Importance of vertical mixing and barrier layer variation on seasonal mixed layer heat balance in the Bay of Bengal

Abstract. Time series measurements from the Research Moored Array for African-Asian-Australian Monsoon Analysis and Prediction (RAMA) moorings at 15° N, 90° E; 12° N, 90° E; 8° N, 90° E; 4° N, 90° E; 1.5° N, 90° E; 0° N, 90° E are used to investigate the seasonal mixed-layer heat balance and the importance of barrier layer thickness (BLT) and vertical mixing (Q−h) in the Bay of Bengal (BoB). It is found that the BLT, Q−h and mixed-layer heat balance all have a strong seasonality in the central BoB. Sea surface temperature (SST), salinity and wind are important for the observed strongest seasonal cycle of BLT in the central BoB, and wind is more important than the SST in the southern BoB. The heat storage rate (HSR) is primarily driven by latent heat flux and shortwave radiation (QSW and QL). Seasonal variations and the magnitudes of longwave radiation (QLW), sensible heat flux (QS), and horizontal mixed-layer heat advection are much weaker compared to those of QSW and QL. Q−h follows a pronounced seasonal cycle in the central BoB and is significantly positively correlated with the seasonal cycle of BLT at each mooring location. The seasonal variability of the stability favors the Q−h during winter and summer monsoon and suppress Q−h during monsoon transition periods. We found that Q−h plays the secondary role in the seasonal mixed-layer heat balance in the BoB. It is evident from the analysis that Q−h associated with temperature inversion (∆T) warms the mixed layer during winter and cools the mixed layer during summer. The warming tendency during winter is strong in the central BoB and weakens towards the equator, indicating a cooling tendency around the year. Our analysis further indicates the weakening of Q−h during monsoon transition periods favors the existence of warmer SST in the BoB, associated with thermal and salinity stratification in the central BoB.


Introduction
The Bay of Bengal (BoB) is a semi-enclosed basin with unique characteristics due to the influences of Asian Monsoon and freshwater influx.It is distinguished by a strongly stratified surface layer and seasonally reversing circulation 60 (Shetye et al., 1996;Schott and McCreary, 2001), and also forced remotely by seasonal winds in the equatorial Indian Ocean (McCreary et al., 1993).Remote equatorial Kelvin waves influence the BoB via direct contact along the eastern boundary of the bay (Yu, 2003).Also, the BoB receives a large quantity of fresh water via precipitation and river runoff which exceed evaporation (Harenduprakash and Mitra, 1988), which makes thesurface layer buoyant and maintains strong stratification in the upper BoB (Shetye et al., 1996;Agarwal et al., 2012).This strong stratification 65 maintains the stability in the surface layer (Chowdary et al., 2016) and supports the formation of a barrier layer (BL), a unique layer between the base of the mixed layer and the top of the isothermal layer (Lukas and Lindstrom, 1991;Sprintall and Tomczak, 1992;Girishkumar et al., 2013).The presence of a BL restricts the mixing within the mixed layer, and also affects sea surface temperature (SST) by reducing the mixing of cool thermocline water in the mixed layer (Vialard and Delecluse, 1998b;Foltz and McPhaden, 2009), hence playing an important role in surface mixed-70 layer heat balance (Lukas and Lindstrom, 1991).Using a model simulation, Montegut et al., (2007b) suggested that thicker BLs are linked to positive SST anomalies in the northern Indian Ocean.Thus, understanding BL formation and its variabiltiy is important to explain the energy balance in the upper layer of the BoB.
Formation and variability of BL depend on the variability of isothermal layer depth (ILD), which is related to shoaling thermocline, Ekman pumping (Thadathil et al., 2008), mixed layer depth (MLD) (Vinayachandran et al., 75 2002), and wave propagation in the BoB (Chacko et al., 2012).Wind stress acts against the formation of a thick BL (Bosc et al., 2009) by deepening the mixed layer.Freshwater flux facilitates a thick BL (Cronin and McPhaden, 2002) by reducing MLD through stratification at the surface.Thus, the seasonality of BL thickness (BLT), which can influence vertical mixing (Wang et al., 2011), is an important phenomenon in the tropical oceans' surface-layer energy balance.

