Editorial Type:
Article
Article Type:
Research Article

Wind and Heat Forcings of the Seasonal and Interannual Sea Level Variabilities in the Southwest Pacific

Lina Yang aLaboratory for Coastal Ocean Variation and Disaster Prediction, College of Ocean and Meteorology, Guangdong Ocean University, Zhanjiang, China
bKey Laboratory of Climate, Resources and Environment in Continental Shelf Sea and Deep Sea of Department of Education of Guangdong Province, Zhanjiang, China
cKey Laboratory of Space Ocean Remote Sensing and Application, Ministry of Natural Resources, Beijing, China

Search for other papers by Lina Yang in
Current site
Google Scholar
PubMed Close
https://orcid.org/0000-0002-8253-0245
,
Xinyang Zhao aLaboratory for Coastal Ocean Variation and Disaster Prediction, College of Ocean and Meteorology, Guangdong Ocean University, Zhanjiang, China

Search for other papers by Xinyang Zhao in
Current site
Google Scholar
PubMed Close
,
Peng Liang aLaboratory for Coastal Ocean Variation and Disaster Prediction, College of Ocean and Meteorology, Guangdong Ocean University, Zhanjiang, China
bKey Laboratory of Climate, Resources and Environment in Continental Shelf Sea and Deep Sea of Department of Education of Guangdong Province, Zhanjiang, China
cKey Laboratory of Space Ocean Remote Sensing and Application, Ministry of Natural Resources, Beijing, China

Search for other papers by Peng Liang in
Current site
Google Scholar
PubMed Close
,
Tianyu Zhang aLaboratory for Coastal Ocean Variation and Disaster Prediction, College of Ocean and Meteorology, Guangdong Ocean University, Zhanjiang, China
bKey Laboratory of Climate, Resources and Environment in Continental Shelf Sea and Deep Sea of Department of Education of Guangdong Province, Zhanjiang, China
cKey Laboratory of Space Ocean Remote Sensing and Application, Ministry of Natural Resources, Beijing, China

Search for other papers by Tianyu Zhang in
Current site
Google Scholar
PubMed Close
,
Lingling Xie aLaboratory for Coastal Ocean Variation and Disaster Prediction, College of Ocean and Meteorology, Guangdong Ocean University, Zhanjiang, China
bKey Laboratory of Climate, Resources and Environment in Continental Shelf Sea and Deep Sea of Department of Education of Guangdong Province, Zhanjiang, China
cKey Laboratory of Space Ocean Remote Sensing and Application, Ministry of Natural Resources, Beijing, China

Search for other papers by Lingling Xie in
Current site
Google Scholar
PubMed Close
, and
Raghu Murtugudde dEarth System Science Interdisciplinary Center, University of Maryland, College Park, College Park, Maryland
eIndian Institute of Technology Bombay, Mumbai, India

Search for other papers by Raghu Murtugudde in
Current site
Google Scholar
PubMed Close
Online Publication:
06 Sep 2023
Print Publication:
01 Sep 2023
DOI:
https://doi.org/10.1175/JPO-D-23-0018.1
Page(s):
2171–2187
Free access

Abstract

Sea level variabilities in the southwest Pacific contribute to the variations of equatorial current bifurcation and the Indonesian Throughflow transport. These processes are closely related to the recharge/discharge of equatorial heat content and dynamic distribution of anthropogenic ocean heating over the Indo-Pacific basin, thus being of profound significance for climate variability and change. Here we identify the major features of seasonal and interannual sea level variabilities in this region, confirming the dominance of the first baroclinic mode in the tropics (contributing 60%–80% of the variances) and higher baroclinic modes in the extratropics (40%–60% of the seasonal variance). Seasonally, except in the western Coral Sea where the Ekman pumping is significant, the wind-driven first-mode baroclinic Rossby waves originating to the east of the date line control the sea level variations over most tropical Pacific regions. In the domain where the 1.5-layer reduced gravity model becomes deficient, the surface heat fluxes dominate, explaining ∼40%–80% of sea level variance. For interannual variability, ∼40%–60% of the variance are El Niño–Southern Oscillation (ENSO) related. The wind-driven Rossby and Kelvin waves east of the date line explain ∼40%–78% of the interannual variance in the tropical Pacific. Outside the tropics, small-scale diffusive processes are presumed critical for interannual variability according to a thermodynamic analysis using an eddy-permitting ocean model simulation. Further process and predictive understandings can be achieved with the coupled climate models properly parameterizing the subgrid-scale processes.

© 2023 American Meteorological Society. This published article is licensed under the terms of the default AMS reuse license. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Lina Yang, ylina_001@sina.com

Abstract

Sea level variabilities in the southwest Pacific contribute to the variations of equatorial current bifurcation and the Indonesian Throughflow transport. These processes are closely related to the recharge/discharge of equatorial heat content and dynamic distribution of anthropogenic ocean heating over the Indo-Pacific basin, thus being of profound significance for climate variability and change. Here we identify the major features of seasonal and interannual sea level variabilities in this region, confirming the dominance of the first baroclinic mode in the tropics (contributing 60%–80% of the variances) and higher baroclinic modes in the extratropics (40%–60% of the seasonal variance). Seasonally, except in the western Coral Sea where the Ekman pumping is significant, the wind-driven first-mode baroclinic Rossby waves originating to the east of the date line control the sea level variations over most tropical Pacific regions. In the domain where the 1.5-layer reduced gravity model becomes deficient, the surface heat fluxes dominate, explaining ∼40%–80% of sea level variance. For interannual variability, ∼40%–60% of the variance are El Niño–Southern Oscillation (ENSO) related. The wind-driven Rossby and Kelvin waves east of the date line explain ∼40%–78% of the interannual variance in the tropical Pacific. Outside the tropics, small-scale diffusive processes are presumed critical for interannual variability according to a thermodynamic analysis using an eddy-permitting ocean model simulation. Further process and predictive understandings can be achieved with the coupled climate models properly parameterizing the subgrid-scale processes.

© 2023 American Meteorological Society. This published article is licensed under the terms of the default AMS reuse license. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Lina Yang, ylina_001@sina.com
Keywords: South Pacific Ocean; Ocean dynamics; Planetary waves; Sea level; Interannual variability; Seasonal variability

1. Introduction

Sea level changes are a good indicator of the upper-ocean heat content and geostrophic circulation variations (Qiu and Joyce 1992; Piecuch and Ponte 2011; Albrecht et al. 2019) and thus are closely related to climate evolution in the surrounding regions. To date, there have been many studies concerning global or regional sea level variabilities at different time scales, from intraseasonal variability to long-term trends, using various kinds of products (satellite altimetry, ocean reanalyses, model output, etc.) (e.g., Qiu and Joyce 1992; Vivier et al. 1999; Ho et al. 2000; Bowen et al. 2006; Zhuang et al. 2010; Piecuch and Ponte 2011; Zhang and Qu 2015; Albrecht et al. 2019; Duan et al. 2020; Wang et al. 2021). Mostly, steric sea level, particularly the thermosteric component, plays a dominant role in the tropical and subtropical oceans (Chen et al. 2000; Roemmich et al. 2007; Piecuch and Ponte 2011; Roberts et al. 2016; Wu et al. 2017). The barotropic response to wind forcing becomes pivotal to sea level variability in regions of high latitudes or shallow water depths (Vivier et al. 2005; Zhuang et al. 2010).

The specific process governing sea level variability depends on the region. One influential study prior to the extensive altimetric data, showed that winds can explain most of the interannual variability of sea levels in the deep tropics using a linear ocean model (Busalacchi et al. 1983). Using the sea surface height (SSH) data retrieved in the early years (October 1992–December 1996) of the altimetry era and an analytical wave model, the wind-forced long Rossby waves were estimated to explain more than 70% of the SSH variance within 10° of the equator but only ∼30% between 10° and 30°N (Vivier et al. 1999). Recently, the annual cycle of sea level anomalies (SLAs) at some latitudes of the tropical Pacific off the equatorial band has been effectively simulated by a linear 1.5-layer reduced-gravity model, highlighting the role of the first-mode baroclinic, linear Rossby waves (Liu and Zhou 2020; Yang et al. 2022). At midlatitudes, Stammer (1997) emphasized the dominance of seasonal sea surface heating and cooling. In contrast, Piecuch and Ponte (2011) concluded that the interannual steric sea level variations in most global oceans are caused by oceanic advection and diffusion. At longer time scales, i.e., decadal scales, the wind-forced baroclinic long Rossby waves seem to become important again in the midlatitude regions. For example, the major features of decadal trends of South Pacific sea levels between 30° and 60°S were reproduced well by the first-mode baroclinic, linear Rossby wave model (Qiu and Chen 2006). It was found that the wind stress curl anomalies related to the southern annular mode were crucial to the decadal fluctuation of sea levels in the subtropical southwest Pacific (Qiu and Chen 2006; Zhang and Qu 2015; Sun et al. 2022). Thus, the regional sea level variabilities of different time scales and the underlying mechanisms vary spatially and temporally, deserving a thorough investigation.

El Niño–Southern Oscillation (ENSO) is the leading mode of global interannual climate variability with a global reach, which shows intimate relations with interannual sea level variability in the low-latitude regions (Ren et al. 2020). In the tropical North Pacific along the meridian of 137°E, for instance, the surface dynamic height shows a coherent drop during the ENSO years, accompanied by a meridional shift of the positions of trough and ridge as well as the associated ocean currents (Qiu and Joyce 1992). In the tropical northwest Pacific around Palau Island, the sea level reaches a minimum about two months after the maximum in Niño-3.4 (Andres et al. 2020) (the Niño-3.4 index is defined in section 2). In addition, a recent study argued for the impacts from an independent Indian Ocean dipole (IOD) event (without an accompanying El Niño or La Niña) on the interannual sea level variability in the western equatorial Pacific and the northwestern tropical Pacific (Duan et al. 2020). They found that a positive IOD caused negative SLAs (4–7 cm) during the IOD year but positive anomalies (6–8 cm) in the following year by modulating the tropical Pacific winds. However, the drivers in the South Pacific are still poorly understood due to a primary focus of the previous studies on the Northern Hemisphere.

In this study, we investigate the seasonal and interannual sea level variabilities and underlying mechanisms in the southwest Pacific between 35°–5°S and 140°E–180°, where numerous island communities have been suffering from the threat of extreme sea levels. It was reported that the sea levels there have shown a linear rise of ∼8 mm yr−1 during the recent two decades, corresponding to a spinup of the subtropical gyre circulation and the interdecadal Pacific oscillation (Roemmich et al. 2007; Albrecht et al. 2019; Sun et al. 2022). In this region, the Pacific South Equatorial Current, whose variability is dominated by sea level differences between certain latitudes (Yang et al. 2022), bifurcates off the Australian coast into the northward flowing North Queensland Current and the southward flowing East Australia Current, with the bifurcation latitude moving from ∼15°S near the surface to ∼22°S at the 800-m depth (Qu and Lindstrom 2002). The northern branch, i.e., the North Queensland Current, serves as an essential source of the Indonesian Throughflow (ITF), through which the Pacific and Indian Oceans exchange mass, heat, and anomalous climate signals (e.g., Gordon 2005; Yuan et al. 2011; Yang et al. 2018). Besides, the subsurface temperature anomalies in the western tropical South Pacific may propagate to the equator through the western boundary currents and then turn eastward along the equator, leading to the regime shift of the ENSO-like decadal variability in the tropical Pacific (Luo and Yamagata 2001). Thus the present study is of significance to deepening our knowledge of ITF variability as well as climate variability and change over the Indo-Pacific, which in turn have known impacts on the global climate variability and change. The local impacts on the island states in the region make this study even more timely as a first step toward a focus on region-specific predictions and projections.

