World Climatology. Encyclo-
DeWeaver E, NigamS. 2002. Linearity in ENSO’s atmospheric response. pedia of Earth Sciences Series, Oliver J (ed). Springer: The
J. Clim. 15: 2446–2461. Netherlands, 183–189, doi: 10.1007/1-4020-3266-837, Bodmin,
Durán-Quesada AM. 2012. Sources of Moisture for Central America Cornwall (GB).
and Transport Based on a Lagrangian Approach: Variability, Con- Wang C. 2007. Variability of the Caribbean low-level jet and its relations
tributions to Precipitation and Transport Mechanisms. PhD thesis, to climate. Clim. Dyn. 29: 411–422.
Universidade de Vigo, Ourense, Spain. Wang C, Enfield DB. 2003. A further study of the tropical west-
Gamble DW, Parnell DB, Curtis S. 2008. Spatial variability of the ern hemisphere warm pool. J. Clim. 16: 1476–1493, doi:
Caribbean midsummer drought and relation to north Atlantic 10.1175/1520-0442-16.10.1476.
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
T. MALDONADO et al.
Wang C, Fiedler PC. 2006. ENSO variability and the eastern trop- ENSO-neutral and weak ENSO winter. A Workshop on Prospects for
ical Pacific: a review. Prog. Oceanogr. 69(2–4): 239–266, doi: Improved Forecast and Weather Short-Term Climate Variability on
10.1016/j.pocean.2006.03.004. Subseasonal (2 weeks to 2 months) Times Scales, NASA/Goddard
Wang C, Lee S-K. 2007. Atlantic warm pool, Caribbean low-level jet, Space Flight Center, Greenbelt, Maryland.
and their potential impact on Atlantic hurricanes. Geophys. Res. Lett. Zárate-Hernández E. 2013. Climatología de masas invernales de aire frío
34: L02703, doi: 10.1029/2006GL028579. que alcanzan Centroamérica y el Caribe y su relación con algunos
Xue Y, Higgins W, Kousky V. 2002. Influences of the Madden Julian índices árticos. Top. Meteorol. Oceanogr. 12(1): 35–55.
Oscillation on temperature and precipitation in North America during
© 2015 Royal Meteorological Society Int. J. Climatol. (2015)
Paper II
Manuscript prepared for Adv. Geosci.
with version 2015/04/24 7.83 Copernicus papers of the LATEX class copernicus.cls.
Date: 7 September 2015
Interannual variability of the midsummer drought in
Central America and the connection with sea surface
temperatures
Tito Maldonado1,2,4, Anna Rutgersson2, Eric Alfaro3,4,5, Jorge Amador3,4, and
Björn Claremar2
1Center for Natural Disaster Sciences, Uppsala University, Villav. 16, 752 36, Uppsala, Sweden
2Department of Earth Sciences, Uppsala University, Villav. 16 752 36, Uppsala, Sweden
3School of Physics, University of Costa Rica, San Pedro de Montes de Oca, 11501-2060 San Jose,
Costa Rica
4Center for Geophysical Research, University of Costa Rica, San Pedro de Montes de Oca,
11501-2060 San Jose, Costa Rica
5Centre for Research in Marine Sciences and Limnology, University of Costa Rica, San Pedro de
Montes de Oca, 11501-2060 San Jose, Costa Rica
Correspondence to: Tito Maldonado (tito.maldonado@geo.uu.se)
Abstract. The midsummer drought (MSD) in Central America is characterised in order to create
annual indexes representing the timing of its phases (start, minimum and end), and other features
relevant for MSD forecasting such as the intensity and the magnitude. The MSD intensity is defined
as the minimum rainfall detected during the MSD, meanwhile the magnitude is the total precipitation
divided by the total days between the start and end of the MSD. It is shown that the MSD extends
along the Pacific coast, however, a similar MSD structure was detected also in two stations in the
Caribbean side of Central America, located in Nicaragua. The MSD intensity and magnitude show
a negative relationship with El Niño 3.4 and a positive relationship with the Caribbean low-level jet
(CLLJ) index, however for the Caribbean stations the results were not statistically significant, which
is indicating that other processes might be modulating the precipitation during the MSD over the
Caribbean coast. On the other hand, the temporal variables (start, minimum and end) show low and
no significant correlations with the same indexes.
The results from canonical correlation analysis (CCA) show good performance to study the MSD
intensity and magnitude, however, for the temporal indexes the performance is not satisfactory due
to the low skill to predict the MSD phases. Moreover, we find that CCA allows producing forecasts
of the MSD intensity and magnitude using sea surface temperatures (SST) with leading times of up
to 3 months. Using CCA as diagnostic tool it is found that during June, an SST dipole pattern upon
the neighbouring waters to Central America is the main variability mode controlling the inter-annual
variability of the MSD features. However, there is also evidence that the regional waters are playing
an important role in the annual modulation of the MSD features. The waters in the PDO vicinity
1
might be also controlling the rainfall during the MSD, however, exerting an opposite effect at the
north and south regions of Central America.
1 Introduction
The geographical features of Central America imprint the characteristics of the regional climate and
weather at the isthmus. The region is conformed by a large and high mountain system surrounded
by the Pacific and Atlantic oceans, which induces the maritime climate conditions governing in the
region (Taylor and Alfaro, 2005). The interaction between the easterly winds (trades), which in turns
are the dominant wind regime, and topography divide the region into two different climate areas: the
Pacific located leeward of the main wind and the Caribbean presenting windward conditions.
The annual rainfall cycle for the entire region has already been well documented for Central Amer-
ica (Alfaro, 2002), and for the Eastern Tropical Pacific (ETP, Magaña et al. 1999; Amador et al.
2006). In the Pacific region the annual precipitation cycle exhibits a bimodal behaviour (Magaña
et al., 1999; Taylor and Alfaro, 2005; Amador et al., 2006). The first precipitation maximum occurs
during May-June when the nearby ocean waters have warmed to around 29 ◦C over the eastern Pa-
cific warm pool, and the easterly winds have decreased, enhancing convection in the region. During
July-August a relative precipitation minimum is observed, along with a decrease in the sea surface
temperatures (SSTs) over the eastern Pacific warm pool, and an increase of the easterly winds. This
reduction in the rainfall is known as the midsummer drought (MSD, Magaña et al. 1999; Amador
et al. 2006). The second precipitation maximum occurs during August-September-October (ASO)
and is accompanied by a reduction of the trades and a relative increase of the SSTs up to 28.5 ◦C
over the eastern Pacific warm pool. It is also during this quarter of the year that the hurricane season
is more active in the Tropical North Atlantic (Amador et al., 2006).
On the other side, upon the Caribbean coast of Central America the annual precipitation cycle
contrasts the one in the Pacific mainly during the boreal winter months, when wetter and more humid
conditions are found over the eastern coast of Central America (Taylor and Alfaro, 2005; Amador
et al., 2006). A similar bimodal cycle, however, has been extensively reported over the Greater
Antilles (Chen and Taylor, 2002; Taylor et al., 2002; Spence et al., 2004; Ashby et al., 2005). The
reduction in rainfall experienced during the summer months over the western Caribbean is argued
to have a similar pattern to that in the Pacific (Martin and Schumacher, 2011). Little is known about
the origins of the MSD in both, the Pacific and Caribbean. The processes involved in such rainfall
decrease are most probably different to those operating in the western Central America and Mexico,
however, both MSDs appear to share the Caribbean low-level jet (CLLJ) as main process increasing
the moisture flux in the Caribbean, which suppresses convection and decreases rainfall (Magaña
et al., 1999; Wang, 2007; Muñoz et al., 2008; Amador, 2008; Whyte et al., 2008). In this study, we
2
focus on the variability of the MSD in the Pacific, nevertheless, we also examine two stations in the
Caribbean side, to contrast the features present in both MSDs.