80
Seasonal variations in the parameters controlling the mixed-layer heat balance are associated with changes in the monsoonal winds over the BoB.McPhaden and Foltz (2013) suggested that in the tropics radiative fluxes dominate the surface heat flux variation; and Chacko et al., (2012) pointed out the significance of atmospheric forcing which influence the mixed-layer temperature/SST in the central BoB.Further, Chacko et al., (2012) suggested the importance of wind-induced heat loss and vertical entrainment in the mixed-layer heat balance.The net radiation increases from 85 a minimum in winter to a maximum in April-May, and thereafter decreases sharply with the start of summer monsoon.
Generally, during pre-summer monsoon the SST over the BoB is higher than 29°C (Chacko et al., 2012).The surplus energy at the sea surface warms the surface layer with shoaling ILD.Heat loss during winter cools the surface layer associated with a thicker BL in the BoB.Using the Research Moored Array for African-Asian-Australian Monsoon Analysis and Prediction (RAMA) mooring data in the southern BoB (8° N, 90° E), Girishkumar et al. (2013) where S is salinity, T is temperature and P is pressure.ILD is calculated as the depth where temperature is 0.8°C lower than the SST (∆ = 0.8°) (Du et al., 2005).BLT is defined as the difference of ILD and MLD. =  −  (Sprintall and Tomczak, 1992;Girishkumar et al., 2013).Chacko et al. (2012)  Ocean-Atmosphere Response Experiment (COARE) bulk algorithms (Fairall et al., 2003).As only the mooring at (15°N, 90°E) had longwave radiation (  ) measurements, net longwave radiation from the TropFlux is used (Kumar et al., 2011) for other mooring locations.Net shortwave radiation (  ) is estimated from downwelling shortwave 130 radiation measured at the mooring sites, corrected for albedo (6%) at the surface.The heat flux into the mixed layer due to air-sea exchange (  ) is estimated using the equation (Eq.2) below, With the convention that heat flux is positive when it is into the ocean.The penetrating shortwave radiation below the mixed layer is estimated using   = 0.47 ×   . −ℎ (Jouanno et al., 2011), considering a constant  − folding 135 depth of 25 m (k=0.04) and h (MLD).
To address the seasonal variability of the mixed-layer heat balance at each mooring location, we consider the following expression (Rao and Sivakumar, 2000;Foltz and McPhaden, 2009;Girishkumar et al., 2013;McPhaden and Foltz, 2013), The terms in (Eq. 3) represent, from left to right, heat storage rate (HSR), net surface heat flux, horizontal mixed-layer heat advection, and the combination of entrainment and vertical turbulent heat fluxes at the base of the mixed layer.The error estimation (∈) includes errors in the estimation of the terms in (Eq.3), which are associated with data sources and unrepresented/unresolved physical processes (Foltz and McPhaden, 2009).In (Eq.3),  is averaged mixed-layer temperature,  is density of seawater ( = 1024   −3 ),   is specific heat capacity of 145 seawater (  = 4000   −1  −1 ), ℎ is MLD,  is time, and  0 is the surface heat flux adjusted for the penetrative shortwave radiation through the base of the mixed layer.Heat fluxes at the mooring locations are estimated using the terms in (Eq.2).The zonal (u) and meridional (v) component of current measurements at 10-m depth are obtained from the moorings.Assuming that the temperature is uniformly distributed in the entire mixed layer, we use Optimallyinterpolated SST (OI SST) product with 0.25° × 0.25° resolution (Figure 2) to compute the horizontal advection term 150 in Eq. (3).OI SSTs averaged over 50 km on either side of the mooring locations are used to estimate the horizontal gradient of SST (Vialard et al., 2008;Girishkumar et al., 2013).

