1. Introduction
The oceanic mixed layer, typically a layer with vertically uniform temperature, salinity, and density near the sea surface, is the media between the atmosphere and the ocean interior. The mixed layer depth (MLD), remarking the base of the surface mixed layer, is critical for the heat budget of the surface ocean and modulates sea surface temperature (SST) variability and atmospheric convection (e.g., Shinoda and Hendon 1998; Alexander et al. 2000; Dommenget and Latif 2002; Schott et al. 2009). For example, the tropical cyclone intensity depends critically upon the initial MLD in its formation region (e.g., Chang and Anthes 1978; Bender et al. 1993; Chan et al. 2001; Zhao and Chan 2017). The MLD also represents the volume of seawater over which surface heat, freshwater, and momentum inputs would be distributed (e.g., Chen et al. 1994), with profound impacts on ocean biology (e.g., Fasham 1995; Polovina et al. 1995; Behrenfeld 2010) and near-surface acoustic propagation (Sutton et al. 1993). The change of MLD can cause downward and upward mass fluxes across the mixed layer base, dubbed as detrainment and entrainment, respectively (e.g., De Szoeke 1980; Cushman-Roisin 1987). Detrainment/entrainment, carrying momentum, energy, and carbon dioxide fluxes, drives interior ocean motions and conveys climate change signals between the atmosphere and subsurface ocean. Therein, the mass flux between the mixed layer and the permanent pycnocline is described as subduction/obduction (Huang 1988; Qiu and Huang 1995), corresponding to the formation/erosion of ventilation-related water masses and serving as the vertical limbs of global ocean circulation.
The theory of subduction and ventilation was initially based on the water mass formation framework postulated by Iselin (1939) and Stommel (1979), which were followed by a number of ideal-fluid pycnocline models for the diagnosis of subduction/obduction rate (e.g., Woods 1985; Huang 1990; Williams 1991; Qiu and Huang 1995; Marshall 1997). A widely used diagnosis framework was proposed by Qiu and Huang (1995), in which only the climatological annual cycle of MLD is considered, and subduction and obduction are assumed to occur only in late winter and early winter, respectively. This framework has been widely utilized to estimate the subduction/obduction rate and its variability (e.g., Karstensen and Quadfasel 2002a,b; Qu and Chen 2009; Trossman et al. 2009; Liu and Huang 2012). Recent studies have worked on the refinement of the subduction/obduction diagnosing method by taking into account more realistic MLD variability. It has been demonstrated that the semiannual variability of MLD in regions controlled by monsoons can give rise to an additional subduction/obduction window in the summer monsoon season and weaken the subduction/obduction in the winter monsoon season (Liu et al. 2018, 2020).
A remarkable advance in the understanding of subduction/obduction is brought by the finding that the high-frequency MLD variability associated with mesoscale eddies can greatly enhance subtropical subduction, particularly for the mode waters, as revealed by in situ observations (e.g., Uehara et al. 2003; Kouketsu et al. 2011; Xu et al. 2014, 2016) and numerical simulations (Qu et al. 2002; Rainville et al. 2007; Nishikawa et al. 2010). By contrast, there were few studies to appreciate the impact of high-frequency MLD variability on the obduction phase, which mainly takes place in the subpolar and tropical regions. The tropical oceans are subjected to strong modulation effects of atmospheric intraseasonal oscillations. Considering that the Madden–Julian oscillation (MJO; Madden and Julian 1971) is the dominant mode of atmospheric intraseasonal oscillations, we use the “MJO” as a collective representation for convenience in the study. Under strong forcing effects of MJOs, the tropical MLD exhibits prominent intraseasonal variations (e.g., Drushka et al. 2012, 2014; Keerthi et al. 2016; Li et al. 2017) and possibly affects the tropical obduction. It is still unknown whether the MJO-induced MLD variability can significantly enhance the obduction, in analogy to the effect of mesoscale eddies on subtropical subduction. Exploring this problem provides an opportunity to further refine the theoretical understanding and diagnosing methods of the thermocline water obduction in tropics. These efforts may also provide implications for understanding climate change, given that the obduction brings anomalies of temperature, salinity, carbon dioxide, and nutrients to the sea surface and potentially modulates air–sea interactions.
The equatorial Indian Ocean (EIO) is home to most strong MJO events (Zhang 2005) and exhibits strong intraseasonal variability (ISV) in SST (e.g., Duvel et al. 2004; Han et al. 2007; Duncan and Han 2009; Jayakumar et al. 2011), ocean salinity (e.g., Matthews et al. 2010; Grunseich et al. 2011; Li et al. 2015), and upper-ocean circulation (e.g., Schiller and Godfrey 2003; Masumoto et al. 2005; Chen et al. 2015, 2019). Intraseasonal SST variability is actively coupled to the MJO’s atmospheric variabilities (Hendon and Glick 1997; Wang and Xie 1998; Woolnough et al. 2001). The EIO is, therefore, an ideal region to examine the MJO’s possible effect on MLD and subsurface water obduction. Estimations based on in situ observations suggested that in this region, the MLD change during strong MJO events achieves a peak-to-peak difference of more than 15 m in boreal winter (Drushka et al. 2012, 2014). Keerthi et al. (2016) explored large-scale intraseasonal MLD variability in the Indian Ocean and found that the summertime MLD in the EIO primarily responds to the intraseasonal active/break spells of the summer monsoon, while the MJO is the primary cause for wintertime intraseasonal MLD anomalies. Using ocean model simulations, Li et al. (2017) showed that intraseasonal MLD variability achieves amplitudes of ~10 m in the EIO and the Bay of Bengal in summer. In spite of existing knowledge, a comprehensive investigation is required to clarify how MJO events cause MLD variability in the EIO, such as the relative importance of surface wind forcing and buoyancy forcing exerted by MJOs.
Given the prominent intraseasonal MLD variability, it is expected that strong entrainment events can occur during the active phase of the MJO when strong winds and radiation-induced cooling drive dramatic thickening of the mixed layer (Drushka et al. 2012, 2014; Li et al. 2017). These entrainment events may greatly enhance the equatorial obduction of the cold, nutrient-rich thermocline water, which could be fundamental for understanding SST variability, air–sea interactions, and marine biological processes (e.g., Chen et al. 2016). Results of ocean model simulations presented in this study, however, suggest quite the opposite. Taking into consideration the MJO forcing effects actually reduces the annual obduction rate. It is an interesting and debatable question as to how and why MJOs suppress the equatorial obduction of thermocline water.
The overall purpose of the present study is to reveal the key processes through which MJOs cause MLD variability in the EIO and explore the possible impacts of MJO-induced high-frequency MLD variability on the equatorial obduction. This is pursued mainly by performing and analyzing ocean general circulation model (OGCM) experiments. The rest of the paper is organized as follows. Section 2 describes the data and model utilized in this study, especially the OGCM experiments designed for our analysis. Section 3 presents the characteristics and mechanisms of intraseasonal MLD variability. Section 4 provides evaluations for the MJO forcing effects on processes of entrainment and obduction, and a conceptual schematic is postulated to demonstrate how the MJO-induced MLD variability affects obduction. Section 5 provides a summary of the main findings of the paper.
2. Data and models
a. Data
In situ temperature measurements during 2012–15 by two moored buoys of the Research Moored Array for African–Asian–Australian Monsoon Analysis and Prediction (RAMA; McPhaden et al. 2009) at 0°, 80.5°E and 1.5°S, 80.5°E are used to estimate the ISV of MLD and validate the OGCM results. Daily temperature data collected typically at 1-, 10-, 13-, 20-, 40-, 60-, 80-, and 100-m depths are linearly interpolated into 1-m vertical intervals. Because of insufficient sampling coverage of salinity data, MLD is computed as the depth at which the temperature decreases 0.5°C from the surface value based on RAMA temperature data (Monterey and Levitus 1997).
