Marine heatwaves (MHWs), analogous to atmospheric heatwaves, are increasing in frequency, intensity, and duration due to global warming. A significant MHW occurred in the Northeast Pacific Ocean during summer 2019, reaching record-breaking sea surface temperatures (SSTs) and causing considerable ecological damage, including impacting the Bering Sea snow crab stock. Understanding the drivers of these events and projecting their future behavior under a warming climate is crucial for effective conservation and management strategies. The 2019 event was characterized by weakened North Pacific high pressure, reduced surface winds, shallower mixed layer depths (MLDs), and reduced low-cloud cover. These factors contributed to decreased ocean mixing and increased absorption of solar radiation. This study addresses two key questions: 1) Was the 2019 MHW influenced by anthropogenic climate change? and 2) How would such an event unfold in an even warmer future? The study employs a novel approach to answer these questions, separating the long-term mean temperature change from event-specific temperature changes resulting from local processes, using storyline simulations.

Traditional attribution studies of extreme events like MHWs rely on statistical methods analyzing changes in frequency, duration, and intensity across different climate periods. These studies consistently show that the likelihood of SSTs like those observed in summer 2019 has increased due to anthropogenic warming. However, probabilistic methods require extensive data and large model ensembles, introducing computational costs and challenges in defining meaningful event boundaries and thresholds. They can also cluster diverse event subtypes with varying responses to climate change. This study adopts an alternative approach using storyline simulations, which nudge a coupled climate model to observed atmospheric wind patterns while allowing other variables to evolve freely under different climate conditions (preindustrial, present-day, +4°C warmer). This isolates the thermodynamic effects of climate change, reducing uncertainties related to natural atmospheric variability and providing a high signal-to-noise ratio at lower computational cost.

The study utilizes the AWI-CM-1-MR coupled climate model, which has been validated extensively and used in CMIP6. Two types of simulations were performed: free-running simulations spanning 1850-2100 with natural and anthropogenic forcings, and nudged storyline simulations. The storyline simulations used initial conditions from the free runs and imposed the observed wind evolution from 2017-2020, creating analogues of the 2019 MHW under different climate scenarios. The atmospheric circulation was constrained by nudging the vorticity and divergence with a relaxation timescale of 24 hours, affecting levels between 700 hPa and 100 hPa. The storylines were initialized using states from the free-running simulations representing preindustrial, present-day, and +4°C warmer climates. Each storyline consisted of five ensemble members. SST anomalies were calculated relative to a 1984-2014 climatology, both for ERA5 reanalysis data and model simulations. The storyline approach allows for a comparison of the MHW under different climates, separating the total warming signal into regional mean warming (from free-running simulations) and event-specific warming (the difference between storyline warming and regional mean warming). This methodology allows assessing if event-specific processes amplify or dampen warming.

The nudged storyline simulations realistically reproduced the spatial and temporal patterns of the 2019 Northeast Pacific MHW, validating the approach. Comparing preindustrial and present-day storylines revealed a ubiquitous increase in summer SSTs across the Northeast Pacific due to anthropogenic warming. The storyline warming exceeded the global mean SST increase (1°C between preindustrial and present-day), reaching 1.4 ± 0.2 °C in the MHW core and coastal region, and 1.7 ± 0.2 °C in the central North Pacific. In the MHW core and coastal regions, the regional mean warming largely accounted for the storyline warming, with event-specific processes showing a slight dampening effect. Conversely, in the central North Pacific, event-specific processes significantly amplified the warming. Projections for a +4°C warmer world showed a substantial increase in MHW intensity (2-3°C warmer than present-day conditions), reaching 23.7°C in the core, exceeding the projected global mean SST increase (1.9°C) by 1°C (52%). This amplification was primarily driven by changes in surface heat fluxes (increased net shortwave radiation due to reduced low-cloud cover), with a relatively minor contribution from changes in MLD. However, event-specific MLD shoaling significantly contributed to the warming amplification in the coastal region and central North Pacific. Anomalous northerly winds advecting warmer air from subpolar regions may also play a role in the central North Pacific amplification.

The study's findings demonstrate that event-specific processes significantly amplify future MHWs beyond the projected regional mean warming. The amplification is primarily driven by changes in surface heat fluxes, particularly reduced low cloud cover and increased shortwave radiation, highlighting a positive low-cloud SST feedback. The shoaling of the mixed layer also plays a crucial role, particularly in regions peripheral to the MHW core. While the model successfully reproduces the 2019 event, some limitations exist. Underestimation of MHW intensity in CMIP6 models like AWI-CM-1, likely due to coarse ocean resolution, may affect the quantitative results. The study's focus on the 2019 event limits the generalizability of the findings; however, the storyline approach offers a valuable tool for understanding how various MHWs respond to climate change. The interplay between MLD shoaling, changes in surface heat flux, and air advection varies across the MHW region, underscoring the complexity of MHW dynamics. The study highlights the critical need for further research into ocean dynamical processes and small-scale features to enhance MHW projections.

This study demonstrates, using a novel storyline approach, that future Northeast Pacific summer marine heatwaves will be amplified beyond the projected regional mean warming. This amplification is primarily driven by changes in surface heat fluxes related to reduced low-cloud cover and mixed layer shoaling, combined with air advection patterns. These findings underscore the substantial risk of intensified and more extensive MHWs under a warming climate, emphasizing the importance of continuing research to improve projections and inform mitigation and adaptation strategies. Further work should investigate the role of ocean dynamics and small-scale processes in modifying MHW behavior.

The study's primary limitation is its focus on a single MHW event. While the storyline approach is valuable, it's important to consider that the atmospheric circulation patterns causing the 2019 event may not be representative of all future MHWs. Furthermore, the medium resolution of the AWI-CM-1-MR model might not fully capture small-scale ocean features that can influence MHW development and intensity. Future studies should analyze multiple MHW events and utilize higher-resolution models to obtain more comprehensive insights into the complex processes driving their evolution under climate change.