Introduction

Understanding the nature and behaviour of cyclones is essential for effective disaster preparedness and public safety. Investigating the systematic air-sea coupling processes responsible for the formation, intensification, and movement of these storms is crucial. Numerous studies have examined the ocean-atmosphere interactions that lead to the development of tropical cyclones across various regions of the tropical belt1,2.

Fundamental prerequisites for cyclone formation, as identified by Palmen and Grey include the Coriolis force, low-level relative vorticity, low vertical wind shear, and ocean temperatures exceeding 26 °C to a depth of 60 m, along with variations in equivalent potential temperature from the surface to 500 mb and mid-tropospheric relative humidity greater than 60–70%3,4,5,6. These storms typically form over tropical oceans, driven primarily by heat transfer from the ocean. Emanuel described tropical cyclone intensity as analogous to a Carnot heat engine, where intensity increases rapidly with rising sea surface temperature (SST), with a critical threshold of 26 °C1. Kleinschmidt emphasized the critical role of latent heat released from the ocean as the energy source for typhoons7. The upper ocean’s influence on cyclone intensification is dominant, particularly concerning the dynamics of heat fluxes8. Therefore, understanding changes in upper ocean heat content is vital for assessing cyclone intensity.

Several factors influence upper ocean heat content, including freshwater influx and oceanic eddies. During the monsoon and post-monsoon seasons, heavy rainfall across the Indian subcontinent typically results in substantial freshwater influx into the Bay of Bengal9,10,11. Major rivers such as the Ganges-Brahmaputra, Irrawaddy, Mahanadi, Godavari, Krishna, and Kaveri contribute significant freshwater from August to November12. This influx creates high stratification in the surface layers of the bay, promoting cyclone intensification by reducing vertical mixing13,14,15. Furthermore, freshwater flux eddies generated by the East Indian Coastal Current (EICC) significantly affect cyclone intensification and dissipation over the Bay of Bengal16,17,18,19,20.

The rise in greenhouse gases leads to atmospheric warming and extreme precipitation events21,22,23,24 all contributing to enhanced stratification and increased ocean heat content25,26. The ocean absorbs 90% of the Earth’s excess heat from global warming, positioning it as a critical indicator of climate change.

While the effects of monsoon-related freshwater influx on cyclone intensification have been extensively studied, there is a gap in understanding the role of pre-monsoon freshwater inputs, particularly those from spring discharge fromthe Ganges-Brahmaputra rivers, on cyclone behaviour. The impact of salinity-induced stratification caused by these freshwater inputs is a crucial but often overlooked aspect of cyclone intensification, especially in the northern Bay of Bengal. This study aims to fill this gap by investigating how the salinity stratification resulting from springtime freshwater contributions from major rivers influences the development and intensification of pre-monsoon cyclones.

In this study, we analyzed the intensification of Cyclone Yaas in the northern Bay of Bengal.

Cyclone Yaas formed as a low-pressure area over the east-central Bay of Bengal on the morning of 22 May. It moved in a north-northwesterly direction and intensified into a Very Severe Cyclonic Storm by 25 May. The cyclone crossed the coast approximately 20 km south of Balasore, Odisha, on 26 May, with maximum sustained wind speeds of 130–140 km/h (IMD).

Cyclone Amphan and Cyclone Mocha, both intense tropical cyclones, originated over the Bay of Bengal and had significant impacts on the surrounding regions. Amphan formed as a low-pressure area over the southeast Bay of Bengal on 16 May 2020, rapidly intensified into a Super Cyclonic Storm by 18 May, and made landfall near Bakkhali, West Bengal, on 20 May, with wind speeds of 155–165 km/h. Cyclone Mocha developed as a low-pressure area over the southeast Bay of Bengal on 8 May 2023, intensified into an Extremely Severe Cyclonic Storm by 13 May, and made landfall near Sittwe, Myanmar, on 14 May, with sustained wind speeds of 150–160 km/h (IMD).

Data and methods

This study employs an integrative approach, combining a blend of observational datasets and advanced numerical models to investigate the intensification of Cyclone Yaas in the northern Bay of Bengal. The primary objective is to capture the interaction between atmospheric and oceanic processes that contributed to the cyclone’s significant intensification, with a particular focus on the pre monsoon season.

