Abstract

This study investigates sea surface salinity (SSS) variations in the tropical Indian Ocean (IO) using the Aquarius/Satelite de Aplicaciones Cientificas-D (SAC-D) and the Soil Moisture and Ocean Salinity (SMOS) satellite data and the Argo observations during July 2010–July 2014. Compared to the Argo observations, the satellite datasets generally provide SSS maps with higher space–time resolution, particularly in the regions where Argo floats are sparse. Both Aquarius and SMOS well captured the SSS variations associated with the Indian Ocean dipole (IOD) mode. Significant SSS changes occurred in the central equatorial IO, along the Java–Sumatra coast, and south of the equatorial IO, due to ocean circulation variations. During the negative IOD events in 2010, 2013, and 2014, westerly wind anomalies strengthened along the equator, weakening coastal upwelling off Java and Sumatra and decreasing SSS. South of the equatorial IO, an anomalous cyclonic gyre changed the tropical circulation, which favored the eastward high-salinity tongue along the equator and the westward low-saline tongue in the south. An upwelling Rossby wave favored the increase of SSS farther to the south. During the positive IOD events in 2011 and 2012, the above-mentioned processes reversed, although the decrease of SSS was weaker in magnitude.

1. Introduction

Ocean salinity is an essential variable of ocean dynamics, playing an important role in global climate variability (Lagerloef 2002). Sea surface salinity (SSS) is highly related to local evaporation (E) and precipitation (P) in the global ocean and to river discharges in the coastal region. In the tropical Indian Ocean (TIO), the SSS shows significant spatial distribution, featuring an east–west contrast and significant SSS tongues at equatorial and southern Indian Ocean (IO). Basically, the east–west contrast is due to large evaporation in the Arabian Sea and abundant rainfall and river discharges in the Bay of Bengal and the eastern TIO (Schott et al. 2002), while the SSS tongues are influenced by ocean circulations (Han and McCreary 2001).

Ocean dynamics plays an equally important role in SSS distribution. North of the equator, monsoon currents dominate the water exchange between the Arabian Sea and the Bay of Bengal, which advect high-salinity water to the Bay of Bengal in summer or conversely low-salinity water to the Arabian Sea in winter (Jensen 2003; Zhang and Du 2012). Along the equator, the Wyrtki jet carries high-salinity water eastward, forming a high SSS tongue in the monsoon transition seasons of April–May and October–November (Yoshida 1959; Wyrtki 1973). South of the equator, the Indonesian Throughflow (ITF) conveys fresher water from the Indonesian seas to the southeastern TIO, maintaining a low SSS (Gordon et al. 1997). The South Equatorial Current (SEC) extends the low-salinity water to the west, forming a low SSS tongue (Song and Gordon 2004).

Earlier modeling studies simulated the impact of precipitation/freshwater flux on the ocean stratification and equatorial circulation in the TIO (Masson et al. 2002, 2003) and pointed out that the salinity stratification strongly affected the ocean dynamics and air–sea interaction (Masson et al. 2004). Among them, layer model experiments revealed that the IO circulations, including Indonesian Throughflow, redistributed the rainfall input and river discharges in the Bay of Bengal to adjacent ocean and induced changes in SST, mixed layer, barrier layer, and thermocline (Han and McCreary 2001; Han et al. 2001; Perigaud et al. 2003). Recent modeling and observational studies revealed that the SSS in the TIO has a significant year-to-year variability (Thompson et al. 2006; Illig and Perigaud 2007; Vinayachandran and Nanjundiah 2009; Subrahmanyam et al. 2011).

A major part of the variability is associated with the tropical climate modes, such as El Niño–South Oscillation (ENSO) and the Indian Ocean dipole (IOD; Saji et al. 1999; Webster et al. 1999). In the southern TIO, Phillips et al. (2005) suggested that the interannual variability of freshwater advection by ITF might be related to ENSO event. During positive IOD events, low and high anomalous SSS centers appear in the central equatorial IO and at the Java–Sumatra coast, respectively, which are attributed to anomalous easterly winds along the equator and enhanced upwelling off Java and Sumatra, respectively (Y. Zhang et al. 2013). Because of the synchronous variation of salinity with the IOD, Grunseich et al. (2011) proposed a salinity IOD index as a complementary to the definition based on sea surface temperature (SST) and found that it well characterizes the Bjerknes (1969) feedback. Furthermore, Y. Zhang et al. (2013) distinguished the difference between the influences of IOD and ENSO, and found that the IOD is more important on the SSS variation in the equatorial IO.

So far, ocean salinity studies were mainly based on in situ observations, either from traditional conductivity, temperature, and depth (CTD) profiles or newly employed Array for Real-Time Geotropic Oceanography (Argo) floats, and the ocean models. Recently, two satellites are providing global ocean surface salinity. One is the Soil Moisture and Ocean Salinity (SMOS) mission by the European Space Agency (ESA), and the other is the Aquarius/Satelite de Aplicaciones Cientificas-D (SAC-D) satellite mission by National Aeronautics and Space Administration (NASA) and the Space Agency of Argentina (Kerr et al. 2001; Lagerloef et al. 2008). Over four years of SSS data from space with global coverage are available at present, providing a good opportunity to advance our knowledge in freshwater budget over the ocean.

In this study, we combine SSS datasets from both SMOS and Aquarius with the Argo observations to investigate the TIO SSS variations associated with the IOD events since 2010. The results suggest that the two satellites generally well capture the year-to-year variability of SSS in the TIO. In particular, the change of tropical ocean circulation south of the equator is identified as the most important ocean dynamic process to the SSS variations associated with the IOD events.

We arrange the remainder of the paper as follows. After a description of the datasets and methods in section 2, we compare the general features of satellite SSS in the TIO and validate the satellite SSS datasets using Argo observations in section 3. Section 4 presents the SSS variations in the two recent IOD events and the associated ocean dynamics. A summary and discussion are given in section 5.

2. Data and methods

a. Aquarius and SMOS

The Aquarius/SAC-D satellite mission, launched on 10 June 2011, is a collaboration between the NASA and the Argentinean Space Agency (Lagerloef et al. 2008, 2013). The mission has a primary goal of measuring SSS of the global ocean from space for better understanding both climate change and the global water cycle. Aquarius is operating in a sun-synchronous polar orbit at 657-km altitude that repeats every seven days; it carries three radiometers and one scatterometer. Both the data collected by the radiometer and SST from another platform are used to derive salinity, and then the data are corrected by surface roughness using data from the scatterometer. We use the Aquarius level-3 standard version 3.0 monthly SSS with 1° horizontal resolution, provided by the Ocean Data Processing System (ODPS) of NASA Goddard Space Flight Center (GSFC) (Lagerloef et al. 2013). The data used in this study are from August 2011 to July 2014.

The SMOS satellite mission, launched on 2 November 2009, is a joint ESA/French Space Agency Earth Observation Program (Kerr et al. 2001). The SMOS uses an aperture synthesis radiometer to capture microwave emissions from Earth’s surface for measuring land soil moisture and ocean salinity. At an altitude of 758 km, the SMOS can achieve global coverage in every three days. The monthly composite of ESA level-3 SSS is available at a quarter-degree resolution. We use the data from January 2010 to June 2014 obtained from the Centre Aval de Traitement des Donnees SMOS (CATDS). The data processing details can be found in Reul et al. (2011). A detailed comparison/validation between SMOS SSS and Argo SSS can be found in Banks et al. (2012).

b. Argo and other satellite measurements

We use the Argo temperature and salinity (T/S) profiles, and gridded monthly Argo product. We also use the climatological monthly temperature and salinity of the World Ocean Atlas 2009 (WOA09) (Antonov et al. 2010; Locarnini et al. 2010) for comparative analysis. The Argo T/S profiles are available at the Global Ocean Data Assimilation Experiment (GODAE), and the monthly Argo product is provided by the Scripps Institution of Oceanography (Roemmich and Gilson 2009).

