Search

Among the documented consequences of anthropogenic global warming are the increased frequency and duration of marine heatwaves in the global ocean. The literature dedicated to Arctic marine heatwaves corroborates those results, but fails to identify the heat sources and sinks. Because of the numerous feedbacks impacting polar regions, understanding the processes triggering and dissipating those extreme events is particularly important to predict their occurrence in a fast changing ocean. A three‐dimensional regional ice‐ocean numerical model is used to calculate a surface mixed layer heat budget and to investigate mechanisms generating and dissipating marine heatwaves. The majority of the marine heatwaves are onset by surface heat fluxes and decayed by bottom and surface heat fluxes. The dominant processes are spatially and seasonally heterogeneous: lateral heat flux can become the primary process when advecting heat anomalies at the main Arctic gateways or by triggering temperature extremes in winter. Using a Reynolds decomposition, it can be determined that the shoaling of the surface mixed layer induced by ice melt can significantly lengthen and intensify Arctic marine heatwaves. In winter, the analysis of marine heatwaves poses unique challenges, with the long term freshening of the Arctic inducing a positive trend of 0.1°C per decade for the freezing point. Arctic marine heatwaves are expected to keep increasing in duration and intensity due to the increased trend of the primary process, the surface heat flux, and their dissipation by bottom heat flux provides a pathway for heat from the atmosphere to the Arctic subsurface water masses. Extreme warm temperature anomalies in the ocean, called marine heatwaves are becoming more frequent, longer and more intense around the globe. In the Arctic Ocean, the complex links between sea ice, the atmosphere and the ocean could both amplify or hinder the impact of those events. Yet, understanding where the heat comes from and where it goes is critical to anticipate further changes in an already fast‐changing environment. To this aim, we calculate the sources and sinks of heat in a numerical model simulating the ocean and sea ice. Most marine heatwaves in the Arctic are generated by taking atmospheric heat up and disappear when this heat is transferred to the deeper ocean. In doing so, marine heatwaves provide a pathway for heat from the atmosphere to the subsurface ocean. Oceanic circulation can also propagate marine heatwaves along the continental shelf and toward the marginal ice zone. When ice melts, it shoals the upper layer of the ocean, concentrating the atmospheric heat and lengthening and intensifying marine heatwaves. The excess heat stored in the subsurface can be expected to resurface later in the season and will then delay the freezing of new sea ice. A heat budget analysis is applied to a regional ice‐ocean model to determine the dominating drivers of marine heatwaves in the ArcticMarine heatwaves in the Arctic provide a pathway for heat from the atmosphere to the subsurface oceanSea ice melt shoals the surface mixed layer, and this prolongs and intensifies marine heatwaves A heat budget analysis is applied to a regional ice‐ocean model to determine the dominating drivers of marine heatwaves in the Arctic Marine heatwaves in the Arctic provide a pathway for heat from the atmosphere to the subsurface ocean Sea ice melt shoals the surface mixed layer, and this prolongs and intensifies marine heatwaves

The global ocean has warmed substantially over the past century, with far-reaching implications for marine ecosystems1. Concurrent with long-term persistent warming, discrete periods of extreme regional ocean warming (marine heatwaves, MHWs) have increased in frequency2. Here we quantify trends and attributes of MHWs across all ocean basins and examine their biological impacts from species to ecosystems. Multiple regions in the Pacific, Atlantic and Indian Oceans are particularly vulnerable to MHW intensification, due to the co-existence of high levels of biodiversity, a prevalence of species found at their warm range edges or concurrent non-climatic human impacts. The physical attributes of prominent MHWs varied considerably, but all had deleterious impacts across a range of biological processes and taxa, including critical foundation species (corals, seagrasses and kelps). MHWs, which will probably intensify with anthropogenic climate change3, are rapidly emerging as forceful agents of disturbance with the capacity to restructure entire ecosystems and disrupt the provision of ecological goods and services in coming decades.

