Contributions of Surface Heat Fluxes and Oceanic Processes to Tropical SST Changes: Seasonal and Regional Dependence

Zhuoqi He State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, China

Search for other papers by Zhuoqi He in
Current site
Google Scholar
PubMed
Close
,
Renguang Wu Center for Monsoon System Research, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

Search for other papers by Renguang Wu in
Current site
Google Scholar
PubMed
Close
,
Weiqiang Wang State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, China

Search for other papers by Weiqiang Wang in
Current site
Google Scholar
PubMed
Close
,
Zhiping Wen Center for Monsoon and Environment Research, School of Atmospheric Sciences, Sun Yat-sen University, Guangzhou, China

Search for other papers by Zhiping Wen in
Current site
Google Scholar
PubMed
Close
, and
Dongxiao Wang State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, China

Search for other papers by Dongxiao Wang in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

The present study employs six surface heat flux datasets and three ocean assimilation products to assess the relative contributions of surface heat fluxes and oceanic processes to the sea surface temperature (SST) change in the tropical oceans. Large differences are identified in the major terms of the heat budget equation. The largest discrepancies among different datasets appear in the contribution of vertical advection. The heat budget is nearly balanced in the shortwave-radiation- and horizontal-advection-dominant cases but not balanced in some of the latent-heat-flux- and vertical-advection-dominant cases. The contributions of surface heat fluxes and ocean advections to the SST tendency display remarkable seasonal and regional dependence. The contribution of surface heat fluxes covers a large geographical area. The oceanic processes dominate the SST tendency in the near-equatorial regions with large values but small spatial scales. In the Pacific and Atlantic Oceans, the SST tendency is governed by the dynamic and thermodynamic processes, respectively, while a wide variety of processes contribute to the SST tendency in the Indian Ocean. Several regions have been selected to illustrate the dominant contributions of individual terms to the SST tendency in different seasons. The seasonality and regionality of the interannual air–sea relationship indicate a physical connection with the mean state.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: Weiqiang Wang, weiqiang.wang@scsio.ac.cn

Abstract

The present study employs six surface heat flux datasets and three ocean assimilation products to assess the relative contributions of surface heat fluxes and oceanic processes to the sea surface temperature (SST) change in the tropical oceans. Large differences are identified in the major terms of the heat budget equation. The largest discrepancies among different datasets appear in the contribution of vertical advection. The heat budget is nearly balanced in the shortwave-radiation- and horizontal-advection-dominant cases but not balanced in some of the latent-heat-flux- and vertical-advection-dominant cases. The contributions of surface heat fluxes and ocean advections to the SST tendency display remarkable seasonal and regional dependence. The contribution of surface heat fluxes covers a large geographical area. The oceanic processes dominate the SST tendency in the near-equatorial regions with large values but small spatial scales. In the Pacific and Atlantic Oceans, the SST tendency is governed by the dynamic and thermodynamic processes, respectively, while a wide variety of processes contribute to the SST tendency in the Indian Ocean. Several regions have been selected to illustrate the dominant contributions of individual terms to the SST tendency in different seasons. The seasonality and regionality of the interannual air–sea relationship indicate a physical connection with the mean state.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: Weiqiang Wang, weiqiang.wang@scsio.ac.cn

1. Introduction

As a large thermal and energy reservoir, the tropical ocean plays an important role in global climate variability and predictability (Klein et al. 1999; Lau and Nath 2003; Wu and Kirtman 2005). The air–sea interaction processes generate sea surface temperature (SST) anomalies in the tropics, which modify tropical heating and, in turn, affect climate in neighboring and remote regions through atmospheric teleconnections. Thus, studying the tropical SST variability and tropical air–sea interaction processes may help us understand climate variability and improve regional climate prediction.

The SST provides an essential communication between the upper ocean and the lower atmosphere via surface turbulent and radiative fluxes. The interannual SST variability is governed by both atmospheric processes (the cloud–radiation effect and the wind–evaporation effect) and oceanic processes (horizontal advection and vertical advection) (Qu 2001; Deser et al. 2010; He and Wu 2013a,b). The thermodynamic processes are important to the SST change. Enhanced (reduced) incoming shortwave radiation associated with less (more) cloudiness is one of the major factors in warming (cooling) the upper ocean, hereafter named the cloud–radiation effect (Norris and Leovy 1994; Klein et al. 1999; Xie 2004). The radiation contribution to the SST change has been noticed in the tropical Indian Ocean (TIO) during spring (Klein et al. 1999; Wu and Kirtman 2007), in the South China Sea (SCS) and northwest Pacific during summer (Klein et al. 1999; Trenberth and Shea 2005; Wu and Kirtman 2007; He and Wu 2013b), in the southeast Indian Ocean during late winter (Wu and Kirtman 2007), in the tropical North Atlantic during spring (Klein et al. 1999), and in the equatorial Pacific during the El Niño–Southern Oscillation (ENSO) developing phase (Klein et al. 1999; Lau and Nath 2003). Surface wind changes can affect SST via modification of surface evaporation, termed the wind–evaporation effect (Wu et al. 2006; He and Wu 2013b). This process is dominant in the midlatitude regions during cold seasons (Cayan 1992; Wu and Kinter 2010; Deser et al. 2010). It is also a crucial process in the development of the Indian Ocean asymmetric mode (Wu et al. 2008; Wu and Yeh 2010) and the seasonal SST evolution in the South China Sea (Chen et al. 2003; He and Wu 2013a). In addition, the upper-ocean heat transportation, including the horizontal and vertical advection, plays a nonnegligible role in the SST variability. These ocean dynamic processes are associated with changes in ocean currents, thermodynamic structure, and the mixed layer depth (MLD) (Qu 2001; Tomita and Nonaka 2006; Deser et al. 2010). The heat transportation is demonstrated to be prominent in the South China Sea during winter and summer (Qu 2001; He and Wu 2013b), and in the coastal North Atlantic and the central and coastal North Pacific during winter (Liu et al. 2005; Deser et al. 2010; Wu and Kinter 2010). It is also important in the development of the Indian Ocean dipole mode (Saji et al. 1999) and the tropical Indian Ocean south–north asymmetric mode (Wu and Yeh 2010).

Previous studies indicate that the tropical SST variation may be governed by different processes in different regions and seasons, but most of these studies only focused on some specified cases, regions, or processes. For example, Trenberth and Shea (2005) and Wu and Kirtman (2007) indicated that the rainfall–cloud–SST relationship displays an obvious seasonality. Cayan (1992) addressed the importance of latent and sensible heat fluxes in driving SST anomalies over the northern oceans. Kitoh et al. (1999) identified different contributions of latent heat flux and shortwave radiation to SST anomalies in the Pacific Ocean. These studies mainly highlighted the importance of the thermodynamic processes without comparison to the dynamic contributions. He and Wu (2013b) underscored the combined effects of both atmospheric and oceanic processes yet only focused on the South China Sea. There is still a lack of systematic research on how the tropical SST responds to dynamic and thermodynamic processes in different regions and seasons. The first purpose of the present study is to identify the relative roles of atmospheric and oceanic processes in the interannual SST variation in the tropical oceans. Typical processes that are seldom mentioned in previous studies will be selected for illustration of different cases.

Furthermore, there may be discrepancies among different datasets as they are derived through a variety of retrieval algorithms, based on different data assimilation systems, or using different input data with different resolutions (Bony et al. 1997; Gulev et al. 2000; Bourras 2006; Large and Yeager 2009; Smith et al. 2011). Bourras (2006) and Smith et al. (2011) compared several turbulent heat flux products and noted their agreements and disagreements. The differences among datasets may lead to spurious information in the heat budget analysis. However, there are seldom comments on the applicability of these datasets to the surface heat budget in the study of interannual air–sea relationship. To understand the uncertainties and to improve the reliability of obtained results, it is necessary to compare the performance of different datasets regarding the contributions of different terms to the SST change and understand the diversities and similarities among different surface heat flux and ocean products.

The present study elucidates the diversity of atmospheric and oceanic processes, contrasts their relative contributions to the interannual variations of tropical SST, and evaluates the performance of six air–sea flux datasets and three ocean products in estimating the surface heat budget. Section 2 describes the datasets and methods. Section 3 documents the relative importance of dynamic and thermodynamic processes and compares the performance of different products. Section 4 utilizes five typical cases to explore how atmospheric and oceanic processes contribute to the regional SST tendency. Summary and discussion is presented in section 5.

2. Datasets and methods

a. Datasets

The present study employs the following datasets for regression and correlation analysis in sections 3 and 4: 1) Monthly mean SST from the National Oceanic and Atmospheric Administration (NOAA) optimum interpolation (OI), version 2 (Reynolds et al. 2002; http://www.esrl.noaa.gov/psd/), on a 1° × 1° grid from December 1981 to September 2015. 2) Monthly mean precipitation from the Global Precipitation Climatology Project (GPCP), version 2.2 (Adler et al. 2003; http://www.esrl.noaa.gov/psd/), on a 2.5° × 2.5° grid for the period of January 1979 to August 2015. 3) Monthly mean 10-m wind from the National Centers for Environmental Prediction–Department of Energy (NCEP–DOE) Reanalysis 2 (Kanamitsu et al. 2002; http://www.esrl.noaa.gov/psd/) on a Global T62 Gaussian grid from 1979 to July 2015. 4) Monthly mean net surface shortwave radiation, net surface longwave radiation, surface latent heat flux, surface sensible heat flux, 10-m wind speed, SST, and 2-m air specific humidity from the Woods Hole Oceanographic Institute (WHOI) objectively analyzed air–sea fluxes (OAFlux) on a 1° × 1° grid (Yu et al. 2008; http://oaflux.whoi.edu). The shortwave radiation covers the period from July 1983 to December 2009. Other variables are available from January 1958 to September 2014. 5) Monthly mean horizontal current, geometric vertical velocity, and potential temperature from the NCEP Global Ocean Data Assimilation System (GODAS) reanalysis data (Ji et al. 1995; http://www.esrl.noaa.gov/psd/) from January 1980 to October 2014. It has a horizontal resolution of ⅓° latitude × 1.0° longitude grid and 40 vertical levels with a 10-m resolution in the upper 200 m.