The development of the MSD has been explained in terms of the interaction of changes in the
divergent (convergent) low-level winds over the warm waters nearby Mexico and Central America
(Magaña et al., 1999). Recent studies such as Karnauskas et al. (2013) and Herrera et al. (2014)
have proposed different hypothesis about the origin of the MSD. Karnauskas et al. (2013) argue that
the MSD originates as a response to one single precipitation-enhancing mechanism occurring twice,
contrary to a suppressing mechanism. The latitudinal dependence in the two peaks of precipitation
surge as response to the biannual crossing of the solar declination, which leads to the two peaks in
convective instability and hence rainfall. This hypothesis, nevertheless, does not explain the almost
simultaneous occurence of the MSD in both southern Mexico and Central America (Magaña et al.,
1999). On the other hand, Herrera et al. (2014) describe the origins of the MSD studying the air-sea
interaction between the CLLJ (Amador, 1998, 2008; Muñoz et al., 2008; Maldonado et al., 2015)
and the ocean waters in the neighbourhood of the Pacific coast. Their findings show that the CLLJ
peaks in July, provokes a cooling of the SSTs over the Pacific and a westward shift of low-level
moisture convergence that combined with subsidence produce the MSD.
The understanding of the variability of the MSD is still a scientific challenge (Herrera et al., 2014),
due to its high spatial and temporal variability (Amador et al., 2006; Amador, 2008). Previous stud-
ies have analysed the influence of the SST anomalies in the precipitation regime in Central America.
Enfield and Alfaro (1999) have shown that the magnitude of the precipitation during the rainy sea-
son (from May to October) in Central America is highly modulated by the SST anomalies at both
oceans. Alfaro (2007) and Fallas-López and Alfaro (2012b) studied the predictability of the rainfall
using canonical correlation analysis (CCA) with SST as predictor. Both studies found that the SST
modify the response on precipitation in different ways during each maximum, the first rainfall peak
being modulated mainly by the SST over the tropical Atlantic, and the secondary maximum being
controlled by a dipole in the SST anomalies over the neighbour oceans to the Central American
coast. Moreover, Fallas-López and Alfaro (2012a) found that the MSD intensity is dependent on
several global climate indexes such as the Atlantic Multi-Decadal Oscillation (AMO), Niño 3 (N3),
and Pacific Decadal Oscillation (PDO). Maldonado and Alfaro (2011) and Maldonado et al. (2013)
utilised CCA to explore the distribution of precipitation extreme events during the second precipita-
tion maximum finding that during the second rainfall maximum, the total precipitation and temporal
distribution of precipitation events are controlled by a dipole in the SSTs between the Pacific and
the Caribbean-North Atlantic waters. Given the above evidence, it would be expected that SST is
involved in the MSD variability at inter-annual scales. In fact, Alfaro (2014) found that warm (cool)
conditions of Niño 3.4 tend to be associated to drier (wetter) MSD events in some regions of the
North Pacific and Central Valley of Costa Rica.
3
Other climate processes occur simultaneously with the MSD, and could also alter the precipitation
during June-August (JJA) e.g the intensification of the North Atlantic Sub-tropical High (NASH,
Wang 2007), the Western Hemisphere Warm Pools (WHWM, Wang and Enfield 2001, 2003), the
North AmericanMonsoon (Vera et al., 2006) and the CLLJ (Amador, 1998, 2008). Recently, Hidalgo
et al. (2015) also found that the latitudinal position of the precipitation centre of mass is correlated
with the intensity of the CLLJ, i.e a weaker (stronger) jet is associated to the northward (southward)
shift of the latitudinal precipitation centre of mass, resulting in rainfall above normal during JJA that
might be also associated with variations of the MSD events.
A better understanding of the variability of the MSD would then contribute to the knowledge
about the weather and climate dynamics in the region, also to the information given to the economic
and social sectors. These sectors, such as water resources, agriculture, and hydro-electrical power
production are benefited in terms of obtaining better data for planning, mitigation and prevention to
severe dry conditions, e.g. the droughts reported in the North Pacific region of Costa Rica during
July 2014 (Chinchilla-Ramírez, 2014). Therefore, the targets of this study are first to contribute to
the understanding of the climate in the region, and second to provide a systematic method for MSD
forecasting.
This study consists of an analysis of the connection between the features of the MSD observed
in the Pacific and Caribbean coast of Central America and the anomalies in the SST of the regional
waters by means of a combination of principal component analysis (PCA) and canonical correlation
analysis (CCA). The precipitation observations, however, are first examined in order to determine
the data quality and to characterised the MSD per station. The MSD is described in terms of start,
timing, end, duration, depth (intensity) and total precipitation (magnitude). This representation is
also used to produce annual indexes of those variables. Previous reports have already characterised
the existence of the MSD in Costa Rica (Ramírez, 1983; Hernández, 2013; Alfaro, 2014; Solano,
2015), nonetheless, in this study the portray is expanded to include the entire the region.
This manuscript is organised as follows: in Section 2 is described all the information relevant
to the databases used in this study. Section 3 is a explanation of the methodology applied in this
analysis. In Section 4 the result and Section 5 the discussion and concluding remarks are presented.
2 Data
A total of 25 gauge stations with daily observations of precipitation provided by the meteorological
services in Central America are used. Their location is shown in Fig. 1a, and the information about
the coordinates of each station is given in Table 1. Note that the majority (23) of the stations are
situated along the Pacific coast with only two stations located along the Caribbean coast. The two
Caribbean stations possess relevant features to describe the precipitation behaviour in the Caribbean
slope, such as their location at the CLLJ exit and their contrasting annual precipitation cycle to
4
7.5
●
5.0 ● ●
●
● ●
● ●● ●
2.5 ●
● ● ●
●
●●
0.0 ●● ●●●
●
7.5 ●
5.0
−95 −90 −85 −80 −75
(a)
(b)
Figure 1. a) Spatial distribution of the gauge stations in Central America. b) SST domain used as predictor
for CCA. The area captures the most important climate modulators for Central America, see Maldonado et al.
(2013).
that in the Pacific. Thus, in order to achieve a better representation of the complete region, we
consider the Caribbean stations for the analysis. The gaps in the time-series are filled using the
methodology described in Alfaro and Soley (2009), which uses autoregressive models of order 1.
This method can filter persistent signals comparable to the length of the filter, and the estimated
values are consistent with the statistical properties of the time series without external superposition
of the data. The precipitation time series length is from 1961 to 2012.
5
Table 1.Geographical position of the gauge stations in degrees north for latitude and degrees west for longiutde.
Station LAT LON
La Argentina 10.03 84.35
Fabio Baudrit 10.00 84.43
Juan Santa Maria 10.00 84.17
Liberia 10.58 85.58
Nicoya 10.15 85.45
Santa Cruz 10.17 85.25
Bagaces 10.53 85.25
CIGEFI 9.94 84.04
Bluefields 12.02 83.78
Ocotal 13.62 86.47
Chinandega 12.63 87.13
Juigalpa 12.10 85.37
Managua 12.14 86.16
Puerto Cabezas 14.05 83.38
Rivas 11.44 85.83
David 8.40 82.43
Los Santos 7.42 80.42
San Miguel 13.43 88.15
Asuncion Mita 14.33 89.71
Huehuetenango 15.32 91.50
LaborOvalle 14.87 91.48
La Fragua 14.96 89.58
Amapala 13.29 87.65
Choluteca 13.32 87.15
Tegucigalpa 14.06 87.22
The extended reconstructed sea surface temperatures (ERSSTv3b, Xue et al. 2003; Smith et al.
2007) are used in this study. The SST anomalies are constructed using a combination of observed
data along with models and historical sampling grids. This global database has a horizontal resolu-
tion of 2.5 by 2.5 degrees. The domain bounded by -22◦ to 63◦ N and 111◦ to 15◦ E (Fig. 1b) is
considered in order to capture the signal of the most important climate variability modes for the Cen-
tral American isthmus such as El Niño Southern Oscillation (ENSO), the Pacific Decadal Oscillation
(PDO), the Atlantic Multi decadal Oscillation (AMO), the North Atlantic Oscillation (NAO), and the
Tropical North Atlantic (TNA), which in turns, have shown to be relevant in terms of variability of
rainfall during the season JJA. The SST anomalies are used as predictors in the CCA models.