For MOAA GPV, SODA, and OGCM outputs that contain salinity records, MLD is defined as the depth at which potential density is larger than the sea surface value by 0.125 kg m−3 (Levitus 1982). Unless otherwise noted, the intraseasonal anomaly of an oceanic or atmospheric variable is obtained through removing the monthly climatology and applying a 30–105-day bandpass Lanczos digital filter. The results are generally insensitive to choices of the upper and lower bounds of the bandpass window.
b. HYCOM and experiments
The ocean model used in this study is the Hybrid Coordinate Ocean Model (HYCOM; e.g., Wallcraft et al. 2009), version 2.2.18, which has been successfully utilized to investigate MJO-related oceanic variations (e.g., Li et al. 2014, 2015, 2017; Chen et al. 2015, 2019). For this study, HYCOM is configured to the Indian Ocean within 30°–122.5°E, 50°S–30°N, with horizontal resolutions of 0.25° × 0.25° and 26 hybrid layers in the vertical direction. Thickness of the top layer is ~2.9 m near the sea surface. Three 5° sponge layers are applied at the western, eastern, and southern open-ocean boundaries, where the model temperature and salinity are relaxed to monthly climatology. The diffusion/mixing parameters are identical to those adopted in Li et al. (2013).
Surface atmospheric forcing fields include surface winds from CCMP for 2000–11 and ASCAT for 2012–14, 1° × 1° surface shortwave radiation (SWR) and longwave radiation (LWR) from Clouds and the Earth’s Radiant Energy System (CERES; Wielicki et al. 1996; Loeb et al. 2001), 0.25° × 0.25° precipitation from the Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA) level 3B42 product (Kummerow et al. 1998; Huffman et al. 2007), and 0.75° × 0.75° 2-m air temperature and humidity from ERA-Interim (Dee et al. 2011). Note that wind speed and wind stress are fed separately to the model. Wind stress affects the model ocean through driving advection, upwelling/downwelling, and turbulent mixing, while wind speed affects the ocean mainly through evaporation and turbulent heat flux. HYCOM computes evaporation and turbulent heat flux with wind speed, air temperature, specific humidity, and surface temperature using the bulk formula of the Coupled Ocean–Atmosphere Response Experiment (COARE 3.0) algorithm (Kara et al. 2005). For river discharges, we use the satellite-based monthly discharge records of the Ganga–Brahmaputra from Papa et al. (2010) and monthly discharge data of Dai et al. (2009) for other rivers.
The model is spun up from a state of rest for 30 years using monthly climatological forcing, and then it is integrated forward from March 2000 to December 2014. Six experiments are designed to isolate the effects of different processes (Table 1; see also Li et al. 2015). The main run (MR) is forced by original daily forcing fields, and its solution is assumed to contain complete processes and used for validation against observational data. To exclude the MJO-related forcings, we performed the NoMJO experiment in which all atmospheric forcing fields are low-passed with a 105-day Lanczos digital filter. Thus, ISV in NoMJO originates from the oceanic internal processes, such as mesoscale eddies and tropical instability waves. The difference (MR − NoMJO) measures the overall effects of MJOs on the ocean. In the third experiment, NoSWR, only SWR is 105-day low-passed, and therefore its difference from MR, MR − NoSWR, measures the effect of intraseasonal SWR anomalies on the ocean. In NoWND, both wind speed and wind stress are low-pass filtered, whereas in NoSTRS only wind stress is low-pass filtered. The difference between MR and NoSTRS (MR − NoSTRS) measures the effect of wind stress, and NoSTRS − NoWND reflects the effect of wind speed. In the NoPRCP experiment, we use only 105-day low-passed filtered precipitation, and MR − NoPRCP isolates the precipitation effect. Outputs of the six experiments are stored as 3-day mean data from March 2000 to December 2014. The outputs in 2000 and 2001 are discarded to exclude the transient effects from the spinup run, and the 13-yr data of 2002–14 are used for our analysis.
Summary of HYCOM experiments.
3. Intraseasonal variability of mixed layer depth
Measurements by RAMA buoys at B1 (0°, 80.5°E) and B2 (1.5°S, 80.5°E) show large-amplitude variability of MLD, with evident intraseasonal fluctuations superimposed on the seasonal cycle (Fig. 1). The MLD generally reaches the deepest position in summer, as the response to the mature stage of the Indian summer monsoon. The seasonal difference of MLD achieves 40–60 m (Figs. 1a,b). There are noticeable intraseasonal deepening events, especially during the winter monsoon season. In the strong deepening event during November 2014–January 2015 at B1, the MLD can reach ~100 m, a depth equivalent to the summertime MLD (Fig. 1a). As quantified by the 30–105-day bandpass filtered MLD anomalies, ISV of MLD can reach as large as 15 m in amplitude, with a standard deviation (STD) of 6.7 m at B1 (Fig. 1c). At B2, intraseasonal fluctuations of ~20 m are found, and the STD is 7.3 m (Figs. 1b,d). It is noted that the intensity of ISV of MLD may be different in different periods. Taking B2 for example, the STD of intraseasonal MLD anomalies between March 2009 and March 2011 achieves 10.4 m (figure not shown). These results derived from RAMA buoys confirm the existence of prominent intraseasonal MLD variability in the EIO region.
Daily temperature (color shading; °C) and MLD (black lines; m) measured by RAMA buoys (a) B1 (0°, 80.5°E) and (b) B2 (1.5°S, 80.5°E). Also shown is 30–105-day bandpass filtered MLD anomalies at (c) B1 and (d) B2. The positive MLD anomaly indicates a deeper MLD than the average.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
Because of the nonlinear nature of ocean stratification, the background mean state and seasonal variation of MLD can affect ISV. Thus, before analyzing the ISV, we first examine the mean state and seasonal variation of MLD in the EIO. As shown in Fig. 2, the spatial structures of climatological MLD in the MR are generally consistent with MOAA GPV, such as the deep MLDs south of 13°N and in the Arabian Sea and the shallow MLDs in the southwest Indian Ocean. There are overly deep MLDs in the Bay of Bengal in HYCOM (Figs. 2a,b). This is likely a common bias in the ocean model simulations (e.g., Duncan and Han 2009; Li et al. 2013, 2014, 2015; Felton et al. 2014) related to the complicated thermal and saline stratification near the sea surface, including the shallow MLD and the thick barrier layer (e.g., Thadathil et al. 2007; Li et al. 2017). Nevertheless, in the EIO on which our analysis focuses, the simulated MLD is close to observation. The monthly anomalies of MLD, relative to the climatological mean MLD, averaged over 3°S–3°N are used to check the seasonal evolution in the EIO (Figs. 2c,d). The MLD is deeper in summer than in winter, in agreement with RAMA buoy measurements (Fig. 1a). The seasonal variation of MLD reaches ~40 m, close to RAMA results, and the largest MLD occurs in June–August. These seasonal features are simulated well by HYCOM.
