Three-Dimensional Heterogeneity of Salinity Extremes Modulated by Mesoscale Eddies Around the Hawaiian Islands

Abstract

What are the main findings?

Mesoscale eddies modulate salinity extremes with a depth-reversed pattern which is driven by vertical displacement of the subsurface salinity maximum layer.

High (low) salinity extremes exhibit single-core (dipole) distribution in cyclonic (anticyclonic) eddies, matching the distribution of salinity anomalies.

What is the implication of the main finding?

These findings may advance understanding of ocean salinity dynamics in eddy-rich regions.

The results provide an understanding of how mesoscale processes modulate extreme events.

Salinity extremes (SEs) play a crucial role in marine ecosystems, ocean circulation, and climate variability. Understanding their distribution and drivers is essential for predicting changes in ocean salinity under climate change, particularly in dynamic regions such as the Hawaiian Islands, where mesoscale eddies significantly modulate water mass properties. This study investigates the three-dimensional characteristics of SEs and their responses to mesoscale eddies using mooring observations and sea surface salinity data. We find that high salinity extremes (HSEs) generally occur more frequently than low salinity extremes (LSEs) in the study region, though LSEs exhibit greater duration and intensity. Mesoscale eddies modulate SEs significantly—anticyclonic eddies (AEs) enhance LSEs, whereas cyclonic eddies (CEs) promote HSEs in the upper layer. This relationship reverses in the deeper layer, with AEs favoring HSEs and CEs enhancing LSEs. These opposing effects are driven by a vertical displacement of the subsurface salinity maximum layer, where CEs lift high-salinity subsurface water to the upper ocean via upwelling, creating HSEs in the upper layer and LSEs in the deeper layer, while AEs subduct high-salinity water downward, reducing upper-layer salinity (LSEs) but increasing deeper-layer salinity (HSEs) via downwelling. Spatially, CEs exhibit a single-core high-salinity anomaly, displaced westward by 0.3 times of the eddy radius from the eddy center, with HSEs peaking in frequency and intensity near the core. In contrast, AEs display a dipole salinity anomaly (low northwest/high southeast), aligning with LSE frequency distribution, while HSEs show an inverse pattern. Mooring data further reveal that AE-LSE co-occurrence is highest within 1.2 times of the eddy radius, whereas CE-HSE probability declines with eddy intensity. Notably, AE-HSE and CE-LSE probabilities, though initially weaker, surpass AE-LSE and CE-HSE at certain depths, underlining the complexity of depth-dependent eddy modulation. These findings may advance understanding of ocean salinity dynamics and provide insights into how mesoscale processes modulate extreme events, with implications for marine biogeochemistry and climate modeling.

1. Introduction

Extreme events in the ocean, such as marine heatwaves (MHWs), marine cold spells (MCSs), and salinity extremes (SEs), have profound impacts on marine ecosystems, biogeochemical cycles, and climate systems. MHWs/MCSs can trigger coral bleaching [1,2], alter oceanic heat content [3], and seriously endanger marine ecosystems [4,5,6]. SEs, defined as periods of significant salinity anomalies [7,8], play a critical role in modulating density-driven processes, affecting both regional and global ocean dynamics, such as water mass properties [9,10], stratification [11,12,13], and thermohaline circulation [14,15]. Given their cascading effects on marine environments and climate systems, understanding the drivers and distribution of SEs is essential for predicting how ocean salinity may respond to ongoing climate change, particularly in dynamic regions like the western boundary current and Antarctic circumpolar current, where mesoscale processes dominate ocean variability [16].

Despite their importance, SEs have received less attention compared with temperature extremes, with most studies focusing on large-scale salinity trends rather than episodic anomalies. Ocean circulation, wind-driven processes [17], and freshwater fluxes [18,19] play dominant roles in driving the distribution of salinity in the ocean. Existing research has primarily examined SEs globally [20], which finds that mesoscale eddies drive sea surface SEs globally with peaks in duration of 5 to 10 days and intensities of 0.2 to 0.3 g kg⁻¹. Furthermore, freshwater fluxes and mean currents modulate the occurrence of prolonged and intense sea surface SEs in tropical and extratropical oceans, respectively. Anomalously saline water advected by mean currents contributed to a significant low salinity extreme (LSE) in the eastern North Atlantic subpolar region from 2012 to 2016 [21]. Notably, a pronounced high salinity extreme (HSE) occurred in the upper ocean of the central tropical Indian Ocean from autumn 2010 to spring 2011, influenced by La Niña and a negative Indian Ocean Dipole [22,23]. However, regional-scale SEs, particularly those modulated by mesoscale processes, remain poorly understood.

Similar to the definition of a marine heatwave [24], SEs can be characterized as prolonged deviations from a climatological mean, but no universal threshold exists. An HSE is defined using a percentile threshold. An HSE is defined as a period of at least five consecutive days with daily salinity observations greater than the 90th percentile threshold. Similarly, an LSE is defined as a period of at least five consecutive days with daily salinity observations less than the 10th percentile threshold [8]. Based on this definition, it has been found that MHWs can lead to LSEs, likely through increased precipitation [8]. A study of HSEs and LSEs in the global ocean also found that mesoscale eddies drive most short-lived extremes, while long-term events are linked to freshwater fluxes and mean currents [20].

Mesoscale eddies (50–500 km in diameter) significantly influence ocean salinity through three-dimensional dynamic processes, including vertical pumping, horizontal advection, and turbulent mixing [25,26]. Previous studies have shown that the water within mesoscale eddies differs from the surrounding ambient water [27,28,29,30]. Salinity anomalies occur within eddies because anticyclonic eddies deviate the background salinity profile, while cyclonic eddies uplift it, altering the salinity distribution [26,28]. Furthermore, there is rich geographic and vertical variability of the resultant salinity anomalies induced by eddies of both polarities [31]. Eddies also play a crucial role in global salt transport, exhibiting strong eddy-driven salt fluxes [26,32,33]. Additionally, subsurface eddies disrupt thermohaline staircases through shear-induced turbulent mixing, overwhelming double-diffusive processes and enhancing vertical salt exchange [34]. These salinity anomalies impact ocean stratification, marine ecosystems, and air–sea interactions, with fresher surface layers in eddies potentially suppressing evaporation and altering regional climate feedbacks [35,36,37,38].

The Hawaiian Islands, situated in the central North Pacific, experience complex currents, intense air–sea interactions, and frequent mesoscale eddy activity [16,39,40]. Additionally, mooring buoy observations below the sea surface are available here. This study focuses on the waters surrounding the Hawaiian Islands (168°W–148°W, 17°N–27°N), characterized by high mean eddy kinetic energy (EKE) and frequent eddy occurrence (Figure 1a–c). Previous research indicated that mesoscale eddies uniquely enhance high-chlorophyll events around the Hawaiian Islands [41]. Regarding salinity, whether these eddies significantly influence SEs, and whether their effects differ between LSEs and HSEs, warrants further investigation.

This study investigates the three-dimensional characteristics of SEs and their modulation by mesoscale eddies around the Hawaiian Islands, combining long-term OceanSITES mooring observations and sea surface salinity (SSS) data. We first quantify the occurrence, duration, and intensity of SEs, then explore their responses to mesoscale eddies across depth layers. Finally, using composite analyses, we reveal the spatial and vertical heterogeneity of eddy-induced salinity anomalies. By clarifying the eddy modulation of SEs, this work aims to advance the understanding of extreme salinity variability in dynamic oceanic regimes.