The isothermal and mixed layer depths, as well as the barrier layer thickness, were calculated from Argo temperature and salinity. The isothermal depth was calculated from temperature with a criterion of 0.8°C decrease from sea surface. The mixed layer depth was specified from a difference in potential density from the surface value, which is equivalent to a 0.8°C decrease in temperature (Kara et al. 2000). This criterion takes into account both temperature and salinity stratification. The barrier layer is an intermediate layer between the base of the mixed layer and the top of the thermocline (Lukas and Lindstrom 1991), the thickness of which indicate the influence of freshwater on the stratification.

In addition, we use precipitation, ocean surface wind, ocean surface currents, SST, sea surface height (SSH), and ocean color [Chlorophyll-a (Chl-a)] in this study. The Global Precipitation Climatology Project (GPCP) version 2.2 monthly data on 2.5° grids are available at NASA GSFC (Adler et al. 2003; Huffman et al. 2009). The monthly Advanced Scatterometer (ASCAT) surface wind stress on 0.25° grids is provided by the Institut Français de Recherche pour l’Exploitation de la MER (IFREMER)/Centre d’Exploitation et de Recherche Satellitaire (CERSAT) (Bentamy and Croize Fillon 2012). The monthly satellite field-derived ocean surface currents on 1° grid are provided by the Ocean Surface Current Analysis Real-Time (OSCAR) of the National oceanic and Atmospheric Administration (NOAA) (Bonjean and Lagerloef 2002). The NOAA version 2 Optimum Interpolation Sea Surface Temperature (OISST) data on 1° grids are obtained from the Physical Sciences Division (PSD) of the Earth System Research Laboratory (ESRL) (Reynolds et al. 2002). Weekly sea surface height anomaly (SSHA) data are from the Archiving, Validation, and Interpretation of Satellite Oceanographic data (AVISO) merged altimeter product on ⅓° grids. Monthly Moderate Resolution Imaging Spectroradiometer (MODIS)–Aqua Chl-a is provided by the Ocean Biology Processing Group on 9-km grids (Meister et al. 2005).

Our study focuses on the period from July 2010 to July 2014, although we use all available dataset to obtain climatological mean and seasonal cycle. We use Argo data from 2004 to 2013, GPCP rain rate from 1979 to 2013, OSCAR currents from 1993 to 2013, ASCAT wind stress from 2008 to 2013, AVISO SSHA from 1993 to 2013, OISST from 1982 to 2013, and MODIS–Aqua Chl-a from 2003 to 2012. The SMOS SSS values before June 2010 were very changeable due to variable configurations testing the functionalities of the instrument and low-level procedures for data acquisition and handling during the in-orbit commissioning phase (Corbella et al. 2011). The SSS anomalies from the two satellites were obtained by removing the multiyear average from each dataset, and then subtracting the climatological monthly mean of the Argo SSS in the period 2004–13, from which the multiyear average was already removed. In short, the satellite SSS anomalies are with respect to the Argo climatological seasonal cycle, but the mean differences were removed first. The anomalies of other variables are with respect to their own climatologies (over different periods).

3. General features of satellite SSS

a. Climatological mean

The averages of Aquarius and SMOS SSS are compared with the Argo and World Ocean Atlas 2009 (WOA09) climatology. The Aquarius SSS is averaged over three years, August 2011–July 2014; the SMOS SSS is averaged over four years, July 2010–June 2014; and the Argo SSS is averaged over 10 years, January 2004–December 2013. The Aquarius and SMOS SSS maps well capture the east–west contrast of SSS in the TIO (Fig. 1), generally constrained by E − P. The ocean circulation explains the misplacement between the SSS and rainfall between the latitude band of 15°S–5°N. The equatorial current contributes to the eastward extension of the high-salinity tongue in the 5°S–5°N band, while the SEC contributes to the westward extension of the low-salinity tongue in the 15°–5°S band. Besides, both the Aquarius and SMOS SSS maps show finer spatial structures than the WOA09 and Argo, such as the mesoscale structure south of India and Sri Lanka.

Fig. 1.

Multiyear averaged SSS (shaded) of Aquarius, SMOS, Argo, and WOA09, superimposed with annual mean GPCP precipitation in each period (contours in mm month−1): (a) Aquarius from August 2011 to July 2014, (b) SMOS from July 2010 to June 2014, (c) Argo from January 2004 to December 2013, and (d) WOA09 climatological annual mean. Superimposed in (c) are the OSCAR surface currents (vectors in m s−1) for the period of 1993–2013. The rectangles in (a) show two selected areas, the central equatorial IO (2°S–2°N, 70°–90°E) and the Java–Sumatra coast (12°–5°S, 100°–112°E) for analysis.

Fig. 1.

Multiyear averaged SSS (shaded) of Aquarius, SMOS, Argo, and WOA09, superimposed with annual mean GPCP precipitation in each period (contours in mm month−1): (a) Aquarius from August 2011 to July 2014, (b) SMOS from July 2010 to June 2014, (c) Argo from January 2004 to December 2013, and (d) WOA09 climatological annual mean. Superimposed in (c) are the OSCAR surface currents (vectors in m s−1) for the period of 1993–2013. The rectangles in (a) show two selected areas, the central equatorial IO (2°S–2°N, 70°–90°E) and the Java–Sumatra coast (12°–5°S, 100°–112°E) for analysis.

b. Satellite versus Argo SSS

There are differences between satellite and Argo SSS on the order of 0.3 psu in these datasets. Figure 2 shows the difference of data pairs. The periods for each data pair used in comparison are the same. Large discrepancies between SMOS and Argo SSS appear in the regions close to the continent, mainly because SMOS suffers from so-called land-induced contaminations (e.g., Reul et al. 2012). Radio frequency interferences (RFIs) and orbit orientation (Banks et al. 2012; Boutin et al. 2012; Kainulainen et al. 2012), as well as the smoothing of Argo objective interpolation in regions with inadequate samplings (Boutin et al. 2013), also result in the differences.

Fig. 2.

Mean SSS differences (shadings) and the corresponding RMS (contours) between (a) Aquarius and Argo for the period from August 2011 to July 2014, (b) SMOS and Argo from July 2010 to June 2014, and (c) Aquarius and SMOS from August 2011 to June 2014. (d)–(f) Time series corresponding to (a)–(c) in the central equatorial IO (2°S–2°N, 70°–90°E; red curves) and the Java–Sumatra coast (12°–5°S, 100°–112°E; blue curves), superimposed with the GPCP precipitation (mm month−1) in the central equatorial IO [dashed red in (d)] and the Java–Sumatra coast [dashed blue in (e)].

Fig. 2.

Mean SSS differences (shadings) and the corresponding RMS (contours) between (a) Aquarius and Argo for the period from August 2011 to July 2014, (b) SMOS and Argo from July 2010 to June 2014, and (c) Aquarius and SMOS from August 2011 to June 2014. (d)–(f) Time series corresponding to (a)–(c) in the central equatorial IO (2°S–2°N, 70°–90°E; red curves) and the Java–Sumatra coast (12°–5°S, 100°–112°E; blue curves), superimposed with the GPCP precipitation (mm month−1) in the central equatorial IO [dashed red in (d)] and the Java–Sumatra coast [dashed blue in (e)].

The differences between Aquarius and Argo SSS show a basinwide low, with an average of about −0.2 psu, and a maximum of about −0.6 psu in the Bay of Bengal and the region west of Java and Sumatra (Fig. 2a). A tongue structure of differences extends from the southeastern TIO to the central equatorial IO, resembling the pattern of the annual mean rainfall and suggesting a relationship between Aquarius and Argo SSS. The spatial pattern of standard deviation of the differences [root-mean-square (RMS)] is similar to the pattern of the mean. This result is consistent with previous studies that showed large differences in the region of low SSS under intensive rainfall (Henocq et al. 2010; Grodsky et al. 2012; Reverdin et al. 2012; Boutin et al. 2013; Qu et al. 2014). The differences in the central equatorial IO and Java–Sumatra coast show seasonal dependence, roughly correlated with the seasonal rainfall; for example, off the Java–Sumatra coast, the correlation coefficient is 0.39 between the Aquarius-minus-Argo SSS difference and rainfall, and is 0.51 between the SMOS-minus-Argo SSS difference and rainfall, both having the statistical significance at the 95% confidence level on a t test (Figs. 2d,e). Off Java and Sumatra, the differences are the maximum when the rainfall reaches the maximum in the austral summer and fall seasons (Fig. 2e).

