See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270338233

A composite sea surface temperature record of the northern South China Sea
for the past 2500 years: A unique look into seasonality and seasonal climate
changes during warm and cold...

Article  in  Earth-Science Reviews · February 2015

DOI: 10.1016/j.earscirev.2014.12.003

CITATIONS READS

23 844

3 authors, including:

Hong Yan
Institute of Earth Environment---Chinese Academy of Sciences
52 PUBLICATIONS   591 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Tridacna Geochemistry and Paleoclimate Research View project

Forum on Nuclear Cooperation In Asia (FNCA) View project

All content following this page was uploaded by Hong Yan on 12 February 2018.

The user has requested enhancement of the downloaded file.

 Earth-Science Reviews 141 (2015) 122–135

Contents lists available at ScienceDirect

Earth-Science Reviews
journal homepage: www.elsevier.com/locate/earscirev

A composite sea surface temperature record of the northern South China

Sea for the past 2500 years: A unique look into seasonality and seasonal
climate changes during warm and cold periods
Hong Yan a,⁎, Willie Soon b, Yuhong Wang c
a
State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences, Xi'an 710075, China
b
Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138, USA
c
Department of Chemistry and Laser Chemistry Institute, Fudan University, Shanghai 200433, China

a r t i c l e i n f o a b s t r a c t

Article history: High-resolution late Holocene climate records that can resolve seasonality are essential for confirming past cli-
Received 4 July 2014 matic dynamics, understanding the late 20th century global warming and predicting future climate. Here a
Accepted 7 December 2014 new composite record of the sea surface temperature, SST, variation in the northern South China Sea (SCS) during
Available online 13 December 2014
the late Holocene is constructed by combining seven seasonally-resolved coral and Tridacna gigas Sr/Ca-based
SST time-windows with the instrumental SST record from modern interval between 1990 and 2000. This com-
Keywords:
South China Sea
posite multi-proxy marine record, together with the reconstructions from mainland China and tropical Western
Coral Pacific, indicates that the late Holocene warm periods, the Roman Warm Period (RWP) and Medieval Warm Pe-
Tridacna gigas riod (MWP), were prominently imprinted and documented in the climatic and environmental history of the East
Sea surface temperature Asia–Western Pacific region. Meanwhile, substantial and significant SST seasonality variations during the late Ho-
Seasonal climate locene were observed in the composite record. The observed increase in seasonality (or amplitude of seasonal cy-
Late Holocene cles) during the cold periods around our study area was probably caused by the different amplitudes between
winter versus summer SST variations in northern SCS, with much larger SST variation during winters than during
summers for the late Holocene. In addition, the distinctive warm, cold and neutral climatic episodes identified in
our northern SCS composite SST record correspond well with other paleo reconstructions from mainland China
and especially well with the Northern Hemisphere-wide composites by Moberg et al. (2005) and Ljungqvist
(2010). The overall agreement however also calls for more information and insights on how seasonal tempera-
tures and their ranges vary on decadal–centennial timescales.
© 2014 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
2. Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
2.1. Regional setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
2.2. Data sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
2.3. The method of compositing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1. A composite proxy-SST record in the northern SCS during the late Holocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.2. Intercomparison with other SST proxies from tropical Western Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.3. Distribution of RWP and MWP in the Asia-Pacific region with a comparison to other regions . . . . . . . . . . . . . . . . . . . . . . 127
3.4. Variation of the SST seasonality in the SCS during the late Holocene and its relationship with the mean climate state . . . . . . . . . . . 128
3.5. Different amplitudes of summer and winter SST variations during the late Holocene . . . . . . . . . . . . . . . . . . . . . . . . . . 130
3.6. Mechanism for the different variation amplitudes of the summer and winter SST during the late Holocene . . . . . . . . . . . . . . . . 131
4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

⁎ Corresponding author.
E-mail address: yanhong@ieecas.cn (H. Yan).

http://dx.doi.org/10.1016/j.earscirev.2014.12.003
0012-8252/© 2014 Elsevier B.V. All rights reserved.
 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135
1. Introduction
High-resolution climate reconstructions during the late Holocene,
especially during the past 2000 years, can provide a timely opportunity
to place the current global warming in a long-term context and help un-
derstand the importance of natural and anthropogenic forcings on past
and future climate changes. Many recent works have focused on the cli-
matic changes during this time period, but the complete characteristic
of the seasonal climate during the late Holocene, as well as the historical
status of the late 20th century global warming, remain debated (IPCC,
2001; Esper et al., 2002; Ge et al., 2003; Soon et al., 2003; IPCC, 2007;
Wang et al., 2007). The insufficiencies of the currently available proxy
data for the late Holocene are the major cause of ongoing debates and
controversies. For example, although more than 1200 proxy reconstruc-
tions were used in the compilation of Mann et al. (2008), only 25 re-
cords cover the last 2000 years. Meanwhile, most of the proxy records
are derived from the mid and high latitudes of the Northern Hemi-
sphere, and the temperature reconstructions from the Southern Hemi-
sphere and tropics are less common (IPCC, 2007; Mann et al., 2008,
2009). Furthermore, while tree-ring-proxy records during the late Holo-
cene have been built for the vast terrestrial area over the last 20 years,
high-resolution proxy ocean temperature records of the late Holocene
are still very limited. Thus a more concerted study on proxy records of
the late Holocene for the past 2000–3000 years, in particular high-
resolution, seasonally-resolved, ocean temperature records from South-
ern Hemisphere and tropics, are needed.
In addition to the variations of mean climatic state [see e.g., two early
papers for discussion on the inadequacy of the current definition and
framework basing on the statistical 30-year averages (Landsberg,
1972; Guttman, 1989)], the seasonal and sub-seasonal changes of cli-
mate conditions are the primary component and physical descriptor of
Earth's climate system (Kukla and Gavin, 2004; Denton et al., 2005;
Soon, 2009). The information of past seasonal and sub-seasonal changes
in different weather regimes and climatic states is the key for a compre-
hensive understanding of past climatic dynamics and for predicting fu-
ture climate. However, except the coral records in tropical area, most of
proxy-based climate records have temporal resolution insufficient for
quantifying temperature seasonality.
The fast-growing marine biogenic carbonates, such as corals and
mollusks, are sensitive to surrounding environmental changes in most
aquatic settings, and they have the potential to sample the full seasonal
range of past sea surface temperature (SST) and were used extensively
to provide high-resolution seasonal/monthly climate records from tro-
pics to high latitudes for the past 30 years (Williams et al., 1982; Cole
et al., 1993; Gagan et al., 1994; McCulloch et al., 1994; Schöne et al.,
2004a,b; Watanabe et al., 2004; Schöne et al., 2005; Wanamaker et al.,
2008; Aubert et al., 2009; Elliot et al., 2009; Maier and Titschack,
2010; Batenburg et al., 2011; Wanamaker et al., 2011). These seasonal
and sub-seasonal proxy-climate records, together with lower resolution
marine sediment records, have greatly improved our understanding of
the past climate changes in global ocean during the late Holocene. How-
ever, the time spans of the marine biogenic carbonate records, typically
less than 200 years for most of the reported records, are too short to
cover the whole late Holocene, and thus limiting the use of high resolu-
tion marine carbonate records to reconstruct the climatic history of late
Holocene and assess the past climatic dynamics.
In order to overcome the short time-duration limit of the bio-
carbonate records, several recent studies have tried to concatenate
some short bio-carbonate archives together in order to produce longer
composite climate records (Cobb et al., 2003; Yu et al., 2005; Patterson
et al., 2010; Wanamaker et al., 2011; Giry et al., 2013). For example,
Cobb et al. (2003) reconstructed the El Niño-Southern Oscillation
(ENSO) variability covering the past millennium using several coral
δ18O time windows from mid-Pacific (Cobb et al., 2003). Yu et al.
(2005) discussed the mid-to-late Holocene monsoonal climate in
South China Sea based on the Sr/Ca chemical ratio and δ18O of several
123
coral records. Recently, the North Atlantic seawater temperature sea-
sonality variations during the past two millennia, concatenated from
about 26 time windows, were investigated using some bivalve shell ar-
chives (Patterson et al., 2010; Wanamaker et al., 2011). These studies
highlighted the potential of compositing bio-carbonate records in
order to reconstruct high-resolution climate history of the world
ocean and for providing the most important historical and geological
context for judging the current climatic changes. We followed similar
approach in producing our composite SST record for northern SCS cov-
ering the past 2500 years.
South China Sea, SCS, is the largest marginal sea of the western trop-
ical Pacific (Fig. 1). The seasonal climate variation in the SCS is largely
controlled by the East Asian Monsoon; but on inter-annual timescales,
it is probably dominated by tropical Pacific ENSO conditions (Yan
et al., 2010, 2011). Climate records from SCS are thus important for in-
vestigating the interactions between tropical climate and the Northern
Hemisphere high latitudes. However, climate characteristics of the SCS
during the late Holocene are still somewhat unclear because there are
insufficient numbers and types of paleoclimate proxies available.
The coral Sr/Ca-based SST reconstructions in SCS have been system-
atically developed in recent decades and a series of paleo-SST records is
reported to reflect high-time-resolution Holocene climate history (Wei
et al., 2004a,b; Yu et al., 2004, 2005; Wei et al., 2007; Deng et al., 2009).
In addition to the coral records, Tridacna gigas, the biggest marine bi-
valves in the SCS region, was also used to reconstruct the high resolution
climate history in the South China Sea. T. gigas shell is the largest bivalve
species in global ocean; it can grow to over 1 m in length and live up to
100 years (Rosewater, 1965). T. gigas has hard and dense aragonite
shells with daily growth lines and its growth rate is usually higher
than 1 mm per year. The microdrill sampling method can provide
monthly resolution δ18O and Sr/Ca profiles easily (Watanabe and Oba,
1999; Watanabe et al., 2004; Elliot et al., 2009; Yan et al., 2013, 2014a,
b,c). The calibrations between modern T. gigas Sr/Ca and instrumental
SST confirmed that the T. gigas Sr/Ca can be used as a high resolution
(resolving time changes, at least, on a month-to-month basis) SST
proxy comparable to the coral Sr/Ca measure (Yan et al., 2013, 2014a)
and indeed several T. gigas Sr/Ca-based Holocene proxy-SST records
were recently reported (Yan et al., 2014b). These high resolution bicar-
bonate SST records, although each covering only specific time intervals
for certain warm and cold periods, offer a new opportunity in gaining a
broader overall knowledge and detailed insight by coming up with a
high-resolution composite SST series in the South China Sea during
the late Holocene.
In this study, we will combine seven high-resolution, seasonally-
resolved, coral and T. gigas proxy-SST records together with the instru-
mental SST record from 1990 to 2000 interval to provide a 2500-year
long documentation and view of the climatic variations in the northern
SCS during the late Holocene. The changes of temperature seasonality
and the climatic dynamics in the northern SCS during the warm and
cold periods of the late Holocene will also be highlighted.
2. Material and methods
2.1. Regional setting
The SCS is located in the far western tropical Pacific, it is a semi-
enclosed marginal sea with an area of ~3.6 million km2 (Fig. 1). SCS con-
sists of hundreds of islands, coral reefs, atolls, and shoals. The largest is-
land is Hainan Island. Most of the others are much smaller and divided
into four archipelagos: Nansha Islands (Spratly Islands), Xisha Islands
(Paracel Islands), Zhongsha Islands (Macclesfield Islands), and Dongsha
Islands (Pratas Islands).
The coral and T. gigas records used in this study are from the Leizhou
Peninsula (20°15′ N, 109°56′ E), Yongxing Island (16°50′ N, 112°20′ E),
Shidao Island (16°50′, 112°20′) and Dongdao Island (16°40′ N, 112°44′
E). The Leizhou Peninsula is located in the north of Hainan Island and
124 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135

