publications

We investigate changes in the vertical structure of the ocean temperature annual cycle amplitude (TEMP AC ) down to a depth of 300 meters, providing important insights into the relative contributions of anthropogenic and natural influences. Using observations and CMIP6 model simulations, we perform a detection and attribution analysis by applying a standard pattern-based “fingerprint” method to zonal-mean TEMP AC anomalies for three major ocean basins. In all model historical simulations and observational datasets, TEMP AC increases significantly in the surface layer, except in the Southern Ocean, and weakens within the subsurface ocean. There is a decrease in TEMP AC below the annual-mean mixed layer depth, mainly due to a deep-reaching winter warming signal. The temporal evolution of signal-to-noise (S/N) ratios in observations indicates an identifiable anthropogenic fingerprint in both surface and interior ocean annual temperature cycles. These findings are consistent across three different observational datasets, with variations in fingerprint detection time likely related to differences in dataset coverage, interpolation method, and accuracy. Analysis of CMIP6 single-forcing simulations reveals the dominant influence of greenhouse gases and anthropogenic aerosols on TEMP AC changes. Our identification of an anthropogenic TEMP AC fingerprint is robust to the selection of different analysis periods. S/N ratios derived with model data only are consistently larger than ratios calculated with observational signals, primarily due to model-versus-observed TEMP AC differences in the Atlantic. Human influence on the seasonality of surface and subsurface ocean temperature may have profound consequences for fisheries, marine ecosystems, and ocean chemistry.

Gaining insight into anthropogenic influence on seasonality is of scientific, economic and societal importance. Here we show that a human-caused signal in the seasonal cycle of sea surface temperature (SST) has emerged from the noise of natural variability. Geographical patterns of changes in SST seasonal cycle amplitude (SSTAC) reveal two distinctive features: an increase at Northern Hemisphere mid-latitudes related to mixed-layer depth changes and a robust dipole pattern between 40° S and 55° S that is mainly driven by surface wind changes. The model-predicted pattern of SSTAC change is identifiable with high statistical confidence in four observed SST products and in 51 individual model realizations of historical climate evolution. Simulations with individual forcings reveal that GHG increases are the primary driver of changes in SSTAC, with smaller but distinct contributions from anthropogenic aerosol and ozone forcing. The robust human ‘fingerprint’ identified here is likely to have wide-ranging impacts on marine ecosystems.

Temporal and spatial variations in the ocean surface mixed layer are important for the climate and ecological systems. During 1980–2019, the Southern Indian Ocean (SIO) mixed layer depth (MLD) displays a basin‐wide shoaling trend that is absent in the other basins within 40°S–40°N. The SIO MLD shoaling is mostly prominent in austral winter with deep climatology MLD, substantially weakening the MLD seasonality. Moreover, the SIO MLD changes are primarily caused by a southward shift of the subtropical anticyclonic winds and hence ocean gyre, associated with a strengthening of the Southern Annular Mode, in recent decades for both winter and summer. However, the poleward‐shifted subtropical ocean circulation preferentially shoals the SIO MLD in winter when the meridional MLD gradient is sharp but not in summer when the gradient is flat. This highlights the distinct subtropical MLD response to meridional mitigation in winds due to different background oceanic conditions across seasons.

Modeled water-mass changes in the North Pacific thermocline, both in the subsurface and at the surface, reveal the impact of the competition between anthropogenic aerosols (AAs) and greenhouse gases (GHGs) over the past 6 decades. The AA effect overwhelms the GHG effect during 1950–1985 in driving salinity changes on density surfaces, while after 1985 the GHG effect dominates. These subsurface water-mass changes are traced back to changes at the surface, of which \sim70% stems from the migration of density surface outcrops, equatorward due to regional cooling by AAs and subsequent poleward due to warming by GHGs. Ocean subduction connects these surface outcrop changes to the main thermocline. Both observations and models reveal this transition in climate forcing around 1985 and highlight the important role of AA climate forcing on our oceans’ water masses.