Annual-mean climatological MLD (m) during 2002–14 from (a) MOAA GPV data and (b) HYCOM MR, the monthly MLD anomaly (m) averaged over 3°S–3°N from (c) MOAAGPV data and (d) HYCOM MR, and the STD of 30–105-day MLD from (e) SODA and (f) HYCOM MR. The black-outlined rectangle defines the equatorial Indian Ocean region of 70°–95°E, 3°S–3°N.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
The STD of 30–105-day bandpass filtered MLD measures the overall intensity of the intraseasonal MLD variability. The prominent ISV is concentrated in the central-to-eastern portion of the Indian Ocean equator, with the largest STD of ~14 m (Fig. 2f). It is consistent with 10–15-m STD of intraseasonal MLD by Keerthi et al. (2016). Moreover, the STD of intraseasonal MLD in HYCOM MR agrees well with SODA data in both spatial distribution and typical intensity in the EIO box (Figs. 2e,f). The following analysis will focus on this EIO box (70°–95°E, 3°S–3°N). It is noted that we also calculate the intraseasonal MLD with another definition, as the depth with potential density increase from the sea surface value equivalent to a temperature decrease of 0.5°C (de Boyer Montegut et al. 2004). Over the area of interest marked by the black box in Fig. 2, the STD maps based on two MLD definitions agree well with each other in both spatial distribution and typical intensity (figure not shown). Thus, the ISV of MLD in the EIO is insensitive to the choice of methods diagnosing MLD. In general, HYCOM MR can well represent the large-scale distribution and seasonal/intraseasonal variation of MLD in the EIO.
The change of MLD is a complex process, determined by a mutable balance between the destabilizing effect of surface wind stress through turbulent mixing and the stabilizing effect by density stratification that are regulated by surface buoyancy fluxes (heat and freshwater) and ocean dynamics (advection and upwelling/downwelling). It is instructive to clarify the major causes for the intraseasonal MLD variability. Figure 3 shows the intraseasonal MLD anomalies induced by different atmospheric forcing effects and oceanic internal processes in the EIO. The STD of 30–105-day MLD averaged over the EIO region is 6.16 m in the MR (Fig. 3a), roughly consistent in magnitude with RAMA buoy measurements (Fig. 1). The total forcing effect by MJOs, as represented by MR − NoMJO, is a STD of 6.22 m and a correlation of 0.96 with that of MR. The small difference in STD (6.16 vs 6.18 m) is statistically insignificant as suggested by an F test and likely arises from ocean internal instability. On the other hand, the 30–105-day MLD from NoMJO, as the ocean internal processes, has a small STD (1.63 m) and an insignificant correlation with the MR (0.10). The sharp contrasts between MR and NoMJO suggest the dominant role played by the MJO in causing intraseasonal MLD variability and a minor role played by the oceanic internal processes.
The 30–105-day MLD (m) averaged in the EIO region (70°–95°E, 3°S–3°N) from the MR (black), (a) NoMJO (cyan) and MJO effect (MR − NoMJO; pink), (b) wind stress effect (MR − NoSTRS; orange), (c) wind speed effect (NoSTRS − NoWND; green), (d) SWR effect (MR − NoSWR; red), and (e) precipitation effect (MR − NoPRCP; blue). STD and linear correlations are listed in each plot.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
To better understand these forcing effects, we compute the MJO composite based on the real-time multivariate (RMM) MJO index (Wheeler and Hendon 2004) for all useful variables (Fig. 4). The index describes the MJO evolution as an eight-phase cycle based on the location of large-scale outgoing longwave radiation and associated changes in vertical wind shear. For a given MJO phase, the composite properties are formed by averaging the variables over the days with RMM index in that phase and RMM magnitude > 1.5 (indicating significant atmospheric perturbations). In this way, the composite results represent the mean evolution of atmospheric/oceanic conditions during the passage of MJOs, and this method was widely used in previous studies (e.g., Guan et al. 2014; Wang et al. 2016; Zhu et al. 2020). The surface wind stress, which is critical for intraseasonal MLD variability, achieves the maximal magnitude in phase 3 of the composite MJO (Fig. 4a). Phase 3 generally characterizes a stage when the MJO has its fully developed convection center, involving windy and cloudy conditions, lingering over the EIO (Wheeler and Hendon 2004). The forcing effect of surface wind stress on MLD is twofold. The increased wind stress enhances turbulent mixing near the ocean surface through kinetic energy input, while the westerly wind anomalies (Fig. 4b) evoke downwelling equatorial Kelvin waves (e.g., Kessler et al. 1995; Hendon et al. 1998; McPhaden and Yu 1999) which cause upper-ocean convergence (e.g., Somavilla et al. 2017), as indicated by eastward-propagating positive sea level anomalies (SLAs; Fig. 4c). Chen et al. (2015) have pointed out that MJO-driven Kelvin waves can affect the intraseasonal MLD variability in the equatorial Indian Ocean by the adiabatic processes. Likely, the enhanced turbulent mixing and the convergence of surface warm water work mutually to cause the deep MLD in phases 3–5 in MR and MR − NoSTRS (Figs. 4d,e). The westerly wind anomalies exhibit swift eastward propagation in the central and eastern basin. Correspondingly, SLAs and MLD anomalies also show eastward propagation tendencies but with slower transition speed, possibly reflecting the slower phase speed of oceanic Kelvin waves (~2.6 m s−1 for the heaviest baroclinic mode) than that of MJOs (~5.0 m s−1). It is noticeable that the SLA amplitude is much stronger near the eastern boundary than in the EIO interior, while the MLD anomalies in the two areas are comparable. This indicates that turbulent mixing by the increased wind stress does contribute to the MLD deepening, in addition to the downwelling effect of westerly winds. Here we are unable to determine the relative importance of the two processes. Moreover, it is unknown why there is seemingly westward propagation of wind stress anomaly west of 60°E (Fig. 4a). However, it has little impact on our results.
Time–longitude plots of composite MJO cycle based on RMM index: (a) wind stress magnitude (N m−2), (b) zonal wind stress (N m−2), (c) SLA (cm), (d) MR MLD (m), (e) wind stress effect on MLD (MR − NoSTRS), (f) wind speed effect on MLD (NoSTRS − NoWND), (g) SWR effect on MLD (MR − NoSWR), and (h) precipitation effect on MLD (MR − NoPRCP). All variables are 30–105-day bandpass filtered and averaged over 3°S–3°N before performing the composite.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
The close resemblance in anomaly features between Figs. 4d and 4e confirms the importance of wind stress in causing MLD changes. By contrast, anomalies induced by wind speed, SWR, and precipitation are smaller in magnitude and show apparent differences in the timing. MLD anomalies produced by wind speed and SWR tend to lead those of MR by 1–2 phases (Figs. 4f,g), while the precipitation effect is roughly out of phase with MR (Fig. 4h). The negative contribution by precipitation to MLD change reflects the fact that the enhanced rainfall during the active phase of MJO convection acts to shoal the MLD through surface buoyancy input, and this effect is opposite to the wind-driven MLD deepening in MR and MR − NoSTRS.
Next, we explore how the effects of different processes are superimposed to yield the total MLD variability in MR. Because the precipitation contribution is weak and negative, here we only examine the wind stress, wind speed, and SWR effects. As shown in Fig. 5a, the evolutions of MLD are highly consistent between MR (black) and the wind stress effect (orange). They are deep during phases 3–5, reaching the peak in phase 4, and shallow in phase 8. SLA, representing the anomalous upper-ocean convergence/divergence, is primarily controlled by wind stress (Fig. 5b) through ocean dynamical processes such as equatorial Kelvin waves. The wind stress-driven SLAs also show a peak in phase 4 and a trough in phase 8, largely mimicking the change of MLD. The resemblance between MLD and SLA confirms that the upper-ocean convergence/divergence is a major process through which the MJO wind stress causes the MLD change, in addition to wind stress-driven turbulent mixing. In comparison, the MLD deepening anomalies caused by the wind speed effect (green) and SWR effect (red) are only ~1 m and much weaker than the ~4 m deepening caused by the wind stress effect (orange). They tend to occur in phases 2–3 (Fig. 5a), corresponding to the surface heat loss peaks in phase 2–3 (Fig. 5c). This explains why their effects on MLD lead the total MLD anomaly by 1–2 phases. These buoyancy forcing effects exerted by SWR and wind speed-controlled turbulent heat flux are weaker than wind stress forcing and mainly contribute to the initial deepening of MLD in phase 3.