In section 4a, six surface heat flux datasets and three ocean assimilation products are used to assess the relative contributions of heat fluxes and oceanic advection to the SST change. They include the NCEP–DOE Reanalysis 2; the WHOI OAFlux; the GODAS reanalysis data; the National Oceanography Centre, Southampton (NOCS), surface flux dataset, version 2.0 (Berry and Kent 2009; http://rda.ucar.edu/datasets/ds260.3); the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) (Kållberg et al. 2004; http://apps.ecmwf.int/datasets/); the Clouds and the Earth’s Radiant Energy System (CERES) SYN1deg-Month Ed3A product (Wielicki et al. 1996; http://ceres.larc.nasa.gov/order_data.php); the Goddard Satellite-based Surface Turbulent Fluxes (GSSTF), version 3, dataset products (Shie et al. 2012; ftp://measures.gsfc.nasa.gov/data/s4pa/GSSTF/GSSTFM.3); the GFDL ensemble coupled data assimilation (ECDA) product, version 2.0 (Zhang et al. 2007; https://www.gfdl.noaa.gov/ocean-data-assimilation-model-output); and the Simple Ocean Data Assimilation (SODA) reanalysis, version 2.2.4 (Carton and Giese 2008; http://sodaserver.tamu.edu/assim/SODA_2.2.4/). The information about these products is given in Table 1.

Table 1.

A list of six surface heat flux and three ocean assimilation products with corresponding variables, periods, and resolution for estimating the contributions of heat flux and ocean advection to SST tendency based on the heat budget analysis. The variables SST, SWR, LHF, T, v, and w denote the sea surface temperature, shortwave radiation, latent heat flux, oceanic potential temperature, horizontal current, and vertical current, respectively.

Table 1.

b. Methods

This study uses monthly mean anomalies obtained by subtracting climatological seasonal cycle from individual monthly means. Linear interpolation is adopted to convert all the variables to 1° horizontal resolution in accordance with the SST resolution. Correlation and regression analyses are employed to document the interannual atmosphere–ocean relationships [Eqs. (1) and (2)]. The Student’s t test is applied for the statistical significance test, and the critical values at the 90% confidence level may differ as the datasets used in the present study cover different periods.

The regression and correlation coefficients are calculated based on the following formula:
e1
e2
where xi is the time series of the contribution term, SSTdi is the time series of SST tendency, Rex is the regression coefficient, and Corx is the correlation coefficient.
The heat budget for the upper-ocean mixed layer (He and Wu 2013a,b) is adopted to assess the contributions of different terms to the SST tendency:
e3
where ρs is the density of seawater, cp is the specific heat of seawater, h is the MLD, Qnet is net surface heat flux, is the upper-ocean heat transportation, RES is the residual term that represents the sum of frictional diffusion and interior oceanic turbulent heating, T is the ocean temperature within the mixed layer, which is approximately equal to SST and thus replaced by SST on the left side of the equation, and ∂T/∂t denotes the SST tendency, which is calculated as the difference of SST in the succeeding month minus SST in the preceding month and divided by 2. It should be noted that the MLD is defined as the lowest depth with the ocean temperature differing from SST by 0.5°C, and varying this criterion between 0.5° and 1.0°C has little effect on the heat budget estimation at the interannual time scale.
In this study, we expand the right-hand side of Eq. (3) and regress the main components against the left-hand side (SST tendency). Here, and Qnet can be expanded as follows:
e4
e5
e6
The term includes the upper-ocean horizontal advection (ADV) and vertical advection (ADW), v is the horizontal current, w is the vertical current, and denotes the vertical mean of oceanic advection within the mixed layer. The ADW should include the effects of vertical entrainment. The term Qnet includes net shortwave radiation SWR, latent heat flux LHF, net longwave radiation LWR, and sensible heat flux SHF. For SWR, we consider the effect of shortwave radiation penetration through MLD SWRpen, which is calculated following the solar radiation penetration parameterization scheme (Paulson and Simpson 1977) with the magnitude R = 0.58 and the penetration depth Li (i = 1, 2) set as L1 = 0.35 m and L2 = 23 m. We adopt SWRin and LHF directly from the flux datasets. According to Reed (1977), the incoming surface shortwave radiation SWRin is affected by the fractional cloud cover, and this is the way for the cloud–radiation effect to work on the SST change. Following the study by Liu et al. (1979), LHF is determined by both the air–sea humidity difference and surface wind speed, and this is how the wind–evaporation effect can affect the SST change.

For convenience of comparison among the SST tendency, heat flux, and ocean advection, we use the convention of positive values for surface heat flux and ocean advection when they favor the SST warming. In the following discussion, SWR and LHF will be used to represent the atmospheric processes and ADV and ADW to represent the oceanic processes.

3. Interannual atmosphere–ocean relationship in the tropical oceans

We first examine the local relationship of the SST change with surface heat flux and ocean advection. Figure 1 presents the pointwise simultaneous regressions of surface heat flux and ocean advection with respect to the anomalous SST tendency during March–April–May (MAM), June–July–August (JJA), September–October–November (SON), and December–January–February (DJF), respectively, for the period 1984–2009. Note that surface heat flux has been divided by climatological monthly mean MLD ρscph to convert it to the unit of °C month−1. The positive values in Fig. 1 indicate that an increasing heat flux into the ocean or a warm ocean advection is associated with an increase in SST. For example, given SST anomaly increases by 1°C month−1, an Rex between heat flux and SST tendency with the value of 0.7 means that the heat flux anomaly equivalently accounts for 70% of this SST change.

Fig. 1.
Fig. 1.

Pointwise simultaneous regression coefficients Rex with respect to seasonal mean SST tendency: (top) OAFlux net heat flux and (bottom) GODAS ocean advection in (a),(e) MAM, (b),(f) JJA, (c),(g) SON, and (d),(h) DJF for the period 1984–2009. Thick contours indicate that the correlations are statistically significant at the 90% confidence level with the sample size of 26.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

The contributions of surface heat flux and ocean advection to the SST tendency display remarkable regional features and seasonal dependence. The surface heat flux plays an important role in regulating SST in off-equatorial regions. A large Rex of heat flux is observed around the Maritime Continent (MC) in MAM (Fig. 1a), in the western North Pacific (WNP) in JJA (Fig. 1b), in the central North Pacific and the North Atlantic in SON (Fig. 1c), and in the South Pacific in DJF (Fig. 1d). This seasonal evolution of heat flux–SST tendency relationship coincides with the south–north migration of regions with active convection and large surface wind variance (Wu and Kirtman 2007). A negative SST tendency–heat flux relationship is detected in the equatorial oceans (Figs. 1a–d), which indicates a damping effect of surface heat flux on the SST or an SST forcing heat flux process (Wu and Kirtman 2007). In contrast, the contribution of ocean advection to the SST change is significant mainly between 15°S and 15°N, such as the central North Pacific (7.5°–15°N, 180°E–130°W) in JJA (Fig. 1f), the western Indian Ocean (WIO) in SON (Fig. 1g), and the western Pacific in DJF (Fig. 1h), and it is dominant in the south Indian Ocean and the equatorial Pacific–Atlantic from MAM to SON (Figs. 1e–g) where there is strong SST forcing of atmosphere. The ocean dynamic contributions have a relatively small scale, probably owing to local features of the oceanic processes. Another feature to be noted here is that a significant difference in both flux and advection contribution exists between winter and other seasons in the equatorial Pacific Ocean, which is probably associated with the ENSO phase locking.

As introduced in section 1, precipitation and surface wind can imprint upon SST via alteration of surface SWR and LHF, namely, the thermodynamic processes. Meanwhile, the dynamic (i.e., oceanic) processes transport heat both horizontally and vertically. In the following, the four terms SWR, LHF, ADV, and ADW will be highlighted to elucidate the relative importance of atmospheric and oceanic processes in interannual SST variations. Considering the spatial consistency of the SST tendency (figures not shown) and its relationship with surface heat flux and ocean advection (Fig. 1), 10 regions are selected for further comparison (Fig. 2). The NIO (5°–20°N, 60°–100°E), the WNP (0°–20°N, 110°–160°E), the MC (15°S–0°, 105°–160°E), and the North Atlantic Ocean (NAO; 10°–20°N, 90°–30°W) are selected for their representations of the monsoon regions. The WIO (0°–10°N, 45°–60°E), the western south Indian Ocean (WSIO; 12°S–0°, 50°–80°E), the eastern south Indian Ocean (ESIO; 15°S–0°, 80°–105°E), the western equatorial Pacific (WEP; 5°S–5°N, 160°E–180°), the eastern equatorial Pacific (EEP; 5°S–5°N, 160°–90°W), and the equatorial Atlantic Ocean (EAO; 5°S–2°N, 40°–10°W) are the regions with active oceanic processes. It should be noted that the ocean vertical advection is generally a local process, and it works on a much smaller spatial scale with higher variance than the atmospheric process. Thus, to highlight the role of ADW, we narrow down to the following regions when calculating the area-mean ADW–SST tendency correlation (dashed boxes in Fig. 2): WSIO (10°–5°S, 50°–80°E), ESIO (5°S–0°, 95°–105°E), WEP (0.5°S–0.5°N, 160°E–180°), and EEP (0.5°S–0.5°N, 160°–90°W).

Fig. 2.
Fig. 2.

