Progress in Oceanography 134 (2015) 244–255
Contents lists available at ScienceDirect
Progress in Oceanography
journal homepage: www.elsevier.com/locate/pocean
Aeolian particles in marine cores as a tool for quantitative
high-resolution reconstruction of upwelling favorable winds
along coastal Atacama Desert, Northern Chile
Valentina Flores-Aqueveque a,b,⇑, Stéphane Alfaro c, Gabriel Vargas a, José A. Rutllant b,d,
Sandrine Caquineau e
a
Departamento de Geología, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile, Plaza Ercilla 803, Santiago, Chile
Departamento de Geofísica, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile, Blanco Encalada 2002, Santiago, Chile
c
LISA, UMR CNRS/INSU 7583, Universités de Paris-Est Créteil et Paris-Diderot, 61 avenue du General de Gaulle, Créteil, France
d
Centre for Advanced Studies in Arid Zones, Colina El Pino s/n, La Serena, Chile
e
IRD-Sorbonne Universités (UPMC, Univ. Paris 06) – CNRS-MNHN, LOCEAN Laboratory, IRD France-Nord, 32, avenue Henri Varagnat, F-93143 Bondy, France
b
a r t i c l e
i n f o
Article history:
Received 30 September 2014
Received in revised form 27 January 2015
Accepted 17 February 2015
Available online 23 February 2015
a b s t r a c t
Upwelling areas play a major role in ocean biogeochemical cycles and ultimately in global climate,
especially in higly productive regions as the South Eastern Pacific. This work is based on the analysis of
the aeolian lithic particles accumulated in laminated sediments off Mejillones (23°S) in the eastern
boundary Humboldt Current System. It proposes a high-resolution quantitative reconstruction of the
upwelling-favorable southerly wind strength in the past 250 years, comparing its variability with
changes in organic carbon export/preserved changes to the sea bottom. The increase of the intensity and
variability in fluxes of particles larger than 35 lm and 100 lm since the second half of the 19th century
and during the 20th century confirms a general strengthening of southerly winds in the region. Spectral
analysis on the complete time-series of yearly depositional fluxes indicates that sedimentary variability
can be explained by a combination of interannual (ENSO) to decadal (PDO) oscillations similar to the ones
yielded by the analysis of the Interdecadal Pacific Oscillation index. However, when applied separately to
the lithic fluxes of the first and last centuries of the time-series, the method shows that relative to the one of
the interannual mode of variability, the influence of the decadal mode has increased in the recent period.
Based on the presence/absence of particles with sizes larger than 35/100 lm, each year of the time series
is classified as a ‘Low wind’ (<6 m/s), ‘Intermediate wind’ (6–8 m/s), or ‘Strong wind’ (10 to >12 m/s) year.
From the AD 1754–1820 period to the AD 1878–1998 one, the proportion of Low and Intermediate wind
years decreased from 12% and 74% to 3% and 68%, respectively, whereas the proportion of strong wind years
increased from 14% to 29%. For these periods the mean organic carbon also increased 22%, stating the
strong relation between export/preservation productivity rate and southerly wind intensity.
In the recent period (from AD 1950 on) for which the Oceanic Niño Index is available, the strong wind
years (AD 1982, 1983, 1994, and 1997) correspond to large values of this index, suggesting that constructive interferences that result from the interplay between interannual and decadal oscillations modes might
explain in part the reinforcement of the winds along the North Chilean coast.
Ó 2015 Elsevier Ltd. All rights reserved.
Introduction
Paleoclimate reconstructions are one of the most important
tools for understanding the Earth climate system. They provide
essential information about the behavior of certain climatic patterns, including its past changes and its natural variability.
⇑ Corresponding author at: Departamento de Geología, Facultad de Ciencias
Físicas y Matemáticas, Universidad de Chile, Plaza Ercilla 803, Santiago, Chile.
E-mail address: v.fl[email protected] (V. Flores-Aqueveque).
http://dx.doi.org/10.1016/j.pocean.2015.02.003
0079-6611/Ó 2015 Elsevier Ltd. All rights reserved.
Moreover, they offer the possibility of testing and calibrating
numerical models to improve simulations of realistic climate
change, especially if the climate forcing can be appropriately specified and the response is sufficiently well constrained (Jansen et al.,
2007). In theory, these modeling studies allow establishing the causal link between the forcings and the climate response at different
spatial and temporal scales but, for this, high-resolution records
documenting the climate variability are required for this. In the current context of increasing anthropogenic forcing, the recent period
has drawn much attention and a growing number of hemispheric
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
and regional short-term (annual to decadal) multi-proxy reconstructions for the past 500–1000 years are becoming available,
especially for Europe and North America (e.g., Mann et al., 1999;
Luterbacher et al., 2004; Moberg et al., 2005). In comparison, studies focused on the Southern Hemisphere, including South America,
are less numerous and a lack of data has long been evident for this
area (Villalba et al., 2009).
In an attempt to fill in this gap, paleoclimate proxies such as
tree-rings, corals, ice cores, glaciers and boreholes, are becoming
more widely used in this region (e.g., Boninsegna et al., 2009;
Espizua and Pitte, 2009; Le Quesne et al., 2009; Masiokas et al.,
2009; Vimeux et al., 2009). Laminated lacustrine (e.g., Rodbell
et al., 1999; Jenny et al., 2002) and marine sediments (e.g.,
Vargas et al., 2007; Gutiérrez et al., 2011) also provide a continuous
climate record spanning time scales ranging from annual to centennial. However, because their formation and preservation
require very special environmental conditions their occurrence is
not common and they usually cover only relatively short periods
of time.
So far, most of the sediment records have been used to analyze
changes in precipitation and atmospheric temperature, but they
can also provide insights into river flow fluctuations, changes in
oceanic circulation, variations in primary productivity, changes in
sea surface temperature, and variability of paleowind circulation.
Regarding the latter point studies are relatively scarce, and the
few existing works only pose the wind reconstruction from a
qualitative perspective (e.g., Stuut et al., 2002; Chauhan et al.,
2010; Dietrich and Seelos, 2010). In addition, it is important to
improve our knowledge on how efficiently the fluxes of exported/preserved phytoplankton-derived organic carbon from the surface ocean toward the sea bottom (e.g., Grimm et al., 1997),
responds to ocean–climate changes, and ultimately contribute to
biogeochemical cycles and especially to the oceanic pump of carbon dioxide in upwelling areas (Keeling and Shertz, 1992; Berger
et al., 1989).
In the Mejillones Bay (23°S) located in the South Eastern
Pacific on the coast of the Atacama Desert (23°S, Fig. 1), the geological, oceanographic and meteorological conditions are particularly
favorable for the accumulation and preservation of marine
laminated sediments (Ortlieb et al., 2000; Valdés et al., 2004;
Vargas et al., 2004). The characteristics of these sediments are well
adapted to the development of high-resolution (interannual to
centennial) paleoclimate studies aiming to reconstruct the past
atmospheric and oceanographic conditions (Vargas et al., 2004,
2007; Caniupán et al., 2009; Díaz-Ochoa et al., 2011). Recent studies have demonstrated that the lithic particles found in these marine sediments have an aeolian origin, and that their variability can
be related to modifications of the intensity of the strong southerly
winds predominant in the area and produced by the combined
effects of the Southeast Pacific Subtropical Anticyclone (SEPSA)
and of local to regional atmospheric features (Vargas et al., 2004,
2007; Flores-Aqueveque et al., 2014a,b).