6
The Niño 3.4 index (Trenberth, 1997) provided by the International Research Institute for Climate
and Society (IRI, http://iridl.ldeo.columbia.edu/, last view 2015/07/14) is used to estimate the rela-
tionship between this index and relevant features of the MSD by means of empirical contingency
tables for conditional data.
Horizontal wind data at 925 hPa is provided by the NCEP/NCAR reanalysis (Kalnay et al., 1996;
Kistler et al., 2001) which has a horizontal resolution of 2.5 by 2.5 degrees. The wind data is used to
calculate the CLLJ index as in Amador (2008) and Maldonado et al. (2015).
3 Methods
Based on the methodology developed in Alfaro (2014), and also applied by Solano (2015), the daily
precipitation times series are filtered using a running triangular weight average with a window of
31 days, to avoid or minimise interruptions of the MSD due to weakening of the trades and/or
approaching of the ITCZ, as suggested by Ramírez (1983) and Alfaro (2014). An algorithm to sys-
tematically identify the features of the MSD is applied to the filtered daily precipitation time series.
This algorithm seeks for the timing of each phase of the MSD (start, minimum, and end), besides
the intensity and magnitude of the MSD. Notice that this algorithm provides an annual value of
each quantity, which are used as indexes to characterise the MSD later in this study. The start of the
MSD is considered as the moment when the decrease in precipitation initiates, usually after May-
Jun. The end of the MSD is when the precipitation stops increasing, normally taking place around
September-October. The minimum occurs in between the start and end of the MSD. The MSD inten-
sity is defined as the minimum rainfall detected during the MSD (or the depth of the valley in Fig.
2), meanwhile the magnitude is the total precipitation divided by the total number of days between
the start and end of the MSD. The description of the whole MSD detecting process is explained as
follows: first, the algorithm scans for the precipitation minimum in the filtered time series within a
reasonable range of days for the existence of the MSD, determined by the climatologies of each sta-
tion, i.e. typically from 1 June to 30 August. If the precipitation minimum falls outside this period,
it is not considered a MSD event. Second, the nearest inflection points to the minimum are sought
to determine the start and the end of the MSD. The shortest distance allowed between the inflection
points and the minimum is 5 days. Consequently, the shortest MSD would have a duration of 10
days. Moreover, the inflection points have the restriction that the difference of precipitation between
them and the minimum should be at least 20% per day. It is worth mentioning, that not all the years
present the inflection points and the minimum. For the scope of this study, hence those years missing
any of the MSD phases are removed. The causes for those fails in the detection of the MSD phases
are several, however, they can be attributed to anomalous years with a extended dry season at the be-
ginning of the year, combined with an earlier and severe MSD, or anomalous drier condition during
the secondary precipitation maximum, years with no MSD at all, and in most cases due to missing
7
C
●
A
●
B
●
160 180 200 220 240 260 280 300
Days of year
Figure 2. Filtered daily time series for the observed precipitation in station La Argentina during 1990. The dots
represent the beginning (A), minimum (B) and end (C) of the MSD. In this year the MSD started by day 195
(14 July), reached the minimum at day 239 (27 August) and ended by day 280 (7 Oct).
data in the observations. Figure 2 shows the schematics of this algorithm. A statistical summary of
the features of the MSD is shown in Fig. 3 and 4.
The contingency tables method seeks for a significant predictive relationship between two vari-
ables, being one independent and the other dependent (Alfaro et al., 2003; Fallas-López and Alfaro,
2012a). The time series are divided into three categories: below normal, normal and above normal
(BN, N, AN, respectively) using the percentiles 33rd and 67th. Fallas-López and Alfaro (2012a)
have shown that using these categories provides a good measure of the possible states of El Niño
and the MSD indexes, besides, the information obtained is easy to use for the public. Thus, the con-
tingency tables are estimated for conditional events to study the dependence of the MSD features
(intensity and magnitude) with ENSO events (independent variable). As mentioned above, the years
in which the algorithm does not detect any of the phases or no MSD at all, are dismissed. The sta-
tistical significance is estimated by means of bootstrapping with no replace (Gershunov and Barnett,
1998). With bootstrapping, a synthetic time series normally distributed is generated from random
noise, then, we estimated contingency tables for the artificial time series. This process is repeated
several times to obtain a number of random frequencies normally distributed, which in this case is
1000 repetitions, thus, we obtain 1000 contingency tables per station. If the value of the observed
absolute frequency is greater than the 90th of the synthetic frequencies, then the observation is sig-
8
daily precipitation (mm)
0 5 10 15 20
nificant at 0.1 significance level. The same can be said for the 0.05 and 0.01 of significance level.
Moreover, an estimated of the Pearson correlation between the MSD intensity (magnitude) and the
Niño 3.4 index are shown in Fig. 5.
As a tool to investigate the variability of the MSD, we use Canonical correlation analysis (CCA,
Wilks 2011) which is a statistical technique that searches for pairs of patterns in two multivariate
data sets (fields), and constructs sets of transformed data variables by projecting the original data
onto those patterns. These new variables maximise the interrelationships between the two fields.
The new variables can be used, analogously to the regression coefficient in the multiple regression.
The new variables or vector weights are also known as canonical vectors, and their projection with
their respective fields as canonical variates. CCA can be useful in different applications, such as: i)
to obtain diagnostic aspects of the coupled variability of two fields, in the case when the time series
of observations of the two fields are simultaneous, or ii) to perform statistical forecasts, in the case
when the time series of observations of one field precede the other (e.g. Alfaro 2007; Maldonado
and Alfaro 2011; Fallas-López and Alfaro 2012b; Maldonado et al. 2013).
In this study CCA is used as a diagnostic tool, however, models for statistical forecasting can be
built for future use. Consequently, CCA is employed in order to analyse the relationship between the
SSTs monthly anomalies (SSTA, also denoted field X) with each of the indexes (start, minimum,
end, intensity and magnitude of MSD) estimated to represent each of the features of the MSD (each
index would be the fieldY , for the corresponding model), that is, we seek for a statistical relationship
between the large and regional scale features of the SSTA and the characteristics of the MSD in each
of the stations (local scale). The above CCA methodology is based on Maldonado et al. (2013) and
it is implemented as follows: the fields (SSTA, MSD indexes) are first reduced by means of principal
component analysis (PCA) to assure stability in the CCA parameters. A maximum of 17 EOFs
and CCA modes in the filtering stage are allowed. This threshold was suggested by Gershunov and
Cayan (2003) and Alfaro (2007) to avoid overparameterisation. The optimal combination of EOFs
and CCA modes are calculated by means of the goodness index (R2, Fig. 6). Notice that any set
EOFs will produce unique CCA modes for that especific set, then, once the best combination of
EOFs is determined, that set EOF is capturing the maximum variability in each field (X and Y ),
separately, for each specific CCA model (intensity and magnitude). The maximum possible number
of CCA modes, however, is determined first by the minimum number of EOFs between both fields.
Then with the goodness skill, the maximum number of CCA modes is found for the best fit to avoid
any overparameterisation in the model Ŷ = bT ·X , where the elements of b are the ordinary least-
squares regression coefficients computed with CCA, and Ŷ is the predicted value of Y . The R2
is computed using cross-validation models with 1-month window from 1961-2012 for each station
in all the models. This metric also allows identifying the best month to predict any of the MSD
features. It is worth mentioning that at the end the models would not necessarily have 17 EOF
and CCA modes. Stations having indexes with more than 30% of missing data (i.e. fails detecting
9
MSD events) are discarded. In total, 21 stations are left for the CCA models. Precipitation data are
transformed to a normal distribution using percentiles, to achieve better performance in the EOF.
The 21 stations having less than 30% of missing data are filled using the long-term means of each
index. Note that filling the data in this step is not the same procedure described above.