The cyclone track data and regional rainfall departure were obtained from the India Meteorological Department (IMD) (www.imd.gov.in), providing critical information on the storm’s movement and associated rainfall anomalies. For upper ocean dynamics, data was retrieved from the Copernicus Marine Service, which offers a multi-dimensional, 3D observational dataset of temperature and salinity with a spatial resolution of 0.25° x 0.25°. This dataset was instrumental in calculating ocean stratification levels, including parameters like the Brunt-Väisälä frequency (N₀ = 3 cph, Nmax), as well as analyzing vertical sections along the cyclone track to assess the ocean’s thermal structure.

$$\ \text{Stratification\,Factor} \, S = \sqrt {\frac{{N_{{\max }} }}{{N_{0} }}};$$

Where N0 = 3 cph;

$${\text{N}}_{{\max }} = \sqrt { - \frac{g}{\rho }\frac{{d\rho }}{{dZ}}}$$

One of the important parameters of this study is the Tropical Cyclone Heat Potential (TCHP) and the heat content of the top 30 m of the ocean, which provide a more detailed view of the thermal energy available to fuel cyclone intensification. The TCHP was calculated using the following equation:

$$\text{Tropical cyclone heat potential} = \rho C_{p} \int\limits_{0}^{{d26}} {(T - 26)dz}$$

where ρ is the density of seawater, CP is the specific heat capacity, T is temperature and D26 is the depth of the 26 °C isotherm. Similarly, the heat content of the top 30 m was computed to assess the thermal structure of the upper ocean.

$$\text{Heat content of top}\,30 \,\text{m}=\rho C_{p} \int\limits_{0}^{{30}} {(T)dz}$$

In addition to temperature and salinity, the analysis incorporated surface currents, heat fluxes, chlorophyll and wind stress. Surface currents were derived from COPERNICUS-GLOBCURRENT altimetric geostrophic currents data with a 0.25° x 0.25° resolution. Heat fluxes such as shortwave radiation, net longwave radiation, latent heat flux and sensible heat flux were obtained from ERA 5 data. Positive values represented incoming fluxes, while negative values indicated outgoing fluxes, crucial for understanding the energy exchanges at the air-sea interface. Chlorophyll concentration was extracted from MODIS ocean color data, while wind stress was derived from scatterometer data from Copernicus Marine Services, also with a 0.25° x 0.25° resolution.

A significant part of the study focuses on the spatial distribution of surface salinity, a key factor in cyclone intensification, especially in the northern Bay of Bengal, where river runoff and freshwater influxes are prevalent. Surface salinity data was sourced from the Soil Moisture Active Passive (SMAP) satellite, offering 0.25° x 0.25° resolution.

Additionally, time-series data from the Hybrid Coordinate Ocean Model (HYCOM) and Navy Coupled Ocean Data Assimilation (NCODA) were used to analyze ocean dynamics over time. NCODA incorporates a variety of observational data, including satellite altimeter observations, sea surface temperature (SST) and in-situ temperature and salinity profiles from XBTs (Expendable Bathythermographs), Argo floats and moored buoys. The NCODA system projects surface observations into the deeper ocean layers using a method called Improved Synthetic Ocean Profiles27, allowing for a detailed representation of the water column, even when direct measurements are unavailable.

The importance of this study lies in its ability to advance the understanding of cyclone-ocean interactions during the pre-monsoon season through an integrative approach.

Results

We analysed upper ocean variability during the cyclone’s lifecycle from 23rd to 28th May 2021, using in-situ datasets and satellite observations. Cyclone Yaas formed as a depression over the east-central Bay of Bengal on 23 May and intensified into a severe cyclonic storm as it moved northwest (Fig. 1a). The ocean’s upper 100 m layer is critical for cyclone dynamics, particularly in identifying the factors contributing to cyclone intensification. Before Cyclone Yaas, SSTs reached 31–32 °C in the north-central Bay of Bengal, approximately 2 °C higher than the climatological average (Fig. 1c). Additionally, salinity values were 2–3 PSU lower than climatological averages (Fig. 1e, f), with the northern Bay recording salinities between 29 and 30 PSU and the central Bay at 30–33 PSU (Fig. 1d).