MJO composites of (a) MLD (m), (b) SLA (cm), and (c) surface net heat flux (W m−2) derived from MR (black), wind stress effect (MR − NoSTRS; orange), wind speed effect (NoSTRS − NoWND; green), and SWR effect (MR − NoSWR; red). All variables are 30–105-day bandpass filtered and averaged in the EIO box before permitting the composite.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
4. Impacts on obduction
The MJO-induced strong intraseasonal MLD deepening events are expected to cause large entrainment events. In the EIO, the intraseasonal fluctuations of entrainment from the MR can reach as large as 3 × 10−5 m s−1, with the STD of 8.82 × 10−6 m s−1 (Fig. 6a). Similar to the case of MLD, the ISV of entrainment rate is predominantly caused by the MJO. The difference, MR − NoMJO, shows a slightly higher STD (8.93 × 10−6 m s−1) and a high correlation with the MR (0.98). The sum of entrainment events in one year, dubbed as annual entrainment rate, is used to quantify both the frequency and magnitude of the water mass exchange at the mixed layer base. Because of the large entrainment events by MJOs, the annual entrainment rate averaged in the EIO box increases from 435 m yr−1 in NoMJO to 582 m yr−1 in MR. In other words, taking into consideration MJO-related forcings can intensify the upward mass flux at the mixed layer base by 34%. Again, the wind stress effect is the leading contributor to ISV of E, showing a STD of 7.91 × 10−6 m s−1 and a correlation of 0.97 with MR (Fig. 6b). Its overall contribution to the total ISV of E is 86.8%. The SWR effect is of secondary importance, with a STD of 3.17 × 10−6 m s−1 and a correlation of 0.70, and its contribution to MR is 25.4% (Fig. 6c). In comparison, the effects of wind speed and precipitation are relatively weak, and the precipitation effect is negative (Figs. 6d,e).
As in Fig. 3, but for intraseasonal entrainment rate anomaly (10−6 m s−1) of the EIO region.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
Wind stress can affect the entrainment rate through three different physical processes, as represented by the three RHS terms in Eq. (4). First, wind stress causes MLD changes through turbulent mixing and ocean dynamical processes, as discussed in section 3. The MLD change, as quantified in Fig. 7a by the MLD tendency term ∂hm/∂t (orange), is of the largest STD (8.73 × 10−6 m s−1), and its overall contribution to E (black) reaches 92.8%. The positive ∂hm/∂t caused by wind forcing in phases 1–3 (Fig. 7b) gives rise to the maximum MLD in phase 4 (Fig. 5a). Second, the vertical pumping term wmb induced by wind stress can directly influence the entrainment by carrying water mass across the mixed layer base. The amplitude of intraseasonal wmb at the mixed layer base accounts for half of that of E (Fig. 7a), but its overall contribution is only 4.3% because of the mismatch in timing (Fig. 7b). Both ∂hm/∂t and E show the peak in phase 2 and the trough in phase 5. However, the peak of wmb occurs in phase 8, and its trough is found in phase 3, which may be related to the downwelling Kelvin waves. Third, the lateral induction term umb ⋅ ∇hm induced by wind stress can also cause entrainment through horizontal currents across MLD slope. The contribution of umb ⋅ ∇hm is the smallest, with the STD of 1.81 × 10−6 m s−1 (Fig. 7a). The intraseasonal entrainment rate variability also shows prominent seasonality, which mainly arises from the MLD tendency term. It achieves the strongest intensity in March–April with the STD of >10 × 10−6 m s−1 and is weakest in July–August (<5 × 10−6 m s−1) (Fig. 7c). This seasonality is consistent with MJO seasonality that MJO is more energetic in the winter monsoon season than in the summer monsoon season (Zhang and Dong 2004).
(a) Intraseasonal entrainment-rate anomaly (black) and its three components, MLD tendency (orange), vertical pumping at the mixed layer base (cyan), and the lateral induction (pink), averaged in the EIO box. (b) MJO composites of 30–105-day entrainment rate and its three components. (c) Bimonthly STD of intraseasonal entrainment rate (black) and MLD tendency (orange). All results are derived from MR − NoSTRS (10−6 m s−1).
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
As shown in Fig. 8a, effective entrainment only accounts for a small fraction (<20%) of the total entrainment events. This indicates that more than 80% of the entrained water was detrained from the surface mixed layer during the previous 1-yr period. In the EIO, obduction of subsurface water mainly takes place in a small region near the equator between 77° and 92°E (Fig. 8b). The maximal obduction rate averaged from 2002 to 2014 is ~120 m yr−1 at ~83°E. The annual obduction rate integrated over the EIO region is 1.36 Sv (1 Sv ≡ 106 m3 s−1) in HYCOM MR. In addition to the annual obduction rate, we also compute the obduction rates in summer and winter separately taking into consideration the effects of monsoon. Here, summer is defined as the summer monsoon period of May–October, while the winter is defined as November–April (Duncan and Han 2009; Han et al. 2007). That is, the accumulated effective entrainment in May–October is calculated as the summertime obduction rate, and the entrainment for November–April is calculated as the wintertime obduction rate. Unexpectedly, we find that 84.6% of obduction occurs during summer monsoon periods (Fig. 8c), with only a small portion occurring during winter monsoon periods (Fig. 8d). This is primarily determined by the seasonal cycle of MLD (Fig. 2c). The deep MLD in response to the summer monsoon is favorable for harvesting water from the permanent pycnocline, while the relatively shallow MLD in the winter monsoon season makes it difficult for effective entrainment to occur.
(a) Percentage of effective entrainment relative to the total entrainment, and the (b) annual, (c) summertime (May–October), and (d) wintertime (November–April) obduction rates (m yr−1). All properties are averaged from 2002 to 2014 as derived from HYCOM MR. In (b)–(d), obduction-rate values smaller than 20 m yr−1 are left white.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
Next, we examine the impact of MJOs on the obduction of subsurface water in the EIO. Figure 9 compares the annual duration time for effective entrainment and summertime obduction rate averaged over 2002–14 between MR and NoMJO. The two simulations generally achieve similar spatial patterns in the EIO. It is interesting to see that without the MJO-related forcings, the duration time of effective entrainment is extended as compared with the case in MR (Figs. 9a,b). Correspondingly, the summertime obduction is larger in NoMJO (Figs. 9c,d). The EIO-integrated summertime obduction rate derived from NoMJO is 2.45 Sv (Table 2), more than twofold that from MR (1.15 Sv). In other words, our model results suggest that MJOs act to greatly suppress the summertime obduction of subsurface water in the EIO, reducing the summertime obduction rate by as large as 53%. Table 2 specifies the effects of different forcing processes of the MJO on the summertime obduction rate. It is seen that all the MJO-related forcing effects, including wind stress, wind speed, SWR, and precipitation, can suppress the obduction of subsurface water in the EIO, and the wind stress effect is the dominant contributor as expected. The SWR effect plays a secondary role but with a much smaller contribution.
The accumulation time of effective entrainment (days) from (a) MR and (b) NoMJO. Also shown are the summertime obduction rate (m yr−1) from (c) MR and (d) NoMJO. All properties are for 2002–14. Regions with obduction rates smaller than 20 m yr−1 are left white.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
The integrated summertime obduction rate in the EIO averaged from 2002 to 2014 from the MR, NoMJO, MJO effect, wind stress effect, wind speed effect, SWR effect, and precipitation effect. All of the units are Sverdrups (1 Sv ≡ 106 m3 s−1).