4 Results
4.1 General features of the MSD
Figs. 3 and 4 show a statistical summary of the timing of the MSD phases. Each phase exhibits
subtle differences throughout the stations. Those discrepancies suggest to study this phenomenon
separately per station. For example, from Fig. 3, one can see that in the southern part of Central
America (Panama, Costa Rica, Nicaragua) the MSD tends to start, develop and end earlier than in
the northern part (El Salvador, Guatemala, Honduras), in agreement with previous studies such as
Alfaro (2002) and Herrera et al. (2014). In the southern stations, typically, the start of the MSD is
detected by 20 June, the minimum is reached by 19 July and the end by 20 August, meanwhile at
the north the start is observed by 22 June, the minimum by 24 July and the end around 24 August.
Note that the difference can be up to 20 days in each phase from south to north if the stations are
analysed separately. The results in Fig. 4 also show the high temporal and spatial variability existent
throughout the region agreement with Alfaro (2002) and Amador (2008), but that feature is also
observed at local scales as noted by Ramírez (1983); Hernández (2013); Alfaro (2014), and Solano
(2015).
The algorithm captures similar MSD structure in stations located over the eastern Nicaragua
(Bluefields and Puerto Cabezas) at the Caribbean region. In addition, the stations located at the
Caribbean coast present later development of the MSD, starting in average by 29 June, the minimum
in 21 July, and the end at 14 August, therefore, being later and shorter in the Caribbean than in
the Pacific side of Central America, showing a marked difference between the eastern and western
Central America. This MSD structure over the Caribbean has been previously studied by Martin
and Schumacher (2011), which concluded that the MSD in the Caribbean is controlled by different
external forcings than the one operating in the Pacific sector.
This observed latitudinal difference in the timing in each MSD phase is examined using a t-test
(not shown). The results, nevertheless, show no statistical difference among the stations at 90% of
significance level, except for the start index in Bluefields station, indicating that in the Caribbean
slope the MSD is out-of-phase at least at the start. Thus, it can be argued that the MSD over the
Pacific occur simultaneously as found by Magaña et al. (1999).
The intensity and magnitude of the MSD also show high spatial variability (Fig. 5), with a strong
contrast between the Pacific and the Caribbean coasts. The Pacific side presents drier MSDs than in
the Caribbean basin. In the western Central America, the mean intensity is 4.06mm (σ = 1.85mm)
10
A B C
station
250
Huehuetenango
La Fragua
LaborOvalle
Asuncion Mita
200 Tegucigalpa
Puerto Cabezas
Ocotal
San Miguel
Choluteca
150 Amapala
Chinandega
Managua
Juigalpa
Bluefields
100 Rivas
Liberia
Bagaces
Santa Cruz
Nicoya
50
La Argentina
Fabio Baudrit
Juan Santa Maria
CIGEFI
0 David
Los Santos
15.3 9.9 15.3 9.9 15.3 9.9
Latitude
Figure 3. Mean values of the indexes describing the timing of the MSD phases: A) start, B) minimum, and C)
end. The x-axis has latitude coordinates being north to the left.
and mean magnitude of 6.50 mmday−1 (σ = 2.00 mmday−1), with the highest score detected in
David, Panama with an intensity of 7.70 mm (σ = 2.01 mm), and a magnitude of 10.45 mm (σ =
2.20mmday−1). On the Caribbean side, Bluefields and Puerto Cabezas show the highest records in
intensity and magnitude. The mean intensity recorded in Bluefields is 16.64mm (σ = 5.50mm), and
Puerto Cabezas 8.81mm (σ = 3.68mm), whereas the mean magnitude scored is 19.93mmday−1 (σ
= 5.39mmday−1) and 11.31mmday−1 (σ = 3.30mmday−1). As above, using t-test is determined
that the differences in the MSD intensity and magnitude indexes amongst the stations are statistically
significant at 90% of significance level.
Inter-annual variability of the MSD intensity and magnitude is examined in Fig. 5. Correlations
with Niño 3.4 index are shown in Fig. 6a,c. Notice that all the stations show a significant negative
correlation, except one (San Miguel, El Salvador), nevertheless in both cases the correlation is rela-
tively low (0.24 for intensity and 0.18 for magnitude), hence, this result can be discarded. Tables 2
and 3 show a summary of the contingency tables calculated. In both cases, the extreme values of the
11
Days of year
Huehuetenango La Fragua LaborOvalle Asuncion Mita Tegucigalpa
300
250
●
200 ●
150
100
Puerto Cabezas Ocotal San Miguel Choluteca Amapala
300
250
●
200
150
100
Chinandega Managua Juigalpa Bluefields Rivas
300
250
200
150
100
Liberia Bagaces Santa Cruz Nicoya La Argentina
300 ●
250
200
150
100
Fabio Baudrit Juan Santa Maria CIGEFI David Los Santos
300
250
200 ●
●
150 ● ●●
100 ●
A B C A B C A B C A B C A B C
MSD phase
Figure 4. Box-plot of the indexes describing the MSD timing. In the x-axis are each of the events defining the
MSD, start (A), minimum (B), and end (C). In the y-axis are the time-coordinates in Julian days. The upper and
lower limits of each box correspond to the first and third quartiles (the 25th and 75th percentiles). The upper
(lower) whisker extends from the hinge to the highest (lowest) value that is within 1.5 times inter-quartile range
of the hinge. Data beyond the end of the whiskers are outliers and plotted as dots. The median is the bold line
in each box. Note that the panels are distributed from north to south, left to right and top to bottom.
MSD intensity and magnitude present a connection with ENSO events, however, those conditions
are not observed in all the stations, despite some of them have a significant correlation with Niño
3.4 (e.g. Liberia and Juigalpa). Notice that the results for the Caribbean stations are not statistically
significant. Alfaro (2014) and Solano (2015) have found similar results for stations in Costa Rica;
and Fallas-López and Alfaro (2012a) reported the same relationship with other stations in Central
America. This association between the SST anomalies in the Nino 3.4 region and the MSD intensity
and magnitude over the west cost of Central America can be connected with low-level wind anoma-
lies over the Caribbean Sea. Previous studies have found a relationship between the fluctuation of the
CLLJ and precipitation during the summer months (Wang, 2007; Amador, 2008). Figure 6b,d show
the correlation of the intensity and magnitude of the MSD with the CLLJ index. Similarly, most of
the stations show positive significant correlations, except the stations in the Caribbean. It is known
that in the summer months (JJA) during warm (cold) ENSO episodes the CLLJ is stronger (weaker)
than normal (Amador, 2008) with a correlation of -0.53 (p-value < 0.01). Hidalgo et al. (2015) also
12
Julian days
mean
17.5 16
12
●
15.0 ● ●
● 8
● ●
●
●●●
4
12.5 ●
● ● ●
●
●● SD
10.0 ●● ●●
● 1
● ● 2
7.5 ● ● 3
● 4
5.0 ● 5
−95 −90 −85 −80 −75
17.5 mean
16
●
15.0 ● ●
● 12
● ●
●
●●● 8
12.5 ●
● ● ●
● 4
●●
10.0 ●● ●● SD
● 2
●
● 3
7.5 ●
● 4
● 5
5.0
−95 −90 −85 −80 −75
Figure 5. Mean and standard deviation for the MSD intensity (mm, top) and magnitude (mmday−1, bottom)
for the period 1961-2012. Note the color scale is not the same in both figures, since the intensity and magnitude
are not necessarily of the same order of magnitude. The dots size represents the standard deviation.
noted that during stronger (weaker) CLLJ the ITCZ is located south (north) from the normal posi-
tion, leading to a reduction (increase) in the total precipitation during boreal summer (JJA), thus,
affecting the MSD. Therefore, the bottom line of this result is that the ENSO events modulate the
intensity and magnitude of the MSD, and the evidence reveals that the observed changes in these
metrics are connected with the modulation of the CLLJ, functioning as the mechanism that transfers
El Niño signal into the precipitation field, and consequently shaping the MSD. It is worth to high-
light again that from these results, those processes explain only the inter-annual variability of the
MSD in the Pacific coast. The same estimations were done for the other variables (start, minimum,
end), however, the results do not shown significant correlations with El Niño and CLLJ index (not
shown).