90°E 110°E 130°E

6
40°N

4 Leizhou Peninsula
5
30°N

2 Xisha Islands
20°N

1
(b)
10°N

North
0°

3 Tridacna gigas shell

(a) (c) Porites lutea coral (d)
Fig. 1. The locations of the study sites. (a): map of the East Asia from northwest China to tropical Western Pacific. The locations of the reported temperature records are marked: 1 (Wang
et al., 1999; Shintani et al., 2011), 2 (Wu et al., 2012), 3 (Oppo et al., 2009), 4 (He et al., 2013), 5 (Liu et al., 2009), 6 (Tan et al., 2003), and 7 (Ge et al., 2003). (b): Locations of the Xisha
Islands and the Leizhou Peninsula. (c) and (d): the pictures of Porites lutea coral and Tridacna gigas shell from the northern SCS.

the coral reef in Leizhou Peninsula is the only developed and well-
preserved coral fringing reef on the Mainland China (Yu et al., 2005)
(Fig. 1). The Yongxing Island, Shidao Island and Dongdao Island are lo-
cated in the Xisha Islands, and they are elliptical tropic reef islands
with a land area of 2.13 km2, 0.08 km2 and 1.55 km2, respectively.
The seasonal climate of SCS is dominated by East Asia Monsoon
(EAM). EAM is characterized by southwesterly winds (warm and wet)
in boreal summer and northeasterly winds (cool and dry) in boreal win-
ter. The monthly SST of the Leizhou Peninsula and Xisha Islands are
shown in Fig. 2 and the average annual-mean SST (AD 1982–2011)
are 25.41 and 27.59 °C, respectively. The monthly time series of SST
and SST anomalies in Leizhou Peninsula and Xisha Islands are shown
in Fig. 3. As seen in Figs. 2 and 3, the summer SSTs in Leizhou Peninsula
and Xisha Islands are similar, but the winter SSTs in Leizhou Peninsula
are substantially colder than those in Xisha Islands, leading to a larger
seasonality of SST in Leizhou Peninsula than in Xisha Islands. The
SST data are obtained from the NOAA-NCEP-EMC-CMB GLOBAL
30
28
SST (°C)
Reyn_SmithOIv2 database (a global weekly SST data starting from
1982 with a spatial resolution of 1° × 1°); the grid cells of the NOAA
SST data used were selected to include Leizhou Peninsula and Xisha
Islands, respectively.
2.2. Data sources
The Sr/Ca ratios of coral skeletons have long been known to robustly
yield a high-resolution SST proxy and have been extensively utilized in
high-resolution paleo-temperature reconstructions in the tropical
oceans (Beck et al., 1992; McCulloch et al., 1994, 1996; Alibert and
McCulloch, 1997). The coral Sr/Ca-based SST reconstructions in the
SCS have also been developed and reported in recent decades. For ex-
ample, three high-resolution late Holocene paleo-SST records, around
~535 BC, ~495 AD and ~1464 AD, were reported in northern SCS from
Leizhou Peninsula, Hainan Island and Xisha Islands, respectively (Wei
et al., 2004a,b). Yu et al. (2005) reported five mid-late Holocene month-
ly resolution Porites corals Sr/Ca based SST records from Leizhou Penin-
sula (~4789 BC, ~3906 BC, ~3011 BC, ~541 BC, ~487 AD), each covering
a growth history of 9–13 years (Yu et al., 2005). The coral Sr/Ca-based
SST records, covering mid-Holocene time interval, were also reported
from Leizhou Peninsula, Hainan Island and Xisha Island (Wei et al.,
2004a; Yu et al., 2004; Wei et al., 2007; Deng et al., 2009).

26
In the meantime, the Sr/Ca ratios of marine bivalve, such as Tridacna
species, have also been investigated in the northern SCS. The study on
24
Sr/Ca of modern T. gigas specimen from SCS showed that the high-
resolution Sr/Ca ratio profiles determined by the inductively coupled
22 plasma-optical emission spectrometer (ICP-OES) had clear annual cy-
cles, correlating with the local SST variations, and were highly reproduc-
20 ible across the shells regardless of the physiological effects (Yan et al.,
0 2 4 6 8 10 12 2013, 2014a). Thus the Sr/Ca ratio of T. gigas from South China Sea, pri-
Month marily controlled by water temperature, is a useful paleo-temperature
proxy that promises to tell us past summer and winter temperatures
Fig. 2. Monthly mean sea surface temperatures (SST) in Xisha Islands (circles connected by (Yan et al., 2013, 2014a). Based on these new insights, the high resolu-
red line) and Leizhou Peninsula (triangles connected by purple line) constructed from the tion Sr/Ca data of two late Holocene T. gigas specimens were analyzed to
averages over the 1982 to 2008 interval. The source of the SST data is the NOAA NCDC
ERSST ver.2 data, a global monthly SST data from January 1854 with a spatial resolution
reconstruct paleo-SST around ~50 AD and ~990 AD (Yan et al., 2014b).
of 2° × 2°. (For interpretation of the references to color in this figure legend, the reader In this study, a total of seven coral and T. gigas Sr/Ca-based SST time
is referred to the web version of this article.) windows as well as modern instrumental SST in northern SCS were
 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135 125

32
a
r=0.91, p<0.001
30

28

SST (°C)
26

24

22

20

18
1982 1987 1992 1997 2002 2007 2012

2.5
2 b
1.5 r=0.54, p<0.001
SST anomaly (°C)

1
0.5
0
-0.5
-1
-1.5
-2
-2.5
1982 1987 1992 1997 2002 2007 2012
Year

Fig. 3. SST and SST anomaly series from 1982 to 2012 in Xisha Islands (red curve) and Leizhou Peninsula (purple curve). A significant correlation between SST anomalies in Xisha Islands
and Leizhou Peninsula is observed (r = 0.54, p b 0.001). The source of the SST data is the NOAA NCDC ERSST ver.2 data, a global monthly SST data from January 1854 with a spatial res-
olution of 2° × 2°. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

assembled here to produce a proxy-SST record over the last 2500 years.
Coral and T. gigas records meeting the following conditions were includ-
ed in this assembly work. First, the coral and T. gigas records are from
the northern SCS. Second, the time windows of coral and T. gigas records
must cover the late Holocene of the last 3000 years or so. Third, the time
resolutions of the coral and T. gigas records are higher than one month.
Based on these three stringent conditions, one coral record from Xisha
Islands (Wei et al., 2004a), four coral records from the Leizhou Peninsu-
la (Wei et al., 2004a,b; Yu et al., 2005), and two T. gigas Sr/Ca based SST
time series from Xisha Islands (Yan et al., 2014b) were selected (see all
the individual records in Fig. 4). The details and statistics of the coral and
T. gigas proxy-SST for the seven time windows of the late Holocene and

8
BC 541±24 BC 530±24 AD 487±22 AD 489±22
4
SST (°C)

-4

-8
Leizhou, Coral Leizhou, Coral Leizhou, Coral Leizhou, Coral
-12

6
AD 50±40 AD 990±40
AD 1455±51
4
2
SST (°C)

0
-2
-4 Xisha, T. gigas Xisha, T. gigas Xisha, Coral
-6

Fig. 4. All seven coral and T. gigas Sr/Ca-based SST time series used in this study. One coral record from Xisha Islands (Wei et al., 2004a), four coral records from the Leizhou Peninsula (Wei
et al., 2004a,b; Yu et al., 2005), and two T. gigas records from Xisha Islands (Yan et al., 2014b) were selected for our study.
126 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135

Table 1
Detailed information of coral and T. gigas Sr/Ca-based SST time windows in northern SCS during the late Holocene. The calculated SST anomalies and relative change rates in SST were also
presented. Please see Section 2.3 for the methodology for the assembly of our composite SST record combining the seven coral and T. gigas records with the instrumental SST records.