To understand how the ISV of MLD affects obduction in the EIO, we compare the probability density distributions of MLD derived from MR and NoMJO (Fig. 10a). MLD is primarily concentrated in the moderate range regardless of MJO-related forcings. 56% and 62% of MLD values from MR and NoMJO fall into the range of 20–50 m, respectively. By contrasting MR and NoMJO results, it is found that taking into account the MJO forcings can produce more extreme MLD values, i.e., >55-m and <15-m values. Figures 10b and 10c suggest that MR and NoMJO show rather similar seasonal variations in MLD, reaching the deepest in July (~50 m) and the shallowest in March (~30 m). The primary difference arises from the intraseasonal time scale. The intraseasonal MLD variability is much stronger in MR than in NoMJO, owing to the MJO-related forcing effects. The STD of intraseasonal MLD in NoMJO is merely ~1.2–1.9 m (Fig. 10c), much smaller than that of 4.0–9.1 m in MR (Fig. 10b). Moreover, MR shows much stronger intraseasonal MLD variability in spring (April–May) than in other seasons (Fig. 10b), which is consistent with SODA results that the monthly STD of intraseasonal MLD reaches the maximum in April–May and minimum in July–August (figure omitted). The difference in STD of intraseasonal MLD between MR and NoMJO also reaches the maximum in April–May (figure not shown). Therefore, the stronger intraseasonal MLD variability in spring may strongly affect the subsequent obduction in the summer monsoon season by rendering entrainment events ineffective. In the existence of intraseasonal MLD deepening, the possibility becomes much larger that the water entrained into the mixed layer in summer originates from the spring mixed layer (recall Figs. 9a,b).
(a) The probability distribution of MLD of the EIO box from the MR (black) and NoMJO (cyan). Also shown are the monthly climatological MLD (blue) and standard deviation of intraseasonal MLD variability (gray shading) in the EIO box from (b) MR and (c) NoMJO.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
Based on the analysis presented above, a conceptual schematic is postulated to demonstrate how MJO-induced intraseasonal MLD variability suppresses the summertime obduction in the EIO. For the cases without the MJO forcing effect, the obduction can be approximately described by the canonical model based on a smooth seasonal cycle (Qiu and Huang 1995), except that obduction occurs in summer featuring a region dominated by the Indian summer monsoon. As shown in Fig. 11a, although mixed layer entrainment takes place over a wide time window of one year, only a fraction of this entrained water, as indicated by the trajectories between A and B, comes from the permanent pycnocline that is not influenced by mixed layer and is labeled as effective entrainment. With the MJO-forcing effect taken into account, strong intraseasonal MLD fluctuations in spring are superimposed on the seasonal cycle, and the trajectories also show strong intraseasonal displacements in the vertical location (Fig. 11b). One can see that detrainment and entrainment take place alternately corresponding to intraseasonal shoaling and lowering of MLD. More importantly, some water trajectories between B and C are intersected by the deepened mixed layer in spring, rendering the entrainment ineffective. In other words, the bound of the effective entrainment window has to move from B to C because of the MJO-driven MLD variability. As a result, only the entrained water between A and C can be considered as effective entrainment and contribute to the summertime obduction rate. Therefore, the overall role played by the MJO forcing is to shorten the time window for effective entrainment and reduce the obduction of subsurface water in the EIO. According to our model simulations, this reduction effect can be as large as 53%.
The conceptual model of obduction for the cases (a) without and (b) with the MJO effects. The black curves indicate the instantaneous location of the mixed layer base. The slanted and curved thick horizontal arrows in (a) and (b), respectively, denote the trajectories of water mass entering into the mixed layer. The abscissa denotes both time and position of the water parcel entering into the mixed layer. The black dashed lines isolate the time windows of the detrainment and entrainment phases, and the blue and red dashed lines present the time window of effective entrainment for the cases without and with MJOs, respectively. Trajectory A corresponds to the ending of effective entrainment, and trajectories B and C correspond to the starting of effective entrainment for the cases without and with MJOs, respectively.
Citation: Journal of Physical Oceanography 51, 4; 10.1175/JPO-D-20-0179.1
5. Conclusions
The MLD is a key parameter for air–sea interaction, and its variability gives rise to detrainment/entrainment processes and contributes to the subduction/obduction of water masses. Strong entrainment usually implies a significant impact of the thermocline water on the mixed layer temperature/salinity and therefore a stronger coupling between the atmosphere and ocean dynamical processes. Particularly, the obduction near the equator represents the upward mass flux from the permanent pycnocline that conveys the climatic signals originating from the subtropics and higher latitudes and modulates the tropical climate on decadal and longer time scales (e.g., Gu and Philander 1997; Zhang et al. 1998; Li et al. 2012). The EIO is a key region of the initiation and development of the MJO, the dominant mode of the tropical atmospheric intraseasonal oscillations, and exhibits strong intraseasonal variability in response to MJOs. However, mechanisms of intraseasonal MLD variability in the EIO and impacts on the obduction of subsurface water were rarely investigated before. This study provides a comprehensive investigation of this issue.
Measurements by RAMA buoys near the Indian Ocean equator have documented prominent ISV of MLD with periods of 30–105 days. Comparisons with in situ observations reveal that HYCOM can well represent the MLD distribution and its variability in the EIO. Intraseasonal MLD variability is prominent in the EIO, where the STD of 30–105-day MLD reaches ~14 m. Six parallel HYCOM experiments are performed to assess the relative importance of MJO-related forcing effects, including wind stress, wind speed, SWR, and precipitation. It is demonstrated that wind stress is the major driver of the intraseasonal MLD variability in the EIO, which can explain 83.7% of the total ISV. In the active stage of MJOs (phase 3 of the RMM index), the increased surface wind stress causes enhanced turbulent mixing, and westerly wind anomalies drive upper-ocean convergence. The two processes work mutually to cause MLD deepening in the EIO. Besides, the wind speed and SWR effects also contribute to the MLD deepening through sea surface buoyancy forcing, but their roles are secondary. In the EIO, the composite MLD is deep during phases 3–5, reaching the largest depth in phase 4, and is shallow in phase 8. MJO-induced intraseasonal MLD deepening gives rise to high-frequency entrainment events of subsurface water, leading to an increase of the annual entrainment rate of the EIO by 34%. The HYCOM results show that MJO-related wind stress is the primary driver for the intraseasonal entrainment, while the contribution of SWR is secondary. Among the three components of entrainment, the wind stress-forced MLD tendency term plays a more important role than the vertical pumping term and the lateral induction term.
In the EIO, only a small percentage of these entrainment events (<20%) can effectively contribute to the annual obduction rate of 1.36 Sv, with the maximal obduction rate ~120 m yr−1. As determined by the seasonal variation of MLD, 84.6% of the obduction occurs in the summer monsoon season. Taking into account the MJO effects can greatly reduce the summertime obduction in the EIO by ~53%. All MJO-related forcing effects act to suppress the obduction of subsurface water, with the wind stress effect as the dominant contributor. Further analysis indicates that MJOs give rise to prominent intraseasonal MLD variability, which achieves the largest intensity in spring (April–May). These strong intraseasonal MLD deepening events in spring greatly affect the subsequent summertime obduction by rendering entrainment events ineffective. A conceptual schematic is postulated to depict how MJO-induced springtime intraseasonal MLD deepening events narrow down the time windows of effective entrainment and thereby suppress the obduction rate. The present study has clarified the mechanisms of intraseasonal MLD variability in the EIO region and highlighted its noticeable influence of suppressing the summertime obduction of subsurface water. There are many interesting problems to be addressed, such as the interannual and event-by-event variabilities of the ISV. These variabilities may cause corresponding modulations of obduction on different time scales and thereby exert impacts on SST. This topic warrants future exploration.