The rest of this paper is organized as follows. The data and methods are introduced in section 2. The major features of seasonal and interannual sea level variabilities in the southwest Pacific and the independent role of ENSO and IOD are described in section 3. Section 4 diagnoses the driving mechanisms, quantifying the barotropic/baroclinic and thermosteric/halosteric contributions and exploring the roles of the first-mode baroclinic Rossby waves, Ekman pumping, and nonlinear advection and diffusion. The summary and discussion are displayed in section 5.

2. Data and methods

a. Data

1) Sea level observations

The monthly SSH data from January 1993 to December 2018 are averaged from the daily delayed-time global sea level map generated by the Data Unification and Altimeter Combination System (DUACS), using an optimal interpolation after cross-calibrating altimeter measurements from multiple missions. This sea level map has a resolution of 0.25° × 0.25°, delivered by the Copernicus Marine Environment and Monitoring Service (CMEMS) (https://data.marine.copernicus.eu/products). SLAs referenced to the mean over 1993–2018 are calculated to analyze the seasonal and interannual sea level variabilities in the southwest Pacific. The mass-induced SLAs (referenced to the mean over 2004–09), i.e., the barotropic variations, from January 2003 to December 2018, are derived from ocean bottom pressure observations by Gravity Recovery and Climate Experiment (GRACE) using ηbt′=pbot′/(ρ0g), in which the mean seawater density (ρ0) is set to 1025.0 kg m−3 (https://www2.csr.utexas.edu/grace/RL06_mascons.html). The data of ηbt′ are regridded onto a 0.25° × 0.25° grid from the native resolution of roughly 300 km.

2) Climate indices and Argo temperatures and salinities

The Niño-3.4 index is the sea surface temperature anomaly averaged over the central-east equatorial Pacific (5°S–5°N, 170°–120°W). The dipole mode index (DMI) is the difference between the area-mean sea surface temperature anomalies in the western (10°S–10°N, 50°–70°E) and eastern (10°S–0°, 90°–110°E) tropical Indian Ocean. In this study, we use the monthly Niño-3.4 and DMI over 1958–2018, constructed by the National Oceanic and Atmospheric Administration/Earth System Research Laboratories (NOAA/ESRL) using the Hadley Centre Sea Ice and Sea Surface Temperature (https://psl.noaa.gov/gcos_wgsp/Timeseries/), to determine the ENSO and IOD years, which are selected as the indices exceeding one standard deviation (listed in Table 1). In addition, Argo temperature and salinity profiles with a spatial resolution of 1° × 1° over January 2004–December 2018 (Roemmich and Gilson 2009; https://argo.ucsd.edu/data/argo-dataproducts) are utilized to evaluate the contributions of thermosteric and halosteric sea level variations and to verify reanalysis data for vertical mode decomposition of pressure anomalies.

Table 1.

List of El Niño, La Niña, positive IOD, and negative IOD years considered in this study.
Table 1.

3) Surface wind and bottom topography

The wind speed at 10-m height spanning from January 1993 to December 2018 at a resolution of 0.25°, used to force the 1.5-layer reduced-gravity model, are obtained from the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim, https://www.ecmwf.int/en/forecasts/datasets). The Earth Topography and Bathymetry data (ETOPO5) with a resolution of 1/12° (https://www.eea.europa.eu/data-and-maps/data/world-digital-elevation-model-etopo5) are linearly interpolated to a 0.25° × 0.25° grid to determine the land mask in our model.

4) Reanalyses and model simulation

Four reanalysis products, i.e., ECMWF Ocean Reanalysis System 5 (ORAS5, https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-oras5?tab=form; Zuo et al. 2019), Estimating the Circulation and Climate of the Ocean, Phase II (ECCO2, https://ecco.jpl.nasa.gov/drive/files/ECCO2; Menemenlis et al. 2008), Simple Ocean Data Assimilation 3.15.2 (SODA3.15.2, http://dsrs.atmos.umd.edu/DATA/soda3.15.2/ORIGINAL/ocean/; Carton et al. 2018), and National Centers for Environmental Prediction (NCEP) Global Ocean Data Assimilation System (GODAS, https://downloads.psl.noaa.gov/Datasets/godas; Behringer et al. 1998), and one model simulation product, i.e., Ocean General Circulation Model for the Earth Simulator (OFES, http://apdrc.soest.hawaii.edu/dchart/index.html; Masumoto et al. 2004) are used to diagnose the thermosteric contribution of sea surface heat fluxes and temperature advection. The detailed information on each product is summarized in Table 2.

Table 2.

Summary of variable information in the reanalysis and model simulation products. Here, Q is net sea surface heat flux; u, υ, and w are zonal, meridional, and vertical velocities; θ and S are potential temperature and salinity, respectively. The OFES data is the version further integrated with the QuikSCAT winds after a climatological run with NCEP winds for 50 years. The resolutions written in the parentheses in the last column are the original resolution of each heat flux product.
Table 2.

b. 1.5-layer nonlinear reduced-gravity model

The oceanic upper-layer circulation and isopycnal surface heaving in the 1.5-layer nonlinear reduced-gravity model are governed by the following equations:
∂u∂t+ζk×u=−∇E+Ah∇2u+τρ0h,
∂h∂t+∇⋅(hu)=0,
where u and h are the horizontal velocity vector and upper-layer thickness, respectively; ζ, E = g′h + (u2 + υ2)/2, Αh, τ, and ρ0 successively represent the absolute vorticity, total energy, horizontal viscosity coefficient, surface wind stress, and mean seawater density (set to a constant of 1025.0 kg m−3). The sea level (hRG) is derived by the formula hRG=(g′/g)h, where g′ and g are the reduced-gravity (0.033 m s−2) and acceleration due to gravity (9.8 m s−2), respectively. The governing equations are similar to Zhuang et al. (2013) except without the bottom friction. In this study, the model is spun up for 30 years with climatological winds over 1993–2018 and an initial upper-layer thickness of 350 m on a 0.25° × 0.25° grid in the Indo-Pacific (60°S–60°N, 20°E–70°W). Then the model is forced by monthly mean wind stresses from January 1993 to December 2018, denoted as EXP0. The model results are not sensitive to the choices of g′ and Ah. Four sensitivity experiments are designed to test the effects of winds across different basins. The experiment settings are described in Table 3.
Table 3.

Experiments of the 1.5-layer nonlinear reduced-gravity model.
Table 3.

c. Ekman pumping dynamics

The convergence or divergence of anomalous Ekman transport leads to a sea level change:
∂hEK∂t=−g′∇×τρ0gf−εhEK,
hEK(x,y,t)=−g′ρ0gf∫t0t∇×τ(x,y,t′)e−ε(t−t′)dt′,
where hEK denotes the Ekman-pumping-induced sea level change, f the Coriolis parameter, and ε the dissipation coefficient, which makes little difference to the result when ranging from (1 yr)−1 to (6 yr)−1. The other variables are the same as stated in section 2b.

d. Sea level variations dominated by the baroclinic modes

Oceanic baroclinic pressure anomaly (P′) which depends on the vertical structure of density is defined as
P′(z)=ρ0gη′+g∫z0[ρ(z′)−ρ0]′dz′=ρ0gη′+g∫z0ρ′(z′)dz′,
where ρ0 is the mean seawater density (set as 1025 kg m−3); ρ′ and η′ respectively represent the density and sea level anomalies, referenced to the mean over 1993–2018. The linear nth vertical mode function (ψn) is calculated using vertical profiles of Brunt–Väisälä frequency (Dewitte et al. 1999; Yang et al. 2022) and normalized so that ψn(0) = 1. Then SLA associated with each baroclinic mode (ηn′) is obtained through the projection of P′ onto ψn:
ηn′=∫−H0P′ψndz∫−H0ψn2dz/(ρ0g).

e. Thermosteric and halosteric sea level variations

SLAs due to temperature and salinity changes are termed thermosteric (ηT′) and halosteric (ηS′) components, which are calculated using the equations:
ηT′=∫−D0αTT′dz,
ηS′=−∫−D0αSS′dz,
where αT and αS are thermal expansion and saline contraction coefficients, respectively; T′ and S′ are temperature and salinity anomalies relative to the mean during 1993–2018, respectively; D is the lower limit of vertical integration (set to 800 m; the result remains unchanged when set to 2000 m). In this study, we found that the seasonal and interannual sea level variabilities in our study area are well represented by ηT′, which can be further decomposed by
ηT′=αT∫t0tQnetρ0cpdt−∫t0t∫−D0αT(u⋅∇T+w∂T∂z)dzdt+RES.,
where Qnet is net sea surface heat flux and cp the specific heat capacity; u ⋅ ∇T and w(∂T/∂z) are the horizontal and vertical temperature advection, respectively; and RES. represents the residual, mainly the contribution of subgrid-scale processes except for the errors in the computation of heat flux and advection terms.

f. Partial regression method

A partial regression method is employed to evaluate the independent effects of ENSO and IOD on the interannual sea level variability (Cai et al. 2011). After removing the linear relationship of both SSH and Niño-3.4 with DMI, termed SSH|DMI and Niño-3.4|DMI, the linear dependence of SSH|DMI upon the normalized Niño-3.4|DMI represents the corresponding partial regression coefficient. All the time series are first detrended and bandpass-filtered (18–84 months). The partial regression of SSH upon DMI is obtained analogously with the linear dependence of SSH and DMI on the Niño-3.4 index removed.

g. Explained variance percentage

Contributions of the simulated sea level variabilities to the observational ones are evaluated using an explained variance percentage (Var.%):
Var. %=[1−∑(SLAsim−SLAobs)2∑SLAobs2]×100%,
where SLAsim and SLAobs denote the simulated and observed sea level anomalies, respectively.

3. Seasonal and interannual sea level variabilities

To extract the major seasonal and interannual sea level variabilities in the southwest Pacific (35°–5°S, 140°E–180°), we perform an empirical orthogonal function (EOF) analysis of the climatological monthly and bandpass filtered (18–84 months) SSHs, whose linear trends are removed first. Figure 1 displays the first two EOF modes.

Fig. 1.
Fig. 1.

