Emergence of the North Pacific heat storage pattern delayed by decadal wind-driven redistribution

Emergence of the North Pacific heat storage pattern delayed by decadal wind-driven redistribution | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Emergence of the North Pacific heat storage pattern delayed by decadal wind-driven redistribution Yuanlong Li, Jing Duan, Yilong Lyu, Zhao Jing, Fan Wang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4905116/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 14 Jan, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Storage of anthropogenic heat in the oceans is spatially inhomogeneous, impacting regional climates and human societies. Climate models project enhanced heat storage in the mid-latitude North Pacific (MNP) and much weaker storage in the tropical Pacific. However, the observed heat storage during the past half-century shows a more complex pattern, with limited warming in the central MNP and enhanced warming in the northwest tropical Pacific. Based on observational datasets, ocean model experiments, and climate models, we show that emergence of human-induced heat storage is likely postponed in the North Pacific by natural variability to the late-21st century. Specifically, phase shifts of the Pacific Decadal Oscillation (PDO) have vitally contributed to trends in the North Pacific winds during recent decades. Changes in surface winds drove meridional heat redistribution via Rossby wave dynamics, leading to regional warming and cooling structures and a more complex historical heat storage than models project. Despite this, enhanced anthropogenic warming has already been emerging in marginal seas along the North Pacific basin rim, for which we shall prepare for the pressing consequences such as increasing marine heatwaves. Earth and environmental sciences/Ocean sciences/Physical oceanography Earth and environmental sciences/Climate sciences/Ocean sciences/Physical oceanography Earth and environmental sciences/Climate sciences/Climate change/Attribution Earth and environmental sciences/Climate sciences/Climate change/Climate and Earth system modelling Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction The storage of excess heat caused by anthropogenic greenhouse warming in oceans 1 is geographically inhomogeneous 2–4 . Climate models 5,6 suggest that the Southern Ocean 3,7–9 , the North Atlantic 3,8 , and the mid-latitude North Pacific (MNP) warm up more quickly than other oceans ( Fig. 1a ). These ocean warming features accompany accelerated ocean currents 8,10 , poleward shift of storm tracks and westerlies 11,12 , increased extreme atmospheric rivers 13 , and altered marine biodiversity patterns 14 . While the enhanced heat storage in the Southern Ocean and North Atlantic has been witnessed in nature, that in the MNP has hardly emerged ( Fig. 1b ). Indeed, despite significant warming trends in marginal seas along the basin rim, there were weak warming or cooling trends in the central MNP since the 1950s. The tropical Pacific is among the regions with the weakest heat storage in CMIP6 multi-model mean (MMM) ( Fig. 1a ). However, the Northwest Tropical Pacific (NWTP) has exhibited the strongest warming rate of the Pacific Ocean during the past decades ( Fig. 1b ). Increased ocean heat content (OHC) in the NWTP has led to rapid sea-level rise 15,16 , increasing marine heatwave and coral bleaching events 17–19 , and altered tropical cyclone behaviors 20–22 . It also exerted impacts on the downstream marginal seas through western boundary currents 23–25 and the Indian Ocean through the Indonesian Throughflow 26,27 . The Pacific is projected to be the leading reservoir of anthropogenic heat among all oceans by the latter half of the 21 st century 3 . Correctly interpreting the model-observation discrepancies there represents a vital scientific issue. The contrasts between observations and models indicate either systematic model biases in simulating the externally forced heat storage or substantial impacts by natural variability. Although anthropogenic fingerprints have emerged in many aspects of the ocean 3,28,29 , natural variability remains influential in the observed OHC changes 30,31 . For instance, the persistent negative phase of the Pacific Decadal Oscillation (PDO) since 1998 led to heat pile-up in the western tropical Pacific via the strengthened Pacific Walker Circulation 32–34 . If the model-observation discrepancies can be successfully attributed to natural variability, we shall expect an acceleration of warming in the MNP within the coming decades to catch up with the projected rate. Here, we set out to quantify the contributions of various processes to the observed North Pacific heat storage pattern, addressing in particular whether natural variability can largely explain the model-observation discrepancy. Our analysis is based on 1) five observational datasets 35–39 to characterize the historical heat storage, 2) experiments of a forced ocean model to isolate effects of surface heat fluxes and wind-driven ocean dynamics, and 3) 20 CMIP6 models to estimate anthropogenic fingerprints and when they may emerge (Methods). We demonstrate that surface wind changes arising from phase shifts of the PDO have driven basin-scale heat redistribution through Rossby waves and modulations in western boundary currents. This effect complicates the observed heat storage pattern by creating regional warming/cooling structures that conceal anthropogenic fingerprints. According to model projections, the human-induced heat storage pattern will likely hide until the late-21 st century. These results provide useful implications for climate prediction in the North Pacific and marginal seas. Heat storage in the North Pacific We begin with more details of the simulated and observed heat storage patterns in the North Pacific during the 1958–2021 period. In the multi-model mean (MMM) of 20 CMIP6 models, consisting of the historical and Shared Socioeconomic Pathway (SSP) 5-8.5 simulations before and after 2014, respectively, the entire MNP is characterized by enhanced 0-2000 m heat storage, in stark contrast to the warming minimum in the western tropical Pacific (Fig. 1 a). The observation-based heat storage shows more regional structures (Fig. 1 b and Extended Data Fig. 1 ); for example, there are alternating warming and cooling regions in the North Pacific, particularly in the western basin. The NWTP (e.g., 125°-180°E, 8°-18°N) stands out with a warming maximum, whereas a “warming hole” occurs in the MNP interior with insignificant trends and is surrounded by significant warming trends in marginal seas, such as the Bering Sea and the Gulf of Alaska. There is another cooling region near the western boundary of the 18°-30°N, sandwiched by the warming areas of the NWTP and the Kuroshio extension (KE) between 30°-40°N. Cooling trends are also seen beyond the North Pacific, such as the subpolar North Atlantic, southwestern subtropical Pacific, and southwestern subtropical Indian Ocean. CMIP6 MMM also show cooling or slackened warming in these regions (Fig. 1 a). Albeit with weaker intensity and smaller spatial range, the subpolar North Atlantic warming hole 40 is clearly discernible in CMIP6 MMM. Therefore, models and observations are broadly consonant in all major ocean basins except for the North Pacific. The model-observation discrepancies in the North Pacific heat storage are worthy of in-depth investigation. We further examine the temporal evolution of OHC in key regions (Fig. 1 c, d). CMIP6 models suggest an average heat storage rate of 0.72 ± 0.62 W m − 2 (± indicates the one standard deviation range of 20 models) in the MNP of 155°E-150°W, 40°-55°N, close in magnitude to that of the Southern Ocean (0.68 ± 0.27 W m − 2 in MMM for 55°-33°S). In comparison, a much weaker rate of 0.17 ± 0.08 W m − 2 in the MNP is obtained from observation-based datasets (Fig. 1 c; ± indicates the one standard deviation range of 5 datasets). Meanwhile, the NWTP shows enhanced warming of 0.58 ± 0.26 W m − 2 in observations, one order stronger than the simulated rate of 0.06 ± 0.32 W m − 2 in CMIP6 models (Fig. 1 d). Interestingly, CMIP6 agrees with observation in the OHC change of the entire North Pacific (120°E-80°W, 0°-60°N), which are 0.33 ± 0.07 W m − 2 and 0.33 ± 0.17 W m − 2 , respectively, implying a heat redistribution over the North Pacific in the observed realization relative to the simulated pattern. The observed OHCs show prominent interannual and decadal variabilities in the MNP and NWTP (Fig. 1 c, d). Meanwhile, we notice a large inter-model spread relative to the MMM change in the historical simulation of CMIP6 before 2014, which also indicates significant influence from natural variability. The effect of decadal natural variability is particularly notable from the 1990s through the mid-2010s. During this period, the observed OHC of the NWTP was substantially elevated and exceeded the + 1 standard deviation range of CMIP6 models, while that of the MNP did not rise significantly as models expected. In the following, we examine how decadal natural variability affects the North Pacific heat storage. Heat redistribution driven by wind changes To understand how the heat storage pattern is formed, we performed experiments using the Hybrid Coordinate Ocean Model (HYCOM) with a coarse horizontal resolution of 0.5° (Methods) and prescribed reanalysis of atmospheric fields 41 as the surface forcing. Despite regional simulation errors, the control run (CTRL) of HYCOM captured broad-scale features of the observed heat storage – the cooling trends in the central MNP and the enhanced warming of the NWTP ( Fig. 2 a). CTRL also well reproduced the prominent interannual and decadal variabilities (Fig. 2 d, e). During the simulation period of 1958–2019, the correlation coefficients between CTRL and IAP are 0.57 and 0.81 for OHCs in the MNP and NWTP, respectively, both significant at the 99% confidence level. With the aid of HYCOM experiments, we can separate the effects of heat redistribution induced by wind-driven ocean circulation changes and heat uptake through surface heat fluxes in heat storage. The heat-flux run (HTFL) retains changes in surface heat fluxes and keeps wind stress and precipitation invariant (Methods), representing the effect of heat uptake. HTFL produces substantially stronger warming in the MNP than in the tropical Pacific (Fig. 2 b) – a pattern resembling CMIP6 MMM (Fig. 1 a). This highlights the dominance of heat uptake in shaping the model-projected heat storage pattern. Alternatively, the wind run (WND), retaining changes only in surface wind stress and keeping other forcing fields unchanged (Methods), produces basin-wide cooling trends in the MNP and enhanced warming of the NWTP (Fig. 2 c). These wind-driven changes greatly modify the pattern shaped by heat uptake and vitally contribute to the total storage in CTRL. Checking the temporal evolution clearly suggests that while heat fluxes drive quasi-monotonic warming trends in both regions (stronger in MNP), the interannual and decadal fluctuations arise mainly from winds (stronger in NWTP) (Fig. 2 d, e). Critically, winds have induced an OHC decrease in the MNP and an abrupt OHC increase in the NWTP since the late 1990s, which greatly altered the overall trends of 1958–2019. Then, we explore changes in surface winds. Reanalysis datasets 41–44 suggest basin-scale trends of anti-cyclonic winds over the North Pacific since 1958 (Fig. 3 a). Correspondingly, there are easterly winds and negative Ekman pumping velocity ω E (indicating downwelling; Methods) in the NWTP, which causes convergence of the upper-layer warm water 45 and enhances the heat storage there. Meanwhile, we also see westerly winds and positive w E (upwelling) in the eastern tropical Pacific, which might also affect the NWTP through Rossby waves. However, this effect fails to dampen the NWTP warming induced by local wind changes, probably owing to the dissipation of Rossby waves during their transition across the Pacific basin 46 . The anti-cyclone also involves strengthening westerlies north of 40°N, which slackens the heat storage in the subpolar North Pacific through Ekman upwelling and enhances heat pile-up in the 30°-40°N band through Ekman downwelling. In CMIP6 MMM, trends of anti-cyclonic winds are confined north of 30°N and westerly trends occupy the entire tropical Pacific (Fig. 3 b). Therefore, some key features of the observed wind trend are missed, including the easterlies in the NWTP and westerlies in the subpolar North Pacific. These differences in surface winds, critically accounting for the model-observation discrepancies in heat storage, likely arise from natural variability, given that the CMIP6 MMM is assumed to represent externally forced changes. The PDO, as the leading mode of decadal natural variability in the North Pacific, was in its positive phase during 1977–1997 and then shifted to its negative phase of 1998–2014 ( Extended Data Fig. 2 ). A regression onto the negative PDO index (-PDO) (Fig. 3 c) shows easterlies in the NWTP and westerlies in the northeastern subpolar Pacific – key features seen in observation but missed in CMIP6 MMM. The negative PDO also induces positive w E anomalies in the western and central parts of the 18°-30°N band, causing the cooling trends observed near the western boundary (Fig. 1 b). The regression of OHC onto the negative PDO shows some features resembling the observed heat storage, particularly the warming in the NWTP and the 30°-40°N band and cooling in the MNP and the 18°-30°N band (Fig. 3 d). The NWTP heat content shows correlation coefficients of -0.42 and − 0.66 with the 8-year low-passed PDO index and the unfiltered December-January-February Oceanic Niño Index (ONI), respectively ( Extended Data Fig. 2 a). This suggests a strong modulation of natural variability on the NWTP on interannual and decadal timescales, with heat pile-up in the NWTP under negative PDO and La Niña conditions. In the mid-latitudes, the regression cannot explain the observed heat storage. For example, cooling trends occur mainly in the marginal seas in the regression (Fig. 3 d) rather than in the MNP interior as in the trend pattern (Fig. 1 b); the meridional dipole-like structure observed east of Japan, linked to the strengthening geostrophic transport of the KE jet (Fig. 1 b), is replaced by prevailing warming in Fig. 3 d. These discrepancies can be reconciled by considering the time lag of oceanic response to wind forcing through planetary wave adjustments 47 . Lagged regressions show that the strengthened KE east of Japan is established ~ 4 years after the negative peak of PDO, and the central MNP cooling takes ~ 9 years (Fig. 3 e, f). The wind-driven OHC anomaly of the MNP, measured by the WND run of HYCOM, lags the PDO index ( Extended Data Fig. 2 b): the PDO shifted from the negative to the positive phase during the 1970s, and correspondingly there was a warming trend of MNP throughout the 1980s; subsequently, the opposite transition from the late 1990s through ~ 2010 led to a cooling trend of the MNP persisting through the late 2010s. Regional ocean dynamics The above analysis points to the vital role of PDO phase shifts in shaping the historical heat storage pattern through wind-driven redistribution. One question arises as to whether models can correctly simulate the PDO-induced variability, which is critical for predicting regional OHC changes and their climatic and environmental impacts in the upcoming decades. This remains a challenging task for the state-of-the-art models. Even with prescribed reanalysis winds, our 0.5° simulation of HYCOM fails to fully reproduce the observed heat storage in the MNP (Fig. 2 a); for instance, the simulated cooling in the central MNP is stronger in amplitude, broader in spatial extent, and shifted to lower latitudes in CTRL compared to that