of 10
1.
Introduction
The Indo-Pacific Warm Pool (IPWP) is a critical region of the global climate system, associated with the Pacific
and Indian Walker cells and the Asian and Australian-Indonesian monsoon systems. Partin et al. (
2007
) presented
the first stalagmite δ
18
O record used to reconstruct convection changes over the IPWP spanning the region's tran
-
sition from the Last Glacial Maximum to the Holocene. The δ
18
O record was constructed from three individual
stalagmites, collected from Gunung Buda National Park cave systems in northern Borneo (4°N, 115°E). The
strongest millennial-scale δ
18
O signal evident in all three records was a gradual shift through Heinrich Stadial 1
(HS1) to the highest δ
18
O values of the record, interpreted as a signal of weakening convection over the western
tropical Pacific through this period. The record also demonstrated the warm pool's relatively smooth variability
Abstract
Recent paleoclimate reconstructions have suggested millennial-scale variability in the Indo-
Pacific Warm Pool region coincident with events of the last deglaciation. Here, we present a new stalagmite
oxygen isotope record from northern Borneo, which today is located near the center of the region's mean
annual intertropical convergence zone. The record spans the full deglaciation, and reveals for the first time
distinct oxygen isotope variations at this location connected with the Bølling-Allerød onset and the Younger
Dryas event. The full deglaciation in the Borneo stalagmite proxy reconstruction appears remarkably similar
to a 20–11 ka transient simulation of rainfall over the area produced using the isotope-enabled Community
Earth System Model. In this model, periods of weakened Atlantic Ocean meridional overturning circulation
are associated with an anomalous Western North Pacific anticyclone, which is produced in boreal autumn and
shifts south over Borneo during boreal winter, causing dry conditions.
Plain Language Summary
Here, we aim to resolve conflicting evidence of how tropical convection
in Borneo in the Indo-Pacific oceanic region, a critical region associated with El Niño events and the Asian
and Australian-Indonesian monsoon systems, was affected by past changes in the strength of Atlantic Ocean
circulation. Analyzing a 20,000+ year old Borneo cave stalagmite along its growth axis, we find large oxygen
isotope shifts coincident with two prominent millennial-scale periods of reduced Atlantic Ocean overturning
circulation strength. The isotope record is similar to a neighboring southwestern Philippine stalagmite record,
and we interpret the signal as drier mean annual conditions over the region during these events compared
to the background state; an interpretation supported by simulations in three independent state-of-the-art
climate models. One model suggests that these conditions in Borneo were driven by the southward shift of an
anomalous anticyclone in the boreal winter season. By combining geochemical and model evidence, this study
has shown that rainfall decreased for millennial periods during the end of the ice age, when the Atlantic Ocean
overturning circulation weakened. In doing so, we have elucidated further how this region responds to major
changes in global climate and ocean conditions.
BUCKINGHAM ET AL.
© 2022. American Geophysical Union.
All Rights Reserved.
Termination 1 Millennial-Scale Rainfall Events Over the
Sunda Shelf
F. L. Buckingham
1
, S. A. Carolin
1,2
, J. W. Partin
3
, J. F. Adkins
4
, K. M. Cobb
5
,
C. C. Day
1
, Q. Ding
6
, C. He
7
, Z. Liu
7
, B. Otto-Bliesner
8
, W. H. G. Roberts
9
,
S. Lejau
10
, and J. Malang
10
1
Department of Earth Sciences, University of Oxford, Oxford, UK,
2
Department of Earth Sciences, University of Cambridge,
Cambridge, UK,
3
Institute for Geophysics, Jackson School of Geosciences, University of Texas at Austin, Austin, TX, USA,
4
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA, USA,
5
Department
of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA, USA,
6
Department of Geography and
Earth Research Institute, University of California Santa Barbara, Santa Barbara, CA, USA,
7
Department of Geography, The
Ohio State University, Columbus, OH, USA,
8
Climate and Global Dynamics Laboratory, National Center for Atmospheric
Research, Boulder, CO, USA,
9
Geography and Environmental Sciences, Northumbria University, Newcastle-Upon-Tyne, UK,
10
Gunung Mulu National Park, Sarawak, Malaysia
Key Points:
A new stalagmite record reveals for
the first time distinct Bølling-Allerød
and Younger Dryas oxygen isotope
variations in northern Borneo
The stalagmite oxygen isotope pattern
strongly resembles the recent iCESM
transient annual rainfall simulation
from 20 to 11 ka
Borneo drying during Heinrich 1 and
the Younger Dryas may be attributed
to an anomalous boreal winter
Western North Pacific anticyclone
Supporting Information:
Supporting Information may be found in
the online version of this article.
