Varied Response of Western Pacific Hydrology to Climate Forcings over the Last Glacial Period

See allHide authors and affiliations

Science  28 Jun 2013:
Vol. 340, Issue 6140, pp. 1564-1566
DOI: 10.1126/science.1233797

Borneo Paleohydrology

Climate records of the last glacial cycle provide a good picture of how climate changed at high and middle latitudes, but fewer records of the tropics are available. Carolin et al. (p. 1564, published online 6 June) present data from a suite of precisely dated stalagmites from Borneo that reveal how the western tropical Pacific region behaved between 100,000 and 15,000 years ago, a period during which many abrupt climate changes occurred in other parts of the world. While the hydroclimate of Borneo changed in response to precessional forcing, it responded only weakly to the forces that produced glacial-interglacial changes in global climate.


Atmospheric deep convection in the west Pacific plays a key role in the global heat and moisture budgets, yet its response to orbital and abrupt climate change events is poorly resolved. Here, we present four absolutely dated, overlapping stalagmite oxygen isotopic records from northern Borneo that span most of the last glacial cycle. The records suggest that northern Borneo’s hydroclimate shifted in phase with precessional forcing but was only weakly affected by glacial-interglacial changes in global climate boundary conditions. Regional convection likely decreased during Heinrich events, but other Northern Hemisphere abrupt climate change events are notably absent. The new records suggest that the deep tropical Pacific hydroclimate variability may have played an important role in shaping the global response to the largest abrupt climate change events.

The response of the tropical Pacific to changes in Earth’s climate system remains highly uncertain. The most recent glacial-interglacial cycle encompasses several precessional cycles; changes in ice volume, sea level, global temperature, and atmospheric partial pressure of CO2; and millennial-scale climate events, thus providing insights into the tropical Pacific response to a variety of climate forcings. Chinese stalagmites show that East Asian monsoon strength closely tracks precessional insolation forcing over several glacial-interglacial cycles and exhibits prominent millennial-scale variability (1, 2). The timing and structure of these abrupt climate changes are nearly identical to millennial-scale events recorded in the Greenland ice cores [Dansgaard-Oeschger (D/O) events] (3) and in sediment records that document ice-rafted debris across the North Atlantic (Heinrich events) (4, 5). A Borneo stalagmite record spanning the past 27,000 years provides a markedly different view of hydrology in the western tropical Pacific, with the Heinrich 1 excursion and spring-fall precessional insolation forcing explaining much of the variability (6). At its most basic, this finding illustrates the complexity of regional responses to various climate forcings, especially at sites located far from the North Atlantic, and demands a more exhaustive tropical Pacific hydrologic record encompassing a full glacial-interglacial cycle.

Here, we present four overlapping stalagmite oxygen isotopic (δ18O) records from Gunung Buda and Gunung Mulu national parks, located in northern Borneo (4°N, 115°E) (fig. S1), that together span most of the last glacial cycle. The research site is located near the center of the west Pacific warm pool (WPWP), where changes in sea surface temperatures and sea-level pressure have considerable impacts on large-scale atmospheric circulation and global hydrology (7). Using multiple stalagmites from different caves, we distinguish shared climate-related features from cave-specific signals in the overlapping δ18O records.