13
a b
17.5 17.5
15.0 ●● ● 15.0 ● ●
● ●
● ●●
12.5 ● 12.5 ●
● ●●
●
● ●●
10.0 ● 10.0 ●●●
sig.level
● ● 99%
7.5 ● 7.5 ● 95%
90%
5.0 5.0
−95 −90 −85 −80 −75 −95 −90 −85 −80 −75 r
c d 0.8
17.5 17.5 0.4
0.0
15.0 ● ● 15.0
●
● ● −0.4
● ●
●
● ●● −0.8
12.5 ● 12.5 ●
●●
●
●●
10.0 ● ● 10.0 ● ●●
●
7.5 ● 7.5 ●
5.0 5.0
−95 −90 −85 −80 −75 −95 −90 −85 −80 −75
Figure 6. Top Pearson correlation between MSD intensity and a) CLLJ index and b) Niño 3.4 indexes. Bottom
the same metric but for the MSD magnitude with c) CLLJ and d) Niño 3.4 indexes. Niño 3.4 index is taken for
June. The CLLJ index is estimated from the daily time series, using the definition by Amador (2008) for July.
4.2 Canonical correlation analysis
Models based on CCA are tailored for each of the indexes describing the features of the MSD,
defined in Section 3. Using the SST anomalies as predictor (X field), CCA identifies SST patterns
that are related to the perturbations of the MSD features (Y field, predictant). The goodness index
(R2) is shown in Fig. 7. The intensity and magnitude show the best skill score compared to the
other indexes. These results show that the models to study the timing of the MSD phases have
a poor performance using the CCA technique. Consequently, those models are not considered in
the further analysis. Both CCA models for MSD intensity and magnitude show one of the highest
values of R2 in April, which means in operational terms, information concerning the MSD intensity
and magnitude can be retrieve up to 3 months in advance of an event. This would be valuable for
preparation and planning of the societal, economical and agricultural activities during the MSD
14
Table 2. Summary of contingency tables for each station. The tables were estimated using Niño 3.4 index and
the mean intensity per year. The categories were defined using the 33rd (below normal, BN) and 67th (above
normal, AN). Years in which the algorithm failed detecting any of the phases of the MSD, were removed.
The star (*) represents cases in which the observed condition in column 2 and 3 dominates, and has statistical
significance > 90%
Station SST AN and Intensity BN SST BN and Intensity AN r (Kendall) p-value
La Argentina * * -0.38 0.00
Fabio Baudrit -0.05 0.62
Juan Santa Maria * * -0.34 0.00
Liberia -0.21 0.05
Nicoya * -0.24 0.02
Santa Cruz * * -0.26 0.01
Bagaces * -0.30 0.01
CIGEFI * -0.30 0.02
Bluefields -0.04 0.77
Ocotal * -0.27 0.01
Chinandega * -0.33 0.00
Juigalpa -0.21 0.04
Managua * * -0.39 0.00
Puerto Cabezas -0.02 0.87
Rivas -0.13 0.26
David * * -0.32 0.00
Los Santos * -0.22 0.03
San Miguel 0.25 0.01
Asuncion Mita -0.22 0.04
Huehuetenango * -0.14 0.16
Labor Ovalle * -0.35 0.00
La Fragua * * -0.42 0.00
Amapala * -0.40 0.00
Choluteca * -0.28 0.01
Tegucigalpa * * -0.26 0.02
period. Notice that both models also shown the highest results in July, concurrently with the existence
of the MSD, and the CLLJ.
In this study, we analyse the SSTA patterns identified by the model in June, since for that month,
CCA models show a relative high R2, and there is also a 1-month leading time for forecasting the
MSD. Besides, the SST do not change significantly from June to July, when the CCA models have
the best performance, thus, the SST in July is expected to be close to the SST in June. The best com-
15
0.35
0.3
0.25
start
0.2 minimum
end
intensity
0.15 magnitude
0.1
0.05
0
-0.05
JAN FEB MAR APR MAY JUN JUL AUG
  SST  
Figure 7. Goodness index (R2) estimated as the average of the Pearson correlation between synthetic time
series generated by cross-validation models and the observed MSD features per station.
bination of EOF and CCAmodes for the MSD intensity areX = 14, Y = 4 EOFs and 2 CCAmodes,
and for the MSD magnitude X = 11, Y = 6 EOFs and 3 CCA modes, meaning that for each model
(for intensity and magnitude respectively) the best fit is achieved when using 2 and 3 modes (canon-
ical variates) respectively, capturing the maximum influence of the SST on the precipitation field,
and specifically in the modulation of the MSD. Figure 8 shows the Pearson correlation between the
predicted time series generated using cross-validation models and the observations of both intensity
and magnitude in each station. In each case, the station exhibits relative high significant correlation,
about 0.35 in average.
The X loadings (correlation between the canonical vectors of the SST and the SSTA) of the first
mode controlling the MSD intensity shows a dipolar pattern in the correlation with the SSTAs sur-
rounding Central America (Fig. 9a,b). The highest positive correlations are found over the Pacific
in the El Niño region, whereas the highest negative correlation are found over the Tropical North
Atlantic and near to the Brasil coast. The influence of this variability mode on the precipitation in
Central America has been previously studied by Enfield and Alfaro (1999); Maldonado and Alfaro
(2011) and Maldonado et al. (2013), but for the secondary peak of precipitation during August-
October (ASO), thus, this results shows that the rainfall during the MSD is governed by the same
variability mode present during the highest precipitation season in Central America. On the other
hand, for the MSD magnitude, the first mode exerts more influence of the SSTA over the El Niño
region, and the regional waters close the west Mexican coast. Notice that in both cases also high pos-
itive correlations are found near the western coast of Mexico revealing that both models are affected
by the influence of regional waters, varying with the same phase that the superficial temperatures in
the El Niño region. The Y loadings in both models are negatively correlated with most of the stations
(Fig. 9c,d). That means for a state in which the water over the Pacific is warmer (colder) than normal,
16
R2
Table 3. Same as Table 2 but for MSD magnitude.
Station SST AN and Magnitude BN SST BN and Magnitude AN r (Kendall) p-value
La Argentina * -0.38 0.00
Fabio Baudrit -0.04 0.71
Juan Santa Maria * * -0.36 0.00
Liberia * -0.17 0.10
Nicoya * -0.26 0.01
Santa Cruz * * -0.23 0.03
Bagaces * -0.23 0.06
CIGEFI * -0.23 0.07
Bluefields -0.06 0.66
Ocotal -0.06 0.55
Chinandega * * -0.32 0.00
Juigalpa -0.08 0.43
Managua * -0.25 0.01
Puerto Cabezas 0.02 0.87
Rivas -0.24 0.03
David * * -0.32 0.00
Los Santos * -0.26 0.01
San Miguel 0.19 0.06
Asuncion Mita * -0.28 0.01
Huehuetenango * * -0.21 0.01
Labor Ovalle * -0.35 0.00
La Fragua * -0.29 0.01
Amapala * -0.33 0.01
Choluteca * -0.26 0.01
Tegucigalpa -0.21 0.05
plus the SST over the Atlantic colder (warmer) than normal, the intensity and magnitude of the MSD
drier (wetter). This results also connects the MSD period with the second rainfall peak, that is, SSTs
conditions for drier MSDs, would persist leading to less rainfall during the second peak, following
the results in Maldonado et al. (2013). The temporal scores (canonical variates) of this mode in both
models show that this mode has mainly inter-annual variations (Fig. 9e,f).