Therefore, the aim of this work is to take advantage of this correlation between the deposition flux of lithics and wind strength
for achieving a quantitative high-resolution reconstruction of the
southerly wind intensity over the last ca. 250 years, and to compare with exported/preserved organic carbon fluxes toward the
sea bottom. Basically, the method will consist in inverting the relationship between the current wind variability and the sedimentological characteristics (size distribution and abundance) of the
aeolian mineral particles present in the laminated sediment cores
extracted from the Mejillones Bay established previously (Alfaro
et al., 2011; Flores-Aqueveque et al., 2012).
The first section of this paper summarizes the local context of the
study area and details the materials and method used. In particular,
the principles of the model used to link the characteristics of the
245
mineral content of the sediment cores with the intensity of the wind
are reviewed. The second section presents the results of the lithic
content time series as well as the wind intensity reconstruction generated from them. This variability is analyzed and discussed. Finally,
the last section interprets the significance of the wind intensity time
series considering its paleoclimatic implications.
Local context and methods
Climatological and oceanographic setting
In the Southern Hemisphere the subsiding branch of the Hadley
cell determines the presence of a quasi-permanent zone of high
pressures near 30°S, whose development over the Pacific Ocean
is known as the Southeast Pacific Subtropical Anticyclone
(SEPSA). The climatic influence of the SEPSA over South America
is modulated by the El Niño Southern Oscillation (ENSO; Cane,
1998) and the Pacific Decadal Oscillation (PDO; Mantua et al.,
1997; Zhang et al., 1997) which act at different time scales: interannual in the first case and decadal in the second one.
The ENSO cycle corresponds to a coupled ocean–atmosphere
phenomenon characterized by irregular (periodicity from 2 to
7 years) fluctuations between warm (El Niño) and cold (La Niña)
phases (Cane, 1998, 2005). ENSO has a strong direct impact over
coastal Ecuador and Northern Perú in austral summer, where it
produces episodic rainstorms that locally interrupt the extreme
aridity of the area (Vargas et al., 2000), and an indirect effect,
through atmospheric teleconnections, over much of subtropical
South America including higher-latitudes, in austral winter–spring.
The precipitation and temperature anomalies associated with
ENSO are considered as major expressions of the interannual variability in South America (Garreaud et al., 2009). Despite its importance and the multiple studies dedicated to ENSO, the
phenomenon is still not well understood. The variations observed
in the amplitude of the ENSO-related anomalies during the last
century (e.g., Elliott and Angell, 1988; Aceituno and Montecinos,
1993) increase the uncertainty attached to the extrapolation into
the past of the present-day relationship between ENSO and interannual climate variability.
The PDO is a long-lived pattern of Pacific climate variability
(Mantua et al., 1997) described as El Niño-like mainly because its
warm (cold) phases are very similar to those of El Niño (La Niña)
events (Garreaud and Battisti, 1999). Furthermore, though of
smaller amplitude, the PDO-related anomalies of precipitation
and temperature over South America are characterized by a spatial
distribution similar to the one of the ENSO-related anomalies
(Garreaud et al., 2009). However, as already mentioned above,
the two phenomena have very different temporal behaviors: the
PDO events persist over longer periods (20–30 years) than the
ENSO ones (about 18–24 months) (Rasmusson and Carpenter,
1983; Allan, 2000; Horii and Hanawa, 2004). Both phenomena
modulate the SEPSA and, consequently, influence climate over a
large part of South America.
In addition to the SEPSA, the climate of the study area is also
highly influenced by smaller scale (regional and local) effects such
as complex land – ocean–atmosphere interactions. These factors
induce extremely arid conditions fostered by the presence of the
coastal and Andean Ranges (Rutllant et al., 2003), the cold oceanic
waters of the Humboldt Current System (Strub et al., 1998), and
the southerly wind-driven coastal upwelling events occurring
throughout the year (Rutllant et al., 1998; Strub et al., 1998).
During the austral spring/summer season, the sea–land thermal
gradient intensifies, particularly during sunny afternoons, and the
speed of the dominant S–SW winds along the coast strengthens
(Rutllant et al., 2003). This favors in turn the occurrence of (1)
strong upwelling events in Punta Angamos (Fig. 1), and (2) wind
246
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
Fig. 1. Left panel: Average (2000–2004, QSCAT) surface vector wind over the eastern South Pacific in austral summer (H = South-eastern Pacific Subtropical Anticyclone). The
location of the study area in indicated by a red square. Right panel: Northern side of Mejillones peninsula and marine core F981A sampling site (modified from FloresAqueveque et al., 2014b).
erosion events at the surface of the hyperarid pampa Mejillones.
During these erosion events lithic particles are transported toward,
and eventually fall into, the Mejillones Bay, a shallow marine basin
located north of the Mejillones Peninsula (23°S).
Due to the protection offered by the near shore orography in the
peninsula, the Mejillones Bay remains almost unaffected by the
direct S–SW winds (upwelling shadow) for a significant part of
the day and by the coastal currents of the upwelling area (Marín
et al., 2001, 2003). This, in combination with a northerly sea–
breeze from mid morning to early afternoon results in a cyclonic
water circulation around the bay which, combined with the anoxic
conditions prevailing within it, favors the preservation of laminations in the sedimentary record (Ortlieb et al., 2000; Valdés et al.,
2004; Vargas et al., 2004). At water depths greater than 50–75 m
below sea level (b.s.l.), the middle of the basin (Fig. 1) constitutes
a depocenter for hemipelagic sedimentation of biogenic rests,
organic matter and lithic particles (Valdés et al., 2004; Vargas
et al., 2004). The sedimentary sequence is characterized by millimetric to centimetric layers composed of diatom remains (mainly
Chaetoceros sp.) related to the upwelling events, organic matter and
other biogenic rests (50–90%), together with lithic particles (5–10%
on average) of aeolian origin (Vargas et al., 2004, 2007;
Flores-Aqueveque et al., 2014b). The variability in the distribution
of these components reflects the changes in the magnitude of the
upwelling events, themselves forced by variations of the intensity
of the S–SW winds (Vargas et al., 2004, 2007).
Materials and methods
Wind erosion and transport of lithic material to the bay
The observation of lithic material of aeolian origin inside the
laminated marine sediments of Mejillones Bay (Flores-Aqueveque
et al., 2014b) and the idea that it would be possible to invert their
characteristics (quantities and size-distribution) for reconstructing
the past variability of the wind in the area (Vargas et al., 2004,
2007) boosted the interest for this region. For establishing the
quantitative link between the strength of the wind and the characteristics of the deposition flux in the bay, two field campaigns were
designed and carried out in 2006 (EOLOS 2006) and 2008 (EOLOS
2008). The first one documented the response of the surface of
Pampa Mejillones to the stress exerted on it by the prevailing
southerly winds. In particular, the results of this campaign were
used to propose a quantitative model of the sand movements at
the surface of the pampa (Flores-Aqueveque et al., 2010). During
EOLOS 2008, meteorological parameters and sand movements
were also monitored at the surface of the pampa, but in parallel
marine sediment traps were installed in the bay for collecting
the wind eroded particles falling into it. This not only demonstrated the aeolian origin of the lithic particles but also allowed the
assessment of the intensity and size distribution of the flux of
sedimenting particles (Flores-Aqueveque et al., 2014b). The fact
that these measurements were performed in parallel with those
at the particles source also allowed the calibration of a simple
model (Alfaro et al., 2011) designed to account for the uptake,
transport, and deposition within the bay of the particles initially
set into motion by the wind on the pampa’s surface. For testing
the sensitivity of the characteristics of the deposition flux of lithics
to the natural variability of the wind, the statistical distribution of
14 years (1991–2004) of hourly wind speeds measured at the surface of the pampa was determined (Flores-Aqueveque et al., 2012).