The 10 regions for calculating the area-mean correlation coefficients Corx between the SST tendency and the heat flux and ocean advection. The dashed boxes represent regions used for calculating the vertical advection–SST tendency correlation.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

The seasonal evolution of the relationship between the SST tendency and the four atmospheric–oceanic processes is illustrated in Fig. 3, which shows the area-mean correlations of the SST tendency with SWR, LHF, ADV, and ADW in the corresponding 10 regions. Distinct features are noted in different regions. In the equatorial Pacific (WEP and EEP), the oceanic advection contribution is prominent with supplementary heat flux contributions (Figs. 3a,b). In the EEP, the ADV and the ADW are highly correlated with the SST tendency throughout the whole year except for winter when alternatively the SWR (LHF)–SST tendency relationship is significant (Fig. 3b). The ADV works significantly in the WEP as well, but the relationship between ADW and SST tendency is largely weakened and the LHF–SST tendency correlation is enhanced in winter (Fig. 3a). The results provide evidence for the involvement of ocean heat transportation associated with strong west–east SST gradient and distinguished equatorial upwelling in the ENSO variability (Bjerknes 1969; Battisti and Hirst 1989; Jin 1997; Picaut et al. 1997; Weisberg and Wang 1997). In the western Pacific (WNP and MC), the relationship is in contrast to that in the equatorial Pacific. There are significant positive LHF–SST tendency and SWR–SST tendency relationships, whereas the oceanic contributions are negligible (Figs. 3c,d). One feature to be noticed is that the positive SWR–SST tendency correlation reaches the maximum in the WNP from April to November and in the MC from October to April. This feature indicates that the cloud–radiation effect tends to follow the evolution of the local rainy season (Wu and Kirtman 2007; Lau and Nath 2009).

Fig. 3.
Fig. 3.

Annual evolution of 3-month running mean Corx between the SST tendency and OAFlux shortwave radiation (black), OAFlux latent heat flux (red), GODAS horizontal advection (green), and GODAS vertical advection (blue) over the 10 regions as presented in Fig. 2 for the period 1984–2009. Marks indicate that correlations are statistically significant at the 90% confidence level.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

Different from the Pacific Ocean, there are a wide variety of air–sea relationships in the Indian Ocean (Figs. 3e–h). The atmospheric and oceanic processes can exert either individual or combined influence on regional SST change. In the WIO, the SST tendency displays a high correlation with both ADV and LHF from January to July (Fig. 3e). The atmospheric process is active in the NIO, accompanied by relatively high ADW contributions during the Indian summer monsoon onset (Fig. 3f). In the south Indian Ocean (WSIO and ESIO), the atmospheric and oceanic processes play alternative roles. The highest correlation between the SST tendency and SWR (LHF) occurs in austral summer, while the SST tendency is closely related to ocean advection in boreal summer and autumn (Figs. 3g,h).

In the Atlantic Ocean, the air–sea relationship displays features similar to those in the equatorial and western Pacific Ocean (Figs. 3i–j). The ADV–SST tendency correlation is high in the equatorial region, and the SWR (LHF)–SST tendency correlation is dominant north of the equator. However, compared with the Pacific Ocean, both the ADV–SST tendency correlation and the seasonal dependence of SWR–SST tendency relationship are less significant, probably owing to a smaller land–sea thermal contrast in association with a smaller-scale ocean basin. Correlation analysis is applied to examine the relationships between the SST tendency and different terms, yet it should be noted that the term with high Corx does not necessarily mean that its contribution to the SST tendency is high. Hence, regression analysis is employed to estimate the relative importance of different processes in the next section.

4. Analysis of selected cases

a. Comparison of regional heat budget among different datasets

The diverse air–sea relationships (Fig. 3) verify the seasonal and regional dependence of the contributions of atmospheric and oceanic processes to the SST change. Accordingly, we summarize the typical cases for regional heat budget analysis. The corresponding regions and periods are outlined in Fig. 4. Figure 4a represents the heat-flux-dominated cases, Fig. 4b the oceanic-advection-dominated cases, and Fig. 4c the coworking cases. The ADW contributions are calculated over smaller regions to underscore their influences in the June–November (JJASON) EEP case, the AM EP case, and the AM NIO case (see Fig. 4).

Fig. 4.
Fig. 4.

The 23 cases with corresponding regions and periods for estimating the contributions of heat flux and ocean advection to SST tendency based on the heat budget analysis. The boxes indicate the regions. The months are noted in Arabic numerals within the boxes or at the bottom of each panel (e.g., “1–3” means the period from January to March). The cases with different dominant processes are distinguished in different colors. Note that the ADW contributions are estimated over smaller regions (dashed boxes) in three cases.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

Figure 5 presents the upper-ocean heat budget estimation [Eqs. (3)(5)] for the 23 selected cases based on the area-mean regression of SWR, LHF, ADV, and ADW with respect to the anomalous SST tendency. Heat flux derived from the OAFlux, NOCS flux, ERA-40, NCEP–DOE Reanalysis 2, CERES, and GSSTF and ocean advection derived from the GFDL ECDA, GODAS, and SODA are included for comparison with the columns standing for the mean values and the marks for each dataset (for details refer to Table 1). The LWR and SHF are not shown in Fig. 5 owing to their insignificant contributions in these selected cases. Note that the surface heat flux and ocean datasets are derived through various methods using different input data, and thus it is difficult for the heat budget to be fully balanced in the present study. However, the heat budget estimation can still provide information about the relative importance of atmospheric and oceanic processes for the interannual SST variations, as well as the relative performance of different datasets in revealing the heat budget variability.

Fig. 5.
Fig. 5.

Heat budget estimation for the corresponding cases in Table 1: area-mean Rex of shortwave radiation, latent heat flux, horizontal advection, and vertical advection onto anomalous SST tendency. The units are converted into °C month−1. Marks stand for different datasets with columns for the mean value. The abbreviation of each case is given on the x axis. The analysis covers the period 1984–2007 for the OAFlux, NOCS flux, NCEP–DOE Reanalysis 2, GFDL ECDA, GODAS, and SODA; 1988–2008 for GSSTF; 1982–2001 for ERA-40; and 2001–13 for CERES.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

Here, we set the criterion for a significant contribution term as the Rex exceeding 0.25. According to Fig. 5, there are a total of four SWR-dominant cases [April–May (AM) WNP, JJASON WNP, October–December (OND) MC, and December–March (DJFM) TIO], four LHF-dominant cases [DJF WNP, May–July (MJJ) MC, November–February (NDJF) NIO, and DJFM NAO], three SWR- and LHF-dominant cases [January–March (JFM) EEP, JFM MC, and July–October (JASO) NAO], six ADV-dominant cases (JFM WEP, April–July (AMJJ) NCP, JFM WIO, June–September (JJAS) WSIO, JASO ESIO, and July–December (JASOND) EAO], two ADW-dominant cases (SON NCP and MAM WSIO), one ADV- and ADW-dominant case (JJASON EEP), and three coworking cases (JJA ESIO, AM EP, and AM NIO). Given that the SST tendency increases by 1°C month−1 and by comparing all these cases, it can be found that the oceanic advection can attain the maximal influence on the SST tendency as large as 1.05°C month−1 in the JJASON EEP case, while the atmospheric processes are weaker with the maximal contribution of 0.45°C month−1 in the OND MC case. The residual term can be deduced by subtracting the four dominant terms from 1°C month−1. The heat budget is mainly balanced with low residual terms in the SWR-dominant and ADV-dominant cases, but the results do not satisfy our expectations in some LHF-dominant and ADW-dominant cases in which the residual terms can be as large as the leading term, such as the DJF WNP case, the NDJF NIO case, the SON NCP case, and the MAM WSIO case. The reasons for this imbalance may be partly owing to the following reasons. The LHF anomalies are determined by both surface wind speed and air–sea humidity difference, which may likely increase the uncertainties and lead to weak LHF in the heat budget (Bourras 2006; Berry and Kent 2009; Smith et al. 2011). In most cases, the ocean upwelling is a local process with high temporal variations. It is difficult to ensure the integrity of ADW signals based on monthly data with relatively low spatial resolutions.

Comparing the estimations based on different datasets in Fig. 5, it is apparent that the difference among the datasets is large for estimation of the major terms, and there is a high level of consistency in computing the minor terms. For example, in the AM WNP case, the SWR contribution, with a mean value of 0.38°C month−1, ranges from 0.48°C month−1 in the OAFlux to 0.08°C month−1 in the NCEP–DOE, while the minor term—the ADW—differs no more than 0.02°C month−1. Meanwhile, the products show larger discrepancy in estimating the ADW than the other three heat budget terms, which may be related to the small spatial scale of the vertical mixing process. For the SWR, the ERA-40 has larger estimations while the NOCS flux presents lower values compared with other products. The LHF term is noticeably larger in GSSTF than in other datasets. The ocean products display no significant coherence in retrieving the ADV and ADW, which implies a high level of uncertainty in the estimates. Overall, it is hard to conclude which product is most applicable for estimating the contributions to the SST tendency. It depends on the variable, the region, and the season. For example, in the JFM WEP case, the ADV in GFDL is much larger than that in other datasets, while in the JASOND EAO case, the GODAS product deviates markedly from other datasets (Fig. 5). Nevertheless, according to comparison of all the 23 selected cases, it could at least be suggested that the OAFlux displays the smallest difference with other flux data.