The first two EOF modes of (a)–(c) seasonal and (d)–(f) interannual SLAs (cm) in the southwest Pacific. Spatial patterns are shown in (a), (b), (d), and (e), with the explained variances in parentheses. The corresponding principal components are shown in (c) and (f) (solid black line for PC1 and dashed gray line for PC2). Niño-3.4 (light red and blue shading) and DMI (green line) are superimposed. The correlations between principal components and climate indices are displayed in Fig. 2.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

For seasonal variability, the leading EOF mode (Fig. 1a and solid black line in Fig. 1c) accounts for 62% of the variance. It shows that sea levels over 20°–10°S off the Australian coast fall in the first and rise in the second half of the year, reaching the lowest in austral autumn and highest in spring, which is the opposite in other regions. The large-amplitude anomalies are concentrated around the southeast coasts of New Guinea and Australia, having a maximum amplitude of ∼10 cm. The second EOF mode (Fig. 1b and dashed gray line in Fig. 1c), explaining 27% of the seasonal variance, depicts an antiphase variation of sea levels in winter and summer. In most regions, sea levels are anomalously high in austral summer and low in winter, which is contrary to the situation near the western boundary. The maximum amplitude of variation is about half of that in the first EOF mode.

At interannual time scales, the leading EOF mode (Fig. 1d and solid black line in Fig. 1f), with an explained variance of 52%, characterizes oppositely signed SLAs in the northern and southern basins, separated by a northwest–southeast-oriented zero-contour extending from the southeastern tip of New Guinea to 20°S and 180°, which is roughly the mean position of the South Pacific convergence zone (SPCZ) (Folland et al. 2002). The position of the SPCZ shifts northward during El Niño and southward during La Niña (Gouriou and Delcroix 2002). Large-amplitude variations mainly occur to the north of the SPCZ, where the SLAs also appear to be closely related to Niño-3.4 (light red and blue shadings in Fig. 1f) and DMI (green line in Fig. 1f) (bandpass filtered, 18–84 months). For example, during the 1997/98, 2009/10, and 2015/16 El Niño events, sea levels in the northern basin significantly declined by over 20 cm, while during La Niña years, e.g., 1996/97, 1998/99, 1999/00, 2007/08, 2010/11, and 2017/18, the sea levels there rose by ∼10 cm. The lag correlations between the principal components and Niño-3.4/DMI are elaborated in the following paragraph. Note that IODs often occur simultaneously with ENSOs (Table 1), which may obscure their independent role in modulating sea level variations. We will address this issue later. The second EOF mode (Fig. 1e and dashed gray line in Fig. 1f) explains 19% of the interannual variance. It features a mostly coherent phase of sea level variations in the tropical Pacific off the western boundary, with relatively large anomalies occurring over ∼20°–12°S. The spatial pattern with alternate positive and negative anomalies in the subtropical ocean suggests a local response to interannual forcings. Near the southeast Australian coast, the sea level variations are particularly significant, with the amplitudes exceeding 12 cm.

After removing the linear relationship with DMI, the first principal component (PC1), the second principal component (PC2), and Niño-3.4 index are termed PC1|DMI, PC2|DMI, and Niño-3.4|DMI, respectively. Analogously, PC1|Niño-3.4, PC2|Niño-3.4, and DMI|Niño-3.4 refer to the exclusion of the linear dependence on Niño-3.4 index. Then the lag partial correlations between the interannual PCs and climate indices are calculated and shown in Figs. 2a and 2b (solid lines), with the lag correlations without removing the linear effects of the other climate index superimposed for comparison (dashed lines). The correlation significance is measured using the t test, with the effective degrees of freedom being 11 and 10 for seasonal and interannual variabilities, respectively, estimated following Bretherton et al. (1999). Overall, the correlations with DMI (blue lines in Figs. 2a,b) are more sensitive to whether the linear dependence on ENSO is removed. There is a maximum negative correlation of −0.50 (the critical correlation coefficient at the 90% significance level is 0.52) between the PC1 and DMI when the latter leads by five months, whereas the correlation between the PC1|Niño-3.4 and DMI|Niño-3.4 declines to only −0.11. This demonstrates that a significant portion of the sea level changes during IOD events is derived from the concurrent ENSO and are largely related to the convection anomalies over the Maritime Continent (Hendon 2003; Wang et al. 2016). In contrast, the maximum negative correlation between the PC1 and Niño-3.4 index decreases a little from −0.88 (α < 0.01) to −0.81 (α < 0.01) after removing the IOD’s contribution, with the Niño-3.4 index leading by four months. Therefore, ENSO plays a dominant role in controlling the interannual sea level variability in the southwest Pacific. IOD, as denoted in previous studies (e.g., Yuan et al. 2011; Jourdain et al. 2016), may exert impacts on Pacific sea levels by modulating the ENSO life cycle. The second interannual mode also exhibits a closer relationship with ENSO than with IOD. The difference is that the correlations become notably weaker once the linear effects of the other index are retained, especially for that between the PC2 and DMI. It is interesting to note that the correlations between the PC2|DMI and Niño-3.4|DMI have a maximum of 0.59 (α < 0.1) when the former leads by nine months. On the other hand, PC2|Niño-3.4 and DMI|Niño-3.4 correlate at only −0.28 with the former leading by 11 months, suggesting a tight relation between the southwest Pacific sea level variations and the development of the following ENSO events. These correlations need further investigation for causal links which are likely to be via the recharge–discharge mechanisms and the Western Pacific convection (Annamalai et al. 2005; Izumo et al. 2010; Ramesh and Murtugudde 2013).

Fig. 2.
Fig. 2.

Lag correlations between the (a) first and (b) second principal component and Niño-3.4 (red lines) and DMI (blue lines). The solid (dashed) lines represent the correlations after (without) removing the linear dependence on the other climate index. (c) Partial regressions of observational SSHs on Niño-3.4 index at the 95% confidence level, with the Niño-3.4 index leading by four months. (d) Explained variance associated with Niño-3.4 index (only exhibiting the values where the partial regressions pass the 95% confidence level). All the time series are detrended and bandpass filtered (18–84 months).

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

To quantify the interannual variance associated with ENSO, we calculate the lag partial regression coefficients of detrended and bandpass filtered (18–84 months) SSHs on Niño-3.4 index, with the index leading by four months. The corresponding explained variance is then estimated using the regression residuals. The results are displayed in Figs. 2c and 2d. Spatially, the partial regressions resemble the leading EOF mode shown in Fig. 1d. The regression coefficients are most significant in the tropical basin north of 15°S, where the ENSO-related variance accounts for 40%–60% of the total variance. In the southern tropical basin, mainly to the west of 170°E, the proportion of variance explained by ENSO reduces to 20%–40%. In the extratropics, the direct influence of ENSO is so limited that the regression coefficients are insignificant over nearly half the regions, and the explained variance is mostly less than 20%.

4. Driving forces and mechanisms

a. Barotropic and baroclinic components of sea level variabilities

1) Validation of ORAS5 data

In this study, the vertical mode decomposition of baroclinic pressure anomalies is based on AVISO SSHs and ORAS5 temperatures and salinities, which show good coherence with Argo observations during January 1993–December 2018, particularly in the upper 100-m depth, where the correlation coefficients generally exceed 0.8 (Fig. 3). Relatively weak correlations (less than 0.4) mainly occur in the extratropics in deep layers below the 600 m depth (Figs. 3e,f,h,i). Considering the errors of Argo data introduced by the interpolation and gridding method and the mostly significant positive correlations between ORAS5 and AVISO SSHs, the results of vertical mode decomposition using ORAS5 reanalyses are convincing and illuminating.

Fig. 3.
Fig. 3.

Correlation coefficients between variables from ORAS5 and observations over January 1993–December 2018. (a) Comparison of SSHs with AVISO altimetry data. (b)–(i) Comparisons of temperatures (T) and salinities (S) with Argo data.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

2) Barotropic and baroclinic contributions

The mass-induced SLAs, i.e., the barotropic component derived from GRACE ocean bottom pressure measurements, have a very limited contribution in most regions from 2003 to 2018 (Figs. 4a,b). For seasonal variability, there are two domains, east of 160°E between 10° and 15°S and near the western boundary, where the barotropic component explains 20%–60% of the variance (Fig. 4a). At interannual time scales, the barotropic contribution is negligible elsewhere except in the north Papua Gulf and off the east Australian coast around 22°S and 150°E, where the barotropic variations account for ∼40% of the total interannual variance (Fig. 4b). Therefore, the seasonal and interannual sea level variabilities in the southwest Pacific Ocean are primarily baroclinic.

Fig. 4.
Fig. 4.

Explained variance of (a),(b) barotropic (mass-induced) and (c)–(f) baroclinic components of SLAs. (left) Seasonal and (right) interannual variability. The first baroclinic mode is shown in (c) and (d), and the summation of the second to sixth baroclinic modes is shown in (e) and (f). Note that the barotropic component, derived from GRACE ocean bottom pressure measurements, is from 2003 to 2018. The baroclinic component, calculated based on vertical mode decomposition of baroclinic pressure anomalies, is from 1993 to 2018.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

Specifically, the first baroclinic mode predominantly controls the sea level variabilities at seasonal and interannual time scales in most tropical regions, accounting for more than 60%–80% of the variances over 1993–2018 (Figs. 4c,d). Outside the tropics, however, the first baroclinic mode becomes less important, with its explained variances hardly exceeding 20%. For seasonal sea levels within 22°–28°S and interannual sea levels south of 25°S and west of 170°E, the first baroclinic mode even shows a negative contribution. Meanwhile, the higher baroclinic modes stand out in the extratropics. About 40%–60% of the seasonal variance can be explained by the second to sixth baroclinic modes. In contrast, only a maximum of ∼20% of the interannual variance is attributable to the higher baroclinic modes (second–sixth). Due to the limitation of the linearity of vertical mode decomposition, nonlinear advection, diffusion, or meso- and small-scale processes are presumed vital for the interannual variability of extratropical sea levels. In the following analyses, a 1.5-layer reduced-gravity model is employed to explore the dynamical response of sea levels to wind forcings. In the regions where the model becomes deficient, the thermodynamical roles of surface heat fluxes and advective processes are diagnosed based on Eqs. (7)–(9).

b. Dynamical effects of wind forcings

1) 1.5-layer reduced-gravity model

Five numerical model experiments are conducted to diagnose the sensitivity of seasonal and interannual sea level variabilities to wind forcings from the eastern Pacific east of 180° (EXP1Seasonal and EXP1Interannual), the eastern Pacific east of 160°E (EXP2Seasonal), and the equatorial Pacific east of 180° (EXP2Interannual) by comparing with the results from the control runs (EXP0). Further information on model experiment settings is listed in Table 3. The phase correlations between the modeled and observed sea levels and the variances explained by the model experiments are displayed in Fig. 5 (seasonal) and Fig. 6 (interannual).

Fig. 5.
Fig. 5.

(a)–(c) Seasonal phase correlation between observations and model simulations from (a) EXP0 (control run), (b) EXP1Seasonal (no seasonal cycle of wind stress east of the date line over 35°–5°S), and (c) EXP2Seasonal (no seasonal cycle of wind stress east of 160°E over 35°–5°S). (d)–(f) Proportion of seasonal variance explained by the model experiments. The black contours with dots mark the critical correlation coefficient (0.55) passing the 95% confidence level test.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

Fig. 6.
Fig. 6.

