Salinity in the Indonesian seas integrates regional oceanographic and atmospheric processes, such as Indonesian Throughflow (ITF) and monsoon rainfall. Here we present a multicentury (1777–1983) δ18O coral record from Nightcliff Reef, located in the Timor Passage off the coast of northern Australia, which we use to infer local salinity change. We show that Australian monsoon rainfall and ITF influence salinity at the study site. These reconstructed salinity changes in the Timor Passage correlate with changes in Pacific sea surface temperature (SST) modes, including the El Niño Southern Oscillation (ENSO) and the Interdecadal Pacific Oscillation (IPO). While environmental stress creates challenging conditions for coral growth, this record particularly tracks the central Pacific signature of ENSO‐driven interannual variability, in agreement with reconstructions of rainfall across northern Australia. The strength of interannual variance in the record follows fluctuations in other local ENSO‐sensitive rainfall reconstructions, demonstrating a strong regional ENSO signature. However, this regional pattern differs from variance in composite ENSO reconstructions, suggesting that the multi‐site nature of these reconstructions may create biases. Salinity variability on decadal and longer time scales occurs throughout the record. Some of these oscillations are consistent with other ITF‐sensitive coral records. Our new salinity record adds a strongly Pacific‐sensitive record to the existing suite of regional paleoclimate reconstructions. Relationships among these records highlight the complexity of salinity in the Indonesian seas and the controls on its variability.
Coral records of surface‐ocean conditions extend our knowledge of interannual El Niño–Southern Oscillation (ENSO) variability into the preinstrumental period. That said, the wide range of natural variability within the climate system as well as multiple sources of uncertainties inherent to the coral archive produce challenges for the paleoclimate community to detect forced changes in ENSO using coral geochemical records. We present a new coral proxy system model (PSM) of intermediate complexity, geared toward the evaluation of changes in interannual variance. Our coral PSM adds additional layers of complexity to previously published transfer functions of sensor models that describe how the archive responds to sea surface temperature (SST) and salinity. We use SST and salinity output from the Community Earth System Model Last Millennium Ensemble 850 control to model coral oxygen isotopic ratios and SST derived from Sr/Ca. We present a detailed analysis of our PSM using climate model output for sites in the central and southwest Pacific before extending the analyses to span the broader tropical Pacific. We demonstrate how variable growth rates, analytical and calibration errors, and age model assumptions systematically impact estimates of interannual variance and show that the relative magnitude of the change in interannual variance is location dependent. Importantly, however, we find that even with the added uncertainties in our PSM, corals from many circum‐Pacific locations are broadly able to capture decadal and longer (decadal+) changes in ENSO variability. Our code is publicly available on GitHub to facilitate future comparisons between model output and coral proxy data.
more » « less- Award ID(s):
- 1805874
- NSF-PAR ID:
- 10456312
- Publisher / Repository:
- DOI PREFIX: 10.1029
- Date Published:
- Journal Name:
- Paleoceanography and Paleoclimatology
- Volume:
- 35
- Issue:
- 7
- ISSN:
- 2572-4517
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract -
more » « less
Abstract Natural and social systems worldwide are impacted by climate modes such as the El Niño/Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO) and Atlantic Multidecadal Oscillation (AMO), making it imperative to understand their sensitivity to climate change. Paleoclimate studies extend the observational climate baseline, and speleothem records (δ18Ospel) are a common data source. However, relationships between δ18Ospeland climate modes are uncertain; climate models provide a way to test the strength and stability of these relationships. Here, we use the isotope‐enabled Community Earth System Model's Last Millennium Ensemble combined with a forward proxy model to delineate the global expression of modal variability in “pseudo‐stalagmite” (δ18Ospel) records worldwide. The modeled δ18Ospelspatially correlates with modal signatures. However, substantial changes in modal variance only modestly affect individual δ18Ospelvariance. A network of δ18Ospelrecords, particularly one that straddles the Pacific, significantly improves the reconstructability of ENSO variance.