In the following discussion, we select five typical cases to explore how the dynamic and thermodynamic processes contribute to regional SST tendency. The investigation first focuses on the surface-heat-flux-dominant cases, followed by discussion of the oceanic-advection-dominant cases, and finally the coworking cases.

b. Surface-heat-flux-dominant cases

Figure 6 presents the AM WNP case by showing simultaneous regressions with respect to anomalous AM WNP (0°–15°N, 110°–160°E) SST tendency for the period 1984–2009. Previous studies indicated that the upper-ocean thermal condition in this region is crucial to the South China Sea summer monsoon variability (Wu 2010). The regional SST experiences a sharp increase (decrease) before (after) the summer monsoon onset, and this rapid SST change may have year-to-year variations. Here, the AM WNP case elucidates how the atmospheric changes contribute to the WNP SST variations during the summer monsoon onset period. Given positive WNP SST tendency (Fig. 6a), there is anomalous lower-level divergence over the WNP (Fig. 6e), which restrains local convection and precipitation (Fig. 6c). This is conducive to SST warming via enabling more incoming SWR to reach the ocean surface (Fig. 6b) under less cloud cover (Fig. 6d). In this case, the SWR regression is highly coherent with the SST tendency regression. It demonstrates that the cloud–radiation effect is a leading process in governing the AM WNP SST tendency with the SWR contribution up to 0.8°C month−1 (Fig. 6b). The LHF contribution is secondary with a mean magnitude of 0.2°C month−1 (Fig. 6e), and the contribution of ocean advection is almost negligible except for some limited regions along the coastal Philippine and New Guinea Islands (Fig. 6f).

Fig. 6.
Fig. 6.

Simultaneous Rex with respect to anomalous AM WNP SST tendency for the period 1984–2009. (a) OI version 2 SST tendency (°C month−1), (b) OAFlux shortwave radiation (°C month−1), (c) GPCP precipitation (mm day−1), (d) NOCS flux cloud cover (%; divided by 2), (e) OAFlux latent heat flux (shading; °C month−1) and NOAA version 2 10-m wind (m s−1; vector), and (f) GODAS ocean horizontal and vertical advection (converted into °C month−1). Shortwave radiation and latent heat flux have been converted to °C month−1 by dividing by GODAS climatological mixed-layer depth. Thick contours indicate that the correlations are statistically significant at the 90% confidence level. Only wind anomalies that are significant at the 90% confidence level are plotted.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

Figure 7 provides the simultaneous regressions with respect to anomalous MJJ MC (20°–5°S, 100°–150°E) SST tendency. This is a case with important wind–evaporation effect for the SST change. The MJJ MC case occurs in boreal summer when mean surface southeasterly winds reach the maximum in the MC region, which makes the SST more susceptible to surface wind changes. According to the study by Liu et al. (1979), the LHF is determined by two factors—surface wind speed and the air–sea humidity difference. There are anomalous northerly winds prevailing north of Australia (Fig. 7c). Superimposed on the southerly cross-equatorial mean flow, these wind anomalies can result in a decrease in surface wind speed over the South China Sea–New Guinea (Fig. 7c), and they reduce upper-ocean heat loss by suppressing upward LHF (Fig. 7b). Meanwhile, the reduced southeasterly winds can attenuate the dry and cold surface air that diverges from the Australian continent and passes through the MC. Correspondingly, it effectively inhibits surface evaporation north of Australia by diminishing the air–sea humidity difference with reinforced near surface humidity (Fig. 7d). In this case, the surface wind anomalies induce changes in both surface wind speed and air–sea humidity difference, which accounts for the MC SST increase during MJJ with the LHF contribution being 0.5°C month−1 approximately (Fig. 7b). By contrast, the contributions of SWR and ocean advection are small (Figs. 7e,f). According to the discussion in section 3a, the heat budget is unbalanced in the MJJ MC case. It is likely attributed to the data uncertainties in estimating the heat budget and the lack of the SST feedback. He and Wu (2013b) pointed out that once SST warms up, there will be a fast atmospheric response via destabilizing the lower-atmospheric column. This SST feedback is revealed in a positive precipitation–SST relationship and a negative SWR–SST relationship (Fig. 7e), which may counteract part of the cloud–radiation effect.

Fig. 7.
Fig. 7.

As in Fig. 6, but with respect to the MJJ MC SST tendency. (a) OI version 2 SST tendency (°C month−1), (b) OAFlux latent heat flux (°C month−1), (c) NOAA version 2 10-m wind (m s−1; vector) and OAFlux 10-m wind speed (m s−1; divided by 2; shading), (d) OAFlux air–sea humidity difference (g kg−1), (e) OAFlux shortwave radiation (°C month−1), and (f) GODAS ocean horizontal and vertical advection (°C month−1). Thick contours indicate that the correlations are statistically significant at the 90% confidence level. Only wind anomalies that are significant at the 90% confidence level are plotted.

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

In addition, the cloud–radiation effect plays an important role in the SST tendency in the WNP during boreal summer and autumn, in the MC during October–November–December, and in the TIO from December to March. The wind–evaporation is significant over the WNP in DJF, over the NIO from November to February, and over the NAO from November to March. There are cases with combined SWR and LHF contributions, such as the EEP in JFM, the MC in JFM, and the NAO from July to October (Fig. 5). These cases suggest that the mean state is important for the atmospheric changes to trigger SST anomalies. Specifically, the cloud–radiation effect usually rules the SST tendency in rainy seasons when the convective cloud is heavy, and the wind–evaporation effect is significant in seasons with large surface wind when the wind disturbance is active. It should also be stressed that the regional atmospheric changes may be related to remote forcing, such as ENSO (Klein et al. 1999; Wu and Wang 2000; Lau and Nath 2003; He and Wu 2014). Yet this is beyond the scope of the present study and will only be discussed briefly in section 5.

c. Oceanic advection dominant cases

For a detailed investigation of the ocean dynamic processes, we decompose the ocean advection as follows. The change of ADV and ADW can be attributed to the current-regulated component, the temperature-gradient-regulated component, and the higher-order perturbation [Eqs. (7) and (8)]:
e7
e8
where, stands for the anomaly and denotes the climatological mean. It should also be noted that the ADV () includes the meridional advection and the zonal advection .

The JFM WIO case is selected as a representation of the ADV-dominant cases. He et al. (2016) demonstrated that the winter WIO SST can influence the WNP summer rainfall. Wu and Kirtman (2007) suggested that these SST variations cannot be totally explained by surface flux contributions. However, both studies do not provide much detail to the development of the WIO SST anomalies. Thus, the present study explores the physical mechanism for the interannual WIO SST variations by considering the role of regional forcing. Figure 8 displays simultaneous regressions with respect to JFM WIO (0°–10°N, 50°–70°E) SST tendency along with climatological mean SST and surface currents (Figs. 8c,d). During JFM, accompanied by a significant east–west SST gradient (Fig. 8d), there are strong surface currents going westward in the region of 2°S–5°N, 45°–80°E and turning eastward along 5°S (Fig. 8c). It can be inferred that changes in these equatorial currents may induce anomalous heat transportation, which may produce significant SST anomalies. Given positive WIO SST tendency (Fig. 8a), there are anomalous anticyclonic gyres south and north of the equatorial WIO (Fig. 8d), which is accompanied by accelerated equatorial westward surface currents (Fig. 8b). The anomalous currents are induced by both a geostrophic effect associated with anomalous anticyclonic gyre north off the equator (Fig. 8g) and a wind-driven-current effect associated with strong southeasterly wind anomalies over the south Indian Ocean that are caused by the enhanced Mascarene high (Fig. 8a). These enhanced westward equatorial currents, superimposed on climatological SSTs higher in the eastern Indian Ocean (Fig. 8d), result in positive and cause warm advection along the equatorial WIO (Figs. 8e,f). On the other hand, the westward extending warm water over the eastern WIO (5°S–5°N, 55°–75°E) enlarges the zonal SST gradient (Fig. 8c). This increased temperature gradient, in conjunction with westward mean currents (Fig. 8c), leads to positive and facilitates positive advection anomalies in the equatorial WIO (Figs. 8e,f). Overall, changes in both the zonal sea level gradient and the Mascarene high can drive the current-regulated and temperature-gradient-regulated advection anomalies via the geostrophic effect and the wind-driven-current effect. They account for the WIO SST warming with significant contributions of zonal advection up to 0.8°C month−1 (Fig. 8f) and small-scale contributions of meridional advection north of the equator (Fig. 8e). It should be stressed that during JFM the surface heat flux has a dominant contribution in the south Indian Ocean, and thus it also plays an important role in this case (as indicated in Fig. 8h since the ADW contribution is small). In contrast to the horizontal advection, the heat flux works on a much larger spatial scale, contributing to an in-phase SST tendency in the whole WIO (15°S–20°N, 40°–75°E).

Fig. 8.
Fig. 8.

As in Fig. 6, but with respect to the JFM WIO SST tendency. (a) OI version 2 SST tendency (°C month−1) and NOAA version 2 10-m wind (m s−1; vector), (b) GODAS surface current (m s−1; multiplied by 2), (c) OI version 2 SST (°C) and climatological JFM-mean GODAS surface current (m s−1; vector), (d) GODAS surface current (m s−1; vector) and climatological JFM-mean OI version 2 SST (°C; divided by 10 and minus 2.5), (e) GODAS meridional advection (°C month−1), (f) GODAS zonal advection (°C month−1), (g) GODAS sea surface height relative to geoid (m), and (h) GODAS vertical advection and OAFlux net heat flux (°C month−1). Thick contours and plotted vectors indicate that the correlations are statistically significant at the 90% confidence level except for the climatological surface current in (c).

