Mega-heatwaves, characterized by long durations and large amplitudes, pose significant threats to ecosystems and human well-being, particularly in regions like central-eastern China, which is experiencing accelerating climate change and an aging population. The 2022 mega-heatwave in central-eastern China, with temperatures exceeding 40°C in some areas, was among the most severe globally, affecting nearly a billion people and causing an energy crisis. Understanding the causal factors of such events and improving heatwave prediction are crucial for effective disaster mitigation. While existing studies attributed the 2022 heatwave to factors like atmospheric circulation, sea surface temperatures, and local feedback mechanisms, a comprehensive quantitative attribution of multiple drivers was lacking. Heatwave development, maintenance, and attenuation involve complex interactions of dynamical and radiative processes. Clear skies associated with high-pressure systems allow increased solar radiation to reach the surface, inducing warming and maintaining the high-pressure system. Land-atmosphere feedback and high-pressure system collapses contribute to heatwave termination. Aerosols and water vapor also modulate heatwave evolution. Existing attribution methods, such as the partial radiative perturbation method (PRP) and coupled atmosphere-surface climate feedback-response analysis method (CFRAM), are suitable for equilibrium states but not for extreme events. Moist static energy (MSE) budgets have been used for intra-seasonal oscillations but not comprehensively for heatwave attribution. This study aims to address this gap by developing a new framework.

Numerous studies have examined heatwaves, highlighting their devastating impacts and the role of various factors in their formation and dissipation. Research has focused on the impact of high-pressure systems, clear skies, surface warming, and land-atmosphere interactions. The influence of aerosols and water vapor has also been investigated. However, most studies lack a comprehensive quantitative attribution of multiple dynamical and radiative drivers during various phases of a heatwave’s lifecycle. Existing attribution methods have limitations when applied to the non-equilibrium conditions of extreme events like mega-heatwaves. The study addresses the need for a framework that can account for the complex interplay of factors that influence the evolution of heatwaves.

This study developed a comprehensive process-resolving, energetics-based attribution framework (PREAF) to quantify the contributions of various physical processes to the 2022 mega-heatwave. The framework uses the total energy budget equation, which considers the balance between temperature change-induced longwave cooling and various physical processes. The PREAF quantifies the contributions from atmospheric dynamics (horizontal and vertical advection), surface heat fluxes (latent and sensible), and radiative drivers (solar insolation, ozone, albedo, temperature, water vapor, cloud, and aerosols). Observational datasets included daily maximum 2m air temperature and precipitation (CPC), Standardized Precipitation Index (SPI), surface evaporation and soil moisture (GLEAM), cloud fraction, precipitable water, and aerosol optical depth (MODIS). Reanalysis data from MERRA-2 provided atmospheric variables. The NCEP CFSv2 sub-seasonal prediction model was used to evaluate the framework's ability to identify prediction biases. Heatwaves were defined as periods when daily maximum 2m air temperature exceeded the 90th percentile of the 2003-2022 climatology for at least three consecutive days. The PREAF decomposes the total energy tendency into contributions from individual processes, using linearized radiative energy perturbations and offline radiative transfer calculations (RRTMG and Monte Carlo ICA). The study focused on the most intense and prolonged heatwave event, from late July to late August 2022. The lifecycle was divided into developing, mature, and decaying phases based on temperature and total energy anomalies.

The PREAF analysis revealed the following key findings regarding the 2022 mega-heatwave: **Developing Phase:** Cloudless skies and atmospheric dynamics (vertical and horizontal advection) dominated the positive total energy tendency of the land surface. Reduced aerosols contributed positively to surface incoming radiation, partially compensated by longwave cooling. Surface latent and sensible heat fluxes transported energy from the land surface to the atmosphere. Atmospheric energy increase was mainly driven by horizontal advection and surface heat fluxes, with radiative processes contributing negatively. **Mature Phase:** Radiative processes (clouds and aerosols) continued to contribute positively to land surface energy. The longwave effect of increased water vapor became more significant. Surface latent heat flux weakened due to drying soil, while sensible heat flux intensified. In the atmosphere, surface heat fluxes contributed positively, along with vertical advection. The positive total energy tendency was partially offset by negative radiative processes. **Decaying Phase:** Total energy decreased rapidly. Increased low-level clouds terminated land surface radiative heating. Upward sensible heat flux and downward latent heat flux contributed to negative land surface energy tendency. Horizontal advection (cold advection) dominated the negative atmospheric energy tendency, partially offset by positive vertical advection and surface heat fluxes. Radiative process effects largely compensated each other. Analysis of the NCEP CFSv2 model showed that it accurately reproduced the developing phase, but predicted an earlier-than-observed decay. The model's fidelity in representing various physical processes accounted for the good prediction skill in the developing phase. However, misrepresentations of atmospheric dynamics (vertical advection), hydrological processes (land-atmosphere coupling and water vapor effect), and surface heat fluxes were responsible for the poor prediction skill during the decaying phase. The model overestimated the soil moisture response to precipitation, leading to an inaccurate representation of surface latent heat flux.

The study’s findings highlight the importance of considering both near-surface temperature and energetics, along with land-atmosphere coupling, when defining and attributing heatwaves. The PREAF effectively quantifies the contributions of multiple processes to heatwave evolution. Atmospheric dynamics, particularly eastward eddy shedding and cold air outbreaks, played key roles in heatwave initiation and termination. Land-atmosphere interaction was crucial for maintaining the heatwave's longevity. Reduced anthropogenic aerosols, likely due to emission reduction policies, contributed positively to the heatwave, suggesting that pollution mitigation efforts might inadvertently increase future heatwave intensity. The comparison of the PREAF results with the CFSv2 model demonstrates its utility in identifying sources of sub-seasonal prediction biases, highlighting the importance of accurate representation of land-atmosphere interactions, atmospheric dynamics, and hydrological processes in improving heatwave prediction models.

This study demonstrates the effectiveness of the PREAF framework in quantitatively attributing the 2022 mega-heatwave's lifecycle to multiple physical processes. The findings underscore the complex interplay of atmospheric dynamics, land-atmosphere coupling, and radiative forcing, particularly the unexpected positive contribution of reduced aerosols. The application of PREAF to sub-seasonal model forecasts reveals key areas for improvement in heatwave prediction. Future research could apply PREAF to other heatwave events and models, exploring its applicability to a wider range of weather and climate phenomena. Further investigation is needed to assess the robustness of these findings across different models and events.

Several limitations warrant consideration. The accuracy of the quantitative attribution is constrained by the uncertainties inherent in observational datasets, particularly regarding clouds and aerosols. The absence of an aerosol module in the CFSv2 model limited the assessment of the model's performance regarding aerosol effects. Furthermore, this study presents a case study of a single mega-heatwave and evaluates only one member of the CFSv2. While the main findings are expected to hold true for other events, further research is needed to confirm their generalizability across various heatwaves and models.