It was shown that the wind speed values of each individual year
can be distributed into three Weibull modes. The winds of the lowest of these modes are too weak to move the sand grains at the surface of the pampa, but this is not the case for the winds of the
intermediate and high speed modes which are able to erode the
arid surface and transport particles to the bay. The interannual
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
247
variability of the wind speed distributions is essentially the result
of the variations of the importance of the high speed mode (HSM,
which mobilizes particles >100 lm). The model of Alfaro et al.
(2011) shows that in periods when this HSM is not observed, only
particles with sizes between 35 and 100 lm can reach the point of
the bay from which the sediment cores analyzed in this study were
retrieved. Conversely, in other years particles larger than 100 lm
can be transported to the bay and their flux is directly linked to
the weight of the HSM in the overall distribution of the yearly
hourly wind speeds. For instance, between 1991 and 2004 this flux
is predicted to vary between 0 and close to 400 particles/cm2/year
whereas in the same time the flux of particles >35 lm varies
between 30 and 800 particles/cm2/year. Note that if the flux of particles >100 lm can be considered as a quantitative tracer of the
importance of the high speed mode at the surface of the pampa,
the variations of the flux of particles >35 lm are more complex
to interpret because they cannot be ascribed unequivocally to
either the intermediate or high speed modes. In the following,
we will consider the flux of particles >35 lm as an indicator of
the wind erosive activity at the surface of the pampa.
series, and S(f) is generally a continuous function of f, which means
that all the frequencies of the domain are necessary to reconstruct
the signal. One of the main interests of the method is that
examination of the frequency spectrum usually allows distinguishing individual peaks superimposed on a blurrier pattern. Though
not always easy to interpret physically, these frequencies play a
particular role in the temporal variability of the studied signal.
The application of the method requires that a value of the signal
is available in each temporal bin of the time window. As will be
shown below this condition is met by the time series of the deposition flux of particles >35 lm for which a value has been determined for each of the 245 years separating the bottom of the
core (AD 1754) from its top (AD 1998). Note that in order to detect
a possible evolution in the temporal structure of the signal, the
spectral analysis can be applied to time windows narrower than
the global one. For instance we will apply the method to the first,
and to the last, hundreds years of the period covered by the core.
Marine sediment core F981A
The core F981A was extracted in 1998 at a water depth of ca.
75 m b.s.l. in the central part of the basin (23°040 S–70°270 W)
(Fig. 1), which is to say in the area dominated by the hemipelagic
sedimentation of biogenic rests. A gravity box corer of 20 cm width
and approximately 50 cm length was used for its extraction. Then
the core was preserved at 2 °C before it was submitted to different
analyses.
Geochronological determinations from detailed 14C and 210Pb
data were obtained by Vargas et al. (2007). The determination of
the high resolution variability of the mineral content and its grain
size distribution was performed through a polarized-light microscope using an ocular lens of 10 and a magnification of 4 in four
thin sections prepared every 10 cm according to the resin replace
procedure described in Vargas et al. (2004). Due to dehydration,
each thin section had its length reduced by approximately 2 cm
after treatment.
A millimetric graduated grid located behind the thin section
helped demarcate the area to be studied.
These optical analyses complemented the X-ray observations
and provided valuable information about lamination thickness
and sedimentological structures.
The quantification of lithic particles, mainly quartz, feldspar and
amphibole, was completed millimetre to millimetre over an area of
1 mm2 approximately (Fig. 2A). Because they are strongly correlated to the strength of the southerly winds (see above), the counting
was focused on particles with sizes >35 lm and also on those
>100 lm. This counting was performed using an automated image
analysis software (Image Pro Plus, v6.2). Because the core was dated every 5 mm, the determination of the accumulation speed (in
mm/year) is easy to determine at each depth by interpolation.
Combining the number of particles counted in each mm2 with this
speed directly yields the deposition flux of particles (in particles/mm2/year) of each of the two counted size classes. Note that
if several values of the flux were obtained in the same year, these
values were averaged.
Stratigraphy of marine laminated sediments
Spectral analysis of the times series
Several methods can be used to extract crucial information from
univariate time-series (see Ghil et al., 2002, for a detailed review of
these methods and their pros and cons). One of the most popular is
based on the Fourier theory which stipulates that any given time
series can be associated to a function of the frequency, S(f), called
its spectral density. The frequency domain is bounded by a lower
limit whose value is the inverse of the length of the initial time
Results and discussion
The microscope observations indicate that the general structure
of these deposits is characterized by a succession of dark and light
laminae, of millimetric to centrimetric thickness with variable content in diatom and other biogenic rests, organic matter and lithic
particles, agreeing with previous similar observations described
by Vargas et al. (2004, 2007), and with those from Caniupán
et al. (2009) who found similar characteristics in core BC3D (23°
03.30 S–70°27.40 W), extracted near the core F981A. These layers
are limited by sharp and often irregular contacts (Fig. 2B). In some
dark layers it is possible to find lenses or bands of diatoms
(Fig. 2D), whereas in light layers there are nodules or agglomerates
of organic matter (Fig. 2C).
The lack of bioturbation and the sharpness of the contacts suggest that changes in lamination style are due to variations in
hemipelagic sedimentation processes in response to upwelling
events and the concomitant increase of primary production, as previously proposed by Vargas et al. (2004). The good state of preservation of the layers indicates that low oxygen levels in bottom
waters intercepted the sediments, thus restricting bioturbation
(Valdés et al., 2004; Vargas et al., 2004).
Dark laminae are characterized by higher contents of diatoms of
the Chaetoceros r.s. type (species related to upwelling events),
agglomerates of organic matter, foraminifer remains, and lithic
particles showing maximum sizes between 35 and 250 lm
approximately (Fig. 2A and C). Conversely, light laminae are
dominated by pennate and centric diatoms (Fig. 2D), and show a
lower lithics content with particles of smaller sizes. These observations coincide with those of Caniupán et al. (2009) and Vargas et al.
(2004) who also showed that the dark layers are relatively denser
and have lower water content than the light layers, which are composed of a more porous material. Dark laminae are generally
enriched in lithic particles, biogenic rests, organic carbon-content
and fish scales, suggesting increased upwelling intensity, phytoplankton production rates and oceanic productivity with respect
to light layers (Vargas et al., 2004; Valdés et al., 2008; Caniupán
et al., 2009).