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

The MAM WSIO case provides information for the ADW-dominant process in Fig. 9. In boreal spring, the thermocline dome develops in the WSIO (10°–2°S, 55°–70°E), which is characterized by a cyclonic gyre with strong open-ocean Ekman upwelling and a cyclonic lower-level circulation with equatorial westerlies (Figs. 9c,d). Compared with other seasons, the mixed layer in MAM is very shallow with a mean depth of nearly 30 m (Fig. 9d), surface wind speed is still small, and convective precipitation is largely weakened as the intertropical convergence zone (ITCZ) shrinks to the eastern Indian Ocean (Fig. 9c). It implies that the upper-ocean temperature may be vulnerable to the oceanic forcing under such a shallow mixed layer since the atmospheric changes are inactive. In this case, anomalous easterly surface winds along the equator are accompanied by an anomalous anticyclonic curl over the WSIO (Fig. 9a), which directly drives an anomalous anticyclonic gyre with significant downward current anomalies around 10°S–0°, 50°–70°E (Fig. 9e). Correspondingly, this anomalous Ekman downwelling deepens the mixed layer via reduction of the vertical temperature gradient and effectively suppresses the ocean mixing (Fig. 9b). Thus, both the current-regulated and temperature-gradient-regulated Ekman pumping supports the upper-ocean warming (Fig. 9a) with the ADW contributions of 0.4°C month−1 (Fig. 9f). In contrast, the ADV and heat flux contributions are relatively small (Figs. 9g,h). The results agree with the study by Xie et al. (2002), who proposed an alternative explanation by highlighting the role of the westward-propagating Rossby wave in inducing winter SST anomalies over the tropical south Indian Ocean. They pointed out that these south Indian Ocean SST anomalies are related to the preceding ENSO and Sumatra remote forcing. Here, we underscore the importance of the synchronous regional wind-induced oceanic processes for the development of the MAM WSIO SST anomalies, which can work either independently from or in conjunction with the remote forcing.

Fig. 9.
Fig. 9.

As in Fig. 6, but with respect to the MAM WSIO SST tendency. (a) OI version 2 SST tendency (°C month−1) and NOAA version 2 10-m wind (m s−1; vector); (b) GODAS MLD (m; divided by 10) and surface current (m s−1; vector); (c) climatological MAM-mean GPCP precipitation (mm day−1; minus 5) and NOAA version 2 10-m wind (m s−1; vector); (d) climatological MAM-mean GODAS MLD (m; contour), vertical current velocity within the mixed layer (m s−1; multiplied by 106; shading; upward positive), and surface current (m s−1, vector); (e) GODAS vertical current velocity within the mixed layer (m s−1; multiplied by 105); (f) GODAS vertical advection (°C month−1); (g) GODAS horizontal advection (°C month−1); and (h) OAFlux net surface heat flux (°C month−1). Thick contours and plotted vectors indicate that the correlations are statistically significant at the 90% confidence level except for the climatological means in (c) and (d).

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

As presented in Fig. 5, the ADV-dominant cases usually occur at low latitudes where there are strong upper-ocean currents and horizontal temperature gradients, including the JFM WEP case, the AMJJ NCP case, the JJAS WSIO case, the JASO ESIO case, and the JASOND EAO case. The ADW is important for the SST tendency in the regions with significant equatorial-coastal upwelling or Ekman pumping (e.g., the SON NCP case). The JJASON EEP case can typically represent the ocean forcing process in which the SST tendency is under overwhelming control of horizontal and vertical advection (Bjerknes 1969; Suarez and Schopf 1988; Battisti and Hirst 1989; Jin 1997; Picaut et al. 1997; Weisberg and Wang 1997). These cases reveal that the ocean dynamic process is also closely related to the climatological mean state.

d. Coworking cases

The AM NIO (4°–10°N, 70°–80°E) case is used to illustrate the combined effects of dynamic and thermodynamic processes on regional SST change. Figure 10 displays the simultaneous regression with respect to the AM NIO SST tendency. The case occurs during the Indian summer monsoon onset period when the precipitation begins to increase, the surface wind turns toward the northeast, and the coastal upwelling develops over the southern Indian Peninsula under a “V shaped” coastal topography (Fig. 10h). The transition season, in coordination with the topographic effect, provides favorable conditions for triggering active atmospheric and oceanic perturbations over the NIO. In this case, given weakened Indian summer monsoon, there are easterly and northeasterly wind anomalies prevailing over the NIO (Fig. 10f), which cause a series of changes in the lower atmosphere and the upper ocean. First, these surface wind anomalies superposed on eastward mean wind (Fig. 10h) reduce upward LHF (Fig. 10c) by decreasing surface wind speed (Fig. 10f). Second, it attenuates the transportation of warm and wet air from the lower latitudes, which, on one hand, is effective at enhancing the incoming SWR (Fig. 10b) via suppression of the convective precipitation (Fig. 10e) and, on the other hand, offsets part of the positive LHF contributions by increasing air–sea humidity difference (figure not shown). Finally, the northeasterly wind anomalies can drive anomalous onshore currents (Fig. 10d), deepen the mixed layer, and inhibit the ocean mixing with anomalous coastal downwelling in the region of 5°–10°N, 70°–90°E (Fig. 10g). The cloud–radiation effect and the wind–evaporation effect have large-scale influences on the SST tendency with the SWR and LHF contributions being 0.4° and 0.35°C month−1, respectively (Figs. 10b,c). The wind-induced coastal upwelling effect is apparent over the southern Indian Peninsula with the ADW contribution being 0.35°C month−1, approximately (Fig. 10d). Overall, the AM NIO SST variability is highly coherent with the Indian summer monsoon activities. The V-shaped topography allows the oceanic forcing to exert a large impact on regional SST anomalies via the coastal upwelling (downwelling) effect. This is distinguished from the AM WNP case (Fig. 6) in which the SST tendency is dominated by the atmospheric processes during the SCS summer monsoon onset period.

Fig. 10.
Fig. 10.

As in Fig. 6, but with respect to the AM NIO SST tendency. (a) OI version 2 SST tendency (°C month−1), (b) OAFlux shortwave radiation (°C month−1), (c) OAFlux latent heat flux (°C month−1), (d) GODAS vertical advection (°C month−1), (e) GPCP precipitation (mm day−1), (f) NOAA version 2 10-m wind vector and OAFlux wind speed (m s−1; multiplied by 2), (g) GODAS MLD (m; the contour interval is 6 m) and vertical current velocity within the mixed layer (m s−1; multiplied by 106; shading; upward positive), and (h) climatological MAM-mean NOAA version 2 10-m wind vector (m s−1), GODAS vertical current velocity within the mixed layer (m s−1; multiplied by 106; shading), and GPCP precipitation (mm day−1; contour). Thick contours and plotted vectors indicate that the correlations are statistically significant at the 90% confidence level except for the climatological means in (h).

Citation: Journal of Climate 30, 11; 10.1175/JCLI-D-16-0500.1

Besides this NIO example, there are other coworking cases (the last row in Fig. 5). For example, the cloud–radiation effect is comparable to the ocean advection effect in the equatorial Pacific during AM when the regional convection becomes relatively active. Both the wind-induced-evaporation and wind-induced-advection processes are important for the summer ESIO SST tendency that is associated with the development of the Indian Ocean dipole (Saji et al. 1999; Webster et al. 1999; Yu and Rienecker 1999; Murtugudde et al. 2000; Li et al. 2002; Xie et al. 2002; Li et al. 2003; Saji and Yamagata 2003; Tokinaga and Tanimoto 2004).

5. Summary and discussion

The present study made a systematic investigation of contributions of surface heat flux and ocean advection to the SST change in the tropical oceans. We underscore that atmospheric and oceanic contributions to the SST tendency display remarkable seasonal and regional dependence. The surface heat flux contribution covers a large spatial scale in the off-equatorial regions, while the oceanic advection with small-scale features dominates SST variations in the equatorial regions. The air–sea relationship is distinguished in different oceans. The SST tendency is governed by dynamic and thermodynamic processes alternatively in the Pacific Ocean. The relationship in the Atlantic Ocean is similar but not as significant as that in the Pacific Ocean. In contrast, a wide variety of processes contribute to the SST change in the Indian Ocean. This is seldom mentioned in previous studies. Five typical cases are further selected to illustrate the corresponding physical processes. The AM WNP case, the MJJ MC case, the JFM WIO case, the MAM WSIO case, and the AM NIO case are governed by the cloud–radiation effect, the wind–evaporation effect, the wind-driven-ADV effect, the wind-driven-ADW effect, and the SWR–LHF–ADW coforcing effect, respectively. Overall, the cloud–radiation effect is dominant in rainy seasons, the wind–evaporation effect is significant in seasons with large winds, the ADV rules the SST tendency in regions with strong upper-ocean currents or horizontal temperature gradients, and the ADW is important in regions with obvious equatorial-coastal upwelling or Ekman pumping. It generally agrees with the studies by Wu and Kirtman (2007) and Deser et al. (2010). The seasonality and regionality of the interannual air–sea relationship indicates a physical connection with the mean state.

Six air–sea flux products (OAFlux, NOCS flux, ERA-40, NCEP–DOE Reanalysis 2, CERES, and GSSTF) and three ocean products (GFDL ECDA, GODAS, and SODA) are employed to evaluate the surface heat budget. According to a comparison of 23 selected cases, the heat budget is nearly balanced in the SWR- and ADV-dominant cases but apparently not balanced in some of the LHF- and ADW-dominant cases. The datasets show large differences in estimating the major terms, and they have a high level of consistency in computing the minor ones. The discrepancies among different products are larger in estimating the contribution of ADW than other heat budget terms, probably owing to the small spatial scale of ocean mixing process. The study suggests that no products are ideal for studying the interannual air–sea relationship in the tropical oceans. For a better understanding of the heat budget, further improvements in surface heat flux and ocean products are highly encouraged.

