In this study, the best-track data for the medicane Daniel was obtained from the Zivipotty Cyclone Report database (https://zivipotty.hu/tcr.html, last access: 13 May 2026). Eddy identification was based on daily SLA fields sourced from the CMEMS. Specifically, the dataset SEALEVEL_EUR_PHY_L4_NRT_OBSERVATIONS_008 _060, with a spatial resolution of 0.125°, was utilized. To detect and characterize the MHW, daily SST data from the NOAA Optimum Interpolation SST V2 dataset (Reynolds et al., 2007) were used. This dataset has a spatial resolution of 0.25° and covers the period from 1981 to the present. The key atmospheric variables, including total column water (which represents the sum of water vapor, liquid water, cloud ice, rain, and snow in a column extending from the surface of the Earth to the top of the atmosphere), total precipitation, vertically integrated moisture divergence, mean sea level pressure (MSLP), 10 m zonal and meridional wind components, daily radiative fluxes of shortwave and longwave radiations, (denoted as QSW and QLW respectively), and turbulent heat fluxes of latent and sensible flux (denoted as Qlat and Qsen respectively) were retrieved from the ERA5 reanalysis (Hersbach et al., 2020). The surface net heat flux (Qnet) was derived from a combination of radiative and turbulent fluxes (Menna et al., 2023). 1Qnet=QSW-QLW-Qlat-Qsen High-resolution SLA observations were obtained from the SWOT Level 3 satellite product, which offers 2 km spatial resolution, provided by Archiving, Validation and Interpretation of Satellite Oceanographic data (AVISO; https://www.aviso.altimetry.fr/en/data, last access: 13 May 2026).

Lastly, biogeochemical variables such as chlorophyll, phytoplankton, nutrients, and dissolved oxygen were accessed via the CMEMS from the product MEDSEA_MULTIYEAR_BGC_006_008, available at a 4–5 km spatial resolution, and 1 h temporal resolution (10.48670/mds-00374). Daily satellite-derived chlorophyll products (OCEANCOLOUR_MED_BGC_L4_NRT) with a spatial resolution of 1 km were used in this study (Volpe et al., 2019; Volpe et al., 2018; Berthon and Zibordi, 2004). These datasets are archived by the Copernicus Marine Service (10.48670/moi-00298) and are provided by the Italian National Research Council (CNR, Rome, Italy), with data availability from January 2023 to the present. The multi-sensor product integrates observations from SeaWiFS, MODIS, MERIS, VIIRS, and OLCI, and includes key biogeochemical variables such as chlorophyll a (Chl a), diffuse attenuation coefficient at 490 nm, and primary production.

Despite their advantages, these datasets are subject to uncertainties arising from atmospheric correction errors, cloud contamination, aerosol effects, and reduced accuracy in optically complex coastal waters. In addition, the gap-filling procedure used to generate continuous fields may introduce smoothing in regions with persistent data gaps. Nevertheless, these products provide robust, high-resolution information on surface biogeochemical variability associated with cyclone-induced ocean processes.

In this study, we investigate the intensification, structure, and impacts of medicane Daniel. This section outlines the methodologies used to analyze the oceanic and atmospheric conditions associated with the event. The analysis focuses on key processes influencing cyclone formation and intensification, including the role of oceanic features such as WCEs and MHWs, the contribution of ocean heat content as a source of subsurface thermal energy, and atmospheric variables such as moisture and wind fields. The methods used for the identification of the sea eddies and MHWs, as well as the computation of ocean heat content, are described below.

Figure 1a–b shows the SLA along the medicane's track. At first, the medicane passed over a cluster of three anticyclonic eddies between 4 and 6 September, presumably contributing to its intensification (Jangir et al., 2023). This intensification between 8 and 10 September was also supported by the presence of a moderate MHW, as shown in Fig. 1c–e (see also Fig. S1 in the Supplement). Furthermore, just before making landfall, the medicane arrived in a region characterized mostly by anticyclonic mesoscale WCE activity (Fig. 1f–h). OHC anomaly shown in Fig. 1i–k indicates higher heat content in the southern part of the domain close to landfall location, also supporting the intensification of the medicane. In addition to the maps in panels (c)–(k) of Fig. 1, the analysis was carried out by computing MHW, SST, and OHC anomalies at each 6 h track position (Fig. 1l–n), with particular emphasis on the maximum cyclone intensity (Max-CI) location. We observed from the analysis that the maximal values of MHW (∼0.6), SST (∼0.83 °C), and OHC (∼16.30 KJ cm⁻²) anomalies occurred in the last 2 d before Max-CI, indicating that medicane Daniel intensified over thermodynamically favourable conditions that were among the most anomalous along its track.