Correspondence to:
S. A. Carolin,
sac219@cam.ac.uk
Citation:
Buckingham, F. L., Carolin, S. A.,
Partin, J. W., Adkins, J. F., Cobb, K. M.,
Day, C. C., et al. (2022). Termination
1 millennial-scale rainfall events over
the Sunda Shelf.
Geophysical Research
Letters
,
49
, e2021GL096937.
https://doi.
org/10.1029/2021GL096937
Received 8 NOV 2021
Accepted 31 JAN 2022
Author Contributions:
Conceptualization:
F. L. Buckingham,
S. A. Carolin
Data curation:
S. A. Carolin
Formal analysis:
F. L. Buckingham,
S. A. Carolin, J. W. Partin, J. F. Adkins,
K. M. Cobb, C. C. Day, Q. Ding, C. He,
Z. Liu, B. Otto-Bliesner, W. H. G. Roberts
Funding acquisition:
S. A. Carolin
Investigation:
F. L. Buckingham, S. A.
Carolin, C. He, S. Lejau, J. Malang
Methodology:
F. L. Buckingham, S. A.
Carolin, C. C. Day, Q. Ding
Project Administration:
S. A. Carolin
Resources:
F. L. Buckingham, S. A.
Carolin, J. W. Partin, S. Lejau, J. Malang
Supervision:
S. A. Carolin
Validation:
F. L. Buckingham, S. A.
Carolin
Visualization:
F. L. Buckingham, S. A.
Carolin, W. H. G. Roberts
10.1029/2021GL096937
RESEARCH LETTER
1 of 10
Geophysical Research Letters
BUCKINGHAM ET AL.
10.1029/2021GL096937
2 of 10
on both orbital and millennial timescales, in contrast to the abrupt shifts seen in midlatitude Chinese stalagmite
δ
18
O records (e.g., Y. J. Wang et al.,
2001
).
A significant feature of the northern Borneo stalagmite record was the lack of a δ
18
O signal coincident with the
Younger Dryas (YD) event, despite a clear HS1 signal (Partin et al.,
2007
). This appeared to suggest convec
-
tion in the IPWP responded differently to the HS1 versus YD North Atlantic events (Partin et al.,
2007
), both
of which are associated with a weakening of the Atlantic Meridional Overturning Circulation (AMOC) (e.g.,
McManus et al.,
2004
; Böhm et al.,
2015
; Ng et al.,
2018
). The YD conclusion at the time was based on a single
stalagmite record, SCH02, as vertical growth in the other two Borneo stalagmites stopped prior to the YD event,
and resumed after the event. A more recent study identified U-Th dating disparities in stalagmite SCH02 during
Heinrich Stadial 4 (Carolin et al.,
2013
), warranting additional study of the Borneo rainfall response to millennial
events of the last deglaciation. Indeed, other stalagmite records found in the IPWP in Sumatra, Indonesia to the
west (0°S, 100°E) (Wurtzel et al.,
2018
), Palawan, Philippines to the northeast (10°N, 119°E) (Partin et al.,
2015
),
and Flores, Indonesia to the south (8.5°S, 120°E) (Griffiths et al.,
2009
; Ayliffe et al.,
2013
) all reveal distinct
δ
18
O shifts coincident with the YD event.
Tropical climate perturbations associated with a weakened AMOC are proposed to result from processes such as
a southward shift of the intertropical convergence zone (ITCZ) over the Pacific (Zhang & Delworth,
2005
,
2007
)
or cross-basin interactions between the Atlantic and the El Niño-Southern Oscillation (ENSO) or Pacific Decadal
Oscillation (Timmermann et al.,
2007
; C. Wang,
2019
). To better understand underlying dynamics linking the
AMOC with the local circulation change over the IPWP, a thorough analysis of the various proposed processes is
needed. Improved climate proxy information in observation records that can be used in model-data comparison
contributes to this aim.