The study provides some information about the difference and consistency of different datasets. However, there are two issues to be noticed. First, the evaluation is based on only the 23 selected cases. Although they cover most types of influences of atmospheric and oceanic processes on the SST change, they cannot represent all the cases in the tropical oceans. Second, there are other air–sea flux and ocean products available—for instance, the in situ–based Florida State University fluxes, version 3 (FSU3 flux; Bourassa et al. 2005); the Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data, version 3.2 (HOAPS 3.2; Fennig et al. 2012); the Coordinated Ocean Research Experiments, version 2 (CORE2; Large and Yeager 2009); the Japanese Ocean Flux Datasets with Use of Remote Sensing Observations, version 2 (J-OFURO2, Kubota et al. 2002); and the Japanese 55-year Reanalysis (JRA-55; Kobayashi et al. 2015). Here, we select only the nine most-used datasets for the analysis. Further research is encouraged to include more new updated products for a better comprehensive assessment.

We adopt the regression analysis to estimate the heat budget. Nevertheless, the surface heat budget cannot be fully balanced, which may be attributed to different reasons. The study includes in situ, satellite, reanalysis, and hybrid datasets, which are derived based on various retrieval methods and input data sources on different temporal and spatial resolutions. There may be both bias and random uncertainties in these products. For example, the LHF in the NOCS flux includes bias uncertainty ranging from 10 to 80 W m−2 compared with the mean LHF ranging from 60 to 200 W m−2 in the tropical oceans (figures not shown). The period used for heat budget calculation (Fig. 5) is different for different datasets based on data availability—for example, 1984–2007 for the OAFlux, NOCS flux, NCEP–DOE Reanalysis 2, GFDL ECDA, GODAS, and SODA; 1988–2008 for GSSTF3; 1982–2001 for ERA-40; and 2001–13 for CERES. Thus, the heat budget estimation may involve a certain extent of inconsistency due to the differences in the temporal coverage. We divide the surface heat flux by climatological monthly mean MLD to highlight the contribution of heat flux to the SST tendency. However, the MLD anomaly-regulated component also affects the SST tendency (Lau and Nath 2003), and this is not considered in the heat budget estimation. The present study does not consider the SST feedback to the atmosphere and ocean. The SST anomaly, on one hand, is under the influence of atmospheric and oceanic process. On the other hand, it can in turn induce changes in local convection and surface wind via thermodynamic adjustment of the troposphere, which may further drive corresponding changes in upper-ocean currents and heat redistribution (Matsuno 1966; Gill 1980; Lindzen and Nigam 1987; Wallace et al. 1989). This SST feedback may damp part of the atmospheric and oceanic effects. Thus, the contributions of surface heat flux and ocean advection may be underestimated in some cases when the SST feedback is strong. These SST response and feedback processes are noticed in the previous studies (Wu and Kirtman 2007; He and Wu 2013a,b). Yet it should be noted that the SST forcing cannot be completely decomposed based on only observational analysis. Numerical experiments need to be conducted for this purpose in the future study.

The present study addresses the relative importance of atmospheric and oceanic processes by highlighting the cloud–radiation effect, the wind–evaporation effect, and the ocean advection effect, without detailed discussions on the roles of LWR, SHF, and MLD. The LWR contribution is mainly proportional to the SST anomaly and partly affected by the cloud cover (Clark et al. 1974). The SHF is not considered for its negligible contribution on the interannual scale. The MLD change may play an indirect role in the SST change (Tomita and Nonaka 2006; Deser et al. 2010). Furthermore, the present study focuses on the simultaneous regional forcing of the SST anomaly without consideration of the remote forcing. Remote forcing also exhibits pronounced influence on the SST variability via modification of atmospheric and oceanic processes. For example, ENSO, as the strongest short-term climatic signal in the tropical oceans, is demonstrated to be important for regional SST variability over the South China Sea, the TIO, the tropical North Atlantic, the western North Pacific, and the central North Pacific (Klein et al. 1999; Alexander et al. 2002; Wu and Wang 2000; Wang et al. 2006; Wu et al. 2008; Deser et al. 2010; Xie et al. 2009; He and Wu 2014; He et al. 2016). According to a lead–lag correlation analysis (figures not shown) and previous studies (Klein et al. 1999; Wu and Kirtman 2007; Xie et al. 2009), the AM WNP case, the MAM WSIO case, and the AM NIO case may be associated with the remote eastern Pacific SST variability. Further research is in progress to detect the relationship between the regional forcing and remote forcing, as well as their corresponding effects on regional climate variability.

Acknowledgments

We express our gratitude to Ruixing Huang for his insightful comments. This study is jointly supported by the Funds for Creative Research Groups of China (41521005), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA11010301), the National Natural Science Foundation of China (Grants 41506003, 41676013, and 41506004), and the Independent Research Project Program of State Key Laboratory of Tropical Oceanography (LTOZZ1603). RW acknowledges the support of the National Natural Science Foundation of China (Grants 41530425, 41475081, and 41275081). ZW acknowledges the support of the National Natural Science Foundation of China (Grant 41530530).

REFERENCES

  • Adler, R. F., and Coauthors, 2003: The version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeor., 4, 11471167, doi:10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Alexander, M. A., I. Bladé, M. Newman, J. R. Lanzante, N. C. Lau, and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans. J. Climate, 15, 22052231, doi:10.1175/1520-0442(2002)015<2205:TABTIO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Battisti, D. S., and A. C. Hirst, 1989: Interannual variability in a tropical atmosphere–ocean model: Influence of the basic state, ocean geometry and nonlinearity. J. Atmos. Sci., 46, 16871712, doi:10.1175/1520-0469(1989)046<1687:IVIATA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Berry, D. I., and E. C. Kent, 2009: A new air–sea interaction gridded dataset from ICOADS with uncertainty estimates. Bull. Amer. Meteor. Soc., 90, 645656, doi:10.1175/2008BAMS2639.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bjerknes, J., 1969: Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97, 163172, doi:10.1175/1520-0493(1969)097<0163:ATFTEP>2.3.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bony, S., Y. Sud, K. M. Lau, J. Susskind, and S. Saha, 1997: Comparison and satellite assessment of NASA/DAO and NCEP–NCAR reanalysis over tropical ocean: Atmospheric hydrology and radiation. J. Climate, 10, 14411462, doi:10.1175/1520-0442(1997)010<1441:CASAON>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bourassa, M. A., R. Romero, S. R. Smith, and J. J. O’Brien, 2005: A new FSU winds climatology. J. Climate, 18, 36863698, doi:10.1175/JCLI3487.1.

  • Bourras, D., 2006: Comparison of five satellite-derived latent heat flux products to moored buoy data. J. Climate, 19, 62916313, doi:10.1175/JCLI3977.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Carton, J. A., and B. S. Giese, 2008: A reanalysis of ocean climate using Simple Ocean Data Assimilation (SODA). Mon. Wea. Rev., 136, 29993017, doi:10.1175/2007MWR1978.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cayan, D. R., 1992: Latent and sensible heat flux anomalies over the northern oceans: Driving the sea surface temperature. J. Phys. Oceanogr., 22, 859881, doi:10.1175/1520-0485(1992)022<0859:LASHFA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, J. M., C. P. Chang, and T. Li, 2003: Annual cycle of the South China Sea surface temperature using the NCEP/NCAR reanalysis. J. Meteor. Soc. Japan, 81, 879884, doi:10.2151/jmsj.81.879.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Clark, N. E., L. Eber, R. M. Laurs, J. A. Renner, and J. F. T. Saur, 1974: Heat exchange between ocean and atmosphere in the eastern North Pacific for 1961–71. NOAA Tech. Rep. NMFS SSRF-682, 108 pp.

  • Deser, C., M. A. Alexander, S. P. Xie, and A. S. Phillips, 2010: Sea surface temperature variability: Patterns and mechanisms. Annu. Rev. Mar. Sci., 2, 115143, doi:10.1146/annurev-marine-120408-151453.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fennig, K., A. Andersson, S. Bakan, C. Klepp, and M. Schröder, 2012: Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data, HOAPS 3.2 - monthly means/6-hourly composites. Satellite Application Facility on Climate Monitoring, accessed 30 January 2014, doi:10.5676/EUM_SAF_CM/HOAPS/V001.

    • Crossref
    • Export Citation
  • Gill, A., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447462, doi:10.1002/qj.49710644905.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gulev, S., and Coauthors, 2000: Intercomparison and validation of ocean–atmosphere energy flux fields. Joint WCRP/SCOR Working Group on Air–Sea Fluxes Final Rep., 306 pp. [Available online at https://eprints.soton.ac.uk/69522/1/wgasf_final_rep.pdf.]

  • He, Z., and R. Wu, 2013a: Coupled seasonal variability in the South China Sea. J. Oceanogr., 69, 5769, doi:10.1007/s10872-012-0157-1.

  • He, Z., and R. Wu, 2013b: Seasonality of interannual atmosphere–ocean interaction in the South China Sea. J. Oceanogr., 69, 699712, doi:10.1007/s10872-013-0201-9.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • He, Z., and R. Wu, 2014: Indo-Pacific remote forcing in summer rainfall variability over the South China Sea. Climate Dyn., 42, 23232337, doi:10.1007/s00382-014-2123-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • He, Z., R. Wu, and W. Wang, 2016: Signals of the South China Sea summer rainfall variability in the Indian Ocean. Climate Dyn., 46, 31813195, doi:10.1007/s00382-015-2760-5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ji, M., A. Leetmaa, and J. Derber, 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123, 460481, doi:10.1175/1520-0493(1995)123<0460:AOASFS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jin, F., 1997: An equatorial ocean recharge paradigm for ENSO. Part I: Conceptual model. J. Atmos. Sci., 54, 811829, doi:10.1175/1520-0469(1997)054<0811:AEORPF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kållberg, P., A. Simmons, S. Uppala, and M. Fuentes, 2004: The ERA-40 archive. ERA-40 Project Rep. Series 17, 32 pp. [Available online at http://www.ecmwf.int/sites/default/files/elibrary/2004/10595-era-40-archive-revised-october-2007.pdf.]