Separating the climate response to external forcing from internal climate variability is a key challenge. While most previous studies have focused on surface responses, here we examine zonal-mean patterns of North Pacific subsurface temperature responses. In particular, the changes since 1950 driven by anthropogenic aerosol emissions are found by using a pattern recognition method. Based on the single-forcing large-ensemble simulations from two models, we show that aerosol forcing caused a nonmonotonic temporal response and a characteristic zonal-mean pattern within North Pacific, which is distinct from the pattern associated with internal variability. The aerosol-forced pattern with the nonmonotonic temporal feature shows a substantial temperature change in subpolar regions and a reversed change on the southern flank of the subtropical gyre. A similar characteristic pattern and nonmonotonic time evolution are extracted from the subsurface observations, which likely reflect the subsurface responses to the aerosol forcing, although differences exist with the simulated responses.

During recent decades, both greenhouse gases (GHGs) and anthropogenic aerosols (AAs) drove major changes in the Earth’s energy imbalance. However, their respective fingerprints in changes to ocean heat content (OHC) have been difficult to isolate and detect when global or hemispheric averages are used. Based on a pattern recognition analysis, we show that AAs drive an interhemispheric asymmetry within the 20°‐35° latitude band in historical OHC change due to the southward shift of the atmospheric and ocean circulation system. This forced pattern is distinct from the GHG‐induced pattern, which dominates the asymmetry in higher latitudes. Moreover, it is found that this significant aerosol‐forced OHC trend pattern can only be captured in analyzed periods of 20 years or longer and including 1975–1990. Using these distinct spatiotemporal characteristics, we show that the fingerprint of aerosol climate forcing in ocean observations can be distinguished from both the stronger GHG‐induced signals and internal variability.

The surface ocean mixed layer (OML) is critical for climate and biological systems. Changes in ocean mixed layer depth (MLD) of the Indian Ocean under global warming are examined utilizing outputs from 24 climate models in the Coupled Model Intercomparison Project phase 6 (CMIP6) models and the Community Earth System Model 1.0 with Community Atmosphere Model version 5 (CESM1–CAM5). The results show that the MLD generally decreases in low- and high-emissions Shared Socioeconomic Pathway (SSP) scenarios (ssp126 and ssp585). In ssp126 and ssp585, the multi-model ensemble-mean OML, respectively shoals about 5 and 10% over both the northern tropics and southern subtropics, with high model consistency. This robust OML shoaling appears in the 1980s and is closely associated with increased surface buoyancy forcing and weakened winds. In contrast, the OML in the south equatorial Indian Ocean slightly deepens and displays large intermodel differences in the sign and magnitude of the changes. The effects of direct CO 2 increase and wind changes on OML changes are further quantified by CESM1–CAM5 partially coupled experiments. The results show that the increased surface net heat flux from direct CO 2 increase dominates OML shoaling in the northern tropics. In the southern subtropics, the increased surface heat flux, reduced wind speed, and wind-driven divergence all facilitate the OML shoaling. In the south equatorial Indian Ocean, wind changes generally deepen the OML, consistent with the CMIP6 results. Moreover, the OML shoaling-related upper ocean stratification changes are contributed by both temperature and salinity changes in the northern tropics but dominated by temperature changes south of 10°S. These results highlight the regional differences in MLD changes and their forcing, which is important for understanding regional climate changes and corresponding changes in extreme events and biological systems under global warming.

Unlike greenhouse gases (GHGs), anthropogenic aerosol (AA) concentrations have increased and then decreased over the past century or so, with the timing of the peak concentration varying in different regions. To date, it has been challenging to separate the climate impact of AAs from that due to GHGs and background internal variability. We use a pattern recognition method, taking advantage of spatiotemporal covariance information, to isolate the forced patterns for the surface ocean and associated atmospheric variables from the all-but-one forcing Community Earth System Model ensembles. We find that the aerosol-forced responses are dominated by two leading modes, with one associated with the historical increase and future decrease of global mean aerosol concentrations (dominated by the Northern Hemisphere sources) and the other due to the transition of the primary sources of AA from the west to the east and also from Northern Hemisphere extratropical regions to tropical regions. In particular, the aerosol transition effect, to some extent compensating the global mean effect, exhibits a zonal asymmetry in the surface temperature and salinity responses. We also show that this transition effect dominates the total AA effect during recent decades, e.g., 1967–2007.