-
The El Nino Southern Oscillation (ENSO) is a semi-periodic fluctuation in sea surface temperature (SST) over the tropical central and eastern Pacific Ocean that influences interannual variability in regional hydrology across the world through long-range dependence or teleconnections. Recent research has demonstrated the value of Deep Learning (DL) methods for improving ENSO prediction as well as Complex Networks (CN) for understanding teleconnections. However, gaps in predictive understanding of ENSO-driven river flows include the black box nature of DL, the use of simple ENSO indices to describe a complex phenomenon and translating DL-based ENSO predictions to river flow predictions. Here we show that eXplainable DL (XDL) methods, based on saliency maps, can extract interpretable predictive information contained in global SST and discover novel SST information regions and dependence structures relevant for river flows which, in tandem with climate network constructions, enable improved predictive understanding. Our results reveal additional information content in global SST beyond ENSO indices, develop new understanding of how SSTs influence river flows, and generate improved river flow predictions with uncertainties. Observations, reanalysis data, and earth system model simulations are used to demonstrate the value of the XDL-CN based methods for future interannual and decadal scale climate projections.more » « less
-
more » « less
Abstract Climate variability has distinct spatial patterns with the strongest signal of sea surface temperature (SST) variance residing in the tropical Pacific. This interannual climate phenomenon, the El Niño-Southern Oscillation (ENSO), impacts weather patterns across the globe via atmospheric teleconnections. Pronounced SST variability, albeit of smaller amplitude, also exists in the other tropical basins as well as in the extratropical regions. To improve our physical understanding of internal climate variability across the global oceans, we here make the case for a conceptual model hierarchy that captures the essence of observed SST variability from subseasonal to decadal timescales. The building blocks consist of the classic stochastic climate model formulated by Klaus Hasselmann, a deterministic low-order model for ENSO variability, and the effect of the seasonal cycle on both of these models. This model hierarchy allows us to trace the impacts of seasonal processes on the statistics of observed and simulated climate variability. One of the important outcomes of ENSO’s interaction with the seasonal cycle is the generation of a frequency cascade leading to deterministic climate variability on a wide range of timescales, including the near-annual ENSO Combination Mode. Using the aforementioned building blocks, we arrive at a succinct conceptual model that delineates ENSO’s ubiquitous climate impacts and allows us to revisit ENSO’s observed statistical relationships with other coherent spatio-temporal patterns of climate variability—so called empirical
modes of variability . We demonstrate the importance of correctly accounting for different seasonal phasing in the linear growth/damping rates of different climate phenomena, as well as the seasonal phasing of ENSO teleconnections and of atmospheric noise forcings. We discuss how previously some of ENSO’s relationships with other modes of variability have been misinterpreted due to non-intuitive seasonal cycle effects on both power spectra and lead/lag correlations. Furthermore, it is evident that ENSO’s impacts on climate variability outside the tropical Pacific are oftentimes larger than previously recognized and that accurately accounting for them has important implications. For instance, it has been shown that improved seasonal prediction skill can be achieved in the Indian Ocean by fully accounting for ENSO’s seasonally modulated and temporally integrated remote impacts. These results move us to refocus our attention to the tropical Pacific for understanding global patterns of climate variability and their predictability. -
null (Ed.)Abstract Scientific understanding of low-frequency tropical Pacific variability, especially responses to perturbations in radiative forcing, suffers from short observational records, sparse proxy networks, and bias in model simulations. Here, we combine the strengths of proxies and models through coral-based paleoclimate data assimilation. We combine coral archives ( δ 18 O, Sr/Ca) with the dynamics, spatial teleconnections, and intervariable relationships of the CMIP5/PMIP3 Past1000 experiments using the Last Millennium Reanalysis data assimilation framework. This analysis creates skillful reconstructions of tropical Pacific temperatures over the observational era. However, during the period of intense volcanism in the early nineteenth century, southwestern Pacific corals produce El Niño–Southern Oscillation (ENSO) reconstructions that are of opposite sign from those from eastern Pacific corals and tree ring records. We systematically evaluate the source of this discrepancy using 1) single-proxy experiments, 2) varied proxy system models (PSMs), and 3) diverse covariance patterns from the Past1000 simulations. We find that individual proxy records and coral PSMs do not significantly contribute to the discrepancy. However, following major eruptions, the southwestern Pacific corals locally record more persistent cold anomalies than found in the Past1000 experiments and canonical ENSO teleconnections to the southwest Pacific strongly control the reconstruction response. Furthermore, using covariance patterns independent of ENSO yields reconstructions consistent with coral archives across the Pacific. These results show that model bias can strongly affect how proxy information is processed in paleoclimate data assimilation. As we illustrate here, model bias influences the magnitude and persistence of the response of the tropical Pacific to volcanic eruptions.more » « less
