Air–Sea Interaction During Ocean Frontal Passage: A Case Study from the Northern South China Sea

Abstract

The northern South China Sea has abundant frontal systems near coastal and island regions, which play crucial roles in regional ocean dynamics and ecosystem. While previous studies have established preliminary understanding of their spatial distribution, seasonal variability, and dynamic characteristics, the atmospheric response to these frontal systems remains poorly understood. This study integrates observations from a moored buoy deployed on the continental shelf of the South China Sea with satellite remote sensing data to analyze oceanic and atmospheric variations during frontal passage. The results reveal that the ocean front can not only induce pronounced oceanic changes characterized by significant cooling, saltiness, and surface current acceleration, but also exert substantial influence on the overlying atmosphere, with consistent decreasing trends in air temperature, humidity, and atmospheric pressure, all of which rapidly recovered following frontal retreat. Notably, when the front directly traversed the buoy location, diurnal temperature cycles were markedly suppressed, while turbulent heat flux and downfront wind-stress curl reached peak magnitudes. These findings demonstrate that ocean fronts and associated sea surface temperature gradients can trigger intense air–sea exchange processes at the ocean–atmosphere interface.

1. Introduction

The oceanic front is defined as narrow transitional zones marking the boundaries between distinct water masses with contrasting physical, chemical, and biological properties [1,2,3]. These frontal systems have spatial scales ranging from O(1) to O(1000) km and serve as regions of intense dynamical activity where energy transfers strongly from large-scale to small-scale motions through processes including frontal instabilities, internal wave generation, and intense turbulent mixing [4,5,6]. The enhanced vertical circulation associated with frontal dynamics facilitates the upward transport of heat and material from ocean interior into surface layer, creating conditions that significantly enhance local biological productivity [2,7,8].

Beyond their oceanic impacts, frontal systems can also exert considerable influence on atmospheric boundary-layer processes through multiple mechanisms [9,10,11,12,13]. Based on satellite observations and model simulations, ocean mesoscale fronts and eddies have been suggested to drive substantial variabilities in local surface wind [14,15,16,17], precipitation [13,18], and cloud formation [14,19]. The impacts of these fronts are not confined to the local area and can propagate to remote regions by forcing planetary waves [20]. The mechanisms for this air–sea interaction typically involve changes in the atmospheric vertical mixing or surface pressure above a strong sea surface temperature (SST) gradient in the frontal region [9,10,16,21]. Submesoscale fronts are also associated with a strong SST gradient and could potentially drive significant changes in the atmosphere, especially in the marine atmospheric boundary layer (MABL) [22,23].

The South China Sea (SCS) is the largest marginal sea in the Western Pacific Ocean, encompassing a complex oceanographic environment [24,25,26,27,28,29,30]. This semi-enclosed basin serves as a critical component of the global ocean circulation system, where multiple oceanic and atmospheric forcing mechanisms converge to create intricate patterns of water mass distribution and frontal dynamics [31,32,33,34,35]. The northern region of the South China Sea exhibits typical frontal systems that demonstrate significant spatiotemporal variability under the influence of multiple physical processes. These include monsoon forcing, tidal mixing, river plume dynamics, coastal upwelling, and topographic interactions [36,37,38,39]. The interplay of these diverse mechanisms results in the formation of distinct water masses with contrasting thermohaline properties along coastal regions and in the vicinity of islands. The winter frontal formation is predominantly driven by the interaction between two distinct water masses: the cooler Guangdong coastal current induced by the northeast monsoon and the warmer waters transported shoreward through Ekman transport mechanisms [40,41,42]. Conversely, during the summer monsoon season, frontal activity diminishes significantly. The southwest monsoon generates coastal upwelling along the Vietnamese coast and the eastern side of Hainan Island, bringing cooler, nutrient-rich subsurface waters to the surface and forming oceanic fronts [36,42].