This study investigates trends and interannual variability of the marine climate across the continental shelf off eastern Tasmania for 1993–2016. This region is a hotspot for global warming and biodiversity. Eastern Tasmania lies at the boundary between two ocean currents (the East Australian Current Extension, or EAC Extension, and the Zeehan Current, ZC) leading to the local marine climate exhibiting trends and variability from both boundary currents. A numerical ocean model is used to provide high‐resolution (∼2 km) estimates of the temperature, salinity, and circulation for the region. Results indicate significant positive trends in temperature, salinity, and southward flow over the shelf, consistent with an increasing EAC Extension. These trends are particularly strong in autumn, indicating a lengthening of the warm season. The interannual variability in the EAC Extension and ZC was quantified by a simple index, based on a modal analysis of surface circulation, indicating the relative dominance of each current. Strong EAC years were related to significantly more summertime marine heat wave days. Large‐scale remote drivers of variability were considered, and we found weak but significant links with El Niño–Southern Oscillation and Tasman Sea Blocking. El Niño–Southern Oscillation was found to modulate the EAC Extension in summer with a El Niño leading to enhanced southward flow and warming over the shelf. Tasman Sea Blocking was found to drive enhanced southward surface flow, particularly in winter. Nonetheless, large‐scale forcing modes explain less than 25% of the total variability in the EAC‐ZC system indicating that most of the variability is internally generated. The near‐surface waters off eastern Tasmania represent both a global warming and a biodiversity hotspot, and marine ecosystems there are under significant stress. The sensitivity of local ecosystems to both warming and transient marine heat wave events has been observed but spatiotemporal patterns of historical temperature and salinity variability are less well known. A numerical ocean model is used here to provide high‐resolution estimates of the seawater temperatures, salinity, and circulation over the eastern Tasmanian continental shelf. We perform a systematic analysis of multidecadal‐scale trends and interannual variability off eastern Tasmania over the 24‐year period from 1993 to 2016. We find a significant increase in East Australian Current (EAC) Extension penetration into the region with an associated increase in temperature and salinity throughout the water column, particularly in summer and autumn. The variability of the circulation on the shelf is found to be dominated by the interplay of the EAC Extension and Zeehan Current, and an annual index is developed indicating the relative strength of these two currents over time. In both summer and winter we find that an anomalously strong EAC Extension (Zeehan Current) leads to an increased (decreased) probability of marine heat wave days. Warming and salinification over the shelf is consistent with an enhanced East Australian Current (EAC) ExtensionThe relative strengths of the EAC Extension and Zeehan Current relate significantly to marine heat wave occurrenceEl Niño‐Southern Oscillation and the Tasman Sea Blocking are weakly related to interannual variations in circulation and temperature

The effects of climate-driven ocean change on reef habitat-forming species are diverse1,2and can be deleterious to the structure and functioning of seafloor communities3–5. Although responses of shallow coral- or seaweed-based reef communities to environmental changes are a focus of ecological research in the coastal zone1,4–6, the ecology of habitat-forming organisms on deeper mesophotic reefs remains poorly known. These reefs are typically highly biodiverse7,8and productive as a result of massive nutrient recycling9. Based on seafloor imagery obtained from an autonomous underwater vehicle8, we related change in community composition on deep reefs (30–90 m) across a latitudinal gradient (25–45° S) in southeastern Australia to high-resolution environmental and oceanographic data, and predicted future changes using downscaled climate change projections for the 2060s10–12. This region is recognized as a global hotspot for ocean warming13. The models show an overall tropicalization trend in these deep temperate reef communities, but different functional groups associate differentially to environmental drivers and display a diversity of responses to projected ocean change. We predict the emergence of novel deep-reef assemblages by the 2060s that have no counterpart on reefs today, which is likely to underpin shifts in biodiversity and ecosystem functioning.