Gray scale contact prints show that generally the layers are
arranged nearly horizontally, especially in the lower 30 cm. At
the top of the core, the layers are sloping gently (Fig. 3). Based
on the style of lamination, two different segments can be identified
(Fig. 3): the first 15 cm approximately (from the base to the top)
display massive, relatively dark laminations. This segment is followed by a sequence of fine to medium-thickness alternations of
248
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
Fig. 2. Features and components of the marine laminated sediments (thin sections of core F981A): (a) microscope images of core thin section taken at different depths, (b)
sharp and irregular contact between light (up) and dark layers (down), (c) typical dark layer containing lithic particles (back arrows), organic matter agglomerate (white
arrows), diatoms of Chaetoceros r.s. type (gray arrow) and foraminifera remains (dark gray arrow), (d) diatom lenses of Coscinodiscus sp.
dense dark and porous light layers in similar proportions but with
a dominance of dark facies toward the top. According to Bull and
Kemp (1996) the variations in the style of lamination could indicate that marine sediments record the changes in ocean–climate
conditions occurring at centennial to decadal, or even interannual
and seasonal, time-scales. In the present case, these changes could
represent variations in the frequency of the upwelling events and
of the associated primary productivity rate.
Considering the geochronological determination from this sediment core (Vargas et al., 2007), this segmentation suggests that
from at least the mid-18th century (AD 1754; bottom of the core)
until the second half of the 19th century (ca. AD 1878), the atmospheric and oceanographic conditions were relatively more stable
than later on. These conditions were characterized by comparatively long (15–70 years) periods of relatively weak winds
and moderate upwelling. This relatively steady period was followed by a phase, between AD 1878 and the top of the core (AD
1998), in which the conditions favoring the hemipelagic sedimentation processes became more frequent, especially toward the end
of the record (second half of the 20th century). The higher frequency of laminations in the first part of this segment, namely between
AD 1878 and AD 1900, suggests more rapid (<10 years) changes of
sedimentation conditions with a dominance of dark layers (Fig. 3).
This has been interpreted as being the result of relatively short
events of deposition driven by ENSO-like conditions during this
period (Vargas et al., 2007). From AD 1900 to the end of the record
(Fig. 3) the layers become thicker and dark facies dominate toward
the top. This suggests that from the second half of the 19th century
the conditions favoring an increase in primary productivity with
higher fluxes of upwelling-related biogenic species, as well as
relatively coarser terrestrial supply and increasing hypoxia within
the bay were dominant (Vargas et al., 2004).
High-resolution quantification of lithogenic components
As mentioned before (Section ‘Wind erosion and transport of
lithic material to the bay’), among the particles contained in the
laminated sedimentary record only those with sizes >35 and
>100 lm will be used for the reconstruction of the variability of
the southerly winds intensity in the coastal Atacama Desert.
The flux of particles >35 lm (Fig. 4A) tends to increase with
time, or equivalently from the bottom of the core to its top. The
10 years running mean of this flux is less than 2 particles/mm2/
year at the base but reaches values close to 8 particles/mm2/year
near the top. A similar increasing trend was observed by Vargas
et al. (2004) who determined the FTIR (Fourier Transform
InfraRed) mineral flux of two cores extracted from the centre of
Mejillones Bay.
Based on the values of the annual flux of particles >35 lm, three
different periods can be roughly distinguished. The first one (P1) corresponds to the lower 10 cm and extends approximately from AD
1754 to 1820. With a mean of less than 2.05 particles/mm2/year, a
mode of 2.35 particles/mm2/year, and a maximum of 4.69 particles/mm2/year (Fig. 4A), this period is characterized by its low and
relatively stable flux of lithics. The next period (P2) extending
approximately from AD 1820 to 1878 seems to constitute a
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
249
Fig. 3. Chronology and stratigraphy of core F981A. White and gray zones represent light and dark layers, respectively.
transition period leading to the long and most recent phase (P3) of
120 years (AD 1870–1998, approximately). P3 presents the highest
flux of lithics with a mean of 4.68 particles/mm2/year, a mode of
5.68 particles/mm2/year and a maximum of 10.72 particles/mm2/
year (Fig. 4A). This segmenting is somewhat arbitrary but it is consistent with the intervals proposed by Vargas et al. (2007) who, on the
basis of the mineral flux FTIR analysis of the Mejillones Bay cores,
defined two stages separated by an upwelling transition period running from AD 1820 to 1878.
The time series of particles >100 lm also displays an increase of
abundance toward the end of the 20th century (Fig. 4A). From the
beginning of the record to AD 1900 approximately, the frequency
of coarse particle deposition as well as the magnitude of the flux
are very low. Coarse particles are even totally absent in the periods
AD 1754–1770 and AD 1883–1903. Between AD 1903 and 1963 the
frequency of particles >100 lm increases as denoted by the
increase of the 10 years running mean, but the yearly flux itself
remains relatively low, with a maximum of 1.42 particles/mm2/
year. From AD 1963 to the end of the record the deposition of
coarse particles shows an important increase in both frequency
and amounts, reaching a maximum of 4.00 particles/mm2/year
(Fig. 4A). Noteworthy is the fact that these values and those of
the flux of particles >35 lm are within the ranges of fluxes predicted (see above) by the transport model of Alfaro et al. (2011).
250
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
Fig. 4. (A) Yearly values of the deposition fluxes of lithic particles larger than 35 lm (red line), and than 100 lm (blue line). The 10 years running means of the two time series
are also represented (note that for the sake of readability the values of the running mean for particles >100 lm are reported on the secondary vertical axis). (B) IPO annual
mean time series (dark gray bands highlight the cyclicity). (C) Mineral flux (FTIR) and organic carbon flux time series. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
A comparison of the segments defined independently on the
basis of the (1) lithic content quantification, and (2) stratigraphic
description with those proposed by Vargas et al. (2007) indicates
that there is a good correlation between the direct quantification
and the FTIR mineral flux analyses. However, the periods defined
by the stratigraphic analysis show important differences relative
to the other two methods indicating that if the stratigraphic analysis is a good help to interpret the information contained in cores it
must also be complemented with more precise data.
In order to assess more precisely the time-scales involved in the
changes of wind intensity regimes indicated by the variations of
the magnitude and size-distribution of the flux of lithic particles
>35 lm, the spectral analysis method has been applied to the
245 year-long (AD 1754–1998) time series. The periodogram
(Fig. 5A) displays three marked peaks centered on 16, 27, and
82 years showing the importance of these decadal periods in the
reconstruction of the observed signal. Secondary peaks corresponding to periods shorter than 10 years are also present indicating the influence of interannual oscillations of the ENSO type.
Finally, a significant portion of the spectrum is occupied by periods
larger than 100 years. This seems to suggest the presence of a centennial transition in the last 250 years but a core covering a longer
time period would be necessary to ascertain this point.
The causes of these oscillations are not known but they can
probably be linked to already documented large scales oscillations,
such as the PDO. In an attempt to detect commonalities in the temporal scales of the deposition of lithics and the PDO, the spectral
analysis has been applied to the annual mean of the Interdecadal
Pacific Oscillation (IPO). The concept of the IPO was introduced by
Power et al. (1999) based on the work of Folland et al. (1999).