Despite considerable advances in understanding the spatial distribution, seasonal variability, and dynamical characteristics of these coastal frontal systems, a significant knowledge gap remains regarding their potential impact on atmospheric processes in marginal sea environments. Previous satellite observations and reanalysis data have revealed statistically significant positive SST–wind correlations in the northern South China Sea coastal region, with coupling coefficients that exceed those observed at mid-latitude oceans but remain smaller than equatorial regions [43]. Oceanic fronts can induce baroclinic adjustment of perturbation pressure within the marine atmospheric boundary layer, with pressure gradient adjustments representing the dominant term in horizontal momentum budgets over frontal zones [44]. However, the lack of high-frequency observational data has hindered our ability to further investigate the instantaneous variations in atmospheric responses to these fontal processes. In this study, we use observational data collected from a moored buoy system deployed along the eastern coast of Hainan Island to investigate the influence of oceanic frontal systems on air–sea interaction. The in situ measurements from hydrographic instrumentation were combined with satellite remote sensing data to successfully identify and characterize a frontal passage event. By examining atmospheric changes occurring before and after the frontal passage, it is found that the oceanic front can influence local atmospheric conditions and contribute to air–sea interaction processes in complex coastal environments.

The rest of this paper is organized as follows. In Section 2, the buoy and satellite observations are introduced, including detailed descriptions of instrumentation, data processing methodologies, and quality control procedures. In Section 3, the frontal passage process and its significant atmospheric impacts are described. The conclusion and discussion are presented in Section 4.

2. Materials and Methods

2.1. Buoy Observation

On 10 April 2025, a buoy system was deployed at a water depth of ~105 m, approximately 50 km east of Hainan Province, China (Figure 1). The buoy configuration was designed to capture both atmospheric and oceanic parameters across multiple depths to facilitate detailed analysis of air–sea interaction processes. The atmospheric measurement suite was positioned 10 m above the sea surface and comprised sensors for temperature, humidity, atmospheric pressure, and wind speed. The sample interval of all sensors was set to one minute. To minimize measurement uncertainties and random noise effects, all meteorological data were subsequently processed to hourly averages. Below the sea surface, five Conductivity–Temperature–Depth (CTD) sensors and five single-point current meters were positioned at depths of 5 m, 20 m, 30 m, 50 m, and 75 m with a sampling interval of one hour. Unfortunately, several technical issues affected data availability during the deployment period. The CTD at 30 m depth experienced an unidentified malfunction prior to 25 April, resulting in the absence of valid measurements from this level. Additionally, the current meter deployed at 75 m depth failed shortly after deployment. After applying a 24 h sliding window analysis for each measured variable, all observations exceeding 3 standard deviations from the local mean within each window were flagged as anomalous values and subsequently removed from the dataset. Finally, the remaining 65% of dataset extended from 10 April to 10 May are used in this study.

2.2. Satellite Observation

To clarify the regional oceanic conditions surrounding the deployment site, SST fields from the Operational Sea Surface Temperature and Sea Ice Analysis (OSTIA) product were incorporated here. The OSTIA dataset provides a global, daily, gap-free foundation SST analysis at a high horizontal resolution of 0.05° on a regular grid, providing sufficient detail to resolve small-scare oceanic features such as fronts and eddies in the vicinity of coastal area. This level-4 (L4) product, covering the period from 1 October 1981, integrates multiple satellite observations alongside in situ measurements. The processing chain involves extracting and quality-controlling observations, applying satellite bias adjustments, assimilating the corrected inputs along with reprocessed sea ice concentration data, and performing an objective analysis. Each daily analysis incorporates data from a three-day window centered on the analysis date, with reduced weighting for surrounding days, resulting in global maps of analyzed SST and its uncertainty field. It should be noted that the three-day temporal averaging may dampen small-scale and short-lived oceanic events, such as frontal passages. This temporal smoothing acts as a low-pass filter, potentially reducing the amplitude and sharpness of rapid SST changes compared to point measurements from buoys or other high-frequency observations. Additionally, hourly wind-field data from scatterometer observations and the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis were utilized to characterize the wind variability in the vicinity of the buoy deployment site. The following discussion is within the domain of 17.5–21.5°N and 109.5–113.5°E.