Locations Materials References Dating Length SST anomalies SST relative change rates
(BC/AD) (years)
Annual SST (°C) Winter SST Summer SST SST Annual Winter Summer SST
(°C) (°C) seasonality SST SST SST seasonality
(°C) (in %) (in %) (in %) (in %)

Northern SCS Instrumental Xisha AD 1990–2000 11 0 ± 0.37 0 ± 0.56 0 ± 0.32 0 ± 0.50 0 0 0 0

data Leizhou 0 ± 0.35 0 ± 0.89 0 ± 0.38 0 ± 1.01
Xisha Island Coral Wei et al. AD 1405 ± 51 3 −1.08 ± 0.36 −0.8 ± 1.00 −1.45 ± 0.14 −0.65 ± 0.89 −3.9 −3.2 −4.9 −13.3
(2004a)
Xisha Island T. gigas Yan et al. AD 990 ± 40 11 0.89 ± 0.65 1.39 ± 0.64 0.3 ± 0.97 −1.09 ± 1.02 3.2 5.6 1.0 −22.2
(2014b)
Leizhou Coral Wei et al. AD 495 ± 22 11 −2 ± 1.58 −2.9 ± 2.46 −1 ± 1.03 1.9 ± 1.01 −7.8 −14.1 −3.4 21.1
Peninsula (2004b)
Leizhou Coral Yu et al. AD 487 ± 22 13 −2.2 ± 1.59 −3.8 ± 2.48 −0.7 ± 1.01 3.1 ± 1.18 −8.6 −18.4 −2.4 34.5
Peninsula (2005)
Xisha Island T. gigas Yan et al. AD 50 ± 40 15 1.55 ± 0.64 1.88 ± 0.73 1.13 ± 0.76 −0.75 ± 0.79 5.6 7.6 3.8 −15.3
(2014b)
Leizhou Coral Wei et al. BC 535 ± 24 9 0.1 ± 1.53 0.4 ± 2.48 −0.2 ± 1.32 −0.6 ± 1.21 0.4 1.9 −0.7 −6.7
Peninsula (2004b)
Leizhou Coral Yu et al. BC 541 ± 24 11 −0.4 ± 1.54 −0.8 ± 2.49 −0.1 ± 1.31 0.7 ± 1.26 −1.6 −3.9 −0.3 7.8
Peninsula (2005)

the instrumental SST data are summarized in Table 1. Our method of SST anomaly is 0.89 °C (28.54–27.65 = 0.89), and the relative change
compositing explicitly assumed that each of the time window, though rate is thus calculated to be 3.2% (0.89/27.65 = 0.032). All relative
relatively short in the full extent of their actual time coverages, can be change rates are also given in Table 1.
representative of the major climatic warm, cold and neutral periods
studied here. We believe that this assumption is reasonable since the
3. Results and discussion
dating accuracy of the coral and T. gigas samples is relatively accurate
so there can be no doubt that our coral and bivalve samples actually
3.1. A composite proxy-SST record in the northern SCS during the late
sampled and recorded the SST conditions during the historical time
Holocene
windows when they were alive.
Both SST anomalies and relative change rates of the composite SST
2.3. The method of compositing
record, including winter, summer and annual-mean SSTs, in northern
SCS during the late Holocene are plotted in Fig. 5. Both plots cover a
In this study, the time resolution of all coral and T. gigas Sr/Ca series
total of 8 time windows (see summary in Table 1 and Fig. 5), including
was homogenously resampled to one month using linear interpolation
5 coral-based time windows around BC 541, BC 535, AD 487, AD 495
from original raw data records. For each series, the proxy-SST was calcu-
and AD 1464, two T. gigas-based time windows around AD 50 and AD
lated from Sr/Ca ratio using the calibration equations between instru-
990, and the instrumental SST time window covering AD 1990–2000.
mental SST and modern coral/T. gigas Sr/Ca ratios (Wei et al., 2004a,b;
The annual-mean SST anomalies for the 8 time windows are ranging
Yu et al., 2005; Yan et al., 2013). For each proxy-SST time series, the win-
from − 2.2 °C (− 8.6%) to 1.55 °C (5.6%) during the last 3000 years
ter SST value was calculated using the average of annual SST minima
(Fig. 5). The SSTs of warm period time-windows around AD 990 ± 40,
and the summer SST was calculated using the average of annual SST
AD 50 ± 40 and BC 535 ± 24 are 0.89 °C (3.2%), 1.55 °C (5.6%) and 0.1
maxima. The SST seasonality of each time windows was defined as the
°C (0.4%), higher than that during modern instrumental period of AD
difference between winter and summer SST values. Then, the average
1990–2000 (Fig. 5). The SSTs of cold period time-windows around AD
of the winter SST and summer SST values was re-calculated and
1464 ± 51, AD 495 ± 22, AD 487 ± 22 and BC 541 ± 24 are 1.08 °C
employed as the annual-mean SST of each time windows. We note
(−3.9%), 2 °C (−7.8%) and 2.2 °C (−8.6%) and 0.4 °C (−1.6%), lower
that the use of such arithmetic means will minimize the impact of ana-
than that of the modern instrumental SST decade (Fig. 5). It is worthwhile
lytical errors, growth rate and other shorter-term variability on the final
to note that the warmer SSTs in northern SCS around AD 990 ± 40 (3.2%)
longer term SST estimates (Yu et al., 2005; Yan et al., 2013).
and AD 50 ± 40 (5.6%) occurred within the notable warm episodes of
As noted, the modern SSTs in Xisha Islands and Leizhou Peninsula
Medieval Warm Period (MWP, ≈AD 800–13001) and Roman Warm Pe-
are different (27.59 vs. 25.41 °C), thus making the combination difficult
riod (RWP, ≈BC 400–AD 200), respectively. Conversely, the cooler SSTs
if using the reconstructed SST data directly. Therefore, we have chosen
around AD 1464 ± 51, AD 495 ± 22 and AD 487 ± 22 are consistent
to use SST anomalies to combine the seven proxy-SST records and in-
with the European cold periods of Little Ice Age (LIA, ≈AD 1400–1850)
strumental SST record. We first calculated the instrumental winter,
and Dark Age Cold Period (DACP, ≈AD 400–800).
summer and mean SST in Xisha Islands and Leizhou Peninsula during
the 1990–2000 using the same method of the paleo-SST. The anomalies
were calculated by subtracting the winter, summer and annual-mean 3.2. Intercomparison with other SST proxies from tropical Western Pacific
SSTs of each paleo time windows from the individual values of instru-
mental winter, summer and annual-mean SSTs (Table 1). Although the high resolution temperature reconstructions in north-
Because the variation range of the modern climate in Leizhou Penin- ern SCS are scarce, the nearby low resolution marine sediment SST
sula is much larger than that in Xisha Islands; the relative change rate,
1
therefore, was used to further normalize our data. The relative change Please see pp. 235–239 of Soon et al. (2003) for a detailed discussion on the appropri-
rate is defined as the percentage of the paleo SST anomalies to the mod- ate use of the terminology “Medieval Warm Period”. We also propose here that the timing
of 1300 AD for the transition between MWP and LIA may be not entirely arbitrary nor ac-
ern instrumental SSTs. For example, the reconstructed annual-mean SST cidental but instead has a physical root and link to the timing of 1246 AD being the last
in Xisha Islands around AD 990 is 28.54 °C, the instrumental annual- time in which the perihelion of the Sun–earth orbit coincided with Northern Hemisphere
mean SST in Xisha Islands during the 1990–2000 is 27.65 °C, the mean winter solstice (see e.g., Kukla and Gavin, 2004).
 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135 127

in Northern SCS (°C) a BC 541±24 AD 50±40 AD 487±22 AD 990±40 so by different authors: 2700–1600 BP in southwest Greenland
2 BC 530±24 AD 489±22 AD 1455±51
(Seidenkrantz et al., 2007), 2400–1600 BP in the North Sea (Hass,
SST anomaly

1996) and 2600–1600 BP in southern Spain (Martín-Puertas et al.,

0 2009). It is still unclear if the warm climate during the Roman times oc-
curred globally but it is clear from the literatures, even within dating im-
-2 Annual precision, the warm period probably did not occurred or initiated
Summer synchronously all across the globe. From the perspective of multi-
Winter
-4 century scale variation of glacier mass fluctuations, Joerin et al.
-1000 -500 0 500 1000 1500 2000 (2006), for example, through the accurate dating of wood and peat frag-
BC/AD ments, suggested a warm and major glacier recession in the Swiss Alps
centered around the 300-year long period from 2150 to 1850 BP. Al-
10
b though many paleoclimate reconstructions from the Eurasia–Atlantic
in Northern SCS (%)