Despite the increasing number of stalagmite sites and proxy records, it remains challenging to interpret IPWP
stalagmite δ
18
O records as changes in past rainfall amount and perturbations to large-scale atmospheric systems
over the last deglaciation. This is largely due to the complexity of the maritime continent region. Under modern
conditions, individual sites around the IPWP have unique rainfall patterns contributing to the mean annual rain
-
fall δ
18
O value, as a result of the seasonal ITCZ migration, the intra-annual Madden-Julian Oscillation, and
the Pacific and/or Indian Ocean Walker circulation cells' interannual variability (e.g., Moerman et al.,
2013
;
Belgaman et al.,
2017
; Wurtzel et al.,
2018
; Konecky et al.,
2019
). During the last deglaciation, the Sunda
Shelf was widely exposed (e.g., Hanebuth et al.,
2000
), placing the paleo-coastline further from some cave
sites than modern, such as those in northern Borneo, and likely reorganizing regional circulation patterns (e.g.,
DiNezio et al.,
2016
; Windler et al.,
2020
). Fully coupled general circulation model simulations with enabled
water isotopes is a tool that can be utilized to assist with IPWP stalagmite δ
18
O interpretations (e.g., Windler
et al.,
2020
; Du et al.,
2021
).
Here, we present a new northern Borneo stalagmite δ
18
O record, the first from the area with continuous growth
through the last deglaciation, to test for climate responses in the IPWP to changes in AMOC and associated
climate events in the North Atlantic. The stalagmite contains clean calcite that is ideal for U-Th dating, and has
previously demonstrated its ability to capture millennial scale events through the last glacial period (Carolin
et al.,
2013
). We use a recent transient simulation of the evolution of global climate and water isotopes over
the last deglaciation (C. He et al.,
2021c
) to investigate if the new Borneo stalagmite δ
18
O response and other
stalagmite δ
18
O responses around the IPWP are present in a simulated climate with water isotopes, and if so what
atmospheric mechanism causes rainfall variability over the Sunda Shelf in the model simulation. We compare
the recent isotope-equipped transient simulation's Heinrich Stadial event response with two other state-of-the-art
climate model simulations to review model consistency. We aim to reconcile evidence learned from the real world
and model world to gain new physical insights on how IPWP convection responded to orbital forcing as well as
significant changes to the AMOC over the North Atlantic.
2.
Methods
2.1.
Site Location
The stalagmite presented in this study, SC02, was collected from Secret Cave in Gunung Mulu National Park,
which is southwest-adjacent to the Gunung Buda National Park in northern Borneo (Text S1, Figure S1 in the
Supporting Information
S1
). The site sits in the center of the tropical convection zone, today receiving ∼5,000 mm
Writing – original draft:
F. L.
Buckingham, S. A. Carolin
Writing – review & editing:
F. L.
Buckingham, S. A. Carolin, J. W. Partin,
J. F. Adkins, K. M. Cobb, C. C. Day, Q.
Ding, C. He, Z. Liu, B. Otto-Bliesner,
W. H. G. Roberts
19448007, 2022, 5, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2021GL096937 by California Inst of Technology, Wiley Online Library on [06/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters
BUCKINGHAM ET AL.
10.1029/2021GL096937
3 of 10
rainfall annually, with little seasonal variability in either temperature or rainfall amount. Over a decade of daily
rainwater and biweekly drip water monitoring shows a small deviation to higher rainwater δ
18
O values in the
February-March-April months (Figure
1
). The linear regression correlation coefficient between annually aver
-
aged δ
18
O and rainfall amount at the study site is −0.67, suggesting ∼45% of the variance in annual rainfall δ
18
O
can be explained by variability in local interannual rainfall amount (Ellis et al.,
2020
). Dry local conditions asso
-
ciated with El Niño events are responsible for the largest interannual rainfall δ
18
O excursions at this site under
modern conditions (Moerman et al.,
2013
; Ellis et al.,
2020
).
2.2.
Geochemical Methods
The U-Th age model for this study was constructed on the upper portion of stalagmite SC02 from 18 U-Th
samples. Carolin et al. (
2013
,
2016
) reported U-Th ages measured in the lower portion of stalagmite SC02,
105–31 ka (thousand years before 1950 C.E.). In the upper portion, four of the 18 U-Th samples were analyzed
in the Division of Geological and Planetary Sciences at the California Institute of Technology, and 14 samples
were analyzed in the Department of Earth Sciences at the University of Oxford, following the methods of Caro
-
lin et al. (
2016
,
2019
). A detailed description of the U-Th age sampling, multicollector inductively coupled-
plasma mass spectrometer (MC-ICP-MS) measurements, and sample age error analysis is found in Text S2 in the
Supporting Information
S1
.