Shortly after these preliminary works, Folland et al. (2002) showed
the near equivalence of the PDO and IPO. A recently updated (2008)
version of the monthly IPO index for years 1871–2008 is distributed
online (www.iges.org/c20c/IPO_v2.doc) by the UK Met Office. Note
that because we do not want to discard the possibility of observing
the interannual variability, we are using in this work the unfiltered
version of the IPO time series rather than the low-pass (11 years)
filtered one also proposed online. The obtained periodogram
(Fig. 5B) is characterized by the presence of major peaks centered
approximately on 20–28, and on 50 years. More rapid (interannual)
oscillations are also present in the spectrum. The presence of similar decadal oscillations with periods between 20 and 30 years in the
periodograms of the lithics and IPO time series suggest an influence
of the IPO on the wind regimes at the surface of Pampa Mejillones.
This also can be observed in Fig. 4B where a coincidence between
the cycles of both series is noticeable.
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
251
Fig. 5. Spectral analyses of the lithic fluxes and IPO. (A) Annual flux of particles >35 lm for the entire core, (B) non filtered time series of the IPO index (annual values) for the
period (1871–2008), (C) annual flux of particles >35 lm for the period (1754–1853 AD), (D) annual flux of particles >35 lm for period (1899–1998 AD).
If one takes into account that the theoretical upper limit of the
periods which can be retrieved by the analysis of the IPO time series is shorter (138 years) than in the case of the lithics (245 years),
the presence of a multidecadal oscillation in the two periodograms
(80 years for the lithics and 50 for the IPO) also seems to support
the hypothesis of an influence of the global scale on regional winds.
When applied to the first (1754–1853 AD) and to the last (AD
1899–1998) hundred years of the period, the spectral analysis
yields significantly different periodograms (Fig. 5C and D). In the
most recent period, almost all the variability of the lithic flux can
be explained by decadal oscillations, and interannual variability
seems to play only a secondary role. Conversely, in the first
100 years of the period, if the slow multidecadal increase of the
flux is well denoted by the importance in the periodogram of periods larger than 50 years, the relative weights of the interannual
and decadal oscillations are much more balanced than in the
20th century. This shows that the increase in the deposition of
lithics >35 lm which occurred in Mejillones Bay during the 19th
century was accompanied by the reinforcement of the influence
of the decadal variability and the relative decreasing importance
of the interannual one. Unfortunately, the lack of IPO monthly values before AD 1871 prevents checking if the change in the temporal modes of lithics deposition also affected the Pacific Oscillation.
Implications for the variability of the wind stress
In good agreement with the physics underlying the modeling
(Alfaro et al., 2011) of the transport and deposition of lithic particles into the Mejillones Bay, the examination of the yearly fluxes
(Fig. 4A) shows that when particles >35 lm are not observed,
particles >100 lm are also absent. This corresponds to years when
the wind stress remained constantly below the threshold of erosion over the pampa’s surface. The fact that there are 7 such years
(AD 1760, 1761, 1769, 1777, 1783, 1791, and 1797) of particularly
low winds (hereinafter referred to as ‘low wind years’, or LWY) in
the second half of the eighteenth century that compare with only 3
(AD 1813, 1847 and 1876) and 4 (AD 1900, 1916, 1946, and 1975)
in the course of the nineteenth and twentieth centuries, respectively, is consistent with the general intensification of wind erosion
already commented. In a similar manner, the years when erosion
did occur can be classified into two categories: (1) years of intermediate winds only (IWY), during which particles >35 lm are
transported to the bay but not particles >100 lm, and (2) years
with strong winds (SWY) when particles >35 and >100 lm are
observed simultaneously. As shown by Flores-Aqueveque et al.
(2012), the statistical distributions of hourly wind speeds of SWY
years are characterized by a significant contribution of the
high-speed mode, which is to say 10-m windspeeds exceeding
12 m/s. Table 1 presents the relative distribution of LWY, IWY
and SWY years in the three defined time intervals.
Again, P2 (AD 1820–1878) century appears as a period of transition between (1) the second half of the 18th century until AD
1820 when LWYs were particularly frequent (12% of the cases)
and SWYs scarce (14%), and (2) the 20th century (AD 1878–
1998) when very few cases of LWYs are observed (3%). In the latter
case the proportion of SWYs over the total (120 years) reaches 29%
(Fig. 6A).
As it can be seen from the flux of particles >100 lm depicted in
Fig. 4A, the frequency of observation of SWYs increases from AD
1754 to approximately AD 1984 but their intensity remains in
252
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
Table 1
Numbers (and proportions (%)) of low winds (LWY), intermediate winds (IWY), and
strong winds (SWY) years detected in the three defined periods. The classification is
based on the observation (or not) of particles >35 and >100 lm in the laminae of the
marine core (see text for details).
Period (AD)
Duration (yrs)
LW years
IW years
SW years
1754–1820
1820–1878
1878–1998
66
58
120
8 (12%)
2 (3%)
4 (3%)
49 (74%)
45 (78%)
81 (68%)
9 (14%)
11 (19%)
35 (29%)
the same order of magnitude with an average flux of 1.30 ± 0.10
part >100 lm per mm2/year for each SWY in this period. In the last
30 years of the period of study, four particular years (AD 1982,
1983, 1994, and 1997) are well above this average with a flux of
3.03 ± 0.65 particles/mm2/year. This indicates that in addition to
their frequency, the intensity of the SWY has also started to
increase in the last 30 years of the 20th century. The fact that all
the 4 years cited above coincide with El Niño events classified as
‘strong’ (1982–1983, 1997–1998) or ‘moderate’ (1994–1995) by
the NOAA (http://www.cpc.ncep.noaa.gov/) might be coincidental
but it also suggests that constructive interferences of the decadal
and interannual oscillations might create in the coastal area of
the subtropical southeastern Pacific conditions for the presence
of stronger than usual winds. This is consistent with reduced
low-cloud cover after the cold to warm PDO transition in the
mid-1970s associated with a slight increase in sea-surface temperature at Antofagasta (Rutllant et al., 1998), that resulted in
stronger southerlies. A similar strengthening of the winds associated with a decrease in low-cloud cover during El Niño was reported
for Lima, Perú by Enfield (1981).
Paleoclimatic and paleoceanographic linkages
While there is little doubt about the expected general correlation
between sea surface temperature variability and primary productivity in coastal upwelling areas, controversial views arise when
projecting upwelling-favorable wind variability and trends in climate change simulations along the tropical and subtropical west
coast of South America induced by increased concentrations of
greenhouse gases in the atmosphere.
Vecchi et al. (2006) observe and predict a weakening of the
Walker circulation that could possibly result in a weakening of
the SEPSA. However, as mentioned in the preceding paragraph, a
similar weakening also characterizes the warm phase of the
ENSO cycle during which stronger winds have been observed along
the coast of Perú (e.g., Enfield, 1981), results that are consistent
within El Niño-like longer time scale oscillations in northern
Chile (Rutllant et al., 1998; Vargas et al., 2007). In all these cases,
an increase in the land–sea thermal contrast enhancing alongshore
wind intensity has been suggested in connection with a decrease in
low-cloud cover, departing from the classical Bakun’s (1990)
hypothesis.
Strengthening of the upwelling-favorable winds concomitant
with lower sea surface temperatures (SSTs) along the coast have
been reported for subtropical Chile by Falvey and Garreaud
(2009) since the last quarter of the 20th century. In these cases,
as also documented in a recent coastal wind climatology by Rahn
and Garreaud (2013), the main mechanism driving alongshore
winds in subtropical and tropical Chile would be the alongshore
pressure gradients conditioned by the poleward expansion of the
subtropical edge of the Hadley cell (e.g., Johanson and Fu, 2009).