2. Materials and Methods

2.1. Data

2.1.1. Satellite Data

SSS, with a spatial resolution of 0.125° × 0.125° and daily temporal resolution, is obtained from the Copernicus Multi-observation Global Ocean Sea Surface Salinity and Sea Surface Density product, which integrates multi-source observations from satellites (e.g., SMAP, SMOS) and in situ measurements using a multivariate optimal interpolation algorithm [42]. SSS data are used to interpolate to the mooring buoy location to provide surface layer salinity. The SSS data are also used to statistically analyze the probability and intensity of SEs in mesoscale eddies.

Geostrophic velocity anomalies are accessed from the Global Ocean Gridded L4 Sea Surface Height and Derived Variables Reprocessing dataset in the Copernicus Marine Environment Monitoring Service (CMEMS). This dataset, which has a 0.125° × 0.125° spatial resolution and daily temporal resolution, is employed for calculating EKE. EKE is a vital indicator of mesoscale eddy activities, which is given as $EKE={\displaystyle \frac{1}{2}}{\rho }_0({u^{\prime }}^2+{v^{\prime }}^2)$, where $u^{\prime }$ and $v^{\prime }$ are zonal and meridional geostrophic velocity anomalies, respectively; ${\rho }_0=1025\mathrm{kg}\cdot {\mathrm{m}}^{-3}$ is the reference seawater density [43]. Obviously, EKE is high in the western boundary currents of the North Pacific but relatively low in the open ocean. However, high EKE values are also observed around the Hawaiian Islands (Figure 1b), indicating vigorous mesoscale eddy activity in this region (Figure 1c).

2.1.2. Mooring Buoy Data

To describe the long-term characteristic of SEs around the Hawaiian Islands, we obtained the mooring buoy data of the WHOTS station provided by OceanSITES (Table 1), which is located at 22.75°N, 157.98°W (northeast of Hawaii) and records daily observations of temperature and salinity from 15 August 2004 to 6 June 2024. The mooring buoy is deployed into 16 layers from 7 to 155 m (Table 1).

Data of 7 m depths (less than 50% coverage) and data points with bad quality flags were excluded. More importantly, missing values are filled sequentially using the following methods:

Gaps are filled via linear interpolation between adjacent depths;

Gaps at 155 m, as well as other short-term gaps of two or more consecutive levels (<1 month) that cannot be filled using method 1 are addressed through linear interpolation between the surrounding observations;

Gaps exceeding one month were filled with the climatological mean for that depth.

2.1.3. Mesoscale Eddy Data

To evaluate the response of SEs to mesoscale eddies, we extracted eddy characteristics—including center coordinates, radius, and amplitude—from the Mesoscale Eddy Trajectory Atlas 3.2 (META3.2), which covers the period from 1993 to the present [44].

2.2. Methods

2.2.1. Definition of SEs

This study adapts the methodological framework established by Hobday et al. for marine heatwave definition to identify SEs [24]. HSEs (LSEs) are salinity anomalies that exceed (below) one threshold ($S_{\mathrm{thre}}$), persisting for at least 5 consecutive days. To confirm the threshold, we need to calculate the climatology salinity ($S_{\mathrm{clim}}$) first. $S_{\mathrm{clim}}$ is the average salinity of an 11-day window centered on the target day of the study analysis period, which is given by

(1)$S_{\mathrm{clim}}(t)={\displaystyle \frac{{\displaystyle \sum _{y=y_s}^{y_e}{\displaystyle \sum _{d=t-5}^{t+5}S(y,d)}}}{11\times (y_e-y_s+1)}}$

where $y_s$ and $y_e$ are the start and end years for the climatology calculation, respectively, and $S(y,d)$ is the observed salinity on the day $d$ of the year $y$.

The analysis period used in this study is from 15 August 2004 to 6 June 2024. This time frame represents the overlapping period across all datasets. With an 11-day window across 20 years, 220 samples were used to compute the climatology for each calendar day.

$S_{\mathrm{thre}}$ is the 90th (10th for MCSs) percentile of the anomaly between the observed salinity and the climatology salinity (based on 220 data points).

2.2.2. Categorization of SEs

Similar to the classification of MHWs and MCSs [45], SEs are each divided into four categories—moderate (1–2×), strong (2–3×), severe (3–4×), and extreme (>4×) —according to the threshold represented by multiples of the value indicating the local difference between the climatological mean and the climatological 90th percentile (Figure 2). The expression of threshold is

(2)$S_{i\times }(t)=S_{\mathrm{clim}}(t)+i\times (S_{\mathrm{thre}}(t)-S_{\mathrm{clim}}(t))\hspace{1em}i=1,2,3,4$

where $S_{1\times }$ is equivalent to $S_{\mathrm{thre}}$.

The duration and intensity are two main characteristics of SEs. The duration of a SE is the number of days the observed salinity exceeds the threshold $S_{\mathrm{thre}}$. Then, the intensity is the average of the differences between the observed salinity and the threshold during its duration. Therefore, the intensities of HSEs are always positive, while the intensities of LSEs are negative.

2.2.3. Assessing Impacts of Eddies on SEs

We employ lift (i.e., co-occurrence probability) as a metric to assess the impacts of mesoscale eddies on SEs. For example, the lift value quantifying the relationship between AEs and LSEs is defined as follows:

(3)$Lif{t}_{\mathrm{AE}-\mathrm{LSE}}={\displaystyle \frac{P(\mathrm{LSE}\cap \mathrm{AE})}{P(\mathrm{LSE})\times P(\mathrm{AE})}}$

where P(LSE∩AE) denotes the co-occurrence probability of AEs and LSEs, and P(AE) and P(LSE) represent the occurrence probabilities of the AE and LSE, respectively. Probability is calculated using the frequency with which it occurs. It equals the percentage of total observation days during which SEs, eddies, or SE–eddy co-occurrences are present.

When the lift is greater (less) than 1, which indicates a positive (negative) correlation between AEs and LSEs, implying that AEs promote (suppress) the occurrence of LSEs. The smaller the lift value (i.e., the closer to 0), the stronger the suppressing effect. When the lift is equal to 1, it suggests statistical independence between AEs and LSEs.

2.2.4. Normalization of the Observed Data

To facilitate the study of SEs in different eddies, the distance between the SEs and the eddy center is normalized by the eddy radius [46]. For example, if an LSE is 45 km away from an AE center and the eddy radius is 30 km, the normalized distance is recorded as 1.5R. For the mooring buoy observational data, it is assumed that it is only affected by the nearest eddy [47].

3. Results

3.1. Statistical Characteristics of SEs

This study uses mooring buoy data across 15 observational layers to identify and statistically analyze SEs, complemented by surface salinity data from satellite observations at the buoy site. The occurrence probabilities of HSEs and LSEs are 8.10% and 9.34%, respectively. Although HSEs (572 times) occur more frequently than LSEs (408 times), LSEs have greater intensities (LSEs: −0.39 psu, HSEs: 0.25 psu) and longer durations (LSEs: 26.51 days, HSEs: 16.41 days). Most SEs are in moderate categories, and no extreme SEs are observed. Strong HSEs account for only 2.27% of total events, whereas strong LSEs account for 20.34%, and rare severe LSEs are also observed (Table 2). Notably, the stronger the intensity, the shorter the duration.

3.2. Vertical Distribution of SEs

SEs exhibit distinct vertical distributions, as quantified in Figure 3. HSEs consistently outnumber LSEs across most depths, except at 155 m where a significant reversal occurs—47 LSEs and 33 HSEs occur at this depth (Figure 3a). Notably, LSE occurrences increase markedly below 100 m. Depth-dependent patterns further demonstrate that both HSEs and LSEs decrease monotonically near the surface (0–50 m), stabilize between 50 and 100 m, and diverge below 100 m. This pattern aligns with the annual mean frequency of MHWs and MCSs (1–3 events/year) [48].