Acknowledgments
HYCOM experiments were performed on the Yellowstone supercomputer of the National Center for Atmospheric Research CISL and the INDOPAC machine of the University of Colorado. We thank two anonymous reviewers for providing insightful comments. This work is supported by National Natural Science Foundation of China under Grant 41976009, the Strategic Priority Research Program of Chinese Academy of Sciences (XDB42000000 and XDB40000000), the National Key Research and Development Program of China under Grant 2017YFC1404002, the Shandong Provincial Natural Science Foundation (ZR2020JQ17), and the Key Deployment Project of CAS Centre for Ocean Mega-Science (COMS2019Q07).
Data availability statement
In situ observational data from RAMA are provided by the NOAA Pacific Marine Environmental Laboratory (https://www.pmel.noaa.gov/gtmba/pmel-theme/indian-ocean-rama); MOAA GPV data are kindly provided by Dr. Hosoda (
REFERENCES
Alexander, M. A., J. D. Scott, and C. Deser, 2000: Processes that influence sea surface temperature and ocean mixed layer depth variability in a coupled model. J. Geophys. Res., 105, 16 823–16 842, https://doi.org/10.1029/2000JC900074.
Atlas, R., J. Ardizzone, and R. N. Hoffman, 2008: Application of satellite surface wind data to ocean wind analysis. Proc. SPIE, 7087, 70870B, https://doi.org/10.1117/12.795371.
Behrenfeld, M. J., 2010: Abandoning Sverdrup’s critical depth hypothesis on phytoplankton blooms. Ecology, 91, 977–989, https://doi.org/10.1890/09-1207.1.
Bender, M. A., I. Ginis, and Y. Kurihara, 1993: Numerical simulations of tropical cyclone-ocean interaction with a high-resolution coupled model. J. Geophys. Res., 98, 23 245–23 262, https://doi.org/10.1029/93JD02370.
Bentamy, A., and D. Croize-Fillon, 2012: Gridded surface wind fields from Metop/ASCAT measurements. Int. J. Remote Sens., 33, 1729–1754, https://doi.org/10.1080/01431161.2011.600348.
Carton, J. A., G. A. Chepurin, and L. Chen, 2018: SODA3: A new ocean climate reanalysis. J. Climate, 31, 6967–6983, https://doi.org/10.1175/JCLI-D-18-0149.1.
Chan, J. C. L., Y. Duan, and L. K. Shay, 2001: Tropical cyclone intensity change from a simple ocean–atmosphere coupled model. J. Atmos. Sci., 58, 154–172, https://doi.org/10.1175/1520-0469(2001)058<0154:TCICFA>2.0.CO;2.
Chang, S. W., and R. A. Anthes, 1978: Numerical simulations of the ocean’s nonlinear, baroclinic response to translating hurricanes. J. Phys. Oceanogr., 8, 468–480, https://doi.org/10.1175/1520-0485(1978)008<0468:NSOTON>2.0.CO;2.
Chen, D., A. J. Busalacchi, and L. M. Rothstein, 1994: The roles of vertical mixing, solar radiation, and wind stress in a model simulation of the sea surface temperature seasonal cycle in the tropical Pacific Ocean. J. Geophys. Res., 99, 20 345–20 359, https://doi.org/10.1029/94JC01621.
Chen, G. X., W. Han, Y. Li, D. X. Wang, and T. Shinoda, 2015: Intraseasonal variability of upwelling in the equatorial eastern Indian Ocean. J. Geophys. Res. Oceans, 120, 7598–7615, https://doi.org/10.1002/2015JC011223.
Chen, G. X., W. Han, Y. Q. Shu, Y. L. Li, D. X. Wang, and Q. Xie, 2016: The role of Equatorial Undercurrent in sustaining the eastern Indian Ocean upwelling. Geophys. Res. Lett., 43, 6444–6451, https://doi.org/10.1002/2016GL069433.
Chen, G. X., W. Han, Y. Li, J. L. Yao, and D. X. Wang, 2019: Intraseasonal variability of the equatorial undercurrent in the Indian Ocean. J. Phys. Oceanogr., 49, 85–101, https://doi.org/10.1175/JPO-D-18-0151.1.
Cushman-Roisin, B., 1987: Subduction. Dynamics of the Oceanic Surface Mixed Layer: Proc. ‘Aha Huliko‘a Hawaiian Winter Workshop, Honolulu, HI, University of Hawai‘i at Mānoa, 181–196, http://www.soest.hawaii.edu/PubServices/1987pdfs/Cushman_Roisin.pdf.
Dai, A., T. Qian, K. E. Trenberth, and J. D. Milliman, 2009: Changes in continental freshwater discharge from 1948 to 2004. J. Climate, 22, 2773–2792, https://doi.org/10.1175/2008JCLI2592.1.
de Boyer Montegut, C., G. Madec, A. S. Fischer, A. Lazar, and D. Iudicone, 2004: Mixed layer depth over the global ocean: An examination of profile data and a profile-based climatology. J. Geophys. Res., 109, C12003, https://doi.org/10.1029/2004JC002378.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, https://doi.org/10.1002/qj.828.
De Szoeke, R. A., 1980: On the effects of horizontal variability of wind stress on the dynamic of the ocean mixed layer. J. Phys. Oceanogr., 10, 1439–1454, https://doi.org/10.1175/1520-0485(1980)010<1439:OTEOHV>2.0.CO;2.
Dommenget, D., and M. Latif, 2002: Analysis of observed and simulated SST spectra in the midlatitudes. Climate Dyn., 19, 277–288, https://doi.org/10.1007/s00382-002-0229-9.
Drushka, K., J. Sprintall, and S. T. Gill, 2012: In situ observations of Madden–Julian oscillation mixed layer dynamics in the Indian and western Pacific Oceans. J. Climate, 25, 2306–2328, https://doi.org/10.1175/JCLI-D-11-00203.1.
Drushka, K., J. Sprintall, and S. T. Gill, 2014: Subseasonal variations in salinity and barrier-layer thickness in the eastern equatorial Indian Ocean. J. Geophys. Res. Oceans, 119, 805–823, https://doi.org/10.1002/2013JC009422.
Duncan, B., and W. Han, 2009: Indian Ocean intraseasonal sea surface temperature variability during boreal summer: Madden-Julian Oscillation versus submonthly forcing and processes. J. Geophys. Res., 114, C05002, https://doi.org/10.1029/2008JC004958.
Duvel, J. P., R. Roca, and J. Vialard, 2004: Ocean mixed layer temperature variations induced by intraseasonal convective perturbations over the Indian Ocean. J. Atmos. Sci., 61, 1004–1023, https://doi.org/10.1175/1520-0469(2004)061<1004:OMLTVI>2.0.CO;2.
Fasham, M. J. R., 1995: Variations in the seasonal cycle of biological production in subarctic oceans: A model sensitivity analysis. Deep-Sea Res. I, 42, 1111–1149, https://doi.org/10.1016/0967-0637(95)00054-A.
Felton, C. S., B. Subrahmanyam, V. S. N. Murty, and J. F. Shriver, 2014: Estimation of the barrier layer thickness in the Indian Ocean using Aquarius salinity. J. Geophys. Res. Oceans, 119, 4200–4213, https://doi.org/10.1002/2013JC009759.