(a)–(c) Interannual phase correlation between observations and model simulations from (a) EXP0 (control run), (b) EXP1Interannual (no interannual signal in the Pacific wind stress over 35°–5°S and 180°–70°W), and (c) EXP2Interannual (no interannual signal in the Pacific wind stress over 35°S–0° and 180°–70°W). (d)–(f) Proportion of interannual variance explained by the model experiments. The black contours with dots mark the critical correlation coefficient (0.60) passing the 95% confidence level test.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

In the tropical regions except around the southern New Guinea and western boundary and east of the Solomon Islands at ∼10°S, the seasonal phases of sea levels are captured well by the model control run (EXP0, Figs. 5a,d), with the correlation generally between 0.6 and 0.9 (α < 0.05), explaining 60%–90% (generally >80% over 19°–13°S) of the variance, highlighting the significance of the wind-driven component of the first baroclinic mode. In the extratropical regions south of 30°S, though the seasonal phases are reproduced well [correlation (Corr.) > 0.8, α < 0.01], the explained variance decreases to less than ∼40%. Between ∼21° and 30°S, the model results become severely deficient, with generally negative correlation and explained variance. The sensitivity experiment (EXP1Seasonal, Figs. 5b,e) shows that the wind stress curl east of the date line is most significant to the seasonal sea level variability in the tropical regions north of 10°S and east of 160°E at 20°–10°S, since the corresponding correlation and explained variance decline substantially when the seasonal variability of the winds in the remote eastern Pacific basin is not considered. On the contrary, the sea levels in the western Coral Sea (22°–14°S) and midlatitude regions (south of 30°S) are not sensitive to the wind variations east of the date line but are influenced by the winds to the east of 160°E (EXP2Seasonal, Figs. 5c,f), demonstrating the primary effects of the winds over 160°E–180°.

At interannual time scales, except in some local areas (to the south of southern New Guinea and near the western boundary), the model simulations of the tropical sea levels are well correlated with observations (Corr. > 0.60, α < 0.05), accounting for 40%–78% of the variance (EXP0, Figs. 6a,d). Nonetheless, the interannual simulations are generally inferior to the seasonal ones except around the regions to the east of southern New Guinea. In the extratropics, the model fails to reproduce the interannual sea level variabilities, suggesting that other factors may have to be considered. Over the entire tropical basin, the winds to the east of the date line act as a more significant contributor than those to the west, explaining more than half of the interannual variance (Figs. 6b,e). The discrepancy between the results of EXP1interannual (Figs. 6b,e) and EXP2interannual (Figs. 6c,f) indicates that the equatorial interannual variations could affect the interannual sea levels around 5°S through Kelvin waves and Rossby waves. Figure 7 displays the composite SLAs (from EXP2interannual) in El Niño years (see Table 1) from month −12 to month 12 at the equator (from 180° to the eastern boundary), along the eastern boundary (from the equator to 5°S), and at 5°S (from the eastern boundary toward west to the western boundary). Here, month 0 refers to the month when Niño-3.4 index reaches the maximum during all El Niño events, and the negative and positive months represent those leading and lagging month 0, respectively, by corresponding months. During the mature phase of an El Niño, a strong positive SLA signal propagates from the equator to 5°S through eastward equatorial Kelvin waves (CK ≈ 2.8 m s−1, Fig. 7a), southward coastal Kelvin waves (Fig. 7b), and westward Rossby waves (CR ≈ 1.1 m s−1, Fig. 7c). Hence, the interannual sea level variability at low latitudes in the tropical southwest Pacific is closely related to the equatorial dynamics.

Fig. 7.
Fig. 7.

Composite SLAs (cm) in EXP1interannual during El Niño years along (a) the equator, (b) the eastern boundary, and (c) 5°S. Month 0 represents the one when Niño-3.4 reaches the maximum during each El Niño event. Months −12 to −1 (1–12) represent those leading (lagging) month 0 by corresponding months.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

Overall, the 1.5-layer reduced-gravity model produces consistent patterns of seasonal and interannual sea level variations as the first baroclinic mode does (see Figs. 4c,d), only with the contributions slightly reduced, since other factors, e.g., the buoyancy forcings, may also generate sea level variations through the first baroclinic mode. In addition, the configuration of latitude-dependent reduced-gravity (g′) and viscosity (Ah) coefficients cannot improve the model results, suggesting the essential role of the other processes, which will be diagnosed in section 4c.

2) Local Ekman pumping

To assess the local effects of wind stress curl, we calculate the Ekman-pumping-induced SLAs and compare them with observations, as displayed in Fig. 8. At seasonal time scales, significant positive correlation (Corr. > 0.55, α < 0.05) is evident in most tropical regions and south of ∼30°S, with notable explained variance (40%–80%) mainly occurring over 10°–20°S (Figs. 8a,b). Remembering the results from EXP1Seasonal (Figs. 5b,e), the seasonal sea level variations in the western Coral Sea and at higher latitudes south of 30°S are mainly attributable to Ekman pumping, while in the other tropical regions, Rossby waves arising from east of the date line are the most critical factor. In the region east of 160°E over 10°–20°S, the effects of Ekman pumping counteract those of Rossby waves arising from west of the date line. As the seasonal signals are filtered, the interannual correlation reaches the statistical significance (Corr. > 0.60, α < 0.05) only in a limited area to the north of ∼12°S, where the explained variance is generally less than 20% (Figs. 8c,d), emphasizing the role of remote Rossby waves. Local forcings along the Rossby wave paths may have to be considered for matching the observations better.

Fig. 8.
Fig. 8.

(a) Seasonal and (c) interannual correlations between the Ekman-pumping-induced and observational SLAs. Proportions of (b) seasonal and (d) interannual variances explained by the Ekman pumping dynamics. The black contours with dots mark the critical correlation coefficient (0.55 for seasonal and 0.60 for interannual correlations) passing the 95% confidence level test.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

c. Thermodynamic effects on steric sea level variations

The thermosteric and halosteric sea level variations are calculated based on the observational Argo temperatures and salinities. Their contributions to the seasonal and interannual sea level variabilities are shown in Fig. 9, in which the predominance of the thermosteric component is evident, with significant positive correlations (Figs. 9a,e) and more than 60%–80% of the seasonal and interannual variances (Figs. 9c,g) over majority of the southwest Pacific. Nonetheless, in some local areas south of 20°S, around the northeast Australian coast, and at the zonal band of ∼10°S (for seasonal variability), and in the southwest corner of our study area (for interannual variability), the thermosteric component presents negative explained variances mainly due to larger magnitudes of variations than the observations. Meanwhile, the halosteric component plays a weak compensating role (Figs. 9b,d,f,h). Only in some sporadic areas, e.g., to the east of 165°E around 10°S, can the halosteric component account for more than 20% of the variances (Figs. 9d,h). Although the relatively coarse resolution (1° × 1°) of gridded Argo data introduces some uncertainties, the fact that the steric sea level variations mainly arise from temperature changes is credible. Therefore, the thermosteric sea level variations due to local sea surface heat fluxes (ηQ) and horizontal (ηHΑDV) and vertical temperature advection (ηVΑDV) are further examined based on four reanalyses and one eddy-permitting model simulation. Note that the thermal advection effects are not independent of the results from the 1.5-layer reduced-gravity model since the Rossby-wave-induced thermocline heaving could also induce temperature anomalies and thermosteric sea level variations. Thus, ηHΑDV and ηVΑDV are not concerned about where the 1.5-layer reduced-gravity model works well.

Fig. 9.
Fig. 9.

(a),(b) Seasonal and (e),(f) interannual correlations between thermosteric/halosteric and observational SLAs. (c),(d) Seasonal and (g),(h) interannual corresponding explained variances. (left) Thermosteric and (right) halosteric components and total sea level variations.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

1) Seasonal thermosteric sea level variability

Despite discrepancies in the details, all the reanalyses (ORAS5, ECCO2, SODA3.15.2, and GODAS) and the OFES model simulation show a consistent result. Hence, only the results from ECCO2 and OFES are displayed in Figs. 10 and 11 (results from the other datasets can be found in the Figs. S1–S3 in the online supplemental material). For seasonal variability, ηQ acts as a good complement to the wind-driven first baroclinic sea level variations in the regions equatorward of 10°S and poleward of ∼20°S, where ηQ shows coincident phases with the observations over a large area (Corr. > 0.55, α < 0.05) (Fig. 10a). The explained variance of ηQ increases from ∼20% at 5°S to ∼60% at 10°S and are mostly within 40%–80% at latitudes south of 20°S (Fig. 10e). Therefore, in the regions where the 1.5-layer reduced-gravity model is severely inadequate to reproduce the seasonal signals, e.g., around the southern tip of New Guinea and the northeast Australian coast, east of the Solomon Islands, and south of 20°S, the surface heat fluxes make a major contribution. In these regions, both ηHΑDV and ηVΑDV generally play a minor role, with positive correlations and contributions only occurring in some local spots (Figs. 10b,c,f,g). The consequence is that taking the effects of horizontal and vertical advection into consideration leads to an even worse simulation than ηQ in the regions where the 1.5-layer reduced-gravity model fails (Figs. 10d,h). Note that in Fig. 11, the comparison of each component is referenced to the model simulated sea levels, thus allowing for a closed budget of the thermosteric sea level variations. Figures 11d and 11h suggest that except for the surface heat fluxes, the small-scale diffusive processes are also vital to complement the opposite role advection plays.

Fig. 10.
Fig. 10.

Contributions of (a),(e) local sea surface heat fluxes, (b),(f) horizontal, and (c),(g) vertical temperature advection, and (d),(h) their summation to the seasonal sea level variations based on ECCO2. (top) The correlation coefficient between each component and observational sea level variations. (bottom) The corresponding explained variance. The black contours with dots mark the critical correlation coefficient (0.55) passing the 95% confidence level test.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

Fig. 11.
Fig. 11.

As in Fig. 10, but based on OFES simulations.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

2) Interannual thermosteric sea level variability

In this section, we only focus on the extratropics (south of 25°S), where the 1.5-layer reduced-gravity model becomes invalid. All the four reanalyses and OFES model simulation reveal a negligible contribution of local sea surface heat fluxes and advective temperature transport (Fig. 12, based on ECCO2; Fig. 13, based on OFES; results from the other reanalyses are in Figs. S4–S6). Considering the dynamical consistency in the model producing the OFES simulation, a closed budget requires the diffusive processes as the most crucial factor dominating the interannual sea level variability outside the tropics, which differs from the case associated with the seasonal sea level variability. Thus, a proper parameterization of the subgrid-scale processes is vital to successfully predict the interannual variability in the southwest Pacific.

Fig. 12.
Fig. 12.

As in Fig. 10, but for the interannual sea level variability.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

Fig. 13.
Fig. 13.

As in Fig. 12, but based on OFES simulations.