More recently Belmadani et al. (2013) claim a projected strengthening of the winds off central Chile south of 30–35°S and a weakening elsewhere along the arid west coast of South America.
Moreover, high-resolution alkenone time series document a
decrease in SST along the main upwelling areas off Northern
Chile and Peru (23°S and 14°S, respectively) (Vargas et al.,
2007; Gutiérrez et al., 2009, 2011). This reconstructed cooling,
especially observed since the mid-20th century, coincides in sign
and magnitude with the changes of coastal SST measured directly
since 1979 (Falvey and Garreaud, 2009) suggesting that the entire
region responds to the same regional-scale forcing processes that
could be related with an intensification of the SEPSA (Falvey and
Garreaud, 2009). In the study area, proxies of primary productivity
and lithic mineral fluxes in marine sediment cores (F981A and
BC3D) from the Mejillones Bay, show an increasing trend in phytoplankton productivity since the late 19th century (Vargas et al.,
Fig. 6. Comparison between the frequencies of the strong wind years (SWY) and mean organic carbon variations.
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
253
Fig. 7. Windstress (historical data) and number of lithic particles (>35 + >100 lm) time series comparison since 1959 (black lines superimposed to lithic particles time series
indicate shifts in mean values).
2007; Caniupán et al., 2009). In particular, they defined two main
lapses of productivity: a pre-1820 period characterized by low
organic carbon and low ‘‘chlorins’’ indicating relatively low primary productivity and a post-1877 period where the proxies showed
an increase in variability as well as in magnitude (Fig. 4C; Vargas
et al., 2007; Caniupán et al., 2009).
As depicted in Fig. 4, these periods are consistent with our results,
confirming an increase of the southerly wind intensity and, consequently, of the coastal upwelling in the studied area after the late
1870s. Moreover, when comparing the frequency of strong wind
years (>12 m/s) determined from the lithic content (ratio
>100/>35 lm) and the mean organic carbon for these periods
(Fig. 6) a remarkable relationship can be noted. From the first (AD
1754–1820) to the second period (AD 1820–1878) an increase of
5.3% in frequency of SWYs can be associated with an increase of
0.197 mg/cm2/yr (9.5%) in the average organic carbon, whereas in
the transition from the second to the third period (AD 1878–1998)
a rise of 10.2% in the SWYs relates to an intensification of
0.261 mg/cm2/yr (11.4%) in the average organic carbon. These
results indicate that the export/preserved productivity increased
22% between the pre-1820 and post-1878 periods, related with the
increase in southerly winds (114% in the frequency of SWYs), and
also with the establishment of oceanographic conditions that
favored the preservation of increased fluxes of fish scales to the sea
bottom since AD 1878 (Valdés et al., 2008; Díaz-Ochoa et al., 2011).
Furthermore, when analyzing in detail the aeolian lithic particles (>35 lm) time series for the period for which historical wind
data (windstress) are available (since 1959) (Fig. 7), it is possible
to see that the documented shift in windstress (and low-cloud cover) since the year 1976 (Rutllant et al., 1998) is consistent with the
rise in lithic content from the mid-1970s.
Conclusion
Aeolian particles have been widely used to interpret past
ocean–climate variability from many sites and sedimentary
records, however, we have few opportunities for developing quantitative reconstructions of paleowind intensity and indeed paleocli
mate–paleoceanography. The examination of time series of lithic
particles >35 lm from laminated sediments in Mejillones Bay
(23°S), spanning the period between AD 1754–1998, confirms that
the 19th century in the broad sense has marked a transition period
between the second half of the 18th century, characterized by
relatively moderate winds, and the 20th century in which erosion
events on the surrounding hyperarid coastal Atacama Desert were
more frequent, and possibly stronger. This last assumption is confirmed by the time series of particles >100 lm which features an
intensification during the 20th century, and particularly in the last
30 years, of both frequency and amplitude of strong erosive events.
Spectral analysis on the lithic particles-time series shows that
the 19th century constitutes a transition from a period when the
influence of interannual ENSO and interdecadal PDO variability
were relatively well balanced, to a period since AD 1878 when decadal variability dominates. This reinforcement of the decadal oscillations in the last centuries is most probably at the origin of the
general strengthening and larger variability of the eroding winds
along the coast of the Atacama Desert. However, as denoted by
the fact that in the second half of the 20th century the years of
larger depositional fluxes of particles >100 lm coincide with El
Niño, it is suggested that constructive interference between decadal and interannual oscillations might partly explain the occurrence of unusually strong winds along Northern Chile. Our
quantitative reconstruction of upwelling favorable wind variability
agrees with previous results based on organic and inorganic proxies which recognized a period of lower phytoplankton productivity
before AD 1820 and increased productivity since AD 1878 (Vargas
et al., 2004, 2007; Caniupán et al., 2009), in the context of decreasing trend in SSTs during the 20th century off Northern Chile and
Central Peru, associated to an intensification of the SEPSA and to
enhanced alongshore winds due to increased sea–land temperature gradients (Vargas et al., 2007; Gutiérrez et al., 2011).
When comparing the frequency of SWYs (winds >12 m/s) and
the sedimentary organic carbon fluxes, we evidence a relationship
in which an increase from 5.3% to 10.2% in SWYs, between the
pre-1820 and post-1878 periods, relates to a rise from 9.5% to
11.4% in the mean organic carbon content in laminated sediments.
These results indicate that the export/preserved productivity rate
can be quantitatively related with the southerly wind intensity,
and then with the ocean–climate system.
Acknowledgements
This work was carried out within the framework of scientific
international collaboration between ECOS-SUD and CONICYT, and
funded by project FONDECYT N° 11121543. Authors also thank
M. Torres, R. Melon, M. Otu and A. Turbina.
254
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
References
Aceituno, P., Montecinos, A., 1993. Stability analysis of the relationship between the
Southern Oscillation and rainfall in South America. Bulletin de l’Institut Français
d’Etudes Andines 22, 53–64.
Alfaro, S., Flores-Aqueveque, V., Foret, G., Caquineau, S., Vargas, G., Rutllant, J., 2011.
A simple model accounting for the uptake, transport, and deposition of
wind-eroded mineral particles in the hyperarid coastal Atacama Desert of
northern Chile. Earth Surface Processes and Landforms 36 (7), 923–932.
Allan, R., 2000. ENSO and climatic variability in the past 150 years. In: Diaz, H.,
Markgraf, V. (Eds.), El Niño and the Southern Oscillation: Multiscale Variability
and Global and Regional Impacts. Cambridge University Press, Cambridge, pp.
3–35.
Bakun, A., 1990. Global climate change and intensification of coastal ocean
upwelling. Science 247, 198–201.
Belmadani, A., Echevin, V., Codron, F., Takahashi, K., Junquas, C., 2013. What
dynamics drive future wind scenarios for coastal upwelling off Peru and Chile?
Climate Dynamics, 1–22. http://dx.doi.org/10.1007/s00382-013-2015-2.
