A coupled modeling approach to understand ocean coupling and energetics of tropical cyclones in the Bay of Bengal basin

Highlights

• The coupled simulations outperform the standalone simulations with 2% improvement in intensity and realistic represention of the rapid intensification phase.

• The coupling strength is due partly to latent heat release and partly to higher latent energy flux at the surface boundaries, and SST variations along with VWS modulates the intensity.

• As VWS increases, high low-level θ_egets advected cyclonically from the downshear-right to the downshear-left quadrants of the storm enhancing convection.

• Bulk energetically growing stage of the tropical cyclone can be termed as a moisture replenishment phase, as the intensity reduces in favour of reorganizing the storm core.

• A cyclone can intensify even in the presence of strong VWS due to enhanced coupling strength.

Abstract

In this study, a coupled model is implemented in the Bay of Bengal basin for simulating tropical cyclones Phailin and Hudhud, with the latter experiencing strong vertical wind shear (VWS). The experiments conducted include a control run (CTL) with the WRF model and a coupled run (CPL) with the COAWST model. Validations prove that the coupled simulations perform better in terms of intensity, track error and equitable threat score. CPL_Hudhud shows enhanced atmosphere-ocean coupling with respect to CTL_Hudhud, while the effect is minute in the case of Phailin. The peak VWS period of tropical cyclone Hudhud shows that high low-level θ_e gets advected cyclonically from the downshear-right of the storm into the cyclone's inflow layer located in the downshear-left half and is marked by strong radar echoes. Storm energetics reveal that bulk kinetic energy and bulk latent energy have opposite trends during the bulk-energetically growing stage of the storm, and kinetic energy increases at the expense of total potential energy through cross-isobaric flows because of rapid moisture consumption by deep convection. SST and VWS are the crucial factors that determine the atmosphere-ocean coupling, but CPL_Hudhud intensified even in the presence of strong VWS due to enhanced moisture consumption through deep convection, enhanced surface fluxes of moisture and kinetic energy generation. The thermodynamic efficiency of tropical cyclones prove to be quite low as most of the latent heat released is used to warm the atmosphere. In short, this study highlights the importance of a coupled model in simulating extreme weather events and the significance of VWS along with SST in dictating the storm energetics.

Introduction

Sea surface temperature (SST) and its relative importance in tropical cyclone maintenance, and intensification is a well-established fact (Emanuel, 1999, Emanuel, 1987; Gray, 1998; Miller, 1958). SST plays a crucial role in the heat flux exchange with the atmosphere, vertical mixing, and upwelling by Ekman pumping (Yablonsky and Ginis, 2009). Miller (1958) found the maximum attainable intensity of a tropical cyclone to be a function of SST. Numerical experiments have also shown the intensity of tropical cyclones to be sensitive towards SST (Emanuel, 1987; Lavender et al., 2018; Warner et al., 2010). The heat fluxes from the ocean that drive tropical cyclones are enhanced by turbulence due to strong winds and increased saturation mixing ratio due to a pressure drop at the surface (Schade, 2000). Since tropical cyclones interact with the upper ocean and not just the surface, SST can sometimes be misleading as it does not represent the actual energy available for cyclone intensification (Ali et al., 2013). Mogensen et al. (2017) conducted numerical experiments with a coupled model and concluded that the upper ocean stratification plays a crucial role in determining the coupled feedbacks and the evolution of tropical cyclones. They further stated that a strong atmosphere-ocean coupled response is observed whenever a shallow warm ocean layer is present, and a weak response is observed in the presence of a thick warm layer.

