Vertical structures of marine heatwaves

Marine heatwaves (MHWs), defined as prolonged periods of unusually warm ocean temperatures, pose a significant threat to marine ecosystems and human societies. These events, lasting from weeks to months and spanning thousands of kilometers, have increased substantially in frequency and intensity over the past century, a trend projected to worsen with ongoing global warming. The ecological impacts are widespread and profound, ranging from coral bleaching and mortality to declines in seagrass and kelp forests, affecting crucial ecosystem services like carbon sequestration and habitat provision for numerous species. These impacts cascade through food webs, causing widespread mortality of various marine organisms and impacting fisheries, tourism, and other socioeconomic sectors. The consequences extend beyond the ocean surface, impacting the atmosphere and deep ocean, underscoring the need for a comprehensive understanding of MHWs' vertical extent and dynamics. MHW formation, persistence, and decay are complex processes driven by a combination of atmospheric and oceanic factors, including air-sea heat flux, horizontal and vertical heat advection, and mixing. These processes are further modulated by large-scale climate modes like ENSO, IOD, NAO, PDO, and SAM, highlighting the intricate interplay between MHWs and climate variability. While surface MHWs have been extensively studied using satellite SST data, knowledge of their vertical structure and global patterns remains limited. This study leverages Argo float data to address this gap by identifying distinct vertical structures of MHWs globally and examining their spatiotemporal characteristics.

Previous research has primarily focused on surface MHWs, utilizing satellite-derived sea surface temperature (SST) data. Recent studies have begun to explore the vertical extent of MHWs, particularly in specific regions like the East Australian Current System, categorizing them into shallow, intermediate, and deep events based on their vertical reach. These studies have indicated that deeper MHWs are driven by advection, while shallower events are dominated by surface fluxes. However, inconsistencies exist, with some shallow and intermediate MHWs showing cooling anomalies below the surface warming. Moreover, not all intermediate and deep events exhibit maximum warming in the subsurface. These observations suggest a greater complexity in the vertical structure of MHWs and a need for a more comprehensive global analysis. The availability of high-resolution temperature and salinity profiles from Argo floats presents an excellent opportunity to investigate this complexity on a global scale.

This study utilizes a combination of satellite and in-situ data to analyze the vertical structure and spatiotemporal characteristics of marine heatwaves (MHWs). First, MHWs were identified using the NOAA Optimum Interpolation (OI) SST V2.1 global daily gridded SST data (1982-2021), following the definition of Hobday et al. (2016). This involves identifying periods where SST exceeds the 90th percentile of a 30-year average seasonal climatology (1982-2011) for at least five consecutive days. The data were upscaled to 2° grid resolution for consistency with subsurface data. To characterize the vertical structure, temperature and salinity profiles from Argo floats (2001-2020) obtained from the China Argo Real-time Data Center (CARDC) were used. These profiles, spanning depths from 0 to 2000 dbar, were mapped onto a 2° grid and linearly interpolated onto 58 depth levels. Individual temperature anomalies were calculated by subtracting seasonal climatological profiles from the CSIRO Atlas of Regional Seas (CARS) 2009. MHW vertical structures were determined by averaging temperature anomalies from Argo float profiles within MHW time intervals. The impact depth of each MHW was defined as the depth at which 85% of the maximum vertically cumulative temperature anomaly is located. Based on this impact depth and characteristics of the temperature anomaly profiles, MHWs were classified into four types: shallow, subsurface-reversed, subsurface-intensified, and deep. The mixed layer depth (MLD) was defined as the depth where potential density differs from the 10-dbar value by 0.03 kg m⁻³, and the thermocline depth was defined as the depth of maximum vertical temperature gradient. To investigate the drivers of different vertical structures, a mixed-layer heat budget analysis was performed using the equation: ∂T/∂t = (Qs − Qb)/(ρwhCp) − ∇ ⋅ (uT) + ∇²T + Res, where the terms represent temperature tendency, surface flux forcing, horizontal advection, and residual components. Data sources for this analysis include NOAA OISST, NCEP-DOE Reanalysis 2 air-sea heat fluxes, GLORYS12V1 MLD, OSCAR horizontal currents, and AVISO mesoscale eddy data. Finally, long-term trends in MHW area and depth were analyzed using linear regression and the bootstrap method.