How the ocean circulation changes in a warming climate is an important but poorly understood problem. Using a global ocean model, we decompose the problem into distinct responses to changes in sea surface temperature, salinity, and wind. Our results show that the surface warming effect, a robust feature of anthropogenic climate change, dominates and accelerates the upper ocean currents in 77% of the global ocean. Specifically, the increased vertical stratification intensifies the upper subtropical gyres and equatorial currents by shoaling these systems, while the differential warming between the Southern Ocean upwelling zone and the region to the north accelerates surface zonal currents in the Southern Ocean. In comparison, the wind stress and surface salinity changes affect regional current systems. Our study points a way forward for investigating ocean circulation change and evaluating the uncertainty.

The Southern Ocean (>30° S) has taken up a large amount of anthropogenic heat north of the Subantarctic Front (SAF) of the Antarctic Circumpolar Current (ACC). Poor sampling before the 1990s and decadal variability have heretofore masked the ocean’s dynamic response to this warming. Here we use the lengthening satellite altimetry and Argo float records to show robust acceleration of zonally averaged Southern Ocean zonal flow at 48° S–58° S. This acceleration is reproduced in a hierarchy of climate models, including an ocean-eddy-resolving model. Anthropogenic ocean warming is the dominant driver, as large (small) heat gain in the downwelling (upwelling) regime north (south) of the SAF causes zonal acceleration on the northern flank of the ACC and adjacent subtropics due to increased baroclinicity; strengthened wind stress is of secondary importance. In Drake Passage, little warming occurs and the SAF velocity remains largely unchanged. Continued ocean warming could further accelerate Southern Ocean zonal flow.

Observations show that since the 1950s, the Southern Ocean has stored a large amount of anthropogenic heat and has freshened at the surface. These patterns can be attributed to two components of surface forcing: poleward-intensified westerly winds and increased buoyancy flux from freshwater and heat. Here we separate the effects of these two forcing components by using a novel partial-coupling technique. We show that buoyancy forcing dominates the overall response in the temperature and salinity structure of the Southern Ocean. Wind stress change results in changes in subsurface temperature and salinity that are closely related to intensified residual meridional overturning circulation. As an important result, we show that buoyancy and wind forcing result in opposing changes in salinity: the wind-induced surface salinity increase due to upwelling of saltier subsurface water offsets surface freshening due to amplification of the global hydrological cycle. Buoyancy and wind forcing further lead to different vertical structures of Antarctic Circumpolar Current (ACC) transport change; buoyancy forcing causes an ACC transport increase (3.1 ± 1.6 Sv; 1 Sv ≡ 10 6 m 3 s −1 ) by increasing the meridional density gradient across the ACC in the upper 2000 m, while the wind-induced response is more barotropic, with the whole column transport increased by 8.7 ± 2.3 Sv. While previous research focused on the wind effect on ACC intensity, we show that surface horizontal current acceleration within the ACC is dominated by buoyancy forcing. These results shed light on how the Southern Ocean might change under global warming, contributing to more reliable future projections.

Ocean uptake of anthropogenic heat over the past 15 years has mostly occurred in the Southern Ocean, based on Argo float observations. This agrees with historical simulations from phase 5 of the Coupled Model Intercomparison Project (CMIP5), where the Southern Ocean (south of 30°S) accounts for 72% ± 28% of global heat uptake, while the contribution from the North Atlantic north of 30°N is only 6%. Aerosols preferentially cool the Northern Hemisphere, and the effect on surface heat flux over the subpolar North Atlantic opposes the greenhouse gas (GHG) effect in nearly equal magnitude. This heat uptake compensation is associated with weakening (strengthening) of the Atlantic meridional overturning circulation (AMOC) in response to GHG (aerosol) radiative forcing. Aerosols are projected to decline in the near future, reinforcing the greenhouse effect on the North Atlantic heat uptake. As a result, the Southern Ocean, which will continue to take up anthropogenic heat largely through the mean upwelling of water from depth, will be joined by increased relative contribution from the North Atlantic because of substantial AMOC slowdown in the twenty-first century. In the RCP8.5 scenario, the percentage contribution to global uptake is projected to decrease to 48% ± 8% in the Southern Ocean and increase to 26% ± 6% in the northern North Atlantic. Despite the large uncertainty in the magnitude of projected aerosol forcing, our results suggest that anthropogenic aerosols, given their geographic distributions and temporal trajectories, strongly influence the high-latitude ocean heat uptake and interhemispheric asymmetry through AMOC change.