Invited reviewThe Indian Ocean Zonal Mode over the past millennium in observed and modeled precipitation isotopes
Introduction
The Indian Ocean Zonal Mode (IOZM; also known as the Indian Ocean Dipole; Table 1) is an important mode of interannual rainfall variability in the circum-Indian Ocean region (Saji et al., 1999). IOZM events occur during boreal Fall (September–December; SOND) every 2–7 years, causing pronounced rainfall anomalies in East Africa, western Indonesia, Australia, and India (e.g., Ashok et al., 2001, Black et al., 2003). Under normal conditions, a strong gradient in sea surface temperature (SST) between the warmer Eastern Equatorial Indian Ocean (EEIO) and the cooler Western Equatorial Indian Ocean (WEIO) causes weak westerly winds along the equator, contributing to the normal Indian Ocean Walker Circulation, with rising air and deep atmospheric convection over Indonesia and descending air and dry conditions in East Africa. During IOZM positive (IOZM+) events, a breakdown of the normal SST gradient and equatorial wind field leads to anomalously dry conditions in western Indonesia while enhancing the October–December (OND) rainy season in East Africa (Fig. 1). The event ceases when the normal boreal summer monsoon circulation develops, cooling the WEIO and relaxing anomalous easterly winds (Webster et al., 1999, Schott and McCreary, 2001).
In addition to its direct effects on East African and Indonesian precipitation, the IOZM has the ability to effect large changes in regional hydrology through its connection with the Indian monsoon and the El Niño-Southern Oscillation (ENSO) (Ashok et al., 2001, D'Arrigo and Smerdon, 2008, Ummenhofer et al., 2011). Both an IOZM+ event and an El Niño event necessitate anomalous cooling in the Indo-Pacific warm pool and weakened vertical ascent and convection over Indonesia, and in some but not all years the two co-occur (Saji et al., 1999, Saji and Yamagata, 2003a), possibly triggered by ENSO-induced zonal shifts in Walker cell anomalies (Fischer et al., 2005). IOZM+ events can also be triggered by meridional Hadley cell perturbations in the absence of ENSO dynamics (Fischer et al., 2005), and numerous other studies have established that the IOZM's interannual fluctuations are distinct from ENSO and should be considered a separate Indian Ocean phenomenon (e.g., Saji and Yamagata, 2003b, Behera et al., 2006, Schott et al., 2009).
While the IOZM operates independently from ENSO, teleconnections to the Pacific Ocean can play an important role in triggering and enhancing IOZM events via local air-sea interactions, ocean dynamics, and atmospheric teleconnections (Li et al., 2003, Behera et al., 2006, Gnanaseelan and Vaid, 2010). Interannual to decadal variability in Pacific SSTs can induce stronger/more frequent IOZM+ events by preconditioning the EEIO with a shallower thermocline (Annamalai et al., 2005, Schott et al., 2009). This preconditioning is associated with changes in the Indonesian Throughflow as well as an atmospheric teleconnection to equatorial winds over the Indian Ocean (Annamalai et al., 2005). Air-sea interactions in the warm pool are also integral to the generation and termination of IOZM events (Cai et al., 2013 and refs therein). SST variability in the EEIO induces a positive feedback with wind speed and evaporative heat loss, further enhancing cold anomalies. However, cool SSTs also suppress clouds and convection, warming the EEIO via increased shortwave radiation (Cai et al., 2013).
In addition to pronounced interannual variability, the IOZM also exhibits low-frequency variability. The power spectrum of observed IOZM SSTs exhibits a spectral peak at ∼10 years (90% significance level), suggesting that periodicity at decadal and potentially longer timescales is intrinsic to the IOZM system (Ashok et al., 2004a, Ashok et al., 2003). Twentieth century SST observations reveal multi-decadal modulations of the IOZM, with periods of time characterized by more frequent/intense IOZM negative (IOZM-) events (∼1880–1920), more frequent/intense IOZM positive (IOZM+) events (∼1960–2000), and periods of lower overall activity (∼1920–1950) (Kripalani and Kumar, 2004, Ihara et al., 2008). Multi-decadal variations in the equatorial Indian Ocean windfield, SST, and thermocline depth are also apparent in observations and in modeling simulations (Saji and Yamagata, 2003b, Ashok et al., 2004a, Annamalai et al., 2005, Tozuka et al., 2007).