165
Before examining the variability of conditions at the RAMA mooring locations, we examine the climatological seasonal conditions in the BoB. Figure 1 shows the seasonal climatologies of BLT, precipitation and winds in the BoB. Figure 2 illustrates the seasonal climatology of salinity and SST anomalies in the bay.During winter monsoon (December-February), relatively larger BLT is present from the central to northern BoB (Figure 1a), when the SST cooling is the largest (Figure 2a) and there is relatively low precipitation.Seasonal cycle of salinity shows 170 freshening in the northern BoB during winter (Figure 2a), which indicates the importance of river runoff associated with surface cooling for the formation of thicker BL in the BoB.BL almost disappears during pre-summer monsoon (March-April) associated with low precipitation (Figure 1b), low surface freshening and warmer SST (Figure 2b).
During summer monsoon (May-September), BLT is relatively thicker in the eastern boundary of the BoB, associated with higher precipitation (Figure 1c) and surface freshening (Figure 2c).During post-summer monsoon (October-

175
November), BLT varies between 0-30 m (Figure 1d), associated with low salinity in the northern BoB and with SST cooling in the southern BoB (Figure 2d).Seasonal climatology of wind clearly indicates the weakening or strengthening of surface winds over the BoB (Figure 1), which is one of the important factors for the formation of thicker or thinner BL (Bosc et al., 2009).
Next, we examine the stratification in the upper 140 m at the RAMA locations in the BoB.The moorings are 180 located from the central BoB to the equator, which are in a region of strong seasonal variability of BLT (Figure 1).
We have selected the mooring at 15°N, 90°E (Figures 3a, 3b), which has the longest data availability and the significant seasonal cycle of BLT and the mooring at 12°N, 90°E (Figures 3c, 3d exhibits a prominent seasonal variation (Figure 6d) with surface freshening and wind forcing, reaching a maximum in July (42±8 m) when the wind is at its maximum (Babu et al., 2004).ILD varies out of phase with SST, reaching its maximum (87±18 m) in February when SST is minimum and reaching its minimum (16±18 m) in April when SST is at its maximum (Figure 6).The seasonal cycle of BLT varies with ILD, reaching its maximum (69±19 m) in February (highest ILD) and its minimum (2±19 m) in April (lowest ILD) (Figure 3b) then with MLD during summer.Towards 190 the southern BoB, the variability of BLT varies in phase with MLD.Thus, it indicates that SST, salinity and wind are important for the observed strongest seasonal cycle of BLT in the central BoB, and that wind is more important than SST for the variability of BLT in the southern BoB (Girishkumar et al., 2011;Felton et al., 2014).Then we use the NCOM model estimations to compare the seasonal variability of the conditions observed at 15°N, 90°E in the BoB (Figure 4).Monthly climatology of MLD is over estimated in NCOM compared to that of

195
RAMA, where the difference is larger during winter (Figure 4a).The estimated ILD agrees well indicating that the effect from temperature is similar in both data sources (Figure 4b), hence it is evident that the difference observed in the MLD estimations are associated with the effect due to salinity.NCOM under estimate the BLT during post-summer monsoon and winter, which indicates the effect of MLD estimation (Figure 4c).Though there are differences in the magnitudes, the observed seasonal variability in the upper layer stratification at 15°N, 90°E by the RAMA mooring 200 is evident from the NCOM model estimations.The computed vertical temperature gradient (dT/dz) (Figure 5a) and salinity gradient (dS/dz) (Figure 5b) illustrate the seasonal variability of sub surface conditions.It is evident from both data sources, the existence of homogenous layer during summer which coincide with maximum wind speed and the highest MLD.The thermal stratification is largest during pre-summer monsoon (Figure 5a) which coincides with the time period of higher insolation and the salinity stratification is largest during post-summer monsoon (Figure 5b).The 205 estimated upper ocean stability illustrates that the upper ocean layers at 15°N, 90°E are more stable during monsoon transition periods compared to that of winter and summer (Figure 5c).Thus the results pointed out that winter and summer favors the vertical mixing (Thangaprakash et al., 2016) with the presence of more unstable layers in the central BoB, and pre and post-summer monsoon tends to inhibit the vertical mixing due to the presence of more stable water layers. 210