The differences can also be found between SMOS and Argo SSS (Fig. 2b). Although many efforts have been made to reduce RFIs, residual effects are clearly seen in SMOS SSS. Aquarius and SMOS SSS maps have different patterns. It is probably due to the differences in their azimuth angles, orientation of the satellite footprints, and retrieval algorithms (e.g., Bindlish et al. 2013; Lagerloef et al. 2013). There are significant differences to note. Unlike the Aquarius SSS, there is no evident SSS difference in the central equatorial IO between SMOS and Argo observations (Fig. 2). Large SSS discrepancies occur in the western TIO between SMOS and Argo, largely due to the land contaminations in the SMOS SSS (Reul et al. 2012). Off Java and Sumatra, both the Aquarius and SMOS SSS values are consistently lower than the Argo observations (Figs. 2 and 3). The differences between SMOS and Argo SSS showed significant changes after February 2011, reduced from <−0.6 psu in September–November 2010 to about −0.1 psu afterward. The time series of rainfall suggests that the largest differences in September–November 2010 might be associated with the negative IOD event in 2010, when warmer SST and enhanced deep convection induced much heavier rainfall (Qiu et al. 2012). We will come back to this point in the following sections.

Fig. 3.

Scatter diagrams of SSS between satellite and Argo data in the central equatorial IO (red dots) and between satellite and Argo data at the Java–Sumatra coast (blue dots): (a) Aquarius and Argo SSS, (b) SMOS and Argo SSS, and (c) Aquarius and SMOS SSS. (d)–(f) As in (a)–(c), but for SSS with its multiyear average removed in each corresponding period at each location.

Fig. 3.

Scatter diagrams of SSS between satellite and Argo data in the central equatorial IO (red dots) and between satellite and Argo data at the Java–Sumatra coast (blue dots): (a) Aquarius and Argo SSS, (b) SMOS and Argo SSS, and (c) Aquarius and SMOS SSS. (d)–(f) As in (a)–(c), but for SSS with its multiyear average removed in each corresponding period at each location.

Paired comparison of Aquarius and Argo SSS, as well as that of SMOS and Argo SSS, is illustrated in Fig. 3 over two selected regions as indicated by the two boxes in Fig. 1. These two regions are sensitive to the interannual variation associated with the IOD (Grunseich et al. 2011; Durand et al. 2013; Y. Zhang et al. 2013). In the central equatorial IO, the three datasets generally are consistent with each other. Off Java and Sumatra, Aquarius and SMOS SSS values have better consistency with each other, especially when the symmetric biases are removed (Figs. 3c,f).

Before doing analysis on the interannual variation, the symmetric biases have to be removed. In other words, we removed the mean from each SSS dataset (e.g., Figs. 35) and then removed the seasonal cycle of Argo SSS from all three datasets to obtain the interannual anomalies (e.g., Figs. 68, 11, and 13), since satellite SSS is too short to create its own climatological seasonal cycle.

4. Features in IOD events

a. SSS variation in the equatorial region

The two time series in the central equatorial IO and the region off Java and Sumatra demonstrate salinity variation in the interannual time scale (Fig. 4). Note that in Figs. 4b–d and 4f–h the seasonal cycle is included. The negative (positive) anomalies refer to the part of the curve lower (higher) than Argo climatological seasonal cycle. There were two positive (July–September 2011 and June–September 2012) and three negative (September–November 2010, May–August 2013, and an ongoing event from July 2014) IOD events since 2010 (Fig. 4a). In the central equatorial IO, Aquarius and SMOS observed SSS changes in the IOD as Argo did (Figs. 4b–d). The equatorial current dominated the zonal advection (Wyrtki 1973). The decrease of SSS in positive IOD event reflected fresher water advection from the east, and vice versa in negative IOD event. Off Java and Sumatra, Aquarius and SMOS observed the SSS anomalies but Argo did not (Figs. 4f–h). It is probably due to inadequate samplings in the coastal region, since Argo also did not observe significant SST change as captured by OISST.

Fig. 4.

Time series of SSS (with multiyear average removed in each corresponding period) and indices/variables for the period from July 2010 to July 2014. On the left are (a) DMI (black curve) and ASCAT zonal wind (blue curve) in the central equatorial IO, and the SSS of (b) Aquarius (black curve), (c) SMOS (red curve), and (d) Argo (blue curve) in the central equatorial IO. The dashed curves in (b)–(d) show the climatological seasonal cycle of Argo SSS over 2004–13 in the region. On the right are (e) OISST (black curve) and Argo SST (dashed curve) anomalies in the Java–Sumatra coast. (f)–(h) As in (b)–(d), but in the Java–Sumatra coast. Values of DMI in (a) and OISST in (e) larger than 0.7 times the standard deviation are highlighted. The dashed curves in (f)–(h) show the climatological seasonal cycle of Argo SSS over 2004–13 in the region.

Fig. 4.

Time series of SSS (with multiyear average removed in each corresponding period) and indices/variables for the period from July 2010 to July 2014. On the left are (a) DMI (black curve) and ASCAT zonal wind (blue curve) in the central equatorial IO, and the SSS of (b) Aquarius (black curve), (c) SMOS (red curve), and (d) Argo (blue curve) in the central equatorial IO. The dashed curves in (b)–(d) show the climatological seasonal cycle of Argo SSS over 2004–13 in the region. On the right are (e) OISST (black curve) and Argo SST (dashed curve) anomalies in the Java–Sumatra coast. (f)–(h) As in (b)–(d), but in the Java–Sumatra coast. Values of DMI in (a) and OISST in (e) larger than 0.7 times the standard deviation are highlighted. The dashed curves in (f)–(h) show the climatological seasonal cycle of Argo SSS over 2004–13 in the region.

The scatter diagrams further demonstrate the relationship between SSS and IOD (Fig. 5). The satellite SSS correlation with IOD mode index (DMI) averaged in the central equatorial IO is opposite to that on the Java–Sumatra coast. In Fig. 5, based on linear regression, 1°C DMI corresponds to about a 0.6-psu change in the region off Java and Sumatra, or to a −0.4-psu change in the central equatorial IO, which means that the SSS in either region could indicate IOD events. Since Argo floats can last much longer time in the open oceans, it implies that the DMI can be defined and IOD can be monitored by SSS in the central equatorial IO (Grunseich et al. 2011).

Fig. 5.

Scatter diagrams of DMI vs Aquarius, SMOS, and Argo SSS: (a)–(c) the central equatorial IO (red dots) and (d)–(f) the Java–Sumatra coast (blue dots). The multiyear average in each SSS dataset and each box is removed.

Fig. 5.

Scatter diagrams of DMI vs Aquarius, SMOS, and Argo SSS: (a)–(c) the central equatorial IO (red dots) and (d)–(f) the Java–Sumatra coast (blue dots). The multiyear average in each SSS dataset and each box is removed.

A longitude–time diagram of the anomalous SSS along the equator is shown in Fig. 6. IOD-related SSS variations are evident in all three datasets. In 2010 negative IOD event, SSS anomalies lasted more than half a year, from June 2010 to March 2011 (Figs. 6b,c). Recent studies suggested that the 2010 event might be the most significant negative IOD event in the recent decade (e.g., Qiu et al. 2012). Two weak positive IOD events occurred successively in 2011 and 2012, reflected in SSS. Aquarius captured more subseasonal time scale processes (Fig. 6a). Large differences exist between satellite and Argo SSS in the western equatorial region where the number of Argo floats was much less and the subseasonal time scale processes were not well observed (Fig. 6d).

Fig. 6.

Time–longitude diagrams of SSS anomalies (shadings) in (a) Aquarius, (b) SMOS, and (c) Argo, and (d) number of Argo profiles, between 2°S and 2°N. The SSS anomalies are obtained by removing the multiyear average in each corresponding period and the seasonal cycle in Argo SSS of 2004–13 (the seasonal cycle is obtained after its multiyear average is removed). Superimposed in (c) are the OSCAR zonal velocity anomalies (contours in m s−1).

Fig. 6.