Low-frequency behavior makes the IOZM potentially important to and detectible in low-resolution paleoclimate proxy reconstructions. Indeed, paleoclimate studies have invoked “IOZM-like” dynamics to explain low-frequency patterns in regional paleoclimate from decadal up to orbital timescales (e.g., Stager et al., 2005, Griffiths et al., 2010, Gupta et al., 2010). These “IOZM-like” dynamics are not directly analogous to those of an individual IOZM event, whose onset and cessation are directly tied to a given year's seasonal cycle. However, a low-frequency, “IOZM-like” mode could arise from periods of more frequent/intense IOZM events (e.g., Tozuka et al., 2007), intrinsic low-frequency variability in Indian Ocean surface ocean characteristics (e.g., Annamalai et al., 2005), and/or feedbacks with the Indian monsoon circulation and the Indonesian Throughflow (Ashok et al., 2001, Annamalai et al., 2003, Annamalai et al., 2005).
Understanding low-frequency variations in the IOZM is critical for decadal-to century-scale climate prediction in the circum-Indian Ocean region. This is particularly true under 21st century global climate change scenarios, as a recent increase in IOZM intensity has been linked to increased greenhouse gas concentrations and warming (Cai et al., 2009) and increasingly strong positive feedbacks with the Asian monsoon circulation (Abram et al., 2008). However, future variability of the IOZM is uncertain due to uncertainties in ENSO and Asian monsoon teleconnections (Abram et al., 2008, Cai et al., 2009), which are strongly affected by interdecadal variability in the Pacific Ocean (Zhang et al., 1997, Ashok et al., 2004a). In some model simulations, changes in the mean state of the tropical Indian and Pacific Oceans are associated with increased frequency of IOZM+ events, particularly events occurring in consecutive years (Saji et al., 2006, Cai et al., 2009). However, IOZM event frequency may not increase in the 21st century, despite a mean state of the Indian Ocean that is projected to become more “IOZM+ -like” in its mean wind and SST structure (Cai et al., 2013).
Paleoclimate records can help to illuminate the interactions between IOZM intensity, low-frequency variability, and changes in mean state. On paleoclimate timescales, corals, lake level records, tree rings, and output from paleoclimate model simulations have all been interpreted to reflect IOZM variability on interannual to multidecadal time-scales (e.g., Stager et al., 2005, Zhao et al., 2005, Abram et al., 2008, Abram et al., 2007, D'Arrigo et al., 2008). However, recent paleoclimate proxy reconstructions and model simulations suggest that continental rainfall in the circum-Indian Ocean region has a nonstationary relationship with Indian Ocean SST variability on decadal and longer timescales (Zinke et al., 2009, Coats et al., 2013). The relationship between Indian Ocean regional rainfall and ENSO, as well as IOZM teleconnections, may also be nonstationary on these timescales (Ashok et al., 2001, Ashok et al., 2004b, Timm et al., 2005). Unstable behavior in the IOZM itself, and/or nonstationarity in the IOZM's relationship with rainfall, must be addressed in order to improve our understanding of the role of the IOZM in paleoclimate as well as its role in regulating regional precipitation variability in the future.
Oxygen and hydrogen isotopes of precipitation (δ18Oprecip and δDprecip) are an increasingly important proxy for modern and past climate processes in the tropics (e.g., Vuille et al., 2012, Conroy et al., 2013, Moerman et al., 2013). In the Indo-Pacific, changes in the Walker circulation associated with the IOZM alter rainfall as well as the O and H isotopic composition of rainfall in East Africa and Indonesia. These anomalies, observed in precipitation and precipitation isotopic data from stations distributed throughout East Africa and Indonesia, were reproduced by an isotope-enabled GCM simulation from 1950 to 1994 (Vuille et al., 2005a). The IOZM produces isotopic anomalies by altering the Walker circulation over the Indian Ocean, changing the rainout and distillation processes that accompany anomalous vertical ascent (descent), upper-level divergence (convergence), and increased (decreased) convection over East Africa (Indonesia) during the SOND season. Because both sides of the Walker cell are affected in an opposite manner, a distinctive east/west spatial pattern characterizes the precipitation isotopic response to the IOZM, with significant changes of opposite direction occurring in East Africa and Indonesia. IOZM+ events are associated with negative (18O-, D-depleted) anomalies in East Africa and positive (18O-, D-enriched) anomalies in western Indonesia.