We categorize HSEs and LSEs into four types, respectively. The number of moderate HSEs is generally higher than that of moderate LSEs throughout the vertical profile (Figure 3b), but the opposite is true for strong SEs, where LSEs outnumber HSEs (Figure 3c). Severe LSEs are predominantly concentrated around 85 m, with no severe HSEs observed at any depth (Table 2, Figure 3d).

Figure 4 illustrates the depth-dependent characteristics of SEs, revealing the discrepancies between HSEs and LSEs. Regarding duration, LSEs generally persist longer than HSEs at most depths (26.5 days versus 16.4 days on average). However, this pattern undergoes a notable reversal below 120 m—LSE duration decreases markedly (reduce by 38% at 155 m), while HSE duration increases slightly.

For a rigorous comparison between the intensities of HSEs and LSEs, we consider their absolute values. Obviously, the intensities of LSEs (−0.39 psu) are greater than that of HSEs (0.25 psu) at all depths. Vertically, SE intensity remains constant in the mixed layer (0–50 m) and decreases with depth in the subsurface layer (50–155 m). Evidently, both duration and intensity (Figure 4) have opposite patterns with SE frequencies (Figure 3a).

3.3. The Co-Occurrence of Eddies and SEs

In this study, we identified a significant correlation between the occurrence of SEs and mesoscale eddies. Notably, the relationship between AEs, CEs, and SEs exhibits depth-dependent variations (Figure 1d and Figure 5). At a 45 m depth, AEs frequently co-occur with LSEs (Figure 5a), whereas at a 155 m depth, they predominantly associate with HSEs (Figure 5b). That is, mesoscale eddies exhibit opposite effects on SEs in the mixed layer and the subsurface layer.

To reveal the modulation of mesoscale eddies on SEs, we create a composition of mooring observation data under the eddy-centric coordinate. Statistically, 152,962 mesoscale eddies are detected in our study region. Then, mesoscale eddy centers are analyzed using a 1° × 1° longitude–latitude grid framework for statistical quantification. Evidently, frequent AE activities are observed in the south and the north (where the mooring buoy site is located) of the Hawaiian Islands (Figure 6a), while CEs are only active in the south of the Hawaiian Islands (Figure 6b). After compositing the mooring buoy data, we find that most of mooring observations are confined within the region of 2R, especially around the edge of mesoscale eddies (Figure 6c), which ensures sufficient samples for the statistical analysis of eddy–SE co-occurrences.

Figure 7a shows the proportion of SEs co-occurring with mesoscale eddies relative to the total SEs occurrences at different depths, and the larger proportion indicates that mesoscale eddies contribute positively to SEs. Within the upper 75 m, about 60% of LSEs co-occur with AEs, which is 3.89 times greater than that of HSEs. That is, AEs favor (suppress) the development of LSEs (HSEs). Conversely, 21.42% (5.88%) of HSEs (LSEs) co-occur with CEs, indicating that CEs exhibit the opposite pattern with promoting HSEs but inhibiting LSEs. Below 75 m, the promoting effect decreases, and the suppressing effect increases. Subsequently, the influence of mesoscale eddies on SEs reverses. At a 155 m depth, 65.43% of HSEs are observed with AEs, and 49.74% of LSEs co-occur with CEs.

Further, we defined lift to assess co-occurrence probability of mesoscale eddies on SEs. The pattern of lift resembles that of the proportion (Figure 7a,b). The lift of AEs (CEs) on LSEs (HSEs) is about two (1.52) within 100 m (75 m), indicating that AEs (CEs) promote the occurrence of LSEs (HSEs). When depth decreases beyond 100 m (135 m), lifts of CEs-HSEs and AEs-LSEs fall below one, which means CEs (AEs) switch from enhancing to suppressing HSEs (LSEs). That is, CEs stimulate the development of LSEs below 100 m, while AEs stimulate the development of HSEs below 135 m.

Both the depth-dependent proportion and lift demonstrate that mesoscale eddies show depth-reversed modulation on SEs—AEs (CEs) enhance LSEs (HSEs) in the upper layer but suppress them in the deeper layer. However, the underlying mechanisms governing this distribution pattern require climatological salinity analysis, which will be discussed in detail in Section 3.4.

3.4. Modulations of Eddies on SEs

To explain the opposite response of SEs to mesoscale eddies, we calculate the climatological salinity profile within AEs, CEs, and total observations (Figure 7c). Notably, the pronounced subsurface salinity maximum persists regardless of AEs, CEs, and total observations, and their corresponding depths are 135 m, 85 m, and 120 m, respectively. That is, upwelling induced by CEs will uplift the subsurface salinity maximum layer, leading to increased salinity in the upper ocean and decreased salinity in the deeper layers. Therefore, CEs stimulate the generation of HSEs in the upper ocean and LSEs in the deeper layer (Figure 7a,b). Conversely, AE-induced downwelling causes the sinking of the subsurface salinity maximum layer, which triggers decreased (increased) salinity in the upper (deeper) layer. As a result, AEs favor the generation of LSEs (HSEs) in the upper (deeper) layer.

Previous studies suggested that the intensity of the upwelling or downwelling exhibits spatial heterogeneity across different regions of mesoscale eddies [49,50], consequently leading to varying probabilities of SE occurrences. Therefore, we create the composition and normalization of mooring observation data and SSS data. In the study region, CEs exhibit a mononuclear high-salinity anomaly (Figure 8a), with its center displacing westward by 0.3R from the eddy center. In contrast, AEs display a dipole-salinity anomaly pattern characterized by low salinity in the northwest and high salinity in the southeast (Figure 8d).

Within CEs, the occurrence frequency of HSEs has a similar pattern with salinity anomalies, with higher occurrence frequency, which can reach 7.7% (Figure 8b) and a stronger intensity of about 0.284 psu around the high-salinity core (Figure 8c). However, the occurrence frequency of LSEs is lower around the high-salinity core and higher outside CEs (beyond R, Figure A1a, Appendix A), while their intensities are weak around the high-salinity core (Figure A1b). The occurrence frequency at a given point is calculated as the number of days when SEs occur within a small window of 0.2R × 0.2R centered on that point, divided by the total number of eddy observation days.

On the contrary, the occurrence frequency of LSEs within AEs also has a similar dipole pattern with salinity anomalies, with higher occurrence frequencies in the low-salinity core and lower occurrence frequencies in the high-salinity core (Figure 8e). For HSEs, their spatial distribution of occurrence frequencies exhibits inverse patterns compared with LSEs (Figure A1c). However, unlike the occurrence frequency, intensities of LSEs and HSEs exhibit a random distribution (Figure 8f and Figure A1d).

In addition to SSS observations, we conduct standardized statistical averaging of mooring measurement data. AEs and LSEs exhibit a high co-occurrence probability within 1.2R region (except at 155 m, Figure 9a). However, the spatial extent of this high AE-LSE co-occurrence probability gradually decreases with increasing depth. Conversely, AEs and HSEs tend to co-occur preferentially beyond 1.2R (Figure 9b). CEs facilitate the generation of HSEs within the eddy (<1R) at depths shallower than 55 m, whereas they tend to promote the generation of HSEs outside the eddy (>1R) at depths deeper than 55 m (Figure 9d). Therefore, mesoscale eddies modulate SEs with significant three-dimensional heterogeneity, exhibiting both vertical and horizontal spatial variability in their impacts (Figure 8 and Figure 9).