2.3. Numerical Model Simulation

To further investigate the atmospheric response to ocean fronts, we conducted high-resolution numerical simulations using the Weather Research and Forecasting Model (WRF) version 3.4, developed by the National Center for Atmospheric Research. The model domain was configured over 18.5–21.5°N and 115–118°E, with a horizontal resolution of 1 km and 30 vertical layers (Figure 2). Initial and boundary conditions were derived from ERA5 reanalysis data, temporally and spatially averaged over the corresponding study region during April 2025. For the bottom boundary conditions, we prescribed an idealized north–south oriented sea surface temperature (SST) front, with temperatures of 21 °C and 23 °C on the western and eastern sides, respectively. The transition zone was set to 4 km width, corresponding to a thermal gradient intensity of 5 × 10⁻⁴ °C m⁻¹. Initial wind conditions were specified as southerly flow. The model was integrated with hourly output frequency, reaching steady states after approximately three days. Our analysis focuses on the averaged results from the subsequent five-day period to ensure statistical robustness.

3. Results

3.1. Moored Buoy and Satellite Observations of an Oceanic Front

During the one-month observation period from 10 April to 10 May, the CTD measurements from the buoy reveal distinct vertical stratification and temporal variability of upper 75 m of water mass (Figure 3). Prior to 12 April, the potential temperature values remained relatively stable around 22–25 °C across all depths, while salinity showed minimal variation around 34.2–34.4 psu, and potential density anomaly (σ_θ) maintained consistent values near 23.8–24.0 kg m⁻³, indicating a well-mixed condition in the whole water column. Subsequently, the water temperature at 15 m progressively increased from approximately 24 °C in mid-April to nearly 29 °C by 10 May, representing a remarkable 5 °C warming over less than one month (Figure 3a). Concurrently, a significant freshening process occurred, with salinity decreasing from 34.3 psu to approximately 33.0 psu (Figure 3b). In contrast, the temperature and salinity at 50–75 m remained consistently around 22–23 °C and 34.5 psu, respectively, until the onset of fluctuations in late April, which suggests the development of a strong thermocline that effectively isolated the surface mixed layer from deeper waters. The combined effects of surface warming and freshening resulted in a pronounced decrease in σ_θ at shallow depths, declining from approximately 23.8 kg m⁻³ to values approaching 21.0 kg m⁻³ by early May (Figure 3c). The resulting density stratification represents a classic example of ocean stratification, transitioning from the well-mixed winter conditions to strongly stratified summer conditions.

To eliminate temperature and salinity disturbances caused by vertical motions of CTDs on the buoy, we interpolate all variables onto two reference depths (13 m and 68 m) (Figure 4). These two depths are obtained by averaging the time-averaged pressure values from the upper two CTDs and lower two CTDs, respectively. This approach minimizes the difference between the measured pressure and the chosen constant reference depths [45]. Beyond the overall warming trend and stratification processes, the temperature time series clearly show a pronounced cooling event occurring around 25 April, characterized by temperature decreases of approximately 2 °C at both depths within less than two days. Accompanying the temperature anomaly, significant changes in salinity and density patterns, with their values at 13 m increased by 1 psu and 1 kg m⁻³, respectively. At the 68 m, salinity and density also increased, though the magnitude was considerably smaller, to only 20–30% of the 13 m values. After 2 days, all variables at both depth levels began to return toward their pre-event levels, implying that this abrupt and short-lived cooling event may be not caused by a simple surface cooling process such as atmospheric heat flux modification, but rather involved the advection or upwelling of cooler water masses from laterally or below into the observation area.