5 regions revealed an obvious warm climate around AD 0, there were

SST anomaly

0 also some temperature records from the Northern and Southern Hemi-
-5
-10
Annual
-15 Summer
Winter
-20
-1000 -500 0 500 1000 1500 2000
BC/AD
Fig. 5. The composite proxy-SST records. Winter (triangles connected by pink curve), sum-
mer (triangles connected by blue curve) and annual-mean (diamonds connected by red
curve) SSTs, in northern SCS during the late Holocene are indicated. Both the SST anoma-
lies (a) and SST relative change rates (b) are shown. The composite SST records consist of 8
time windows (see summary in Table S1), including 5 coral time-windows around BC 541,
BC 535, AD 487, AD 495 and AD 1464, two T. gigas time-windows around AD 50 and AD
990, and a instrumental SST time-window covering AD 1990–2000. (For interpretation
of the references to color in this figure legend, the reader is referred to the web version
of this article.)
records (Wang et al., 1999; Shintani et al., 2011), which yield similar
variation pattern with our composite SST record, provide some support
to our composite 8 time windows SST record (Fig. 6). First to note is the
larger variability of winter SST relative to summer SST, which is clearly
shown in our composite high resolution SST record (Fig. 5), was also ob-
served in the marine sediment SST record (Fig. 6) (Wang et al., 1999;
Shintani et al., 2011). The SST reconstruction of Indonesia derived
from the Mg/Ca of planktonic foraminifera (Oppo et al., 2009; Steinke
et al., 2014) and SST record of southern Okinawa Trough derived from
TEX86 (Wu et al., 2012) also showed similar time variation pattern
with our composite SST record during the late Holocene (Fig. 6). These
marine sediment SST records that are more continuous in time coverage
largely supported the reliability of our composite SST record in northern
SCS that was constructed from seven proxy-SST and the instrumental
SST time windows. It may be important to note that the relative
agreement of our Sr/Ca-based SST proxies with the SST deduced from
different proxies and methodologies in Fig. 6 serve as a very good inde-
pendent check of our Northern SCS seasonal SST records.
3.3. Distribution of RWP and MWP in the Asia-Pacific region with a compar-
ison to other regions
The composite proxy-SST record gives us an important overview and
scientific context for the climate changes in northern SCS during the late
Holocene (Fig. 5). As seen in Fig. 5, the reconstructed SST record re-
vealed six climatically recognizable periods: two warm periods around
AD 990 and AD 50 corresponding to the MWP and RWP, two cold pe-
riods around AD 1464 ± 51, AD 495 ± 22 and AD 487 ± 22 correspond-
ing to the LIA and DACP, and two “normal” periods around BC 541 ± 24,
BC 535 ± 24 and AD 1990–2000.
The RWP has been proposed as a period of unusually warm weather
in Europe and the North Atlantic (Lamb, 1977). The time span for RWP
is generally considered to cover broadly between 2500 and 1600 BP, al-
though the exact time period varies in different regions and stated to be
spheres (Cook et al., 2000; Hantemirov and Shiyatov, 2002) as well as
Antarctica (Masson et al., 2000), that showed no significant warm con-
dition during the RWP 2000 years ago. Similarly, there were some ef-
forts to synthesize various records from different regions to produce a
wider areal, if not hemispheric, averages, but the results were also con-
troversial. For example, one early multi-proxy compilation of Northern
Hemisphere did not find the obvious warm period during the Roman
times (Moberg et al., 2005), but a more recent synthesis of extratropical
Northern Hemisphere temperature showed the obvious warm climate
around AD 0 (Ljungqvist, 2010) (see Fig. 7).
The annual-mean proxy-SST for the time window during the RWP
(around AD 50 ± 40) is the warmest one among the eight time win-
dows of our northern SCS composite record, with 1.55 °C warmer than
that of recent decades (Fig. 4). The warm climate in northern SCS during
the RWP is consistent with the nearby SST reconstruction surrounding
marine-continental region of Indonesia (Fig. 6) (Oppo et al., 2009) and
in China (Fig. 7a) (Yang et al., 2002). However, the warm RWP does
not seem to be uniformly warm for the whole East Asian Monsoon af-
fected domain. The historical documentary and stalagmite reconstruc-
tions from the mid-east (Fig. 7b) and north China (Fig. 7c) suggested a
relatively warm period (Ge et al., 2003; Tan et al., 2003), but the tree
ring (Fig. 7d) and lake sediment records (Fig. 7e) from the northwest
China indicated either neutral or even a cold period during the Roman
times (Liu et al., 2009; He et al., 2013). The regional difference, the er-
rors of age models and nonlinear proxy transfer functions or relative in-
sensitivity of certain proxies could all provide some explanations for the
discrepancy. Overall, we find that although the presence of warm cli-
mate in northwest China during the Roman times remains to be
substantiated, the relatively prominent and geographically wide distri-
bution of the Roman warm period in the Eastern margin of China and
the western tropical Pacific region is well established.
In contrast to RWP, there are more comprehensive proxy records
that support the MWP, a warm period during AD 800–1300, and the ex-
istence of MWP is demonstrated by climate data from not only in north-
ern SCS and northern Europe but also in many parts of the world
(Hughes and Diaz, 1994; Soon et al., 2003; Mann et al., 2008, 2009;
Esper and Frank, 2009; Ljungqvist, 2010; Graham et al., 2011). The aver-
aged SST of the time window during the MWP (around AD 990 ± 40)
was higher than that during the DACP and the LIA (Fig. 5 and Table 1),
confirming the existence of a warm period during the MWP in the
northern SCS. The warm medieval times was also found in the
Indonesian record, as well as some records from China (Figs. 6 and 7).
Although there are also some difference in the timing and the tempera-
ture change amplitude in these diverse types of proxy records, but a
warm climate during the MWP seems to be certain in Asia-Pacific re-
gion, covering broad geographical areas from Northwest China to the
Western Pacific.
Our composite proxy-SST record also reveals two cold periods, one is
between RWP and MWP corresponding to the European DACP well
noted in historical documents and the other is between MWP and cur-
rent warm period corresponding to the relatively better-studied LIA in-
terval (Matthes, 1939, 1940; Grove, 1988). The cold climate during the
128 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135

a
28

Annual mean SST (°C)

17940-Northern SCS
27

26

25 Annual SST, SD= 0.64 °C

b

Summer/Winter mean SST (°C)

30

17940-Northern SCS
Summer SST, SD= 0.28 °C
27

24

Winter SST, SD= 1.19 °C 21

c
29
West Pacific Warm Pool
SST (°C)

28

d
28

Southern Okinawa Trough

27

27

SST (°C)
26

25
RWP MWP
24
-1000 -500 0 500 1000 1500 2000
BC/AD

Fig. 6. The comparison between composite SST series in northern SCS (triangles connected by red line) and other marine sediment-based temperature records from northern SCS (Wang
et al., 1999; Shintani et al., 2011), Indonesia (Oppo et al., 2009) and southern Okinawa Trough (Wu et al., 2012). (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)

Dark Age was pointed out specifically by Hubert Lamb (Lamb, 1977).
Later studies further confirmed a cold and wet period in Europe during
the DACP. The cold climate during the DACP was also clearly recorded in
various proxies from the Western Pacific to the Northwestern China
(Figs. 6 and 7), such as the marine sediment record (Oppo et al.,
2009), the ancient Chinese literature record (Ge et al., 2003), the cave
stalagmite record (Tan et al., 2003) and the lake sediment record (He
et al., 2013), confirming that the cold climate during the DACP was in-
deed widely distributed all over the East Asia–Western Pacific region.
Despite the relatively more stringent requirements for persistency of
meteorological and climatic forcings and finite time for the growth of
glacier ice mass, the clear imprints of the DACP can also be found in
the glacial record of Sheridan Glacier, around the Copper River Delta,
coastal South-Central Alaska which has its peaked phase of advance
centered at around 530 to 640 AD (Barclay et al., 2013). In addition,
the recent syntheses work of Northern Hemisphere temperature varia-
tions also demonstrated an obvious cold climate during the DACP
(Fig. 7) (Moberg et al., 2005; Ljungqvist, 2010). The anomalous weather
regimes and climatic conditions during the LIA have been heavily stud-
ied in recent decades and the results suggested a cold climate with a
substantial accumulation of land ice covering most high-elevation
areas of the Earth. The relatively low temperature anomaly of 1.08 °C
during the LIA in our composite proxy-SST record was clearly consistent
with the other reconstructions from the East Asia–Western Pacific re-
gion (Figs. 6 and 7).
Overall, the variation pattern of our northern SCS SST record
was similar to that of other temperature reconstructions from East
Asia–Western Pacific regions and two syntheses of the Northern
Hemisphere-wide temperature records during the late Holocene with
some divergences in details especially for the RWP (Fig. 7). These results
demonstrated convincingly that the climate oscillations in northern SCS
were in step with other regional and perhaps even the global climate
changes during the late Holocene.
3.4. Variation of the SST seasonality in the SCS during the late Holocene and
its relationship with the mean climate state
Temperature seasonality, the difference between summer and win-
ter temperatures, is an important aspect of Earth's climate and has sig-
nificant influences on local, regional and global terrestrial and marine
 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135 129

a 1.5

Temperature variation
1.0

in China (°C)
0.5
b
0.0

Winter half-year temperature

1.0

in mid-east China (°C)

-0.5
0.5
-1.0

0.0

c -0.5
Warm season temperature in Beijing

24 -1.0
Northern China (°C)

23

22

21 d
3.0

in mid-east Tebet Platea

North-west China (°C)
Temperature variaion
20 2.0
1.0
0.0
-1.0
e
18 -2.0
North-West China (°C)
Temperature in Gahai

16 -3.0

14

12 f
0.5
10
Northern Hemisphere
Temperature (°C)

g -0.5
extratropical Temperature (°C)

0.2 -1
Northern Hemisphere

0
-1.5
-0.2

-0.4

-0.6
RWP MWP
-0.8
-1000 -500 0 500 1000 1500 2000
BC/AD

Fig. 7. The comparison between composite SST series in northern SCS (triangles connected by red curve) and other temperature records from mainland China (a) (Yang et al., 2002), mid-
east China (b) (Ge et al., 2003), northern China (c) (Tan et al., 2003), north-west China (d, e) (Liu et al., 2009; He et al., 2013), Northern Hemisphere (f) (Moberg et al., 2005), and
extratropical region of Northern Hemisphere (g) (Ljungqvist, 2010). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this
article.)

environments and ecosystems. However, most of the current proxy- record gives us an exciting opportunity to examine the seasonality var-
based climate records are not sufficient to quantify the temperature sea- iation in northern SCS during the late Holocene that are known to cover
sonality due to the limits of the temporal resolution inherent in those several warm and cold weather regimes and climatic periods. We em-
individual paleoclimate proxies. In this study, our high-resolution SST phasize that the current climate understanding and theory are still
130 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135