Mixed-layer heat balance
Measurements from 15°N, 90°E and 12°N, 90°E reveal pronounced seasonal cycles of SST, wind speed, net surface heat flux, MLD, and ILD during 2008-2016 (Figure 6).SST reaches its maximum (30.4±1°C) in pre-summer monsoon and its minimum (26.6±1°C) in winter (Figure 6a).Wind speed reaches its maximum (9.2±1.8 ms -1 ) during summer and its minimum (2.9±1.8 ms -1 ) in pre-summer monsoon (Figure 6b).The surface heat flux follows the 215 seasonal cycle of SST, tends to heat the mixed layer during the pre-summer through post-summer monsoon and tends to cool the mixed layer during winter (Figure 6c).Both MLD and ILD vary out of phase with wind and SST, where the seasonal cycle is strong in the central BoB (Figure 6d).Higher differences in SST (~1°C) are observed during winter, from the central to southern BoB; and less differences in SST (<0.5°C), during pre-to post-summer monsoon.
The mean SST at the mooring locations increases towards the equator, and the moorings located at 15°N and 12°N

220
experience the highest SST during pre-summer monsoon.Wind speed undergoes a more pronounced seasonal cycle at 15°N and 12°N compared to that at 8°N, 4°N, 1.5°N and 0°N, tending to enhance the seasonal cycle of Q L in the central BoB.
Based on the strongest seasonal cycles observed, we consider the mixed-layer heat balance at 15°N, 90°E (Figure 7).The seasonal cycle of SST at 15°N, 90°E is driven primarily by changes in the net surface heat flux.

225
McPhaden and Foltz [2013]  Thus, these results indicate that the pronounced seasonal cycle of mixed-layer heat storage rate is driven primarily by the seasonal variability of      in the central BoB (Figure 7d).Mixed-layer heat storage rate (HSR) reaches its maximum (43±24 W m -2 ) in pre-summer monsoon and its minimum (-42±24 W m -2 ) in summer (Figure 7d).
235 Thangaprakash et al. (2016) suggested that the penetrative component of shortwave radiation (  ) plays a crucial role in the mixed-layer heat balance in the BoB, especially during the pre-summer monsoon and it is evident from our results (Figure 7a).During winter, HSR illustrates a cooling tendency in the central BoB (Figure 7d), but its magnitude is still larger than that of the net surface heat flux.Though HSR is primarily driven by      , the difference observed during summer and winter monsoons brings the importance played by other terms in mixed-layer heat 240 balance as secondary.
The observed mixed-layer heat advection at the mooring locations is much weaker (Figure 8), and the contribution to the mixed-layer heat balance is less significant, illustrating the importance of vertical mixing.To study the importance of entrainment and vertical turbulent heat fluxes (hereafter vertical process) at the base of the mixed layer, we compute the vertical process following the methods used in Foltz and McPhaden (2009)  The residual (∈) term is larger compared to the calculated entrainment and vertical diffusion at the mooring locations, showing the uncertinities due to unpresented/unresolved physical processes in the mixed-layer heat balnce.

250
The  −ℎ term undergoes a strong seasonal cycle at 15°N, 90°E, tending to cool the mixed layer at a rate up to 66±38 Wm -2 during summer and to warm the mixed layer up to 70±38 Wm -2 during winter (Figure 8a).Contribution by the vertical process to the mixed-layer heat balance decreases towards the equator (Figure 8), indicating the dominance of atmospheric forcing in influencing the mixed-layer heat balance.BLT varies in phase with  −ℎ , reaching its maximum of 30-70 m during winter and its minimum of <3 m in pre-summer monsoon (Figures 3b, 7a).The 255 correlation coefficient for daily-averaged BLT and  −ℎ is 0.84 (Table 1).Foltz and McPhaden (2009) suggested the warming and cooling tendencies by  −ℎ are associated with the temperature differences (∆) at the base of the mixed layer and the BLT can exert a significant influence on  −ℎ through its modulation of ∆.Thus, it indicates the importance of ∆ (both positive and negative) associated with the variability of BLT to the mixed-layer heat balance in terms of  −ℎ (Figure 9a).

