An energetics tale of the 2022 mega-heatwave over central-eastern China

Heatwaves pose widespread yet disproportionate threats to the ecosystems and human well-being across the globe, and these associated adverse effects are further exacerbated by mega-heatwaves with long durations and large amplitude. Central-eastern China, with accelerating climate change and population aging, is particularly vulnerable to increasing and intensifying heatwaves. In the mid-to-late summer of 2022, a record-breaking mega-heatwave roasted broad swathes of China, among which the central-eastern region is considered the hardest-hit. This event with scorching temperature exceeding 40 °C at some stations, superimposed by droughts and wildfires, is among the most severe recorded in global history. Close to one billion people were affected by the heat and related water and electricity shortages. Given that more deadly heatwaves are projected in the future, quantifying the causal factors of mega-heatwaves and disentangling the origin of uncertainties in heatwave prediction can provide impact-based decision-making support for disaster mitigation and prevention. While several studies attributed the extremely hot summer in 2022 to extra-tropical atmospheric circulation, tropical sea surface temperature, and local soil moisture-temperature feedback, a comprehensive quantitative attribution of multiple dynamical and radiative drivers on this mega-heatwave has been absent. Typically, the development, maintenance, and attenuation of heatwaves involve both dynamical and radiative processes. During development and maintenance, a high-pressure system produces clear skies allowing more solar radiation to reach the ground, inducing anomalous surface warming. Upward surface heat fluxes are generated, along with adiabatic subsidence, leading to an increase in air temperature, maintaining the local high-pressure system and hence the heatwave. Collapses of the high-pressure system and local land-atmosphere feedback contribute to termination. The evolution may also be modulated by radiative processes associated with aerosols and water vapor. For example, increased aerosols can reduce surface temperature and heatwave probability through absorption or scattering of shortwave radiation, and vice versa. A fundamental question remains: What are the relative roles of those dynamical and radiative processes in different phases of heatwaves? According to the total energy balance equation, temperature change-induced longwave cooling is balanced by multiple physical processes in an equilibrium state. Several climate attribution methods, including the partial radiative perturbation method (PRP) and the coupled atmosphere-surface climate feedback-response analysis method (CFRAM), have been developed to quantify contributions from individual processes to radiative forcings and to temperature changes, but these methods are not applicable in non-equilibrium states where extremes lie. Moist static energy (MSE) budgets with energy unbalanced have been adopted to quantify contributions from physical processes in intra-seasonal oscillations. Furthermore, shortwave and longwave radiative forcings can be decomposed into partial perturbations through linearization as proposed by PRP and CFRAM. Changes in temperature during different phases of heatwaves are directly linked to recharge-discharge of internal energy, the dominant part of total atmospheric energy. This provides insight to dissect heatwaves from an energetics perspective, attributing total energy (MSE plus kinetic energy) change into multiple dynamical and radiative processes according to the total energy budget equation together with linearization concepts from PRP and CFRAM. In this study, we develop a comprehensive process-resolving, energetics-based attribution framework (PREAF) to quantify contributions from atmospheric dynamics (horizontal and vertical advection), surface latent and sensible heat fluxes, and radiative drivers (solar insolation, ozone, surface albedo, temperature, water vapor, cloud, and aerosols) to the lifecycle of the 2022 mega-heatwave over central-eastern China, and to diagnose model prediction biases.

Prior work shows heatwaves involve dynamical drivers (e.g., high-pressure systems, subsidence, advection) and radiative processes (clear-sky shortwave enhancement, longwave effects of temperature and water vapor). Studies specifically on summer 2022 East Asia linked the extreme heat to extra-tropical circulation anomalies, tropical SST forcing (including La Niña), and soil moisture–temperature feedbacks. Aerosol radiative effects modulate surface temperature, with increased aerosols generally reducing heatwaves via shortwave scattering/absorption; reductions in anthropogenic aerosols can have the opposite effect. Traditional attribution tools (PRP, CFRAM) effectively decompose radiative forcings and temperature responses under quasi-equilibrium but are limited for transient extremes; MSE budget approaches have been successful for intraseasonal oscillations. This study builds on these by combining total energy (MSE + kinetic) budgets with radiative decomposition via linearization to attribute the heatwave lifecycle.