simply not mature enough (Essex, 2011, 2013) to answer such a direct
question, so we will have to rely on empirical knowledge systematically
built from such a paleoclimatic study.
The changes in SST seasonality derived from the differences between
winter and summer SST anomalies were shown in Fig. 8. Both the incon-
stant nature of and the large variation in the amplitude of seasonality for
locations like Northern SCS are the major findings in this study. The SST
seasonality anomalies of eight time windows during the late Holocene
varied substantially from −1.09 to 3.1 °C and the relative change rates
of SST seasonality varied from −22.2% to 34.5% (Table 1 and Fig. 8).
We compare the annual-mean SST anomalies with the SST seasonal-
ity anomalies and find a strong and statistically significant negative cor-
relation (r2 = 0.69, p b 0.001, Fig. 8). The seasonality decreased when
the mean SSTs were warmer, such as the time windows centered at
around RWP (SST seasonality anomaly was − 0.75 °C) and MWP (SST
seasonality anomaly was −1.09 °C). The result of decreased seasonality
under warmer climate is consistent with the previous composite bivalve
shell records (i.e., also concatenated from several time windows) from
the Gulf of Maine, North Atlantic, which revealed increased seasonality
with climatic cooling (Wanamaker et al., 2011). This empirical fact
concerning the inverse relation between amplitude of seasonality and
mean climatic condition is also supported by some marine sediment re-
cords. For example, the nearby marine sediment core SCS90-36
(17°59.70′ N, 111°29.64′ E) record presents an obvious decrease of sea-
sonality in northern SCS during the Holocene than during the last glacial
period (Huang et al., 1997). The new and independent results, both in
terms of area of study and methodology, from western Arabian Sea
also found, from analyses of δ 18O in individual shells of planktic forami-
nifera samples from their Ocean Drilling Program Site 723A, that the
seasonality amplitude of SST around western Arabian Sea for the rela-
tively colder time during LIA (ca. 0.5 ka for their samples) is about
four times larger than the seasonal SST amplitude during the warmer
period around RWP (ca. 2.07 ka) (Naidu et al., 2014). Our result can fur-
ther directly verify such an observation for northern SCS where the larg-
est amplitude variation for the seasonality occurs during the cold period
of DACP. We have also checked the relationship between the SST sea-
sonality and the SST anomalies in winter and summer, respectively.
The results suggested that the SST seasonality anomalies were highly
correlated with the winter SST anomalies (r2 = 0.86, p b 0.001, Fig. 8),
but not with the summer SST anomalies (r2 = 0.19, p = 0.29, Fig. 8),
probably indicating that the changes of SST seasonality in northern
SCS mainly depend on the variations of winter climate which is sugges-
tive of the more prevalent effects on seasonality by the land–sea con-
trast of the monsoon-controlled climate regions like East Asia (Ding
et al., 2014).
Similar statistical correlations and physical tendencies can also be
found in instrumental SST data for modern period from 1982 to 2012.
Both the observations from Xisha Island and Leizhou Peninsula revealed
stronger correlations between SST seasonality and winter SST, and weak
ones between SST seasonality and summer SST (Fig. 9). The coefficients
of the correlations between SST seasonality and annual-mean, summer
and winter SST in Xisha Islands were − 0.5 (r2 = 0.25, p b 0.01), 0.1
(r2 = 0.01, p = 0.58) and − 0.82 (r2 = 0.68, p b 0.001), respectively
(Fig. 9 top panels). The coefficients of the correlations between SST sea-
sonality and annual-mean, summer and winter SST in Leizhou Peninsu-
la were − 0.58 (r2 = 0.34, p b 0.001), 0.45 (r2 = 0.21, p = 0.01) and
− 0.92 (r2 = 0.85, p b 0.001), respectively (Fig. 9 bottom panels).
Such an observed reality from instrumental data clearly provides a
good physical grounding for our interpretation of the paleo-SST results.
3.5. Different amplitudes of summer and winter SST variations during the
late Holocene
Meanwhile, the variation of the winter SST anomaly amplitude in
our composite record was much larger than that of summer SST anom-
aly during the late Holocene (Fig. 5). The winter SST anomalies for the
eight time-windows changed from −3.8 °C to 1.88 °C and the summer
values varied from −1.45 °C to 1.13 °C. The regression between annual-
mean SST anomalies and winter SST anomalies of eight time windows
yielded the equation (Fig. 10): winter-SST (°C) = 1.45 × annual-

4 4
b
SST seasonality anomaly

3 a 3
in Northern SCS (°C)
in Northern SCS (°C)

y = -0.93x - 0.03
2 2 2
r = 0.69
SST anomaly

1 1
0
0
-1
-1
-2
Summer -2
-3
Winter -3
-4
Seasonality -4
-5
-1000 -500 0 500 1000 1500 2000 -3 -2 -1 0 1 2
BC/AD Annual SST (°C)

4 4
c d
SST seasonality anomaly

SST seasonality anomaly

3 y = -0.80x + 0.12 3 y = -0.70x - 0.08

in Northern SCS (°C)

in Northern SCS (°C)

2 2
2 r = 0.19 2 r = 0.86
1 1
0 0
-1 -1
-2 -2
-3 -3
-4 -4
-2 -1.5 -1 -0.5 0 0.5 1 1.5 -4 -3 -2 -1 0 1 2
Summer SST (°C) Winter SST (°C)

Fig. 8. The SST seasonality in northern SCS during the late Holocene (a). The linear regressions of SST seasonality versus annual-mean SST anomaly (b), SST seasonality versus summer SST
anomaly (c) and SST seasonality versus winter SST anomaly (d) are also shown.
 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135 131

7 7 7
y = 0.16x + 0.15

SST seasonality (oC)

y = -0.67x + 23.56 y = -0.73x + 23.07
R2 = 0.25 R2 = 0.01 R2 = 0.68
6 6 6

5 5 5

4 4 4

X i s ha a X i s ha b Xisha c
3 3 3
27 27.5 28 28.5 29 28.5 29 29.5 30 30.5 23.5 24.5 25.5 26.5
Annual SST (oC) Summer SST (oC) Winter SST (oC)

12 12 12
y = -1.65x + 51.32 y = 1.18x - 25.70
SST seasonality (oC)

y = -1.03x + 30.3
11 R2 = 0.34 11 R2 = 0.21 11 R2 = 0.85
10 10 10

9 9 9

8 8 8
Leizhou a Leizhou b Leizhou c
7 7 7
24.5 25 25.5 26 26.5 28.5 29 29.5 30 30.5 18.5 19.5 20.5 21.5 22.5
Annual SST (oC) Summer SST (oC) Winter SST (oC)

Fig. 9. The linear regressions of instrumental SST seasonality versus annual-mean SST (a), SST seasonality versus summer SST (b) and SST seasonality versus winter SST (c) in Xisha Islands
(red triangles and line) and Leizhou Peninsula (purple triangles and line) from 1982 to 2012. The source of the SST data is the NOAA NCDC ERSST ver.2 data, a global monthly SST data from
January 1854 with a spatial resolution of 2° × 2°. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

mean-SST (°C) − 0.026 (r2 = 0.96, p b 0.001, n = 8). The regression from Hainan Island and the winter monsoon index/winter SST, also im-
between mean SSTs and summer SSTs yielded the equation: summer- plying a larger variability of winter SST.
SST (°C) = 0.52 × annual-mean-SST (°C) − 0.05 (r2 = 0.74, p b 0.01, The large difference in the variability of the winter SST and summer
n = 8). This empirical result suggests that an increase of annual-mean SST in northern SCS was also observed in the instrumental SST record
SST of 1 °C corresponds to a winter SST increase of 1.45 °C and a summer (Fig. 11). The direct instrumental SST data for the most recent
SST increase of only 0.52 °C (Fig. 10). 50 years in northern SCS suggest that an annual-mean SST increase of
This large difference in the variability of the winter SST and summer 1 °C corresponds to a winter SST increase of 1.64 °C and a summer SST
SST has also been detected in two previously published coral δ18O time increase of 0.62 °C (Fig. 11), similar to the calculated results from the
series from northern SCS (Peng et al., 2003; Sun et al., 2005). For exam- paleo-SST records during the late Holocene even in a quantitative
ple, Sun et al. (2005) found that the interannual variability of winter sense. This fact adds confidence to our paleoclimate study of seasonality
coral δ18O in Hainan Island of northern SCS was dominated by the inter- contrasting the sharp differences between climate warm and cold pe-
annual variation of the winter SST, but the interannual variability of riods of the past 2500 years.
summer coral δ18O was primary controlled by the summer rainfall Thus, we conclude that the larger amplitude in SST variation during
with little impact from the summer SST variation. This independent re- winters than during summers in northern SCS during the late Holocene
sult and interpretation confirm the small interannual variability of sum- and instrumental era could explain the negative correlation between
mer SST relative to winter SST variability. Peng et al. (2003) reported a the SST seasonality and the mean climate state in Figs. 8 and 9. When
significant correlation between the interannual variation of coral δ18O northern SCS climate turned warm, the summer SST increased less
and the winter SST increased disproportionately more, resulting in a re-
2 duction of the amplitude of seasonality. When northern SCS climate was
Winter SST (pink) and Summer SST

undergoing cooling tendencies, the summer SST decreased less and

y = 0.5216x - 0.0543 winter SST decreased more, resulting in a larger seasonality. The stron-
1
2
R = 0.7393 ger correlation between the winter SST and seasonality than that be-
tween summer SST and seasonality could be similarly explained by
0
the relative dominance of the winter season for northern SCS monsoon-
(blue) (°C)

al climate region.
-1
3.6. Mechanism for the different variation amplitudes of the summer and
-2 y = 1.4539x - 0.0262 winter SST during the late Holocene
2
R = 0.9545
-3 Seasonal and sub-seasonal changes are important components of
Earth's multi-scale and multi-component climate system. The mecha-
-4 nism for the different variation amplitudes in winter and summer SST
-2.5 -1.5 -0.5 0.5 1.5
is, therefore, essential both for clarifying and understanding the climatic
SST (mean, °C)
dynamics in the northern SCS. The SST maps in the tropical Western Pa-
cific (including SCS) for the winter (i.e., January, 2000) and summer
Fig. 10. The linear regressions of winter SST versus annual-mean SST anomalies (pink tri-
angles and line) and summer SST versus annual-mean SST anomalies (blue circles and
(i.e., July, 2000), respectively, are shown in Fig. 12. The SCS was covered
line) in northern SCS during the late Holocene. (For interpretation of the references to by the West Pacific Warm-Fresh Pool (WPWFP; consider iso-
color in this figure legend, the reader is referred to the web version of this article.) temperature zones with SST N 28.5 °C (Cravatte et al., 2009)) during
132 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135

29 31 25
a y = 0.0192x - 11.324
2
b y = 0.0109x + 7.6391 c y = 0.0315x - 39.344
2
2

Temperature (°C)
R = 0.4645 R = 0.1323 SD=0.41 R = 0.3015
28 30 24 SD=0.79
SD=0.39

27 29 23

26 28 22
Annual mean Summer Winter
25 27 21
1955 1975 1995 1955 1975 1995 1955 1975 1995
Year Year Year
30.4 26
Summer temperature (°C)