The current velocity measurements from the upper 50 m also show a significant flow change that preceded the cooling event by approximately two days (Figure 5). Throughout most of the observation period, the current magnitudes remained relatively consistent across all depth levels, exhibiting a gradual strengthening trend with predominantly southward flow directions. However, on 21 April, the 15 m current speed increases from approximately 0.5 m/s to nearly 1.0 m/s within 25 h, representing an acceleration rate that far exceeded the background trend. Noted that the flow rotates from southward to eastward direction during the acceleration process, which provides crucial insight into the potential source region of the anomalous water mass, suggesting that the cooler, more saline waters originated from the continental shelf or nearshore coastal areas to the buoy location. Similarly to the rapid return of water properties, the current velocities drop back to the previous levels by 25 April, confirming that the entire sequence of events represents a coherent, short-lived perturbation to the local ocean state. Only energetic mesoscale or smaller-scale features such as oceanic fronts or eddies possess the capacity to generate such dramatic and rapid changes in both the velocity field and water mass properties.

An examination of satellite-derived SST data provides a broader spatial perspective on the regional oceanographic conditions during the observation period (Figure 6). Consistent with the CTD measurement, the SST field reveals a coherent warming trend throughout the surrounding region over the one-month period. However, the spatial distribution of SST exhibits a distinct north–south gradient, with temperatures systematically decreasing from south to north due to the presence of coastal currents that transport cooler, saltier waters along the continental margin. When these cooler, saltier coastal waters encounter the warmer, fresher offshore waters characteristic of the South China Sea, the resulting density contrasts create favorable conditions for the development and intensification of frontal systems [42,46]. On 20 April, a well-defined oceanic front oriented in a northeast-southwest direction was formed to approximately 30 km northwest of the buoy deployment site (Figure 6a,d). This front manifests as a sharp gradient in SST, with temperature differences across the frontal zone reaching ~2 °C over distances of only 15–20 km. Subsequently, the intensification of northeasterly winds drives the coastal current system to transport increasingly larger volumes of cold water toward the Hainan Island, leading to a progressive strengthening of the frontal system and its gradual offshore migration (Figure 6b,e). By 25 April, the frontal system reaches its maximum intensity and extends to the buoy location, spanning approximately 200 km in length with an average cross-frontal temperature gradient reaching magnitudes of 2 × 10⁻⁴ °C m⁻¹ (Figure 6e). Following the peak intensity on 25 April, a shift from northeasterly to southeasterly winds leads to a retreat of the frontal system back toward the coast, allowing the warmer offshore waters to reoccupy the buoy location and explaining the subsequent recovery of temperature and salinity values(Figure 6c,f). The satellite observations thus provide compelling evidence that the rapid cooling event, current acceleration, and water mass property changes recorded at the buoy resulted from the passage of this well-defined oceanic front, rather than from local surface cooling or other small-scale processes.

3.2. Atmospheric Response to the Frontal Passage

Although the moored buoy system cannot directly measure the spatial structure of the oceanic front, the satellite-derived evidence of frontal migration toward the buoy location enables us to treat this event as a cross-frontal observation, offering valuable insights into whether oceanic fronts can exert measurable influence on the overlying atmospheric boundary layer. Figure 7 displays the hourly evolution of atmospheric variables from 22 to 25 April during the frontal approach and passage toward the buoy location. The air temperature time series reveal a progressive cooling trend that closely mirrors the underlying SST evolution (Figure 7a), demonstrating strong coupling between oceanic and atmospheric thermal structures. Initially maintaining relatively stable values around 27 °C on 22–23 April, the air temperature begins a systematic decline as the oceanic front approaches the observation site. Most remarkably, the typical diurnal temperature cycle, which remained clearly visible during 22–23 April with amplitude variations of approximately 0.5 °C, becomes virtually eliminated on 24 April, indicating that the oceanic forcing associated with the frontal passage overwhelms the normal solar heating cycle and creates anomalous atmospheric conditions. The air temperature reaches its minimum value of approximately 26.2 °C at 06:00 UTC on 25 April, coinciding with the timing when satellite observations show the oceanic front achieving maximum intensity and closest proximity to the buoy location, thus establishing a clear temporal correlation between oceanic and atmospheric thermal anomalies. Following this minimum, the atmospheric temperature begins a gradual recovery that parallels the retreat of the oceanic front and the subsequent reoccupation of warm waters.