in observation. These discrepancies probably arise from complex regional ocean dynamics that are not properly represented by coarse-resolution models - such as the “bimodal” variability of the KE 47,48 . In the mid-latitudes, a negative phase of the PDO drives downwelling Rossby waves through negative w E anomalies between 150°-140°W (Fig. 3 c). Along the latitudinal band of the KE (e.g., 31°-36°N), these Rossby waves propagate across the Pacific basin to the KE region east of Japan by ~ 4 years, as manifested in satellite-based sea surface height (SSH) anomalies (Fig. 4 ). These downwelling waves lead to a “stable” state of the KE system characterized by a strengthened KE jet and weakened mesoscale eddy variability 47–49 , as indicated by the enhanced anticyclonic “southern recirculation” locating south of the jet 50,51 (Fig. 4 a). The strengthened geostrophic KE jet manifests as a meridional dipole in OHC east of Japan (Fig. 3 e). In addition to PDO, the KE is also modulated by the upstream Kuroshio path south of Japan. The PDO shifted to a positive phase in ~ 2014, and correspondingly, an unstable dynamical state of the KE was established in early 2017, as indicated by the weakened southern recirculation (Fig. 4 a). However, the occurrence of the Kuroshio large meander in August 2017 - with the Kuroshio following an offshore meandering path in the Shikoku basin south of Japan 52 – compelled the KE to switch back to the stable state in late 2017 and remain stable since then 48,53 . Albeit interrupted in 2017, there has been a persistently stable KE since ~ 2009. This led to an overall trend of the KE toward its stable state during the past decades, given the more unstable KE during the 1980s and 1990s due to the positive PDO phase ( Extended Data Fig. 3 ). This trend manifests as a dipole-like structure in heat storage east of Japan (Fig. 1 b). By affecting the overlying storm tracks and large-scale winds, the stable KE also contributed to the cooling in the central MNP and the warming in the eastern basin 48,54 . The response of KE to Rossby waves is successfully reproduced by a 0.1° simulation of HYCOM (Methods; Extended Data Fig. 3 ) that can better resolve the Kuroshio path and mesoscale eddies. As a result, this 0.1° simulation can realistically represent the heat redistribution driven by PDO winds, including the meridional dipole east of Japan with a lag time of ~ 4 years (Fig. 5 a) and the cooling in the central MNP with a lag time of ~ 9 years (Fig. 5 b) to the negative peak of PDO. This ~ 9-year lag time reflects the slow propagation of upwelling Rossby waves between 40°-55°N ( Extended Data Fig. 4 ) generated by positive w E anomalies near the west coasts of North America (Fig. 3 c). There are also detailed discrepancies between the 0.1° simulation and observation in the KE region (Fig. 5 a, b compared to Fig. 3 e, f), which may result from the lack of feedback between mesoscale eddies and the atmosphere in our standard-alone HYCOM simulation. The eddy-atmosphere feedback has been demonstrated to efficiently damp mesoscale eddies and amend the simulated strength and pathway of the KE 55 . Most CMIP6 models are of horizontal resolutions of ~ 50 or ~ 100 km in their ocean components, close to our 0.5° HYCOM simulation. We examine how the PDO-induced heat redistribution is represented in these models. The regression using individual simulations of CMIP6 models (Methods) suggests that CMIP6 models can capture some large-scale features of the PDO-induced changes in surface winds and OHC, particularly in the mid-latitudes (Fig. 5 c, d and Extended Data Fig. 5 ). The cooling of the central MNP is seen at a lag of ~ 9 years, albeit with extended spatial range (Fig. 5 d). The PDO’s signatures in the NWTP are weaker in models than in observations, linked to the uncertain tropical Pacific wind changes ( Extended Data Fig. 5 ). This indicates a weaker coupling between the PDO and the tropical Pacific climate in models 56 . By prescribing the observed (reanalysis) surface winds, the 0.5° HYCOM simulations can well reproduce the response of NWTP changes associated with the PDO winds ( Extended Data Fig. 6 ). Furthermore, the improved fidelity of the 0.1° HYCOM (Fig. 5 a, b) relative to the 0.5° HYCOM and CMIP6 models highlights the necessity of fine resolution to account for the complex ocean dynamics involved in heat redistribution. To further confirm the PDO’s impact, we select 5 model realizations with the simulated PDO showing the highest positive correlations with the observed PDO (+ 5 models) and 5 model realizations showing the highest negative correlation with the observation (-5 models). As such, contrasting the + 5 and − 5 models mimics impacts of the observed PDO phase shifts on the OHC trends since 1958 ( Extended Data Fig. 7 ). The composite of + 5 minus − 5 models reassures that the PDO phase shifts in observation can dampen the heat storage in the MNP and enhance the warming of the NWTP ( Extended Data Fig. 7c ) through wind-driven heat redistribution ( Extended Data Fig. 7d ). Therefore, despite unresolved regional structures due to coarse model resolutions, discrepancies between CMIP6 models and observations in the North Pacific heat storage pattern, particularly the contrasts in the NWTP and the central MNP, are primarily explained by natural variability represented by the PDO. Emergence of anthropogenic heat storage The above analysis indicates that the emergence of anthropogenic heat storage has been delayed by natural variability in some key regions of the North Pacific and will eventually take place in the future as the emission of greenhouse gases continues. Therefore, it is instructive to estimate the time of emergence (ToE) of anthropogenic heat storage based on CMIP6 simulations (Fig. 6 ), although we fully acknowledge that such estimation is subject to potential influence from inaccurate ocean dynamics as discussed earlier. Here, we utilize the piControl simulation, historical simulation, and projections under three emission scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5) of 20 CMIP6 models to determine the “signal-to-noise” ratio and compute the ToE for the 0-2000 m OHC change (Methods 28,57 ). Under the “middle-of-the-road” scenario of SSP2-4.5, a scenario viewed as where social and economic trends do not shift markedly from historical patterns, the anthropogenic heat storage (Fig. 6 a) has already emerged (e.g., ToE earlier than the 2020s) in the Southern Oceans between 55°-33°S and the tropical and northern Atlantic. In the North Pacific, the ToE is as early as the 2010s in marginal seas along the basin rim, including the western coasts of North America, the Bering Sea, the Okhotsk Sea (most of which is shallower than 2000 m and not estimated for OHC and ToE), and the Japan Sea (Fig. 6 a). By contrast, anthropogenic signals have not emerged by present (the 2020s) in the basin interior and even until the 2050s in the central MNP, the KE, and the NWTP where we see obvious model-observation discrepancies. The ToEs under SSP1-2.6 and SSP5-8.5 scenarios are similar in spatial distribution to those under SSP2-4.5 but postponed and advanced in average time ( Extended Data Fig. 8 ), respectively. In marginal seas of the North Pacific, such as the Bering Sea and the eastern boundary coasts, anthropogenic heat storage has already been emerging. The multi-model median ToE is during the 2010s and 2020s in these regions regardless of the emission scenario (Fig. 6 b, c). By contrast, the central MNP, KE, and NWTP regions show delayed ToEs and relatively high sensitivities to the emission scenario (Fig. 6 d-f). Under the SSP1-2.6 and SSP2-4.5 scenarios, anthropogenic heat storage is projected to emerge in the latter half of this century in these regions, such as the 2080s in the KE region. Under the SSP5-8.5 scenario, the ToE is as early as the 2040s in the central MNP, the 2050s in the KE, and the 2060s in the NWTP. To summarize, despite uncertainties arising from inter-model spread, the North Pacific heat storage is projected to remain under a significant impact of natural variability until the late 21st century, except for marginal seas along the basin rim. Therefore, we expect that the heat storage pattern of the North Pacific will alter but probably still differ from that projected by climate models in the upcoming several decades. Summary and Implications Based on observational datasets, ocean model experiments, and CMIP6 model simulations, we demonstrate a strong modification effect by the PDO on the observed heat storage pattern in the North Pacific since the mid-20th century. Specifically, phase shifts of the PDO in recent decades have altered the trends in surface winds over the North Pacific. Surface wind changes induced meridional heat redistribution through Rossby waves and variability of the KE system, effectively erasing the warming in the central MNP and fueling the heat pile-up in the NWTP. These effects led to a more complex heat storage pattern in observations than in models by creating regional warming/cooling structures that mask the human-induced fingerprints ( Extended Data Fig. 9 ). Recognition of the strong influence of PDO on heat storage provides critical insights into near-term climate prediction. The persistent negative phase of PDO since 1998, critical in shaping the multi-decadal trends of surface winds and heat storage, terminated in the mid-2010s (Fig. 4 c). If the PDO switches to a positive phase in the upcoming decade, the central MNP “warming hole” shall vanish soon and the basin-scale heat storage will evolve toward the pattern projected by models. However, the recent triple La Niña events in 2020-2023 58,59 and the possible developing La Niña condition in 2024 and their extratropical impacts indicate a likelihood of a lengthened negative phase, which acts to maintain the existing heat storage pattern. The broad impacts of the PDO on surface atmospheric and ocean conditions have been well established, such as storm tracks, atmospheric rivers, and Pacific salmon production 60–62 . Here, we add that the PDO is also critical for the transient response of the North Pacific to anthropogenic forcing. Through modulating wind-driven ocean circulations, the PDO fundamentally regulates the distributions of anthropogenic heat, CO 2 , and the Fukushima nuclear effluents in the North Pacific and their spread into the Indian and Arctic Oceans via the Indonesian through-flow 26,27 and Bering Strait 63 , respectively These implications highlight the urgent need for an accurate initialized prediction of the PDO and its far-reaching impacts. Climate models tend to underestimate the PDO variability and its impacts over the historical period 64 . As a result, none of the 20 CMIP6 models has produced a historical heat storage pattern resembling the observation (not shown). Increasing model resolution seems promising. The representation of natural variability in the North Pacific is considerably improved in ~ 10 km simulations compared to ~ 100 km simulations 65,66 ; our analysis based on the 0.1° HYCOM simulation also supports this notion. In addition, increasing ensemble members can better account for the extreme cases of natural variability, such as the 1998–2014 negative phase 67 , and improve the ToE estimate. The human-induced heat storage pattern in the Pacific Ocean is projected to eventually emerge by the late 21st century, which is much later than in the Atlantic and Southern Oceans. Apart from the strong impacts of natural variability, the postponed ToEs in the Pacific are linked to the lack of a deep-reaching meridional overturning circulation, which also holds true for the Indian Ocean. In the North Atlantic and Southern Oceans, overturning circulations can efficiently communicate anthropogenic heat to the deep ocean, enhancing the average heat uptake 68 and the signal-to-noise ratio in heat storage. Nevertheless, human-induced warming has already been emerging in marginal seas along the North Pacific basin rim, such as the Bering Sea, Okhotsk Sea, Japan Sea, Gulf of Alaska, and western coasts of North America. Increasing frequency and severity of marine heatwaves have been extensively reported recently 69–71 , with adverse influences on marine ecosystems and regional socioeconomics 72,73 . Anthropogenic fingerprints in these marginal seas could be enhanced by influence from lands where the air temperature warms quicker than in oceans 74 or coastal processes that are insufficiently resolved by climate models. This calls for extensive investigations to reveal the mechanisms underlying the warming of marginal seas and multi-disciplinary implications, which shall aid the prediction and decision-making. Methods Datasets. To estimate the 0-2000 m ocean heat content (OHC) trend since the mid-20th century, ocean temperature data since 1958 from four observational analyses and one reanalysis product are utilized. The four observational analyses are the Institute of Atmospheric Physics ocean analysis (IAP) for 1958–2021 provided by the Chinese Academy of Sciences 35 , the World Ocean Atlas (WOA) for 1958–2018 provided by the National Oceanic and Atmospheric Administration’s (NOAA’s) National Centers for Environmental Information of the United States 37 ; the Ishii analysis for 1958-2021 36 ; and the version 4.2.2 of the Met Office Hadley Centre ‘‘EN’’ series of datasets (EN4) for 1958–2021 provided by the Met Office of the United Kingdom 38 . The one ocean reanalysis product is the Ocean Reanalysis System 4 (ORA-S4) for 1958-2017 39 from the European Centre for Medium-Range Weather Forecasts (ECMWF). All five datasets provide 1°×1° ocean temperature fields. WOA provides pentad-mean data in annual intervals, and monthly fields of IAP, Ishii, EN4, and ORA-S4 are averaged into annual-mean data. Surface winds of four atmospheric reanalysis products are used, including the 1.25°×1.25° Japanese 55-year Reanalysis (JRA55) product for 1958-2021 41 , the 1°×1° ORA-S4 product as a combination of the 40-year ECMWF Re-Analysis (ERA-40) of 1958-1988 42 and the ECMWF Interim Re-Analysis (ERA-Interim) of 1989-2017 43 , the 2.5°x2.5° National Centers for Environmental Prediction-National Center for Atmospheric Research (NCEP-NCAR) reanalysis for 1958-2021 44 , and the 0.25°x0.25° fifth generation ECMWF reanalysis (ERA5) for 1958-2021 75 . To examine the changes in the Kuroshio Extension (KE) jet, the 0.25°x0.25° sea surface height data distributed by the Copernicus Marine and Environment Monitoring Service (CMEMS) for 1993–2021 is used. The coupled model simulations. Monthly outputs of the pre-industrial control (piControl) (the last 100 years), historical (from 1958 to 2014), Shared Socioeconomic Pathways 126 (SSP1-2.6), SSP2-4.5, and SSP5-8.5 (from 2015 to 2100) simulations of twenty models of the Coupled Model Intercomparison Project Phase 6 (CMIP6) are analyzed. These models include the following: ACCESS-CM2, ACCESS-ESM1-5, BCC-CSM2-MR, CAMS-CSM1-0, CESM2-WACCM, CNRM-CM6-1, CanESM5, EC-Earth3, EC-Earth3-Veg, FGOALS-g3, FIO-ESM-2-0, GFDL-ESM4, GISS-E2-1-G, INM-CM4-8, INM-CM5-0, MPI-ESM1-2-LR, MRI-ESM2-0, NESM3, NorESM2-LM, NorESM2-MM. The multi-model mean (MMM) of 20 CMIP6 models represents the externally forced trend, and the inter-model spread represents the natural climate variability and difference in model physics. Monthly fields of CMIP6 simulations are interpolated onto the same set of 1°×1° grids and averaged into annual-mean data to match observational datasets. HYCOM experiments. We use the Hybrid Coordinate Ocean Model (HYCOM) version-2.3.01 76 to simulate and understand the ocean heat storage pattern. The HYCOM 0.5° simulation adopts a quasi-global ocean domain (78°S-75°N, 30°-389.5°E) with resolutions of 0.5° × 0.5° × 50 levels. The model is spun up from a state of rest for 600 years under the daily JRA55 atmospheric forcing of 1958, and the