Time–longitude diagrams of SSS anomalies (shadings) in (a) Aquarius, (b) SMOS, and (c) Argo, and (d) number of Argo profiles, between 2°S and 2°N. The SSS anomalies are obtained by removing the multiyear average in each corresponding period and the seasonal cycle in Argo SSS of 2004–13 (the seasonal cycle is obtained after its multiyear average is removed). Superimposed in (c) are the OSCAR zonal velocity anomalies (contours in m s−1).

b. Individual cases

The SSS patterns show significant changes between the negative and positive IOD events in 2010 and 2012, respectively (Fig. 7). Along the equator, the high-salinity tongue shifted ~15° in longitude, while along the east coast low-salinity water moved meridionally. The former change was mostly due to enhanced or weakened tropical circulations in the southern TIO, especially in the southeastern TIO (Figs. 7b,f). The later change might relate to the southeastward or northwestward coastal currents (Figs. 7b,f), absent or strengthened upwelling, or positive or negative anomalous precipitation (Figs. 8b and 9c) off Java and Sumatra. The SSS also changed in other regions, such as south and east of Sri Lanka, where an anomalous eastward current formed. Compared with the Argo SSS, Aquarius and SMOS observed much finer SSS structures (Figs. 7a,d,e), such as the eddy structures around the southern tip of the Indian subcontinent.

Fig. 7.

September–October SSS patterns in 2010 (negative IOD event) and in 2012 (positive IOD event). (a) SMOS and (b) Argo SSS, and (c) number of Argo profiles. (d) Aquarius, (e) SMOS, and (f) Argo SSS, and (g) number of Argo profiles. The September–October OSCAR currents (vectors in m s−1) are superimposed in (b) and (f). The triangles for 2010 and circles for 2012 indicate the locations of selected Argo T/S profiles shown in Fig. 12 in the central equatorial IO (black), Java–Sumatra coast (red), and south of the equatorial IO (blue).

Fig. 7.

September–October SSS patterns in 2010 (negative IOD event) and in 2012 (positive IOD event). (a) SMOS and (b) Argo SSS, and (c) number of Argo profiles. (d) Aquarius, (e) SMOS, and (f) Argo SSS, and (g) number of Argo profiles. The September–October OSCAR currents (vectors in m s−1) are superimposed in (b) and (f). The triangles for 2010 and circles for 2012 indicate the locations of selected Argo T/S profiles shown in Fig. 12 in the central equatorial IO (black), Java–Sumatra coast (red), and south of the equatorial IO (blue).

Fig. 8.

Bimonthly anomalies during 2010 (negative IOD event), showing July–August (JA), September–October (SO), November–December (ND), and the following January–February (JF). (a)–(d) SMOS SSS (shadings) and GPCP precipitation (contours in mm month−1); (e)–(h) Argo SSS (shadings), mixed layer depth (contours in m), and OSCAR currents (vectors in m s−1).

Fig. 8.

Bimonthly anomalies during 2010 (negative IOD event), showing July–August (JA), September–October (SO), November–December (ND), and the following January–February (JF). (a)–(d) SMOS SSS (shadings) and GPCP precipitation (contours in mm month−1); (e)–(h) Argo SSS (shadings), mixed layer depth (contours in m), and OSCAR currents (vectors in m s−1).

Fig. 9.

(a)–(h) As in Fig. 8, but for 2012 (positive IOD event) and starting from May–June (MJ). (i)–(l) Bimonthly anomalies of Aquarius SSS and barrier layer thickness (contours in m).

Fig. 9.

(a)–(h) As in Fig. 8, but for 2012 (positive IOD event) and starting from May–June (MJ). (i)–(l) Bimonthly anomalies of Aquarius SSS and barrier layer thickness (contours in m).

Figures 8 and 9 show the observations of SSS, rainfall, currents, and mixed layer depth during the evolution of negative and positive IOD events in 2010 and 2012, respectively. Since Aquarius SSS is available only after 2011, only SMOS SSS is compared with Argo SSS in the negative IOD event in 2010. In the central equatorial IO, significant positive SSS anomalies appeared from July to August 2010, and reached the maximum in November–December. Anomalous currents could explain the increase in SSS, which advected high-salinity water eastward along the equator. Off Java and Sumatra, negative SSS anomalies appeared and reached the maximum in September–October. The convergence in the eastern equatorial IO suppressed the upwelling, which contributed to the decrease of SSS off Java and Sumatra as well as the increased precipitation. A weaker SSS increase in SMOS than in Argo around 80°E in September–October was probably due to the fact that the satellite observes the skin surface salinity, which can be easily changed by the freshwater input from heavy rainfall and river discharges (Fig. 8b), while the uppermost salinity sampling level of Argo float is about 5 m below the surface (Riser et al. 2008). The 5-m salinity is much higher than the surface when a thin freshwater cap forms (Delcroix and McPhaden 2002). South of the equator, an anomalous cyclonic gyre maintained anomalously high SSS in the decay phase of the IOD. This process is consistent with the finding of Y. Zhang et al. (2013). The opposite processes existed in 2012 positive IOD event. Anomalous westward equatorial current advected low-salinity water to the central equatorial region, while enhanced upwelling off Java and Sumatra brought high-salinity water to the surface. South of the equator, an anomalous anticyclonic gyre maintained low SSS there.

In Fig. 8, positive SSS anomalies persisted into January–February 2011 and extended farther to the south of the equator. Since September–October 2010, the rainfall increased rather than decreased, implying that freshwater flux is not the cause of the persistence of positive SSS anomalies in the open ocean south of the equator. Similar negative SSS anomalies existed in 2012 (Fig. 9). It is the anomalous gyre in the southeastern TIO that contributed to the change of SSS, influencing the high-salinity tongue along the equator and the low-salinity tongue in the south (Fig. 7). In 2010 both tongues strengthened, whereas in 2012 the high-salinity tongue disappeared and only the westward low-salinity tongue formed.

c. Impact of Rossby waves

The 2010 negative IOD concurred along with a La Niña in the Pacific. Near the La Niña mature phase (September–December), cyclonic wind circulation was centered south of 10°S in the southeastern TIO (Fig. 10a), associated with enhanced westerly wind near the equator. The positive wind curl forced upwelling oceanic Rossby waves (Fig. 10b). Ekman divergences induced upwelling and brought the thermocline water into the mixed layer, which then increased SSS. High Chl-a indicated there were high-nutrient water from the subsurface (Fig. 10c). A cyclonic ocean gyre provided a large part of the eastward water transport along the equator (Figs. 7b and 8f,g).

Fig. 10.

Anomalies of (a),(d) ASCAT wind stress (2008–13; vectors in N m−2) and Ekman pumping velocity (shadings in m month−1), (b),(e) AVISO SSH (1993–2013; shadings in cm) and OISST (1982–2013; contours in °C), and (c),(f) MODIS–Aqua Chl-a (2003–12; mg m−3) in September–October during (left) 2010 (negative IOD event) and (right) 2012 (positive IOD event).

Fig. 10.

Anomalies of (a),(d) ASCAT wind stress (2008–13; vectors in N m−2) and Ekman pumping velocity (shadings in m month−1), (b),(e) AVISO SSH (1993–2013; shadings in cm) and OISST (1982–2013; contours in °C), and (c),(f) MODIS–Aqua Chl-a (2003–12; mg m−3) in September–October during (left) 2010 (negative IOD event) and (right) 2012 (positive IOD event).

An opposite case occurred in the 2012 positive IOD event. Only weak positive SST anomalies appeared in the central Pacific, not strong enough to form an El Niño (not shown). Even so, an anomalous anticyclonic wind-driven circulation formed in the southeastern TIO, but centered north of 10°S, which was different from the case in 2010. This indicated the different influence from IOD and ENSO (Rao and Behera 2005; Yu et al. 2005). Because of the differences in atmospheric forcing, the anomalous ocean circulations associated with IOD are more confined to the region north of 10°S, whereas ENSO-induced anomalies are dominant south of 10°S (Rao and Behera 2005; Yu et al. 2005). A downwelling oceanic Rossby wave impeded the entrainment at the bottom of the mixed layer and no Chl-a bloom formed. From July–August to November–December (Figs. 7 and 9), the anticyclonic ocean gyre recirculated the low-salinity water and maintained a negative SSS south of the equator.