An increasing number of high-resolution proxy records of δ18Oprecip and δDprecip spanning the past millennium (1000 C.E.-present) is becoming available from sediments, corals, and speleothems in the circum-Indian Ocean region. These records have suggested considerable variations in hydrology over the past millennium (Zinke et al., 2004, Newton et al., 2006, Fleitmann et al., 2007, Partin et al., 2007, Sinha et al., 2007, Abram et al., 2008, Griffiths et al., 2010, Tierney et al., 2010, Tierney et al., 2011, Yan et al., 2011, Konecky et al., 2013, Konecky et al., 2014). Many but not all of these records are characterized by pronounced multi-decadal variability (Zinke et al., 2004, Fleitmann et al., 2007, Konecky et al., 2013), and some show distinctive centennial-scale features or even millennium-long trends (Tierney et al., 2011, Yan et al., 2011, Konecky et al., 2013, Konecky et al., 2014). Changes in the Indian or Pacific Ocean Walker circulation have been proposed to explain O- and H-isotopic variations on all timescales, as well as interactions with the Indian, East Asian, and Australasian monsoon circulation. The Northern Hemisphere Medieval Climate Anomaly (MCA; ∼1000–1200 C.E.) and Little Ice Age (LIA; 1550–1800 C.E.) have received particular attention, with potential ramifications for solar forcing and Northern Hemisphere climate changes to affect regional atmospheric circulation. The timing of these multi-century events varies from site to site, and age model uncertainties often preclude a direct comparison of these variations to each other and to Northern Hemisphere temperature records (Masson-Delmotte et al., 2013). Nonetheless, records from both sides of the Indian Ocean suggest that the Indian Ocean Walker circulation may have responded strongly to external forcings on multi-decadal and longer timescales, leading to both precipitation and isotopic anomalies over the continents (Griffiths et al., 2010, Tierney et al., 2013). This highlights the importance of systematically testing IOZM relationships in isotopic proxy records, especially records that directly track δ18Oprecip/δDprecip.
In this study, we examine low-frequency behavior of the IOZM/δ18Oprecip (and IOZM/δDprecip) relationship over two important time periods: the historical period (1870–2003), and the past millennium (∼1000 C.E. to present). The historical period, for which a critical mass of SST observations is available, provides the longest and best-constrained timeframe to observe IOZM behavior, including several multi-decadal cycles. The past millennium provides an important test case to observe multi-decadal to centennial variability in IOZM behavior and to assess its regional impacts. These proxy records comprise the only network of δ18Oprecip/δDprecip-based proxy reconstructions from the core IOZM-affected regions of East Africa and Indonesia that are of high enough resolution to assess low-frequency timescales of variability over the past millennium. Together, these two timeframes provide important insights into low-frequency behavior of the IOZM in data and proxy records, setting the stage for future data/model comparisons with longer isotope-enabled paleoclimate simulations.
We describe our methods in Section 2. In Section 3, we examine multi-decadal variations in the IOZM/δ18Oprecip relationship during the historical period using an isotope-enabled Atmospheric General Circulation Model (AGCM) experiment forced with observed SSTs. This simulation from the Stable Water Isotope INtercomparison Group (SWING) enables us to test the stationarity of the precipitation and precipitation isotopic response to the IOZM on a multi-decadal timescale, and to investigate the potential local and remote climate relationships that may contribute to enhancing or diminishing the IOZM's signal through time. In Section 4, we assess the role of the IOZM in rainfall variations in the Indian Ocean region over the past millennium using a synthesis of ∼1 ka continental proxy reconstructions from East Africa and western Indonesia. We interpret these results in light of the SWING simulation, focusing especially on potential nonstationarities in the proxy response to the IOZM. In Section 5, we discuss the implications of our model and proxy synthesis results for studies of the IOZM, and for studies of Indian Ocean region climate over the past millennium.