atmospheric surface forcing fields include the wind stress, net shortwave and longwave radiations (SWR and LWR), wind speed, precipitation, and air temperature and humidity. The control run (CTRL) is forced with realistic daily atmospheric forcings of JRA55 reanalysis from 1958 to 2019. CTRL contains complete physical processes of forming ocean heat storage pattern and is compared with observational data to evaluate the model performance. Two sensitivity runs are conducted to explore the effects of different drivers on the ocean heat storage pattern. Only wind stress adopts realistic daily fields in wind stress run (WND), while all the other forcings, such as wind speed and radiations, are fixed to the 1958 fields as in spin-up. In the WND, OHC trends primarily arise from heat redistribution due to wind stress-driven changes in ocean circulation processes. The heat flux run (HTFL) only adopts realistic SWR and LWR radiations, 2-m air temperature, specific humidity, and wind speed from 1958 to 2019 but fixes wind stress and precipitation to the 1958 fields. As such, changes in these fields of HTFL affect OHC trends mainly by altering the heat fluxes (SWR, LWR, sensible, and latent heat fluxes), representing the role of heat-driven processes or ocean heat uptake. To examine the impacts of decadally varying KE jet and its associated mesoscale eddies, we also use the HYCOM version-2.3.34 77 . The HYCOM 0.1° simulation uses the same domain as the HYCOM 0.5° simulation but with resolutions of 0.1° × 0.1° × 50 levels. After a 20-year spin-up, the HYCOM 0.1° simulation is integrated forwards from 1979 to 2021 under daily ERA5 atmospheric forcing. The comparison of the HYCOM 0.1° simulation and the HYCOM 0.5° CTRL allows us to test the possible impact of mesoscale oceanic processes on heat storage pattern in the North Pacific. Definitions The OHC is calculated by integrating ocean temperature T within the upper 2000 m as OHC = \(\:{\int\:}_{0}^{2000}\rho\:{c}_{p}Tdz\) , ( 1 ) where c p = 4096 J (kg °C) −1 is thermal capacity, and ρ = 1025 kg m − 3 is seawater density. The Ekman pumping velocity is calculated as ω E = curl( τ /f)ρ −1 , where τ is the wind stress vector, and f is the Coriolis parameter. The least square method is used to estimate the linear trend. Unless otherwise stated, the statistical significance of the trend is defined as > 95% confidence level based on a modified Mann-Kendall test. The Pacific decadal oscillation (PDO) index is calculated as the leading principle component from the empirical orthogonal function (EOF) analysis of North Pacific (poleward of 20°N) SST anomalies 61 , using the Met Office’s Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST) 78 for observation and historical and SSP5-8.5 simulations of the 20 CMIP6 models. The long-term trend is removed at each grid point before conducting the EOF analysis. The regressions of OHC and wind anomalies on the negative PDO index (-PDO) are first calculated from each of the 20 CMIP6 models, and then the MMM regression maps are derived from the ensemble mean of their coefficients. The Oceanic Niño Index (ONI) is the 3-month running-mean detrended temperature anomalies in the Niño-3.4 region (5°N-5°S, 120°-170°W) based on version 5 of the Extended Reconstructed Sea Surface Temperature 79 , and the monthly ONI is averaged to the annual mean. Time of Emergence (ToE). A “signal-to-noise” method 28,57 is adopted to compute the time of emergence (ToE) of the OHC trend. We use the piControl simulation, historical simulation, and projections under three emission scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5) of 20 CMIP6 models to determine the “signal-to-noise” ratio (SNR). The noise is defined as the interannual standard deviation of the quadratically detrended 80 OHC variability of piControl simulation in the last 100 years, representing the natural noise envelope. The anthropogenic signals from 1958 to 2100 are computed as the OHC differences between the historical plus three emission scenario simulations and the time-mean piControl value. The ToE is defined as the time when the signal last exceeds and remains above a threshold of two times of noise (i.e., SNR > 2.0). The threshold for SNR is chosen as 2.0 because it represents a 95% confidence level of signal emergence 81 . For the regional ToEs, the noise and signal of selected regions are first averaged and then the time with SNR > 2.0 is searched. Declarations Data availability IAP data is available from http://www.ocean.iap.ac.cn/?navAnchor=home. WOA data is available from https://www.nodc.noaa.gov/OC5/indprod.html. Ishii data is available at https://www.data.jma.go.jp/gmd/kaiyou/english/ohc/ohc_global_en.html. EN4 (version 4.2.2) data is available at http://www.metoffice.gov.uk/hadobs/en4/index.html. ORA-S4 temperature and wind fields are available at https://icdc.cen.uni-hamburg.de/daten/reanalysis-ocean/easy-init-ocean/ecmwf-ocean-reanalysis-system-4-oras4.html and https://apps.ecmwf.int/datasets/. JRA55 data is available from https://jra.kishou.go.jp/JRA-55/index_en.html#mirror. EAR5 data can be downloaded at https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5. NCEP-NCAR data is available from https://psl.noaa.gov/data/gridded/tables/monthly.html. CMIP6 simulations are available at https://esgf-node.llnl.gov/projects/cmip6/. HadISST data is available from https://www.metoffice.gov.uk/hadobs/hadisst/. CMEMS data is downloaded at https://data.marine.copernicus.eu/product/SEALEVEL_GLO_PHY_L4_MY_008_047/description. ONI is obtained from https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php. Code availability The HYCOM code used in this word is available via https://github.com/HYCOM. Matlab R2017a is used for plotting. The Matlab code for data analysis and graphing is available upon request. Acknowledgments This work is supported by the Laoshan Laboratory (LSKJ202202601), National Natural Science Foundation of China (42176007), the Strategic Priority Research Program of Chinese Academy of Science (XDB42000000), and the Oceanographic Data Center, the Institute of Oceanology, Chinese Academy of Sciences. Author contributions J. D and Y. L. (Yuanlong Li) drafted the paper. Y. L. (Yuanlong Li) directed this work with contributions from all authors. J. D. and Y. L. (Yilong Lyu) contributed to ocean model data analysis. J. D. contributed to CMIP6 data analysis. All authors discussed the results and commented on the manuscript. Correspondence and requests for materials should be addressed to Y. L. (Yuanlong Li). References IPCC. Climate Change 2021: The Physical Science Basis . (Cambridge University Press, 2021). Fox-Kemper, B. et al. Ocean,cryosphere and sea level change. Climate Change 2021: The Physical Science Basis . (Cambridge University Press, 2021). Cheng, L. et al. Past and future ocean warming. Nat. Rev. Earth Environ. 3, 776–794 (2022). Cheng, L. et al. Another Year of Record Heat for the Oceans. Adv. Atmos. Sci. 40, 963–974 (2023). Eyring, V. et al. Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9, 1937–1958 (2016). Kuhlbrodt, T., Voldoire, A., Palmer, M. D., Geoffroy, O. & Killick, R. E. Historical Ocean Heat Uptake in Two Pairs of CMIP6 Models: Global and Regional Perspectives. J. Clim. 36, 2183–2203 (2023). Gille, S. T. Warming of the Southern Ocean Since the 1950s. Science (80-.). 295, 1275–1277 (2002). Shi, J. R., Xie, S. P. & Talley, L. D. Evolving relative importance of the Southern Ocean and North Atlantic in anthropogenic ocean heat uptake. J. Clim. 31, 7459–7479 (2018). Cai, W. et al. Southern Ocean warming and its climatic impacts. Sci. Bull. 68, 946–960 (2023). Peng, Q. et al. Surface warming–induced global acceleration of upper ocean currents. Sci. Adv. 8, 1–12 (2022). Tamarin-Brodsky, T. & Kaspi, Y. Enhanced poleward propagation of storms under climate change. Nat. Geosci. 10, 908–913 (2017). Swart, N. C. & Fyfe, J. C. Observed and simulated changes in the Southern Hemisphere surface westerly wind-stress. Geophys. Res. Lett. 39, n/a-n/a (2012). Wang, S. et al. Extreme atmospheric rivers in a warming climate. Nat. Commun. 14, 3219 (2023). Ibarbalz, F. M. et al. Global Trends in Marine Plankton Diversity across Kingdoms of Life. Cell 179, 1084–1097.e21 (2019). Han, W. et al. Intensification of decadal and multi-decadal sea level variability in the western tropical Pacific during recent decades. Clim. Dyn. 43, 1357–1379 (2014). Hamlington, B. D. et al. Uncovering an anthropogenic sea-level rise signal in the Pacific Ocean. Nat. Clim. Chang. 4, 782–785 (2014). Hughes, T. P. et al. Spatial and temporal patterns of mass bleaching of corals in the Anthropocene. Science (80-.). 359, 80–83 (2018). Smale, D. A. et al. Marine heatwaves threaten global biodiversity and the provision of ecosystem services. Nat. Clim. Chang. 9, 306–312 (2019). Oliver, E. C. J. et al. Longer and more frequent marine heatwaves over the past century. Nat. Commun. 9, 1324 (2018). Kossin, J. P., Emanuel, K. A. & Vecchi, G. A. The poleward migration of the location of tropical cyclone maximum intensity. Nature 509, 349–352 (2014). Shan, K., Lin, Y., Chu, P., Yu, X. & Song, F. Seasonal advance of intense tropical cyclones in a warming climate. Nature (2023) doi: 10.1038/s41586-023-06544-0 . Mei, W., Xie, S. P., Primeau, F., McWilliams, J. C. & Pasquero, C. Northwestern Pacific typhoon intensity controlled by changes in ocean temperatures. Sci. Adv. 1, (2015). Wu, L. et al. Enhanced warming over the global subtropical western boundary currents. Nat. Clim. Chang. 2, 161–166 (2012). Wang, F. et al. The seas around China in a warming climate. Nat. Rev. Earth Environ. (2023) doi: 10.1038/s43017-023-00453-6 . Hu, D. et al. Pacific western boundary currents and their roles in climate. Nature 522, 299–308 (2015). Lee, S. K. et al. Pacific origin of the abrupt increase in Indian Ocean heat content during the warming hiatus. Nat. Geosci. 8, 445–449 (2015). Li, Y., Han, W. & Zhang, L. Enhanced Decadal Warming of the Southeast Indian Ocean During the Recent Global Surface Warming Slowdown. Geophys. Res. Lett. 44, 9876–9884 (2017). Lyu, K., Zhang, X., Church, J. A., Slangen, A. B. A. & Hu, J. Time of emergence for regional sea-level change. Nat. Clim. Chang. 4, 1006–1010 (2014). Silvy, Y., Sallée, J. B., Guilyardi, E., Mignot, J. & Rousset, C. What Causes Anthropogenic Ocean Warming to Emerge from Internal Variability in a Coupled Model? J. Clim. 35, 3835–3854 (2022). Watanabe, M., Dufresne, J. L., Kosaka, Y., Mauritsen, T. & Tatebe, H. Enhanced warming constrained by past trends in equatorial Pacific sea surface temperature gradient. Nat. Clim. Chang. 11, 3–9 (2020). Wu, B., Lin, X. & Yu, L. North Pacific subtropical mode water is controlled by the Atlantic Multidecadal Variability. Nat. Clim. Chang. 10, 238–243 (2020). England, M. H. et al. Recent intensification of wind-driven circulation in the Pacific and the ongoing warming hiatus. Nat. Clim. Chang. 4, 222–227 (2014). Nieves, V., Willis, J. K. & Patzert, W. C. Recent hiatus caused by decadal shift in Indo-Pacific heating. Science (80-.). 349, 532–535 (2015). Liu, W., Xie, S. P. & Lu, J. Tracking ocean heat uptake during the surface warming hiatus. Nat. Commun. 7, 1–9 (2016). Cheng, L. et al. Improved estimates of ocean heat content from 1960 to 2015. Sci. Adv. 3, (2017). Ishii, M. et al. Accuracy of Global Upper Ocean Heat Content Estimation Expected from Present Observational Data Sets. SOLA 13, 163–167 (2017). Levitus, S. et al. World ocean heat content and thermosteric sea level change (0-2000 m), 1955–2010. Geophys. Res. Lett. 39, n/a-n/a (2012). Good, S. A., Martin, M. J. & Rayner, N. A. EN4: Quality controlled ocean temperature and salinity profiles and monthly objective analyses with uncertainty estimates. J. Geophys. Res. Ocean. 118, 6704–6716 (2013). Balmaseda, M. A., Mogensen, K. & Weaver, A. T. Evaluation of the ECMWF ocean reanalysis system ORAS4. Q. J. R. Meteorol. Soc. 139, 1132–1161 (2013). Chemke, R., Zanna, L. & Polvani, L. M. Identifying a human signal in the North Atlantic warming hole. Nat. Commun. 11, 1–7 (2020). Harada, Y. et al. The JRA-55 reanalysis: Representation of atmospheric circulation and climate variability. J. Meteorol. Soc. Japan 94, 269–302 (2016). Uppala, S. M. et al. The ERA-40 re-analysis. Q. J. R. Meteorol. Soc. 131, 2961–3012 (2005). Dee, D. P. et al. The ERA-Interim reanalysis: configuration and performance of the data assimilation system. Q. J. R. Meteorol. Soc. 137, 553–597 (2011). Kalnay, E. et al. The NCEP/NCAR 40-Year Reanalysis Project. Bull. Am. Meteorol. Soc. 77, 437–472 (1996). Duan, J., Li, Y., Cheng, L., Lin, P. & Wang, F. Heat Storage in the Upper Indian Ocean: The Role of Wind-Driven Redistribution. J. Clim. 36, 2221–2242 (2023). Qiu, B. & Lukas, R. Seasonal and interannual variability of the North Equatorial Current, the Mindanao Current, and the Kuroshio along the Pacific western boundary. J. Geophys. Res. Ocean. 101, 12315–12330 (1996). Qiu, B. & Chen, S. Variability of the Kuroshio Extension Jet, Recirculation Gyre, and Mesoscale Eddies on Decadal Time Scales. J. Phys. Oceanogr. 35, 2090–2103 (2005). Qiu, B., Chen, S., Schneider, N., Oka, E. & Sugimoto, S. On the Reset of the Wind-Forced Decadal Kuroshio Extension Variability in Late 2017. J. Clim. 33, 10813–10828 (2020). Qiu, B. & Chen, S. Effect of Decadal Kuroshio Extension Jet and Eddy Variability on the Modification of North Pacific Intermediate Water. J. Phys. Oceanogr. 41, 503–515 (2011). Taguchi, B. et al. Decadal Variability of the Kuroshio Extension: Observations and an Eddy-Resolving Model Hindcast*. J. Clim. 20, 2357–2377 (2007). Qiu, B. & Chen, S. Interannual-to-Decadal Variability in the Bifurcation of the North Equatorial Current off the Philippines. J. Phys. Oceanogr. 40, 2525–2538 (2010). Kawabe, M. Variations of Current Path, Velocity, and Volume Transport of the Kuroshio in Relation with the Large Meander. J. Phys. Oceanogr. 25, 3103–3117 (1995). Qiu, B., Chen, S. & Oka, E. Why Did the 2017 Kuroshio Large Meander Event Become the Longest in the Past 70 Years? Geophys. Res. Lett. 50, 1–9 (2023). Silva, E. N. S. & Anderson, B. T. Northeast Pacific marine heatwaves linked to Kuroshio Extension variability. Commun. Earth Environ. 4, 367 (2023). Ma, X. et al. Western boundary currents regulated by interaction between ocean eddies and the atmosphere. Nature 535, 533–537 (2016). Newman, M. et al. The Pacific Decadal Oscillation, Revisited. J. Clim. 29, 4399–4427 (2016). Silvy, Y., Guilyardi, E., Sallée, J. B. & Durack, P. J. Human-induced changes to the global ocean water masses and their time of emergence. Nat. Clim. Chang. 10, 1030–1036 (2020). Jones, N. Rare ‘triple’ La Niña climate event looks likely - what does the future hold? Nature vol. 607 21 (2022). Geng, T. et al. Emergence of changing Central-Pacific and Eastern-Pacific El Niño-Southern Oscillation in a warming climate. Nat. Commun. 13, 6616 (2022). Gan, B. & Wu, L. Seasonal and Long-Term Coupling between Wintertime Storm Tracks and Sea Surface Temperature in the North Pacific. J. Clim. 26, 6123–6136 (2013). Mantua, N. J., Hare, S. R., Zhang, Y., Wallace, J. M. & Francis, R. C. A Pacific Interdecadal Climate Oscillation with Impacts on Salmon Production. Bull. Am. Meteorol. Soc. 78, 1069–1079 (1997). Payne, A. E. et al. Responses and impacts of atmospheric rivers to climate change. Nature Reviews Earth and Environment vol. 1 143–157 (2020). Xu, G. et al. High-resolution modelling identifies the Bering Strait’s role in amplified Arctic warming. Nat. Clim. Chang. (2024) doi: 10.1038/s41558-024-02008-z . Farneti, R. Modelling interdecadal climate variability and the role of the ocean. WIREs Clim. Chang. 8, (2017). Chang, P. et al. An Unprecedented Set of High-Resolution Earth System Simulations for Understanding Multiscale Interactions in Climate Variability and Change. J. Adv. Model. Earth Syst. 12, (2020). An, B. et al. The benefits of high-resolution models in simulating the Kuroshio Extension and its long-term changes. Clim. Dyn. 61, 5407–5427 (2023). Chung, E. S. et al. Reconciling opposing Walker circulation trends in observations and model projections. Nat. Clim. Chang. 