The combination of two extreme oceanic preconditions, namely the WCE and the MHW, together with high absolute values of OHC at the intensification site near the coastal region, may have contributed to the deadly outcome of this medicane. Rathore et al. (2022) and Jangir et al. (2024) highlighted the critical role of sudden intensification in Cyclone Amphan over the Indian Ocean and medicane Ianos over the Mediterranean Sea, respectively, in the presence of MHWs along their paths. The findings of the current study are consistent with these observations, demonstrating that while cyclone genesis and intensification can occur independently of such features, the presence of WCEs and MHWs may enhance the rate and magnitude of intensification over a shorter time.

Figure 1 also highlights the potential importance of ocean characteristics such as WCEs and MHWs in cyclone intensification, indicating that cyclone intensity increased in the presence of WCEs and MHWs. This behavior is similar to how cyclones in other ocean basins react to changes in intensification factors related to underlying eddies (Ali et al., 2007; Lin et al., 2013; Jangir et al., 2021, 2023). When a cyclone encounters a WCE, the negative feedback loop between cyclone intensity and SST diminishes. Normally, cyclones extract heat from the ocean, resulting in surface cooling due to enhanced mixing and evaporation, which acts to reduce cyclone intensity. However, if a WCE or MHWs are present, the high SST persists longer, intensifying the cyclone and reducing the negative feedback effect (Bender et al., 1993; Jangir et al., 2024).

Furthermore, atmospheric cyclones draw a significant portion of their energy from warm, deep ocean waters; therefore, quantifying the amount of this warm, deep water provides a more accurate measure of the energy available to the storm. OHC serves as this metric, indicating how much warm water a cyclone can convert into energy. Studies have shown that OHC is a far superior predictor compared to SST alone (Wada and Usui, 2007; Sharma and Ali, 2014; Lin et al., 2013; Law, 2011). Analysis of the OHC revealed a significant amount of OHC at the intensification locations, providing a favorable upper-ocean thermal reservoir for medicane Daniel to intensify. Approximately 120 KJ cm⁻² of heat was available from 4 to 9 September, even before the cyclone's intensification (Fig. 1i–k and Fig. S2). This accumulated heat at the intensification location is attributed to the presence of the WCE and the MHW, which decreases after the passage of the medicane on 11 and 12 September 2023 (Fig. S2). The presence of heat in the form of SST anomaly (Fig. 1f–h) and OHC anomaly (Fig. 1i–k) along the path of medicane maintains intensity by reducing negative feedback that occurs due to the passage of the cyclone (Jangir et al., 2023; Jangir et al., 2024).

The MSLP and wind speed, computed using the Cressman averaging method (Cressman, 1959) along the cyclone's track, are shown in Fig. 2a. The MSLP indicates moderate intensification from 1006 to 1002 hPa when the storm passes over the WCE on 5 September 2023 (Fig. 2a). Subsequently, from 6 to 7 September 2023, the medicane passes over a CCE region, and its intensity is reduced to 1004 hPa. Upon reaching the vicinity of Libya's coast on 8 September 2023, it quickly intensified. An additional WCE was present in the vicinity of the cyclone's path at that time, potentially further contributing to its fast intensification. This fast intensification is indicated by a drop in the MSLP and a sudden increase in wind speed near the eddy and MHW location (Fig. 2a). The cutoff low, which was supported by high net heat flux, persisted for 2 d (Figs. 2b–d, S3), even after the cyclone made landfall in Libya.