During austral summer 2015–2016, prolonged extreme ocean warming events, known as marine heatwaves (MHWs), occurred in the waters around tropical Australia. MHWs arose first in the southeast tropical Indian Ocean in November 2015, emerging progressively east until March 2016, when all waters from the North West Shelf to the Coral Sea were affected. The MHW maximum intensity tended to occur in March, coinciding with the timing of the maximum sea surface temperature (SST). Large areas were in a MHW state for 3–4 months continuously with maximum intensities over 2°C. In 2016, the Indonesian‐Australian Basin and areas including the Timor Sea and Kimberley shelf experienced the longest and most intense MHW from remotely sensed SST dating back to 1982. In situ temperature data from temperature loggers at coastal sites revealed a consistent picture, with MHWs appearing from west to east and peaking in March 2016. Temperature data from moorings, an Argo float, and Slocum gliders showed the extent of warming with depth. The events occurred during a strong El Niño and weakened monsoon activity, enhanced by the extended suppressed phase of the Madden‐Julian Oscillation. Reduced cloud cover in January and February 2016 led to positive air‐sea heat flux anomalies into the ocean, predominantly due to the shortwave radiation contribution with a smaller additional contribution from the latent heat flux anomalies. A data‐assimilating ocean model showed regional changes in the upper ocean circulation and a change in summer surface mixed layer depths and barrier layer thicknesses consistent with past El Niño events. From the southeast tropical Indian Ocean to the Coral Sea, waters across tropical Australia experienced extreme warming during austral summer 2015–2016. This widespread warming was mainly caused by changes in the atmospheric conditions related to the strong El Niño. These regions had reduced cloud cover and weakened winds, leading to enhanced oceanic heating by the atmosphere. In the Indonesian‐Australian Basin and surroundings areas, the shallower mixed layer depth further intensified the near‐surface warming. We used satellite sea surface temperature and a range of instruments to document the warming events near the coast and in the deeper ocean. The warmer than average conditions persisted through austral winter and weakened in spring. During 2016, the southeast tropical Indian Ocean experienced the longest and most intense marine heatwave on record from 1982Around tropical Australia, coastal warming was caused by shortwave radiation and latent heat flux anomalies in January and February 2016The 2015–2016 air‐sea heat flux anomalies, weakened monsoon winds and mixed layer depth anomalies were consistent with past El Niños

The northwest Atlantic Ocean is an important sink for carbon dioxide produced by anthropogenic activities. However the strong seasonal variability in the surface waters paired with the sparse and summer biased observations of ocean carbon makes it difficult to capture a full picture of its temporal variations throughout the water column. We aim to improve the estimation of temporal trends of dissolved inorganic carbon (DIC) due to anthropogenic sources using a new statistical approach: a time series generalization of the extended multiple linear regression (eMLR) method. Anthropogenic increase of northwest Atlantic DIC in the surface waters is hard to quantify due to the strong, natural seasonal variations of DIC. We address this by separating DIC into its seasonal, natural and anthropogenic components. Ocean carbon data is often collected in the summer, creating a summer bias, however using monthly averaged data made our results less susceptible to the strong summer bias in the available data. Variations in waters below 1000m have usually been analyzed on decadal time scales, but our monthly analysis showed the anthropogenic carbon component had a sudden change in 2000 from stationary to an increasing trend at the same rate as the waters above. All depths layers had similar rates of anthropogenic increase of ∼0.57µmol kg−1year−1, and our uncertainty levels are smaller than with eMLR results. Integration throughout the water column (0–3,500 m) gives an anthropogenic carbon storage rate of 1.37 ± 0.57 mol m−2year−1, which is consistent with other published estimates. We need to measure the ocean sink for the CO2emitted by industrialized societies, and it is particularly important for the northwest Atlantic Ocean. The rate of carbon increase is often overshadowed by natural and seasonal variability. We introduce new statistical approaches to better estimate the rate of anthropogenic carbon that has accumulated due to human activities. Ocean carbon data is often collected in the summer, creating a summer bias, however using monthly averaged data made our results less susceptible to the strong summer bias in the available data. From 1993 to 2015 in the northwest Atlantic Ocean, anthropogenic carbon increased at ∼0.57 μmol kg−1year−1within all depth‐layers. Integration of results throughout the water column (0–3,500 m) gives an anthropogenic carbon storage rate of 1.37 ± 0.57 mol m−2year−1. A time series generalization of the extended multiple linear regression (eMLR) method is developed to produce monthly estimates with uncertainties of anthropogenic ocean carbonThe rate of anthropogenic carbon increase in northwest Atlantic is roughly the same for all depth layers, at 0.57 µmol/kg/yearOur method produces estimates of anthropogenic carbon increase that are comparable to those from eMLR, but with smaller uncertainties A time series generalization of the extended multiple linear regression (eMLR) method is developed to produce monthly estimates with uncertainties of anthropogenic ocean carbon The rate of anthropogenic carbon increase in northwest Atlantic is roughly the same for all depth layers, at 0.57 µmol/kg/year Our method produces estimates of anthropogenic carbon increase that are comparable to those from eMLR, but with smaller uncertainties