9, 405–412 (2019). Ren, Q. et al. Substantial Warming of the Atlantic Ocean in CMIP6 Models. J. Clim. (2024) doi: 10.1175/JCLI-D-23-0418.1 . Di Lorenzo, E. & Mantua, N. Multi-year persistence of the 2014/15 North Pacific marine heatwave. Nat. Clim. Chang. 6, 1042–1047 (2016). Amaya, D. J., Miller, A. J., Xie, S. & Kosaka, Y. Physical drivers of the summer 2019 North Pacific marine heatwave. Nat. Commun. 11, 1903 (2020). Song, Q., Wang, C., Yao, Y. & Fan, H. Unraveling the Indian monsoon’s role in fueling the unprecedented 2022 Marine Heatwave in the Western North Pacific. npj Clim. Atmos. Sci. 7, 90 (2024). Cavole, L. et al. Biological Impacts of the 2013–2015 Warm-Water Anomaly in the Northeast Pacific: Winners, Losers, and the Future. Oceanography 29, (2016). Nielsen, J. M. et al. Responses of ichthyoplankton assemblages to the recent marine heatwave and previous climate fluctuations in several Northeast Pacific marine ecosystems. Glob. Chang. Biol. 27, 506–520 (2021). Lima, F. P. & Wethey, D. S. Three decades of high-resolution coastal sea surface temperatures reveal more than warming. Nat. Commun. 3, 704 (2012). Hersbach, H.. et al. ERA5 monthly averaged data on single levels from 1940 to present. Copernicus Clim. Chang. Serv. Clim. Data Store (2023) doi: 10.24381/cds.f17050d7 . Chassignet, E. P., Smith, L. T., Halliwell, G. R. & Bleck, R. North Atlantic Simulations with the Hybrid Coordinate Ocean Model (HYCOM): Impact of the Vertical Coordinate Choice, Reference Pressure, and Thermobaricity. J. Phys. Oceanogr. 33, 2504–2526 (2003). Wallcraft, a. J., Metzger, E. J. & Carrol, S. N. Software Design Description for the HYbrid Coordinate Ocean Model (HYCOM). 149 (2009). Rayner, N. A. Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res. 108, 4407 (2003). Huang, B. et al. Extended Reconstructed Sea Surface Temperature, Version 5 (ERSSTv5): Upgrades, Validations, and Intercomparisons. J. Clim. 30, 8179–8205 (2017). Cai, W. et al. Butterfly effect and a self-modulating El Niño response to global warming. Nature 585, 68–73 (2020). Hawkins, E. & Sutton, R. Time of emergence of climate signals. Geophys. Res. Lett. 39, n/a-n/a (2012). Additional Declarations There is NO Competing Interest. Supplementary Files ExtendedDataFig.docx Cite Share Download PDF Status: Published Journal Publication published 14 Jan, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4905116","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":340786008,"identity":"6939fefb-78fb-4135-acd7-9c2d708c304b","order_by":0,"name":"Yuanlong Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4UlEQVRIiWNgGAWjYBACPmYwZQHEzAcYeAyI0MIG0SIBYiYQqYUBrgWonocYh7Gx85hJ/NwhIWfOv+bjjTcFNgz87QcYPxfgdRiPmWTvGQljyxlvN1vOMUhjkDiTwCw9g4AWCd42icQNN85uk+YxOMzAcAMkSMiWv2AtZ56BtcgTo0UabMv5HjawFgPCWtiKrWXbJIwNbrAZg/zCY3gmsVkanxZ+/sMbb75ts5EzOH/44Y03f2zk5I4fPvgZf2hzQKNPIgESOwwMjA14NTAwsD+A2ncArGUUjIJRMApGAQYAACncPpHVpLO1AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-7239-5756","institution":"Institute of Oceanology","correspondingAuthor":true,"prefix":"","firstName":"Yuanlong","middleName":"","lastName":"Li","suffix":""},{"id":340786009,"identity":"bbe41b8b-afa1-4f11-bb6d-05769cbbcb49","order_by":1,"name":"Jing Duan","email":"","orcid":"","institution":"Key Laboratory of Ocean Circulation and Waves, Institute of Oceanology, Center for Ocean Mega-Science, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Jing","middleName":"","lastName":"Duan","suffix":""},{"id":340786010,"identity":"06829fa0-3744-4f45-87ab-67d775ae66d6","order_by":2,"name":"Yilong Lyu","email":"","orcid":"https://orcid.org/0000-0001-5943-5760","institution":"Institute of Oceanology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Yilong","middleName":"","lastName":"Lyu","suffix":""},{"id":340786011,"identity":"ef290d7c-81f9-4bcd-b0a8-30d7e35c0ce3","order_by":3,"name":"Zhao Jing","email":"","orcid":"https://orcid.org/0000-0002-8430-9149","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Zhao","middleName":"","lastName":"Jing","suffix":""},{"id":340786012,"identity":"69b7dd23-db37-4dd3-873d-371aae29123f","order_by":4,"name":"Fan Wang","email":"","orcid":"","institution":"Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Fan","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2024-08-13 07:55:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4905116/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4905116/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-56005-7","type":"published","date":"2025-01-14T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":63948316,"identity":"2f4b15e9-3443-4c27-8de9-6ff9278e890a","added_by":"auto","created_at":"2024-09-04 06:29:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":576857,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOcean heat storage pattern and the evolution in key regions. a, b, \u003c/strong\u003ePatterns of 0-2000 m ocean heat content (OHC) trend (W m\u003csup\u003e-2\u003c/sup\u003e) of 1958-2021 based on the multi-model mean (MMM) of 20 CMIP6 historical and SSP5-8.5 simulations (\u003cstrong\u003ea\u003c/strong\u003e) and the Institute of Atmospheric Physics ocean analysis (IAP)\u003csup\u003e35\u003c/sup\u003e (\u003cstrong\u003eb\u003c/strong\u003e). Stippling indicates significant trends at the 95% confidence level based on a Mann-Kendall test. The black boxes remark the regions with enhanced heat uptake: the mid-latitude North Pacific (MNP; 155°E-150°W, 40°-55°N) and the northwest tropical Pacific (NWTP; 125°-180°E, 8°-18°N). \u003cstrong\u003ec, d, \u003c/strong\u003eRegional OHCs (GJ m\u003csup\u003e-2\u003c/sup\u003e, 1 GJ = 10\u003csup\u003e9\u003c/sup\u003e J) of MNP (\u003cstrong\u003ec\u003c/strong\u003e) and NWTP (\u003cstrong\u003ed\u003c/strong\u003e) derived from observations and CMIP6 simulations. The observations include IAP, WOA, Ishii, EN4, and ORA-S4. Shadings denote the one standard deviation range.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/da9f19210863bf44d69d31bb.png"},{"id":63948314,"identity":"9bf83bf5-6084-47be-90dc-645c12ca9204","added_by":"auto","created_at":"2024-09-04 06:29:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":551554,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffect of wind forcing on heat storage. a-c, \u003c/strong\u003eLinear OHC trends of 1958-2019 derived from CTRL (\u003cstrong\u003ea\u003c/strong\u003e), HTFL (\u003cstrong\u003eb\u003c/strong\u003e), and WND (\u003cstrong\u003ec\u003c/strong\u003e) runs of the HYCOM 0.5° simulation. Stippling indicates significant trends at the 95% confidence level. \u003cstrong\u003ed, e,\u003c/strong\u003e Regional OHCs of the MNP (\u003cstrong\u003ed\u003c/strong\u003e) and NWTP (\u003cstrong\u003ee\u003c/strong\u003e) derived from CTRL, HTFL, and WND of HYCOM 0.5° simulation, referenced to those derived from IAP.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/328161a871c502919db2edb1.png"},{"id":63948310,"identity":"a819131a-0e6f-44a2-b54f-ce7f4370c560","added_by":"auto","created_at":"2024-09-04 06:29:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":648391,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChanges in surface winds and impacts of the Pacific Decadal Oscillation (PDO). a, b, \u003c/strong\u003eTrends\u003cstrong\u003e \u003c/strong\u003eof wind stress (arrows; Pa yr\u003csup\u003e-1\u003c/sup\u003e) and Ekman pumping velocity \u003cem\u003eω\u003c/em\u003e\u003csub\u003e\u003cem\u003eE\u003c/em\u003e\u003c/sub\u003e (color shading; 10\u003csup\u003e-7\u003c/sup\u003e m s\u003csup\u003e-1\u003c/sup\u003e yr\u003csup\u003e-1\u003c/sup\u003e) for 1958-2021 based on the ensemble average of four atmospheric reanalysis products (JRA55, ERA5, NCEP-NCAR, and ORA-S4; addressed as “observed”) (\u003cstrong\u003ea\u003c/strong\u003e) and CMIP6 MMM (\u003cstrong\u003eb\u003c/strong\u003e). \u003cstrong\u003ec,\u003c/strong\u003e Regressions of observed wind stress (Pa) and \u003cem\u003eω\u003c/em\u003e\u003csub\u003e\u003cem\u003eE\u003c/em\u003e\u003c/sub\u003e (10\u003csup\u003e-6\u003c/sup\u003e m s\u003csup\u003e-1\u003c/sup\u003e) on the negative PDO index (-PDO) based on HadISST for 1958-2021. \u003cstrong\u003ed,\u003c/strong\u003e Regression map of OHC anomalies derived from IAP on -PDO for 1958-2021. Stippling indicates the fields exceeding the 95% confidence level based on an \u003cem\u003eF\u003c/em\u003e-test. The black boxes denote NWTP and MNP. \u003cstrong\u003ee, f,\u003c/strong\u003e Same as \u003cstrong\u003ed\u003c/strong\u003e, but with the OHC anomalies lagging -PDO by 4 (\u003cstrong\u003ee\u003c/strong\u003e) and 9 years (\u003cstrong\u003ef\u003c/strong\u003e).\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/82e93b7a846189f354ffc44b.png"},{"id":63949572,"identity":"6763f2af-d1e3-4a5f-9ed4-4efec78fe7e7","added_by":"auto","created_at":"2024-09-04 06:45:56","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":814474,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImpacts of the PDO on the Kuroshio Extension (KE). a, \u003c/strong\u003eThe normalized sea surface height (SSH) anomaly of the southern recirculation region (140°-165°E, 31°-36°N). \u003cstrong\u003eb, \u003c/strong\u003eTime-longitude plot of SSH anomaly along the zonal band of 32°-34°N. \u003cstrong\u003ea, b\u003c/strong\u003e are based on monthly CMEMS data. The black arrow in (\u003cstrong\u003eb\u003c/strong\u003e) indicates the propagation of Rossby waves. \u003cstrong\u003ec, \u003c/strong\u003eThe normalized monthly PDO index based on HadISST data.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/bd7f7408a7c8a17afcf28262.png"},{"id":63948312,"identity":"ada26077-e1c8-494e-9100-95d567cd084a","added_by":"auto","created_at":"2024-09-04 06:29:56","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":769501,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImpacts of the PDO on OHC. a, b, \u003c/strong\u003eLagged regression map of OHC on -PDO derived from HYCOM-0.1° simulation of 1979-2021, with the former lagging the latter by 4 (\u003cstrong\u003ea\u003c/strong\u003e) and 9 years (\u003cstrong\u003eb\u003c/strong\u003e). \u003cstrong\u003ec, d, \u003c/strong\u003eSame as\u003cstrong\u003e a, b\u003c/strong\u003e, but based on the MMM of 20 CMIP6 models. Stippling indicates the fields exceeding the 95% confidence level.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/1f15c9460f06507ea0a5f7c6.png"},{"id":63948967,"identity":"1438a939-6208-4e58-b238-20e2b9746dd1","added_by":"auto","created_at":"2024-09-04 06:37:56","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":251104,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTime of emergence (ToE) for anthropogenic heat storage. a, \u003c/strong\u003eMulti-model median ToE of the anthropogenic OHC trend under the SSP2-4.5 scenario. The ToE is the year when the trend exceeds the range of the twofold natural interannual standard deviation (Methods). Grey shading denotes the signal that has not emerged by 2099.\u003cstrong\u003e b-f, \u003c/strong\u003eBox-and-whisker plots of ToEs of regional OHCs in the Bering Sea (160°E-170°W, 52°-60°N) (\u003cstrong\u003eb\u003c/strong\u003e), Eastern boundary (132°-117°W, 30°-55°N) (\u003cstrong\u003ec\u003c/strong\u003e), central MNP (170°E-155°W, 40°-48°N) (\u003cstrong\u003ed\u003c/strong\u003e), Kuroshio Extension (KE; 141°-165°E, 33°-40°N) (\u003cstrong\u003ee\u003c/strong\u003e), and NWTP (\u003cstrong\u003ef\u003c/strong\u003e). The ToEs \u0026gt; 2099 are not shown.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/d2fd5dae69bb6699c6927a52.png"},{"id":73840392,"identity":"00a04d8b-309f-476f-aabb-6247f9913c53","added_by":"auto","created_at":"2025-01-15 08:09:38","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4138228,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/f32d1c97-f008-4f43-92cf-f4b749f7c80c.pdf"},{"id":63948969,"identity":"c0c0fcfc-ff8b-43f4-ae48-88677695c8d0","added_by":"auto","created_at":"2024-09-04 06:37:56","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":7351623,"visible":true,"origin":"","legend":"","description":"","filename":"ExtendedDataFig.docx","url":"https://assets-eu.researchsquare.com/files/rs-4905116/v1/0578f47807872655dc583377.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Emergence of the North Pacific heat storage pattern delayed by decadal wind-driven redistribution","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe storage of excess heat caused by anthropogenic greenhouse warming in oceans\u003csup\u003e1\u003c/sup\u003e is geographically inhomogeneous\u003csup\u003e2\u0026ndash;4\u003c/sup\u003e. Climate models\u003csup\u003e5,6\u003c/sup\u003e suggest that the Southern Ocean\u003csup\u003e3,7\u0026ndash;9\u003c/sup\u003e, the North Atlantic\u003csup\u003e3,8\u003c/sup\u003e, and the mid-latitude North Pacific (MNP) warm up more quickly than other oceans (\u003cstrong\u003eFig. 1a\u003c/strong\u003e). These ocean warming features accompany accelerated ocean currents\u003csup\u003e8,10\u003c/sup\u003e, poleward shift of storm tracks and westerlies\u003csup\u003e11,12\u003c/sup\u003e, increased extreme atmospheric rivers\u003csup\u003e13\u003c/sup\u003e, and altered marine biodiversity patterns\u003csup\u003e14\u003c/sup\u003e. While the enhanced heat storage in the Southern Ocean and North Atlantic has been witnessed in nature, that in the MNP has hardly emerged (\u003cstrong\u003eFig. 1b\u003c/strong\u003e). Indeed, despite significant warming trends in marginal seas along the basin rim, there were weak warming or cooling trends in the central MNP since the 1950s.\u003c/p\u003e\n\u003cp\u003eThe tropical Pacific is among the regions with the weakest heat storage in CMIP6 multi-model mean (MMM) (\u003cstrong\u003eFig. 1a\u003c/strong\u003e). However, the Northwest Tropical Pacific (NWTP) has exhibited the strongest warming rate of the Pacific Ocean during the past decades (\u003cstrong\u003eFig. 1b\u003c/strong\u003e). Increased ocean heat content (OHC) in the NWTP has led to rapid sea-level rise\u003csup\u003e15,16\u003c/sup\u003e, increasing marine heatwave and coral bleaching events\u003csup\u003e17\u0026ndash;19\u003c/sup\u003e, and altered tropical cyclone behaviors\u003csup\u003e20\u0026ndash;22\u003c/sup\u003e. It also exerted impacts on the downstream marginal seas through western boundary currents\u003csup\u003e23\u0026ndash;25\u003c/sup\u003e and the Indian Ocean through the Indonesian Throughflow\u003csup\u003e26,27\u003c/sup\u003e. The Pacific is projected to be the leading reservoir of anthropogenic heat among all oceans by the latter half of the 21\u003csup\u003est\u003c/sup\u003e century\u003csup\u003e3\u003c/sup\u003e. Correctly interpreting the model-observation discrepancies there represents a vital scientific issue.\u003c/p\u003e\n\u003cp\u003eThe contrasts between observations and models indicate either systematic model biases in simulating the externally forced heat storage or substantial impacts by natural variability. Although anthropogenic fingerprints have emerged in many aspects of the ocean\u003csup\u003e3,28,29\u003c/sup\u003e, natural variability remains influential in the observed OHC changes\u003csup\u003e30,31\u003c/sup\u003e. For instance, the persistent negative phase of the Pacific Decadal Oscillation (PDO) since 1998 led to heat pile-up in the western tropical Pacific via the strengthened Pacific Walker Circulation\u003csup\u003e32\u0026ndash;34\u003c/sup\u003e. If the model-observation discrepancies can be successfully attributed to natural variability, we shall expect an acceleration of warming in the MNP within the coming decades to catch up with the projected rate.