The relative humidity and atmospheric pressure fields (Figure 7b,c) exhibit similarly coherent responses to the frontal passage, but their minimum values occur approximately 15 h earlier than the temperature minimum, at around 15:00 UTC on 24 April. This temporal offset suggests that humidity and pressure respond more rapidly to the initial approach of the cooler, denser air mass associated with the oceanic front. The wind speed had a sustained weakening trend throughout the entire observation period, with a reduction in more than 50% (Figure 7d). Although the wind speed does not recover to its initial values following the frontal passage, a pronounced wind speed perturbation occurs at 20:00 UTC on 24 April. This wind speed anomaly likely represents the instantaneous atmospheric response generated by the sharp horizontal temperature gradients associated with the oceanic front.

The modifications to the air–sea exchange processes across the front are further quantified by the surface heat flux based on COARE 3.0 algorithm [47]. The temporal evolution of sensible heat flux (SHF), latent heat flux (LHF), and turbulent heat flux (THF) from 22 to 25 April demonstrates that the heat exchange is predominantly governed by sensible heat transfer (Figure 8). Before 22–23 April, the turbulent heat fluxes display characteristic diurnal cycling patterns, with THF reaching maximum positive values of approximately 12–13 W m⁻² during nighttime hours around 05:00 UTC when the air–sea temperature differential favors upward heat transfer from the relatively warm ocean surface to the cooler atmospheric boundary layer. The diurnal minima typically occur during daytime hours around 14:00 UTC when solar heating elevates air temperature, resulting in sign reversal of flux values. However, the approach and passage of the oceanic front on 24 April fundamentally disrupts these established heat exchange patterns. The THF maintains persistently negative values, indicating a sustained period of oceanic heat absorption from the atmosphere. The magnitude of this thermal forcing becomes particularly evident when examining the timing and intensity of the daily flux maximum, which undergoes both temporal displacement and substantial amplification during the frontal passage. The trough heat flux on 24 April is delayed by approximately 6 h, not reaching its minimum until around 20:00 UTC on 24 April. This temporal shift coincides with the period of maximum frontal intensity and closest proximity to the observation site. More remarkably, the magnitude of this delayed trough reaches approximately −40 W m⁻², representing nearly a two-fold amplification compared to the typical values of 10–20 W m⁻². This doubling of heat flux intensity underscores the extraordinary efficiency of oceanic fronts in enhancing air–sea exchange processes through their ability to maintain steep temperature gradients and sustained thermal disequilibrium between oceanic and atmospheric boundary layers.

An additional complexity in the heat flux evolution appears at 19:00 UTC on 24 April, precisely coinciding with the sudden wind speed reduction documented in the atmospheric measurements. The wind speed perturbation at this time likely represents the atmospheric boundary-layer response to the passage of the most intense portion of the oceanic front. Given that the oceanic front maintained a relatively consistent northeast-southwest orientation throughout its approach and passage, with satellite analyses revealing an average frontal orientation of 30° east of north during 22–25 April, we projected the measured wind vectors onto this front-relative coordinate system to isolate the downfront and cross-front wind components, thereby enabling a more precise characterization of the atmospheric response to frontal forcing (Figure 9a,b). This coordinate transformation shows that the wind speed reduction manifests predominantly in the downfront component rather than the cross-front component. To quantify the dynamical implications of these wind-field modifications and assess their relationship to oceanic frontal forcing, we computed the temporal derivatives of wind-stress components and interpreted these quantities as approximations for the cross-front components of wind-stress curl and divergence, respectively (Figure 9c,d). The calculated curl component (Figure 9c) reveals a dramatic intensification between 18:00 and 22:00 on 24 April, with magnitude increases of approximately three times of magnitude compared to pre-frontal conditions. Although the values oscillate between positive and negative extremes reflects the limitation of calculating only one component of the wind-stress curl, it is somewhat indicative of a positive correlation between the cross-front SST gradient intensity and curl magnitude, consist with the previous satellite observation studies [17,43,48].