The Madden‐Julian oscillation (MJO) significantly impacts North Atlantic hurricanes, with increased hurricane activity occurring when the MJO enhances convection over Africa and the tropical Indian Ocean and suppressed hurricane activity occurring when the MJO enhances convection over the tropical Pacific. Using data from 1905 to 2015, we find more tropical cyclones (TCs) make landfall in the continental United States when the MJO enhances tropical Indian Ocean convection. In addition, when the MJO enhances Western Pacific and Western Hemisphere convection, TC activity is preferentially favored in the Caribbean, leading to more Gulf Coast landfalls. As MJO‐enhanced convection moves to the Indian Ocean and Maritime Continent, more storms form in the tropical Atlantic, favoring Florida Peninsula and East Coast landfalls. The MJO's TC steering wind modulation appears to be secondary to its genesis location modulation. The Madden‐Julian oscillation (MJO) is a large‐scale atmospheric signal of winds, rainfall, and surface pressure along the equator that circles the globe every 30–70 days. The MJO's location affects wind patterns that can then increase or decrease North Atlantic hurricane activity. When the MJO increases thunderstorms (convection) over Africa and the Indian Ocean, there tend to be more North Atlantic hurricanes. We show that the MJO patterns that increase North Atlantic hurricane activity also make hurricane landfalls more likely in the continental United States. We are likely to see more landfalls from Texas to the Florida Panhandle than from the Florida Peninsula to Maine when the MJO increases thunderstorms over the Western Pacific and Western Hemisphere. Landfalls from the Florida Peninsula to Maine tend to increase compared with Texas to Florida Panhandle landfalls when the MJO increases thunderstorms over the Indian Ocean. We believe this shift in landfall location is related to where the MJO helps storms form and less related to how the storms move after they form. The Madden‐Julian oscillation (MJO) significantly impacts continental US hurricane landfall frequencyGulf Coast hurricane landfalls are favored when the MJO is enhancing Western Pacific/Western Hemisphere convectionEast Coast hurricane landfalls are favored when the MJO is enhancing Indian Ocean/Maritime Continent convection The Madden‐Julian oscillation (MJO) significantly impacts continental US hurricane landfall frequency Gulf Coast hurricane landfalls are favored when the MJO is enhancing Western Pacific/Western Hemisphere convection East Coast hurricane landfalls are favored when the MJO is enhancing Indian Ocean/Maritime Continent convection

The Tasman Sea is a hot spot of ocean warming, that is linked to the increased poleward influence of the East Australian Current (EAC) over recent decades. Specifically, the EAC produces mesoscale eddies which have significant impacts on the physical, chemical, and biological properties of the Tasman Sea. To effectively consider and explain potential eddy changes in the next 50 years, we use high‐resolution dynamically downscaled climate change simulations to characterize the projected future marine climate and mesoscale eddies in the Tasman Sea through the 2060s. We assess changes in the marine climate and the eddy field using bulk statistics and by detecting and tracking individual eddies. We find that the eddy kinetic energy is projected to increase along southeast Australia. In addition, we find that eddies in the projected future climate are composed of a higher proportion of anticyclonic eddies in this region and that these eddies are longer lived and more stable. This amounts to nearly a doubling of eddy‐related southward temperature transport in the upper 200 m of the Tasman Sea. These changes are concurrent with increases in baroclinic and barotropic instabilities focused around the EAC separation point. This poleward transport and increase in eddy activity would be expected to also increase the frequency of sudden warming events, including ocean temperature extremes, with potential impacts on marine fisheries, aquaculture, and biodiversity off Tasmania's east coast, through direct warming or competition/predation from invasive migrating species. Projected increase in the number and of warm‐core eddies in the EAC ExtensionRelated to changes in baroclinic and barotropic instabilities near the EAC separation pointLeads to nearly a doubling of eddy‐related southward heat transport by the 2060s