Over the study region, no significant SST changes were found in 2010 negative IOD event. It seems that the shoaling of mixed layer (Fig. 8), which causes surface layer to warm up more easily, compensated a major part of the cooling effect of Ekman pumping (Fig. 10b). However, a rather large increase of SST appeared in 2012 positive IOD event (Fig. 10e). It can probably be attributed to the low-salinity water advected from the east, which had a shoaling effect on the mixed layer and favored SST warming. Figure 10e shows the downwelling Rossby wave in 2012 with similar amplitude as the upwelling Rossby wave in 2010, but the change of the mixed layer depth in 2012 was much smaller (Figs. 8 and 9). Figures 9e–h indicate that the mixed layer even shoaled in some regions. Figures 9i–l show that the barrier layer formed in those regions, which favored SST warming.

Through shoaling and deepening thermocline and mixed layer, Rossby waves can influence SSS. The SSH, depth of 22°C water (Z22), and SSS show westward copropagation in the longitude–time diagrams (Fig. 11). The phase speed was about 0.2 m s−1, reflecting Rossby wave propagation at this latitude, consistent with previous studies (Xie et al. 2002; Chowdary et al. 2009). In both negative (2010, 2013, and 2014) and positive (2011 and 2012) IOD events, significant changes were captured by satellite SSS and SSHA, and by Argo-based Z22 and mixed layer depth.

Fig. 11.

Anomalies of (a) Aquarius, (b) SMOS, and (c) Argo SSS, (d) AVISO SSH (shadings in cm), (e) depth of 22°C (shading in m), and (f) mixed layer depth (shadings in m) averaged in 12°–8°S. Anomalies of GPCP precipitation (contours in mm month−1), ASCAT Ekman pumping (contours in m month−1), barrier layer thickness (contours in m), and OISST (contours in °C) are superimposed in (c)–(f), respectively.

Fig. 11.

Anomalies of (a) Aquarius, (b) SMOS, and (c) Argo SSS, (d) AVISO SSH (shadings in cm), (e) depth of 22°C (shading in m), and (f) mixed layer depth (shadings in m) averaged in 12°–8°S. Anomalies of GPCP precipitation (contours in mm month−1), ASCAT Ekman pumping (contours in m month−1), barrier layer thickness (contours in m), and OISST (contours in °C) are superimposed in (c)–(f), respectively.

Significant positive SSS anomalies persisted from August 2010 to July 2011 (Figs. 11b,c). These can be separated into two periods. In August–December 2010, thermocline and mixed layer depth shoaled more than 20 m, corresponding to 20-cm depression of SSH, induced by divergence in the upper ocean (Figs. 11d–f). The uplifted thermocline not only shoaled the mixed layer, but also favored high-salinity water entering the mixed layer and thus increased SSS. Although precipitation increased east of 80°E (Fig. 11c), SSS increased there. In the second period when Rossby wave propagated to the west, it enhanced the open ocean upwelling in the southwestern TIO, where wind shear maintained a thermocline dome (Figs. 11d,e) (Xie et al. 2002; Du et al. 2009). Thus, significant high SSS persisted to July 2011. The decrease of SST in the second period confirms that this upwelling Rossby wave phase-locked with the seasonal upwelling peak in February–August (Fig. 11f). The negative SST anomalies persisted until July 2011 with a peak of −0.9°C in May 2011. The cooling effect had a feedback to the precipitation decreasing from October 2010 to July 2011 (Fig. 11c). A similar case occurred in 2013. The processes were robust in both satellite and Argo observations.

Downwelling Rossby wave had an opposite effect on the SSS. Aquarius, SMOS, and Argo datasets show the evolution of SSS anomalies in 2011 and 2012. In 2011, negative SSS signature was weak, probably due to weak wind forcing and the remaining effect of the previous negative IOD event in 2010. In 2012, without concurrent El Niño, the effect of downwelling Rossby waves terminated in December, as did the SSS variation. The reemergence of negative SSS anomalies in March 2013 was due to the seasonal enhancement of precipitation (Fig. 11c). Local rainfall seemed important to keep the low SSS. A break in SSS and SSH in February 2013 suggests the possible impact of intreaseaonal variation, where the Madden–Julian oscillation (MJO) is a dominant climate mode (Madden and Julian 1972; C. Zhang et al. 2013), with favorable high-SST condition (Figs. 11e,f).

d. Vertical structure

Individual Argo profiles in 2010 negative and 2012 positive IOD events are compared along with their seasonal climatology in the three regions, namely, the central equatorial IO, the Java–Sumatra coast, and south of the equatorial IO (Fig. 12). These sites are marked in Fig. 7. The values of satellite SSS are marked in Fig. 12.

Fig. 12.

Individual Argo temperature (blue curves in °C)/salinity (black curves; T/S) profiles in the three regions: (a),(d) the central equatorial IO, (b),(e) the Java–Sumatra coast, and (c),(f) south of the equatorial IO (8°–12°S, 80°–95°E) in (top) October 2010 and (bottom) October 2012. The dashed curves indicate the climatological T/S profiles in the 1° × 1° grids nearest to the individual Argo profiles in the three regions. Black horizontal line indicates the mixed layer depth (m). The locations and times of the profiles for (a)–(f) are 1.108°N, 79.985°E on 10 Oct 2010; 5.480°S, 100.312°E on 15 Oct 2010; 9.119°S, 85.003°E on 27 Oct 2010; 1.787°S, 76.103°E on 26 Oct 2012; 10.482°S, 106.241°E on 28 Oct 2012; and 10.908°S, 86.849°E on 2 Oct 2012, respectively. Satellite SSS over each profile is marked by dots for Aquarius and triangles for SMOS.

Fig. 12.

Individual Argo temperature (blue curves in °C)/salinity (black curves; T/S) profiles in the three regions: (a),(d) the central equatorial IO, (b),(e) the Java–Sumatra coast, and (c),(f) south of the equatorial IO (8°–12°S, 80°–95°E) in (top) October 2010 and (bottom) October 2012. The dashed curves indicate the climatological T/S profiles in the 1° × 1° grids nearest to the individual Argo profiles in the three regions. Black horizontal line indicates the mixed layer depth (m). The locations and times of the profiles for (a)–(f) are 1.108°N, 79.985°E on 10 Oct 2010; 5.480°S, 100.312°E on 15 Oct 2010; 9.119°S, 85.003°E on 27 Oct 2010; 1.787°S, 76.103°E on 26 Oct 2012; 10.482°S, 106.241°E on 28 Oct 2012; and 10.908°S, 86.849°E on 2 Oct 2012, respectively. Satellite SSS over each profile is marked by dots for Aquarius and triangles for SMOS.

In October 2010, an increase of salinity reached 90-m depth in the central equatorial IO, with the maximum of 0.6 psu at the surface (Fig. 12a). Under enhanced westerly wind, the mixed layer deepened from 47 to 89 m. The temperature did not change much. Off Java and Sumatra, no Argo profile was found in the strong upwelling area, so the nearest profile was selected (Fig. 12b). The salinity and temperature did not change much at the surface, but the mixed layer deepened more than 50 m, from 36 to 94 m, reflecting strong convergence induced by enhanced westerly wind. The less saline water deepened by 30 m, from 35 to 65 m, much shallower than the isothermal layer depth, indicating a thicker barrier layer there (Fig. 12b). The eastward high-salinity water did not reach this region, while fresher water flushed the region along the coast from the north (Fig. 7).

A significant change in salinity appeared south of the equatorial IO (Fig. 12c). With an upwelling Rossby wave, the mixed layer was 29 m deep in October 2010, shoaling by more than 30 m than in the climatology. The maximum salinity was located at the depth of 45 m, 1.1 psu higher than its seasonal climatology; the salinity was 0.3 psu higher than its seasonal climatology at 100 m. The maximum salinity water at the subsurface did not come from deeper layers; rather, it came from the west (Figs. 7 and 8). The salinity in the surface layer was higher than its seasonal climatology. High Chl-a implies high-salinity water at the surface coming from the subsurface, bringing in high nutrients or higher subsurface Chl-a (Fig. 10e).