\u003c/p\u003e\n\u003cp\u003eHere, we set out to quantify the contributions of various processes to the observed North Pacific heat storage pattern, addressing in particular whether natural variability can largely explain the model-observation discrepancy. Our analysis is based on 1) five observational datasets\u003csup\u003e35\u0026ndash;39\u003c/sup\u003e to characterize the historical heat storage, 2) experiments of a forced ocean model to isolate effects of surface heat fluxes and wind-driven ocean dynamics, and 3) 20 CMIP6 models to estimate anthropogenic fingerprints and when they may emerge (Methods). We demonstrate that surface wind changes arising from phase shifts of the PDO have driven basin-scale heat redistribution through Rossby waves and modulations in western boundary currents. This effect complicates the observed heat storage pattern by creating regional warming/cooling structures that conceal anthropogenic fingerprints. According to model projections, the human-induced heat storage pattern will likely hide until the late-21\u003csup\u003est\u003c/sup\u003e century. These results provide useful implications for climate prediction in the North Pacific and marginal seas.\u003c/p\u003e\n\u003ch3\u003eHeat storage in the North Pacific\u003c/h3\u003e\n\u003cp\u003eWe begin with more details of the simulated and observed heat storage patterns in the North Pacific during the 1958\u0026ndash;2021 period. In the multi-model mean (MMM) of 20 CMIP6 models, consisting of the historical and Shared Socioeconomic Pathway (SSP) 5-8.5 simulations before and after 2014, respectively, the entire MNP is characterized by enhanced 0-2000 m heat storage, in stark contrast to the warming minimum in the western tropical Pacific (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). The observation-based heat storage shows more regional structures (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb and \u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e); for example, there are alternating warming and cooling regions in the North Pacific, particularly in the western basin. The NWTP (e.g., 125\u0026deg;-180\u0026deg;E, 8\u0026deg;-18\u0026deg;N) stands out with a warming maximum, whereas a \u0026ldquo;warming hole\u0026rdquo; occurs in the MNP interior with insignificant trends and is surrounded by significant warming trends in marginal seas, such as the Bering Sea and the Gulf of Alaska. There is another cooling region near the western boundary of the 18\u0026deg;-30\u0026deg;N, sandwiched by the warming areas of the NWTP and the Kuroshio extension (KE) between 30\u0026deg;-40\u0026deg;N. Cooling trends are also seen beyond the North Pacific, such as the subpolar North Atlantic, southwestern subtropical Pacific, and southwestern subtropical Indian Ocean. CMIP6 MMM also show cooling or slackened warming in these regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). Albeit with weaker intensity and smaller spatial range, the subpolar North Atlantic warming hole\u003csup\u003e40\u003c/sup\u003e is clearly discernible in CMIP6 MMM. Therefore, models and observations are broadly consonant in all major ocean basins except for the North Pacific. The model-observation discrepancies in the North Pacific heat storage are worthy of in-depth investigation.\u003c/p\u003e \u003cp\u003eWe further examine the temporal evolution of OHC in key regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, d). CMIP6 models suggest an average heat storage rate of 0.72\u0026thinsp;\u0026plusmn;\u0026thinsp;0.62 W m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e (\u0026plusmn;\u0026thinsp;indicates the one standard deviation range of 20 models) in the MNP of 155\u0026deg;E-150\u0026deg;W, 40\u0026deg;-55\u0026deg;N, close in magnitude to that of the Southern Ocean (0.68\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27 W m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e in MMM for 55\u0026deg;-33\u0026deg;S). In comparison, a much weaker rate of 0.17\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08 W m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e in the MNP is obtained from observation-based datasets (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec; \u0026plusmn; indicates the one standard deviation range of 5 datasets). Meanwhile, the NWTP shows enhanced warming of 0.58\u0026thinsp;\u0026plusmn;\u0026thinsp;0.26 W m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e in observations, one order stronger than the simulated rate of 0.06\u0026thinsp;\u0026plusmn;\u0026thinsp;0.32 W m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e in CMIP6 models (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Interestingly, CMIP6 agrees with observation in the OHC change of the entire North Pacific (120\u0026deg;E-80\u0026deg;W, 0\u0026deg;-60\u0026deg;N), which are 0.33\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07 W m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e and 0.33\u0026thinsp;\u0026plusmn;\u0026thinsp;0.17 W m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e, respectively, implying a heat redistribution over the North Pacific in the observed realization relative to the simulated pattern.\u003c/p\u003e \u003cp\u003eThe observed OHCs show prominent interannual and decadal variabilities in the MNP and NWTP (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, d). Meanwhile, we notice a large inter-model spread relative to the MMM change in the historical simulation of CMIP6 before 2014, which also indicates significant influence from natural variability. The effect of decadal natural variability is particularly notable from the 1990s through the mid-2010s. During this period, the observed OHC of the NWTP was substantially elevated and exceeded the +\u0026thinsp;1 standard deviation range of CMIP6 models, while that of the MNP did not rise significantly as models expected. In the following, we examine how decadal natural variability affects the North Pacific heat storage.\u003c/p\u003e \u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003eHeat redistribution driven by wind changes\u003c/h2\u003e \u003cp\u003eTo understand how the heat storage pattern is formed, we performed experiments using the Hybrid Coordinate Ocean Model (HYCOM) with a coarse horizontal resolution of 0.5\u0026deg; (Methods) and prescribed reanalysis of atmospheric fields\u003csup\u003e41\u003c/sup\u003e as the surface forcing. Despite regional simulation errors, the control run (CTRL) of HYCOM captured broad-scale features of the observed heat storage \u0026ndash; the cooling trends in the central MNP and the enhanced warming of the NWTP \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). CTRL also well reproduced the prominent interannual and decadal variabilities (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed, e). During the simulation period of 1958\u0026ndash;2019, the correlation coefficients between CTRL and IAP are 0.57 and 0.81 for OHCs in the MNP and NWTP, respectively, both significant at the 99% confidence level.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWith the aid of HYCOM experiments, we can separate the effects of heat redistribution induced by wind-driven ocean circulation changes and heat uptake through surface heat fluxes in heat storage. The heat-flux run (HTFL) retains changes in surface heat fluxes and keeps wind stress and precipitation invariant (Methods), representing the effect of heat uptake. HTFL produces substantially stronger warming in the MNP than in the tropical Pacific (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb) \u0026ndash; a pattern resembling CMIP6 MMM (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). This highlights the dominance of heat uptake in shaping the model-projected heat storage pattern. Alternatively, the wind run (WND), retaining changes only in surface wind stress and keeping other forcing fields unchanged (Methods), produces basin-wide cooling trends in the MNP and enhanced warming of the NWTP (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). These wind-driven changes greatly modify the pattern shaped by heat uptake and vitally contribute to the total storage in CTRL. Checking the temporal evolution clearly suggests that while heat fluxes drive quasi-monotonic warming trends in both regions (stronger in MNP), the interannual and decadal fluctuations arise mainly from winds (stronger in NWTP) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed, e). Critically, winds have induced an OHC decrease in the MNP and an abrupt OHC increase in the NWTP since the late 1990s, which greatly altered the overall trends of 1958\u0026ndash;2019.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThen, we explore changes in surface winds. Reanalysis datasets\u003csup\u003e41\u0026ndash;44\u003c/sup\u003e suggest basin-scale trends of anti-cyclonic winds over the North Pacific since 1958 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Correspondingly, there are easterly winds and negative Ekman pumping velocity \u003cem\u003eω\u003c/em\u003e\u003csub\u003e\u003cem\u003eE\u003c/em\u003e\u003c/sub\u003e (indicating downwelling; Methods) in the NWTP, which causes convergence of the upper-layer warm water\u003csup\u003e45\u003c/sup\u003e and enhances the heat storage there. Meanwhile, we also see westerly winds and positive \u003cem\u003ew\u003c/em\u003e\u003csub\u003eE\u003c/sub\u003e (upwelling) in the eastern tropical Pacific, which might also affect the NWTP through Rossby waves. However, this effect fails to dampen the NWTP warming induced by local wind changes, probably owing to the dissipation of Rossby waves during their transition across the Pacific basin\u003csup\u003e46\u003c/sup\u003e. The anti-cyclone also involves strengthening westerlies north of 40\u0026deg;N, which slackens the heat storage in the subpolar North Pacific through Ekman upwelling and enhances heat pile-up in the 30\u0026deg;-40\u0026deg;N band through Ekman downwelling.\u003c/p\u003e \u003cp\u003eIn CMIP6 MMM, trends of anti-cyclonic winds are confined north of 30\u0026deg;N and westerly trends occupy the entire tropical Pacific (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). Therefore, some key features of the observed wind trend are missed, including the easterlies in the NWTP and westerlies in the subpolar North Pacific. These differences in surface winds, critically accounting for the model-observation discrepancies in heat storage, likely arise from natural variability, given that the CMIP6 MMM is assumed to represent externally forced changes. The PDO, as the leading mode of decadal natural variability in the North Pacific, was in its positive phase during 1977\u0026ndash;1997 and then shifted to its negative phase of 1998\u0026ndash;2014 (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). A regression onto the negative PDO index (-PDO) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec) shows easterlies in the NWTP and westerlies in the northeastern subpolar Pacific \u0026ndash; key features seen in observation but missed in CMIP6 MMM. The negative PDO also induces positive \u003cem\u003ew\u003c/em\u003e\u003csub\u003eE\u003c/sub\u003e anomalies in the western and central parts of the 18\u0026deg;-30\u0026deg;N band, causing the cooling trends observed near the western boundary (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003eThe regression of OHC onto the negative PDO shows some features resembling the observed heat storage, particularly the warming in the NWTP and the 30\u0026deg;-40\u0026deg;N band and cooling in the MNP and the 18\u0026deg;-30\u0026deg;N band (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). The NWTP heat content shows correlation coefficients of -0.42 and \u0026minus;\u0026thinsp;0.66 with the 8-year low-passed PDO index and the unfiltered December-January-February Oceanic Ni\u0026ntilde;o Index (ONI), respectively (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). This suggests a strong modulation of natural variability on the NWTP on interannual and decadal timescales, with heat pile-up in the NWTP under negative PDO and La Ni\u0026ntilde;a conditions. In the mid-latitudes, the regression cannot explain the observed heat storage. For example, cooling trends occur mainly in the marginal seas in the regression (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed) rather than in the MNP interior as in the trend pattern (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb); the meridional dipole-like structure observed east of Japan, linked to the strengthening geostrophic transport of the KE jet (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb), is replaced by prevailing warming in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed. These discrepancies can be reconciled by considering the time lag of oceanic response to wind forcing through planetary wave adjustments\u003csup\u003e47\u003c/sup\u003e. Lagged regressions show that the strengthened KE east of Japan is established\u0026thinsp;~\u0026thinsp;4 years after the negative peak of PDO, and the central MNP cooling takes\u0026thinsp;~\u0026thinsp;9 years (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee, f). The wind-driven OHC anomaly of the MNP, measured by the WND run of HYCOM, lags the PDO index (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb): the PDO shifted from the negative to the positive phase during the 1970s, and correspondingly there was a warming trend of MNP throughout the 1980s; subsequently, the opposite transition from the late 1990s through ~\u0026thinsp;2010 led to a cooling trend of the MNP persisting through the late 2010s.