Eurasian mid-latitude jet stream bridges an Atlantic to Asia summer teleconnection in heat extremes

In this study, we use the daily maximum surface air temperature, sea surface temperature (SST) and zonal wind, meridional wind, vertical velocity, air temperature, geopotential height on pressure levels from the fifth generation of atmospheric reanalysis in European Centre for Medium-Range Weather Forecasts (ERA5 reanalysis) (Hersbach et al 2020). The data analyzed are on the $1^\circ\times1^\circ$ longitude-latitude grids for the period of June 1979–2023.

The early-summer Northeast Asian heat extremes are identified if the daily maximum surface air temperature averaged over Northeast Asia (110^∘ E–130^∘ E, 35^∘ N–50^∘ N) is above the 95th percentile of historical daily value during June 1979–2023. The results are robust if daily mean/minimum temperature is used. 69 heat extreme days have been detected for the total 45 yr. The thermal budget and circulation characteristics associated with heat extremes are investigated through lagged composites of these fields. Daily anomaly data of all fields are used throughout this paper by removing the daily climatology and long-term trend. All significance testing of composite analysis is conducted with a two-tailed t test.

To help understand the local temperature evolution associated with the heat extremes, the thermal budget analysis at 850 hPa is employed as in Nie et al (2022), which we present below.

$$\begin{eqnarray} &&\frac{\partial T}{\partial t} = -\textbf{V}\cdot\nabla T+\omega\left(\kappa\frac{T}{p}-\frac{\partial T}{\partial p}\right)+\frac{Q}{c_p}, \end{eqnarray} \tag{ 1 }$$

where T is the temperature, V is the horizontal wind vector, ω is the vertical velocity in pressure coordinates, c_p is the specific heat capacity of air at constant pressure, R_d is the gas constant for dry air, $\kappa = R_d/c_p$, and Q is the diabatic heating. By equation (1), the local temperature change is determined by the horizontal advection of temperature, changes in temperature due to adiabatic expansion or compression due to vertical motion, and diabatic processes. Note that the diabatic heating term is estimated as a residual of equation (1).

To examine the wave propagation associated with the Northeast Asian heat extremes, the horizontal component of the wave activity flux derived by Takaya and Nakamura (2001), hereafter TN01 is diagnosed.

$$\begin{align} W = \frac{p\mathrm{cos}\phi}{2\left|\bar{V}\right|} \left\{ \begin{array}{ll} \frac{U}{a^2\mathrm{cos}^2\phi}\left[\left(\frac{\partial\Psi^{^{\prime}}}{\partial\lambda}\right)^2-\Psi^{^{\prime}}\frac{\partial^2\Psi^{^{\prime}}}{\partial\lambda^2}\right] +\frac{V}{a^2\mathrm{cos}\phi}\left[\frac{\partial\Psi^{^{\prime}}}{\partial\lambda}\frac{\partial\Psi^{^{\prime}}}{\partial\phi}-\Psi^{^{\prime}}\frac{\partial^2\Psi^{^{\prime}}}{\partial\lambda\partial\phi}\right]\\ \frac{U}{a^2\mathrm{cos}^2\phi}\left[\frac{\partial\Psi^{^{\prime}}}{\partial\lambda}\frac{\partial\Psi^{^{\prime}}}{\partial\phi}-\Psi^{^{\prime}}\frac{\partial^2\Psi^{^{\prime}}}{\partial\lambda\partial\phi}\right] +\frac{V}{a^2}\left[\left(\frac{\partial\Psi^{^{\prime}}}{\partial\phi}\right)^2-\Psi^{^{\prime}}\frac{\partial^2\Psi^{^{\prime}}}{\partial\phi^2}\right]\\ \end{array}\right\}, \end{align} \tag{ 2 }$$

where $\Psi^{^{\prime}}$ is the deviation of the stream function from the daily climatology, U and V are the climatological background flow. The wave activity flux measures the eddy propagation in accordance with the background flow, with its direction parallel to the local group velocity of Rossby waves. Through this diagnostic, the source and sink of wave packets and the wave energy propagation relative to/in accordance with the mean flow associated with Northeast Asian heat extremes can be explicitly identified.