In 2012, the above changes reversed. In the central equatorial IO, the salinity decreased by 0.3 psu in the top 30 m (Fig. 12d). However, the salinity increased at 50–120 m, implying an enhanced high-salinity water intrusion to the east in the subsurface layer, probably due to a stronger undercurrent. It further increased the salinity in the subsurface layer off Java/Sumatra, with 0.1–0.2 psu higher in 50–150 m, while not much change was found at the surface (Fig. 12e). The opposite changes of salinity in the central equatorial IO suggest an anomalous zonal shallow overturning along the equator (Fig. 12d), which needs further investigation.

South of the equatorial IO, downwelling Rossby wave deepened the thermocline with fresher water in the surface layer, decreasing the salinity all the way to at least 200 m, with a maximum at 100 m (Fig. 12f). The scatter diagram implies a close relationship between the SSS in the mixed layer and the thermocline depth; it also indicates nonsymmetric effects between upwelling and downwelling Rossby waves (Fig. 13), either on SSS or on Chl-a (Fig. 10). Upwelling brings more significant effects on SSS than downwelling does, due to the stratification, particularly the characteristics of the mixed layer. When upwelled high-salinity water entered the surface layer, mixing made water properties uniform in the vertical quickly. For downwelling, convergence deepened the mixed layer and the thermocline at the same time, so the salinity in the mixed layer hardly changed.

Fig. 13.

Scatter diagrams of Argo SSS and subsurface salinity at 100 m south of the equatorial IO (12°–8°S, 80°–95°E) over the period of January 2010–July 2014: (a) salinity and (b) anomalous salinity with respect to 2004–13 mean seasonal cycle.

Fig. 13.

Scatter diagrams of Argo SSS and subsurface salinity at 100 m south of the equatorial IO (12°–8°S, 80°–95°E) over the period of January 2010–July 2014: (a) salinity and (b) anomalous salinity with respect to 2004–13 mean seasonal cycle.

5. Summary and discussion

Satellite salinity measurements, by the SMOS and Aquarius missions, provide a global coverage of SSS from the space for the first time. Based on over four years of satellite SSS, we investigated the salinity variations associated with the IOD mode. The observed salinity variation indicated a possible relationship with the IOD-related ocean circulation.

Compared with Argo SSS, the SMOS/Aquarius SSS values were about 0.2 psu lower in the TIO as a whole, partially reflecting the difference between the skin salinity by satellites and 5-m salinity by Argo floats. The difference of <−0.6 psu from Argo SSS were found in the coastal and deep convection regions, due to land-induced contamination and intensive rainfall, and there was no significant difference in the open ocean between satellite and Argo SSS. Both SMOS and Aquarius well captured the interannual variation of SSS during 2010–14, during which two positive and three negative IOD events occurred. This study confirms the capability of satellite in observing SSS. In particular, the satellite can capture the SSS anomaly in the upwelling region off the Java–Sumatra coast, where Argo floats are sparse.

SMOS and Aquarius well captured the development, evolution, and decay of the SSS variation associated with the IOD since 2010. The schematic diagrams in Fig. 14 illustrate the key processes. During the positive IOD event, particularly in a strong event, the westerly wind reversed to be easterly along the equator, impeding the eastward advection of high-salinity water originated from the Arabian Sea. Surface flow advected fresher water westward. Off Java and Sumatra, coastal upwelling brought high-salinity water into the surface layer. In the southern TIO, an anomalous anticyclonic circulation weakened the tropical circulation, which impeded the SEC but favored the water exchange in the meridional direction. Both the high-salinity tongue along the equator, driven by the Wyrtki jet, and the low-salinity tongue in the south, driven by the SEC, became weak. In the southern TIO, downwelling Rossby wave deepened the thermocline and prevented high-salinity water from entering the mixed layer, which favored low SSS on the east flank of the anomalous anticyclonic gyre. Along with above ocean processes, precipitation increased in the central equatorial IO and decreased in the southeastern TIO during the mature phase of IOD, which further decreased and increased the SSS in the central equatorial IO and Java–Sumatra coast, respectively.

Fig. 14.

Schematic diagrams of the forcing mechanisms explaining the SSS variability during (a) positive and (b) negative IOD events. The background map (shadings) uses SMOS SSS in September–October of 2012 and 2010 in (a) and (b), respectively, with needed smoothing for better illustration.

Fig. 14.

Schematic diagrams of the forcing mechanisms explaining the SSS variability during (a) positive and (b) negative IOD events. The background map (shadings) uses SMOS SSS in September–October of 2012 and 2010 in (a) and (b), respectively, with needed smoothing for better illustration.

The above processes reversed during negative IOD event (Fig. 14b). The enhanced Wyrtki jet advected high-salinity water to the Java–Sumatra coast, where coastal downwelling weakened the seasonal upwelling and maintained low SSS. In the south, an anomalous cyclonic gyre enhanced the SEC. Both low- and high-salinity tongues were strengthened and extended. Upwelling Rossby waves shoaled the thermocline, favored high-salinity water entering the mixed layer, and thus maintained the high SSS on the west flank of the anomalous cyclonic gyre. High Chl-a confirmed high-nutrient water coming from the subsurface (Fig. 10c). The scatter diagrams indicate a close relationship between surface and subsurface salinity (Fig. 13). In addition, they reveal that the relationship was more robust in the negative IOD event, due to the nature of water exchange between the mixed layer and thermocline. With surface mixing, upwelled high-salinity water could increase SSS easily.

An issue worth further consideration is the barrier layer. The region we studied is characterized by intense convective precipitation, where a fresher water cap and salinity stratified barrier layer commonly exist (e.g., Masson et al. 2004; Qu and Meyers 2005). A few studies revealed that the barrier layer is important to the heat exchange between the mixed layer and thermocline, thus having impact on the evolution of the IOD (e.g., Du et al. 2005; Du and Xie 2008; Qiu et al. 2012), like its role in the ENSO dynamics (e.g., Delcroix and McPhaden 2002; Qu et al. 2014). This work suggests there was a feedback between low-salinity water, thin mixed layer, and SST increase. Through monitoring the SSS variation with SMOS and Aquarius, together with the Argo and other satellite observations, including rainfall, SST, and wind, the role of barrier layer or fresher water cap in changing heat fluxes can be better understood.

Acknowledgments

We thank Dr. Yi.Chao and Dr. Shang-Ping Xie for their valuable and constructive comments, which helped to improve the manuscript. We acknowledge the NASA/GSFC/ODPS for providing the Aquarius salinity data (http://oceandata.sci.gsfc.nasa.gov) and the LOCEAN/IPSL researchers for producing and making available the SMOS salinity data (http://www.salinityremotesensing.ifremer.fr). The Argo product was provided by the International Argo Program (http://www.argo.ucsd.edu). The GPCP precipitation was obtained from the NASA GSFC (http://precip.gsfc.nasa.gov). The ASCAT wind was available at IFREMER/CERSAT (http://cersat.ifremer.fr). The OSCAR currents were provided by NOAA’s Ocean Surface Current Analyses Real Time (http://www.oscar.noaa.gov/index.html). The OISST was obtained from the NOAA/Earth System Research Laboratory (http://www.esrl.noaa.gov). The SSHA was provided by the AVISO (http://www.aviso.oceanobs.com/en). The MODIS-Aqua Chl-a was obtained from the OceanColor WEB at the NASA GSFC (http://oceancolor.gsfc.nasa.gov). This work is supported by the Ministry of Science and Technology (2010CB950302, 2012CB955603), the Chinese Academy of Sciences (XDA11010103, XDA11010203), the National Science Foundation of China (41176024), and the CAS/SAFEA International Partnership Program for Creative Research Teams.