Even though considerable progress has been made in forecasting the tropical cyclone track, intensity forecasting, and factors governing it still remains a challenge (Emanuel and Zhang, 2016). The effect of environmental flow (vertical wind shear) on tropical cyclone intensity is far more complicated than previously thought (Zhu et al., 2004). A common explanation for the effect of VWS on tropical cyclone intensity is “ventilation” of the hurricane warm core. Simpson and Riehl (1958) first hypothesised that under sheared conditions, lateral advection of environmental air cools the core, and thus the temperature is lower than the value of a parcel's moist adiabatic ascent from the surface. Multiple studies have also shown the negative impact of VWS on tropical cyclone intensity (Gray, 1968; Zeng et al., 2008, Zeng et al., 2007). Various alternative mechanisms have been proposed in the recent past based on equivalent potential temperature (θ_e) and potential vorticity (PV). Downdrafts outside the eyewall cause drying of the boundary layer underneath (low θ_e), and injection of this low θ_e air from the storm's inflow layer into the eyewall reduces the thermodynamic efficiency as the eyewall becomes less buoyant with respect to the environment (Riemer et al., 2010; Tang and Emanuel, 2010). DeMaria (1996) showed that in the presence of VWS, the PV of the vortex circulation gets tilted in the vertical. As a conservative quantity, this PV tilting requires an increase in the mid-level temperature perturbation near the vortex which inhibits convection and storm development. Thus, rather than ventilation, this effect of VWS on tropical cyclones can also be viewed as an act of “tilting and stabilisation”.

In spite of the negative impact of VWS on tropical cyclone intensity, an analysis of 139 storms over the Atlantic basin confirmed the fact that developing cyclones exhibit VWS (900–200 hPa) of the order of 10 ms⁻¹ (Bracken and Bosart, 2000). Hurricane Jimena (1991) achieved category 4 intensity in 13–20 ms⁻¹ vertical shear, Hurricane Olivia (1994) intensified under 8 ms⁻¹ vertical shear, Hurricane Bonnie's (1998) maintenance phase encountered shear exceeding 15 ms⁻¹, and tropical cyclones Omar (1992) and Steve (1993) could intensify in a sheared environment in excess of 12.5 ms⁻¹ (Black et al., 2002; Elsberry and Jeffries, 1996; Zhu et al., 2004). The complex interplay between downshear convection, downshear vorticity, and interaction of the storm vortex with a new downshear vortex in tropical storm Gabrielle (2001) led to its intensification in the presence of substantial VWS by a process known as downshear reformation (Molinari et al., 2006). Hurricane Bonnie (1998) was documented as having successive episodic convection on the downshear side of the vortex which moved cyclonically around the downshear left of the storm, and a warm-core due to episodic convective cells, along with a balance of VWS and instability helped in the development of supercells which lasts longer than convective cells (Heymsfield et al., 2001; Molinari and Vollaro, 2008). A maximum in downshear updrafts was described in terms of vorticity conservation by Willoughby Jr. and Feinberg (1984). Assuming the vortex to be embedded in the steering flow, they argued that vorticity advection due to the relative motion of the vortex with respect to the steering flow must be balanced by stretching (compression) of vorticity associated with convergence (divergence). Evans et al. (2010) highlighted that in an environment of low inertial stability, the asymmetric flow could favour intensification through the divergent branch of secondary circulation as it faces less resistance to outflow, thus favouring strong updrafts through mass continuity.

As tropical cyclones draw most of their energy from the warm oceans, a series of energy conversions take place to sustain its vigour. The moist air from the planetary boundary layer (PBL) rises and condenses, releasing latent heat in the process. This latent energy (LE) release is stored as an intermediary phase, known as total potential energy (TPE), and the warming thus caused creates a pressure drop below. In an attempt to balance the pressure gradient force, kinetic energy (KE) must increase at the expense of TPE, and it does so via adiabatic cross-isobaric flow (Hogsett and Zhang, 2009; hereafter HZ09). Thus to understand the effect of VWS on storm energetics, two cyclones are studied in the Bay of Bengal (BoB) basin: (1) Phailin (2013), which did not encounter significant shear during its intensification phase, and (2) Hudhud (2014), in which the influence of VWS was prominent. Another objective of this study is to understand the conditions under which it becomes crucial to implement an interactive atmosphere-ocean-wave coupled model rather than its standalone counterpart.

The next section describes the model and the experiments conducted, followed by validation of the simulations with observations. The differences in ocean coupling between the experiments are presented in the atmosphere-ocean coupling segment. Quadrant-wise analysis of the storms and storm energetics in a quasi-Lagrangian framework are discussed in sections 5 and 6, respectively. The final section summarises and highlights the findings.