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eRegional ocean dynamics\u003c/h3\u003e\n\u003cp\u003eThe above analysis points to the vital role of PDO phase shifts in shaping the historical heat storage pattern through wind-driven redistribution. One question arises as to whether models can correctly simulate the PDO-induced variability, which is critical for predicting regional OHC changes and their climatic and environmental impacts in the upcoming decades. This remains a challenging task for the state-of-the-art models. Even with prescribed reanalysis winds, our 0.5\u0026deg; simulation of HYCOM fails to fully reproduce the observed heat storage in the MNP (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea); for instance, the simulated cooling in the central MNP is stronger in amplitude, broader in spatial extent, and shifted to lower latitudes in CTRL compared to that in observation. These discrepancies probably arise from complex regional ocean dynamics that are not properly represented by coarse-resolution models - such as the \u0026ldquo;bimodal\u0026rdquo; variability of the KE\u003csup\u003e47,48\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn the mid-latitudes, a negative phase of the PDO drives downwelling Rossby waves through negative \u003cem\u003ew\u003c/em\u003e\u003csub\u003eE\u003c/sub\u003e anomalies between 150\u0026deg;-140\u0026deg;W (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Along the latitudinal band of the KE (e.g., 31\u0026deg;-36\u0026deg;N), these Rossby waves propagate across the Pacific basin to the KE region east of Japan by ~\u0026thinsp;4 years, as manifested in satellite-based sea surface height (SSH) anomalies (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). These downwelling waves lead to a \u0026ldquo;stable\u0026rdquo; state of the KE system characterized by a strengthened KE jet and weakened mesoscale eddy variability\u003csup\u003e47\u0026ndash;49\u003c/sup\u003e, as indicated by the enhanced anticyclonic \u0026ldquo;southern recirculation\u0026rdquo; locating south of the jet\u003csup\u003e50,51\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). The strengthened geostrophic KE jet manifests as a meridional dipole in OHC east of Japan (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee). In addition to PDO, the KE is also modulated by the upstream Kuroshio path south of Japan. The PDO shifted to a positive phase in ~\u0026thinsp;2014, and correspondingly, an unstable dynamical state of the KE was established in early 2017, as indicated by the weakened southern recirculation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). However, the occurrence of the Kuroshio large meander in August 2017 - with the Kuroshio following an offshore meandering path in the Shikoku basin south of Japan\u003csup\u003e52\u003c/sup\u003e \u0026ndash; compelled the KE to switch back to the stable state in late 2017 and remain stable since then\u003csup\u003e48,53\u003c/sup\u003e. Albeit interrupted in 2017, there has been a persistently stable KE since ~\u0026thinsp;2009. This led to an overall trend of the KE toward its stable state during the past decades, given the more unstable KE during the 1980s and 1990s due to the positive PDO phase (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). This trend manifests as a dipole-like structure in heat storage east of Japan (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). By affecting the overlying storm tracks and large-scale winds, the stable KE also contributed to the cooling in the central MNP and the warming in the eastern basin\u003csup\u003e48,54\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe response of KE to Rossby waves is successfully reproduced by a 0.1\u0026deg; simulation of HYCOM (Methods; \u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) that can better resolve the Kuroshio path and mesoscale eddies. As a result, this 0.1\u0026deg; simulation can realistically represent the heat redistribution driven by PDO winds, including the meridional dipole east of Japan with a lag time of ~\u0026thinsp;4 years (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea) and the cooling in the central MNP with a lag time of ~\u0026thinsp;9 years (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb) to the negative peak of PDO. This\u0026thinsp;~\u0026thinsp;9-year lag time reflects the slow propagation of upwelling Rossby waves between 40\u0026deg;-55\u0026deg;N (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) generated by positive \u003cem\u003ew\u003c/em\u003e\u003csub\u003eE\u003c/sub\u003e anomalies near the west coasts of North America (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). There are also detailed discrepancies between the 0.1\u0026deg; simulation and observation in the KE region (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, b compared to Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee, f), which may result from the lack of feedback between mesoscale eddies and the atmosphere in our standard-alone HYCOM simulation. The eddy-atmosphere feedback has been demonstrated to efficiently damp mesoscale eddies and amend the simulated strength and pathway of the KE\u003csup\u003e55\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eMost CMIP6 models are of horizontal resolutions of ~\u0026thinsp;50 or ~\u0026thinsp;100 km in their ocean components, close to our 0.5\u0026deg; HYCOM simulation. We examine how the PDO-induced heat redistribution is represented in these models. The regression using individual simulations of CMIP6 models (Methods) suggests that CMIP6 models can capture some large-scale features of the PDO-induced changes in surface winds and OHC, particularly in the mid-latitudes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec, d and \u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The cooling of the central MNP is seen at a lag of ~\u0026thinsp;9 years, albeit with extended spatial range (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). The PDO\u0026rsquo;s signatures in the NWTP are weaker in models than in observations, linked to the uncertain tropical Pacific wind changes (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This indicates a weaker coupling between the PDO and the tropical Pacific climate in models\u003csup\u003e56\u003c/sup\u003e. By prescribing the observed (reanalysis) surface winds, the 0.5\u0026deg; HYCOM simulations can well reproduce the response of NWTP changes associated with the PDO winds (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Furthermore, the improved fidelity of the 0.1\u0026deg; HYCOM (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, b) relative to the 0.5\u0026deg; HYCOM and CMIP6 models highlights the necessity of fine resolution to account for the complex ocean dynamics involved in heat redistribution.\u003c/p\u003e \u003cp\u003eTo further confirm the PDO\u0026rsquo;s impact, we select 5 model realizations with the simulated PDO showing the highest positive correlations with the observed PDO (+\u0026thinsp;5 models) and 5 model realizations showing the highest negative correlation with the observation (-5 models). As such, contrasting the +\u0026thinsp;5 and \u0026minus;\u0026thinsp;5 models mimics impacts of the observed PDO phase shifts on the OHC trends since 1958 (\u003cb\u003eExtended Data Fig.\u0026nbsp;7\u003c/b\u003e). The composite of +\u0026thinsp;5 minus \u0026minus;\u0026thinsp;5 models reassures that the PDO phase shifts in observation can dampen the heat storage in the MNP and enhance the warming of the NWTP (\u003cb\u003eExtended Data Fig.\u0026nbsp;7c\u003c/b\u003e) through wind-driven heat redistribution (\u003cb\u003eExtended Data Fig.\u0026nbsp;7d\u003c/b\u003e). Therefore, despite unresolved regional structures due to coarse model resolutions, discrepancies between CMIP6 models and observations in the North Pacific heat storage pattern, particularly the contrasts in the NWTP and the central MNP, are primarily explained by natural variability represented by the PDO.\u003c/p\u003e\n\u003ch3\u003eEmergence of anthropogenic heat storage\u003c/h3\u003e\n\u003cp\u003eThe above analysis indicates that the emergence of anthropogenic heat storage has been delayed by natural variability in some key regions of the North Pacific and will eventually take place in the future as the emission of greenhouse gases continues. Therefore, it is instructive to estimate the time of emergence (ToE) of anthropogenic heat storage based on CMIP6 simulations (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e), although we fully acknowledge that such estimation is subject to potential influence from inaccurate ocean dynamics as discussed earlier. Here, we utilize the piControl simulation, historical simulation, and projections under three emission scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5) of 20 CMIP6 models to determine the \u0026ldquo;signal-to-noise\u0026rdquo; ratio and compute the ToE for the 0-2000 m OHC change (Methods\u003csup\u003e28,57\u003c/sup\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eUnder the \u0026ldquo;middle-of-the-road\u0026rdquo; scenario of SSP2-4.5, a scenario viewed as where social and economic trends do not shift markedly from historical patterns, the anthropogenic heat storage (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea) has already emerged (e.g., ToE earlier than the 2020s) in the Southern Oceans between 55\u0026deg;-33\u0026deg;S and the tropical and northern Atlantic. In the North Pacific, the ToE is as early as the 2010s in marginal seas along the basin rim, including the western coasts of North America, the Bering Sea, the Okhotsk Sea (most of which is shallower than 2000 m and not estimated for OHC and ToE), and the Japan Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). By contrast, anthropogenic signals have not emerged by present (the 2020s) in the basin interior and even until the 2050s in the central MNP, the KE, and the NWTP where we see obvious model-observation discrepancies. The ToEs under SSP1-2.6 and SSP5-8.5 scenarios are similar in spatial distribution to those under SSP2-4.5 but postponed and advanced in average time (\u003cb\u003eExtended Data Fig.\u0026nbsp;8\u003c/b\u003e), respectively.\u003c/p\u003e \u003cp\u003eIn marginal seas of the North Pacific, such as the Bering Sea and the eastern boundary coasts, anthropogenic heat storage has already been emerging. The multi-model median ToE is during the 2010s and 2020s in these regions regardless of the emission scenario (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb, c). By contrast, the central MNP, KE, and NWTP regions show delayed ToEs and relatively high sensitivities to the emission scenario (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ed-f). Under the SSP1-2.6 and SSP2-4.5 scenarios, anthropogenic heat storage is projected to emerge in the latter half of this century in these regions, such as the 2080s in the KE region. Under the SSP5-8.5 scenario, the ToE is as early as the 2040s in the central MNP, the 2050s in the KE, and the 2060s in the NWTP. To summarize, despite uncertainties arising from inter-model spread, the North Pacific heat storage is projected to remain under a significant impact of natural variability until the late 21st century, except for marginal seas along the basin rim. Therefore, we expect that the heat storage pattern of the North Pacific will alter but probably still differ from that projected by climate models in the upcoming several decades.\u003c/p\u003e"},{"header":"Summary and Implications","content":"\u003cp\u003eBased on observational datasets, ocean model experiments, and CMIP6 model simulations, we demonstrate a strong modification effect by the PDO on the observed heat storage pattern in the North Pacific since the mid-20th century. Specifically, phase shifts of the PDO in recent decades have altered the trends in surface winds over the North Pacific. Surface wind changes induced meridional heat redistribution through Rossby waves and variability of the KE system, effectively erasing the warming in the central MNP and fueling the heat pile-up in the NWTP. These effects led to a more complex heat storage pattern in observations than in models by creating regional warming/cooling structures that mask the human-induced fingerprints (\u003cb\u003eExtended Data Fig.\u0026nbsp;9\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eRecognition of the strong influence of PDO on heat storage provides critical insights into near-term climate prediction. The persistent negative phase of PDO since 1998, critical in shaping the multi-decadal trends of surface winds and heat storage, terminated in the mid-2010s (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec). If the PDO switches to a positive phase in the upcoming decade, the central MNP \u0026ldquo;warming hole\u0026rdquo; shall vanish soon and the basin-scale heat storage will evolve toward the pattern projected by models. However, the recent triple La Ni\u0026ntilde;a events in 2020-2023\u003csup\u003e58,59\u003c/sup\u003e and the possible developing La Ni\u0026ntilde;a condition in 2024 and their extratropical impacts indicate a likelihood of a lengthened negative phase, which acts to maintain the existing heat storage pattern. The broad impacts of the PDO on surface atmospheric and ocean conditions have been well established, such as storm tracks, atmospheric rivers, and Pacific salmon production\u003csup\u003e60\u0026ndash;62\u003c/sup\u003e. Here, we add that the PDO is also critical for the transient response of the North Pacific to anthropogenic forcing. Through modulating wind-driven ocean circulations, the PDO fundamentally regulates the distributions of anthropogenic heat, CO\u003csub\u003e2\u003c/sub\u003e, and the Fukushima nuclear effluents in the North Pacific and their spread into the Indian and Arctic Oceans via the Indonesian through-flow\u003csup\u003e26,27\u003c/sup\u003e and Bering Strait\u003csup\u003e63\u003c/sup\u003e, respectively These implications highlight the urgent need for an accurate initialized prediction of the PDO and its far-reaching impacts.\u003c/p\u003e \u003cp\u003eClimate models tend to underestimate the PDO variability and its impacts over the historical period\u003csup\u003e64\u003c/sup\u003e. As a result, none of the 20 CMIP6 models has produced a historical heat storage pattern resembling the observation (not shown). Increasing model resolution seems promising. The representation of natural variability in the North Pacific is considerably improved in ~\u0026thinsp;10 km simulations compared to ~\u0026thinsp;100 km simulations\u003csup\u003e65,66\u003c/sup\u003e; our analysis based on the 0.1\u0026deg; HYCOM simulation also supports this notion. In addition, increasing ensemble members can better account for the extreme cases of natural variability, such as the 1998\u0026ndash;2014 negative phase\u003csup\u003e67\u003c/sup\u003e, and improve the ToE estimate.\u003c/p\u003e \u003cp\u003eThe human-induced heat storage pattern in the Pacific Ocean is projected to eventually emerge by the late 21st century, which is much later than in the Atlantic and Southern Oceans. Apart from the strong impacts of natural variability, the postponed ToEs in the Pacific are linked to the lack of a deep-reaching meridional overturning circulation, which also holds true for the Indian Ocean. In the North Atlantic and Southern Oceans, overturning circulations can efficiently communicate anthropogenic heat to the deep ocean, enhancing the average heat uptake\u003csup\u003e68\u003c/sup\u003e and the signal-to-noise ratio in heat storage. Nevertheless, human-induced warming has already been emerging in marginal seas along the North Pacific basin rim, such as the Bering Sea, Okhotsk Sea, Japan Sea, Gulf of Alaska, and western coasts of North America. Increasing frequency and severity of marine heatwaves have been extensively reported recently\u003csup\u003e69\u0026ndash;71\u003c/sup\u003e, with adverse influences on marine ecosystems and regional socioeconomics\u003csup\u003e72,73\u003c/sup\u003e. Anthropogenic fingerprints in these marginal seas could be enhanced by influence from lands where the air temperature warms quicker than in oceans\u003csup\u003e74\u003c/sup\u003e or coastal processes that are insufficiently resolved by climate models. This calls for extensive investigations to reveal the mechanisms underlying the warming of marginal seas and multi-disciplinary implications, which shall aid the prediction and decision-making.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eDatasets.\u003c/strong\u003e To estimate the 0-2000 m ocean heat content (OHC) trend since the mid-20th century, ocean temperature data since 1958 from four observational analyses and one reanalysis product are utilized. The four observational analyses are the Institute of Atmospheric Physics ocean analysis (IAP) for 1958\u0026ndash;2021 provided by the Chinese Academy of Sciences\u003csup\u003e35\u003c/sup\u003e, the World Ocean Atlas (WOA) for 1958\u0026ndash;2018 provided by the National Oceanic and Atmospheric Administration\u0026rsquo;s (NOAA\u0026rsquo;s) National Centers for Environmental Information of the United States\u003csup\u003e37\u003c/sup\u003e; the Ishii analysis for 1958-2021\u003csup\u003e36\u003c/sup\u003e; and the version 4.2.2 of the Met Office Hadley Centre \u0026lsquo;\u0026lsquo;EN\u0026rsquo;\u0026rsquo; series of datasets (EN4) for 1958\u0026ndash;2021 provided by the Met Office of the United Kingdom\u003csup\u003e38\u003c/sup\u003e. The one ocean reanalysis product is the Ocean Reanalysis System 4 (ORA-S4) for 1958-2017\u003csup\u003e39\u003c/sup\u003e from the European Centre for Medium-Range Weather Forecasts (ECMWF). All five datasets provide 1\u0026deg;\u0026times;1\u0026deg; ocean temperature fields. WOA provides pentad-mean data in annual intervals, and monthly fields of IAP, Ishii, EN4, and ORA-S4 are averaged into annual-mean data. Surface winds of four atmospheric reanalysis products are used, including the 1.25\u0026deg;\u0026times;1.25\u0026deg; Japanese 55-year Reanalysis (JRA55) product for 1958-2021\u003csup\u003e41\u003c/sup\u003e, the 1\u0026deg;\u0026times;1\u0026deg; ORA-S4 product as a combination of the 40-year ECMWF Re-Analysis (ERA-40) of 1958-1988\u003csup\u003e42\u003c/sup\u003e and the ECMWF Interim Re-Analysis (ERA-Interim) of 1989-2017\u003csup\u003e43\u003c/sup\u003e, the 2.5\u0026deg;x2.5\u0026deg; National Centers for Environmental Prediction-National Center for Atmospheric Research (NCEP-NCAR) reanalysis for 1958-2021\u003csup\u003e44\u003c/sup\u003e, and the 0.25\u0026deg;x0.25\u0026deg; fifth generation ECMWF reanalysis (ERA5) for 1958-2021\u003csup\u003e75\u003c/sup\u003e. To examine the changes in the Kuroshio Extension (KE) jet, the 0.25\u0026deg;x0.25\u0026deg; sea surface height data distributed by the Copernicus Marine and Environment Monitoring Service (CMEMS) for 1993\u0026ndash;2021 is used.