In addition, we assess the modulation of mesoscale eddy intensity on SEs, using the sea-level anomaly (SLA) amplitude of the eddy to represent intensity. In the upper ocean, the co-occurrence probability of AE-LSE increases with eddy intensity, reaching a peak of approximately 23.60% at an intensity of 0.08 m (Figure 10a,c). In contrast, the co-occurrence probability of CE-HSE decreases with increasing intensity (Figure 10b,d). The AE-HSE and CE-LSE co-occurrence probabilities are significantly smaller than AE-LSE and CE-HSE. However, AE-HSE and CE-LSE co-occurrence probabilities are both increased with eddy intensity, which are larger than AE-LSE and CE-LSE (Figure 10e,f), indicating that the modulation of mesoscale eddies and SEs varies with depth.

4. Discussion

Our findings reveal a complex, depth-dependent modulation of SEs by mesoscale eddies around the Hawaiian Islands. Specifically, we demonstrate that AEs promote LSEs in the upper layer but enhance HSEs at depths, whereas CEs exhibit the opposite behavior, favoring HSEs in the upper ocean and LSEs in deeper layers. This depth-reversed pattern is primarily governed by the vertical displacement of the subsurface salinity maximum layer—a feature largely associated with Subtropical Underwater (STUW) or North Pacific Tropical Water (NPTW) [51,52]. AEs induce downwellings that subduct the subsurface salinity maximum layer, inducing lower salinity in the upper ocean and higher salinity at deeper layers. This process facilitates the generation of LSEs in the upper layer and HSEs in the deeper layer. Conversely, CEs drive upwellings that increase upper-ocean salinity while decreasing deeper-layer salinity, thereby enhancing surface HSEs and deep LSEs. However, as shown in Figure 5, not all SEs correspond directly with mesoscale eddies. Figure 7 further demonstrates that many SEs occur independently of mesoscale eddy presence. This discrepancy arises from two key factors: First, the upwelling/downwelling velocities associated with eddy-induced geostrophic convergence/divergence are relatively slow [53,54], likely causing a temporal lag between mesoscale eddies and SEs. Second, other dynamic processes can also drive rapid salinity variations, including freshwater fluxes [55], typhoon-induced entrainment [56], and submesoscale processes [57,58], which may contribute to the generation of SEs (patchiness in Figure A1d).

The vertical distribution characteristics of SEs differ fundamentally from those of MHWs and MCSs. MHWs and MCSs both show increasing occurrences and intensities but decreasing durations with depth, but SE intensity is weakened with depth. Notably, the occurrence/duration relative relationship between HSEs and LSEs reverses with depth—a phenomenon not observed in MHWs/MCSs. These differences likely arise from distinct vertical thermohaline structures—temperature decreases monotonically with depth, whereas salinity typically peaks in the subsurface. However, the observations from the mooring buoy, limited to a 155 m depth, may not fully capture deeper variability patterns, suggesting the need for future studies to use deeper moorings to verify the prevalence and mechanisms of these inversion phenomena [49].

The composite analyses further reveal distinct spatial structures of eddy-induced salinity anomalies. The salinity anomalies associated with AEs exhibit a dipole distribution pattern, with low salinity in the northwest and high salinity in the southeast (Figure 8a), which is consistent with the spatial structures of temperature [59] and chlorophyll [46]. This dipole may arise from the eddy-induced horizontal advection dominated by the interaction between eddy-induced currents and background flow within AEs [60]. Meanwhile, CEs exhibit a single-core high-salinity core displaced westward by approximately 0.3 times the eddy radius, consistent with the nonlinear dynamics and westward propagation of mesoscale eddies [46]. The single-core distribution of CEs may be dominated by the vertical upwelling process. Delcroix et al. proposed that mesoscale eddies at the largest near-surface vertical salinity gradients and largest eddy amplitudes show an single-core SSS anomaly [61]. More energetic eddies are more effective at trapping high-salinity water in the nonlinear cores of the eddies [60].

Notably, HSEs occurred more frequently than LSEs in our study region, particularly in the upper ocean. Although the number of AEs exceeds that of CEs in our study region, the probability of CEs generating HSEs is higher than that of AEs generating LSEs. Consequently, the number of HSEs observed by the mooring is greater than that of LSEs. Furthermore, our study area is situated in the subtropical North Pacific, a region characterized by a mean state of high salinity due to STUW/NPTW [51,52] and persistent net evaporation [62]. In such a saline background state, generating a HSE is thermodynamically and dynamically less demanding.

Previous research has demonstrated significant influences of mesoscale eddies on marine heatwaves and cold spells [63,64,65]. However, these studies primarily characterized eddy features using only sea surface height data. The structure of the mesoscale is ignored while in reality, mesoscale eddies exhibit complex three-dimensional structures with horizontal deformation [66,67] and vertical tilt [68,69]. Nevertheless, the lack of subsurface high-resolution observational data has limited the accurate characterization of three-dimensional eddy structures. Consequently, current data can only reveal statistical patterns of eddy modulation on SEs. To investigate finer eddy impacts on SEs, targeted long-term high-resolution observations are required.

5. Conclusions

This study investigates the distribution of SEs and their responses to mesoscale eddies around the Hawaiian Islands, utilizing long-term mooring observations and SSS datasets. We found that HSEs generally occur more frequently than LSEs overall. For moderate events, their vertical distribution is consistent with the overall patterns. However, the distribution reverses for strong events. In terms of duration and intensity, LSEs typically persist longer and show greater magnitudes than HSEs. Notably, the relative occurrence frequency and duration relationship between HSEs and LSEs reverses in deeper layers. However, the intensities of both types of SEs remain stable in the upper layer and decrease with depth in the deeper layer. Notably, mesoscale eddies exhibit a depth-reversed modulation on SEs, with AEs promoting LSEs and suppressing HSEs in the upper layer, while CEs show the opposite effect. In the deeper layer, this pattern reverses, as AEs begin enhancing HSEs and CEs favor LSEs. The opposite effects of mesoscale eddies on SEs are governed by the vertical displacement of the subsurface salinity maximum layer—CEs (AEs) enhance upper-layer HSEs (LSEs) and deeper-layer LSEs (HSEs) via upwelling (downwelling). After compositing mooring and SSS data, we find mesoscale eddies modulate SEs with three-dimensional heterogeneity. CEs feature a westward-shifted (0.3R) mononuclear high-salinity anomaly, with HSEs showing co-located frequency peaks (7.7%) and a stronger intensity (0.284 psu) near the core. Conversely, AEs exhibit a dipole-salinity structure (low northwest/high southeast) mirrored by LSE frequency distribution, whereas HSEs display inverse spatial patterns. AEs predominantly co-occur with LSEs within 1.2R, though this spatial correlation diminishes with depth, while AEs and HSEs associate beyond 1.2R. CEs promote HSEs inside the eddy (<1R) at the upper layer but shift to favoring HSEs outside (>1R) in the deeper layer. Eddy intensity further influences these patterns: AE-LSE probability strengthens with intensity, whereas CE-HSE probability declines. Notably, AE-HSE and CE-LSE probabilities, though initially weaker, increase with intensity and exceed AE-LSE/CE-HSE at certain depths, highlighting depth-dependent modulation dynamics.

Author Contributions

Conceptualization: Z.Y. and S.L.; methodology: Z.Y., Q.S., and W.Z.; software: S.L.; validation: Z.Y., H.W., and X.G.; formal analysis: S.L. and J.S.; investigation: S.L. and X.G.; resources: Z.Y. and H.G.; data curation: S.L. and J.C.; writing—original draft preparation: S.L.; writing—review and editing: Z.Y., Q.S., H.W., and J.W.; visualization: S.L.; supervision: Z.Y.; project administration: Z.Y. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The SSS data and geostrophic velocity anomaly data can be obtained from the Copernicus Marine Data Store. The SSS data are at https://doi.org/10.48670/moi-00051 (accessed on 15 February 2025), and the geostrophic velocity anomaly data are at https://doi.org/10.48670/moi-00148 (accessed on 15 February 2025). Mooring buoy data from OceanSITES are available at https://dods.ndbc.noaa.gov/oceansites/ (accessed on 15 February 2025). The eddy products, META 3.2DT and META 3.2NRT, were provided by the AVISO at https://www.aviso.altimetry.fr/en/data/products/value-added-products/global-mesoscale-eddy-trajectory-product.html (accessed on 15 February 2025).