260
The combination of surface cooling and a thicker BL is important for the generation of temperature inversion at the base of the mixed layer.During winter, temperature inversion is prominent (Montegut et al., 2007b;Girishkumar et al., 2011;Girishkumar et al., 2013)  The effect from the vertical process to the mixed-layer heat balance during monsoon transition periods 275 remains relatively small compared to the other seasons, while during June vertical pocess is more important compared to other terms in mixed-layer heat balance (Figure 9c).The role of the vertical processes during monsoon transition periods is important due to the presence of persistent warm SST in the BoB, higher than 28°C (Shenoi et al., 2002), which is generally considered as the threshold for atmospheric convection (Johnson and Xie, 2010).Figure 9b illustrates the terms in the vertical process, computed following Girishkumar et al. (2013) and Zeng and Wang (2016), averaged using a 7-day running mean filter to identify the variation. −ℎ term calculated following Foltz and McPhaden (2009) indicates the seasonal cycle of the vertical process at 15°N, 90°E.During pre-summer monsoon,  −ℎ changes its phase from warming (winter) through cooling (summer), indicating the seasonal variability associated with temperature inversion and BLT (Figure 9b).From August to September, the contribution of  −ℎ indicates a warming tendency; and during October, there is a cooling tendency with less significance.The contributions from 285 entrainment and vertical diffusion illustrate a cooling tendency during August-September, which indicates a missing source of warming in the central BoB.Foltz and McPhaden (2009) pointed out that the missing sources of warming and cooling found at the PAIRATA moorings in the tropical Atlantic are mainly due to differences in the estimation of horizontal eddy heat advection and penetrative shortwave radiation.

Importance of vertical process during post-summer monsoon in the central BoB 290
The RAMA moorings at 15°N, 90°E; 12°N, 90°E and 8°N, 90°E illustrate a positve heat storage rate during August-October, which is the period that the central BoB shows the second highest SST due to positive net surface heat flux (Figure 6).Late phase of summer monsoon and post-summer monsoon are an important period for the formation of deep depressions and cyclones in the region.Deepening of the ocean mixed layer associated with SST cooling is thought to inhibit cyclone intensification (Emanuel, 1999;Wentz et al., 2000).Shay et al. (2000) pointed

295
out that if SST does not cool because the upper ocean has a deep warm water layer, a cyclone could intensify rapidly and sustains its intensity longer.Our results illustrate the shallowest MLD (Figure 6d), associated with a moderate BL (Figure 8), forms during this season in the central BoB.Thus, stratification helps to maintain a warm surface layer, inhibiting SST cooling, and contribution from vertical mixing remains minimum during this period (Figures 8a, 9b).and that may be the reason for the observed differences during cyclone events.Observations from moored RAMA buoys revealed that the importance of seasonal vertical process in SST cooling/warming associated with BLT in the 305 BoB.