Section snippets
Model and observational data
SWING Experiment S1B (http://atoc.colorado.edu/∼dcn/SWING/index.php) is an atmosphere-only GCM (AGCM) simulation forced with observed monthly SST from the HadISST v1.1 global dataset. Stable water isotope tracers are fully incorporated into the atmosphere, allowing isotopic fractionation to occur during water phase changes and to be advected through the atmospheric water cycle. The ECHAM4 model configuration used for SWING S1B is as described in Vuille et al. (2005a). Several AGCMs have
Simulation of observed IOZM-driven rainfall and δ18Oprecip variations with ECHAM4
Vuille et al. (2005a) demonstrated that ECHAM4 was able to reasonably simulate the SOND precipitation (P) and δ18Oprecip response to the IOZM in a shorter SWING experiment from 1950 to 1994. The longer SWING S1B experiment also provides a reasonable simulation of the seasonal cycle of precipitation and of interannual precipitation variability (not shown). The model simulates the P and δ18Oprecip response to the IOZM adequately, although it underestimates the spatial extent of the response in
Low-frequency IOZM behavior during the past millennium in precipitation isotopic proxy archives
Our results from the SWING simulation support previous observational and model evidence that the IOZM's spatial influence is broader in East African and SW Indonesian δ18Oprecip and δDprecip than in its imprint on precipitation amount alone (Vuille et al., 2005a). Given this broad regional influence on δ18Oprecip/δDprecip (Section 3.1 and Vuille et al., 2005a), it follows that a collection of spatially distributed proxy archives of δ18Oprecip/δDprecip from both poles of the IOZM may be
A data/model perspective on low-frequency IOZM variability
Both our model and our proxy synthesis results indicate the importance of distinguishing whether the IOZM itself varied over the past millennium, or whether its correlation with P and δ18Oprecip/δDprecip varied. The ECHAM4 SWING experiment shows that when the 134 years are considered as a whole, the ZWI is moderately but significantly (p < 0.05) correlated with both P and δ18Oprecip in East Africa and SW Indonesia. The IOZM influences δ18Oprecip across a broad region, extending beyond its
Conclusions
It is difficult, if not impossible, to assess IOZM-like variability in paleoclimate time series based on individual records, or based on East Africa–or Indonesia-specific groupings. Many climatic processes influence variability in P and δ18Oprecip/δDprecip in each respective region. SOND season precipitation anomalies are highly influential on interannual P and δ18Oprecip/δDprecip in both East Africa and in SW Indonesia. However, tropical δDwax and δ18Ocave proxies by nature integrate a
Acknowledgments
This project was supported by NOAA grant NOAA- NA090AR4310107 to J. Russell, NOAA grant NA09OAR4310090 and NSF AGS grant 1003690 to M. Vuille, and by an NSF Graduate Research Fellowship to B. Konecky. We thank three anonymous reviewers for helpful feedback on this manuscript.
References (96)
- et al.
Coupled dynamics over the Indian Ocean: spring initiation of the Zonal Mode
Deep-Sea Res. Pt II
(2003) - et al.
Holocene changes in eastern tropical Pacific climate inferred from a Galápagos lake sediment record
Quat. Sci. Rev.
(2008) - et al.
Isotopic reconstruction of the African Humid period and Congo air boundary migration at Lake Tana, Ethiopia
Quat. Sci. Rev.
(2014) - et al.
Holocene ITCZ and Indian monsoon dynamics recorded in stalagmites from Oman and Yemen (Socotra)
Quat. Sci. Rev.
(2007) - et al.
Evidence for Holocene changes in Australian-Indonesian monsoon rainfall from stalagmite trace element and stable isotope ratios
Earth Planet. Sci. Lett.
(2010) - et al.
Impact of monsoons, temperature, and CO2 on the rainfall and ecosystems of Mt. Kenya during the Common Era
Palaeogeogr. Palaeoclimatol. Palaeoecol.
(2014) - et al.
D/H variation in terrestrial lipids from Santa Barbara Basin over the past 1400years: a preliminary assessment of paleoclimatic relevance
Org. Geochem.
(2011) - et al.
Diurnal to interannual rainfall δ18O variations in northern Borneo driven by regional hydrology
Earth Planet. Sci. Lett.
(2013) - et al.
A severe drought during the last millennium in East Java, Indonesia
Quat. Sci. Rev.
(2013) - et al.
The climate of Socotra Island (Yemen): a first-time assessment of the timing of the monsoon wind reversal and its influence on precipitation and vegetation patterns
J. Arid Environ.
(2010)
The monsoon circulation of the Indian Ocean
Prog. Oceanogr.
Late Quaternary behavior of the East African monsoon and the importance of the Congo Air Boundary
Quat. Sci. Rev.
An isotopic and modelling study of flow paths and storage in Quaternary calcarenite, SW Australia: implications for speleothem paleoclimate records
Quat. Sci. Rev.
ENSO and Indian Ocean subtropical dipole variability is recorded in a coral record off southwest Madagascar for the period 1659 to 1995
Earth Planet. Sci. Lett.
Recent intensification of tropical climate variability in the Indian Ocean
Nat. Geosci.
Seasonal characteristics of the Indian Ocean Dipole during the Holocene epoch
Nature
Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature
Int. J. Climatol.
Identifying coherent spatiotemporal modes in time-uncertain proxy paleoclimate records
Clim. Dyn.
Effect of preconditioning on the extreme climate events in the tropical Indian Ocean
J. Clim.
Decadal variability of the Indian Ocean Dipole
Geophys. Res. Lett.
Individual and combined influences of ENSO and the Indian Ocean Dipole on the Indian summer monsoon
J. Clim.
Impact of the Indian Ocean Dipole on the relationship between the Indian monsoon rainfall and ENSO
Geophys. Res. Lett.
A look at the relationship between the ENSO and the Indian Ocean Dipole
J. Meteorol. Soc. Jpn.
Millennial-length forward models and pseudoproxies of stalagmite δ18O: an example from NW Scotland
Clim. Past Discuss.