Winter temperature (°C)

y = 0.6219x + 12.644
2
d 25
y = 1.6415x - 20.797 e
30 R = 0.341 2
R = 0.6499
24
29.6
23
29.2
22
28.8 21
28.4 20
25.5 26 26.5 27 27.5 28 25.5 26 26.5 27 27.5 28
Annual mean temperature (°C) Annual mean temperature (°C)

Fig. 11. The variation trends of annual-mean (a), summer (b) and winter (c) SST in Xisha Islands from 1958 to 2005. The linear regressions of summer SST versus annual-mean SST (d) and
winter SST versus annual-mean SST (e) from 1958 to 2005 are also shown.

summer when the WPWFP migrated north (right panel in Fig. 12). The
summer SST in northern SCS was, therefore, consistent with the SST of
WPWFP. As a result, the changes of summer SST in northern SCS were
closely controlled by and connected to the climate conditions of the
WPWFP. However, the situation was quite different during winter
when the WPWFP migrated south (left panel in Fig. 12). The SCS now
seemed to be free from the predominance of the WPWFP and an obvi-
ous “cold tongue” lied on the northern SCS (Fig. 12). The “cold tongue”
was formed by the southward migration of the cold water along the
northwest coast of the northern SCS driven by the strong East Asian
Winter Monsoonal circulation (EAWM). The “cold tongue” makes the
SST in northern SCS about 3–5 °C lower than the SST in the eastern
part of the Philippines under the same latitude.
The summer and winter SSTs in northern SCS are controlled by
oceanographic conditions of the WPWFP and the strength of the
EAWM atmospheric circulation, respectively. In modern meteorological
observation, the impacts of EAWM on the northern SCS are manifested
through the pulse type of cold waves. The winter temperature around
the region could drop significantly by more than five degrees rather
abruptly when a cold wave sets in from the Siberian High, and the sea
surface temperature also could recover quickly after the cold waves dis-
sipated (Ding et al., 2014). In contrast, the WPWFP oceanographic