Summary and Conclusions
In this study, we examine the seasonal mixed-layer heat balance and the importance of the vertical process and BLT in the BoB, using time series measurements recorded at 15°N, 90°E; 12°N, 90°E; 8°N, 90°E; 4°N, 90°E; 1.5°N, 90°E; 0°N, 90°E by the RAMA moorings.At all the mooring locations, it is found that the seasonal changes 310 in mixed-layer HSR is primarily driven by shortwave radiation (  ) and latent heat flux (  ).The seasonality of HSR is more pronounced in the central BoB.Seasonal variations and magnitudes of longwave radiation (  ), sensible heat flux (  ) are smaller compared to those of      .The horizontal mixed-layer heat advection also weaker compared to that of vertical mixing.The vertical mixing at the base of the mixed layer ( −ℎ ), estimated as the residual in the heat balance following Foltz and McPhaden (2009), also follows a pronounced seasonal cycle in 315 the central BoB, and is correlated positively with the seasonal cycle of BLT at each mooring location.We find that  −ℎ plays the secondary role in mixed-layer heat balance in the BoB.It is evident from the analysis that the vertical mixing associated with temperature inversion (∆) warms the mixed layer during winter and cools the mixed layer during summer.The warming tendency during winter is strong in the central BoB and weakens towards the equator, indicating a cooling tendency around the year.The impact of BLT on  −ℎ is the strongest at 15°N, 90°E where the 320 seasonal cycle of BLT is the strongest, which is consistent with the results of Foltz and McPhaden (2009) in the central tropical Atlanctic.
To examine the importance of entrainment and vertical diffusion in the vertical process, we estimated vertical mixing following Girishkumar et al. (2013), and found that entrainment is more important in the vertical process.We Thus, it illustrates the imporatnce of the seasonal cycle of BLT on the mixed-layer heat balnce in the central BoB.
The results of this study thus indicate the importance of BLT and vertical mixing on the seasonal mixed-layer heat balance in the BoB.Late phase of summer monsoon and post-summer monsoon are a period of active air-sea interaction in the BoB, and it is possible that weakening of vertical mixing and strong stratification (higher stability)