The four stalagmite records span portions of the last glacial cycle with many intervals of overlap, based on U-series dates (Fig. 1). Stalagmites were recovered from Secret Cave at Gunung Mulu [SC02, 37 to 94 thousand years before the present (ky B.P.), and SC03, 32 to 100 ky B.P.] and from Bukit Assam (BA02, 15 to 46 ky B.P.) and Snail Shell Cave (SCH02, 31 to 73 ky B.P.) at Gunung Buda, 25 km from Gunung Mulu (fig. S2). The deglacial and Holocene δ18O records from stalagmite SCH02 were presented in (6). Eighty-six new U/Th dates measured across the four stalagmites fall in stratigraphic order within 2σ errors (8). Large uncertainties in the 230Th/232Th ratio of the contaminant phases translate into large uncertainties associated with the correction for detrital thorium contamination. Fourteen isochrons measured across stalagmites from three separate caves give initial 230Th/232Th atomic ratios of 56 ± 11 (2σ) for Bukit Assam Cave, 59 ± 13 (2σ) for Snail Shell Cave, and 111 ± 41 (2σ) × 10−6 for Secret Cave (8), which fall within the range of previously published values from our site (6). Absolute age errors for each U/Th date were calculated with a Monte Carlo approach that combined multiple sources of error. The resulting dating errors average ±200, ±250, ±400, and ±500 years (2σ) for BA02, SCH02, SC02, and SC03, respectively. Age models were initially constructed by linearly interpolating between each date and were refined by aligning five major millennial-scale δ18O excursions visible across all four records within age error (8). The fact that both chronologies fall nearly completely within the StalAge (9) algorithm’s 95% confidence interval (figs. S3 to S6) adds credibility to our assigned chronologies and associated error estimates. With our 1-mm sampling interval, the temporal resolution of the associated δ18O records averages 60 years per sample for the faster-growing stalagmites BA02 and SCH02 and 200 years per sample for the slower-growing stalagmites SC02 and SC03. During the 50– to 38–ky B.P. interval, SC02 and SC03 were sampled at 0.5-mm to achieve a resolution of ~100 years per sample. Ultraslow growth intervals (<10 μm/year for the faster-growing stalagmites and <3 μm/year for the slower growing stalagmites) may represent unresolved hiatuses and, as such, were excluded from the resulting paleoclimate reconstructions (8), following (6).

Fig. 1 Comparison of four overlapping stalagmite δ18O records from northern Borneo.

(A) δ18O records from SC02 (blue), SC03 (red), SCH02 (green), and BA02 (purple) are overlain after aligning five major millennial-scale δ18O excursions shared across all four stalagmites to within 2σ dating errors (8), plotted with previously published stalagmite δ18O data from our site (black) (6). SC03 and SC02 mean δ18O have been offset +0.2 per mil (‰), and BA02 mean δ18O has been offset –0.45‰ to match the absolute value of SCH02, consistent with the prior use of SCH02 as a benchmark for the deglacial–Holocene Borneo records (6). (B) The δ18O record for SC02, plotted using its raw age model (blue), shown with the three other overlapping Borneo stalagmite δ18O records using their raw age models (gray). (C) Same as (B), but for SC03 (red). (D) Same as (B), but for SCH02 (green). (E) Same as (B), but for BA02 (purple). The U-Th–based age model was used to construct the aligned δ18O record plotted in corresponding colors at the top, shown with 2σ uncertainty limits (8). The x axis indicates age in thousand years before the present (kybp).

The stalagmite δ18O records provide reconstructions of rainfall δ18O variability at the research site, which, in turn, tracks the strength of regional convective activity (10). Consistent with the tropical amount effect (11, 12), rainfall δ18O variations measured at the site from 2006 to 2011 are significantly anticorrelated with regional precipitation amount and closely track the El Niño–Southern Oscillation on monthly time scales (10). A weak semi-annual seasonal cycle in rainfall δ18O is characterized by relative minima in June to July and November to January and relative maxima in February to April and August to October. Such a pattern suggests that the twice-yearly passage of the Intertropical Convergence Zone (ITCZ) over the site is associated with shifts in the moisture sources and/or trajectories that drive the observed seasonal fractionations (10). Dripwater δ18O values match rainfall δ18O values averaged over the preceding 2 to 6 months (13), suggesting a short residence time of dripwater δ18O relative to our centennial-scale sampling of stalagmite δ18O. Time series of Buda and Mulu stalagmite δ18O are highly reproducible (6, 14), strongly supporting their interpretation as rainfall δ18O reconstructions and, by extension, as records of past regional convective activity.