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe coupled model simulations.\u003c/strong\u003e Monthly outputs of the pre-industrial control (piControl) (the last 100 years), historical (from 1958 to 2014), Shared Socioeconomic Pathways 126 (SSP1-2.6), SSP2-4.5, and SSP5-8.5 (from 2015 to 2100) simulations of twenty models of the Coupled Model Intercomparison Project Phase 6 (CMIP6) are analyzed. These models include the following: ACCESS-CM2, ACCESS-ESM1-5, BCC-CSM2-MR, CAMS-CSM1-0, CESM2-WACCM, CNRM-CM6-1, CanESM5, EC-Earth3, EC-Earth3-Veg, FGOALS-g3, FIO-ESM-2-0, GFDL-ESM4, GISS-E2-1-G, INM-CM4-8, INM-CM5-0, MPI-ESM1-2-LR, MRI-ESM2-0, NESM3, NorESM2-LM, NorESM2-MM. The multi-model mean (MMM) of 20 CMIP6 models represents the externally forced trend, and the inter-model spread represents the natural climate variability and difference in model physics. Monthly fields of CMIP6 simulations are interpolated onto the same set of 1\u0026deg;\u0026times;1\u0026deg; grids and averaged into annual-mean data to match observational datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHYCOM experiments.\u003c/strong\u003e We use the Hybrid Coordinate Ocean Model (HYCOM) version-2.3.01\u003csup\u003e76\u003c/sup\u003e to simulate and understand the ocean heat storage pattern. The HYCOM 0.5\u0026deg; simulation adopts a quasi-global ocean domain (78\u0026deg;S-75\u0026deg;N, 30\u0026deg;-389.5\u0026deg;E) with resolutions of 0.5\u0026deg; \u0026times; 0.5\u0026deg; \u0026times; 50 levels. The model is spun up from a state of rest for 600 years under the daily JRA55 atmospheric forcing of 1958, and the atmospheric surface forcing fields include the wind stress, net shortwave and longwave radiations (SWR and LWR), wind speed, precipitation, and air temperature and humidity. The control run (CTRL) is forced with realistic daily atmospheric forcings of JRA55 reanalysis from 1958 to 2019. CTRL contains complete physical processes of forming ocean heat storage pattern and is compared with observational data to evaluate the model performance. Two sensitivity runs are conducted to explore the effects of different drivers on the ocean heat storage pattern. Only wind stress adopts realistic daily fields in wind stress run (WND), while all the other forcings, such as wind speed and radiations, are fixed to the 1958 fields as in spin-up. In the WND, OHC trends primarily arise from heat redistribution due to wind stress-driven changes in ocean circulation processes. The heat flux run (HTFL) only adopts realistic SWR and LWR radiations, 2-m air temperature, specific humidity, and wind speed from 1958 to 2019 but fixes wind stress and precipitation to the 1958 fields. As such, changes in these fields of HTFL affect OHC trends mainly by altering the heat fluxes (SWR, LWR, sensible, and latent heat fluxes), representing the role of heat-driven processes or ocean heat uptake. To examine the impacts of decadally varying KE jet and its associated mesoscale eddies, we also use the HYCOM version-2.3.34\u003csup\u003e77\u003c/sup\u003e. The HYCOM 0.1\u0026deg; simulation uses the same domain as the HYCOM 0.5\u0026deg; simulation but with resolutions of 0.1\u0026deg; \u0026times; 0.1\u0026deg; \u0026times; 50 levels. After a 20-year spin-up, the HYCOM 0.1\u0026deg; simulation is integrated forwards from 1979 to 2021 under daily ERA5 atmospheric forcing. The comparison of the HYCOM 0.1\u0026deg; simulation and the HYCOM 0.5\u0026deg; CTRL allows us to test the possible impact of mesoscale oceanic processes on heat storage pattern in the North Pacific.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDefinitions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe OHC is calculated by integrating ocean temperature \u003cem\u003eT\u003c/em\u003e within the upper 2000 m as\u003c/p\u003e\n\u003cp\u003eOHC = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\int\\:}_{0}^{2000}\\rho\\:{c}_{p}Tdz\\)\u003c/span\u003e\u003c/span\u003e, (\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\n\u003cp\u003ewhere \u003cem\u003ec\u003c/em\u003e\u003csub\u003ep\u003c/sub\u003e = 4096 J (kg \u0026deg;C)\u003csup\u003e\u0026minus;1\u003c/sup\u003e is thermal capacity, and \u003cem\u003e\u0026rho;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1025 kg m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e is seawater density. The Ekman pumping velocity is calculated as \u003cem\u003e\u0026omega;\u003c/em\u003e\u003csub\u003e\u003cem\u003eE\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003ecurl(\u003c/em\u003e\u003cstrong\u003e\u0026tau;\u003c/strong\u003e\u003cem\u003e/f)\u0026rho;\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;1\u003c/em\u003e\u003c/sup\u003e, where \u003cstrong\u003e\u0026tau;\u003c/strong\u003e is the wind stress vector, and \u003cem\u003ef\u003c/em\u003e is the Coriolis parameter. The least square method is used to estimate the linear trend. Unless otherwise stated, the statistical significance of the trend is defined as \u0026gt;\u0026thinsp;95% confidence level based on a modified Mann-Kendall test. The Pacific decadal oscillation (PDO) index is calculated as the leading principle component from the empirical orthogonal function (EOF) analysis of North Pacific (poleward of 20\u0026deg;N) SST anomalies\u003csup\u003e61\u003c/sup\u003e, using the Met Office\u0026rsquo;s Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST)\u003csup\u003e78\u003c/sup\u003e for observation and historical and SSP5-8.5 simulations of the 20 CMIP6 models. The long-term trend is removed at each grid point before conducting the EOF analysis. The regressions of OHC and wind anomalies on the negative PDO index (-PDO) are first calculated from each of the 20 CMIP6 models, and then the MMM regression maps are derived from the ensemble mean of their coefficients. The Oceanic Ni\u0026ntilde;o Index (ONI) is the 3-month running-mean detrended temperature anomalies in the Ni\u0026ntilde;o-3.4 region (5\u0026deg;N-5\u0026deg;S, 120\u0026deg;-170\u0026deg;W) based on version 5 of the Extended Reconstructed Sea Surface Temperature\u003csup\u003e79\u003c/sup\u003e, and the monthly ONI is averaged to the annual mean.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTime of Emergence (ToE).\u003c/strong\u003e A \u0026ldquo;signal-to-noise\u0026rdquo; method\u003csup\u003e28,57\u003c/sup\u003e is adopted to compute the time of emergence (ToE) of the OHC trend. We use the piControl simulation, historical simulation, and projections under three emission scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5) of 20 CMIP6 models to determine the \u0026ldquo;signal-to-noise\u0026rdquo; ratio (SNR). The noise is defined as the interannual standard deviation of the quadratically detrended\u003csup\u003e80\u003c/sup\u003e OHC variability of piControl simulation in the last 100 years, representing the natural noise envelope. The anthropogenic signals from 1958 to 2100 are computed as the OHC differences between the historical plus three emission scenario simulations and the time-mean piControl value. The ToE is defined as the time when the signal last exceeds and remains above a threshold of two times of noise (i.e., SNR\u0026thinsp;\u0026gt;\u0026thinsp;2.0). The threshold for SNR is chosen as 2.0 because it represents a 95% confidence level of signal emergence\u003csup\u003e81\u003c/sup\u003e. For the regional ToEs, the noise and signal of selected regions are first averaged and then the time with SNR\u0026thinsp;\u0026gt;\u0026thinsp;2.0 is searched.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIAP data is available from http://www.ocean.iap.ac.cn/?navAnchor=home. WOA data is available from https://www.nodc.noaa.gov/OC5/indprod.html. Ishii data is available at https://www.data.jma.go.jp/gmd/kaiyou/english/ohc/ohc_global_en.html. EN4 (version 4.2.2) data is available at http://www.metoffice.gov.uk/hadobs/en4/index.html. ORA-S4 temperature and wind fields are available at https://icdc.cen.uni-hamburg.de/daten/reanalysis-ocean/easy-init-ocean/ecmwf-ocean-reanalysis-system-4-oras4.html and https://apps.ecmwf.int/datasets/. JRA55 data is available from https://jra.kishou.go.jp/JRA-55/index_en.html#mirror. EAR5 data can be downloaded at https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5. NCEP-NCAR data is available from https://psl.noaa.gov/data/gridded/tables/monthly.html. CMIP6 simulations are available at https://esgf-node.llnl.gov/projects/cmip6/. HadISST data is available from https://www.metoffice.gov.uk/hadobs/hadisst/. CMEMS data is downloaded at https://data.marine.copernicus.eu/product/SEALEVEL_GLO_PHY_L4_MY_008_047/description. ONI is obtained from https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php.\u003c/p\u003e\n\u003cp\u003eCode availability\u003c/p\u003e\n\u003cp\u003eThe HYCOM code used in this word is available via https://github.com/HYCOM. Matlab R2017a is used for plotting. The Matlab code for data analysis and graphing is available upon request.\u003c/p\u003e\n\u003cp\u003eAcknowledgments\u003c/p\u003e\n\u003cp\u003eThis work is supported by the Laoshan Laboratory (LSKJ202202601), National Natural Science Foundation of China (42176007), the Strategic Priority Research Program of Chinese Academy of Science (XDB42000000), and the Oceanographic Data Center, the Institute of Oceanology, Chinese Academy of Sciences.\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u003c/p\u003e\n\u003cp\u003eJ. D and Y. L. (Yuanlong Li) drafted the paper. Y. L. (Yuanlong Li) directed this work with contributions from all authors. J. D. and Y. L. (Yilong Lyu) contributed to ocean model data analysis. J. D. contributed to CMIP6 data analysis. All authors discussed the results and commented on the manuscript.\u003c/p\u003e\n\u003cp\u003eCorrespondence and requests for materials should be addressed to Y. L. (Yuanlong Li).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eIPCC. \u003cem\u003eClimate Change 2021: The Physical Science Basis\u003c/em\u003e. (Cambridge University Press, 2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFox-Kemper, B. \u003cem\u003eet al. Ocean,cryosphere and sea level change. Climate Change 2021: The Physical Science Basis\u003c/em\u003e. (Cambridge University Press, 2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng, L. \u003cem\u003eet al.\u003c/em\u003e Past and future ocean warming. Nat. Rev. Earth Environ. 3, 776\u0026ndash;794 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng, L. \u003cem\u003eet al.\u003c/em\u003e Another Year of Record Heat for the Oceans. Adv. Atmos. Sci. 40, 963\u0026ndash;974 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEyring, V. \u003cem\u003eet al.\u003c/em\u003e Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9, 1937\u0026ndash;1958 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuhlbrodt, T., Voldoire, A., Palmer, M. D., Geoffroy, O. \u0026amp; Killick, R. E. Historical Ocean Heat Uptake in Two Pairs of CMIP6 Models: Global and Regional Perspectives. J. Clim. 36, 2183\u0026ndash;2203 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGille, S. T. Warming of the Southern Ocean Since the 1950s. Science (80-.). 295, 1275\u0026ndash;1277 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi, J. R., Xie, S. P. \u0026amp; Talley, L. D. Evolving relative importance of the Southern Ocean and North Atlantic in anthropogenic ocean heat uptake. J. Clim. 31, 7459\u0026ndash;7479 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCai, W. \u003cem\u003eet al.\u003c/em\u003e Southern Ocean warming and its climatic impacts. Sci. Bull. 68, 946\u0026ndash;960 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng, Q. \u003cem\u003eet al.\u003c/em\u003e Surface warming\u0026ndash;induced global acceleration of upper ocean currents. Sci. Adv. 8, 1\u0026ndash;12 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTamarin-Brodsky, T. \u0026amp; Kaspi, Y. Enhanced poleward propagation of storms under climate change. Nat. Geosci. 10, 908\u0026ndash;913 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSwart, N. C. \u0026amp; Fyfe, J. C. Observed and simulated changes in the Southern Hemisphere surface westerly wind-stress. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e 39, n/a-n/a (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, S. \u003cem\u003eet al.\u003c/em\u003e Extreme atmospheric rivers in a warming climate. Nat. Commun. 14, 3219 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIbarbalz, F. M. \u003cem\u003eet al.\u003c/em\u003e Global Trends in Marine Plankton Diversity across Kingdoms of Life. Cell 179, 1084\u0026ndash;1097.e21 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan, W. \u003cem\u003eet al.\u003c/em\u003e Intensification of decadal and multi-decadal sea level variability in the western tropical Pacific during recent decades. Clim. Dyn. 43, 1357\u0026ndash;1379 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamlington, B. D. \u003cem\u003eet al.\u003c/em\u003e Uncovering an anthropogenic sea-level rise signal in the Pacific Ocean. Nat. Clim. Chang. 4, 782\u0026ndash;785 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHughes, T. P. \u003cem\u003eet al.\u003c/em\u003e Spatial and temporal patterns of mass bleaching of corals in the Anthropocene. Science (80-.). 359, 80\u0026ndash;83 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmale, D. A. \u003cem\u003eet al.\u003c/em\u003e Marine heatwaves threaten global biodiversity and the provision of ecosystem services. Nat. Clim. Chang. 9, 306\u0026ndash;312 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOliver, E. C. J. \u003cem\u003eet al.\u003c/em\u003e Longer and more frequent marine heatwaves over the past century. Nat. Commun. 9, 1324 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKossin, J. P., Emanuel, K. A. \u0026amp; Vecchi, G. A. The poleward migration of the location of tropical cyclone maximum intensity. Nature 509, 349\u0026ndash;352 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShan, K., Lin, Y., Chu, P., Yu, X. \u0026amp; Song, F. Seasonal advance of intense tropical cyclones in a warming climate. Nature (2023) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-023-06544-0\u003c/span\u003e\u003cspan address=\"10.1038/s41586-023-06544-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMei, W., Xie, S. P., Primeau, F., McWilliams, J. C. \u0026amp; Pasquero, C. Northwestern Pacific typhoon intensity controlled by changes in ocean temperatures. Sci. Adv. 1, (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu, L. \u003cem\u003eet al.\u003c/em\u003e Enhanced warming over the global subtropical western boundary currents. Nat. Clim. Chang. 2, 161\u0026ndash;166 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, F. \u003cem\u003eet al.\u003c/em\u003e The seas around China in a warming climate. Nat. Rev. Earth Environ. (2023) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s43017-023-00453-6\u003c/span\u003e\u003cspan address=\"10.1038/s43017-023-00453-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu, D. \u003cem\u003eet al.