Berger, W.H., Smetacek, V.S., Wefer, G., 1989. Productivity of the Ocean: Present and
Past, vol. xviii. Wiley-Interscience, New York, 470 pp.
Boninsegna, J.A., Argollo, J., Aravena, J.C., Barichivich, J., Christie, D., Ferrero, M.E.,
Lara, A., Le Quesne, C., Luckman, B.H., Masiokas, M., Morales, M., Oliveira, J.M.,
Roig, F., Srur, A., Villalba, R., 2009. Dendroclimatological reconstructions in
South America: a review. Palaeogeography, Palaeoclimatology, Palaeoecology
281, 210–228.
Bull, D., Kemp, A.E.S., 1996. Composition and origins of laminae in late Quaternary
and Holocene sediments from the Santa Barbara Basin. In: Kemp, A.E.S. (Ed.),
Palaeoclimatology and Palaeoceanography from Laminated Sediments, Geol.
Soc. Spec. Pub., vol. 116. The Geological Society Publishing House, Bath, UK, pp.
143–146.
Cane, M.A., 1998. Climate change – a role for the tropical Pacific. Science 282 (5386),
60–61.
Cane, M.A., 2005. The evolution of El Niño, past and future. Earth and Planetary
Science Letters 230 (3–4), 237–240.
Caniupán, M., Villaseñor, T., Pantoja, S., Lange, C.B., Vargas, G., Muñoz, P., Salamanca,
M., 2009. Sedimentos laminados de la Bahía Mejillones como registro de
cambios temporales en la productividad fitoplanctónica de los últimos 200
años. Revista Chilena de Historia Natural 82, 83–96.
Chauhan, O., Vogelsang, E., Basavaiah, N., Kader, U.S.A., 2010. Reconstruction of the
variability of the southwest monsoon during the past 3 ka, from the continental
margin of the southeastern Arabian Sea. Journal of Quaternary Science 25 (5),
798–807.
Díaz-Ochoa, J.A., Pantoja, S., De Lange, G.J., Lange, C.B., Sánchez, G.E., Acuña, V.R.,
Muñoz, P., Vargas, G., 2011. Oxygenation variability in Mejillones Bay, off
northern Chile, during the last two centuries. Biogeosciences 8, 137–146.
Dietrich, S., Seelos, K., 2010. The reconstruction of easterly wind directions for the
Eifel region (Central Europe) during the period 40.3–12.9 ka BP. Climate of the
Past 6, 145–154.
Elliott, W., Angell, J., 1988. Evidence for changes in Southern Oscillation
relationships during the last 100 years. Journal of Climate 1, 729–737.
Enfield, D., 1981. Thermally driven wind variability in the planetary boundary layer
above Lima, Peru. Journal of Geophysical Research 86 (C3), 2005–2016.
Espizua, L.E., Pitte, P., 2009. The Little Ice Age glacier advance in the Central Andes
(35°S), Argentina. Palaeogeography, Palaeoclimatology, Palaeoecology 281,
345–350.
Falvey, M., Garreaud, R.D., 2009. Regional cooling in a warming world: recent
temperature trends in the southeast Pacific and along the west coast of
southtropical South America (1979–2006). Journal of Geophysical Research
114, D04102.
Flores-Aqueveque, V., Alfaro, S., Muñoz, R., Rutllant, J., Caquineau, S., Le Roux, J.,
Vargas, G., 2010. Aeolian erosion and sand transport over the Mejillones
Pampa in the coastal Atacama Desert of northern Chile. Geomorphology 120,
312–325.
Flores-Aqueveque, V., Alfaro, S., Caquineau, S., Foret, G., Vargas, G., Rutllant, J., 2012.
Inter-annual variability of surface winds in the coastal part of the Atacama
Desert (23°S, north Chile): implications for the export of aeolian material to a
nearby bay. Sedimentology 59, 990–1000.
Flores-Aqueveque, V., Caquineau, S., Alfaro, S., Valdés, J., Vargas, G., 2014a. Using
image-based size-analysis for determining the size-distribution and flux of
aeolian particles sampled in coastal northern Chile (23°S). Journal of
Sedimentary Research 84, 238–244.
Flores-Aqueveque, V., Caquineau, S., Alfaro, S., Valdés, J., Vargas, G., 2014b.
Assessing the origin and variability of aeolian lithic material for
high-resolution paleoceanographic reconstructions off northern Chile. Journal
of Sedimentary Research 84, 897–909.
Folland, C.K., Parker, D.E., Colman, A.W., Washington, R., 1999. Large scale modes of
ocean surface temperature since the late nineteenth century. In: Navarra, A.
(Ed.), Beyond El Nino: Decadal and Interdecadal Climate Variability, vol. 73-120.
Springer-Verlag, Berlin, p. 374.
Folland, C.K., Renwick, J.A., Salinger, M.J., Mullan, A.B., 2002. Relative influences of
the interdecadal Pacific Oscillation and ENSO on the South Pacific Convergence
Zone. Geophysical Research Letters 29 (13), 21-1–21-4.
Garreaud, R.D., Battisti, D.S., 1999. Interannual (ENSO) and interdecadal (ENSO-like)
variability in the Southern Hemisphere tropospheric circulation. Journal of
Climate 2, 2113–2123.
Garreaud, R.N., Vuille, M., Compagnucci, R., Matengo, J., 2009. Present-day South
American climate. Palaeogeography, Palaeoclimatology, Palaeoecology 281,
180–195.
Ghil, M., Allen, M.R., Dettinger, M.D., Ide, K., Kondrashov, D., Mann, M.E., Robertson,
A.W., Saunders, A., Tian, Y., Varadi, F., Yiou, P., 2002. Advanced spectral methods
for climatic time series. Reviews of Geophysics 40 (1), 1003.
Grimm, K.A., Lange, C.B., Gill, A.S., 1997. Self-sedimentation of phytoplankton
blooms in the geologic record. Sedimentary Geology 119, 151–161.
Gutiérrez, D., Sifeddine, A., Field, D.B., Ortlieb, L., Vargas, G., Chavez, F.P., Velazco, F.,
Ferreira, V., Tapia, P., Salvatteci, R., Boucher, H., Morales, M.C., Valdés, J., Reyes,
J.-L., Campusano, A., Boussafir, M., Mandeng-Yogo, M., García, M., Baumgartner,
T., 2009. Rapid reorganization in ocean biogeochemistry off Peru towards the
end of the little ice age. Biogeosciences 6, 835–848.
Gutiérrez, D., Bouloubassi, I., Sifeddine, A., Purca, S., Goubanova, S., Graco, M., Field,
D., Méjanelle, L., Velazco, F., Lorre, A., Salvatteci, R., Quispe, D., Vargas, G.,
Dewitte, B., Ortlieb, L., 2011. Coastal cooling and increased productivity in the
main upwelling zone off Peru since the mid-twentieth century. Geophysical
Research Letters 38, L07603.
Horii, T., Hanawa, K., 2004. A relationship between the timing of El Niño onset and
subsequent evolution. Geophysical Research Letters 31, 1–4.
Jansen, E., Overpeck, J., Briffa, K.R., Duplessy, J.-C., Joos, F., Masson-Delmotte, V.,
Olago, D., Otto-Bliesner, B., Peltier, W.R., Rahmstorf, S., Ramesh, R., Raynaud, D.,
Rind, D., Solomina, O., Villalba, R., Zhang, D., 2007. Palaeoclimate. In: Solomon,
S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller,
H.L. (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change. Cambridge University Press, pp. 433–497.