30N
January 2000 July 2000
29.5
28.5

20N 27.5
26.5

25.5
24.5
10N
23.5

22.5
21.5
0 20.5

19.5
18.5

10S
85E 95E 105E 115E 125E 135E 85E 95E 105E 115E 125E 135E

Fig. 12. The spatial distributions of SST in SCS in January 2000 (left) and July 2000 (right). The source of the SST data is the NOAA NCDC ERSST ver.2 data, a global monthly SST data from
January 1854 with a spatial resolution of 2° × 2°.
 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135
conditioning during summers is usually kept more stable with less var-
iability. That is why the instrumental SST has a relatively larger variation
amplitude during winters than during summers (see Fig. 11).
From a broader and even hemispheric view, the pulse-type cold
wave is likely one major cause for the larger variability of winter tem-
peratures (Ding et al., 2014). For example, in regions like East Asia and
the east coast of North America, which are strongly influenced by the
winter pulse-type cold waves, hence winter temperature variability
was much larger than summer one (Fig. 13). Ding et al. (2014) recently
provided a comprehensive description and summary of the multi-scale
and multi-modal natures of the EAWM. These authors explained that
EAWM is modulated on timescale of several decades while being repre-
sented by at least three major meteorological modes that are concur-
rently operating and competing. On the prominent multidecadal
timescale, Ding et al. (2014)'s illuminating discussion of physical pro-
cesses covered not only ENSO but also the co-workings of AO, PDO
and AMO phenomena that are not necessarily local to East Asia or
South China Sea.
The modern instrumental data could provide us some additional
clues about the seasonal variations of our composite SST record during
the late Holocene. The similar patterns of winter and summer SST vari-
ations in northern SCS probably hinted at a close inverse relation be-
tween the EAWM and WPWFP conditions during the late Holocene,
with a weak EAWM corresponding to a warm WPWFP. This coupling
between EAWM and tropical Pacific has also been reported by many
previous studies using instrumental data and observations (Lau and Li,
1984; Masumoto and Yamagata, 1991; Li and Mu, 2000; Wang et al.,
2000).
The larger variability of atmosphere-controlled EAWM than ocean-
controlled WPWFP conditions derived from the instrumental data ex-
plains the relatively larger variation amplitude of winter SST than that
of summer SST in northern SCS during the late Holocene. We compare
our composite record with the SST reconstruction from the Indonesian
region of WPWFP. The variability of the winter proxy-SST in our com-
posite record was much larger than that of the proxy-SST record from
the Indonesia region of WPWFP, but the amplitude of the summer SST
60N
40N
20N
EQ
20S
40S
60S
0 60E 120E
0.2 0.25 0.33 0.5
Fig. 13. A comparison of the temperature variability during winters versus during summers. The red zones indicate areas with larger winter temperature variability. The values were cal-
culated from the ratio of the standard deviation of January temperatures to the standard deviation of July temperatures from 1982 to 2012. The source of the SST data is the NOAA NCDC
ERSST ver.2 data, a global monthly SST data from January 1854 with a spatial resolution of 2° × 2°. (For interpretation of the references to color in this figure legend, the reader is referred to
the web version of this article.)
133
variations in both our northern SCS and Indonesian WPWFP records
were comparable. This fact suggests the existence of different mecha-
nisms for the summer and winter SST variations in northern SCS during
the late Holocene.
The large differences in the amplitude of summer and winter SST
variations found in our study imply a potentially serious problem for
other syntheses of multi-proxy paleoclimate records. In many of the
current synthesis works (IPCC, 2001, 2007; Mann et al., 2008, 2009),
most of the proxy records considered and adopted have a time-
resolution too low to take seasonal effect into consideration and these
proxy records tend to reflect the climate of a particular season. For ex-
ample, most tree ring growth and increment in Northern Hemisphere
occur during boreal springs and summers and the tree ring width
index is basically recording summer-climate information (Briffa et al.,
1990; Büntgen et al., 2005). The climate information derived from the
phenological records and historical winter snow-day records in East
Asia region mainly reflect the early-spring and winter temperatures, re-
spectively (Ge et al., 2003; Aono and Kazui, 2008). Another example is
recently pointed out by Ekaykin et al. (2014) where these authors con-
vincingly demonstrated that their stacked δD isotopic proxy record for
the past three centuries from Vostok station in East Antarctica best rep-
resented summer temperature variations and that summer-sensitive
proxy behavior is distinctly different from variations in annual-mean
temperature (Ekaykin et al., 2014). The synthesis of these climate re-
cords may result in substantial uncertainty if the seasonal cycle ampli-
tude is large (i.e., especially over mid-to-high latitude regions;
including even highly sensitive tropical regions like northern SCS de-
scribed above) and that seasonal bias is a significant factor for the
proxies.
4. Conclusions
In this study, a composite multi-proxy record of the SST variation in
the northern SCS during the late Holocene is constructed by the combi-
nation of seven high-resolution coral and T. gigas Sr/Ca-based SST time
windows and the instrumental SST from 1990 to 2000 interval. The
180 120W 60W 0
1 2 3 4 5 7 10
134 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135
variation pattern and characteristic of the composite SST record is con-
sistent with other reconstructions from Eastern to Northwestern China
and tropical Western Pacific; they all indicate a prominent high-
amplitude and geographically wide distribution of the warm climate
during the RWP and MWP in the East Asian–Western Pacific region.
Meanwhile, we also observe the substantial difference in seasonal SST
variations in northern SCS during the late Holocene, with much larger
SST amplitude during winters than during summers, and the climatic
seasonality of northern SCS over the last 2500 years appears to be signif-
icantly impacted by East Asian Winter Monsoon. The importance of sea-
sonality for past climate states and climatic variations highlights the
potential uncertainty in other current synthesis works that adopted cli-
mate proxies that cannot resolve seasonal SST.
Acknowledgment
Financial support for this research was provided by the
Natural Science Foundation of China (NSFC) (41403018), Major
State Basic Research Development Program of China (973 Program)
(2013CB955900), the West Light Foundation of The Chinese Acade-
my of Sciences (29Y42909101) and the Key Programs of the Chinese
Academy of Sciences (55ZZBS1304101). We are grateful to Professor
Kefu Yu and Gangjian Wei for the constructive comments on earlier
versions of the manuscript, as well as useful comments from the ed-
itor and three anonymous reviewers.
References
Alibert, C., McCulloch, M.T., 1997. Strontium/calcium ratios in modern Porites corals from
the Great Barrier Reef as a proxy for sea surface temperature: calibration of the ther-
mometer and monitoring of ENSO. Paleoceanography 12 (3), 345–363.
Aono, Y., Kazui, K., 2008. Phenological data series of cherry tree flowering in Kyoto, Japan,
and its application to reconstruction of springtime temperatures since the 9th centu-
ry. Int. J. Climatol. 28 (7), 905–914.
Aubert, A., Lazareth, C., Cabioch, G., Boucher, H., Yamada, T., Iryu, Y., Farman, R., 2009. The
tropical giant clam Hippopus hippopus shell, a new archive of environmental condi-
tions as revealed by sclerochronological and δ18O profiles. Coral Reefs 28 (4),
989–998.
Barclay, D.J., Yager, E.M., Graves, J., Kloczko, M., Calkin, P.E., 2013. Late Holocene glacial
history of the Copper River Delta, coastal south-central Alaska, and controls on valley
glacier fluctuations. Quat. Sci. Rev. 81, 74–89.
Batenburg, S.J., Reichart, G.J., Jilbert, T., Janse, M., Wesselingh, F.P., Renema, W., 2011. In-
terannual climate variability in the Miocene: high resolution trace element and stable
isotope ratios in giant clams. Palaeogeogr. Palaeoclimatol. Palaeoecol. 306 (1–2),
75–81.
Beck, J., Edwards, R., Ito, E., Taylor, F., Recy, J., Rougerie, F., Joannot, P., Henin, C., 1992. Sea-
surface temperature from coral skeletal strontium/calcium ratios. Science 257
(5070), 644–647.
Briffa, K.R., Bartholin, T.S., Eckstein, D., Jones, P.D., Karlén, W., Schweingruber, F.H.,
Zetterberg, P., 1990. A 1,400-year tree-ring record of summer temperatures in
Fennoscandia. Nature 346, 434–439.
Büntgen, U., Esper, J., Frank, D.C., Nicolussi, K., Schmidhalter, M., 2005. A 1052-year tree-
ring proxy for Alpine summer temperatures. Clim. Dyn. 25 (2), 141–153.
Cobb, K.M., Charles, C.D., Cheng, H., Edwards, R.L., 2003. El Nino/southern oscillation and
tropical Pacific climate during the last millennium. Nature 424 (6946), 271–276.
Cole, J., Fairbanks, R., Shen, G., 1993. Recent variability in the southern oscillation: isotopic
results from a Tarawa atoll coral. Science 260 (5115), 1790–1793.
Cook, E., Buckley, B., D'Arrigo, R., Peterson, M., 2000. Warm-season temperatures since
1600 BC reconstructed from Tasmanian tree rings and their relationship to large-
scale sea surface temperature anomalies. Clim. Dyn. 16 (2–3), 79–91.
Cravatte, S., Delcroix, T., Zhang, D., McPhaden, M., Leloup, J., 2009. Observed freshening
and warming of the western Pacific warm pool. Clim. Dyn. 33 (4), 565–589.
Deng, W.F., Wei, G.J., Li, X.H., Yu, K.F., Zhao, J.X., Sun, W.D., Liu, Y., 2009. Paleoprecipitation
record from coral Sr/Ca and delta O-18 during the mid Holocene in the northern
South China Sea. The Holocene 19 (6), 811–821.
Denton, G.H., Alley, R.B., Comer, G.C., Broecker, W.S., 2005. The role of seasonality in
abrupt climate change. Quat. Sci. Rev. 24 (10), 1159–1182.
Ding, Y., Liu, Y., Liang, S., Ma, X., Zhang, Y., Si, D., Liang, P., Song, Y., Zhang, J., 2014.
Interdecadal variability of the East Asian winter monsoon and its possible links to
global climate change. J. Meteorol. Res. 28, 693–713.
Ekaykin, A., Kozachek, A., Lipenkov, V.Y., Shibaev, Y.A., 2014. Multiple climate shifts in the
Southern Hemisphere over the past three centuries based on central Antarctic snow
pits and core studies. Ann. Glaciol. 55 (66), 259–266.
Elliot, M., Welsh, K., Chilcott, C., McCulloch, M., Chappell, J., Ayling, B., 2009. Profiles of
trace elements and stable isotopes derived from giant long-lived Tridacna gigas bi-
valves: potential applications in paleoclimate studies. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 280 (1–2), 132–142.
Esper, J., Frank, D., 2009. The IPCC on a heterogeneous Medieval warm period. Clim.
Chang. 94 (3–4), 267–273.
Esper, J., Cook, E., Schweingruber, F., 2002. Low-frequency signals in long tree-ring chronol-
ogies for reconstructing past temperature variability. Science 295 (5563), 2250–2253.
Essex, C., 2011. Climate theory versus a theory for climate. Int. J. Bifurcation Chaos 21
(12), 3477–3487.
Essex, C., 2013. Does laboratory‐scale physics obstruct the development of a theory for cli-
mate? J. Geophys. Res. Atmos. 118 (3), 1218–1225.
Gagan, M., Chivas, A., Isdale, P., 1994. High-resolution isotopic records from corals using
ocean temperature and mass-spawning chronometers. Earth Planet. Sci. Lett. 121
(3–4), 549–558.
Ge, Q., Zheng, J., Fang, X., Man, Z., Zhang, X., Zhang, P., Wang, W., 2003. Winter half-
year temperature reconstruction for the middle and lower reaches of the Yellow
River and Yangtze River, China, during the past 2000 years. The Holocene 13 (6),
933–940.
Giry, C., Felis, T., Kolling, M., Wei, W., Lohmann, G., Scheffers, S., 2013. Controls of Caribbe-
an surface hydrology during the mid-to late Holocene: insights from monthly re-
solved coral records. Clim. Past 9 (2), 841–858.
Graham, N., Ammann, C., Fleitmann, D., Cobb, K., Luterbacher, J., 2011. Support for global
climate reorganization during the ‘Medieval Climate Anomaly’? Clim. Dyn. 37 (5–6),
1217–1245.
Grove, J., 1988. The Little Ice Age. Methuen, New York.
Guttman, N.B., 1989. Statistical descriptors of climate. Bull. Am. Meteorol. Soc. 70 (6),
602–607.