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:

SEs	Salinity Extremes
LSEs	Low Salinity Extremes
HSEs	High Salinity Extremes
CEs	Cyclonic Eddies
AEs	Anticyclonic Eddies
MHWs	Marine Heatwaves
MCSs	Marine Cold Spells
EKE	Eddy Kinetic Energy
SSS	Sea Surface Salinity
STUW	Subtropical Underwater
NPTW	North Pacific Tropical Water

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1 Overview of the study area. (a) Location of the study area. The yellow boxes in (a,b) mark the research area. (b) Distribution of the time-mean eddy kinetic energy (EKE) (2004–2022) in the North Pacific Ocean. (c) Mesoscale eddies near the mooring buoy on 7 September 2007. Red solid lines denote anticyclonic eddies (AEs), blue solid lines denote cyclonic eddies (CEs), and the yellow star indicates the mooring buoy location. (d) Salinity anomaly time series at the mooring buoy before and after 7 September 2007. Fill color represents salinity anomalies ($S^{\prime }=S_{\mathrm{obs}}-S_{\mathrm{clim}}$), where $S_{\mathrm{clim}}$ is defined in Equation (1). The gray dashed arrow points to the observation on 7 September (marked by the black dashed line). Blue shading above indicates periods when a CE was centered over the buoy; red shading indicates an AE. The shaded red area identifies high salinity extremes (HSEs), with the red line indicating its range. Blue shading corresponds to low salinity extremes (LSEs).

Figure 2 Definition and classification of SEs. The left shading area represents an HSE, and the right shading area represents an LSE. The dashed line represents the observed salinity, the thick solid line represents the climatology salinity, and thin solid lines represent the threshold for different types of SEs.

Figure 3 Vertical spatial distribution of the number of SEs. (a) The total number of all types of HSEs and LSEs at each observational depth during the observation period, with red representing the number of HSEs and blue representing the number of LSEs. The depths of the observational layers are marked with dots. (b–d) as in (a), but for moderate, strong, and severe SEs, respectively.

Figure 4 Vertical spatial distribution of the average duration and intensity absolute value of SEs. (a) Vertical spatial distribution of the average durations of HSEs and LSEs. (b) Box plot of the intensity absolute value of SEs. The line chart represents the average value of the intensity absolute.

Figure 5 Observations of SEs and eddies. The shaded areas represent the times when the buoy was inside an eddy, with light red indicating AEs and light blue indicating CEs. The shaded area between the lines represents the identified SEs, with blue for LSEs and red for HSEs. The black solid line represents the observed salinity values from the mooring buoy, the gray solid line represents the climatology, the dark red solid line represents the HSE threshold, and the dark blue solid line represents the LSE threshold. (a) At a 45 m depth. (b) At a 155 m depth.

Figure 6 Distribution of data. (a) Distribution of the centers of all AEs used for composite analysis. Each grid cell value represents localized eddy observation numbers within the window. The yellow stars in (a,b) represent the position of the mooring buoy. (b) same as for (a) but for CEs. (c) Distribution of mooring buoy observation values relative to the nearest eddy each day, with different colors indicating the data density of the surrounding data points.

Figure 7 The proportion of simultaneous occurrence of eddies and SEs during SEs, the lift of eddies on SEs, and salinity profiles within eddies. (a) The impact of eddies on the occurrence frequency of SEs. The percentage represents the proportion of SEs observed in eddies. The red dashed line indicates the proportion of time when AEs were observed, the red dotted line shows the proportion of time when AEs were observed during LSEs, and the red solid line represents the proportion of time when AEs were observed during HSEs. The three blue curves correspond to CEs. (b) The lift of eddies on SEs. Lift represents the degree of impact of eddies on SEs. (c) Salinity profiles within eddies. The black solid line represents the average salinity profile. The red dashed line represents the salinity profile within AEs, and the blue dashed line represents the salinity profile within CEs.

Figure 8 Distribution of salinity anomalies, the occurrence frequencies and intensities of sea surface SEs within eddies, with the horizontal and vertical coordinates representing the normalized distance from the eddy center R. The gray dashed circles indicate the radii at one and two times the distance from the center of the eddies. (a) Salinity anomalies within CEs. (b) Occurrence frequencies of HSEs within CEs. (c) Intensities of HSEs within CEs. (d) Salinity anomalies within AEs. (e) Occurrence frequencies of LSEs within AEs. (f) Intensities of LSEs within AEs.

Figure 9 Impact of normalized distance on the occurrence frequencies of SEs at all observation depths, including only data with eddy amplitudes exceeding 5 cm. (a) Variation in the occurrence frequencies of LSEs within AEs. (b) Variation in the occurrence frequencies of HSEs within AEs. (c) Variation in the occurrence frequencies of LSEs within CEs. (d) Variation in the occurrence frequencies of HSEs within CEs.

Figure 10 Impact of eddy amplitude on the occurrence frequency of SEs, including only the data that falls within one radius from the eddy center. (a) Changes in the occurrence frequencies of SEs with increasing AE amplitude at sea surface. (b) Changes in the occurrence of SEs with increasing CE amplitude at sea surface. (c,d) are same as (a,b) but at a 55 m depth. (e,f) are same as (a,b) but at a 155 m depth.

Table 1

Basic information of the WHOTS mooring buoy.

Information	Content
Position	23°N, 158°W
Observed period	15 August 2004–6 June 2024
Observed level	7, 15, 25, 35, 40, 45, 50, 55, 65, 75, 85, 95, 105, 120, 135, 155
Observed elements	Salinity, Temperature
Time resolution	Daily

Table 2

Statistics of SEs.

Category	Number	Percentage of Time	Intensity (psu)	Duration (Days)
HSEs	572	/	0.25	16.41
LSEs	408	/	−0.39	26.51
Moderate HSEs	559	97.73%	0.25	15.90
Strong HSEs	13	2.27%	0.31	3.61
Moderate LSEs	316	77.45%	−0.30	14.05
Strong LSEs	83	20.34%	−0.39	12.41
Severe LSEs	9	2.21%	−0.38	7.11

Appendix A

Figure A1 Distribution of occurrence frequencies, and intensities of SEs near eddies. The gray dashed circles indicate the radii at one and two times the distance from the center of the eddies. (a) Occurrence frequencies of LSEs within CEs. (b) Intensities of LSEs within CEs. (c) Occurrence frequencies of HSEs within AEs. (d) Intensities of HSEs within AEs.