Grunseich, G., B. Subrahmanyam, and A. Arguez, 2011: Influence of the Madden-Julian Oscillation on sea surface salinity in the Indian Ocean. Geophys. Res. Lett., 38, L17605, https://doi.org/10.1029/2011GL049047.
Gu, D., and S. G. H. Philander, 1997: Interdecadal climate fluctuations that depend on exchange between the tropics and extratropics. Science, 275, 805–807, https://doi.org/10.1126/science.275.5301.805.
Guan, B., T. Lee, D. J. Halkides, and D. E. Waliser, 2014: Aquarius surface salinity and the Madden-Julian Oscillation: The role of salinity in the surface layer density and potential energy. Geophys. Res. Lett., 41, 2858–2869, https://doi.org/10.1002/2014GL059704.
Han, W., D. Yuan, W. T. Liu, and D. Halkides, 2007: Intraseasonal variability of Indian Ocean sea surface temperature during boreal winter: Madden-Julian Oscillation versus submonthly forcing and processes. J. Geophys. Res., 112, C04001, https://doi.org/10.1029/2006JC003791.
Hendon, H., and J. Glick, 1997: Intraseasonal air–sea interaction in the tropical Indian and Pacific Oceans. J. Climate, 10, 647–661, https://doi.org/10.1175/1520-0442(1997)010<0647:IASIIT>2.0.CO;2.
Hendon, H., B. Liebmann, and J. D. Glick, 1998: Oceanic Kelvin waves and the Madden–Julian oscillation. J. Atmos. Sci., 55, 88–101, https://doi.org/10.1175/1520-0469(1998)055<0088:OKWATM>2.0.CO;2.
Hosoda, S., T. Ohira, and T. Nakamura, 2008: A monthly mean dataset of global oceanic temperature and salinity derived from Argo float observations. JAMETC Rep. Res. Dev., 8, 47–59, https://doi.org/10.5918/jamstecr.8.47.
Huang, R. X., 1988: On boundary value problems of the ideal-fluid thermocline. J. Phys. Oceanogr., 18, 619–641, https://doi.org/10.1175/1520-0485(1988)018<0619:OBVPOT>2.0.CO;2.
Huang, R. X., 1990: On the three-dimensional structure of the wind-driven circulation in the North Atlantic. Dyn. Atmos. Oceans, 15, 117–159, https://doi.org/10.1016/0377-0265(90)90006-H.
Huffman, G. J., and Coauthors, 2007: The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 38–55, https://doi.org/10.1175/JHM560.1.
Iselin, C. O. D., 1939: The influence of vertical and lateral turbulence on the characteristics of the waters at mid-depths. Trans. Amer. Geophys. Union, 20, 414–417, https://doi.org/10.1029/TR020i003p00414.
Jayakumar, A., J. Vialard, M. Lengaigne, C. Gnanaseelan, J. P. McCreary, and B. Praveen Kumar, 2011: Processes controlling the surface temperature signature of the Madden-Julian Oscillation in the thermocline ridge o the Indian Ocean. Climate Dyn., 37, 2217–2234, https://doi.org/10.1007/s00382-010-0953-5.
Kara, A. B., H. E. Hurlburt, and A. J. Wallcraft, 2005: Stability-dependent exchange coefficients for air–sea fluxes. J. Atmos. Oceanic Technol., 22, 1080–1094, https://doi.org/10.1175/JTECH1747.1.
Karstensen, J., and D. Quadfasel, 2002a: Water subducted into the Indian Ocean subtropical gyre. Deep-Sea Res. II, 49, 1441–1457, https://doi.org/10.1016/S0967-0645(01)00160-6.
Karstensen, J., and D. Quadfasel, 2002b: Formation of Southern Hemisphere thermocline waters: Water mass conversion and subduction. J. Phys. Oceanogr., 32, 3020–3038, https://doi.org/10.1175/1520-0485(2002)032<3020:FOSHTW>2.0.CO;2.
Keerthi, M. G., M. Lengaigne, K. Drushka, J. Vialard, C. de Boyer Montegut, S. Pous, M. Levy, and P. M. Muraleedharan, 2016: Intraseasonal variability of mixed layer depth in the tropical Indian Ocean. Climate Dyn., 46, 2633–2655, https://doi.org/10.1007/s00382-015-2721-z.
Kessler, W. S., M. J. McPhaden, and K. M. Weickmann, 1995: Forcing of intraseasonal Kelvin waves in the equatorial Pacific. J. Geophys. Res., 100, 10 613–10 631, https://doi.org/10.1029/95JC00382.
Kouketsu, S., H. Thmits, E. Oka, S. Hosoda, T. Kobayashi, and K. Sato, 2011: The role of meso-scale eddies in mixed layer deepening and mode water formation in the western North Pacific. J. Oceanogr., 68, 63–77, https://doi.org/10.1007/s10872-011-0049-9.
Kummerow, C., W. Barnes, T. Kozu, J. Shiue, and J. Simpson, 1998: The Tropical Rainfall Measuring Mission (TRMM) sensor package. J. Atmos. Oceanic Technol., 15, 809–817, https://doi.org/10.1175/1520-0426(1998)015<0809:TTRMMT>2.0.CO;2.
Levitus, S., 1982: Climatological Atlas of the World Ocean. NOAA Prof. Paper 13, 173 pp.
Li, Y., F. Wang, and Y. Sun, 2012: Low-frequency spiciness variations in the tropical Pacific Ocean observed during 2003-2012. Geophys. Res. Lett., 39, L23601, https://doi.org/10.1029/2012GL053971.
Li, Y., W. Han, T. Shinoda, C. Wang, R. C. Lien, J. N. Moum, and J. W. Wang, 2013: Effects of the diurnal cycle in solar radiation on the tropical Indian Ocean mixed layer variability during wintertime Madden-Julian Oscillations. J. Geophys. Res. Oceans, 118, 4945–4964, https://doi.org/10.1002/jgrc.20395.
Li, Y., W. Han, T. Shinoda, C. Wang, M. Ravichandran, and J. W. Wang, 2014: Revisiting the wintertime intraseasonal SST variability in the tropical Indian Ocean: Impact of the ocean interannual variation. J. Phys. Oceanogr., 44, 1886–1907, https://doi.org/10.1175/JPO-D-13-0238.1.
Li, Y., W. Han, and T. Lee, 2015: Intraseasonal sea surface salinity variability in the equatorial Indo-Pacific Ocean induced by Madden-Julian oscillations. J. Geophys. Res. Oceans, 120, 2233–2258, https://doi.org/10.1002/2014JC010647.
Li, Y., W. Han, W. Wang, M. Ravichandran, T. Lee, and T. Shinoda, 2017: Bay of Bengal salinity stratification and Indian summer monsoon intraseasonal oscillation: 1. Intraseasonal variability and causes. J. Geophys. Res. Oceans, 122, 4291–4311, https://doi.org/10.1002/2017JC012691.
Liu, L.-L., and R. X. Huang, 2012: The global subduction/obduction rates, their interannual and decadal variability. J. Climate, 25, 1096–1115, https://doi.org/10.1175/2011JCLI4228.1.
Liu, L.-L., R. X. Huang, and F. Wang, 2018: Ventilation of a monsoon-dominated oceans: Subduction and obduction in the North Indian Ocean. J. Geophys. Res. Oceans, 123, 4449–4463, https://doi.org/10.1029/2017JC013719.
Liu, L.-L., R. X. Huang, and F. Wang, 2020: Ventilation in the South China Sea: Subduction versus obduction. Deep-Sea Res. I, 158, 103227, https://doi.org/10.1016/j.dsr.2020.103227.
