Role of external forcing on the seasonal and interannual variability of mixed layer depth over the Bay of Bengal using reanalysis datasets during 1980–2015

Highlights

• The seasonal variability related physical processes are responsible for deepening (shallowing) of MLD during summer (winter) monsoon in the BoB.

• The equatorial Kelvin waves associated with IOD modulates interannual variability of MLD in the BoB.

• Interannual variability indicates that the NIOD (PIOD) event experiences the deepening (shallowing) of MLD in the eastern part of the BoB.

Abstract

The monsoon reversal winds in different seasons and high influx of freshwater from various rivers make the Bay of Bengal (BoB) a unique region. Thus, the knowledge of the dynamics of the mixed layer over this region is very important to assess the climatic variation of the Indian subcontinent. A comprehensive study of the role of external forcing on the seasonal and interannual mixed layer depth (MLD) variability over the BoB is carried out for 36 years (1980–2015) using reanalysis products. A weak and strong seasonality of MLD is observed over the northern and the southern BoB (NBoB and SBoB) respectively. The partial correlation suggests that the net heat flux (Qnet) is the major contributor to the deepening of MLD over the NBoB, whereas the wind stress controls the deepening over the SBoB. The seasonal variability reveals the deepening of MLD during summer and winter monsoon and the shallowing during pre- and post-monsoon over the BoB. The relation of the interannual MLD variability and the different phases of the Indian Ocean Dipole (IOD) reveals that the negative phase of IOD is associated with deeper MLD over BoB while the positive phase of IOD depicts shallower MLD. In addition, the opposing characteristic of MLD is highly prominent during October-December. This is majorly contributed by variations related to the second downwelling Kelvin and associated Rossby waves over BoB for the opposing phases of the IOD years.

Introduction

The high latitudes and regions like the Bay of Bengal (BoB) with high freshwater influx shows an exception with the Mixed Layer Depth (MLD) to be saltier than the deeper layers. The study of mixed layer dynamics is essential for several oceanic and atmospheric processes like the heat transportation from the atmosphere to deep oceans, cyclogenesis, fishery studies, delineating good/bad monsoon years (Shukla, 1975; Rao and Goswami, 1988; Mohanty and Mohan Kumar, 1990) and the prediction of cyclone track (Goni and Trinanes, 2003; Price, 2002). The information of MLD is also used in underwater acoustic propagation for scientific and defense applications (Sutton et al., 2005). In the North Indian Ocean (NIO) the mixed layer varies on several temporal scales ranging from diurnal to intraseasonal and seasonal (Weller et al., 2002; Babu et al., 2004). The thickness of the MLD determines the heat content and inertia that directly interact with the atmosphere (Udaya Bhaskar et al., 2006). In the NIO a shallow (deeper) MLD has lesser (higher) heat content and hence is more (less) reactive to the atmospheric fluxes (Keerthi et al., 2013). Shaji et al. (2003) found that the Arabian Sea (AS) and BoB have very distinct near-surface heat budgets while observing the response of the upper ocean to the seasonal cycle. The study using a 1-D turbulent closure model concluded that the major cause for differential MLD and Sea Surface Temperature (SST) in the AS and BoB is due to the differences in the buoyancy forcing and surface wind stress rather than the vertical salinity gradient (Prasad, 2004). The circulation of the AS and BoB is affected by the latent heat fluxes which are majorly controlled by the wind variation over the surface (De Boyer Montégut et al., 2007). A recent study found that the mixed layer variation over NIO is subjected to the surface circulation which is majorly contributed by the near-surface wind forcing (Srivastava et al., 2018). Narvekar and Prasanna Kumar (2014) examined the seasonal variability of MLD in association with biomass over the BoB and concluded that there is a coupling between the MLD variability and chlorophyll in certain seasons. A study relating the effect of seasonal river input over the BoB using the Regional Ocean Modelling System (ROMS) showed that the wind stress (buoyancy flux) plays a crucial role during summer (winter) in the variation of MLD over the BoB (Jana et al., 2015). The MLD is an important factor in determining the intraseasonal variability and strength over the BOB and has a strong seasonal evolution during the boreal summer (Liu and Wang, 2012). A recent study by Liu et al. (2020) developed an upgraded subseasonal prediction system which considers the seasonal evaluation of MLD change even in one season.

Interannual MLD variability over the NIO for the summer season is limited to the AS and is due to the changes in the intensities of the monsoonal winds (Carton et al., 2008). Hong and Li (2010) found that there is a negative SST skewness over the Eastern Equatorial Indian Ocean which attributes to the asymmetry of the Mixed Layer Temperature tendency during the Indian Ocean Dipole (IOD) developing months (June – August) which is primarily contributed by the horizontal temperature advection. In the case of tropical cyclones (TCs), over the NIO the stormy winds thicken the MLD by mixing the heat, deep into the ocean (Vissa et al., 2013; Roy Chowdhury et al., 2020). Li et al. (2015) observed that the TCs activity increases over the Southeast Equatorial Indian Ocean mainly during the negative Indian Ocean Dipole (NIOD) event. Over the BoB the TCs show a significant interannual variability between the NIOD and positive IOD (PIOD) events but no prominent variation is found between the El Niño and La Niña events (Li et al., 2016). Hong et al. (2010) observed that there is an asymmetry in the MLD between the warm and the cold IOD events which contributes to a positive skewness in the Indian Ocean Basin. Furthermore, they found that the MLD over the eastern (western) Indian Ocean is shallower (deeper) during the warm IOD events than the cold IOD events. An interannual MLD variability study over the NIO reported that the El Nino-Southern Oscillation (ENSO) has less role in the MLD variability and is mainly confined to the AS. They concluded that the IOD has a significant effect on the interannual variability of MLD over the NIO (Keerthi et al., 2013). The seasonal variability of MLD and the impact of external forcing have been well investigated whereas, very little work can be found for interannual variability over the BoB. Thus, we confine our interannual variability study to the different phases of the IOD years.

In the tropical zone, the BoB is a region with a constant vertical salinity gradient and a thick barrier layer. This is due to the high freshwater influx and heavy precipitation in this region (Schiller and Godfrey, 2003; Waliser, 2006; Girishkumar et al., 2017). The study of variability over the BoB is a significant factor in assessing climate change in the Indian subcontinent. Thus, to govern the climatic variability over this region, it is very imperative to study the dynamics of the mixed layer over the BoB. In general, the BoB experiences a weaker seasonal MLD variation in comparison to other parts of the NIO (Shenoi, 2002; Prasad, 2004; Keerthi et al., 2016). In this study, the latitudinal area ranges from $5°N\text{to}24°N$ and longitudinal area ranges from $80°E\text{to}100°E$. Over the BoB shows a distinct MLD structure and depicts a clear difference between the north-south regions with $15°\text{N}$ the limiting latitude (Fig. 1) (Narvekar and Prasanna Kumar, 2014). As per previous studies, it is concluded that in the world ocean the Net Heat Flux (Qnet), wind stress, and Evaporation minus Precipitation (E-P) are the external local forces, which affect the mixed layer dynamics. Thus, we divide the study area into the Northern BoB (NBoB) and Southern BoB (SBoB) and investigate the role of external forcing on the seasonal and interannual variation of MLD over the BoB for 36 years from 1980 to 2015 using reanalysis products. The second section deals with the obtained data and the methods used. The results are discussed in the third section with seasonal and interannual studies. The fourth section concludes and summarizes the outcome of the study.