Data: Observations included CPC daily maximum 2 m air temperature and precipitation (0.5°), GLEAM v3.7b daily evaporation and 0–10 cm soil moisture (0.25°), and MODIS Terra/Aqua Level-3 (1°) cloud fraction (cloud mask), precipitable water (IR), cloud liquid/ice water paths, and aerosol optical depth at 550 nm (Dark Target + Deep Blue). Period: 2003–2022. Reanalysis: NASA MERRA-2 (0.5°×0.625°, 42 levels) for meteorological and aerosol variables (temperatures, humidity, winds, geopotential, potential vorticity, surface fluxes, cloud properties, precipitation, soil moisture, precipitable water, AOD). CPC precipitation was used to correct MERRA-2 precipitation in low/mid-latitudes. Prediction model: NCEP CFSv2 (T126 atmosphere, 64 hybrid levels; MOM4 ocean ~0.5°, 40 levels; Noah LSM; interactive sea ice). A 45-day forecast initialized at 0600 UTC 26 July 2022 (member 1) was analyzed for process-level predictability assessment using PREAF. Heatwave and climatology definitions: A heatwave day is when daily max 2 m temperature exceeds the 90th percentile (2003–2022) for ≥3 consecutive days. Climatology is 2003–2022 daily means. Time series use 3-day means from 24 July–12 September (MERRA-2) and 27 July–6 September (CFSv2). The column top p_top is 150 hPa, the top of significant warming. PREAF framework: The total atmospheric energy per layer is E = c_p T + gz + L_v q + ½(u²+v²+w²). The vertically integrated atmospheric energy tendency difference between state B (event) and state A (climatology) is decomposed into contributions from horizontal advection, vertical advection, shortwave and longwave radiative heating, surface latent (LH) and sensible (SH) heat fluxes, and surface friction (negligible): ∂E/∂t = −∇_h·(V_h E) − ∂(ωE)/∂p + Δ(SW_ptop−SW_surface) + Δ(LW_ptop−LW_surface) + ΔLH + ΔSH − ΔFric. Land surface energy tendency is ∂E/∂t = ΔSW_surface + ΔLW_surface − ΔLH − ΔSH (surface heat storage). Radiative decomposition: Net SW and LW flux changes are linearized into partial contributions from solar insolation (SR), ozone (O3), surface albedo (AL), temperature (T), water vapor (WV), clouds (CLD), and aerosols (BC, OC, sulfate, sea salt, dust), plus residuals (ΔresSW, ΔresLW) accounting for offline calculation errors and unrepresented forcings (e.g., CO2, CH4, N2O). Aerosol effects represent direct radiative effects; indirect aerosol effects are included via cloud terms. Radiative transfer used RRTM v5 and MCICA with maximum-random cloud overlap; aerosol optical properties from MAM4. Daily mean inputs drove offline radiative calculations for states A and B. To reduce cloud-radiative biases from time-mean cloud properties, CLD effects were taken as all-sky minus clear-sky surface net radiative flux changes from MERRA-2 and CFSv2 directly. Due to limited temporal resolution (3-hourly MERRA-2, 6-hourly CFSv2), horizontal advection was computed as a residual of the atmospheric energy budget after other terms were evaluated. Lifecycle phases: Based on area-averaged anomalies over 105°–123°E, 25°–34°N, the event was separated into developing (30 Jul–13 Aug), mature (14–25 Aug), and decaying (26 Aug–6 Sep) phases. Column integrals and tendencies were computed from surface to 150 hPa for atmosphere, and for the land surface layer.