Previous studies have found a wide range of initial detrital
230
Th/
232
Th values in Buda and Mulu stalagmites,
which are used to correct for any detrital
230
Th captured within the calcite crystals during formation. Individ
-
ual isochrons suggest values from 45 to 200 ppm as atom ratios (or 8–37 as activity ratios) (Partin et al.,
2007
;
Carolin et al.,
2013
,
2016
; Chen et al.,
2016
), several times larger than the commonly used bulk surface sili
-
cate
230
Th/
232
Th value of ∼5 ppm. The initial detrital
230
Th/
232
Th value of 111 ± 41 ppm was previously calcu
-
lated for SC02 based on two isochrons measured in separate stalagmites from Secret Cave in Mulu (methods and
calculations of Carolin et al.,
2013
in the Supporting Information
S1
), and is used here to correct the U-Th ages
in the upper portion of SC02 for initial detrital Th contamination.
Stable isotope samples were milled along the growth axis of SC02 in the Department of Earth Sciences at the
University of Oxford using a New Wave Micromill. Samples were milled in a continuous trench style at 0.2 mm
intervals around the YD calcite section and at 1 mm intervals in the other sections (Text S3 in the Supporting
Information
S1
). Samples were analyzed on a Delta V Advantage isotope ratio mass spectrometer coupled to
Figure 1.
Location of study site (circled) and other stalagmite paleoclimate sites in the text: Palawan, 10°N (Partin
et al.,
2015
); Borneo, 4°N (this study); Sumatra, 0°S (Wurtzel et al.,
2018
); Sulawesi, 5°S (Krause et al.,
2019
); Flores, 8.5°S
(Griffiths et al.,
2009
; Ayliffe et al.,
2013
); Western Australia, 17°S (Denniston et al.,
2013
). Triangle symbols indicate
cave site is predominantly boreal (up) or austral (down) summer rainfall. Circle symbol sites have little rainfall seasonality.
Contours show mean annual sea surface temperatures (Hersbach et al.,
2019
). Black filled shapes show modern landmasses,
with the gray contours at 60 m below sea level modern bathymetry, indicating the estimated Younger Dryas event coastlines.
Long term average (2006–2018) monthly rainfall δ
18
O with 1σ error bars at this study site are shown on the right (Moerman
et al.,
2013
; Ellis et al.,
2020
). Note the
y
-axis is inverted.
19448007, 2022, 5, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2021GL096937 by California Inst of Technology, Wiley Online Library on [06/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters
BUCKINGHAM ET AL.
10.1029/2021GL096937
4 of 10
either a Kiel IV carbonate device or Gas Bench II introductory system at the
University of Oxford. Oxygen and carbon isotope results are given in parts
per mil (‰) relative to the Vienna Pee Dee Belemnite Standard (VPDB).
Average measurement error is less than 0.05‰ (1σ) for δ
13
C and 0.10‰
(1σ) for δ
18
O. Interpolated ages at all stable isotope sample locations between
individual U-Th age sample trenches were computed using the Poisson-pro
-
cess deposition model feature in OxCal (v4.4) (Bronk Ramsey,
2008
,
2009
;
Bronk Ramsey & Lee,
2013
) (Text S4 in the Supporting Information
S1
).
2.3.