The Madden‐Julian oscillation (MJO) has been documented in previous studies to significantly impact tropical cyclone activity in all ocean basins. Most of these studies have utilized the Wheeler‐Hendon index. This index is only available since 1974, the period over which remotely sensed outgoing longwave radiation data has been available. Our study utilizes a long reconstructed MJO index, based on surface pressures, which extends back to 1905. We document consistent modulation of tropical cyclone activity by the MJO in all basins over this time period. These modulations are shown to be remarkably stable over the entire analysis period. We also examine the combined impacts of El Niño‐Southern Oscillation and the MJO on tropical cyclone activity in each basin over multidecadal time scales. Surface pressure derived MJO index used to look at long‐term TC/MJO relationshipModulation of global TC activity by the MJO is stable back to the mid‐1900sMJO and ENSO combined cause large modulations in TC activity for certain basins

Deep convection and associated deep water formation are key processes for climate variability, since they impact the oceanic uptake of heat and trace gases and alter the structure and strength of the global overturning circulation. For long, deep convection in the subpolar North Atlantic was thought to be confined to the central Labrador Sea in the western subpolar gyre (SPG). However, there is increasing observational evidence that deep convection also has occurred in the eastern SPG south of Cape Farewell and in the Irminger Sea, in particular, in 2015–2018. Here we assess this recent event in the context of the temporal evolution of spatial deep convection patterns in the SPG since the mid‐twentieth century, using realistic eddy‐rich ocean model simulations. These reveal a large interannual variability with changing contributions of the eastern SPG to the total deep convection volume. Notably, in the late 1980s to early 1990s, the period with highest deep convection intensity in the Labrador Sea related to a persistent positive phase of the North Atlantic Oscillation, the relative contribution of the eastern SPG was small. In contrast, in 2015–2018, deep convection occurred with an unprecedented large relative contribution of the eastern SPG. This is partly linked to a smaller north‐westward extent of deep convection in the Labrador Sea compared to previous periods of intensified deep convection, and may be a first fingerprint of freshening trends in the Labrador Sea potentially associated with enhanced Greenland melting and the oceanic advection of the 2012–2016 eastern North Atlantic fresh anomaly. The subpolar North Atlantic is one of the few oceanic regions where deep convection occurs, a process by which surface waters become dense enough to mix downwards. Thereby, heat and carbon dioxide are transported into the deep ocean, helping to slow global warming. For long, deep convection in the subpolar North Atlantic was thought to be confined to the central Labrador Sea. However, recently, deep convection was also observed in the Irminger Sea. In this study, we use ocean model simulations to reconstruct when and where deep convection occurred since the mid‐twentieth century. Our simulations show a large variability in strength and spatial patterns of deep convection, with changing relative importance of the Labrador and Irminger Seas. Most notably, in the mid 2010's, deep convection shifted eastwards, so that the relative importance of the Irminger Sea and Labrador Sea were anomalously large and small, respectively. Freshening trends in the Labrador Sea suggest that this shift partially resulted from enhanced melting of Greenland's glaciers and changes in the North Atlantic circulation. Our results stress the need for monitoring and correctly modeling deep convection variability in the Labrador and Irminger Seas to better understand and predict the looming impact of future Greenland melting. Eddy‐rich ocean model simulations (1958–2019) feature large variability of spatial deep convection patterns in subpolar North AtlanticIn 2015–2018, deep convection showed exceptional large and small relative contributions of eastern and western subpolar gyre, respectivelySmall western contribution is potentially associated with enhanced Greenland melting and recent eastern North Atlantic fresh anomaly Eddy‐rich ocean model simulations (1958–2019) feature large variability of spatial deep convection patterns in subpolar North Atlantic In 2015–2018, deep convection showed exceptional large and small relative contributions of eastern and western subpolar gyre, respectively Small western contribution is potentially associated with enhanced Greenland melting and recent eastern North Atlantic fresh anomaly