  • Kanamitsu, M., W. Ebisuzaki, J. Woollen, S. K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643, doi:10.1175/BAMS-83-11-1631.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kitoh, A., T. Motoi, and H. Koide, 1999: SST variability and its mechanism in a coupled atmosphere–mixed layer ocean model. J. Climate, 12, 12211239, doi:10.1175/1520-0442(1999)012<1221:SVAIMI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Klein, S. A., B. J. Soden, and N. C. Lau, 1999: Remote sea surface temperature variations during ENSO: Evidence for a tropical atmospheric bridge. J. Climate, 12, 917932, doi:10.1175/1520-0442(1999)012<0917:RSSTVD>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 548, doi:10.2151/jmsj.2015-001.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kubota, M., N. Iwasaka, S. Kizu, M. Konda, and K. Kutsuwada, 2002: Japanese ocean flux data sets with use of remote sensing observations (J-OFURO). J. Oceanogr., 58, 213225, doi:10.1023/A:1015845321836.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., and S. G. Yeager, 2009: The global climatology of an interannually varying air–sea flux data set. Climate Dyn., 33, 341364, doi:10.1007/s00382-008-0441-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lau, N. C., and M. J. Nath, 2003: Atmosphere–ocean variations in the Indo-Pacific sector during ENSO episodes. J. Climate, 16, 320, doi:10.1175/1520-0442(2003)016<0003:AOVITI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lau, N. C., and M. J. Nath, 2009: A model investigation of the role of air–sea interaction in the climatological evolution and ENSO-related variability of the summer monsoon over the South China Sea and western North Pacific. J. Climate, 22, 47714797, doi:10.1175/2009JCLI2758.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, T., Y. Zhang, E. Lu, and D. Wang, 2002: Relative role of dynamic and thermodynamic processes in the development of the Indian Ocean dipole: An OGCM diagnosis. Geophys. Res. Lett., 29, 2110, doi:10.1029/2002GL015789.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, T., B. Wang, C. P. Chang, and Y. Zhang, 2003: A theory for the Indian Ocean dipole–zonal mode. J. Atmos. Sci., 60, 21192135, doi:10.1175/1520-0469(2003)060<2119:ATFTIO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lindzen, R. S., and S. Nigam, 1987: On the role of sea surface temperature gradients in forcing low-level winds and convergence in the tropics. J. Atmos. Sci., 44, 24182436, doi:10.1175/1520-0469(1987)044<2418:OTROSS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, Q. Y., S. P. Xie, L. Li, and N. A. Maximenko, 2005: Ocean thermal advective effect on the annual range of sea surface temperature. Geophys. Res. Lett., 32, L24604, doi:10.1029/2005GL024493.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, W. T., K. B. Katsaros, and J. A. Businger, 1979: Bulk parameterization of air–sea exchanges of heat and water vapor including the molecular constraints at the interface. J. Atmos. Sci., 36, 17221735, doi:10.1175/1520-0469(1979)036<1722:BPOASE>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Matsuno, T., 1966: Quasi-geostrophic motions in the equatorial area. J. Meteor. Soc. Japan, 44, 2543.

  • Murtugudde, R., J. P. McCreary, and A. J. Busalacchi, 2000: Oceanic processes associated with anomalous events in the Indian Ocean with relevance to 1997–1998. J. Geophys. Res., 105, 32953306, doi:10.1029/1999JC900294.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Norris, J. R., and C. B. Leovy, 1994: Interannual variability in stratiform cloudiness and sea surface temperature. J. Climate, 7, 19151925, doi:10.1175/1520-0442(1994)007<1915:IVISCA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Paulson, C. A., and J. J. Simpson, 1977: Irradiance measurements in the upper ocean. J. Phys. Oceanogr., 7, 952956, doi:10.1175/1520-0485(1977)007<0952:IMITUO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Picaut, J., F. Masia, and Y. D. Penhoat, 1997: An advective–reflective conceptual model for the oscillatory nature of the ENSO. Science, 277, 663666, doi:10.1126/science.277.5326.663.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Qu, T., 2001: Role of ocean dynamics in determining the mean seasonal cycle of the South China Sea surface temperature. J. Geophys. Res., 106, 69436955, doi:10.1029/2000JC000479.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reed, R. K., 1977: On estimating insolation over the ocean. J. Phys. Oceanogr., 7, 482485, doi:10.1175/1520-0485(1977)007<0482:OEIOTO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., N. A. Rayner, T. M. Smith, D. C. Stokes, and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15, 16091625, doi:10.1175/1520-0442(2002)015<1609:AIISAS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saji, N. H., and T. Yamagata, 2003: Structure of SST and surface wind variability during Indian Ocean dipole mode events: COADS observations. J. Climate, 16, 27352751, doi:10.1175/1520-0442(2003)016<2735:SOSASW>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saji, N. H., B. N. Goswami, P. N. Vinayachandran, and T. Yamagata, 1999: A dipole mode in the tropical Indian Ocean. Nature, 401, 360363.

  • Shie, C. L., K. Hilburn, L. S. Chiu, R. Adler, I. I. Lin, E. Nelkin, J. Ardizzone, and S. Gao, 2012: Goddard satellite-based surface turbulent fluxes, 0.25 × 0.25 deg, daily grid V3, version 3. Goddard Earth Sciences Data and Information Services Center (GES DISC), accessed 9 December 2014. [Available online at http://disc.gsfc.nasa.gov/datacollection/GSSTF_3.html.]

  • Smith, S. R., P. J. Hughes, and M. A. Bourassa, 2011: A comparison of nine monthly air–sea flux products. Int. J. Climatol., 31, 10021027, doi:10.1002/joc.2225.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Suarez, M. J., and P. S. Schopf, 1988: A delayed action oscillator for ENSO. J. Atmos. Sci., 45, 32833287, doi:10.1175/1520-0469(1988)045<3283:ADAOFE>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tokinaga, H., and Y. Tanimoto, 2004: Seasonal transition of SST anomalies in the tropical Indian Ocean during El Niño and Indian Ocean dipole years. J. Meteor. Soc. Japan, 82, 10071018, doi:10.2151/jmsj.2004.1007.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tomita, T., and M. Nonaka, 2006: Upper-ocean mixed layer and wintertime sea surface temperature anomalies in the North Pacific. J. Climate, 19, 300307, doi:10.1175/JCLI3616.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., and D. J. Shea, 2005: Relationships between precipitation and surface temperature. Geophys. Res. Lett., 32, L14703, doi:10.1029/2005GL022760.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., T. P. Mitchell, and C. Deser, 1989: The influence of sea surface temperature on surface wind in the eastern equatorial Pacific: Seasonal and interannual variability. J. Climate, 2, 14921499, doi:10.1175/1520-0442(1989)002<1492:TIOSST>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, C., W. Wang, D. Wang, and Q. Wang, 2006: Interannual variability of the South China Sea associated with El Niño. J. Geophys. Res., 111, C03023, doi:10.1029/2005JC003333.

    • Search Google Scholar
    • Export Citation
  • Webster, P. J., A. M. Moore, J. P. Loschnigg, and R. R. Leben, 1999: Coupled ocean–atmosphere dynamics in the Indian Ocean during 1997–98. Nature, 401, 356360, doi:10.1038/43848.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weisberg, R. H., and C. A. Wang, 1997: A western Pacific oscillator paradigm for the El Niño–Southern Oscillation. Geophys. Res. Lett., 24, 779782, doi:10.1029/97GL00689.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wielicki, B. A., B. R. Barkstrom, E. F. Harrison, R. B. Lee, G. L. Smith, and J. E. Cooper, 1996: Clouds and the Earth’s Radiant Energy System (CERES): An Earth observing system experiment. Bull. Amer. Meteor. Soc., 77, 853868, doi:10.1175/1520-0477(1996)077<0853:CATERE>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, R., 2010: Subseasonal variability during the South China Sea summer monsoon onset. Climate Dyn., 34, 629642, doi:10.1007/s00382-009-0679-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, R., and B. Wang, 2000: Interannual variability of summer monsoon onset over the western North Pacific and the underlying processes. J. Climate, 13, 24832501, doi:10.1175/1520-0442(2000)013<2483:IVOSMO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, R., and B. P. Kirtman, 2005: Roles of Indian and Pacific Ocean air–sea coupling in tropical atmospheric variability. Climate Dyn., 25, 155170, doi:10.1007/s00382-005-0003-x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, R., and B. P. Kirtman, 2007: Regimes of seasonal air–sea interaction and implications for performance of forced simulations. Climate Dyn., 29, 393410, doi:10.1007/s00382-007-0246-9.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, R., and J. L. Kinter, 2010: Atmosphere–ocean relationship in the midlatitude North Pacific: Seasonal dependence and east-west contrast. J. Geophys. Res., 115, D06101, doi:10.1029/2009JD012579.

    • Search Google Scholar
    • Export Citation
  • Wu, R., and S. W. Yeh, 2010: A further study of the tropical Indian Ocean asymmetric mode in boreal spring. J. Geophys. Res., 115, D08101, doi:10.1029/2009JD012999.

    • Search Google Scholar
    • Export Citation
  • Wu, R., B. P. Kirtman, and K. Pegion, 2006: Local air–sea relationship in observations and model simulations. J. Climate, 19, 49144932, doi:10.1175/JCLI3904.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, R., B. P. Kirtman, and V. Krishnamurthy, 2008: An asymmetric mode of tropical Indian Ocean rainfall variability in boreal spring. J. Geophys. Res., 113, D05104, doi:10.1029/2007JD008931.