The overlapping Borneo stalagmite δ18O records show orbital-scale variability related to precessional insolation forcing and glacial-interglacial (G-I) changes (Fig. 2). The similarity of our stalagmite δ18O time series to indices of G-I variability greatly diminishes after removing the mean δ18O of seawater due to changes in ice volume (8, 15) from the Borneo δ18O records (Fig. 2 and fig. S7). After this correction, Last Glacial Maximum (LGM) δ18O values are nearly identical to δ18O values at ~85 ky B.P., despite the presence of substantially larger ice sheets, cooler regional temperatures (16, 17), and a completely exposed Sunda Shelf during the LGM. In particular, Sunda Shelf emergence has been implicated in shaping glacial western tropical Pacific hydroclimate in previous studies (6, 18, 19). However, we find little correspondence between Borneo stalagmite δ18O and an index of Sunda Shelf areal extent over the entire glacial cycle (fig. S7). For example, Borneo stalagmite δ18O variations in the 70– to 90–ky B.P. interval bear little resemblance to reconstructed sea-level changes, especially from ~71 to 76 ky B.P. (20), when a large drop in sea level almost doubled the size of the exposed shelf (fig. S7). As such, the new Borneo δ18O records suggest that the cumulative influence of G-I boundary conditions, including changes in global temperature and CO2, did not drive considerable changes in rainfall δ18O at our site. However, given the complexities of influences on rainfall δ18O (10), LGM climate may have been characterized by two or more competing influences on regional rainfall δ18O. For example, regional drying during the LGM inferred from WPWP sediment cores (21) and modeling studies (19) may have increased rainfall δ18O, whereas longer moisture trajectories associated with the emergence of the Sunda Shelf may have decreased rainfall δ18O.

Fig. 2 Comparison of Borneo stalagmite δ18O records to climate forcings and records of paleoclimate from key regions.

(A) Greenland NGRIP (North Greenland Ice Core Project) ice core δ18O (gray) (31) with 100-year averages (black), plotted using the GICC05modelext age model (32). (B) Hulu–Sanbao cave stalagmite δ18O records from China (1, 2) (Sanbao has been offset by +1.6‰ to match Hulu), plotted with July insolation at 65°N (33). (C) Borneo stalagmite δ18O records, plotted with age models aligned and adjusted to account for ice-volume–related changes in global seawater δ18O (8). Also plotted are October insolation at 0°N (black) (33) and Borneo stalagmite δ18O records (gray) that have not been corrected for ice volume. (D) Coral-based estimates of paleo–sea-level record (20, 34, 35) (black symbols) and global mean sea-level record (15) (solid line, average; dotted lines, minimum and maximum). (E) Sulu Sea planktonic foraminifera δ18O (24), plotted with a revised age model using updated IntCal09 calibration curve 41–ky B.P. modern and aligning 60–ky B.P. δ18O excursion to the Hulu-Sanbao stalagmite δ18O records. (F) EPICA (European Project for Ice Coring in Antarctica) Dronning Maud Land (EDML) ice core δ18O (gray) (28) with 7-year averages (black). Vertical blue bars indicate the timing of Heinrich events H1 to H6 (5), as recorded by the Hulu-Sanbao stalagmite δ18O records (1, 2).

The Borneo stalagmite δ18O records vary in phase with insolation at the equator during boreal fall in stage 5 and the Holocene, when precessional forcing is relatively strong (Fig. 2C). The impact of precessional forcing on Borneo stalagmite δ18O is weak during stage 3, in part owing to reduced precessional amplitude during this time. Precessional forcing is also apparent in older glacial-interglacial stalagmite δ18O reconstructions from Borneo (14). Taken together, the Borneo records suggest that precession may be the dominant source of orbital-scale hydroclimate variability in the WPWP. The implied sensitivity of northern Borneo hydrology to boreal fall insolation is consistent with results from a previous modeling study (22). Moreover, results from a long-term rainfall δ18O monitoring program at Mulu demonstrate that mean annual rainfall δ18O values depend, in part, on the magnitude of rainfall δ18O enrichments during the boreal spring-fall seasons (10). In this sense, the observed sensitivity to boreal fall insolation may represent a direct response of mean annual rainfall δ18O to local changes in seasonal moisture sources and trajectories. However, El Niño–Southern Oscillation and the Madden-Julian Oscillations (23) have large impacts on modern Mulu rainfall δ18O variability (10), such that Borneo stalagmite δ18O signals may represent a combination of one or more climatic influences.