References

1. Doherty, M.L.; Johnson, J.V.; Goodbody-Gringley, G. Widespread Coral Bleaching and Mass Mortality during the 2023–2024 Marine Heatwave in Little Cayman. PLoS ONE; 2025; 20, e0322636. [DOI: https://dx.doi.org/10.1371/journal.pone.0322636]

2. Hughes, T.P.; Kerry, J.T.; Baird, A.H.; Connolly, S.R.; Dietzel, A.; Eakin, C.M.; Heron, S.F.; Hoey, A.S.; Hoogenboom, M.O.; Liu, G. . Global Warming Transforms Coral Reef Assemblages. Nature; 2018; 556, pp. 492-496. [DOI: https://dx.doi.org/10.1038/s41586-018-0041-2] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29670282]

3. Behrens, E.; Fernandez, D.; Sutton, P. Meridional Oceanic Heat Transport Influences Marine Heatwaves in the Tasman Sea on Interannual to Decadal Timescales. Front. Mar. Sci.; 2019; 6, 228. [DOI: https://dx.doi.org/10.3389/fmars.2019.00228]

4. Holbrook, N.J.; Sen Gupta, A.; Oliver, E.C.J.; Hobday, A.J.; Benthuysen, J.A.; Scannell, H.A.; Smale, D.A.; Wernberg, T. Keeping Pace with Marine Heatwaves. Nat. Rev. Earth Environ.; 2020; 1, pp. 482-493. [DOI: https://dx.doi.org/10.1038/s43017-020-0068-4]

5. Smith, K.E.; Burrows, M.T.; Hobday, A.J.; Sen Gupta, A.; Moore, P.J.; Thomsen, M.; Wernberg, T.; Smale, D.A. Socioeconomic Impacts of Marine Heatwaves: Global Issues and Opportunities. Science; 2021; 374, eabj3593. [DOI: https://dx.doi.org/10.1126/science.abj3593]

6. Josey, S.A.; Hirschi, J.J.-M.; Sinha, B.; Duchez, A.; Grist, J.P.; Marsh, R. The Recent Atlantic Cold Anomaly: Causes, Consequences, and Related Phenomena. Annual Review of Marine Science; Carlson, C.A.; Giovannoni, S.J. Annual Reviews: Palo Alto, CA, USA, 2018; Volume 10, pp. 475-501.

7. Malan, N.; Roughan, M.; Hemming, M.; Ingleton, T. Quantifying Coastal Freshwater Extremes during Unprecedented Rainfall Using Long Timeseries Multi-Platform Salinity Observations. Nat. Commun.; 2024; 15, 424. [DOI: https://dx.doi.org/10.1038/s41467-023-44398-2]

8. Liu, H.; Nie, X.; Cui, C.; Wei, Z. Compound Marine Heatwaves and Low Sea Surface Salinity Extremes over the Tropical Pacific Ocean. Environ. Res. Lett.; 2023; 18, 064001. [DOI: https://dx.doi.org/10.1088/1748-9326/acd0c4]

9. Barton, B.I.; Lique, C.; Lenn, Y.-D. Water Mass Properties Derived from Satellite Observations in the Barents Sea. J. Geophys. Res. Ocean.; 2020; 125, e2019JC015449. [DOI: https://dx.doi.org/10.1029/2019JC015449]

10. Carter, L.; Bostock-Lyman, H.; Bowen, M. Chapter 4—Water Masses, Circulation and Change in the Modern Southern Ocean. Antarctic Climate Evolution; 2nd ed. Florindo, F.; Siegert, M.; Santis, L.D.; Naish, T. Elsevier: Amsterdam, The Netherlands, 2022; pp. 165-197. ISBN 978-0-12-819109-5

11. Li, G.; Cheng, L.; Zhu, J.; Trenberth, K.E.; Mann, M.E.; Abraham, J.P. Increasing Ocean Stratification over the Past Half-Century. Nat. Clim. Change; 2020; 10, pp. 1116-1123. [DOI: https://dx.doi.org/10.1038/s41558-020-00918-2]

12. Wang, T.; Geyer, W.R. The Balance of Salinity Variance in a Partially Stratified Estuary: Implications for Exchange Flow, Mixing, and Stratification. J. Phys. Oceanogr.; 2018; 48, pp. 2887-2899. [DOI: https://dx.doi.org/10.1175/JPO-D-18-0032.1]

13. Peng, S.; Wang, Q. Fast Enhancement of the Stratification in the Indian Ocean over the Past 20 Years. J. Clim.; 2024; 37, pp. 2231-2245. [DOI: https://dx.doi.org/10.1175/JCLI-D-23-0255.1]

14. Doos, K.; Nilsson, J.; Nycander, J.; Brodeau, L.; Ballarotta, M. The World Ocean Thermohaline Circulation. J. Phys. Oceanogr.; 2012; 42, pp. 1445-1460. [DOI: https://dx.doi.org/10.1175/JPO-D-11-0163.1]

15. Hátún, H.; Sando, A.B.; Drange, H.; Hansen, B.; Valdimarsson, H. Influence of the Atlantic Subpolar Gyre on the Thermohaline Circulation. Science; 2005; 309, pp. 1841-1844. [DOI: https://dx.doi.org/10.1126/science.1114777] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16166513]

16. Chelton, D.B.; Schlax, M.G.; Samelson, R.M. Global Observations of Nonlinear Mesoscale Eddies. Prog. Oceanogr.; 2011; 91, pp. 167-216. [DOI: https://dx.doi.org/10.1016/j.pocean.2011.01.002] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/40933104]

17. Ponte, R.M.; Vinogradova, N.T. An Assessment of Basic Processes Controlling Mean Surface Salinity over the Global Ocean. Geophys. Res. Lett.; 2016; 43, pp. 7052-7058. [DOI: https://dx.doi.org/10.1002/2016GL069857]

18. Du, Y.; Zhang, Y.; Shi, J. Relationship between Sea Surface Salinity and Ocean Circulation and Climate Change. Sci. China Earth Sci.; 2019; 62, pp. 771-782. [DOI: https://dx.doi.org/10.1007/s11430-018-9276-6]

19. Liu, C.; Liang, X.; Ponte, R.M.; Vinogradova, N.; Wang, O. Vertical Redistribution of Salt and Layered Changes in Global Ocean Salinity. Nat. Commun.; 2019; 10, 3445. [DOI: https://dx.doi.org/10.1038/s41467-019-11436-x]

20. Liu, H.; Xu, T.; Li, D.; Nie, X.; Wei, Z. Sea Surface Salinity Extremes over the Global Ocean. Environ. Res. Lett.; 2023; 18, 114020. [DOI: https://dx.doi.org/10.1088/1748-9326/ad0165]

21. Siddiqui, A.H.; Haine, T.W.N.; Nguyen, A.T.; Buckley, M.W. Controls on Upper Ocean Salinity Variability in the Eastern Subpolar North Atlantic During 1992–2017. J. Geophys. Res. Ocean.; 2024; 129, e2024JC020887. [DOI: https://dx.doi.org/10.1029/2024JC020887]

22. Liu, J.; Wang, D.; Zu, T.; Huang, K.; Zhang, O.Y.W. Either IOD Leading or ENSO Leading Triggers Extreme Thermohaline Events in the Central Tropical Indian Ocean. Clim. Dyn.; 2023; 60, pp. 2113-2129. [DOI: https://dx.doi.org/10.1007/s00382-022-06413-y]

23. Sun, Q.; Du, Y.; Zhang, Y.; Feng, M.; Chowdary, J.S.; Chi, J.; Qiu, S.; Yu, W. Evolution of Sea Surface Salinity Anomalies in the Southwestern Tropical Indian Ocean During 2010–2011 Influenced by a Negative IOD Event. J. Geophys. Res. Ocean.; 2019; 124, pp. 3428-3445. [DOI: https://dx.doi.org/10.1029/2018JC014580]

24. Hobday, A.J.; Alexander, L.V.; Perkins, S.E.; Smale, D.A.; Straub, S.C.; Oliver, E.C.J.; Benthuysen, J.A.; Burrows, M.T.; Donat, M.G.; Feng, M. . A Hierarchical Approach to Defining Marine Heatwaves. Prog. Oceanogr.; 2016; 141, pp. 227-238. [DOI: https://dx.doi.org/10.1016/j.pocean.2015.12.014]

25. Zhang, Z.; Wang, W.; Qiu, B. Oceanic Mass Transport by Mesoscale Eddies. Science; 2014; 345, pp. 322-324. [DOI: https://dx.doi.org/10.1126/science.1252418] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25035491]