TheX loadings of the second mode are shown in Fig. 10a and b. The MSD intensity is dominated
by negative correlations with SST anomalies in the Tropical North Atlantic, the Caribbean and Gulf
of Mexico and the eastern Pacific, being more important than the first region. The X loadings for the
MSD magnitude on the other hand, reveal a tripolar configuration in the North Pacific. This tripolar
setting is located in the region where the PDO develops. The PDO has shown more influence in the
17
Intensity
17.5
15.0 ●
●
12.5
10.0 ● ● sig.level
● 90%
7.5 95%
99%
5.0 r
Magnitude
0.6
17.5
0.4
15.0 0.2
●
0.0
12.5
●
10.0
7.5
5.0
−95 −90 −85 −80 −75
Figure 8. Pearson correlation of the CCA models with the time series of intensity (top) and magnitude (bottom)
of MSD.
southern USA and northern Mexico than in Central America (Muñoz et al., 2010; Maldonado et al.,
2015), and that could be the reason that its influence is observed in the second mode. The Y loadings
in both cases (Fig. 10c,d) show a non-uniform distribution of the correlation. For the MSD intensity
(Fig. 9c), the stations are positively correlated with this mode, meanwhile, the MSD magnitude (Fig.
10d) shows a clear division between northern and southern Central America, being the former region
(Guatemala, El Salvador, Honduras and west Nicaragua) negatively correlated and the latter region
(Costa Rica, Panama and east Nicaragua) positively correlated. The temporal scores for the intensity
model exhibit mainly inter-annual variability with a negative trend after 1990 (Fig. 9e), while the
temporal scores of the magnitude (Fig. 10f) present a decadal variation, that again possibly relates
to the influence of the PDO.
The MSD magnitude is the only feature with a third mode, the X loadings (Fig. 11a) reveal that
this mode is dominated by regional SSTs in the Pacific and Tropical North Atlantic. The north and
south Pacific show a dipole being positive and negative correlations respectively. The Y loadings
18
111W 150E 180 150W 120W 90W 60W 30W 0E 111W 150E 180W 150W 120W 90W 60W 30W 0E
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
a) b)
95W 90W 85W 80W 75W 95W 90W 85W 80W 75W
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
c) d)
e) f)
Figure 9. CCA mode 1 for both MSD intensity and magnitude. In the left column are the X (upper) and
Y(middle) loadings, and the time scores (bottom) for the intensity. In the right column the same but for the
magnitude.
(Fig. 11b) show a north/south division being positively correlated to the north and negatively to the
south, while the temporal scores (Fig. 11c) reveal an interannual variation of this mode.
5 Discussion and conclusions
The MSD is characterised using daily time series of rainfall for 25 stations located mainly in the
Pacific coast of Central America during the 1961 – 2012 period. An algorithm to detect the MSD
phases is developed in order to compare the timing of the MSD start, date of minimum, and end
in each station. This algorithm clearly captures the development of the MSD over the Pacific, but
also a similar MSD-like structure in two stations located in the Caribbean coast. The results confirm
19
−20 0 20 40 60
6 8 10 12 14 16 18
−20 0 20 40 60
6 8 10 12 14 16 18
111W 150E 180W 150W 120W 90W 60W 30W 0E 111W 150E 180W 150W 120W 90W 60W 30W 0E
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
a) b)
95W 90W 85W 80W 75W 95W 90W 85W 80W 75W
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
c) d)
e) f)
Figure 10. Same as Fig. 9 but for Mode 2.
previous observations showing the high spatial and temporal variability of the MSD, thus, suggesting
the need of a more detailed analysis. We , thus, analyse the features of the MSD individually per
station, using a combination of empirical analysis (contingency tables) and a more sophisticated
statistical method as the canonical correlation analysis to diagnose the variability of the MSD in
terms of changes in the SSTA of the surrounding waters.
The MSD intensity and magnitude are correlated with the Niño 3.4 and the CLLJ index in sta-
tions located over the western side of Central America. These results reveal a significant negative
correlation in almost all the stations with ENSO phases, i.e. warm (cold) anomalies in the Niño 3.4
region, corresponding to an enhanced (decreased) CLLJ intensity, are associated with drier (wetter)
MSD. The results of the correlation with the CLLJ show also a positive association of such process,
indicating that with a stronger (weaker) jet the MSD intensifies (decreases). This indicates that the
20
−20 0 20 40 60
6 8 10 12 14 16 18
−20 0 20 40 60
6 8 10 12 14 16 18
111W 150E 180W 150W 120W 90W 60W 30W 0E
−1.0 −0.5 0.0 0.5 1.0
(a)
95W 90W 85W 80W 75W
−1.0 −0.5 0.0 0.5 1.0
(b)
(c)
Figure 11. CCA mode 3 for MSD magnitude. X (a) and Y(b) loadings, and (c) the time scores.
21
6 8 10 12 14 16 18 −20 0 20 40 60
CLLJ is involved in the mechanism connecting the SST anomalies in the Pacific with precipitation
fluctuations, however, other processes such the migration of the ITCZ, the North American Monsoon
and perturbations in the SLP fields over the Amazon could also be engaged with anomalous MSD
(Hidalgo et al., 2015). The two stations placed in the Caribbean side do not reveal results statisti-
cally significant, hence, suggesting that other elements are involved explaining the annual rainfall
variability during their MSD period.
The outputs of CCA show that the MSD intensity is mainly modulated by a dipolar configuration
in the SST anomalies formed between the Pacific and the Tropical North Atlantic and Caribbean sea,
meanwhile, the first mode controlling the MSD magnitude shows more influence from El Niño and
the regional waters close to the western coast of Mexico and Central America. This variability mode
in both models has an inter-annual scale, with negative effects in precipitation, i.e. when this mode
is positive for both intensity and magnitude, a reduction in rainfall is observed in almost all the sta-
tions. Similar SST dipole has been reported to influence the anomalies in rainfall during the months
of the highest precipitation events (August-November), and it has been associated with intensifica-
tion/reduction of the trades, leading to a decrease/increase of precipitation in Central America during
the second rainfall period, however, these results show this mode is present prior to the quarter ASO
and is also modulating the precipitation during boreal summer.
The second mode in both models presents more complex structures, and it is difficult to distinguish
a general pattern affecting the intensity as well as magnitude of the MSD. For the intensity model,
the SST show a negative correlation pattern over the equatorial Pacific and North Atlantic waters,
nevertheless, the latter region dominates; the time series for this mode show a negative trend after
1990. Meanwhile, for the magnitude, a tripolar configuration over the vicinity of the PDO develop-
ment region is found. Also negative correlations are found over the Gulf of Mexico. The influence of
the PDO should be taken with caution and needs more analysis since the PDO has shown more influ-
ence in the southern USA and northern Mexico than in Central America. The time series exhibited
inter-decadal variability. As previously mentioned, the correlations between both variables (intensity
and magnitude) and this mode, however, do not depict a clear pattern in the stations; for the MSD
intensity the correlations reveal a positive association with the SSTA, meanwhile, for the MSD mag-
nitude shows a division between the north and south Central America, being the north (Guatemala,
Honduras, El Salvador and west Nicaragua) affected negatively and the south and Caribbean (Costa
Rica, Panama and east Nicaragua) positively. It is clear that the second mode is not controlling the
precipitation in both models in the same way . The North Atlantic waters become more important
for the MSD intensity, hence, could be associated with the same controlling mechanism present dur-
ing the first precipitation maximum (Alfaro, 2007; Fallas-López and Alfaro, 2012b). While for the
MSD magnitude the tripolar configuration might suggest the influence of the PDO, also noted in the
contrasting effects of this mode between north and south Central America.
22
The MSD magnitude is the only variable with higher CCA modes. The third mode reveals the
influence of regional waters close to the Central America Pacific and a dipole formed between the
north (positive) and south (negative) Pacific, affecting with opposite sign the eastern (negative) and
western (positive) part of Central America. This mode presents an inter-annual variability scale. The
CCA allows identifying patterns of the SST affecting variables describing theMSD such as the inten-
sity and magnitude, however, it shows a poor performance related to the temporal variables. Another
benefit achieved using CCA was the identification of particular months suitable for prediction of the
intensity and magnitude of the MSD, being capable of forecasting up to 3 months in advance, which
is a reasonable time in terms of practical matters related to prevention and planing for the season.