Hantemirov, R.M., Shiyatov, S.G., 2002. A continuous multimillennial ring-width chronol-
ogy in Yamal, northwestern Siberia. The Holocene 12 (6), 717–726.
Hass, H.C., 1996. Northern Europe climate variations during late Holocene: evidence from
marine Skagerrak. Palaeogeogr. Palaeoclimatol. Palaeoecol. 123 (1), 121–145.
He, Y., Liu, W., Zhao, C., Wang, Z., Wang, H., Liu, Y., Qin, X., Hu, Q., An, Z., Liu, Z., 2013. Solar
influenced late Holocene temperature changes on the northern Tibetan Plateau. Chin.
Sci. Bull. 57 (1), 1–7.
Huang, C.-Y., Wu, S.-F., Zhao, M., Chen, M.-T., Wang, C.-H., Tu, X., Yuan, P.B., 1997. Surface
ocean and monsoon climate variability in the South China Sea since the last glacia-
tion. Mar. Micropaleontol. 32 (1), 71–94.
Hughes, M.K., Diaz, H.F., 1994. Was there a ‘Medieval Warm Period’, and if so, where and
when? Clim. Chang. 26 (2–3), 109–142.
IPCC, 2001. Climate Change 2001: The Scientific Basis. Cambridge University Press,
Cambridge.
IPCC, 2007. Climate Change 2007: The Physical Science Basis. Cambridge University Press,
Cambridge.
Joerin, U.E., Stocker, T.F., Schlüchter, C., 2006. Multicentury glacier fluctuations in the
Swiss Alps during the Holocene. The Holocene 16 (5), 697–704.
Kukla, G., Gavin, J., 2004. Milankovitch climate reinforcements. Glob. Planet. Chang. 40
(1), 27–48.
Lamb, H.H., 1977. Climate: present, past and future. Climatic history and the future vol. 2.
Methuen & Company.
Landsberg, H.E., 1972. Weather “normals” and normal weather. NOAA Environmental
Data Service, pp. 8–13 (October issue).
Lau, K.-M., Li, M.-T., 1984. The monsoon of East Asia and its global associations—a survey.
Bull. Am. Meteorol. Soc. 65 (2), 114–125.
Li, C., Mu, M., 2000. Relationship between East Asian winter monsoon, warm pool situa-
tion and ENSO cycle. Chin. Sci. Bull. 45 (16), 1448–1455.
Liu, Y., An, Z., Linderholm, H.W., Chen, D., Song, H., Cai, Q., Sun, J., Tian, H., 2009. Annual
temperatures during the last 2485 years in the mid-eastern Tibetan Plateau inferred
from tree rings. Sci. China Ser. D Earth Sci. 52 (3), 348–359.
Ljungqvist, F.C., 2010. A new reconstruction of temperature variability in the extra-
tropical Northern Hemisphere during the last two millennia. Geogr. Ann. Ser. A
Phys. Geogr. 92 (3), 339–351.
Maier, E., Titschack, J., 2010. Spondylus gaederopus: a new Mediterranean climate archive—
based on high-resolution oxygen and carbon isotope analyses. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 291 (3–4), 228–238.
Mann, M.E., Zhang, Z., Hughes, M.K., Bradley, R.S., Miller, S.K., Rutherford, S., Ni, F., 2008.
Proxy-based reconstructions of hemispheric and global surface temperature varia-
tions over the past two millennia. Proc. Natl. Acad. Sci. 105 (36), 13252–13257.
Mann, M., Zhang, Z., Rutherford, S., Bradley, R., Hughes, M., Shindell, D., Ammann, C.,
Faluvegi, G., Ni, F., 2009. Global signatures and dynamical origins of the little ice
age and medieval climate anomaly. Science 326 (5957), 1256–1260.
Martín-Puertas, C., Valero-Garcés, B.L., Brauer, A., Mata, M.P., Delgado-Huertas, A., Dulski,
P., 2009. The Iberian-Roman Humid Period (2600–1600 cal yr BP) in the Zoñar Lake
varve record (Andalucía, southern Spain). Quat. Res. 71 (2), 108–120.
Masson, V., Vimeux, F., Jouzel, J., Morgan, V., Delmotte, M., Ciais, P., Hammer, C., Johnsen,
S., Lipenkov, V.Y., Mosley-Thompson, E., 2000. Holocene climate variability in
Antarctica based on 11 ice-core isotopic records. Quat. Res. 54 (3), 348–358.
Masumoto, Y., Yamagata, T., 1991. Response of the western tropical Pacific to the Asian
winter monsoon: the generation of the Mindanao Dome. J. Phys. Oceanogr. 21 (9),
1386–1398.
Matthes, F.E., 1939. Report of committee on glaciers, April 1939. Trans. Am. Geophys.
Union 20, 518–523.
Matthes, F.E., 1940. Committee on glaciers, 1939–40. Trans. Am. Geophys. Union 21,
396–406.
McCulloch, M., Gagan, M., Mortimer, G., Chivas, A., Isdale, P., 1994. A high-resolution Sr/Ca
and δ18O coral record from the Great Barrier Reef, Australia, and the 1982–1983 El
Nino. Geochim. Cosmochim. Acta 58 (12), 2747–2754.
McCulloch, M., Mortimer, G., Esat, T., Xianhua, L., Pillans, B., Chappell, J., 1996. High reso-
lution windows into early Holocene climate: Sr/Ca coral records from the Huon Pen-
insula. Earth Planet. Sci. Lett. 138 (1–4), 169–178.
 H. Yan et al. / Earth-Science Reviews 141 (2015) 122–135
Moberg, A., Sonechkin, D., Holmgren, K., Datsenko, N., Karlén, W., 2005. Highly variable
Northern Hemisphere temperatures reconstructed from low-and high-resolution
proxy data. Nature 433 (7026), 613–617.
Naidu, P., Niitsuma, N., Naik, S., 2014. Oxygen isotopic analyses of individual planktic fo-
raminifera species: implications for seasonality in the western Arabian Sea. Clim. Past
Discuss. 10 (5), 3661–3688.
Oppo, D.W., Rosenthal, Y., Linsley, B.K., 2009. 2,000-year-long temperature and hydrology
reconstructions from the Indo-Pacific warm pool. Nature 460 (7259), 1113–1116.
Patterson, W.P., Dietrich, K.A., Holmden, C., Andrews, J.T., 2010. Two millennia of North
Atlantic seasonality and implications for Norse colonies. Proc. Natl. Acad. Sci. 107
(12), 5306–5310.
Peng, Z.C., Chen, T.G., Nie, B.F., Head, M.J., He, X.X., Zhou, W.J., 2003. Coral delta O-18 re-
cords as an indicator of winter monsoon intensity in the South China Sea. Quat.
Res. 59 (3), 285–292.
Rosewater, J., 1965. The family Tridacnidae in the Indo-Pacific. Indo Pac. Mollusca 1,
347–396.
Schöne, B.R., Dunca, E., Mutvei, H., Norlund, U., 2004a. A 217-year record of summer
air temperature reconstructed from freshwater pearl mussels (M-margarifitera,
Sweden). Quat. Sci. Rev. 23 (16–17), 1803–1816.
Schöne, B.R., Freyre Castro, A.D., Fiebig, J., Houk, S.D., Oschmann, W., Kroncke, I., 2004b.
Sea surface water temperatures over the period 1884–1983 reconstructed from oxy-
gen isotope ratios of a bivalve mollusk shell (Arctica islandica, southern North Sea).
Palaeogeogr. Palaeoclimatol. Palaeoecol. 212 (3–4), 215–232.
Schöne, B.R., Fiebig, J., Pfeiffer, M., Gless, R., Hickson, J., Johnson, A.L.A., Dreyer, W.,
Oschmann, W., 2005. Climate records from a bivalved Methuselah (Arctica islandica,
Mollusca; Iceland). Palaeogeogr. Palaeoclimatol. Palaeoecol. 228 (1–2), 130–148.
Seidenkrantz, M.-S., Aagaard-Sørensen, S., Sulsbrück, H., Kuijpers, A., Jensen, K.G.,
Kunzendorf, H., 2007. Hydrography and climate of the last 4400 years in a SW Green-
land fjord: implications for Labrador Sea palaeoceanography. The Holocene 17 (3),
387–401.
Shintani, T., Yamamoto, M., Chen, M.-T., 2011. Paleoenvironmental changes in the north-
ern South China Sea over the past 28,000 years: a study of TEX86-derived sea surface
temperatures and terrestrial biomarkers. J. Asian Earth Sci. 40 (6), 1221–1229.
Soon, W.W.-H., 2009. Solar arctic-mediated climate variation on multidecadal to centen-
nial timescales: empirical evidence, mechanistic explanation, and testable conse-
quences. Phys. Geogr. 30 (2), 144–184.
Soon, W., Baliunas, S., Idso, C., Idso, S., Legates, D.R., 2003. Reconstructing climatic and en-
vironmental changes of the past 1000 years: a reappraisal. Energy Environ. 14 (2),
233–296.
Steinke, S., Prange, M., Feist, C., Groeneveld, J., Mohtadi, M., 2014. Upwelling variability off
southern Indonesia over the past two millennia. Geophys. Res. Lett. http://dx.doi.org/
10.1002/2014GL061450.
Sun, D.H., Gagan, M.K., Cheng, H., Scott-Gagan, H., Dykoski, C.A., Edwards, R.L., Sua, R.X.,
2005. Seasonal and interannual variability of the Mid-Holocene East Asian monsoon
in coral delta O-18 records from the South China Sea. Earth Planet. Sci. Lett. 237
(1–2), 69–84.
Tan, M., Liu, T., Hou, J., Qin, X., Zhang, H., Li, T., 2003. Cyclic rapid warming on centennial-
scale revealed by a 2650-year stalagmite record of warm season temperature.
Geophys. Res. Lett. 30 (12), 1617.
Wanamaker, A.D., Kreutz, K.J., Schöne, B.R., Pettigrew, N., Borns, H.W., Introne, D.S.,
Belknap, D., Maasch, K.A., Feindel, S., 2008. Coupled North Atlantic slope water forc-
ing on Gulf of Maine temperatures over the past millennium. Clim. Dyn. 31 (2),
183–194.
Wanamaker, A.D., Kreutz, K.J., Schöne, B.R., Introne, D.S., 2011. Gulf of Maine shells reveal
changes in seawater temperature seasonality during the Medieval Climate Anomaly
and the Little Ice Age. Palaeogeogr. Palaeoclimatol. Palaeoecol. 302 (1), 43–51.
Wang, L., Sarnthein, M., Erlenkeuser, H., Grimalt, J., Grootes, P., Heilig, S., Ivanova, E.,
Kienast, M., Pelejero, C., Pflaumann, U., 1999. East Asian monsoon climate during
View publication stats
135
the Late Pleistocene: high-resolution sediment records from the South China Sea.
Mar. Geol. 156 (1), 245–284.
Wang, B., Wu, R., Fu, X., 2000. Pacific-East Asian teleconnection: how does ENSO affect
East Asian climate? J. Clim. 13 (9), 1517–1536.
Wang, S.W., Wen, X.Y., Luo, Y., Dong, W.J., Zhao, Z.C., Yang, B., 2007. Reconstruction of
temperature series of China for the last 1000 years. Chin. Sci. Bull. 52 (23),
3272–3280.
Watanabe, T., Oba, T., 1999. Daily reconstruction of water temperature from oxygen iso-
topic ratios of a modern Tridacna shell using a freezing microtome sampling tech-
nique. J. Geophys. Res. Oceans 104 (C9), 20667–20674.
Watanabe, T., Suzuki, A., Kawahata, H., Kan, H., Ogawa, S., 2004. A 60-year isotopic record
from a mid-Holocene fossil giant clam (Tridacna gigas) in the Ryukyu Islands: physi-
ological and paleoclimatic implications. Palaeogeogr. Palaeoclimatol. Palaeoecol. 212
(3–4), 343–354.
Wei, G., Yu, K., Li, X., Zhao, J., Sun, Y., Nie, B., 2004a. Coralline Sr/Ca and Mg/Ca thermom-
eter for the northern South China Sea: calibration and primary application on high
resolution SST reconstructing. Quat. Sci. 24 (003), 325–331 (In Chinese with English
Abstract).
Wei, G.J., Yu, K.F., Zhao, J.X., 2004b. Sea surface temperature variations recorded on coral-
line Sr/Ca ratios during Mid-Late Holocene in Leizhou Peninsula. Chin. Sci. Bull. 49
(17), 1876–1881.
Wei, G.J., Deng, W.F., Yu, K.F., Li, X.H., Sun, W.D., Zhao, J.X., 2007. Sea surface temperature
records in the northern South China Sea from mid-Holocene coral Sr/Ca ratios.
Paleoceanography 22 (3), PA3206.
Williams, D.F., Arthur, M.A., Jones, D.S., Healy-Williams, N., 1982. Seasonality and mean
annual sea surface temperatures from isotopic and sclerochronological records. Na-
ture (296), 432–434.
Wu, W., Tan, W., Zhou, L., Yang, H., Xu, Y., 2012. Sea surface temperature variability in
southern Okinawa Trough during last 2700 years. Geophys. Res. Lett. 39 (14), L14705.
Yan, H., Sun, L., Liu, X., Qiu, S., 2010. Relationship between ENSO events and regional cli-
mate anomalies around the Xisha Islands during the last 50 years. J. Trop. Oceanogr.
29 (5), 29–35 (In Chinese with English abstract).
Yan, H., Sun, L., Oppo, D.W., Wang, Y., Liu, Z., Xie, Z., Liu, X., Cheng, W., 2011. South China
Sea hydrological changes and Pacific walker circulation variations over the last mil-
lennium. Nat. Commun. 2, 293.
Yan, H., Shao, D., Wang, Y., Sun, L., 2013. Sr/Ca profile of long-lived Tridacna gigas bivalves
from South China Sea: a new high-resolution SST proxy. Geochim. Cosmochim. Acta
112, 52–65.
Yan, H., Shao, D., Wang, Y., Sun, L., 2014a. Sr/Ca differences within and among three
Tridacnidae species from the South China Sea: implication for paleoclimate recon-
struction. Chem. Geol. 390, 22–31.
Yan, H., Sun, L., Shao, D., Wang, Y., 2014b. Higher sea surface temperature in the northern
South China Sea during the natural warm periods of late Holocene than recent de-
cades. Chin. Sci. Bull. 59, 4115–4122.
Yan, H., Wang, Y., Sun, L., 2014c. High resolution oxygen isotope and grayscale records of a
medieval fossil giant clam (Tridacna gigas) in the South China Sea: physiological and
paleoclimatic implications. Acta Oceanol. Sin. 33 (8), 18–25.
Yang, B., Braeuning, A., Johnson, K.R., Yafeng, S., 2002. General characteristics of temper-
ature variation in China during the last two millennia. Geophys. Res. Lett. 29 (9)
(38-1-38-4).
Yu, K., Zhao, J., Liu, T., Wei, G., Wang, P., Collerson, K., 2004. High-frequency winter cooling
and reef coral mortality during the Holocene climatic optimum. Earth Planet. Sci. Lett.
224 (1–2), 143–155.
Yu, K., Zhao, J., Wei, G., Cheng, X., Wang, P., 2005. Mid-late Holocene monsoon climate re-
trieved from seasonal Sr/Ca and δ18O records of Porites lutea corals at Leizhou Penin-
sula, northern coast of South China Sea. Glob. Planet. Chang. 47 (2–4), 301–316.