To examine the roles of jet stream in modulating the Northeast Asian heat extremes, we perform two sets of idealized experiments using the Geophysical Fluid Dynamics Laboratory (GFDL) spectral dynamical core. The model is run at T42 resolution with 40 vertical sigma levels and a time step of 800 s. The model is forced by the Newtonian relaxation to zonally asymmetric equilibrium temperature T_eq and damped by the Rayleigh friction in the planetary boundary layer. The model also uses realistic topography that allows for the excitation of a rather realistic planetary-scale stationary wave pattern. Model parameters are identical to those used in Wu and Reichler (2018), except where noted.

The first family of experiments uses the model default configuration. We optimize T_eq to let the simulated temperature climatology converging toward that of the reanalysis in early summer (June) and late summer (August) respectively. These experiments are referred to early-summer control experiment and late-summer control experiment. In the second set of experiments, a thermal forcing experiment (heating) is performed. We add a simple gaussian-like thermal forcing in the subpolar North Atlantic region centering at (20^∘ W, 55^∘ N) as follows:

$$\begin{align} F & = F_0\text{exp}\left\{-\frac{1}{2}\left[\left(\frac{\lambda+20^\circ}{20^\circ}\right)^2+\left(\frac{\phi-55^\circ}{4^\circ}\right)^2\right.\right.\nonumber\\ &\quad\left.\left.+\,\left(\frac{p-p_s}{200\,hPa}\right)^2\right]\right\}, \end{align} \tag{ 3 }$$

where F₀ is the amplitude of heating (2 K day⁻¹), λ and φ denote the longitude and latitude. For all experiments, the model is integrated under perpetual conditions for 2000 days with the first 500 days of spinup discarded.

A series of record-breaking heatwaves hit Northern China during June 2023. Figure 1(a) shows the daily evolution of surface maximum air temperature in Beijing from June 1 to June 30, 2023. There were three typical heatwaves with air temperature exceeding 35 ^∘C. The longest and strongest heatwaves were observed from June 21 to June 30, with the maximum temperature even exceeding 40 ^∘C, which broke the historical record since 1961. The regional-averaged temperature anomalies over Northeast Asia (110–130^∘ E, 35–50^∘ N) exceeded 3 ^∘C from June 21 to June 27 (figure 1(b)). During the same period, the UK experienced the hottest June on record (Met Office 2023). The Northern European surface maximum air temperature during June 2023 break the historical record as shown in figure 1(c). The daily maximum air temperature anomaly exceeded 3 ^∘C from June 10 to June 25 (figure 1(d)).

Zoom In Zoom Out Reset image size

Figures 1(e) and (f) further display the geographical distribution of surface air temperature anomalies and circulation characteristics during 21–30 June 2023. It is clear that the surface air temperature shows strong warm anomalies over Northern Europe, east Siberia and Northeast Asia. Such clustering warm anomalies in different regions may be related to planetary Rossby wave packets (Ding and Wang 2005). The upper-tropospheric geopotential height displays positive anomalies over Northern Europe, negative anomalies over Ural region and positive anomalies over east Siberia and Northeast Asia, which matches well with the surface air temperature anomalies. Such alternatively positive and negative geopotential height anomalies are associated with Rossby wave packets originating from the subpolar North Atlantic (see the arrows in figure 1(e)), indicating an important role of Rossby wave packets in forming the Northeast Asian heat extremes. As suggested by previous studies (Nie et al 2022), the early summer Rossby wave packets are often organized by the jet waveguide. As shown in figure 1(f), the zonal wind displays a stronger Eurasian jet during the heatwave period, which provides a favorable guide for the wave packets.

We next systematically explore the formation mechanisms responsible for the historical heat extremes over Northeast Asia during early summer by diagnosing the thermal budget and examining the evolution of circulation anomalies. Figure 2 shows the 3 d averaged composite maps of each term in the 850 hPa thermal budget of equation (1) prior to the peak dates of Northeast Asian heat extremes. As shown in figure 2(a), the temperature tendency exhibits strong positive values over Northeast Asia (black box), indicating a strong warming tendency. Decomposition of temperature tendency shown in figures 2(b)–(d) illustrates that horizontal temperature advection plays an important role in the western region of Northeast Asia. The horizontal advection of temperature is mainly determined by the anomalous flow advection on mean temperature (figure S1). Adiabatic term makes a major contribution to the warming over the eastern region of Northeast Asia. Diabatic heating acts to warm the central Northeast Asia. This indicates that anticyclonic circulation anomalies above the warm anomalies dominate the heat extreme formation over Northeast Asia. On the one hand, the anticyclonic wind anomaly acts to blow the warm air from the upstream to Northeast Asia. On the other hand, maintenance of anticyclonic flow anomalies leads to subsidence, which acts to exaggerate the surface warm anomaly through adiabatic process (figure S2). The diabatic heating exerts a direct heating effect to the region mainly through clear-sky conditions and increased incoming solar radiation.