Finally, it is worth pointing out that this analysis also provide a systematic method to study the MSD
features, that can be used for statistical forecasts of such phenomenon.
Acknowledgements. This research was carried out within the CNDS research school, supported by the Swedish
International Development Cooperation Agency (Sida) through their contract with the International Science
Programme (ISP) at Uppsala University (contract number: 54100006). The authors would like to thank the
Centre for Natural Disaster Science (CNDS) in Uppsala University, the National Centers for Environmental Pre-
diction (NCEP), and the National Center for Atmospheric Research (NCAR) for the reanalysis data. We would
also like to acknowledge support via project VI-805-A9-532, CIGEFI-UCR, provided via a SIDA-CSUCA
agreement.
23
References
Alfaro, E.: Some Characteristics of the Annual Precipitation Cycle in Central America and their Relationships
with its Surrounding Tropical Oceans, Tópicos Meteorológicos y Oceanográficos, 9, 1–13, http://imn.ac.cr/
publicaciones/revista/2002/Diciembre/1-Alfaro-Dic02.pdf, 2002.
Alfaro, E.: Uso del análisis de correlación canónica para la predicción de la precipitación pluvial en Cen-
troamérica, Ingeniería y Competitividad, 9, 33–48, http://bibliotecadigital.univalle.edu.co/xmlui/handle/
10893/1622, 2007.
Alfaro, E.: Caracterización del “veranillo” en dos cuencas de la vertiente del Pacífico de Costa Rica,
América Central (Characterization of the Mid Summer Drought in two Pacific slope river basins
of Costa Rica, Central America). Spanish., International Journal of Tropical Biology, 62, 1–15,
https://www.academia.edu/9493294/Caracterizaci%C3%B3n_del_veranillo_en_dos_cuencas_de_la_
vertiente_del_Pac%C3%ADfico_de_Costa_Rica_Am%C3%A9rica_Central_Characterization_of_the_
Mid_Summer_Drought_in_two_Pacific_slope_river_basins_of_Costa_Rica_Central_America_._Spanish,
2014.
Alfaro, E. and Soley, J.: Descripción de dos métodos de rellenado de datos ausentes en series de tiempo me-
teorológicas, Revista de Matemáticas: Teoría y Aplicaciones., 16, 59–74, http://revistas.ucr.ac.cr/index.php/
matematica/article/view/1419, 2009.
Alfaro, E., Soley, J., and Enfield, D.: Uso de una Tabla de Contingencia para Aplicaciones Climáticas (Use of a
Contingency Table for Climatic Applications), p. 51, ESPOL and FUNDESPOL, Guayaquil, Ecuador, 2003.
Amador, J. A.: A Climatic Feature of the Tropical Americas: The Trade Wind Easterly Jet, Top. Meteor.
Oceanogr., 5, 91–102, http://imn.ac.cr/publicaciones/revista/1998/diciembre/1-Amador,Jorge-dic98.pdf.pdf,
1998.
Amador, J. A.: The Intra-Americas Sea Low-level Jet Overview and Future Research, Annals of the New York
Academy of Sciences, 1146, 153–188, doi:10.1196/annals.1446.012, http://onlinelibrary.wiley.com/doi/10.
1196/annals.1446.012/abstract, 2008.
Amador, J. A., Alfaro, E. J., Lizano, O. G., and Magaña, V. O.: Atmospheric forcing of the eastern tropical
Pacific: A review, Progress In Oceanography, 69, 101–142, doi:10.1016/j.pocean.2006.03.007, http://www.
sciencedirect.com/science/article/pii/S0079661106000292, 2006.
Ashby, S. A., Taylor, M. A., and Chen, A. A.: Statistical models for predicting rainfall in the Caribbean,
Theor. Appl. Climatol., 82, 65–80, doi:10.1007/s00704-004-0118-8, http://link.springer.com/article/10.
1007/s00704-004-0118-8, 2005.
Chen, A. A. and Taylor, M. A.: Investigating the link between early season Caribbean rainfall and the El Niño+
1 year, Int. J. Climatol., 22, 87–106, doi:10.1002/joc.711, 2002.
Chinchilla-Ramírez, G.: Resumen Meteorológico Julio 2014. Boletín Meteorológico Mensual, http://www.imn.
ac.cr/boletin_meteo/historial/2014/BMET072014.pdf, 2014.
Enfield, D. B. and Alfaro, E. J.: The Dependence of Caribbean Rainfall on the Interaction of
the Tropical Atlantic and Pacific Oceans, Journal of Climate, 12, 2093–2103, doi:10.1175/1520-
0442(1999)012<2093:TDOCRO>2.0.CO;2, http://journals.ametsoc.org/doi/abs/10.1175/1520-0442(1999)
012%3C2093%3ATDOCRO%3E2.0.CO%3B2, 1999.
24
Fallas-López, B. and Alfaro, E. J.: Uso de herramientas estadísticas para la predicción estacional del campo de
precipitación en América Central como apoyo a los Foros Climáticos Regionales. 2: Análisis de Correlación
Canónica., Revista de Climatología, 12, http://webs.ono.com/reclim8/reclim12g.pdf, 2012a.
Fallas-López, B. and Alfaro, E. J.: Uso de herramientas estadísticas para la predicción estacional del campo de
precipitación en América Central como apoyo a los Foros Climáticos Regionales. 1: Análisis de tablas de
contingencia., Revista de Climatología, 2012, 12: 61-79, http://webs.ono.com/reclim7/reclim12e.pdf, 2012b.
Gershunov, A. and Barnett, T.: ENSO influence on intraseasonal extreme rainfall and temperature frequencies in
the contiguous United States: Observations and model results, J. Climate, 11, 1575–1586, doi:10.1175/1520-
0442(1998)011<1575:EIOIER>2.0.CO;2, 1998.
Gershunov, A. and Cayan, D. R.: Heavy Daily Precipitation Frequency over the Contiguous United
States: Sources of Climatic Variability and Seasonal Predictability, Journal of Climate, 16, 2752–
2765, doi:10.1175/1520-0442(2003)016<2752:HDPFOT>2.0.CO;2, http://journals.ametsoc.org/doi/abs/10.
1175/1520-0442%282003%29016%3C2752%3AHDPFOT%3E2.0.CO%3B2, 2003.
Hernández, K.: Estudio de la evaporación en Costa Rica y su aplicación para determinar el inicio y la conclusión
de la época seca y lluviosa., Tesis de Grado, Licenciatura, Escuela de Física, Universidad de Costa Rica, San
José, Costa Rica, 2013.
Herrera, E., Magaña, V., and Caetano, E.: Air–sea interactions and dynamical processes associated with the
midsummer drought, Int. J. Climatol., pp. n/a–n/a, doi:10.1002/joc.4077, http://onlinelibrary.wiley.com/doi/
10.1002/joc.4077/abstract, 2014.
Hidalgo, H. G., Durán-Quesada, A.M., Amador, J. A., and Alfaro, E. J.: The Caribbean Low-Level Jet, the Inter-
Tropical Convergence Zone and Precipitation Patterns in the Intra-Americas Sea: A Proposed Dynamical
Mechanism, Geografiska Annaler: Series A, Physical Geography, 97, 41–59, doi:10.1111/geoa.12085, http:
//onlinelibrary.wiley.com/doi/10.1111/geoa.12085/abstract, 2015.
Jury, M. R. and Malmgren, B. A.: Joint modes of climate variability across the inter-Americas, Int. J. Climatol.,
32, 1033–1046, doi:10.1002/joc.2324, http://onlinelibrary.wiley.com/doi/10.1002/joc.2324/abstract, 2012.