\u003c/em\u003e Pacific western boundary currents and their roles in climate. Nature 522, 299\u0026ndash;308 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, S. K. \u003cem\u003eet al.\u003c/em\u003e Pacific origin of the abrupt increase in Indian Ocean heat content during the warming hiatus. Nat. Geosci. 8, 445\u0026ndash;449 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, Y., Han, W. \u0026amp; Zhang, L. Enhanced Decadal Warming of the Southeast Indian Ocean During the Recent Global Surface Warming Slowdown. Geophys. Res. Lett. 44, 9876\u0026ndash;9884 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLyu, K., Zhang, X., Church, J. A., Slangen, A. B. A. \u0026amp; Hu, J. Time of emergence for regional sea-level change. Nat. Clim. Chang. 4, 1006\u0026ndash;1010 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilvy, Y., Sall\u0026eacute;e, J. B., Guilyardi, E., Mignot, J. \u0026amp; Rousset, C. What Causes Anthropogenic Ocean Warming to Emerge from Internal Variability in a Coupled Model? J. Clim. 35, 3835\u0026ndash;3854 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatanabe, M., Dufresne, J. L., Kosaka, Y., Mauritsen, T. \u0026amp; Tatebe, H. Enhanced warming constrained by past trends in equatorial Pacific sea surface temperature gradient. Nat. Clim. Chang. 11, 3\u0026ndash;9 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu, B., Lin, X. \u0026amp; Yu, L. North Pacific subtropical mode water is controlled by the Atlantic Multidecadal Variability. Nat. Clim. Chang. 10, 238\u0026ndash;243 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEngland, M. H. \u003cem\u003eet al.\u003c/em\u003e Recent intensification of wind-driven circulation in the Pacific and the ongoing warming hiatus. Nat. Clim. Chang. 4, 222\u0026ndash;227 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNieves, V., Willis, J. K. \u0026amp; Patzert, W. C. Recent hiatus caused by decadal shift in Indo-Pacific heating. Science (80-.). 349, 532\u0026ndash;535 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, W., Xie, S. P. \u0026amp; Lu, J. Tracking ocean heat uptake during the surface warming hiatus. Nat. Commun. 7, 1\u0026ndash;9 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng, L. \u003cem\u003eet al.\u003c/em\u003e Improved estimates of ocean heat content from 1960 to 2015. Sci. Adv. 3, (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIshii, M. \u003cem\u003eet al.\u003c/em\u003e Accuracy of Global Upper Ocean Heat Content Estimation Expected from Present Observational Data Sets. SOLA 13, 163\u0026ndash;167 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLevitus, S. \u003cem\u003eet al.\u003c/em\u003e World ocean heat content and thermosteric sea level change (0-2000 m), 1955\u0026ndash;2010. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e 39, n/a-n/a (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGood, S. A., Martin, M. J. \u0026amp; Rayner, N. A. EN4: Quality controlled ocean temperature and salinity profiles and monthly objective analyses with uncertainty estimates. J. Geophys. Res. Ocean. 118, 6704\u0026ndash;6716 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalmaseda, M. A., Mogensen, K. \u0026amp; Weaver, A. T. Evaluation of the ECMWF ocean reanalysis system ORAS4. Q. J. R. Meteorol. Soc. 139, 1132\u0026ndash;1161 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChemke, R., Zanna, L. \u0026amp; Polvani, L. M. Identifying a human signal in the North Atlantic warming hole. Nat. Commun. 11, 1\u0026ndash;7 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarada, Y. \u003cem\u003eet al.\u003c/em\u003e The JRA-55 reanalysis: Representation of atmospheric circulation and climate variability. J. Meteorol. Soc. Japan 94, 269\u0026ndash;302 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUppala, S. M. \u003cem\u003eet al.\u003c/em\u003e The ERA-40 re-analysis. Q. J. R. Meteorol. Soc. 131, 2961\u0026ndash;3012 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDee, D. P. \u003cem\u003eet al.\u003c/em\u003e The ERA-Interim reanalysis: configuration and performance of the data assimilation system. Q. J. R. Meteorol. Soc. 137, 553\u0026ndash;597 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKalnay, E. \u003cem\u003eet al.\u003c/em\u003e The NCEP/NCAR 40-Year Reanalysis Project. Bull. Am. Meteorol. Soc. 77, 437\u0026ndash;472 (1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuan, J., Li, Y., Cheng, L., Lin, P. \u0026amp; Wang, F. Heat Storage in the Upper Indian Ocean: The Role of Wind-Driven Redistribution. J. Clim. 36, 2221\u0026ndash;2242 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu, B. \u0026amp; Lukas, R. Seasonal and interannual variability of the North Equatorial Current, the Mindanao Current, and the Kuroshio along the Pacific western boundary. J. Geophys. Res. Ocean. 101, 12315\u0026ndash;12330 (1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu, B. \u0026amp; Chen, S. Variability of the Kuroshio Extension Jet, Recirculation Gyre, and Mesoscale Eddies on Decadal Time Scales. J. Phys. Oceanogr. 35, 2090\u0026ndash;2103 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu, B., Chen, S., Schneider, N., Oka, E. \u0026amp; Sugimoto, S. On the Reset of the Wind-Forced Decadal Kuroshio Extension Variability in Late 2017. J. Clim. 33, 10813\u0026ndash;10828 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu, B. \u0026amp; Chen, S. Effect of Decadal Kuroshio Extension Jet and Eddy Variability on the Modification of North Pacific Intermediate Water. J. Phys. Oceanogr. 41, 503\u0026ndash;515 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaguchi, B. \u003cem\u003eet al.\u003c/em\u003e Decadal Variability of the Kuroshio Extension: Observations and an Eddy-Resolving Model Hindcast*. J. Clim. 20, 2357\u0026ndash;2377 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu, B. \u0026amp; Chen, S. Interannual-to-Decadal Variability in the Bifurcation of the North Equatorial Current off the Philippines. J. Phys. Oceanogr. 40, 2525\u0026ndash;2538 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKawabe, M. Variations of Current Path, Velocity, and Volume Transport of the Kuroshio in Relation with the Large Meander. J. Phys. Oceanogr. 25, 3103\u0026ndash;3117 (1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiu, B., Chen, S. \u0026amp; Oka, E. Why Did the 2017 Kuroshio Large Meander Event Become the Longest in the Past 70 Years? Geophys. Res. Lett. 50, 1\u0026ndash;9 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilva, E. N. S. \u0026amp; Anderson, B. T. Northeast Pacific marine heatwaves linked to Kuroshio Extension variability. Commun. Earth Environ. 4, 367 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa, X. \u003cem\u003eet al.\u003c/em\u003e Western boundary currents regulated by interaction between ocean eddies and the atmosphere. Nature 535, 533\u0026ndash;537 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman, M. \u003cem\u003eet al.\u003c/em\u003e The Pacific Decadal Oscillation, Revisited. J. Clim. 29, 4399\u0026ndash;4427 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilvy, Y., Guilyardi, E., Sall\u0026eacute;e, J. B. \u0026amp; Durack, P. J. Human-induced changes to the global ocean water masses and their time of emergence. Nat. Clim. Chang. 10, 1030\u0026ndash;1036 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones, N. Rare \u0026lsquo;triple\u0026rsquo; La Ni\u0026ntilde;a climate event looks likely - what does the future hold? \u003cem\u003eNature\u003c/em\u003e vol. 607 21 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeng, T. \u003cem\u003eet al.\u003c/em\u003e Emergence of changing Central-Pacific and Eastern-Pacific El Ni\u0026ntilde;o-Southern Oscillation in a warming climate. Nat. Commun. 13, 6616 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGan, B. \u0026amp; Wu, L. Seasonal and Long-Term Coupling between Wintertime Storm Tracks and Sea Surface Temperature in the North Pacific. J. Clim. 26, 6123\u0026ndash;6136 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMantua, N. J., Hare, S. R., Zhang, Y., Wallace, J. M. \u0026amp; Francis, R. C. A Pacific Interdecadal Climate Oscillation with Impacts on Salmon Production. Bull. Am. Meteorol. Soc. 78, 1069\u0026ndash;1079 (1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePayne, A. E. \u003cem\u003eet al.\u003c/em\u003e Responses and impacts of atmospheric rivers to climate change. Nature Reviews Earth and Environment vol. 1 143\u0026ndash;157 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, G. \u003cem\u003eet al.\u003c/em\u003e High-resolution modelling identifies the Bering Strait\u0026rsquo;s role in amplified Arctic warming. Nat. Clim. Chang. (2024) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41558-024-02008-z\u003c/span\u003e\u003cspan address=\"10.1038/s41558-024-02008-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFarneti, R. Modelling interdecadal climate variability and the role of the ocean. WIREs Clim. Chang. 8, (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChang, P. \u003cem\u003eet al.\u003c/em\u003e An Unprecedented Set of High-Resolution Earth System Simulations for Understanding Multiscale Interactions in Climate Variability and Change. J. Adv. Model. Earth Syst. 12, (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAn, B. \u003cem\u003eet al.\u003c/em\u003e The benefits of high-resolution models in simulating the Kuroshio Extension and its long-term changes. Clim. Dyn. 61, 5407\u0026ndash;5427 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChung, E. S. \u003cem\u003eet al.\u003c/em\u003e Reconciling opposing Walker circulation trends in observations and model projections. Nat. Clim. Chang. 9, 405\u0026ndash;412 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRen, Q. \u003cem\u003eet al.\u003c/em\u003e Substantial Warming of the Atlantic Ocean in CMIP6 Models. J. Clim. (2024) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/JCLI-D-23-0418.1\u003c/span\u003e\u003cspan address=\"10.1175/JCLI-D-23-0418.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi Lorenzo, E. \u0026amp; Mantua, N. Multi-year persistence of the 2014/15 North Pacific marine heatwave. Nat. Clim. Chang. 6, 1042\u0026ndash;1047 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmaya, D. J., Miller, A. J., Xie, S. \u0026amp; Kosaka, Y. Physical drivers of the summer 2019 North Pacific marine heatwave. Nat. Commun. 11, 1903 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong, Q., Wang, C., Yao, Y. \u0026amp; Fan, H. Unraveling the Indian monsoon\u0026rsquo;s role in fueling the unprecedented 2022 Marine Heatwave in the Western North Pacific. npj Clim. Atmos. Sci. 7, 90 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCavole, L. \u003cem\u003eet al.\u003c/em\u003e Biological Impacts of the 2013\u0026ndash;2015 Warm-Water Anomaly in the Northeast Pacific: Winners, Losers, and the Future. \u003cem\u003eOceanography\u003c/em\u003e 29, (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNielsen, J. M. \u003cem\u003eet al.\u003c/em\u003e Responses of ichthyoplankton assemblages to the recent marine heatwave and previous climate fluctuations in several Northeast Pacific marine ecosystems. Glob. Chang. Biol. 27, 506\u0026ndash;520 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLima, F. P. \u0026amp; Wethey, D. S. Three decades of high-resolution coastal sea surface temperatures reveal more than warming. Nat. Commun. 3, 704 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHersbach, H.. \u003cem\u003eet al.\u003c/em\u003e ERA5 monthly averaged data on single levels from 1940 to present. Copernicus Clim. Chang. Serv. Clim. Data Store (2023) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.24381/cds.f17050d7\u003c/span\u003e\u003cspan address=\"10.24381/cds.f17050d7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChassignet, E. P., Smith, L. T., Halliwell, G. R. \u0026amp; Bleck, R. North Atlantic Simulations with the Hybrid Coordinate Ocean Model (HYCOM): Impact of the Vertical Coordinate Choice, Reference Pressure, and Thermobaricity. J. Phys. Oceanogr. 33, 2504\u0026ndash;2526 (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWallcraft, a. J., Metzger, E. J. \u0026amp; Carrol, S. N. Software Design Description for the HYbrid Coordinate Ocean Model (HYCOM). 149 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRayner, N. A. Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res. 108, 4407 (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, B. \u003cem\u003eet al.\u003c/em\u003e Extended Reconstructed Sea Surface Temperature, Version 5 (ERSSTv5): Upgrades, Validations, and Intercomparisons. J. Clim. 30, 8179\u0026ndash;8205 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCai, W. \u003cem\u003eet al.\u003c/em\u003e Butterfly effect and a self-modulating El Ni\u0026ntilde;o response to global warming. Nature 585, 68\u0026ndash;73 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHawkins, E. \u0026amp; Sutton, R. Time of emergence of climate signals. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e 39, n/a-n/a (2012).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4905116/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4905116/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eStorage of anthropogenic heat in the oceans is spatially inhomogeneous, impacting regional climates and human societies. Climate models project enhanced heat storage in the mid-latitude North Pacific (MNP) and much weaker storage in the tropical Pacific. However, the observed heat storage during the past half-century shows a more complex pattern, with limited warming in the central MNP and enhanced warming in the northwest tropical Pacific. Based on observational datasets, ocean model experiments, and climate models, we show that emergence of human-induced heat storage is likely postponed in the North Pacific by natural variability to the late-21st century. Specifically, phase shifts of the Pacific Decadal Oscillation (PDO) have vitally contributed to trends in the North Pacific winds during recent decades. Changes in surface winds drove meridional heat redistribution via Rossby wave dynamics, leading to regional warming and cooling structures and a more complex historical heat storage than models project. Despite this, enhanced anthropogenic warming has already been emerging in marginal seas along the North Pacific basin rim, for which we shall prepare for the pressing consequences such as increasing marine heatwaves.\u003c/p\u003e","manuscriptTitle":"Emergence of the North Pacific heat storage pattern delayed by decadal wind-driven redistribution","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-09-04 06:29:51","doi":"10.21203/rs.3.rs-4905116/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"432aa2e1-38f8-401f-b1b6-21ef5c8c2dc6","owner":[],"postedDate":"September 4th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":36107853,"name":"Earth and environmental sciences/Ocean sciences/Physical oceanography"},{"id":36107854,"name":"Earth and environmental sciences/Climate sciences/Ocean sciences/Physical oceanography"},{"id":36107855,"name":"Earth and environmental sciences/Climate sciences/Climate change/Attribution"},{"id":36107856,"name":"Earth and environmental sciences/Climate sciences/Climate change/Climate and Earth system modelling"}],"tags":[],"updatedAt":"2025-01-15T08:09:31+00:00","versionOfRecord":{"articleIdentity":"rs-4905116","link":"https://doi.org/10.1038/s41467-025-56005-7","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2025-01-14 05:00:00","publishedOnDateReadable":"January 14th, 2025"},"versionCreatedAt":"2024-09-04 06:29:51","video":"","vorDoi":"10.1038/s41467-025-56005-7","vorDoiUrl":"https://doi.org/10.1038/s41467-025-56005-7","workflowStages":[]},"version":"v1","identity":"rs-4905116","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4905116","identity":"rs-4905116","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}