The northward flow of the East Indian Coastal Current (EICC), coupled with cold and warm core eddies before Cyclone Yaas, further modulated these conditions (Fig. 1b). Low salinities, high SSTs and warm core eddies promoted conditions facilitating the cyclone’s intensification. Freshwater influx led to strong stratification in the upper layers, preventing mixing and thus trapping heat. The stratification observed over the north-central Bay of Bengal just before Cyclone Yaas was substantially higher than climatological means for the region (Fig. 1g, h).

Fig. 1
figure 1

a Yaas cyclone track; along track and spatial variability of b Currents; c, d Temperature (°C) before cyclone Yaas (19/05/2021) and climatology respectively; e, f Salinity (psu) before cyclone Yaas (19/05/2021) and climatology respectively; g, h Stratification before cyclone Yaas (19/05/2021) and climatology.

Full size image

High stratification in the Bay of Bengal is typically associated with monsoon runoff during the post-monsoon season. However, in this case, low salinity was already present in April–May, likely resulting from freshwater influx. Rivers such as the Ganges, Brahmaputra, and Irrawaddy, fed by spring and early summer hydrological processes, contribute to this freshwater input28. Despite the occurrence of a double-dip La Niña in 202129, the year was the seventh warmest on record30.

During premonsoon season, the main contributors to freshwater input in the central and northern regions of the Bay of Bengal are seasonal river runoff and pre-monsoon rainfall, with cyclones and related rainfall playing a secondary role31,32. In 2021, despite an overall surplus in annual rainfall, there was a deficit in pre-monsoon rainfall compared to normal levels in the eastern and northeastern parts of India (IMD) (Fig. 2a). Since there were no cyclonic events prior to Cyclone Yaas, river runoff remained the primary source of freshwater, as reflected in the salinity data (Fig. 2a, b).

Cyclones in the Bay of Bengal are heavily influenced by elevated sea surface temperatures (SSTs) during the pre-monsoon season and reduced salinity during the post-monsoon season. However, the ongoing effects of global warming have disrupted these natural processes altering seasonal runoff patterns and significantly influencing oceanic and atmospheric coupling21,22,23,24.

This pattern was also observed during the spring of 2020 (Cyclone Amphan) and 2023 (Cyclone Mocha), as shown in Graphs 2 b, c. Both Cyclones Amphan (2020) and Mocha (2023) intensified under similar conditions. Time series data for the central Bay of Bengal illustrate the lowest surface salinity values and elevated sea surface temperatures during the pre-monsoon season of 2020, 2021 and 2023 (Fig. 2b, c), highlighting the impact of fresh water influx on the upper ocean’s thermal and haline structures.

Low salinity waters transported from north to south along the eastern Bay formed strong vertical salinity stratification, trapping heat in the surface layers and preventing mixing with the deeper ocean. This pattern was observed in all three cyclonic events, Amphan (2020), Yaas (2021), and Mocha (2023) (Fig. 2d, e, f ).

In case of cyclone Yaas, a warm core eddy observed on 24th May in the central Bay further amplified this effect by contributing to the intensification of Cyclone Yaas (Fig. 1b). The ocean’s heat content is crucial in intensifying cyclones, providing positive feedback. Warm ocean conditions assist in the ascent of moist air, which condenses into clouds and releases latent heat energy that the cyclone utilizes for further intensification. Compared to the climatological stratification over the region, the stratification just before Cyclone Yaas in May 2021 (Fig. 1g, h) is considerably higher (> 2). This stratification hindered vertical mixing, leading to heat retention in the upper layers, thus fuelling cyclone development. For Cyclones Amphan (2020) and Mocha (2023), the low-salinity waters caused by freshwater influx is the primary drivers of stratification and intensification in the region, negating the need for additional detailed discussion on this matter.

Fig. 2
figure 2

a Percentage departure in spring rainfall over north and northeastern parts of India. Time series surface b salinity and c temperature data at a location (88.3o E; 15.3o N ) over the northern Bay of Bengal from HYCOM data. Spatial variation of sea surface salinity from SMAP and ARMOR data during d April 2020 (Before Amphan), e April 2021 (before Yaas), f April 2023 (before Mocha).

Full size image
Fig. 3
figure 3

a Depth of 26oC isotherm b Tropical cyclone Heat Potential (W/m2) c Heat content in the top 30 m (W/m2) d. Temperature vertical section along the track (south to north). e Salinity vertical section along the track during 19th May, 2021.