Climate Simulation
For this study we compare stalagmite δ
18
O records with iTRACE (C. He
et al.,
2021c
), a 20 ka–11 ka transient simulation of global climate using
the Community Earth System Model version 1.3 with fully coupled water
isotope modules (iCESM) (Brady et al.,
2019
). The iTRACE simulation well
reproduces the deglacial climate and water isotope evolution in the pan-Asian
monsoon regions and Greenland (C. He et al.,
2021a
,
2021c
). The iTRACE
experiments generally follow the strategy of the transient TraCE-21ka simu
-
lations (Liu et al.,
2009
; F. He,
2011
), with meltwater flux designed to be
largely consistent with reconstructed sea level change (Lambeck et al.,
2014
),
continental ice sheets modified at the beginning of each 1,000 years interval
based on the ICE-6G reconstruction (Peltier et al.,
2015
), and greenhouse
gases prescribed throughout following reconstructions (Petit et al.,
1999
;
Lüthi et al.,
2008
; Schilt et al.,
2010
) (Figure
2a
in C. He et al.,
2021c
). At 14
ka, the bathymetry of the Sunda Shelf coast line shifts to the −75 m prein
-
dustrial isobath, and at 12 ka, sea level rises further around the Sunda Shelf,
according to ICE-6G (Peltier et al.,
2015
). As such, any abrupt shifts in the
simulated climate at precisely 14 ka and 12 ka are discounted, as these are
artifacts of the imposed abrupt coastline changes in the model at those times
(C. He et al.,
2021c
). A full description of the model and its experiments can
be found of C. He et al. (
2021c
) in the Supporting Information
S1
.
3.
Results
3.1.
SC02 δ
18
O Record
Growth within the upper portion of SC02 spans the full deglaciation
(Figure
2
, Figure S2 and S3 in the Supporting Information
S1
). The U-Th
ages do not include any age reversals, with all ages within the sequential
order of growth within error (Figure S3, Data Set S1 in the Supporting Infor
-
mation
S1
). Vertical extension rate is slow but relatively constant, between
5 and 15 μm/yr (Figure S3, Data Set S2 in the Supporting Information
S1
).
Over orbital timescales (∼5 kyr), changes in δ
18
O in SC02 match those in the
original Buda δ
18
O records (Partin et al.,
2007
), however, millennial-scale
δ
18
O shifts are most prominent in this study's SC02 record (Figure
2
). The SC02 δ
18
O record's mean value is
also the lowest of the group of four Mulu and Buda stalagmite records. Larger amplitude δ
18
O variability and
lower mean δ
18
O values suggest this study's Mulu Secret Cave stalagmite was least affected by mixing (Moerman
et al.,
2014
; Ellis et al.,
2020
) and kinetic effects (Hendy,
1971
).
The SC02 δ
18
O record looks similar to the original Buda Snail Shell Cave SCH02 δ
18
O record (Partin et al.,
2007
)
on depth scale, however, the 95% CI of the timing of onset and ending of δ
18
O events are not in alignment between
the Mulu SC02 and original Buda SCH02 records (Figure S4 in the Supporting Information
S1
). The disagreement
between the Mulu SC02 and Buda SCH02 event timing is not caused by the initial detrital
230
Th/
232
Th correction
(Figure S4 in the Supporting Information
S1
). We conclude that something inherent to the original Buda SCH02
sample geochemistry caused inaccuracies in calculated U-Th ages on millennial timescales, possibly related to
Figure 2.
(a) SC02 (this study; black) and BA04, SSC01, and SCH02 (Partin
et al.,
2007
; green, red, and blue, respectively) δ
18
O records, plotted on a
single
y
-axis. (b) Individual SC02 (this study), and BA04, SSC01, and SCH02
(Partin et al.,
2007
) δ
18
O records. For both (a and b), the BA04, SSC01, and
SCH02 records are plotted on their originally published age models. U-Th
ages of all samples were corrected using the mean initial detrital
230
Th/
232
Th
as atom ratios listed. Solid black vertical lines in (b) are at 14.7, 12.87, and
11.7 ka.
19448007, 2022, 5, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2021GL096937 by California Inst of Technology, Wiley Online Library on [06/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters
BUCKINGHAM ET AL.
10.1029/2021GL096937
5 of 10
the greater amount of detrital material, larger number of growth hiatuses, slow growth portions, unresolved, short
hiatuses found in this sample during certain time intervals, and/or uranium loss due to open system behavior. The
inaccurate timing of certain millennial scale events in SCH02 could only be recognized through comparison with
other same-site samples after the original publication.
The new SC02 δ
18
O record reveals for the first time evidence of abrupt northern Borneo stalagmite δ
18
O changes,
which overlap within age error, and appear similar to, the North Atlantic Bølling-Allerød onset (∼14.7 ka, e.g.,
Buizert et al.,
2014
), Younger Dryas onset (12.85 ± 0.06 ka, Cheng et al.,
2020
), and Younger Dryas ending
(11.70–11.61 ± 0.04 ka, Cheng et al.,
2020
) (Figure S4 in the Supporting Information
S1
).