Citation: Journal of Physical Oceanography 53, 9; 10.1175/JPO-D-23-0018.1

5. Summary and discussion

Using the altimetric sea levels, Argo temperatures and salinities, GRACE ocean bottom pressure anomalies, and four reanalysis products (ORAS5, ECOO2, SODA3.15.2, and GODAS) and a model simulation (OFES), we examined the seasonal and interannual sea level variabilities and the underlying dynamics in the southwest Pacific over 1993–2018, confirming the difference in dominant drivers varying over space and time. Broadly speaking, most sea level variabilities in the tropics are dominated by the first baroclinic mode forced by winds, while outside the tropics to the surface heat (seasonal) or diffusive fluxes (interannual).

The seasonal variability is mainly characterized by an opposite pattern between a zonal band at 20°–10°S (away from the western boundary) and the other regions, with the highest and lowest sea levels occurring in autumn or spring. Over most of the tropical Pacific, the wind-driven first-mode baroclinic Rossby waves play a dominant role (especially over 19°–13°S), explaining 60%–90% of the total variance, mainly originating from east of the date line, except in the western Coral Sea where the Rossby waves coming into the study region are forced over 160°E–180°. The exceptions are around the southern New Guinea and east Australian coast, some local areas east of the southern Solomon Islands, and south of 20°S, where the 1.5-layer reduced-gravity model becomes too deficient. Over a majority of these regions, the contribution of the surface heat fluxes stands out, accounting for 40%–80% of the variance. Meanwhile, the small-scale diffusive processes of temperature are also indispensable to compensate the advection effects.

The interannual sea levels are dominated by a seesaw variation in the northern and southern tropical basins roughly separated by the SPCZ, with the amplitudes of variation in the northern basin being particularly prominent. Using a partial regression method, we assessed the independent role played by ENSO and IOD in modulating the interannual sea level variability. The dominance of ENSO is confirmed with a maximum negative correlation coefficient of −0.81 (α < 0.01) between the PC1|DMI and the Niño-3.4|DMI when the latter leads by four months. In tropical regions, ∼40%–60% of the total interannual variance are ENSO related. In contrast, IOD’s effect seems to become negligible after removing the linear dependence on ENSO. Some previous studies have revealed that IOD events favor the phase transitions of ENSOs through influencing winds in the western-central tropical Pacific (Izumo et al. 2010, 2014, 2015; Jourdain et al. 2016; Duan et al. 2020). Thus, IODs are likely to affect the Pacific sea levels by modulating the evolution of ENSOs. In this study, the ITF is poorly simulated by the 1.5-layer reduced-gravity model. A coupled climate model with a satisfactory depiction of the ITF is required to resolve the Indian Ocean effects on the interannual sea level variability in the southwest Pacific.

The 1.5-layer reduced-gravity model explains 40%–78% of the interannual variance in the southwestern tropical Pacific (the northern Coral Sea and western boundaries are not included), which are controlled by the first-mode baroclinic Rossby waves and equatorial Kelvin waves, mostly from the east of the date line. At extratropical latitudes, the interannual sea level variability is not sensitive to the surface heat fluxes or horizontal and vertical advection of temperature. Thus, diffusion is probably vital to the interannual variability there, in consideration of the closed budget in dynamically consistent numerical models. The significant role of temperature diffusion in global interannual sea level variability was previously reported based on an observationally constrained oceanic model output (Piecuch and Ponte 2011). We suggest that a more sophisticated coupled climate model with proper parameterization of subgrid-scale processes is necessary for a comprehensive study in the future. Considering the close association between the sea level variabilities in the southwest Pacific and the fluctuations of volume transport and bifurcation latitude of the South Equatorial Current, a thorough investigation of the mechanisms controlling the South Pacific sea level variabilities is of potential significance to advance the understanding of ITF variabilities. Needless to say, this low-latitude component of the global conveyor belt is a critical player in ocean heat storage and rate of global warming. Sea level variabilities in this corner of the Pacific are also likely to provide lead-time clues to the ENSO evolution (Ramesh and Murtugudde 2013).

Acknowledgments.

The authors thank the two anonymous reviewers for their constructive comments. We would like to acknowledge the Data Unification and Altimeter Combination System, the Copernicus Marine Environment and Monitoring Service, the National Oceanic and Atmospheric Administration/Earth System Research Laboratories, the Hadley Centre Sea Ice and Sea Surface Temperature, the European Centre for Medium-Range Weather Forecasts, National Centers for Environmental Prediction, and Asia-Pacific Data-Research Center for sharing the data online. This research is funded by the National Key Research and Development Program of China (2021YFC3101801), the National Natural Science Foundation of China (42006023, 41706033, 41976200, 42276019), the program for scientific research start-up funds of Guangdong Ocean University (R20020, R20023). RM would like to gratefully acknowledge the Visiting Faculty position at IIT Bombay.

Data availability statement.

All the data links can be found in section 2a.

REFERENCES

  • Albrecht, F., O. Pizarro, A. Montecinos, and X. Zhang, 2019: Understanding sea level change in the South Pacific during the late 20th century and early 21st century. J. Geophys. Res. Oceans, 124, 3849–3858, https://doi.org/10.1029/2018JC014828.

    • Search Google Scholar
    • Export Citation

  • Andres, M., R. C. Musgrave, D. L. Rudnick, K. L. Zeiden, T. P. Eacock, and J. H. Park, 2020: On the predictability of sea surface height around Palau. J. Phys. Oceanogr., 50, 3267–3294, https://doi.org/10.1175/JPO-D-19-0310.1.

    • Search Google Scholar
    • Export Citation

  • Annamalai, H., S. P. Xie, J. P. McCreary, and R. Murtugudde, 2005: Impact of Indian Ocean sea surface temperature on developing El Niño. J. Climate, 18, 302–319, https://doi.org/10.1175/JCLI-3268.1.

    • Search Google Scholar
    • Export Citation

  • Behringer, D. W., M. Ji, and A. Leetmaa, 1998: An improved coupled model for ENSO prediction and implications for ocean initialization. Part I: The ocean data assimilation system. Mon. Wea. Rev., 126, 1013–1021, https://doi.org/10.1175/1520-0493(1998)126<1013:AICMFE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Bowen, M. M., P. J. H. Sutton, and D. Roemmich, 2006: Wind-driven and steric fluctuations of sea surface height in the southwest Pacific. Geophys. Res. Lett., 33, L14617, https://doi.org/10.1029/2006GL026160.

    • Search Google Scholar
    • Export Citation

  • Bretherton, C. S., M. Widmann, V. P. Dymnikov, J. M. Wallace, and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 1990–2009, https://doi.org/10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Busalacchi, A. J., K. Takeuchi, and J. J. O’Brien, 1983: On the interannual wind-driven response of the tropical Pacific Ocean. Hydrodynamics of the Equatorial Ocean, J. C. J. Nihoul, Ed., Elsevier Oceanography Series, Vol. 36, Elsevier, 155–195, https://doi.org/10.1016/S0422-9894(08)70634-2.

  • Cai, W. J., P. V. Rensch, T. Cowan, and H. H. Hendon, 2011: Teleconnection pathways of ENSO and the IOD and the mechanisms for impacts on Australian rainfall. J. Climate, 24, 3910–3923, https://doi.org/10.1175/2011JCLI4129.1.

    • Search Google Scholar
    • Export Citation

  • Carton, J. A., G. A. Chepurin, and L. Chen, 2018: SODA3: A new ocean climate reanalysis. J. Climate, 31, 6967–6983, https://doi.org/10.1175/JCLI-D-18-0149.1.

    • Search Google Scholar
    • Export Citation

  • Chen, J. L., C. K. Shum, C. R. Wilson, D. P. Chambers, and B. D. Tapley, 2000: Seasonal sea level change from TOPEX/Poseidon observation and thermal contribution. J. Geod., 73, 638–647, https://doi.org/10.1007/s001900050002.

    • Search Google Scholar
    • Export Citation

  • Dewitte, B., G. Reverdin, and C. Maes, 1999: Vertical structure of an OGCM simulation of the equatorial Pacific Ocean in 1985–94. J. Phys. Oceanogr., 29, 1542–1570, https://doi.org/10.1175/1520-0485(1999)029,1542:VSOAOS.2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Duan, J., Y. Li, L. Zhang, and F. Wang, 2020: Impacts of the Indian Ocean dipole on sea level and gyre circulation of the western tropical Pacific Ocean. J. Climate, 33, 4207–4228, https://doi.org/10.1175/JCLI-D-19-0782.1.

    • Search Google Scholar
    • Export Citation

  • Folland, C. K., J. A. Renwick, M. J. Salinger, and A. B. Mullan, 2002: Relative influences of the interdecadal Pacific Oscillation and ENSO on the South Pacific convergence zone. Geophys. Res. Lett., 29, 1643, https://doi.org/10.1029/2001GL014201.

    • Search Google Scholar
    • Export Citation

  • Gordon, A. L., 2005: Oceanography of the Indonesian Seas and their throughflow. Oceanography, 18, 14–27, https://doi.org/10.5670/oceanog.2005.01.

    • Search Google Scholar
    • Export Citation

  • Gouriou, Y., and T. Delcroix, 2002: Seasonal and ENSO variations of sea surface salinity and temperature in the South Pacific convergence zone during 1976–2000. J. Geophys. Res., 107, 8011, https://doi.org/10.1029/2001JC000830.

    • Search Google Scholar
    • Export Citation

  • Hendon, H. H., 2003: Indonesian rainfall variability: Impacts of ENSO and local air–sea interaction. J. Climate, 16, 1775–1790, https://doi.org/10.1175/1520-0442(2003)016<1775:IRVIOE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Ho, C.-R., Q. Zheng, Y. S. Soong, N.-J. Kuo, and J.-H. Hu, 2000: Seasonal variability of sea surface height in the South China Sea observed with TOPEX/Poseidon altimeter data. J. Geophys. Res., 105, 13 981–13 990, https://doi.org/10.1029/2000JC900001.

    • Search Google Scholar
    • Export Citation

  • Izumo, T., and Coauthors, 2010: Influence of the state of the Indian Ocean dipole on the following year’s El Niño. Nat. Geosci., 3, 168–172, https://doi.org/10.1038/ngeo760.

    • Search Google Scholar
    • Export Citation

  • Izumo, T., M. Lengaigne, J. Vialard, J.-J. Luo, T. Yamagata, and G. Madec, 2014: Influence of Indian Ocean dipole and Pacific recharge on following year’s El Niño: Interdecadal robustness. Climate Dyn., 42, 291–310, https://doi.org/10.1007/s00382-012-1628-1.

    • Search Google Scholar
    • Export Citation

  • Izumo, T., J. Vialard, H. Dayan, M. Lengaigne, and I. Suresh, 2015: A simple estimation of equatorial Pacific response from windstress to untangle Indian Ocean dipole and basin influences on El Niño. Climate Dyn., 46, 2247–2268, https://doi.org/10.1007/s00382-015-2700-4.