Full size image

The 26 °C isotherm lies at a depth of approximately 120 m along the cyclone’s track, particularly in the central and western regions of the Bay of Bengal (Fig. 3a). Along this path, the tropical cyclone heat potential exceeds 120 W/m² and reaches up to 200 W/m² in the western Bay (Fig. 3b, c). Moreover, the heat content within the upper 30 m exceeds 400 W/m², making it higher than any other area in the Bay along the cyclone’s track (Fig. 3f). Temperature and salinity cross-sections along the track reveal elevated temperatures, with values exceeding 30 °C at depths of 25–30 m, alongside the intrusion of low-salinity waters moving from north to south in the upper layers (Fig. 3d, e). This leads to significant stratification in the surface layers of the northern Bay of Bengal, where maximum mixing is confined to depths of 10 m or less in some areas.

The net surface heat flux before Cyclone Yaas, calculated from ERA 5 data, shows an incoming shortwave flux of 250 W/m² and outgoing longwave radiation, latent heat flux, and sensible heat flux are approximately − 50 W/m², -120 W/m², and − 5 W/m² respectively, resulting in a net heat flux of 390 W/m² (Fig. 4a–e). This net heat flux nearly coincides with the heat content in the top 30 m of the ocean, particularly in the north-central and western Bay, where freshwater influx and stratification limited mixing and cyclone-induced upwelling. High humidity, elevated surface fluxes and low wind shear, creating conditions that are favorable for cyclone intensification (Fig. 4f, g). In this study, the influence of oceanic factors was dominant, exerting significant control over the atmospheric parameters. Notably, phenomena such as the Madden-Julian Oscillation (MJO) were inactive during the intensification of Cyclone Yaas, with the RMM index remaining below 1, as observed from both ECMWF and BOM datasets.

Following Cyclone Yaas’s passage on 29th May 2021, satellite data revealed an increase in chlorophyll concentrations (Fig. 5a, b) at locations with the highest wind stress and currents, which were situated away from the cyclone’s center (Fig. 5c). The chlorophyll increase, approximately 0.75 mg/m³, suggests upwelling and nutrient enrichment at these locations.

Fig. 4
figure 4

Surface fluxes of a Short wave (W/m2). b Longwave (W/m2). c Latent heat (W/m2). d Sensible Heat (W/m2). e Net heat (W/m2) f Specific humidity (kg/kg) g Wind shear (m/s) over the Bay of Bengal before cyclone Yaas i.e. on 19th May, 2021.

Full size image
Fig. 5
figure 5

a Altimetry geostrophic currents (m/s) on 25th May, 2021. b East ward Wind stress (Nm− 2) c Chlorophyll (mg/m3).

Full size image

Conclusions

This study emphasizes the key role of pre-monsoon freshwater influx and upper ocean stratification in the intensification of Cyclone Yaas (May 2021) over the northern Bay of Bengal. Enhanced river discharge from the Ganges-Brahmaputra system led to reduced surface salinity, resulting in strong haline stratification that suppressed vertical mixing and allowed significant heat entrapment in the upper 30 m. Sea surface temperatures exceeded 31.5 °C nearby 2 °C above climatological norms providing abundant thermal energy to fuel the cyclone. The presence of a warm-core eddy and a strong EICC further elevated ocean heat content by deepening the thermocline. The stratification also reduced chlorophyll concentrations, indicating suppressed vertical nutrient transport and a passive biological response along the cyclone track.

Similar preconditioning factors were evident during Cyclone Amphan (2020) and Cyclone Mocha (2023), both of which exhibited rapid intensification phases supported by enhanced stratification and anomalously warm surface waters. These findings reveal the thermodynamic advantage offered by pre-existing stratified layers, which enhance cyclone intensity by allowing the storm to draw latent heat from the warm surface layer without cooling it through mixing. The role of pre-monsoon stratification, traditionally underemphasized, emerges as a critical factor in cyclone dynamics. Importantly, the influx of freshwater into the Bay of Bengal during the pre-monsoon period may also be influenced by accelerated snowmelt in the Himalayan region due to changing climatic conditions, a linkage that requires further detailed investigation. This study also suggests that climate-driven changes in early runoff patterns may intensify future pre-monsoon cyclones. It highlights the need to incorporate freshwater dynamics and upper ocean stratification into forecasting models for better cyclone intensity prediction. Improved understanding of these pre-conditioning factors can significantly enhance disaster preparedness and coastal resilience.