The Borneo stalagmite δ18O records are dominated by six millennial-scale increases in δ18O that coincide with Heinrich events, inferring a decrease in regional convection during these abrupt climate changes (Fig. 2). A nearby Sulu Sea sediment core (Fig. 2E) also documents increased planktonic foraminiferal δ18O values during Heinrich events (24), consistent with a reduction in regional convective activity. The dominant paradigm to explain millennial-scale tropical hydroclimate anomalies is that they are driven from the North Atlantic region, either from weakening of the Atlantic thermohaline circulation or from a dramatic albedo change due to sea-ice cover, both of which drive a southward migration of the ITCZ that dries most of the northern tropics (25, 26). A similar chain of events is used to describe D/O abrupt climate changes that are well documented outside of the tropical Pacific, most notably in Chinese and Peruvian stalagmite δ18O records (1, 2, 27) and in a high-resolution ice core δ18O record from the south Atlantic sector of Antarctica (28). However, the Borneo stalagmite δ18O records lack any coherent signature of D/O events (Fig. 2 and fig. S8). The Borneo stalagmite δ18O records show no consistent response to D/O events 8 and 12, the prominent D/O events that occur on the heels of Heinrich events 4 and 5 (fig. S8). Of particular note, the records show little millennial-scale variability from ~30 to 40 ky B.P. across D/O events 5 to 8 (fig. S8). The records do bear a strong resemblance to the Chinese δ18O records during the 50– to 60–ky B.P. interval, as both records contain a distinct δ18O increase at ~55 ky B.P. This shared δ18O enrichment may reflect the influence of an additional Heinrich event, referred to as “Heinrich 5a” in one study (29), or may indicate a regional hydrological sensitivity to the relatively prolonged D/O events that occurred during this time interval. Contrary to inferences drawn from a deglacial Borneo stalagmite δ18O record (6), there is no evidence for a Southern Hemisphere influence on millennial-scale variability in Borneo hydroclimate over the last glacial cycle (fig. S8).

The unambiguous signature of Heinrich events in the Borneo stalagmite δ18O records stands in stark contrast to the lack of consistent D/O-related signals in the records, implying a selective response of WPWP hydrology to high-latitude abrupt climate change forcing. Specifically, the absence of any readily identifiable D/O signals in the Borneo δ18O record represents a clear challenge to our understanding of abrupt climate change mechanisms. The new Borneo records suggest that one of two possibilities must be true: (i) If D/O events reflect a similar mechanism to Heinrich events, then they must not be strong enough to affect northern Borneo hydrology appreciably, or (ii) D/O events and Heinrich events are characterized by fundamentally different climate mechanisms and feedbacks.

The largest millennial-scale anomaly in the Borneo records is not a Heinrich event, but rather an abrupt increase in δ18O that occurs at 73.42 ± 0.30 (2σ) ky B.P., coincident with a similarly large and abrupt increase in Chinese stalagmite δ18O (Fig. 2). Whether this event is associated with the Toba supereruption, dated at 73.88 ± 0.64 (2σ) ky B.P. (30), and/or a prominent early abrupt climate change event visible in Greenland ice core δ18O (Fig. 2A) merits investigation in additional high-resolution paleoclimate records from the Indo-Pacific.

The Borneo composite records demonstrate the sensitivity of western equatorial Pacific hydrology to both high-latitude and low-latitude forcings. However, the response of northern Borneo hydroclimate to these forcings is not uniform: Glacial conditions and D/O events apparently had much smaller impacts on regional hydrology than either insolation or Heinrich-related forcing. Our results imply that once the hydrological response threshold is reached, then climate feedbacks internal to the tropics may serve to amplify and prolong a given climate change event, whether the trigger originates from internal dynamics or external radiative forcing.

Supplementary Materials

Materials and Methods

Figs. S1 to S12

Tables S1 to S4

References (3638)

References and Notes

  1. Materials and methods are available as supplementary materials on Science Online.
  2. Acknowledgments: We thank N. Meckler, J. Partin, and S. Clark (Gunung Mulu National Park) for field assistance; J. Partin for assistance in sample analysis; G. Paris, M. Raven, S. Hines, and A. Subhas for assistance in U-Th dating; and J. Lynch-Stieglitz for providing comments on early versions of the manuscript. S.A.C, K.M.C., and J.F.A. were involved in the writing and design of this study; A.A.T. and B.C. facilitated the fieldwork for this study; S.A.C, K.M.C., S.L., and J.M. collected samples; and S.A.C. analyzed the samples. The research was funded by NSF PECASE Award no. 0645291 to K.M.C., NSF AGS award no. 0903099 to J.F.A., and a NSF Graduate Research Fellowship to S.A.C. Permits for this work were granted by the Malaysian Economic Planning Unit, the Sarawak State Planning Unit, and the Sarawak Forestry Department. All data reported in this paper are archived at the National Climatic Data Center (
View Abstract

Stay Connected to Science

Navigate This Article