The simulation results provide compelling evidence for the atmospheric response to ocean fronts. Figure 10 shows the cross-frontal SST and atmospheric variations. The air temperature and THF exhibit a sharp transition coinciding with the SST front at x = 0 km, demonstrating the strong coupling between oceanic and atmospheric thermal structures. Although the magnitude of LHF is only half of the observed values, the THF increase is still primarily contributed by LHF rather than SHF. The LHF shows a dramatic peak near the frontal position, reaching approximately −15 W m⁻², while the SHF contribution remains relatively modest throughout the cross-frontal section. Furthermore, the wind-stress curl intensity indeed shows significant amplification at the frontal location, consistent with our observational findings. The downfront wind-stress gradient in the cross-front direction (−∂τ_x/∂y) dominates the wind-stress curl magnitude. This enhanced wind-stress curl generation is attributed to the differential surface roughness and stability conditions across the thermal front. Therefore, these results substantiate that the observed atmospheric anomalies can be attributed to the presence and dynamics of oceanic fronts.

4. Conclusions and Discussion

This study presents a comprehensive analysis of oceanic frontal impacts on air–sea interface processes in the northern South China Sea using in situ buoy observations combined with satellite-derived data. Through a one-month deployment period, we successfully captured and characterized the passage of a well-defined oceanic front and its subsequent effects on both water properties and atmospheric boundary-layer conditions. The main conclusions are summarized as follows:

(a) The buoy recorded a distinct cooling and saltiness event characterized by rapid temperature decreases of approximately 2 °C and salinity increases of 1 psu at 13 m depth. The current velocity measurements revealed a significant flow enhancement occurring two days prior to the cooling event, with speeds accelerating from 0.5 m/s to nearly 1.0 m/s, accompanied by a directional shift from southward to eastward flow. This temporal sequence indicates that oceanic dynamical processes, rather than atmosphere forcing, dominated the change in water mass properties.

(b) Satellite-derived SST fields confirmed that the observed anomalies resulted from the passage of an oceanic front. The frontal system, driven by northeasterly winds, transported cooler, saltier coastal waters offshore, reaching maximum intensity on 25 April with cross-frontal temperature gradients of 2 × 10⁻⁴ °C·m⁻¹ spanning approximately 200 km. The subsequent wind direction changes to southeasterly caused frontal retreat and the restoration of pre-frontal conditions.

(c) The oceanic front exerted measurable influence on the overlying atmospheric boundary layer, manifesting through significant decreases in air temperature, humidity, and atmospheric pressure that closely tracked the underlying SST evolution. Most notably, the passage of the frontal system completely suppressed the normal diurnal cycle of air temperature on 24 April, indicating that oceanic forcing overwhelmed solar heating effects. The enhancement of turbulent heat flux and downfront wind stress curl collectively demonstrates the capacity of oceanic fronts to significantly modify air–sea exchange processes.

While our single-point moored buoy observations detected the temporal evolution of frontal passage, the inherently limited spatial coverage constrains our ability to fully characterize the three-dimensional structure and cross-frontal gradients of the oceanic front. The observed atmospheric responses also represent point measurements that may not accurately reflect the spatial heterogeneity of air–sea exchange processes across the frontal zone. Future investigations would benefit from spatially distributed observational arrays or cross-frontal transects to better resolve the horizontal scale dependence of frontal impacts on atmospheric boundary-layer dynamics. In addition, the observed frontal event occurred during the spring-to-summer transition period, when the South China Sea experiences significant changes in monsoon patterns and thermal stratification [40,42,43]. The relationship between frontal intensity and seasonal atmospheric forcing remains unclear from our limited temporal dataset.