    • Search Google Scholar
    • Export Citation

  • Jourdain, N. C., M. Lengaigne, J. Vialard, T. Izumo, and A. Sen Gupta, 2016: Further insights on the influence of the Indian Ocean dipole on the following year’s ENSO from observations and CMIP5 models. J. Climate, 29, 637–658, https://doi.org/10.1175/JCLI-D-15-0481.1.

    • Search Google Scholar
    • Export Citation

  • Liu, X., and H. Zhou, 2020: Seasonal variations of the North Equatorial Current across the Pacific Ocean. J. Geophys. Res. Oceans, 125, e2019JC015895, https://doi.org/10.1029/2019JC015895.

    • Search Google Scholar
    • Export Citation

  • Luo, J. J., and T. Yamagata, 2001: Long-term El Niño-Southern Oscillation (ENSO)-like variation with special emphasis on the South Pacific. J. Geophys. Res., 106, 22 211–22 227, https://doi.org/10.1029/2000JC000471.

    • Search Google Scholar
    • Export Citation

  • Masumoto, Y., and Coauthors, 2004: A fifty-year eddy-resolving simulation of the world ocean-preliminary outcomes of OFES (OGCM for the earth simulator). J. Earth Simul., 1, 35–56, https://doi.org/10.32131/jes.1.35.

    • Search Google Scholar
    • Export Citation

  • Menemenlis, D., J. Campin, P. Heimbach, C. Hill, T. Lee, A. Nguyen, M. Schodlock, and H. Zhang, 2008: ECCO2: High resolution global ocean and sea ice data synthesis. Mercator Ocean Quarterly Newsletter, No. 31, Mercator Ocean, Ramonville Saint-Agne, France, 13–21.

  • Piecuch, C. G., and R. M. Ponte, 2011: Mechanisms of interannual steric sea level variability. Geophys. Res. Lett., 38, L15605, https://doi.org/10.1029/2011GL048440.

    • Search Google Scholar
    • Export Citation

  • Qiu, B., and T. M. Joyce, 1992: Interannual variability in the mid-and low-latitude western North Pacific. J. Phys. Oceanogr., 22, 1062–1079, https://doi.org/10.1175/1520-0485(1992)022<1062:IVITMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Qiu, B., and S. Chen, 2006: Decadal variability in the large-scale sea surface height field of the South Pacific Ocean: Observations and causes. J. Phys. Oceanogr., 36, 1751–1762, https://doi.org/10.1175/JPO2943.1.

    • Search Google Scholar
    • Export Citation

  • Qu, T., and E. J. Lindstrom, 2002: A climatological interpretation of the circulation in the western South Pacific. J. Phys. Oceanogr., 32, 2492–2508, https://doi.org/10.1175/1520-0485(2002)032<2492:ACIOTC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Ramesh, N., and R. Murtugudde, 2013: All flavours of El Niño have similar early subsurface origins. Nat. Climate Change, 3, 42–46, https://doi.org/10.1038/nclimate1600.

    • Search Google Scholar
    • Export Citation

  • Ren, Q., Y. Li, F. Zheng, F. Wang, J. Duan, and R. Li, 2020: Asymmetry of interannual sea level variability in the western tropical Pacific: Responses to El Niño and La Niña. J. Geophys. Res. Oceans, 125, e2020JC016616, https://doi.org/10.1029/2020JC016616.

    • Search Google Scholar
    • Export Citation

  • Roberts, C. D., D. Calvert, N. Dunstone, L. Hermanson, M. D. Palmer, and D. Smith, 2016: On the drivers and predictability of seasonal-to interannual variations in regional sea level. J. Climate, 29, 7565–7585, https://doi.org/10.1175/JCLI-D-15-0886.1.

    • Search Google Scholar
    • Export Citation

  • Roemmich, D., and J. Gilson, 2009: The 2004–2008 mean and annual cycle of temperature, salinity, and steric height in the global ocean from the Argo Program. Prog. Oceanogr., 82, 81–100, https://doi.org/10.1016/j.pocean.2009.03.004.

    • Search Google Scholar
    • Export Citation

  • Roemmich, D., J. Gilson, R. Davis, P. Sutton, S. Wijffels, and S. Riser, 2007: Decadal spinup of the South Pacific subtropical gyre. J. Phys. Oceanogr., 37, 162–173, https://doi.org/10.1175/JPO3004.1.

    • Search Google Scholar
    • Export Citation

  • Stammer, D., 1997: Steric and wind-induced changes in TOPEX/POSEIDON large-scale sea surface topography observations. J. Geophys. Res., 102, 20 987–21 009, https://doi.org/10.1029/97JC01475.

    • Search Google Scholar
    • Export Citation

  • Sun, J., L. Zhang, and D. Hu, 2022: Decadal and long-term variability of sea level in the southwestern Pacific during 1948–2018. Geophys. Res. Lett., 49, e2022GL098747, https://doi.org/10.1029/2022GL098747.

    • Search Google Scholar
    • Export Citation

  • Vivier, F., K. A. Kelly, and L. Thompson, 1999: Contributions of wind forcing, waves, and surface heating to sea surface height observations in the Pacific Ocean. J. Geophys. Res., 104, 20 767–20 788, https://doi.org/10.1029/1999JC900096.

    • Search Google Scholar
    • Export Citation

  • Vivier, F., K. A. Kelly, and M. Harismendy, 2005: Causes of large-scale sea level variations in the Southern Ocean: Analyses of sea level and a barotropic model. J. Geophys. Res., 110, C09014, https://doi.org/10.1029/2004JC002773.

    • Search Google Scholar
    • Export Citation

  • Wang, H., R. Murtugudde, and A. Kumar, 2016: Evolution of Indian Ocean dipole and its forcing mechanisms in the absence of ENSO. Climate Dyn., 47, 2481–2500, https://doi.org/10.1007/s00382-016-2977-y.

    • Search Google Scholar
    • Export Citation

  • Wang, J., J. Li, J. Yin, W. Tan, and Y. Liu, 2021: Sea level seasonal, interannual and decadal variability in the tropical Pacific Ocean. Remote Sens., 13, 3809, https://doi.org/10.3390/rs13193809.

    • Search Google Scholar
    • Export Citation

  • Wu, Q., X. Zhang, J. A. Church, and J. Hu, 2017: Variability and change of sea level and its components in the Indo-Pacific region during the altimetry era. J. Geophys. Res. Oceans, 122, 1862–1881, https://doi.org/10.1002/2016JC012345.

    • Search Google Scholar
    • Export Citation

  • Yang, L., L. Zhou, S. Li, and Z. Wei, 2018: Spreading of the South Pacific tropical water and Antarctic intermediate water over the Maritime Continent. J. Geophys. Res. Oceans, 123, 4423–4446, https://doi.org/10.1029/2018JC013831.

    • Search Google Scholar
    • Export Citation

  • Yang, L., R. Murtugudde, S. Zheng, P. Liang, W. Tan, L. Wang, B. Feng, and T. Zhang, 2022: Seasonal variability of the Pacific South Equatorial Current during the Argo era. J. Phys. Oceanogr., 52, 2289–2304, https://doi.org/10.1175/JPO-D-21-0311.1.

    • Search Google Scholar
    • Export Citation

  • Yuan, D. L., and Coauthors, 2011: Forcing of the Indian Ocean dipole on the interannual variations of the tropical Pacific Ocean: Roles of the Indonesian throughflow. J. Climate, 24, 3593–3608, https://doi.org/10.1175/2011JCLI3649.1.

    • Search Google Scholar
    • Export Citation

  • Zhang, L., and T. Qu, 2015: Low-frequency variability of the South Pacific subtropical gyre as seen from satellite altimetry and Argo. J. Phys. Oceanogr., 45, 3083–3098, https://doi.org/10.1175/JPO-D-15-0026.1.

    • Search Google Scholar
    • Export Citation

  • Zhuang, W., S.-P. Xie, D. Wang, B. Taguchi, H. Aiki, and H. Sasaki, 2010: Intraseasonal variability in sea surface height over the South China Sea. J. Geophys. Res., 115, C04010, https://doi.org/10.1029/2009JC005647.

    • Search Google Scholar
    • Export Citation

  • Zhuang, W., B. Qiu, and Y. Du, 2013: Low-frequency western Pacific Ocean sea level and circulation changes due to the connectivity of the Philippine Archipelago. J. Geophys. Res. Oceans, 118, 6759–6773, https://doi.org/10.1002/2013JC009376.

    • Search Google Scholar
    • Export Citation

  • Zuo, H., M. A. Balmaseda, S. Tietsche, K. Mogensen, and M. Mayer, 2019: The ECMWF operational ensemble reanalysis–analysis system for ocean and sea ice: A description of the system and assessment. Ocean Sci., 15, 779–808, https://doi.org/10.5194/os-15-779-2019.

    • Search Google Scholar
    • Export Citation

Supplementary Materials

Save
Cover Journal of Physical Oceanography
Journal of Physical Oceanography

  • Albrecht, F., O. Pizarro, A. Montecinos, and X. Zhang, 2019: Understanding sea level change in the South Pacific during the late 20th century and early 21st century. J. Geophys. Res. Oceans, 124, 3849–3858, https://doi.org/10.1029/2018JC014828.

    • Search Google Scholar
    • Export Citation

  • Andres, M., R. C. Musgrave, D. L. Rudnick, K. L. Zeiden, T. P. Eacock, and J. H. Park, 2020: On the predictability of sea surface height around Palau. J. Phys. Oceanogr., 50, 3267–3294, https://doi.org/10.1175/JPO-D-19-0310.1.

    • Search Google Scholar
    • Export Citation

  • Annamalai, H., S. P. Xie, J. P. McCreary, and R. Murtugudde, 2005: Impact of Indian Ocean sea surface temperature on developing El Niño. J. Climate, 18, 302–319, https://doi.org/10.1175/JCLI-3268.1.

    • Search Google Scholar
    • Export Citation

  • Behringer, D. W., M. Ji, and A. Leetmaa, 1998: An improved coupled model for ENSO prediction and implications for ocean initialization. Part I: The ocean data assimilation system. Mon. Wea. Rev., 126, 1013–1021, https://doi.org/10.1175/1520-0493(1998)126<1013:AICMFE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Bowen, M. M., P. J. H. Sutton, and D. Roemmich, 2006: Wind-driven and steric fluctuations of sea surface height in the southwest Pacific. Geophys. Res. Lett., 33, L14617, https://doi.org/10.1029/2006GL026160.

    • Search Google Scholar
    • Export Citation

  • Bretherton, C. S., M. Widmann, V. P. Dymnikov, J. M. Wallace, and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 1990–2009, https://doi.org/10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Busalacchi, A. J., K. Takeuchi, and J. J. O’Brien, 1983: On the interannual wind-driven response of the tropical Pacific Ocean. Hydrodynamics of the Equatorial Ocean, J. C. J. Nihoul, Ed., Elsevier Oceanography Series, Vol. 36, Elsevier, 155–195, https://doi.org/10.1016/S0422-9894(08)70634-2.