26. Dong, C.; McWilliams, J.C.; Liu, Y.; Chen, D. Global Heat and Salt Transports by Eddy Movement. Nat. Commun.; 2014; 5, 3294. [DOI: https://dx.doi.org/10.1038/ncomms4294]

27. Della Penna, A.; Llort, J.; Moreau, S.; Patel, R.; Kloser, R.; Gaube, P.; Strutton, P.; Boyd, P.W. The Impact of a Southern Ocean Cyclonic Eddy on Mesopelagic Micronekton. J. Geophys. Res. Ocean.; 2022; 127, e2022JC018893. [DOI: https://dx.doi.org/10.1029/2022JC018893]

28. Frenger, I.; Muennich, M.; Gruber, N.; Knutti, R. Southern Ocean Eddy Phenomenology. J. Geophys. Res. Ocean.; 2015; 120, pp. 7413-7449. [DOI: https://dx.doi.org/10.1002/2015JC011047]

29. Sun, W.; An, M.; Liu, J.; Liu, J.; Yang, J.; Tan, W.; Dong, C.; Liu, Y. Comparative Analysis of Four Types of Mesoscale Eddies in the Kuroshio-Oyashio Extension Region. Front. Mar. Sci.; 2022; 9, 984244. [DOI: https://dx.doi.org/10.3389/fmars.2022.984244]

30. Wang, X.; Zhang, S.; Lin, X.; Qiu, B.; Yu, L. Characteristics of 3-Dimensional Structure and Heat Budget of Mesoscale Eddies in the South Atlantic Ocean. J. Geophys. Res. Ocean.; 2021; 126, e2020JC016922. [DOI: https://dx.doi.org/10.1029/2020JC016922]

31. Mo, D.; He, Q.; Zhan, W.; He, Y.; Zhan, H. A Global Assessment of Eddy-Induced Salinity Anomalies and Salt Transport by Eddy Movement. J. Geophys. Res. Ocean.; 2024; 129, e2023JC020382. [DOI: https://dx.doi.org/10.1029/2023JC020382]

32. Lee, M.-M.; Nurser, A.J.G.; Coward, A.C.; de Cuevas, B.A. Eddy Advective and Diffusive Transports of Heat and Salt in the Southern Ocean. J. Phys. Oceanogr.; 2007; 37, pp. 1376-1393. [DOI: https://dx.doi.org/10.1175/JPO3057.1]

33. Yang, Y.; Wang, D.; Wang, Q.; Zeng, L.; Xing, T.; He, Y.; Shu, Y.; Chen, J.; Wang, Y. Eddy-Induced Transport of Saline Kuroshio Water Into the Northern South China Sea. J. Geophys. Res. Ocean.; 2019; 124, pp. 6673-6687. [DOI: https://dx.doi.org/10.1029/2018JC014847]

34. Yang, S.; Zhang, K.; Song, H.; Ruddick, B.; Liu, M.; Meng, L. Disruptions in Thermohaline Staircases Caused by Subsurface Mesoscale Eddies in the Eastern Caribbean Sea. Commun. Earth Environ.; 2024; 5, 408. [DOI: https://dx.doi.org/10.1038/s43247-024-01577-3]

35. Zhi, H.; Ma, T.; Zhang, R.-H.; Wang, X.; Wu, M. Surface Flux–Induced Salinity Change and Its Effects on Ocean Stratification in Response to Global Warming. Atmos. Ocean. Sci. Lett.; 2025; 100635. [DOI: https://dx.doi.org/10.1016/j.aosl.2025.100635]

36. Röthig, T.; Trevathan-Tackett, S.M.; Voolstra, C.R.; Ross, C.; Chaffron, S.; Durack, P.J.; Warmuth, L.M.; Sweet, M. Human-Induced Salinity Changes Impact Marine Organisms and Ecosystems. Glob. Change Biol.; 2023; 29, pp. 4731-4749. [DOI: https://dx.doi.org/10.1111/gcb.16859]

37. Lukas, R.; Lindstrom, E. The Mixed Layer of the Western Equatorial Pacific Ocean. J. Geophys. Res. Ocean.; 1991; 96, pp. 3343-3357. [DOI: https://dx.doi.org/10.1029/90JC01951]

38. Liu, X.; Ma, X.; Chang, P.; Jia, Y.; Fu, D.; Xu, G.; Wu, L.; Saravanan, R.; Patricola, C.M. Ocean Fronts and Eddies Force Atmospheric Rivers and Heavy Precipitation in Western North America. Nat. Commun.; 2021; 12, 1268. [DOI: https://dx.doi.org/10.1038/s41467-021-21504-w]

39. Calil, P.H.R.; Richards, K.J.; Jia, Y.; Bidigare, R.R. Eddy Activity in the Lee of the Hawaiian Islands. Deep Sea Res. Part II Top. Stud. Oceanogr.; 2008; 55, pp. 1179-1194. [DOI: https://dx.doi.org/10.1016/j.dsr2.2008.01.008]

40. Xie, S.-P.; Liu, W.T.; Liu, Q.; Nonaka, M. Far-Reaching Effects of the Hawaiian Islands on the Pacific Ocean-Atmosphere System. Science; 2001; 292, pp. 2057-2060. [DOI: https://dx.doi.org/10.1126/science.1059781] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11408652]

41. Feloy, K.; Powell, B.S.; Friedrich, T. Remote Impacts of Cyclonic Eddies on Productivity Around the Main Hawaiian Islands. J. Geophys. Res. Ocean.; 2024; 129, e2023JC020670. [DOI: https://dx.doi.org/10.1029/2023JC020670]

42. Droghei, R.; Nardelli, B.B.; Santoleri, R. Combining In Situ and Satellite Observations to Retrieve Salinity and Density at the Ocean Surface. J. Atmos. Ocean. Technol.; 2016; 33, pp. 1211-1223. [DOI: https://dx.doi.org/10.1175/JTECH-D-15-0194.1]

43. Martínez-Moreno, J.; Hogg, A.M.; England, M.H.; Constantinou, N.C.; Kiss, A.E.; Morrison, A.K. Global Changes in Oceanic Mesoscale Currents over the Satellite Altimetry Record. Nat. Clim. Change; 2021; 11, pp. 397-403. [DOI: https://dx.doi.org/10.1038/s41558-021-01006-9]

44. Pegliasco, C.; Delepoulle, A.; Mason, E.; Morrow, R.; Faugère, Y.; Dibarboure, G. META3.1exp: A New Global Mesoscale Eddy Trajectory Atlas Derived from Altimetry. Earth Syst. Sci. Data; 2022; 14, pp. 1087-1107. [DOI: https://dx.doi.org/10.5194/essd-14-1087-2022]

45. Hobday, A.J.; Oliver, E.C.J.; Sen Gupta, A.; Benthuysen, J.A.; Burrows, M.T.; Donat, M.G.; Holbrook, N.J.; Moore, P.J.; Thomsen, M.S.; Wernberg, T. . Categorizing and Naming Marine Heatwaves. Oceanography; 2018; 31, pp. 162-173. [DOI: https://dx.doi.org/10.5670/oceanog.2018.205]

46. Chelton, D.B.; Gaube, P.; Schlax, M.G.; Early, J.J.; Samelson, R.M. The Influence of Nonlinear Mesoscale Eddies on Near-Surface Oceanic Chlorophyll. Science; 2011; 334, pp. 328-332. [DOI: https://dx.doi.org/10.1126/science.1208897] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21921157]

47. He, Q.; Zhan, W.; Feng, M.; Gong, Y.; Cai, S.; Zhan, H. Common Occurrences of Subsurface Heatwaves and Cold Spells in Ocean Eddies. Nature; 2024; 634, pp. 1111-1117. [DOI: https://dx.doi.org/10.1038/s41586-024-08051-2] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39415017]