3.2.
iTRACE IPWP Model-Data Comparison
Figure
3
shows the full iTRACE simulated time series of rainfall and calculated calcite δ
18
O, a function of rain
-
fall δ
18
O and surface temperature (Text S5 in the Supporting Information
S1
), compared with stalagmite δ
18
O
from northern Borneo (stalagmite SC02) and five other sites around the IPWP (see Figure
1
map), on a 50-year
moving mean of the annual average. The shape of the simulated rainfall variability around the northern Borneo
site is remarkably similar to the SC02 stalagmite δ
18
O record (inverse relationship between Figure
3b
left and
middle columns) through the whole deglaciation, and there is also general agreement between the derived model
calcite δ
18
O and the SC02 stalagmite δ
18
O record (Figure
3b
middle and right columns). The directional shift
in δ
18
O in both northern and southern IPWP stalagmites is inversely related to each site's respective simulated
rainfall shift during the deglaciation millennial events, changes consistent with the tropical amount effect (e.g.,
Rozanski et al.,
1993
).
4.
Discussion
The regional coherence of multiple stalagmite δ
18
O deglaciation signals spaced throughout the maritime conti
-
nent of the IPWP (Figure
3
) suggests broad-scale atmospheric dynamics, not local site-specific climate changes,
drove the rainwater and stalagmite δ
18
O pattern. This interpretation is supported by the mean annual rainfall
response simulated in iTRACE, in which the broad area of the Sunda Shelf north of the equator is drier, while
the Indian Ocean south of the Sunda Shelf is wetter, in response to meltwater forcing and AMOC weakening
simulated in the model (Figure
3
, Figure S7 in the Supporting Information
S1
.
The iTRACE IPWP mean annual rainfall response to North Atlantic freshwater forcing and a weakened AMOC
is generally consistent with two other fully coupled general circulation climate model simulations: TraCE-21ka
(Liu et al.,
2009
; F. He,
2011
) and HadCM3 (Roberts & Hopcroft,
2020
) (Figure S8c in the Supporting Infor
-
mation
S1
). TraCE-21ka is a 21 ka-modern transient simulation of global climate using the Community Climate
System Model version 3 (CCSM3) (Collins et al.,
2006
). HadCM3 (Gordon et al.,
2000
; Pope et al.,
2000
) has
been extensively used for climate studies since the late 1990s, and performs reasonably well with respect to mean
climate (Valdes et al.,
2017
), particularly in the IPWP region under glacial conditions (DiNezio & Tierney,
2013
).
While the spatial rainfall anomaly pattern is consistent between all three simulations, the amplitude of the change
in mean annual rainfall during the Heinrich event is much greater in the CESM1 (iTRACE) versus the CCSM3
simulation (Figure S8c in the Supporting Information
S1
). The CESM1 mean annual rainfall response is similar
in amplitude to the HadCM3 0.25 Sv forcing simulation (Figure S8c in the Supporting Information
S1
). DiNezio
et al. (
2016
) previously noted the agreement of IPWP paleoclimate proxies with both the CESM1 and HadCM3
simulated steady-state LGM climate. This study extends this agreement between proxy and model to include
these two models' mean annual response to a North Atlantic freshwater forcing event.
The iTRACE simulated drying over Borneo during freshwater forcing events is largely caused by an anomalous
local anticyclone in this model, which shifts equatorward from the Philippine Sea over Borneo during the boreal
winter months when AMOC is weakened (Figure
4
). This anticyclone was recently recognized in the iTRACE
simulation in C. He et al. (
2021b
). The authors suggested that the anticyclone was yielded and persisted in boreal
autumn by anomalously low moist static air from the midlatitudes, which was advected by mean northeasterly
winds. The low moist static air results from an enhanced meridional sea surface temperature gradient in the
Western North Pacific which is associated with a weakened AMOC state. During the winter months, these mean
northeasterly winds strengthen, pushing the anticyclone southward over Borneo (Figure 4 in C. He et al.,
2021b
),
affecting both northern Borneo and nearby Palawan stalagmite locations in the model world (Figure
4
).
19448007, 2022, 5, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2021GL096937 by California Inst of Technology, Wiley Online Library on [06/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License