REFERENCES

REFERENCES
Adler
,
R. F.
, and Coauthors
,
2003
:
The version 2.1 Global Precipitation Climatology Project (GPCP) Monthly Precipitation Analysis (1979–present)
.
J. Hydrometeor.
,
4
,
1147
1167
, doi:.
Antonov
,
J. I.
, and Coauthors
,
2010
: Salinity. Vol. 2, World Ocean Atlas 2009, NOAA Atlas NESDIS 69, 184 pp.
Banks
,
C. J.
,
C. P.
Gommenginger
,
M. A.
Srokosz
, and
H. M.
Snaith
,
2012
:
Validating SMOS ocean surface salinity in the Atlantic with Argo and operational ocean model data
.
IEEE Trans. Geosci. Remote Sens.
,
50
,
1688
1702
, doi:.
Bentamy
,
A.
, and
D.
Croize Fillon
,
2012
:
Gridded surface wind fields from Metop/ASCAT measurements
.
Int. J. Remote Sens.
,
33
,
1729
1754
, doi:.
Bindlish
,
R.
,
T.
Jackson
,
T.
Zhao
,
G.
Lagerloef
,
D.
Le Vine
,
S.
Yueh
, and
Y.
Kerr
,
2013
: Intercomparison of Aquarius and SMOS brightness temperature observations. SMOS and Aquarius Science Workshop, Brest, France, IFREMER. [Available online at http://aquarius.umaine.edu/cgi/meet_more.htm?id=46.]
Bjerknes
,
J.
,
1969
:
Atmospheric teleconnections from the equatorial Pacific
.
Mon. Wea. Rev.
,
97
,
163
172
, doi:.
Bonjean
,
F.
, and
G. S. E.
Lagerloef
,
2002
:
Diagnostic model and analysis of the surface currents in the tropical Pacific Ocean
.
J. Phys. Oceanogr.
,
32
,
2938
2954
, doi:.
Boutin
,
J.
,
N.
Martin
,
X.
Yin
,
J.
Font
,
N.
Reul
, and
P.
Spurgeon
,
2012
:
First assessment of SMOS data over open ocean: Part II—Sea surface salinity
.
IEEE Trans. Geosci. Remote Sens.
,
50
,
1662
1675
, doi:.
Boutin
,
J.
,
N.
Martin
,
G.
Reverdin
,
X.
Yin
, and
F.
Gaillard
,
2013
:
Sea surface freshening inferred from SMOS and ARGO salinity: Impact of rain
.
Ocean Sci.
,
9
,
183
192
, doi:.
Chowdary
,
J. S.
,
C.
Gnanaseelan
, and
S.-P.
Xie
,
2009
:
Westward propagation of barrier layer formation in the 2006–07 Rossby wave event over the tropical southwest Indian Ocean
.
Geophys. Res. Lett.
,
36
, L04607, doi:.
Corbella
,
I.
,
F.
Torres
,
N.
Duffo
,
V.
González-Gambau
,
M.
Pablos
,
I.
Duran
, and
M.
Martín-Neira
,
2011
:
MIRAS calibration and performance: Results from the SMOS in-orbit commissioning phase
.
IEEE Trans. Geosci. Remote Sens.
,
49
,
3147
3155
, doi:.
Delcroix
,
T.
, and
M. J.
McPhaden
,
2002
:
Interannual sea surface salinity and temperature changes in the western Pacific warm pool during 1992–2000
.
J. Geophys. Res.
,
107
, 8002, doi:.
Du
,
Y.
, and
S.-P.
Xie
,
2008
:
Role of atmospheric adjustments in the tropical Indian Ocean warming during the 20th century in climate models
.
Geophys. Res. Lett.
,
35
, L08712, doi:.
Du
,
Y.
,
T.
Qu
,
G.
Meyers
,
Y.
Masumoto
, and
H.
Sasaki
,
2005
:
Seasonal heat budget in the mixed layer of the southeastern tropical Indian Ocean in a high-resolution ocean general circulation model
.
J. Geophys. Res.
,
110
, C04012, doi:.
Du
,
Y.
,
S.-P.
Xie
,
G.
Huang
, and
K.
Hu
,
2009
:
Role of air–sea interaction in the long persistence of El Niño–induced north Indian Ocean warming
.
J. Climate
,
22
,
2023
2038
, doi:.
Durand
,
F.
,
G.
Alory
,
R.
Dussin
, and
N.
Reul
,
2013
:
SMOS reveals the signature of Indian Ocean dipole events
.
Ocean Dyn.
,
63
,
1203
1212
, doi:.
Gordon
,
A. L.
,
S.
Ma
,
D. B.
Olson
,
P.
Hacker
,
A.
Ffield
,
L. D.
Talley
,
D.
Wilson
, and
M.
Baringer
,
1997
:
Advection and diffusion of Indonesian Throughflow water within the Indian Ocean South Equatorial Current
.
Geophys. Res. Lett.
,
24
,
2573
2576
, doi:.
Grodsky
,
S. A.
, and Coauthors
,
2012
:
Haline hurricane wake in the Amazon/Orinoco plume: Aquarius/SAC-D and SMOS observations
.
Geophys. Res. Lett.
,
39
, L20603, doi:.
Grunseich
,
G.
,
B.
Subrahmanyam
,
V. S. N.
Murty
, and
B. S.
Giese
,
2011
:
Sea surface salinity variability during the Indian Ocean dipole and ENSO events in the tropical Indian Ocean
.
J. Geophys. Res.
,
116
, C11013, doi:.
Han
,
W.
, and
J. P.
McCreary
,
2001
:
Modeling salinity distribution in the Indian Ocean
.
J. Geophys. Res.
,
106
,
859
877
, doi:.
Han
,
W.
,
J. P.
McCreary
, and
K. E.
Kohler
,
2001
:
Influence of precipitation minus evaporation and Bay of Bengal rivers on dynamics, thermodynamics, and mixed-layer physics in the upper Indian Ocean
.
J. Geophys. Res.
,
106
,
6895
6916
, doi:.
Henocq
,
C.
,
J.
Boutin
,
G.
Reverdin
,
F.
Petitcolin
,
S.
Arnault
, and
P.
Lattes
,
2010
:
Vertical variability of near-surface salinity in the tropics: Consequences for L-band radiometer calibration and validation
.
J. Atmos. Oceanic Technol.
,
27
,
192
209
, doi:.
Huffman
,
G. J.
,
R. F.
Adler
,
D. T.
Bolvin
, and
G.
Gu
,
2009
:
Improving the global precipitation record: GPCP version 2.1
.
Geophys. Res. Lett.
,
36
, L17808, doi:.
Illig
,
S.
, and
C.
Perigaud
,
2007
:
Yearly impact of submonthly rain fluctuations on the Indian Ocean salinity
.
Geophys. Res. Lett.
,
34
, L12609, doi:.
Jensen
,
T. G.
,
2003
:
Cross-equatorial pathways of salt and tracers from the northern Indian Ocean: Modelling results
.
Deep-Sea Res. II
,
50
,
2111
2127
, doi:.
Kainulainen
,
J.
, and Coauthors
,
2012
:
Radiometric performance of the SMOS reference radiometers—Assessment after one year of operation
.
IEEE Trans. Geosci. Remote Sens.
,
50
,
1367
1387
, doi:.
Kara
,
A. B.
,
P. A.
Rochford
, and
H. E.
Hurlburt
,
2000
:
An optimal definition for ocean mixed layer depth
.
J. Geophys. Res.
,
105
,
16 803
16 821
, doi:.
Kerr
,
Y.
,
P.
Waldteufel
,
J. P.
Wigneron
,
J. M.
Martinuzzi
,
J.
Font
, and
M.
Berger
,
2001
:
Soil moisture retrieval from space: The Soil Moisture and Ocean Salinity (SMOS) mission
.
IEEE Trans. Geosci. Remote Sens.
,
39
,
1729
1735
, doi:.
Lagerloef
,
G.
,
2002
:
Introduction to the special section: The role of surface salinity on upper ocean dynamics, air–sea interaction and climate
.
J. Geophys. Res.
,
107
, 8000, doi:.
Lagerloef
,
G.
, and Coauthors
,
2008
:
The Aquarius/SAC-D Mission: Designed to meet the salinity remote-sensing challenge
.
Oceanography
,
21
,
68
81
, doi:.
Lagerloef
,
G.
,
Y.
Kerr
,
F.
Cabot
,
T.
Jackson
,
R.
Bindlish
,
H.-Y.
Kao
,
N.
Reul
, and
J.
Font
,
2013