• The 2022 central-eastern China mega-heatwave (late July–late August) featured widespread daily max 2 m temperature anomalies of ~+2 °C and severe drought (SPI < −1.6), with warm anomalies extending through most of the troposphere.
• Energetics perspective: Total energy anomalies closely tracked temperature anomalies; internal energy dominated total energy changes, with latent energy playing a secondary but increasing role.
• Developing phase (30 Jul–13 Aug):
  • Land surface energy tendency was dominated by positive radiative forcing due to clear-sky shortwave enhancement from reduced low-level clouds; aerosol reductions (notably black carbon and sulfate) provided additional positive shortwave contributions; longwave surface cooling from temperature feedback partially offset SW gains. Surface sensible and latent heat fluxes exported energy upward (negative to land).
  • Atmosphere gained energy mainly from positive horizontal advection and upward surface heat fluxes; radiative processes produced a net negative atmospheric contribution due to longwave cooling from reduced cloud water and precipitable water and shortwave effects of reduced absorbing aerosols.
  • Quantitatively (W m−2): Land: SW +33.29, LW −4.64, RAD +28.65, LH −17.51, SH −8.79, E_tend +2.35. Atmosphere: net radiative −14.85 (13.85 at p_top with 28.65 to surface), H_adv +17.02, V_adv −1.83, E_tend +26.69.
  • Dynamical trigger: An anticyclonic eddy shedding from the Asian monsoon anticyclone intensified anticyclonic circulation, promoting subsidence (adiabatic heating) and westward extension of the WNP subtropical high.
• Mature phase (14–25 Aug):
  • Land radiative heating remained positive, dominated by cloud shortwave effects; increasing precipitable water strengthened the longwave greenhouse effect. With continued drying, latent heat flux export weakened while sensible heat flux export strengthened.
  • Atmosphere maintenance was supported by positive vertical advection (subsidence-related adiabatic heating) and continued import from surface heat fluxes, partially offset by negative radiative effects and by horizontal advection.
  • Quantitatively (W m−2): Land: RAD +18.26, LH −5.66, SH −10.86, E_tend +1.74. Atmosphere: V_adv +14.54, LH +5.66, SH +10.86, radiative export (8.53 at p_top vs −18.26 at surface), H_adv −9.83, E_tend +11.50.
• Decaying phase (26 Aug–6 Sep):
  • A transition to low-level cyclonic circulation and northerlies increased low-level cloudiness, terminating radiative surface heating. Persistent warm surface sustained upward sensible heat flux; precipitation recovery and reduced soil dryness led to downward latent heat flux (positive to land surface).
  • Atmospheric total energy decreased rapidly, primarily due to strong cold horizontal advection associated with northerlies; positive vertical advection partially compensated; net radiative effects on the atmosphere were weak as cloud, water vapor, temperature, and aerosol impacts largely offset.
  • Quantitatively (W m−2): Land: RAD +1.83, LH +5.73 (downward), SH −10.49 (upward), E_tend −2.93. Atmosphere: H_adv −54.44, V_adv +15.00, radiative import +2.65 (4.48 at p_top minus 1.83 at surface), surface flux import +4.76, E_tend −32.03.
• Aerosol role: A persistent reduction in anthropogenic aerosols increased surface shortwave heating throughout, implying that pollution mitigation can inadvertently amplify heatwave intensity by reducing aerosol cooling.
• Sub-seasonal prediction diagnosis (CFSv2): The model reproduced the developing phase energetics and circulation well but predicted an earlier decay (15–25 Aug) driven by erroneous upper-level cyclonic circulation and associated vertical advection (adiabatic cooling), plus misrepresented land–atmosphere coupling (soil moisture increased with precipitation leading to opposite-signed latent heat flux and reduced sensible heat flux). Water vapor radiative effects were misrepresented during decay. The aerosol radiative effect was not assessed due to the absence of an aerosol module in CFSv2.

The study addressed the key question of the relative roles of dynamical versus radiative processes across heatwave phases by quantifying energy tendency contributions in a coupled land–atmosphere column. Results show dynamics (eddy shedding-triggered anticyclone, subsequent northerly cold advection) initiate and terminate the event, while radiative clear-sky shortwave heating and land–atmosphere coupling (surface heat fluxes under warm/dry land) sustain and amplify it. The PREAF framework provides a transparent, energy-based diagnostic that disentangles multiple interacting processes and yields phase-dependent attribution. The persistent positive contribution from reduced anthropogenic aerosols underscores an important policy implication: emission controls that decrease aerosol loading can increase the amplitude of heatwaves absent concurrent greenhouse gas mitigation or adaptation measures. The framework also effectively diagnoses sub-seasonal prediction errors by comparing process-level energy tendencies, revealing that misrepresented dynamics and hydrology (vertical advection, soil moisture–flux responses) drove premature decay in CFSv2, while cloud radiative effects were relatively well captured under clear-sky conditions. These insights are relevant for improving prediction systems and for assessing how mitigation policies may affect extremes via radiative pathways.

This work develops and applies a process-resolving, energetics-based attribution framework (PREAF) to quantitatively delineate the lifecycle of the 2022 central-eastern China mega-heatwave. It demonstrates that clear-sky radiative heating (dominated by cloud shortwave effects) and atmospheric dynamics drive rapid energy buildup; land–atmosphere coupling sustains the heatwave; and cold horizontal advection terminates it. Reduced anthropogenic aerosols provided a persistent positive contribution, implying potential amplification of future heatwaves under continued aerosol emission reductions. PREAF also identifies sources of sub-seasonal prediction errors at a process level. Future directions include applying PREAF across multi-scale phenomena (other heat/cold extremes, MJO/BSISO, monsoons, ENSO) and incorporating explicit contributions from solar variability and trace gases (e.g., ozone, CO2, CH4) to quantify additive anthropogenic forcings. Broader multi-event, multi-model assessments can generalize findings and guide improvements to land–atmosphere coupling and dynamical representations in prediction models.

• Observational and reanalysis uncertainties limit attribution accuracy, with especially large challenges for cloud and aerosol properties despite satellite-era improvements.
• The CFSv2 lacks an aerosol module, precluding assessment of aerosol radiative effects in the prediction diagnosis.
• This is a preliminary single-event case study assessing one forecast member; broader analyses across additional mega-heatwaves and models are needed to test generality.