  • Cai, W. J., P. V. Rensch, T. Cowan, and H. H. Hendon, 2011: Teleconnection pathways of ENSO and the IOD and the mechanisms for impacts on Australian rainfall. J. Climate, 24, 3910–3923, https://doi.org/10.1175/2011JCLI4129.1.

    • Search Google Scholar
    • Export Citation

  • Carton, J. A., G. A. Chepurin, and L. Chen, 2018: SODA3: A new ocean climate reanalysis. J. Climate, 31, 6967–6983, https://doi.org/10.1175/JCLI-D-18-0149.1.

    • Search Google Scholar
    • Export Citation

  • Chen, J. L., C. K. Shum, C. R. Wilson, D. P. Chambers, and B. D. Tapley, 2000: Seasonal sea level change from TOPEX/Poseidon observation and thermal contribution. J. Geod., 73, 638–647, https://doi.org/10.1007/s001900050002.

    • Search Google Scholar
    • Export Citation

  • Dewitte, B., G. Reverdin, and C. Maes, 1999: Vertical structure of an OGCM simulation of the equatorial Pacific Ocean in 1985–94. J. Phys. Oceanogr., 29, 1542–1570, https://doi.org/10.1175/1520-0485(1999)029,1542:VSOAOS.2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Duan, J., Y. Li, L. Zhang, and F. Wang, 2020: Impacts of the Indian Ocean dipole on sea level and gyre circulation of the western tropical Pacific Ocean. J. Climate, 33, 4207–4228, https://doi.org/10.1175/JCLI-D-19-0782.1.

    • Search Google Scholar
    • Export Citation

  • Folland, C. K., J. A. Renwick, M. J. Salinger, and A. B. Mullan, 2002: Relative influences of the interdecadal Pacific Oscillation and ENSO on the South Pacific convergence zone. Geophys. Res. Lett., 29, 1643, https://doi.org/10.1029/2001GL014201.

    • Search Google Scholar
    • Export Citation

  • Gordon, A. L., 2005: Oceanography of the Indonesian Seas and their throughflow. Oceanography, 18, 14–27, https://doi.org/10.5670/oceanog.2005.01.

    • Search Google Scholar
    • Export Citation

  • Gouriou, Y., and T. Delcroix, 2002: Seasonal and ENSO variations of sea surface salinity and temperature in the South Pacific convergence zone during 1976–2000. J. Geophys. Res., 107, 8011, https://doi.org/10.1029/2001JC000830.

    • Search Google Scholar
    • Export Citation

  • Hendon, H. H., 2003: Indonesian rainfall variability: Impacts of ENSO and local air–sea interaction. J. Climate, 16, 1775–1790, https://doi.org/10.1175/1520-0442(2003)016<1775:IRVIOE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Ho, C.-R., Q. Zheng, Y. S. Soong, N.-J. Kuo, and J.-H. Hu, 2000: Seasonal variability of sea surface height in the South China Sea observed with TOPEX/Poseidon altimeter data. J. Geophys. Res., 105, 13 981–13 990, https://doi.org/10.1029/2000JC900001.

    • Search Google Scholar
    • Export Citation

  • Izumo, T., and Coauthors, 2010: Influence of the state of the Indian Ocean dipole on the following year’s El Niño. Nat. Geosci., 3, 168–172, https://doi.org/10.1038/ngeo760.

    • Search Google Scholar
    • Export Citation

  • Izumo, T., M. Lengaigne, J. Vialard, J.-J. Luo, T. Yamagata, and G. Madec, 2014: Influence of Indian Ocean dipole and Pacific recharge on following year’s El Niño: Interdecadal robustness. Climate Dyn., 42, 291–310, https://doi.org/10.1007/s00382-012-1628-1.

    • Search Google Scholar
    • Export Citation

  • Izumo, T., J. Vialard, H. Dayan, M. Lengaigne, and I. Suresh, 2015: A simple estimation of equatorial Pacific response from windstress to untangle Indian Ocean dipole and basin influences on El Niño. Climate Dyn., 46, 2247–2268, https://doi.org/10.1007/s00382-015-2700-4.

    • Search Google Scholar
    • Export Citation

  • Jourdain, N. C., M. Lengaigne, J. Vialard, T. Izumo, and A. Sen Gupta, 2016: Further insights on the influence of the Indian Ocean dipole on the following year’s ENSO from observations and CMIP5 models. J. Climate, 29, 637–658, https://doi.org/10.1175/JCLI-D-15-0481.1.

    • Search Google Scholar
    • Export Citation

  • Liu, X., and H. Zhou, 2020: Seasonal variations of the North Equatorial Current across the Pacific Ocean. J. Geophys. Res. Oceans, 125, e2019JC015895, https://doi.org/10.1029/2019JC015895.

    • Search Google Scholar
    • Export Citation

  • Luo, J. J., and T. Yamagata, 2001: Long-term El Niño-Southern Oscillation (ENSO)-like variation with special emphasis on the South Pacific. J. Geophys. Res., 106, 22 211–22 227, https://doi.org/10.1029/2000JC000471.

    • Search Google Scholar
    • Export Citation

  • Masumoto, Y., and Coauthors, 2004: A fifty-year eddy-resolving simulation of the world ocean-preliminary outcomes of OFES (OGCM for the earth simulator). J. Earth Simul., 1, 35–56, https://doi.org/10.32131/jes.1.35.

    • Search Google Scholar
    • Export Citation

  • Menemenlis, D., J. Campin, P. Heimbach, C. Hill, T. Lee, A. Nguyen, M. Schodlock, and H. Zhang, 2008: ECCO2: High resolution global ocean and sea ice data synthesis. Mercator Ocean Quarterly Newsletter, No. 31, Mercator Ocean, Ramonville Saint-Agne, France, 13–21.

  • Piecuch, C. G., and R. M. Ponte, 2011: Mechanisms of interannual steric sea level variability. Geophys. Res. Lett., 38, L15605, https://doi.org/10.1029/2011GL048440.

    • Search Google Scholar
    • Export Citation

  • Qiu, B., and T. M. Joyce, 1992: Interannual variability in the mid-and low-latitude western North Pacific. J. Phys. Oceanogr., 22, 1062–1079, https://doi.org/10.1175/1520-0485(1992)022<1062:IVITMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Qiu, B., and S. Chen, 2006: Decadal variability in the large-scale sea surface height field of the South Pacific Ocean: Observations and causes. J. Phys. Oceanogr., 36, 1751–1762, https://doi.org/10.1175/JPO2943.1.

    • Search Google Scholar
    • Export Citation

  • Qu, T., and E. J. Lindstrom, 2002: A climatological interpretation of the circulation in the western South Pacific. J. Phys. Oceanogr., 32, 2492–2508, https://doi.org/10.1175/1520-0485(2002)032<2492:ACIOTC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation

  • Ramesh, N., and R. Murtugudde, 2013: All flavours of El Niño have similar early subsurface origins. Nat. Climate Change, 3, 42–46, https://doi.org/10.1038/nclimate1600.

    • Search Google Scholar
    • Export Citation

  • Ren, Q., Y. Li, F. Zheng, F. Wang, J. Duan, and R. Li, 2020: Asymmetry of interannual sea level variability in the western tropical Pacific: Responses to El Niño and La Niña. J. Geophys. Res. Oceans, 125, e2020JC016616, https://doi.org/10.1029/2020JC016616.

    • Search Google Scholar
    • Export Citation

  • Roberts, C. D., D. Calvert, N. Dunstone, L. Hermanson, M. D. Palmer, and D. Smith, 2016: On the drivers and predictability of seasonal-to interannual variations in regional sea level. J. Climate, 29, 7565–7585, https://doi.org/10.1175/JCLI-D-15-0886.1.

    • Search Google Scholar
    • Export Citation

  • Roemmich, D., and J. Gilson, 2009: The 2004–2008 mean and annual cycle of temperature, salinity, and steric height in the global ocean from the Argo Program. Prog. Oceanogr., 82, 81–100, https://doi.org/10.1016/j.pocean.2009.03.004.

    • Search Google Scholar
    • Export Citation

  • Roemmich, D., J. Gilson, R. Davis, P. Sutton, S. Wijffels, and S. Riser, 2007: Decadal spinup of the South Pacific subtropical gyre. J. Phys. Oceanogr., 37, 162–173, https://doi.org/10.1175/JPO3004.1.

    • Search Google Scholar
    • Export Citation

  • Stammer, D., 1997: Steric and wind-induced changes in TOPEX/POSEIDON large-scale sea surface topography observations. J. Geophys. Res., 102, 20 987–21 009, https://doi.org/10.1029/97JC01475.

    • Search Google Scholar
    • Export Citation

  • Sun, J., L. Zhang, and D. Hu, 2022: Decadal and long-term variability of sea level in the southwestern Pacific during 1948–2018. Geophys. Res. Lett., 49, e2022GL098747, https://doi.org/10.1029/2022GL098747.

    • Search Google Scholar
    • Export Citation

  • Vivier, F., K. A. Kelly, and L. Thompson, 1999: Contributions of wind forcing, waves, and surface heating to sea surface height observations in the Pacific Ocean. J. Geophys. Res., 104, 20 767–20 788, https://doi.org/10.1029/1999JC900096.

    • Search Google Scholar
    • Export Citation

  • Vivier, F., K. A. Kelly, and M. Harismendy, 2005: Causes of large-scale sea level variations in the Southern Ocean: Analyses of sea level and a barotropic model. J. Geophys. Res., 110, C09014, https://doi.org/10.1029/2004JC002773.

    • Search Google Scholar
    • Export Citation

  • Wang, H., R. Murtugudde, and A. Kumar, 2016: Evolution of Indian Ocean dipole and its forcing mechanisms in the absence of ENSO. Climate Dyn., 47, 2481–2500, https://doi.org/10.1007/s00382-016-2977-y.

    • Search Google Scholar
    • Export Citation

  • Wang, J., J. Li, J. Yin, W. Tan, and Y. Liu, 2021: Sea level seasonal, interannual and decadal variability in the tropical Pacific Ocean. Remote Sens., 13, 3809, https://doi.org/10.3390/rs13193809.

    • Search Google Scholar
    • Export Citation

  • Wu, Q., X. Zhang, J. A. Church, and J. Hu, 2017: Variability and change of sea level and its components in the Indo-Pacific region during the altimetry era. J. Geophys. Res. Oceans, 122, 1862–1881, https://doi.org/10.1002/2016JC012345.

    • Search Google Scholar
    • Export Citation

  • Yang, L., L. Zhou, S. Li, and Z. Wei, 2018: Spreading of the South Pacific tropical water and Antarctic intermediate water over the Maritime Continent. J. Geophys. Res. Oceans, 123, 4423–4446, https://doi.org/10.1029/2018JC013831.

    • Search Google Scholar
    • Export Citation

  • Yang, L., R. Murtugudde, S. Zheng, P. Liang, W. Tan, L. Wang, B. Feng, and T. Zhang, 2022: Seasonal variability of the Pacific South Equatorial Current during the Argo era. J. Phys. Oceanogr., 52, 2289–2304, https://doi.org/10.1175/JPO-D-21-0311.1.