Kalnay, E., Kanamitsu, M., Kistler, R., Collins, W., Deaven, D., Gandin, L., Iredell, M., Saha, S., White, G.,
Woollen, J., Zhu, Y., Leetmaa, A., Reynolds, R., Chelliah, M., Ebisuzaki, W., Higgins, W., Janowiak, J.,
Mo, K. C., Ropelewski, C., Wang, J., Jenne, R., and Joseph, D.: The NCEP/NCAR 40-Year Reanalysis
Project, Bull. Amer. Meteor. Soc., 77, 437–471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2,
http://dx.doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2, 1996.
Karnauskas, K. B., Seager, R., Giannini, A., and Busalacchi, A. J.: A simple mechanism for
the climatological midsummer drought along the Pacific coast of Central America, Atmós-
fera, 26, 261–281, doi:10.1016/S0187-6236(13)71075-0, http://www.sciencedirect.com/science/article/pii/
S0187623613710750, 2013.
Kistler, R., Collins, W., Saha, S., White, G., Woollen, J., Kalnay, E., Chelliah, M., Ebisuzaki, W., Kana-
mitsu, M., Kousky, V., van den Dool, H., Jenne, R., and Fiorino, M.: The NCEP–NCAR 50–Year
Reanalysis: Monthly Means CD–ROM and Documentation, Bull. Amer. Meteor. Soc., 82, 247–267,
doi:10.1175/1520-0477(2001)082<0247:TNNYRM>2.3.CO;2, http://dx.doi.org/10.1175/1520-0477(2001)
082<0247:TNNYRM>2.3.CO;2, 2001.
25
Magaña, V., Amador, J. A., and Medina, S.: The Midsummer Drought over Mexico and Central Amer-
ica, Journal of Climate, 12, 1577–1588, doi:10.1175/1520-0442(1999)012<1577:TMDOMA>2.0.CO;2,
http://journals.ametsoc.org/doi/abs/10.1175/1520-0442%281999%29012%3C1577%3ATMDOMA%3E2.
0.CO%3B2, 1999.
Maldonado, T. and Alfaro, E.: Predicción estacional para ASO de eventos extremos y días con precipitación
sobre las vertientes Pacífico y Caribe de América Central, utilizando análisis de correlación canónica, In-
terSedes, 12, 78–108, http://www.intersedes.ucr.ac.cr/ojs/index.php/intersedes/article/view/301, 2011.
Maldonado, T., Alfaro, E., Fallas-López, B., and Alvarado, L.: Seasonal prediction of extreme precipitation
events and frequency of rainy days over Costa Rica, Central America, using Canonical Correlation Analysis,
Adv. Geosci., 33, 41–52, doi:10.5194/adgeo-33-41-2013, http://www.adv-geosci.net/33/41/2013/, 2013.
Maldonado, T., Rutgersson, A., Amador, J., Alfaro, E., and Claremar, B.: Variability of the Caribbean low-level
jet during boreal winter: large-scale forcings, 2015.
Martin, E. R. and Schumacher, C.: The Caribbean Low-Level Jet and Its Relationship with Precipitation in IPCC
AR4 Models, Journal of Climate, 24, 5935–5950, doi:10.1175/JCLI-D-11-00134.1, http://journals.ametsoc.
org/doi/abs/10.1175/JCLI-D-11-00134.1, 2011.
Muñoz, E., Busalacchi, A. J., Nigam, S., and Ruiz-Barradas, A.: Winter and Summer Structure of the Caribbean
Low-Level Jet, Journal of Climate, 21, 1260–1276, doi:10.1175/2007JCLI1855.1, http://journals.ametsoc.
org/doi/abs/10.1175/2007JCLI1855.1, 2008.
Muñoz, E., Wang, C., and Enfield, D.: The Intra-Americas Sea springtime surface temperature anomaly dipole
as fingerprint of remote influence, J. Climate, 23, 43–56, doi:http://dx.doi.org/10.1175/2009JCLI3006.1,
2010.
Ramírez, P.: Estudio Meteorológico de los Veranillos en Costa Rica., Nota de investigación. 5, Instituto Meteo-
rológico Nacional, Ministerio de Agricultura y Ganadería., San José, Costa Rica, 1983.
Smith, T., Reynolds, R., Peterson, T. C., and Lawrimore, J.: Improvements to NOAA’s Historical Merged
Land–Ocean Surface Temperature Analysis (1880–2006), J. Climate, 21, 2283 – 2296, 2007.
Solano, E.: Análisis del comportamiento de los períodos caniculares en Costa Rica en algunas cuencas del
Pacífico Norte y del Valle Central entre los años 1981 y 2010., Tesis de Grado, Licenciatura, Escuela de
Física, Universidad de Costa Rica, San José, Costa Rica, 2015.
Spence, J. M., Taylor, M. A., and Chen, A. A.: The effect of concurrent sea-surface temperature anomalies in the
tropical Pacific and Atlantic on Caribbean rainfall, Int. J. Climatol., 24, 1531–1541, doi:10.1002/joc.1068,
http://onlinelibrary.wiley.com/doi/10.1002/joc.1068/abstract, 2004.
Taylor, M. A. and Alfaro, E. J.: Climate of Central America and the Caribbean, in: Encyclopedia of World
Climatology, edited by Oliver, J. E., pp. 183–186, Springer, Netherlands, 2005.
Taylor, M. A., Enfield, D. B., and Chen, A. A.: Influence of the tropical Atlantic versus the trop-
ical Pacific on Caribbean rainfall, Journal of Geophysical Research: Oceans, 107, 10–1–10–14,
doi:10.1029/2001JC001097, http://onlinelibrary.wiley.com/doi/10.1029/2001JC001097/abstract, 2002.
Trenberth, K. E.: The Definition of El Niño, Bulletin of the American Meteorological Society, 78, 2771–
2777, doi:10.1175/1520-0477(1997)078<2771:TDOENO>2.0.CO;2, http://journals.ametsoc.org/doi/abs/10.
1175/1520-0477(1997)078%3C2771%3ATDOENO%3E2.0.CO%3B2, 1997.
26
Vera, C., Higgins, W., Amador, J., Ambrizzi, T., Garreaud, R., Gochis, D., Gutzler, D., Lettenmaier, D.,
Marengo, J., and Mechoso, C. R.: Toward a unified view of the American monsoon systems, Journal of
Climate, 19, 4977–5000, http://journals.ametsoc.org/doi/pdf/10.1175/JCLI3896.1, 2006.
Wang, C.: Variability of the Caribbean Low-Level Jet and its relations to climate, Clim Dyn, 29, 411–422,
doi:10.1007/s00382-007-0243-z, http://link.springer.com/article/10.1007/s00382-007-0243-z, 2007.
Wang, C. and Enfield, D. B.: The Tropical Western Hemisphere Warm Pool, Geophysical Research Letters,
28, 1635–1638, doi:10.1029/2000GL011763, http://onlinelibrary.wiley.com/doi/10.1029/2000GL011763/
abstract, 2001.
Wang, C. and Enfield, D. B.: A Further Study of the Tropical Western Hemisphere Warm Pool, J. Cli-
mate, 16, 1476–1493, doi:10.1175/1520-0442(2003)016<1476:AFSOTT>2.0.CO;2, http://journals.ametsoc.
org/doi/abs/10.1175/1520-0442%282003%29016%3C1476%3AAFSOTT%3E2.0.CO%3B2, 2003.
Whyte, F. S., Taylor, M. A., Stephenson, T. S., and Campbell, J. D.: Features of the Caribbean low level jet,
International Journal of Climatology, 28, 119–128, doi:10.1002/joc.1510, http://onlinelibrary.wiley.com/doi/
10.1002/joc.1510/abstract, 2008.
Wilks, D. S.: Statistical Methods in the Atmospheric Sciences, Volume 100, Third Edition, Academic Press,
Amsterdam ; Boston, 3 edition edn., 2011.
Xue, Y., Smith, T., and Reynolds, R.: Interdecadal changes of 30-yr SST normals during 1871-2000, JOURNAL
OF CLIMATE, 16, 1601–1612, doi:10.1175/1520-0442-16.10.1601, 2003.
27