This study identified four distinct vertical structures of marine heatwaves (MHWs) using Argo data from 2001 to 2020: 1. **Shallow MHWs:** Warming is confined to the surface layer and decreases rapidly with depth. These are most prevalent in middle-high latitudes. 2. **Subsurface-reversed MHWs:** Show surface warming with anomalous cooling beneath. Hotspots are found at low latitudes. 3. **Subsurface-intensified MHWs:** Exhibit maximum warming anomalies in the subsurface layers. Hotspots are located at low latitudes. 4. **Deep MHWs:** Display surface warming anomalies that decay slowly with depth, occurring primarily in subtropical-subpolar regions. The spatial distribution of these MHW types varies significantly. Shallow MHWs are concentrated in mid-high latitudes, while subsurface-reversed and subsurface-intensified MHWs are more common in tropical oceans. Deep MHWs are prevalent in subtropical-subpolar regions. Oceanic planetary waves, boundary currents, eddies, and mixing influence the vertical structures and spatial distribution. Hotspots of subsurface-reversed and subsurface-intensified MHWs are observed in low-latitude regions, likely due to active oceanic planetary wave processes associated with climate modes. These waves drive thermocline fluctuations, affecting the subsurface signals of MHWs. In the tropical Pacific, the distribution of these two types is spatially differentiated, with subsurface-reversed MHWs concentrated in the central basin and subsurface-intensified MHWs in the east and west. This pattern is consistent with the correlation between surface and subsurface temperature anomalies. The impact depth of MHWs, defined as the depth where 85% of the maximum vertically cumulative temperature anomaly is located, varies among the four types. Shallow MHWs tend to have shallow impact depths, while subsurface-intensified and deep MHWs have deep impact depths. The spatial distribution of impact depth correlates with mixed layer depth (MLD) and thermocline depth, indicating the importance of oceanic stratification. The study also analyzed the heat budget of MHWs, examining the roles of temperature tendency, surface flux forcing, horizontal advection, and residual terms. In tropical oceans, vertical processes, particularly those driven by oceanic planetary waves, are crucial. In subtropical oceans, air-sea heat flux anomalies associated with high-pressure systems are dominant factors. High-latitude oceans are influenced by heat advection, mesoscale eddies, and upwelling/mixing anomalies. Western boundary current regions exhibit strong horizontal heat advection, contributing to deep MHWs. Eastern boundary currents show a more balanced influence of atmospheric and oceanic forcing. Finally, significant increasing trends in both the area and depth of all MHW types were observed during the Argo era (2001-2020), with the greatest increase in subsurface-intensified MHWs and the least in shallow MHWs. The increasing trends are largely attributed to the long-term warming of the global ocean, although interannual variability related to climate modes like ENSO and IOD is also apparent, particularly in tropical oceans.

This study's findings significantly advance our understanding of MHWs by highlighting the diversity of their vertical structures and their relationships to regional oceanographic processes and climate modes. The identification of four distinct MHW types challenges the previous simplification of MHWs as purely surface phenomena. The spatial distribution of these MHW types underscores the critical role of ocean dynamics in shaping their vertical extent and intensity. The strong correlation between MHW impact depth and oceanic stratification parameters (MLD, thermocline depth) points to the importance of considering these factors in predicting MHWs and assessing their ecological consequences. The heat budget analysis further clarifies the regional differences in the dominant physical drivers of MHWs, emphasizing the interplay between atmospheric forcing (e.g., air-sea heat fluxes) and oceanic processes (e.g., advection, mixing, planetary waves). The observed increasing trends in both the area and depth of all MHW types over the Argo era directly reflect the ongoing warming of the global ocean, a clear indicator of climate change. This underscores the urgency of addressing climate change to mitigate the increasing frequency and severity of MHWs. Future research should focus on improving the understanding of the ecological impacts of subsurface MHWs on marine organisms, particularly mesopelagic species, as they often experience stronger temperature anomalies than surface waters. Moreover, incorporating these findings into numerical ocean models is crucial for more accurate predictions of MHWs in the future, enabling better management strategies for fisheries, aquaculture, and marine conservation.

This study provides a comprehensive analysis of the vertical structures of global MHWs, revealing four distinct types characterized by unique spatial distributions and depths. These findings demonstrate the importance of considering MHWs' vertical extent in understanding their impact on marine ecosystems and highlight the significant influence of ocean dynamics. The observed increasing trends in MHW area and depth underscore the ongoing effects of climate change. Future research should focus on improving the ecological impact assessments of subsurface MHWs, and on refining numerical models to improve MHW predictions.

While this study utilizes a large dataset of Argo float profiles, the relatively sparse spatial and temporal resolution of these measurements compared to satellite SST data could influence the accuracy of vertical structure identification and spatial distribution analysis. The Argo record, spanning from 2001 to 2020, is relatively short, limiting the ability to discern decadal to multidecadal climate variability from long-term trends. Furthermore, the classification scheme for the four MHW types is based on a set of specific criteria, and there could be some subjectivity in classifying events near the boundaries of these categories. Finally, while the study provides insights into the potential drivers of different MHW vertical structures, further research is needed to fully quantify the relative contribution of each factor in different regions.