Loeb, N. G., K. J. Priestley, D. P. Kratz, E. B. Geier, R. N. Green, B. A. Wielicki, P. O. R. Hinton, and S. K. Nolan, 2001: Determination of unfiltered radiances from the Clouds and the Earth’s Radiant Energy System instrument. J. Appl. Meteor., 40, 822–835, https://doi.org/10.1175/1520-0450(2001)040<0822:DOURFT>2.0.CO;2.
Lombardo, C. P., and M. C. Gregg, 1989: Similarity scaling of viscous and thermal dissipation in a convecting surface boundary layer. J. Geophys. Res., 94, 6273–6284, https://doi.org/10.1029/JC094iC05p06273.
Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702–708, https://doi.org/10.1175/1520-0469(1971)028<0702:DOADOI>2.0.CO;2.
Marshall, D., 1997: Subduction of water masses in an eddying ocean. J. Mar. Res., 55, 201–222, https://doi.org/10.1357/0022240973224373.
Masumoto, Y., H. Hase, Y. Kuroda, H. Matsuura, and K. Takeuchi, 2005: Intraseasonal variability in the upper layer currents observed in the eastern equatorial Indian Ocean. Geophys. Res. Lett., 32, L02607, https://doi.org/10.1029/2004GL021896.
Matthews, A. J., P. Singhruck, and K. J. Heywood, 2010: Ocean temperature and salinity components of the Madden–Julian oscillation observed by Argo floats. Climate Dyn., 35, 1149–1168, https://doi.org/10.1007/s00382-009-0631-7.
McPhaden, M. J., and X. Yu, 1999: Equatorial waves and the 1997-98 El Niño. Geophys. Res. Lett., 26, 2961–2964, https://doi.org/10.1029/1999GL004901.
McPhaden, M. J., and Coauthors, 2009: RAMA: The Research Moored Array for African–Asian–Australian Monsoon Analysis and Prediction. Bull. Amer. Meteor. Soc., 90, 459–480, https://doi.org/10.1175/2008BAMS2608.1.
Monin, A. S., and A. M. Obukhov, 1954: Basic turbulent mixing laws in the atmospheric surface layer. Tr. Geofiz. Inst., Akad. Nauk SSSR, 24, 163–187.
Monterey, G., and G. Levitus, 1997: Seasonal variability of mixed layer depth for the World Ocean. NOAA Atlas NESDIS 14, 96 pp.
Nishikawa, S., H. Tsujino, K. Sakamoto, and H. Nakano, 2010: Effects of mesoscale eddies on subduction and distribution of subtropical mode water in an eddy-resolving OGCM of the western North Pacific. J. Phys. Oceanogr., 40, 1748–1765, https://doi.org/10.1175/2010JPO4261.1.
Papa, F., F. Durand, W. B. Rossow, A. Rahman, and S. K. Bala, 2010: Satellite altimeter-derived monthly discharge of the Ganga-Brahmaputra River and its seasonal to interannual variations from 1993 to 2008. J. Geophys. Res., 115, C12013, https://doi.org/10.1029/2009JC006075.
Polovina, J. J., G. T. Mitchum, and G. T. Evans, 1995: Decadal and basin-scale variation in mixed layer depth and the impact on biological production in the central and North Pacific, 1960-88. Deep-Sea Res. I, 42, 1701–1716, https://doi.org/10.1016/0967-0637(95)00075-H.
Qiu, B., and R. X. Huang, 1995: Ventilation of the North Atlantic and North Pacific: Subduction versus obduction. J. Phys. Oceanogr., 25, 2374–2390, https://doi.org/10.1175/1520-0485(1995)025<2374:VOTNAA>2.0.CO;2.
Qu, T., and J. Chen, 2009: A North Pacific decadal variability in subduction rate. Geophys. Res. Lett., 36, L22602, https://doi.org/10.1029/2009GL040914.
Qu, T., S. P. Xie, H. Mitsudera, and A. Ishida, 2002: Subduction of the North Pacific mode waters in a global high-resolution GCM. J. Phys. Oceanogr., 32, 746–763, https://doi.org/10.1175/1520-0485(2002)032<0746:SOTNPM>2.0.CO;2.
Rainville, L., S. R. Jayne, J. L. McClean, and M. E. Maltrud, 2007: Formation of subtropical mode water in a high-resolution ocean simulation of the Kuroshio extension region. Ocean Modell., 17, 338–356, https://doi.org/10.1016/j.ocemod.2007.03.002.
Schiller, A., and J. S. Godfrey, 2003: Indian Ocean intraseasonal variability in an ocean general circulation model. J. Climate, 16, 21–39, https://doi.org/10.1175/1520-0442(2003)016<0021:IOIVIA>2.0.CO;2.
Schott, F. A., S. P. Xie, and J. P. McCreary Jr., 2009: Indian Ocean circulation and climate variability. Rev. Geophys., 47, RG1002, https://doi.org/10.1029/2007RG000245.
Shinoda, T., and H. H. Hendon, 1998: Mixed layer modelling of intraseasonal variability in the tropical western Pacific and Indian Oceans. J. Climate, 11, 2668–2685, https://doi.org/10.1175/1520-0442(1998)011<2668:MLMOIV>2.0.CO;2.
Somavilla, R., C. González-Pola, and J. Fernández-Diaz, 2017: The warmer the ocean surface, the shallower the mixed layer. How much of this is true? J. Geophys. Res. Oceans, 122, 7698–7716, https://doi.org/10.1002/2017JC013125.
Stommel, H. M., 1979: Determination of water mass properties of water pumped down from the Ekman layer to the geostrophic flow below. Proc. Natl. Acad. Sci. USA, 76, 3051–3055, https://doi.org/10.1073/pnas.76.7.3051.
Sutton, P. J., P. F. Worcester, G. Masters, B. D. Cornuelle, and J. F. Lynch, 1993: Ocean mixed layers and acoustic pulse propagation in the Greenland Sea. J. Acoust. Soc. Amer., 94, 1517–1526, https://doi.org/10.1121/1.408128.
Thadathil, P., P. M. Muraleedharan, R. R. Rao, Y. K. Somayajulu, G. V. Reddy, and C. Revichandran, 2007: Observed seasonal variability of barrier layer in the Bay of Bengal. J. Geophys. Res., 112, C02009, https://doi.org/10.1029/2006JC003651.
Trossman, D. S., L. Thompson, K. A. Kelley, and Y.-O. Kwon, 2009: Estimates of North Atlantic ventilation of mode water formation for winters 2002–06. J. Phys. Oceanogr., 39, 2600–2617, https://doi.org/10.1175/2009JPO3930.1.
Uehara, H., T. Suga, K. Hanawa, and N. Shikama, 2003: A role of eddies in formation and transport of North Pacific subtropical mode water. Geophys. Res. Lett., 30, 1705, https://doi.org/10.1029/2003GL017542.
Wallcraft, A. J., E. J. Metzger, and S. N. Carroll, 2009: Software design description for the Hybrid Coordinate Ocean Model (HYCOM) Version 2.2. Naval Research Lab Tech. Memo., 157 pp., https://apps.dtic.mil/dtic/tr/fulltext/u2/a494779.pdf.
Wang, B., and X. S. Xie, 1998: Coupled modes of the warm pool climate system. Part 1: The role of air–sea interaction in maintaining Madden–Julian oscillation. J. Climate, 11, 2116–2135, https://doi.org/10.1175/1520-0442-11.8.2116.
Wang, F., Y. Li, and J. Wang, 2016: Intraseasonal variability of the surface zonal currents in the western tropical Pacific Ocean: Characteristics and mechanisms. J. Phys. Oceanogr., 46, 3639–3660, https://doi.org/10.1175/JPO-D-16-0033.1.
Wheeler,