: SMOS and Aquarius inter-comparison over oceans. Geophysical Research Abstracts, Vol. 15, Abstract EGU2013-6456. [Available online at http://meetingorganizer.copernicus.org/EGU2013/EGU2013-6456.pdf.]
Locarnini
,
R. A.
,
A. V.
Mishonov
,
J. I.
Antonov
,
T. P.
Boyer
,
H. E.
Garcia
,
O. K.
Baranova
,
M. M.
Zweng
, and
D. R.
Johnson
,
2010
: Temperature. Vol. 1, World Ocean Atlas 2009, NOAA Atlas NESDIS 68, 184 pp.
Lukas
,
R.
, and
E.
Lindstrom
,
1991
:
The mixed layer of the western equatorial Pacific Ocean
.
J. Geophys. Res.
,
96
,
3343
3357
, doi:.
Madden
,
R. A.
, and
P. R.
Julian
,
1972
:
Description of global-scale circulation cells in tropics with a 40–50 day period
.
J. Atmos. Sci.
,
29
,
1109
1123
, doi:.
Masson
,
S.
,
P.
Delecluse
,
J.-P.
Boulanger
, and
C.
Menkes
,
2002
:
A model study of the seasonal variability and formation mechanisms of the barrier layer in the eastern equatorial Indian Ocean
.
J. Geophys. Res.
,
107
, 8017, doi:.
Masson
,
S.
,
C.
Menkes
,
P.
Delecluse
, and
J.-P.
Boulanger
,
2003
:
Impacts of salinity on the eastern Indian Ocean during the termination of the fall Wyrtki jet
.
J. Geophys. Res.
,
108
, 3067, doi:.
Masson
,
S.
,
J.-P.
Boulanger
,
C.
Menkes
,
P.
Delecluse
, and
T.
Yamagata
,
2004
:
Impact of salinity on the 1997 Indian Ocean dipole event in a numerical experiment
.
J. Geophys. Res.
,
109
, C02002, doi:.
Meister
,
G.
,
B.
Franz
,
K.
Turpie
, and
C.
McClain
,
2005
: The MODIS Aqua point-spread function for ocean color bands. Ninth Int. Symp. on Physical Measurements and Signatures in Remote Sensing, Vol. 36, Beijing, China, ISPR WG-VII/1, 334–336.
Perigaud
,
C.
,
J. P.
McCreary
Jr.
, and
K. Q.
Zhang
,
2003
:
Impact of interannual rainfall anomalies on Indian Ocean salinity and temperature variability
.
J. Geophys. Res.
,
108
, 3319, doi:.
Phillips
,
H. E.
,
S. E.
Wijffels
, and
M.
Feng
,
2005
:
Interannual variability in the freshwater content of the Indonesian-Australian Basin
.
Geophys. Res. Lett.
,
32
, L03603, doi:.
Qiu
,
Y.
,
W.
Cai
,
L.
Li
, and
X.
Guo
,
2012
:
Argo profiles variability of barrier layer in the tropical Indian Ocean and its relationship with the Indian Ocean dipole
.
Geophys. Res. Lett.
,
39
, L08605, doi:.
Qu
,
T.
, and
G.
Meyers
,
2005
:
Seasonal characteristics of circulation in the southeastern tropical Indian Ocean
.
J. Phys. Oceanogr.
,
35
,
255
267
, doi:.
Qu
,
T.
,
T.
Song
, and
C.
Maes
,
2014
:
Sea surface salinity and barrier layer variability in the equatorial Pacific as seen from Aquarius and Argo
.
J. Geophys. Res.
,
119
,
15
29
, doi:.
Rao
,
S. A.
, and
S. K.
Behera
,
2005
:
Subsurface influence on SST in the tropical Indian Ocean: Structure and interannual variability
.
Dyn. Atmos. Oceans
,
39
,
103
135
, doi:.
Reul
,
N.
, and Coauthors
,
2011
: SMOS level 3 SSS research products: Product validation document reprocessed year 2010. Tech. Doc. CATDS-CECOS-L3-VALDOC, IFREMER, 141 pp.
Reul
,
N.
,
J.
Tenerelli
,
J.
Boutin
,
B.
Chapron
,
F.
Paul
,
E.
Brion
,
F.
Gaillard
, and
O.
Archer
,
2012
:
Overview of the first SMOS sea surface salinity products. Part I: Quality assessment for the second half of 2010
.
IEEE Trans. Geosci. Remote Sens.
,
50
,
1636
1647
, doi:.
Reverdin
,
G.
,
S.
Morisset
,
J.
Boutin
, and
N.
Martin
,
2012
:
Rain-induced variability of near sea-surface T and S from drifter data
.
J. Geophys. Res.
,
117
, C02032, doi:.
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
,
1609
1625
, doi:.
Riser
,
S.
,
L.
Ren
, and
A.
Wong
,
2008
:
Salinity in Argo: A modern view of a changing ocean
.
Oceanography
,
21
,
56
67
, doi:.
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
, doi:.
Saji
,
N. H.
,
B. N.
Goswami
,
P. N.
Vinayachandran
, and
T.
Yamagata
,
1999
:
A dipole in the tropical Indian Ocean
.
Nature
,
401
,
360
363
.
Schott
,
F.
,
M.
Dengler
, and
R.
Schoenefeldt
,
2002
:
The shallow overturning circulation of the Indian Ocean
.
Prog. Oceanogr.
,
53
,
57
103
, doi:.
Song
,
Q.
, and
A. L.
Gordon
,
2004
:
Significance of the vertical profile of the Indonesian Throughflow transport to the Indian Ocean
.
Geophys. Res. Lett.
,
31
, L16307, doi:.
Subrahmanyam
,
B.
,
V. S. N.
Murty
, and
D. M.
Heffner
,
2011
:
Sea surface salinity variability in the tropical Indian Ocean
.
Remote Sens. Environ.
,
115
,
944
956
, doi:.
Thompson
,
B.
,
C.
Gnanaseelan
, and
P. S.
Salvekar
,
2006
:
Variability in the Indian Ocean circulation and salinity and its impact on SST anomalies during dipole events
.
J. Mar. Res.
,
64
,
853
880
, doi:.
Vinayachandran
,
P. N.
, and
R. S.
Nanjundiah
,
2009
:
Indian Ocean sea surface salinity variations in a coupled model
.
Climate Dyn.
,
33
,
245
263
, doi:.
Webster
,
P. J.
,
A. M.
Moore
,
J. P.
Loschnigg
, and
R. R.
Leben
,
1999
:
Coupled oceanic–atmospheric dynamics in the Indian Ocean during 1997–98
.
Nature
,
401
,
356
360
, doi:.
Wyrtki
,
K.
,
1973
:
An equatorial jet in the Indian Ocean
.
Science
,
181
,
262
264
, doi:.
Xie
,
S.-P.
,
H.
Annamalai
,
F. A.
Schott
, and
J. P.
McCreary
Jr.
,
2002
:
Structure and mechanism of south Indian Ocean climate variability
.
J. Climate
,
15
,
864
878
, doi:.
Yoshida
,
K.
,
1959
:
A theory of the Cromwell current (the equatorial undercurrent) and of the equatorial upwelling—An interpretation in a similarity to a coastal circulation
.
J. Oceanogr. Soc. Japan
,
15
,
159
170
.
Yu
,
W.
,
B.
Xiang
,
L.
Liu
, and
N.
Liu
,
2005
:
Understanding the origins of interannual thermocline variations in the tropical Indian Ocean
.
Geophys. Res. Lett.
,
32
, L24706, doi:.
Zhang
,
C.
,
J.
Gottschalck
,
E. D.
Maloney
,
M. W.
Moncrieff
,
F.
Vitart
,
D. E.
Waliser
,
B.
Wang
, and
M. C.
Wheeler
,
2013
:
Cracking the MJO nut
.
Geophys. Res. Lett.
,
40
,
1223
1230
, doi:.
Zhang
,
Y.
, and
Y.
Du
,
2012
:
Seasonal variability of salinity budget and water exchange in the northern Indian Ocean from HYCOM assimilation
.
Chin. J. Oceanol. Limnol.
,
30
,
97
107
, doi:.
Zhang
,
Y.
,
Y.
Du
,
S.
Zheng
,
Y.
Yang
, and
X.
Cheng
,
2013
:
Impact of Indian Ocean Dipole on the salinity budget in the equatorial Indian Ocean
.
J. Geophys. Res. Oceans
,
118
,
4911
4923
, doi:.