Seasonal forecasting of tropical cyclones is a topic of considerable interest to the public, government and private sectors. To improve understanding of the dynamics controlling the predictability of tropical cyclone (TC) activity, and improve the accuracy of forecasts, multiple studies have related TC activity to empirically‐defined indices including the El Niño‐Southern Oscillation, the Atlantic Multidecadal Oscillation, and the North Atlantic Oscillation. These indices were not developed to forecast TC activity but rather summarize other aspects of atmosphere‐ocean variability. In this study we use a statistical approach, based on redundancy analysis, to define two indices related to overall activity and steering of TCs. We focus on North Atlantic TCs that reached tropical storm strength (≥34 kt) between August and October 1948–2016. TC occurrences are binned using an equal area grid that covers the North Atlantic. The redundancy indices are linear combinations of mean sea level pressure for the same season. Cross validation is used to guard against over fitting in the definition of the indices. This approach provides two physically interpretable redundancy indices related to North Atlantic TC activity. The leading redundancy index is used to successfully reconstruct the total number of TCs and the accumulated cyclone energy, over the extended period 1878–2014 using seasonal mean sea level pressures from the National Centers for Environmental Prediction 20th‐Century Reanalysis version 2c. Extensions of the approach for seasonal forecasting are discussed. Relationship between North Atlantic tropical cyclone activity and concurrent mean sea level pressure is studied with redundancy analysisPrimary redundancy index is physically interpreted in terms of overall tropical cyclone activity with ∼50%variance explained from 1948 to 2016Secondary redundancy index is physically interpreted in terms of steering of storm tracks toward the coast of North America Relationship between North Atlantic tropical cyclone activity and concurrent mean sea level pressure is studied with redundancy analysis Primary redundancy index is physically interpreted in terms of overall tropical cyclone activity with ∼50%variance explained from 1948 to 2016 Secondary redundancy index is physically interpreted in terms of steering of storm tracks toward the coast of North America

Rainfall in tropical Australia is a critical resource for the agricultural sector. However, its high variability implores improvements in our understanding of its variability. Australian tropical rainfall is influenced by both the Madden‐Julian Oscillation (MJO) on intraseasonal time scales and El Niño‐Southern Oscillation (ENSO) on interannual time scales. This study examines the joint relationship between the MJO, ENSO, and tropical Australian rainfall variability. We analyze daily precipitation data from stations across tropical Australia during the wet season (November to April). The wet season rainfall response to the MJO is found to be greater during El Niño than La Niña. We demonstrate that this relationship is not due to the statistical relationship between the MJO and ENSO indices but instead due to differences in how the MJO modulates the large‐scale circulation during El Niño versus during La Niña. This manuscript presents studies of the simultaneous relationship between tropical Australian rainfall variability and two modes of climate variability: the Madden‐Julian Oscillation (MJO) and El Niño‐Southern Oscillation (ENSO). We analyze daily rainfall at 43 long‐record (1942–2011) weather stations across tropical Australia during the wet season (November to April). The MJO is shown to increase rainfall in certain phases (phases 5 and 6) and decrease it in other phases (phases 2 and 3), but more importantly, this signal is enhanced during El Niño as compared to La Niña. This is demonstrated to be due to how the MJO influences atmospheric circulation differently in El Niño versus La Niña wet seasons. These results are significant in demonstrating that the MJO and ENSO do not act independently on rainfall; the state of both climate modes need to be known in order to better predict rainfall variability. Agricultural decisions are highly dependent on knowledge of rainfall variability, from crop management on intraseasonal scales to crop types on interannual scales. Overall, the results of this study provide a better understanding of wet season rainfall and its potential predictability in tropical Australia, a major exporter of agriculture and livestock, with implications for improved agricultural decision‐making. The Madden‐Julian Oscillation (MJO) modulates tropical Australian rainfallThis response is enhanced in El Niño periodsThe effect of ENSO on the MJO‐rainfall connection is independent of the MJO‐ENSO statistical connection