A CGCM study on the interaction between IOD and ENSO
J. Clim.
Flexible paleoclimate age-depth models using an autoregressive gamma process
Bayesian Anal.
An observational study of the relationship between excessively strong short rains in coastal East Africa and Indian Ocean SST
Mon. Weather Rev.
Climate change contributes to more frequent consecutive positive Indian Ocean Dipole events
Geophys. Res. Lett.
Projected response of the Indian Ocean Dipole to greenhouse warming
Nat. Geosci.
Interdecadal variability of the relationship between the Indian Ocean Zonal Mode and East African coastal rainfall anomalies
J. Clim.
Stationarity of the tropical Pacific teleconnection to North America in CMIP5/PMIP3 model simulations
Geophys. Res. Lett.
El Nino/Southern oscillation and tropical Pacific climate during the last millennium
Nature
Highly variable El Nino-southern oscillation throughout the Holocene
Science
The twentieth century reanalysis project
Q. J. R. Meteorol. Soc.
Comparison of precipitation isotope variability across the tropical Pacific in observations and SWING2 model simulations
J. Geophys. Res-Atmos.
Pacific and Indian Ocean climate signals in a tree-ring record of Java monsoon drought
Int. J. Climatol.
Tropical climate influences on drought variability over Java, Indonesia
Geophys. Res. Lett.
Stable isotopes in precipitation
Tellus
The backbone of the climate network
Europhys. Lett.
Multiscale variabilities in global sea surface temperatures and their relationships with tropospheric climate patterns
J. Clim.
Pronounced interannual variability in tropical South Pacific temperatures during Heinrich Stadial 1
Nat. Commun.
Two independent triggers for the Indian Ocean Dipole/Zonal Mode in a coupled GCM
J. Clim.
Interannual variability in the Biannual Rossby waves in the tropical Indian Ocean and its relation to Indian Ocean Dipole and El Nino forcing
Ocean Dyn.
Increasing Australian-Indonesian monsoon rainfall linked to early Holocene sea-level rise
Nat. Geosci.
Mid-Brunhes strengthening of the Indian Ocean Dipole caused increased equatorial East African and decreased Australasian rainfall
Geophys. Res. Lett.
Mechanisms of climate anomalies in the equatorial Indian Ocean
J. Geophys. Res.
Atmospheric-hydrospheric mechanisms of climate anomalies in the western equatorial Indian Ocean
J. Geophys. Res.
Indonesian rainfall Variability: impacts of ENSO and local Air–Sea interaction
J. Clim.
Cited by (17)
Atmospheric controls on precipitation isotopes and hydroclimate in high-elevation regions in Eastern Africa since the Last Glacial Maximum
2021, Earth and Planetary Science LettersPalaeoclimate perspectives on the Indian Ocean Dipole
2020, Quaternary Science ReviewsStable isotope proxy records in tropical terrestrial environments
2020, Palaeogeography, Palaeoclimatology, PalaeoecologyCitation Excerpt :The Indo-Pacific Warm Pool (IPWP) represents a critical component of the global coupled ocean-atmosphere system, being the largest source of water vapour to the atmosphere, now and in the past (De Deckker et al., 2002; Niedermeyer et al., 2014). The IPWP interacts with the monsoons, the El Niño - Southern Oscillation (ENSO) to the east (e.g Osborne et al., 2014) and the Indian Ocean Zonal Mode (IOZM) to the west (e.g. Konecky et al., 2014), and with the extra-tropics through atmospheric teleconnections (Alexander et al., 2002). Complex interactions between these phenomena result in significant climate variability across the tropics on timescales ranging from inter-annual to decadal.
Oxygen isotope analyses of ungulate tooth enamel confirm low seasonality of rainfall contributed to the African Humid Period in Somalia
2019, Palaeogeography, Palaeoclimatology, PalaeoecologyHemispherically in-phase precipitation variability over the last 1700 years in a Madagascar speleothem record
2017, Quaternary Science ReviewsCitation Excerpt :Reduced Walker circulation would result in the El Niño like state observed in the regionally wet precipitation pattern in east Africa, but would require a decoupling of Indian and Pacific Walker circulation. Atmospheric modelling studies suggest that low-frequency zonal variability in the Indian Ocean is related to changes in the intensity or frequency of the IOD rather than the underlying mean state (Konecky et al., 2014; Tozuka et al., 2007). However, the IOD is a much more seasonally locked climate mode than ENSO, with maximum impact during the austral spring (SON).
Vanuatu environments from the Holocene to the present day: A state of knowledge
2016, Anthropologie (France)