    • Search Google Scholar
    • Export Citation
  • Xie, S. P., 2004: Satellite observations of cool ocean–atmosphere interaction. Bull. Amer. Meteor. Soc., 85, 195208, doi:10.1175/BAMS-85-2-195.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xie, S. P., H. Annamalai, F. A. Schott, and J. P. McCreary, 2002: Structure and mechanisms of south Indian Ocean climate variability. J. Climate, 15, 864878, doi:10.1175/1520-0442(2002)015<0864:SAMOSI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xie, S. P., K. Hu, J. Hafner, H. Tokinaga, Y. Du, G. Huang, and T. Sampe, 2009: Indian Ocean capacitor effect on Indo–western Pacific climate during the summer following El Niño. J. Climate, 22, 730747, doi:10.1175/2008JCLI2544.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yu, L., and M. M. Rienecker, 1999: Mechanisms for the Indian Ocean warming during the 1997–98 El Niño. Geophys. Res. Lett., 26, 735738, doi:10.1029/1999GL900072.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yu, L., X. Jin, and R. A. Weller, 2008: Multidecade global flux datasets from the objectively analyzed air–sea fluxes (OAFlux) project: Latent and sensible heat fluxes, ocean evaporation, and related surface meteorological variables. Woods Hole Oceanographic Institution OAFlux Project Tech. Rep. OA-2008-01, 64 pp. [Available online at http://oaflux.whoi.edu/pdfs/OAFlux_TechReport_3rd_release.pdf.]

  • Zhang, S., M. J. Harrison, A. Rosati, and A. Wittenberg, 2007: System design and evaluation of coupled ensemble data assimilation for global oceanic climate studies. Mon. Wea. Rev., 135, 35413564, doi:10.1175/MWR3466.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save
  • Adler, R. F., and Coauthors, 2003: The version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeor., 4, 11471167, doi:10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Alexander, M. A., I. Bladé, M. Newman, J. R. Lanzante, N. C. Lau, and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans. J. Climate, 15, 22052231, doi:10.1175/1520-0442(2002)015<2205:TABTIO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Battisti, D. S., and A. C. Hirst, 1989: Interannual variability in a tropical atmosphere–ocean model: Influence of the basic state, ocean geometry and nonlinearity. J. Atmos. Sci., 46, 16871712, doi:10.1175/1520-0469(1989)046<1687:IVIATA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Berry, D. I., and E. C. Kent, 2009: A new air–sea interaction gridded dataset from ICOADS with uncertainty estimates. Bull. Amer. Meteor. Soc., 90, 645656, doi:10.1175/2008BAMS2639.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bjerknes, J., 1969: Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97, 163172, doi:10.1175/1520-0493(1969)097<0163:ATFTEP>2.3.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bony, S., Y. Sud, K. M. Lau, J. Susskind, and S. Saha, 1997: Comparison and satellite assessment of NASA/DAO and NCEP–NCAR reanalysis over tropical ocean: Atmospheric hydrology and radiation. J. Climate, 10, 14411462, doi:10.1175/1520-0442(1997)010<1441:CASAON>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bourassa, M. A., R. Romero, S. R. Smith, and J. J. O’Brien, 2005: A new FSU winds climatology. J. Climate, 18, 36863698, doi:10.1175/JCLI3487.1.

  • Bourras, D., 2006: Comparison of five satellite-derived latent heat flux products to moored buoy data. J. Climate, 19, 62916313, doi:10.1175/JCLI3977.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Carton, J. A., and B. S. Giese, 2008: A reanalysis of ocean climate using Simple Ocean Data Assimilation (SODA). Mon. Wea. Rev., 136, 29993017, doi:10.1175/2007MWR1978.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cayan, D. R., 1992: Latent and sensible heat flux anomalies over the northern oceans: Driving the sea surface temperature. J. Phys. Oceanogr., 22, 859881, doi:10.1175/1520-0485(1992)022<0859:LASHFA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, J. M., C. P. Chang, and T. Li, 2003: Annual cycle of the South China Sea surface temperature using the NCEP/NCAR reanalysis. J. Meteor. Soc. Japan, 81, 879884, doi:10.2151/jmsj.81.879.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Clark, N. E., L. Eber, R. M. Laurs, J. A. Renner, and J. F. T. Saur, 1974: Heat exchange between ocean and atmosphere in the eastern North Pacific for 1961–71. NOAA Tech. Rep. NMFS SSRF-682, 108 pp.

  • Deser, C., M. A. Alexander, S. P. Xie, and A. S. Phillips, 2010: Sea surface temperature variability: Patterns and mechanisms. Annu. Rev. Mar. Sci., 2, 115143, doi:10.1146/annurev-marine-120408-151453.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fennig, K., A. Andersson, S. Bakan, C. Klepp, and M. Schröder, 2012: Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data, HOAPS 3.2 - monthly means/6-hourly composites. Satellite Application Facility on Climate Monitoring, accessed 30 January 2014, doi:10.5676/EUM_SAF_CM/HOAPS/V001.

    • Crossref
    • Export Citation
  • Gill, A., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447462, doi:10.1002/qj.49710644905.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gulev, S., and Coauthors, 2000: Intercomparison and validation of ocean–atmosphere energy flux fields. Joint WCRP/SCOR Working Group on Air–Sea Fluxes Final Rep., 306 pp. [Available online at https://eprints.soton.ac.uk/69522/1/wgasf_final_rep.pdf.]

  • He, Z., and R. Wu, 2013a: Coupled seasonal variability in the South China Sea. J. Oceanogr., 69, 5769, doi:10.1007/s10872-012-0157-1.

  • He, Z., and R. Wu, 2013b: Seasonality of interannual atmosphere–ocean interaction in the South China Sea. J. Oceanogr., 69, 699712, doi:10.1007/s10872-013-0201-9.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • He, Z., and R. Wu, 2014: Indo-Pacific remote forcing in summer rainfall variability over the South China Sea. Climate Dyn., 42, 23232337, doi:10.1007/s00382-014-2123-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • He, Z., R. Wu, and W. Wang, 2016: Signals of the South China Sea summer rainfall variability in the Indian Ocean. Climate Dyn., 46, 31813195, doi:10.1007/s00382-015-2760-5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ji, M., A. Leetmaa, and J. Derber, 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123, 460481, doi:10.1175/1520-0493(1995)123<0460:AOASFS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jin, F., 1997: An equatorial ocean recharge paradigm for ENSO. Part I: Conceptual model. J. Atmos. Sci., 54, 811829, doi:10.1175/1520-0469(1997)054<0811:AEORPF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kållberg, P., A. Simmons, S. Uppala, and M. Fuentes, 2004: The ERA-40 archive. ERA-40 Project Rep. Series 17, 32 pp. [Available online at http://www.ecmwf.int/sites/default/files/elibrary/2004/10595-era-40-archive-revised-october-2007.pdf.]

  • Kanamitsu, M., W. Ebisuzaki, J. Woollen, S. K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643, doi:10.1175/BAMS-83-11-1631.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kitoh, A., T. Motoi, and H. Koide, 1999: SST variability and its mechanism in a coupled atmosphere–mixed layer ocean model. J. Climate, 12, 12211239, doi:10.1175/1520-0442(1999)012<1221:SVAIMI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Klein, S. A., B. J. Soden, and N. C. Lau, 1999: Remote sea surface temperature variations during ENSO: Evidence for a tropical atmospheric bridge. J. Climate, 12, 917932, doi:10.1175/1520-0442(1999)012<0917:RSSTVD>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 548, doi:10.2151/jmsj.2015-001.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kubota, M., N. Iwasaka, S. Kizu, M. Konda, and K. Kutsuwada, 2002: Japanese ocean flux data sets with use of remote sensing observations (J-OFURO). J. Oceanogr., 58, 213225, doi:10.1023/A:1015845321836.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., and S. G. Yeager, 2009: The global climatology of an interannually varying air–sea flux data set. Climate Dyn., 33, 341364, doi:10.1007/s00382-008-0441-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lau, N. C., and M. J. Nath, 2003: Atmosphere–ocean variations in the Indo-Pacific sector during ENSO episodes. J. Climate, 16, 320, doi:10.1175/1520-0442(2003)016<0003:AOVITI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lau, N. C., and M. J. Nath, 2009: A model investigation of the role of air–sea interaction in the climatological evolution and ENSO-related variability of the summer monsoon over the South China Sea and western North Pacific. J. Climate, 22, 47714797, doi:10.1175/2009JCLI2758.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, T., Y. Zhang, E. Lu, and D. Wang, 2002: Relative role of dynamic and thermodynamic processes in the development of the Indian Ocean dipole: An OGCM diagnosis. Geophys. Res. Lett., 29, 2110, doi:10.1029/2002GL015789.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, T., B. Wang, C. P. Chang, and Y. Zhang, 2003: A theory for the Indian Ocean dipole–zonal mode. J. Atmos. Sci., 60, 21192135, doi:10.1175/1520-0469(2003)060<2119:ATFTIO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lindzen, R. S., and S. Nigam, 1987: On the role of sea surface temperature gradients in forcing low-level winds and convergence in the tropics. J. Atmos. Sci., 44, 24182436, doi:10.1175/1520-0469(1987)044<2418:OTROSS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, Q. Y., S. P. Xie, L. Li, and N. A. Maximenko, 2005: Ocean thermal advective effect on the annual range of sea surface temperature. Geophys. Res. Lett., 32, L24604, doi:10.1029/2005GL024493.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, W. T., K. B. Katsaros, and J. A. Businger, 1979: Bulk parameterization of air–sea exchanges of heat and water vapor including the molecular constraints at the interface. J. Atmos. Sci., 36, 17221735, doi:10.1175/1520-0469(1979)036<1722:BPOASE>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Matsuno, T., 1966: Quasi-geostrophic motions in the equatorial area. J. Meteor. Soc. Japan, 44, 2543.

    • Crossref