Moisture processes play an equally important role in cyclone intensification. Their importances been demonstrated in previous studies by Jangir et al. (2023) and Pytharoulis et al. (2018), emphasizing that elevated SST in the form of WCE or MHW is essential for providing moisture to a medicane via surface fluxes, enhancing convection. Additionally, Jangir et al. (2024) highlighted the importance of moisture convergence in the intensification of cyclones and increasing total associated precipitation. Thus, motivated by these findings, we focus in this study on the causes of the intensification of the medicane Daniel and the extreme flood that occurred during the event.

The analysis of moisture convergence (i.e., mean vertically integrated moisture divergence) showed a pattern of moisture convergence along the cyclone's path. Notably, this convergence coincides with the eddy location at the intensification location on 9 September 2023 (Fig. S4). This alignment suggests that the eddy supplied the moisture needed for the cyclone's intensification. The interaction between the eddy and the medicane likely enhanced moisture availability, contributing to the storm's strengthening at that specific point in it's path. Additionally, the total column water was notably high at the WCE and MHW locations. While this total water was present before the intensification location as well, it converged around the eddy at the intensification location (Fig. S5), leading to substantial precipitation in that area (Figs. 2e–g and S6). The severe precipitation near the coastal region highly coincides with the WCE and MHW. The WCE's influence intensified the cyclone by providing additional moisture and heat, leading to heavy rainfall. This heavy precipitation, concentrated near the coast, exacerbated the storm's impact, severely damaging the affected areas.

The SWOT mission offers high-resolution sea surface height anomaly or SLA data with unprecedented spatial detail, enabling precise detection of mesoscale and sub-mesoscale ocean features, such as eddies and fronts (Morrow et al., 2019). This can be valuable for studying cyclones, which interact strongly with oceanic eddies that influence storm intensity. Unlike traditional altimeters, SWOT's wide-swath coverage enables improved detection of mesoscale and sub-mesoscale eddies, frontal gradients, and filaments that regulate ocean heat distribution and air-sea exchanges. These features are often underrepresented in low-resolution datasets, limiting their ability to capture localized processes such as eddy-cyclone interactions and cyclone-induced mixing. By improving the detection of these fine-scale physical structures, SWOT also provides a framework for interpreting biogeochemical responses. While coarse datasets show bulk chlorophyll changes, SWOT helps identify localized regions of enhanced mixing and upwelling that drive nutrient supply and biological variability. This allows for a clearer linkage between physical forcing and biogeochemical response. Overall, SWOT overcomes key limitations of conventional altimetry by preserving high-frequency spatial gradients, enabling a more accurate representation of the ocean state during extreme events such as medicane Daniel.

Here we show the SWOT swath passing over the location of the eddies along the track of medicane Daniel (Fig. 3). In Fig. 3a–b, the eddy initially appears small and low intensity in the CMEMS, and the cyclone is observed nearby. However, SWOT data reveal a more intense and extensive eddy structure than CMEMS, with the cyclone positioned directly above it. These findings report the value of SWOT observations in capturing fine-scale oceanic features and dynamics, offering critical insights into cyclone-eddy interactions.

To investigate the impact of the medicane on oceanic physical (temperature and salinity) and biogeochemical properties (i.e., chlorophyll a, phytoplankton, nitrate and phosphate, and oxygen concentration), we analyzed vertical profiles of key variables along the cyclone's track. The analysis focused on differences between 2 d after the cyclone's passage, minus 2 d before (Fig. 4a–g). The results reveal a notable decrease in temperature along the cyclone path, with a general strong cooling along its path except for a short pause in the cold SST anomaly region in the morning of the 8th (Fig. 1n). This region was also relatively outside the MHW domain (Fig. 1l). The salinity also decreases on the surface presumably due to a massive influx of freshwater from heavy rainfall, again, except for the morning of the 8th, in which the cyclone was outside the influence of WCEs and MHW (Fig. 4b). In contrast, at the subsurface, Chl a and phytoplankton concentrations exhibit a marked dipole (Fig. 4c–d), while nutrients increase (Fig. 4e–f) and oxygen decreases (Fig. 4g). This biological response can be attributed to cyclone-induced Ekman pumping upwelling and high subsurface vertical mixing. Enhanced nutrient availability at the subsurface layer can persist into the mixed layer, allowing for sufficient sunlight, and together with elevated oxygen concentrations (associated with surface cooling), may foster increased surface Chl a and phytoplankton biomass. Signs of this can be seen in the MHW region, where higher Chl a concentrations reach the surface. But, unlike previous results (Jangir et al., 2026), medicane Daniel only shows an increase in Chl a at the surface in the MHW region.