48. Oliver, E.C.J.; Donat, M.G.; Burrows, M.T.; Moore, P.J.; Smale, D.A.; Alexander, L.V.; Benthuysen, J.A.; Feng, M.; Sen Gupta, A.; Hobday, A.J. . Longer and More Frequent Marine Heatwaves over the Past Century. Nat. Commun.; 2018; 9, 1324. [DOI: https://dx.doi.org/10.1038/s41467-018-03732-9]

49. Kahru, M.; Mitchell, B.G.; Gille, S.T.; Hewes, C.D.; Holm-Hansen, O. Eddies Enhance Biological Production in the Weddell-Scotia Confluence of the Southern Ocean. Geophys. Res. Lett.; 2007; 34, L14603. [DOI: https://dx.doi.org/10.1029/2007GL030430]

50. Lévy, M.; Klein, P. Does the Low Frequency Variability of Mesoscale Dynamics Explain a Part of the Phytoplankton and Zooplankton Spectral Variability?. Proc. R. Soc. Lond. Ser. Math. Phys. Eng. Sci.; 2004; 460, pp. 1673-1687. [DOI: https://dx.doi.org/10.1098/rspa.2003.1219]

51. O’Connor, B.M.; Fine, R.A.; Maillet, K.A.; Olson, D.B. Formation Rates of Subtropical Underwater in the Pacific Ocean. Deep Sea Res. Part I Oceanogr. Res. Pap.; 2002; 49, pp. 1571-1590. [DOI: https://dx.doi.org/10.1016/S0967-0637(02)00087-0]

52. Suga, T.; Kato, A.; Hanawa, K. North Pacific Tropical Water: Its Climatology and Temporal Changes Associated with the Climate Regime Shift in the 1970s. Prog. Oceanogr.; 2000; 47, pp. 223-256. [DOI: https://dx.doi.org/10.1016/s0079-6611(00)00037-9]

53. Barcelo-Llull, B.; Pallas-Sanz, E.; Sangra, P.; Martinez-Marrero, A.; Estrada-Allis, S.N.; Aristegui, J. Ageostrophic Secondary Circulation in a Subtropical Intrathermocline Eddy. J. Phys. Oceanogr.; 2017; 47, pp. 1107-1123. [DOI: https://dx.doi.org/10.1175/JPO-D-16-0235.1]

54. Chang, Y.-L.; Miyazawa, Y.; Oey, L.-Y.; Kodaira, T.; Huang, S. The Formation Processes of Phytoplankton Growth and Decline in Mesoscale Eddies in the Western North Pacific Ocean. J. Geophys. Res. Ocean.; 2017; 122, pp. 4444-4455. [DOI: https://dx.doi.org/10.1002/2017JC012722]

55. Vinogradova, N.T.; Ponte, R.M. Clarifying the Link between Surface Salinity and Freshwater Fluxes on Monthly to Interannual Time Scales. J. Geophys. Res. Ocean.; 2013; 118, pp. 3190-3201. [DOI: https://dx.doi.org/10.1002/jgrc.20200]

56. Grodsky, S.A.; Reul, N.; Lagerloef, G.; Reverdin, G.; Carton, J.A.; Chapron, B.; Quilfen, Y.; Kudryavtsev, V.N.; Kao, H.-Y. Haline Hurricane Wake in the Amazon/Orinoco Plume: AQUARIUS/SACD and SMOS Observations. Geophys. Res. Lett.; 2012; 39, L20603. [DOI: https://dx.doi.org/10.1029/2012GL053335]

57. Drushka, K.; Asher, W.E.; Sprintall, J.; Gille, S.T.; Hoang, C. Global Patterns of Submesoscale Surface Salinity Variability. J. Phys. Oceanogr.; 2019; 49, pp. 1669-1685. [DOI: https://dx.doi.org/10.1175/JPO-D-19-0018.1]

58. Zhu, R.; Yang, H.; Li, M.; Chen, Z.; Ma, X.; Cai, J.; Wu, L. Observations Reveal Vertical Transport Induced by Submesoscale Front. Sci. Rep.; 2024; 14, 4407. [DOI: https://dx.doi.org/10.1038/s41598-024-54940-x]

59. Schlegel, R.W.; Oliver, E.C.J.; Hobday, A.J.; Smit, A.J. Detecting Marine Heatwaves with Sub-Optimal Data. Front. Mar. Sci.; 2019; 6, 737. [DOI: https://dx.doi.org/10.3389/fmars.2019.00737]

60. Gaube, P.; Chelton, D.B.; Samelson, R.M.; Schlax, M.G.; O’Neill, L.W. Satellite Observations of Mesoscale Eddy-Induced Ekman Pumping. J. Phys. Oceanogr.; 2015; 45, pp. 104-132. [DOI: https://dx.doi.org/10.1175/JPO-D-14-0032.1]

61. Delcroix, T.; Chaigneau, A.; Soviadan, D.; Boutin, J.; Pegliasco, C. Eddy-Induced Salinity Changes in the Tropical Pacific. J. Geophys. Res. Ocean.; 2019; 124, pp. 374-389. [DOI: https://dx.doi.org/10.1029/2018JC014394]

62. Liu, H.; Wei, Z. Intercomparison of Global Sea Surface Salinity from Multiple Datasets over 2011–2018. Remote Sens.; 2021; 13, 811. [DOI: https://dx.doi.org/10.3390/rs13040811]

63. Bian, C.; Jing, Z.; Wang, H.; Wu, L.; Chen, Z.; Gan, B.; Yang, H. Oceanic Mesoscale Eddies as Crucial Drivers of Global Marine Heatwaves. Nat. Commun.; 2023; 14, 2970. [DOI: https://dx.doi.org/10.1038/s41467-023-38811-z]

64. Song, X.; Sun, R.; He, S.; Zhang, H.; Gu, Y.; Li, P.; Song, J. Impact of Anticyclonic Mesoscale Eddies on the Vertical Structures of Marine Heatwaves in the South China Sea. Remote Sens.; 2025; 17, 370. [DOI: https://dx.doi.org/10.3390/rs17030370]

65. Napitupulu, G. Eddy-Induced Modulation of Marine Heatwaves and Cold Spells in a Tropical Region: A Case Study in the Natuna Sea Area. Ocean Dyn.; 2025; 75, 28. [DOI: https://dx.doi.org/10.1007/s10236-025-01673-8]

66. Ni, Q.; Zhai, X.; Wang, G.; Marshall, D.P. Random Movement of Mesoscale Eddies in the Global Ocean. J. Phys. Oceanogr.; 2020; 50, pp. 2341-2357. [DOI: https://dx.doi.org/10.1175/jpo-d-19-0192.1]

67. Wang, H.; Qiu, B.; Liu, H.; Zhang, Z. Doubling of Surface Oceanic Meridional Heat Transport by Non-Symmetry of Mesoscale Eddies. Nat. Commun.; 2023; 14, 5460. [DOI: https://dx.doi.org/10.1038/s41467-023-41294-7] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37673876]

68. Li, H.; Xu, F.; Wang, G. Global Mapping of Mesoscale Eddy Vertical Tilt. J. Geophys. Res. Ocean.; 2022; 127, e2022JC019131. [DOI: https://dx.doi.org/10.1029/2022JC019131]

69. Zhang, Z.; Tian, J.; Qiu, B.; Zhao, W.; Chang, P.; Wu, D.; Wan, X. Observed 3D Structure, Generation, and Dissipation of Oceanic Mesoscale Eddies in the South China Sea. Sci. Rep.; 2016; 6, 24349. [DOI: https://dx.doi.org/10.1038/srep24349] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27074710]