    • Search Google Scholar
    • Export Citation

  • Yuan, D. L., and Coauthors, 2011: Forcing of the Indian Ocean dipole on the interannual variations of the tropical Pacific Ocean: Roles of the Indonesian throughflow. J. Climate, 24, 3593–3608, https://doi.org/10.1175/2011JCLI3649.1.

    • Search Google Scholar
    • Export Citation

  • Zhang, L., and T. Qu, 2015: Low-frequency variability of the South Pacific subtropical gyre as seen from satellite altimetry and Argo. J. Phys. Oceanogr., 45, 3083–3098, https://doi.org/10.1175/JPO-D-15-0026.1.

    • Search Google Scholar
    • Export Citation

  • Zhuang, W., S.-P. Xie, D. Wang, B. Taguchi, H. Aiki, and H. Sasaki, 2010: Intraseasonal variability in sea surface height over the South China Sea. J. Geophys. Res., 115, C04010, https://doi.org/10.1029/2009JC005647.

    • Search Google Scholar
    • Export Citation

  • Zhuang, W., B. Qiu, and Y. Du, 2013: Low-frequency western Pacific Ocean sea level and circulation changes due to the connectivity of the Philippine Archipelago. J. Geophys. Res. Oceans, 118, 6759–6773, https://doi.org/10.1002/2013JC009376.

    • Search Google Scholar
    • Export Citation

  • Zuo, H., M. A. Balmaseda, S. Tietsche, K. Mogensen, and M. Mayer, 2019: The ECMWF operational ensemble reanalysis–analysis system for ocean and sea ice: A description of the system and assessment. Ocean Sci., 15, 779–808, https://doi.org/10.5194/os-15-779-2019.

    • Search Google Scholar
    • Export Citation
  • <sc>Fig</sc>. 1.
    View in gallery
    Fig. 1.

    The first two EOF modes of (a)–(c) seasonal and (d)–(f) interannual SLAs (cm) in the southwest Pacific. Spatial patterns are shown in (a), (b), (d), and (e), with the explained variances in parentheses. The corresponding principal components are shown in (c) and (f) (solid black line for PC1 and dashed gray line for PC2). Niño-3.4 (light red and blue shading) and DMI (green line) are superimposed. The correlations between principal components and climate indices are displayed in Fig. 2.

  • <sc>Fig</sc>. 2.
    View in gallery
    Fig. 2.

    Lag correlations between the (a) first and (b) second principal component and Niño-3.4 (red lines) and DMI (blue lines). The solid (dashed) lines represent the correlations after (without) removing the linear dependence on the other climate index. (c) Partial regressions of observational SSHs on Niño-3.4 index at the 95% confidence level, with the Niño-3.4 index leading by four months. (d) Explained variance associated with Niño-3.4 index (only exhibiting the values where the partial regressions pass the 95% confidence level). All the time series are detrended and bandpass filtered (18–84 months).

  • <sc>Fig</sc>. 3.
    View in gallery
    Fig. 3.

    Correlation coefficients between variables from ORAS5 and observations over January 1993–December 2018. (a) Comparison of SSHs with AVISO altimetry data. (b)–(i) Comparisons of temperatures (T) and salinities (S) with Argo data.

  • <sc>Fig</sc>. 4.
    View in gallery
    Fig. 4.

    Explained variance of (a),(b) barotropic (mass-induced) and (c)–(f) baroclinic components of SLAs. (left) Seasonal and (right) interannual variability. The first baroclinic mode is shown in (c) and (d), and the summation of the second to sixth baroclinic modes is shown in (e) and (f). Note that the barotropic component, derived from GRACE ocean bottom pressure measurements, is from 2003 to 2018. The baroclinic component, calculated based on vertical mode decomposition of baroclinic pressure anomalies, is from 1993 to 2018.

  • <sc>Fig</sc>. 5.
    View in gallery
    Fig. 5.

    (a)–(c) Seasonal phase correlation between observations and model simulations from (a) EXP0 (control run), (b) EXP1Seasonal (no seasonal cycle of wind stress east of the date line over 35°–5°S), and (c) EXP2Seasonal (no seasonal cycle of wind stress east of 160°E over 35°–5°S). (d)–(f) Proportion of seasonal variance explained by the model experiments. The black contours with dots mark the critical correlation coefficient (0.55) passing the 95% confidence level test.

  • <sc>Fig</sc>. 6.
    View in gallery
    Fig. 6.

    (a)–(c) Interannual phase correlation between observations and model simulations from (a) EXP0 (control run), (b) EXP1Interannual (no interannual signal in the Pacific wind stress over 35°–5°S and 180°–70°W), and (c) EXP2Interannual (no interannual signal in the Pacific wind stress over 35°S–0° and 180°–70°W). (d)–(f) Proportion of interannual variance explained by the model experiments. The black contours with dots mark the critical correlation coefficient (0.60) passing the 95% confidence level test.

  • <sc>Fig</sc>. 7.
    View in gallery
    Fig. 7.

    Composite SLAs (cm) in EXP1interannual during El Niño years along (a) the equator, (b) the eastern boundary, and (c) 5°S. Month 0 represents the one when Niño-3.4 reaches the maximum during each El Niño event. Months −12 to −1 (1–12) represent those leading (lagging) month 0 by corresponding months.

  • <sc>Fig</sc>. 8.
    View in gallery
    Fig. 8.

    (a) Seasonal and (c) interannual correlations between the Ekman-pumping-induced and observational SLAs. Proportions of (b) seasonal and (d) interannual variances explained by the Ekman pumping dynamics. The black contours with dots mark the critical correlation coefficient (0.55 for seasonal and 0.60 for interannual correlations) passing the 95% confidence level test.

  • <sc>Fig</sc>. 9.
    View in gallery
    Fig. 9.

    (a),(b) Seasonal and (e),(f) interannual correlations between thermosteric/halosteric and observational SLAs. (c),(d) Seasonal and (g),(h) interannual corresponding explained variances. (left) Thermosteric and (right) halosteric components and total sea level variations.

  • <sc>Fig</sc>. 10.
    View in gallery
    Fig. 10.

    Contributions of (a),(e) local sea surface heat fluxes, (b),(f) horizontal, and (c),(g) vertical temperature advection, and (d),(h) their summation to the seasonal sea level variations based on ECCO2. (top) The correlation coefficient between each component and observational sea level variations. (bottom) The corresponding explained variance. The black contours with dots mark the critical correlation coefficient (0.55) passing the 95% confidence level test.

  • <sc>Fig</sc>. 11.
    View in gallery
    Fig. 11.

    As in Fig. 10, but based on OFES simulations.

  • <sc>Fig</sc>. 12.
    View in gallery
    Fig. 12.

    As in Fig. 10, but for the interannual sea level variability.

  • <sc>Fig</sc>. 13.
    View in gallery
    Fig. 13.

    As in Fig. 12, but based on OFES simulations.

Fig. 1.

The first two EOF modes of (a)–(c) seasonal and (d)–(f) interannual SLAs (cm) in the southwest Pacific. Spatial patterns are shown in (a), (b), (d), and (e), with the explained variances in parentheses. The corresponding principal components are shown in (c) and (f) (solid black line for PC1 and dashed gray line for PC2). Niño-3.4 (light red and blue shading) and DMI (green line) are superimposed. The correlations between principal components and climate indices are displayed in Fig. 2.
Fig. 2.

Lag correlations between the (a) first and (b) second principal component and Niño-3.4 (red lines) and DMI (blue lines). The solid (dashed) lines represent the correlations after (without) removing the linear dependence on the other climate index. (c) Partial regressions of observational SSHs on Niño-3.4 index at the 95% confidence level, with the Niño-3.4 index leading by four months. (d) Explained variance associated with Niño-3.4 index (only exhibiting the values where the partial regressions pass the 95% confidence level). All the time series are detrended and bandpass filtered (18–84 months).
Fig. 3.

Correlation coefficients between variables from ORAS5 and observations over January 1993–December 2018. (a) Comparison of SSHs with AVISO altimetry data. (b)–(i) Comparisons of temperatures (T) and salinities (S) with Argo data.
Fig. 4.

Explained variance of (a),(b) barotropic (mass-induced) and (c)–(f) baroclinic components of SLAs. (left) Seasonal and (right) interannual variability. The first baroclinic mode is shown in (c) and (d), and the summation of the second to sixth baroclinic modes is shown in (e) and (f). Note that the barotropic component, derived from GRACE ocean bottom pressure measurements, is from 2003 to 2018. The baroclinic component, calculated based on vertical mode decomposition of baroclinic pressure anomalies, is from 1993 to 2018.
Fig. 5.

(a)–(c) Seasonal phase correlation between observations and model simulations from (a) EXP0 (control run), (b) EXP1Seasonal (no seasonal cycle of wind stress east of the date line over 35°–5°S), and (c) EXP2Seasonal (no seasonal cycle of wind stress east of 160°E over 35°–5°S). (d)–(f) Proportion of seasonal variance explained by the model experiments. The black contours with dots mark the critical correlation coefficient (0.55) passing the 95% confidence level test.
Fig. 6.

(a)–(c) Interannual phase correlation between observations and model simulations from (a) EXP0 (control run), (b) EXP1Interannual (no interannual signal in the Pacific wind stress over 35°–5°S and 180°–70°W), and (c) EXP2Interannual (no interannual signal in the Pacific wind stress over 35°S–0° and 180°–70°W). (d)–(f) Proportion of interannual variance explained by the model experiments. The black contours with dots mark the critical correlation coefficient (0.60) passing the 95% confidence level test.
Fig. 7.

Composite SLAs (cm) in EXP1interannual during El Niño years along (a) the equator, (b) the eastern boundary, and (c) 5°S. Month 0 represents the one when Niño-3.4 reaches the maximum during each El Niño event. Months −12 to −1 (1–12) represent those leading (lagging) month 0 by corresponding months.
Fig. 8.

(a) Seasonal and (c) interannual correlations between the Ekman-pumping-induced and observational SLAs. Proportions of (b) seasonal and (d) interannual variances explained by the Ekman pumping dynamics. The black contours with dots mark the critical correlation coefficient (0.55 for seasonal and 0.60 for interannual correlations) passing the 95% confidence level test.
Fig. 9.

(a),(b) Seasonal and (e),(f) interannual correlations between thermosteric/halosteric and observational SLAs. (c),(d) Seasonal and (g),(h) interannual corresponding explained variances. (left) Thermosteric and (right) halosteric components and total sea level variations.
Fig. 10.

Contributions of (a),(e) local sea surface heat fluxes, (b),(f) horizontal, and (c),(g) vertical temperature advection, and (d),(h) their summation to the seasonal sea level variations based on ECCO2. (top) The correlation coefficient between each component and observational sea level variations. (bottom) The corresponding explained variance. The black contours with dots mark the critical correlation coefficient (0.55) passing the 95% confidence level test.