Profiles of temperature and Chl a at the maximum cyclone intensity (Max-CI) location and time (Fig. 4h–i) reveal general cooling after the passage of the cyclone. The deep chlorophyll maximum (DCM) was located far below the mixed layer depth, at around 140 m depth. The subsurface crossings between the profiles and the relatively stationary location of the DCM in Fig. 4i indicate that out of the two processes mentioned above, namely cyclone-induced upwelling and high subsurface vertical mixing, only cyclone-induced subsurface mixing can explain the change. Subsurface mixing mechanism is typically much slower than turbulence in the mixed layer, but under storm conditions may become comparable. The gradual subsurface increase in Chl a, as opposed to the vertical line observed in the mixed layer, indicates weaker but comparable turbulence below the mixed layer.

Figure 4i shows the change in subsurface profiles at the time and location of Max-CI, situated on the periphery of a WCE. To further investigate the underlying mechanisms and compare the medicane's impact on both WCEs and CCEs, vertical cross-sections across the main center of the WCE and adjacent CCE region were analyzed during the pre-storm, during storm, and post-storm phases (Fig. 5). These cross-sections reveal notable eddy-dependent subsurface changes associated with the passage of the medicane. As expected, during and after the passage of the storm, significant surface cooling is observed in Fig. 5 (panels b–d and h–i). In addition to the cooling observed within the mixed layer, three distinct circular-like cooling patterns are evident immediately beneath it. These may indicate a secondary circulation starting at the deep subsurface (below 200 m), which transfers deep cold water to the layers below the mixed layer. In this case, the two patterns on the right indicate an upwelling cell at the WCE boundaries (green arrows in Fig. 5i) and another at the CCE center (purple arrow in Fig. 5i). These circulation cells also create the dipole pattern in the Chl a inside the WCE and the CCE (Fig. 5e–g and j–k).

To explore the dynamical mechanisms, Fig. 6 shows vertical profiles from 1 to 15 September 2023 of Chl a at two locations along the same line as in Fig. 5a, one inside the CCE (panel a) and the other inside the WCE (panel b). Before the arrival of the medicane (1–5 September), the DCM within the CCE was shallower than in the WCE, consistent with the typical vertical structure of these eddies. The DCM in both eddies remains relatively stable until the cyclone approaches the region. As the medicane approaches the eddies, it intensifies the CCE while diminishing the WCE, which, as expected from theory (e.g., Klein and Lapeyre, 2009), supports eddy-induced upwelling. Indeed, during the storm, the DCM is lifted upward in both cases from 5 to 9 September. Yet, the DCM inside the WCE is exhibited significantly greater shoaling than inside the CCE (∼40 m in the WCE relative to ∼15 m in the CCE). The more pronounced response in the WCE likely stems from lower thermal stratification and higher surface wind stress curl that contribute to cyclone-induced Ekman pumping upwelling. Only a few days later, the DCM inside the WCE partially starts to restore its pre-storm condition. The DCM inside the CCE quickly drops back down and even overshoots its depth relative to pre-storm conditions, which can be explained by its stronger stratification and resulting buoyancy restoring force. In addition, we found that cyclone-induced upwelling alone cannot explain the DCM increase in both the WCE and CCE (Fig. 6c), which indicates an increase of about 10–15 m during the medicane influence (5–9 September). Therefore, we conclude that the upwelling may have also been influenced by isopycnal adjustment triggered by the medicane.

Subsurface mixing similar to what was shown under the medicane at Max-CI (Fig. 4i), seems to play a role also here. This role can be indicated by the crossing of the Chl a profiles during the storm (5–9 September) with the profiles after the storm (10–15 September) between MLD (∼-20 m) and the DCM (∼-130 m). However, here, it seems to complement upwelling and to play a more important role at the WCE (as expected, since in general wind stress is higher above WCEs). Also, satellite-based surface observations (Fig. 6d) indicate higher Chl a above the WCE region, which may be explained by the closer-to-the-surface DCM and stronger cyclone-induced upwelling and vertical mixing.