Jenny, B., Valero-Garcés, B., Villa-Martínez, R., Urrutia, R., Geyh, M., Veit, H., 2002.
Early to mid-Holocene aridity in central Chile and the southern westerlies: the
Laguna Aculeo record (34°S). Quaternary Research 58, 160–170.
Johanson, C.M., Fu, Q., 2009. Hadley cell widening: model simulations versus
observations. Journal of Climate 22, 2713–2725. http://dx.doi.org/10.1175/
2008JCLI2620.1.
Keeling, R.F., Shertz, S.R., 1992. Seasonal and interannual variations in atmospheric
oxygen and implications for the global carbon cycle. Nature 358, 723–727.
Le Quesne, C., Acuña, C., Boninsegna, J.A., Rivera, A., Barichivich, J., 2009. Long-term
glaciervariations in the Central Andes of Argentina and Chile, inferred from
historical records and tree-ring reconstructed precipitation. Palaeogeography,
Palaeoclimatology, Palaeoecology 281, 334–344.
Luterbacher, J., Dietrich, D., Xoplaki, E., Grosjean, M., Wanner, H., 2004. European
seasonal and annual temperature variability, trends, and extremes since 1500.
Science 303, 1499–1503.
Mann, M.E., Bradley, R.S., Hughes, M.K., 1999. Northern hemisphere temperatures
during the past millennium: inferences, uncertainties, and limitations.
Geophysical Research Letters 26, 759–762.
Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C., 1997. A Pacific
interdecadal climate oscillation with impacts on salmon production. Bulletin of
the American Meteorological Society 78, 1069–1079.
Masiokas, M.H., Rivera, A., Espizua, L.E., Villalba, R., Delgado, S., Aravena, J.C., 2009.
Glacier fluctuations in extratropical South America during the past 1000 years.
Palaeogeography, Palaeoclimatology, Palaeoecology 281, 242–268.
Moberg, A., Sonechkin, D.M., Holmgren, K., Datsenko, N.M., Karlén, W., 2005.
Highlyvariable Northern Hemisphere temperatures reconstructed from lowand high-resolution proxy data. Nature 433, 613–617.
Ortlieb, L., Escribano, R., Follegati, R., Zúñiga, O., Kong, I., Rodríguez, L., Valdés, J.,
Guzmán, N., Iratchet, P., 2000. Recording ocean–climate changes during the last
2,000 years in a hypoxic marine environment off northern Chile (238S). Revista
Chilena de Historia Natural 73, 221–242.
Power, S., Casey, T., Folland, C.K., Colman, A., Mehta, V., 1999. Inter-decadal
modulation of the impact of ENSO on Australia. Climate Dynamics 15, 319–323.
Rahn, D., Garreaud, R., 2013. A synoptic climatology of the near-surface wind along
the west coast of South America. International Journal of Climatology 34, 780–
792. http://dx.doi.org/10.1002/joc.3724.
Rasmusson, E., Carpenter, T., 1983. The relationship between Eastern equatorial
Pacific sea surface temperatures and rainfall over India and Sri Lanka. Monthly
Weather Review 111, 517–528.
Rodbell, D.T., Seltzer, G.0., Anderson, D.M., Abbott, M.B., Enfteld, D.B., Newman, J.H.,
1999. An 15,000-year record of EI Nino-driven alluviation in southwestern
Ecuador. Science 283, 517–520.
Rutllant, J., Fuenzalida, H., Torres, R., Figueroa, D., 1998. Interacción océano-atmós
fera-tierra en la Región de Antofagasta (Chile, 23°S): Experimento DICLIMA.
Revista Chilena de Historia Natural 71, 405–427.
Rutllant, J., Fuenzalida, H., Aceituno, P., 2003. Climate dynamics along the arid
northern coast of Chile: the 1997–1998 Dinámica del Clima de la Región de
Antofagasta (DICLIMA) experiment. Journal of Geophysical Research –
Atmospheres 108 (D17), 4358. http://dx.doi.org/10.1029/2002JD003357.
Strub, T., Mesías, J., Montecino, V., Rutllant, J., Salinas, S., 1998. Coastal ocean
circulation off western South America. In: Robinson, A.R., Brink, K.H. (Eds.), The
Sea: Coastal Oceans, vol. 11. Wiley, New York, pp. 273–313.
Stuut, J.-B.W., Prins, M.A., Schneider, R.R., Weltje, G.J., Jansen, J.H.F., Postma, G.,
2002. A 300-kyr record of aridity and wind strength in southwestern Africa:
inferences from grain-size distributions of sediments on Walvis Ridge, SE
Atlantic. Marine Geology 180, 221–233.
Valdés, J., Sifeddine, A., Lallier-Verges, E., Ortlieb, L., 2004. Petrographic and
geochemical study of organic matter in superficial laminated sediments from an
V. Flores-Aqueveque et al. / Progress in Oceanography 134 (2015) 244–255
upwelling system (Mejillones del Sur Bay, Northern Chile). Organic
Geochemistry 35, 881–894.
Valdés, J., Ortlieb, L., Gutiérrez, D., Marinovic, L., Vargas, G., Sifeddine, A., 2008. A
250-years record of sardine and anchovy scale deposition in Mejillones Bay,
northern Chile. Progress in Oceanography 79, 198–207.
Vargas, G., Ortlieb, L., Rutllant, J., 2000. Aluviones históricos en Antofagasta y su
relación con eventos El Niño/Oscilación del Sur. Revista Geológica de Chile 27
(2), 155–174.
Vargas, G., Ortlieb, L., Pichon, J.J., Bertaux, J., Pujos, M., 2004. Sedimentary facies and
high resolution primary production inferences from laminated diatomaceous
sediments off northern Chile (23°S). Marine Geology 211, 79–99.
Vargas, G., Pantoja, S., Rutllant, J., Lange, C., Ortlieb, L., 2007. Enhancement of coastal
upwelling and interdecadal ENSO-like variability in the Peru–Chile current
since late 19th century. Geophysical Research Letters 34, L13607. http://
dx.doi.org/10.1029/2006 GL028812.
255
Vecchi, G.A., Soden, B.J., Wittenberg, A., Held, I.M., Leetma, A., Harrison, M.J., 2006.
Weakening of tropical Pacific atmospheric circulation due to anthropogenic
forcing. Nature 425, 73–76.
Villalba, R., Grosjean, M., Kiefer, T., 2009. Long-term multi-proxy climate
reconstructions and dynamics in South America (LOTRED-SA): state of the art
and perspectives. Palaeogeography, Palaeoclimatology, Palaeoecology 281,
175–179.
Vimeux, F., Ginot, P., Schwikowski, M., Vuille, M., Hoffmann, G., Thompson, L.G.,
Schotterer, U., 2009. Climate variability during the last 1000 years inferred from
Andean ice cores: a review of methodology and recent results.
Palaeogeography, Palaeoclimatology, Palaeoecology 281, 229–241.
Zhang, Y., Wallace, J.M., Battisti, D.S., 1997. ENSO-like interdecadal variability:
1900–1993. Journal of Climate 10 (5), 1004–1020.