345
during this period influence the intensity and frequency of BoB cyclones.Moreover, studies with systematic measurements are needed to understand the upper-ocean dynamics, the process of vertical mixing and its influence on mixed-layer temperature in the BoB, which can influence the weather and climate in the region and beyond.
suggested 90 the significance of BL and temperature inversion on mixed-layer heat budget.Using the Pilot Research Moored Array in the Tropical Atlantic (PIRATA) mooring data at three locations (15°N, 38°W; 12°N, 38°W; 8°N, 38°W), Foltz and Ocean Sci.Discuss., https://doi.org/10.5194/os-2017-67Manuscript under review for journal Ocean Sci. Discussion started: 4 September 2017 c Author(s) 2017.CC BY 4.0 License.120 ∆  =   ( + ∆, , ) −   (, , ) ponited out that the estimation of MLD and ILD by 0.8°C criterion is reasonably good for the BoB compared to 0.5°C or 1.0°C criterion.125 Ocean Sci.Discuss., https://doi.org/10.5194/os-2017-67Manuscript under review for journal Ocean Sci. Discussion started: 4 September 2017 c Author(s) 2017.CC BY 4.0 License.Air-sea fluxes at the mooring locations are computed from the daily winds extrapolated to the 10-m height, SST, air temperature and relative humidity.Latent (  ) and sensible (  ) heat fluxes are estimated using the Coupled ) to illustrate the seasonal stratification in the central BoB.The estimated mean errors in MLD and ILD are typically ±2 m and ±3 m with a standard deviation of ±8 m and ±18 m, giving errors in BLT of around ±5 m with a standard deviation of ±19 m.MLD in the central BoB 185 Ocean Sci.Discuss., https://doi.org/10.5194/os-2017-67Manuscript under review for journal Ocean Sci. Discussion started: 4 September 2017 c Author(s) 2017.CC BY 4.0 License.
and Girishkumar et   245   al. (2013).The estimated  −ℎ (Figure8) followingFoltz and McPhaden (2009) is the difference between the mixedlayer heat storage rate and the sum of the first two terms in equation (3).The estimated vertical process followingGirishkumar et al. (2013) is the summation of entrainment and vertical diffusion ( [ ℎ in the region.Girishkumar et al. (2013) pointed out during times when thicker BL and temperature inversion (∆) occur coincidentally, the vertical process shows a strong warming tendency due Ocean Sci.Discuss., https://doi.org/10.5194/os-2017-67Manuscript under review for journal Ocean Sci. Discussion started: 4 September 2017 c Author(s) 2017.CC BY 4.0 License.inversion supported by BL in the BoB affects surface temperature through entrainment of warm subsurface waters into the mixed layer.We observed the warming tendencies by the vertical process during winter at 15°N, 12°N and 8°N in the BoB similar to that ofFoltz and McPhaden (2009) for the tropical Atlantic and ofVialard and Delecluse (1998b) for the western tropical Pacific.Further our results indicate that the vertical process tends to cool the mixed layer at 4°N, 1.5°N and 0°N around the year.During summer, vertical mixing tends to cool the mixed layer in the 270 central BoB (Figure8a), reaching its mximum during June.The rate of cooling by vertical mixing in summer decreases from the central BoB to equatorward.Pre-summer through summer monsoon is a period with high net heat flux, tending to increse the temperature in the mixed layer in the central BoB.Thus, it points out vertical mixing during summer plays the secondary role in mixed-layer heat balance.
Chacko et al. (2012) suggested the siginificance of vertical mixing over surface forcing in inducing mixed-layer Ocean Sci.Discuss., https://doi.org/10.5194/os-2017-67Manuscript under review for journal Ocean Sci. Discussion started: 4 September 2017 c Author(s) 2017.CC BY 4.0 License.temperature variability in the BoB and Sengupta et al. (2008) pointed out that during post-summer monsoon the northern BoB responds quite differently to cyclones than during pre-summer monsoon.Stability during post-summer monsoon due to salinity stratification is stronger compared to that of pre-summer monsoon due to thermal stratification have found a missing source of warming during August-September in the central BoB up to ~25 Wm -2 .The 325 uncertinities are mainly associated with measurement errors, calculation errors and parameterization of the vertical process.Our results further indicate that entrainment is weaker during post-summer monsoon period, which tends to weaken the SST cooling by vertical mixing and helps to maintain warmer surface temperature at all the locations.The seasonal variability of the upper ocean stability favors the  −ℎ during winter and summer monsoon and suppress  −ℎ during monsoon transition periods in the BoB.The surface heat fluxes alone do not account for the changes observed 330 in seasonal mixed-layer heat balance.Thus, it brings the importance of vertical mixing, which influences the seasonal variability of mixed-layer heat balance in the BoB.This study further indicates that MLD, ILD and BLT undergo a strong seasonal cycle in the central BoB.It is evident from our results the change in ILD with SST is important for the change in BLT during winter and presummer monsoons, while the change in MLD with wind and surface freshning is important during summer and post-summer monsoons in the central BoB.The significant positive correlation between BLT and  −ℎ means that vertical mixing is the weakest when the BL is the thickest.We have found that, time periods with the thicker and thinner barrier layers are associated with significant vertical mixing where the moderate BLT supresses the vertical mixing in the central BoB during the periods with strong upper ocean stability.The warming and cooling tendencies by vertical mixing associated with the variability of BLT in the central BoB are consistent with the results of Vialard and 340 Delecluse (1998b) in the western equatorial Pacific and Foltz and McPhaden (2009) in the central tropical Atlantic.

Figure 3 .
Figure 3.Time series of daily averaged data from January 2008 -January 2017 (a, b) daily RAMA buoy data at 15 °N, 90 °E, (c, d) daily RAMA buoy data at 12 °N, 90 °E, (a, c) sub surface temperature, (b, d) sub surface salinity.In the figure (a, c) thick, thin and dashed lines indicate MLD, ILD and D23, (b, d) dashed line indicate BLT.

Figure 6 .
Figure 6.Seasonal cycles of (a) SST, (b) wind speed, (c) net surface heat flux, (d) mixed layer depth (solid) and isothermal layer depth (dashed) from RAMA moorings at 15°N (black), and at 12 °N (red).All-time series have been smoothed using 3-month running mean filter.