Zoom In Zoom Out Reset image size

Since the anticyclonic anomaly over Northeast Asia is crucial for the formation of heat extremes, we next explore its formation processes. Figures 3(a)–(c) show the lagged composites of anomalous surface air temperature, 250 hPa geopotential height and wave activity flux associated with the evolution of Northeast Asian heat extremes. Prior to the heat extremes by 4 to 6 days, the 250 hPa geopotential height displays alternating positive and negative anomalies from subpolar North Atlantic to Northeast Asia (figure 3(a)). It is characterized with an anticyclonic anomaly in the subpolar North Atlantic, a cyclonic anomaly in the Ural region and an anticyclonic anomaly in Northeast Asia. This well corresponds to the warm-cold-warm anomalies of surface air temperature over the same regions. The anomalous wave activity flux shows Rossby wave packets originating from the subpolar North Atlantic and propagating downstream toward the Northeast Asia with a curved pathway, suggesting an important role of Rossby wave train in the initial formation of Northeast Asian heat extremes. Then at lags −3 to −1 days, the Z250 exhibits much stronger positive anomalies in Northeast Asia (figure 3(b)), suggesting a rapid growth of the preexisting anticyclonic high. The anomalous wave activity flux displays a relatively weaker horizontal wave propagation in the Eurasian continent. Following the peak date of heat extremes by 0–2 d (figure 3(c)), the anticyclonic anomalies in the Northeast Asia are maintained and stayed.

Zoom In Zoom Out Reset image size

The above thermal budget and wave activity flux diagnostics suggest that the Northeast Asian anticyclonic anomaly, which dominantly drives the Northeast Asian heat extremes, is primarily initialized and maintained by midlatitude Rossby wave packets over the Eurasian continent. According to the classical stationary wave theory, the Rossby wave train is often organized by the jet waveguide (Hoskins and Karoly 1981, Held et al 2002). The right column of figure 3 shows the lagged composite maps of 250 hPa zonal wind and its anomaly associated with the evolution of Northeast Asian heat extremes. It is evident that preceding the heat extremes by 4–6 days, the Eurasian jet stream is relatively stronger and wave packets propagate along the jet stream. Meanwhile, the zonal wind over Northeast Asia exhibits a north-south dipolar structure in the whole life cycle of heat extremes, with westerly anomaly in the north and easterly anomaly in the south. Such dipolar zonal wind anomalies are in favor of the maintenance for the anticyclonic anomaly.

To quantify the contribution of Eurasian Rossby wave packets to the formation of Northeast Asian heat extremes, we also count the occurrence of wave packets prior to the heat extremes by 1–3 days. Figure 4(a) shows the scatter plot between the geopotential height anomalies averaged over Northeast Asia and those over subpolar North Atlantic and Ural region preceding the heat extremes. 97% of the Northeast Asian heat extremes are associated with the Northeast Asian anticyclonic anomalies. 60% of anticyclonic-type Northeast Asian heat extreme formation are preceded by positive height anomalies over subpolar North Atlantic and negative height anomalies over Ural region. The correlation between geopotential height anomalies over Northeast Asia and those in subpolar North Atlantic and Ural region are 0.28 and −0.61 respectively, which passes the 95% significance level. Furthermore, we define the occurrence of wave packet as the days when the geopotential height in these three key regions (the black boxes in figure 3(b)) are greater than 0.5 times daily standard deviation respectively. 36% of anticyclonic-type Northeast Asian heat extreme formation are preceded by the Eurasian midlatitude Rossby wave trains. The heatwaves in early summer 2023 were extreme events with unusually stronger wave-packet influence (denoted by the bigger circles).

Zoom In Zoom Out Reset image size

The importance of Eurasian jet stream in affecting the Northeast Asian heat extremes is further examined by calculating the lagged correlation between the Eurasian midlatitude jet speed anomalies averaged over (50–120^∘ E, 45–60^∘ N) and Northeast Asian surface air temperature anomalies averaged over ((110-130^∘ E, 35–50^∘ N)). As shown in figure 4(b), the daily correlation between the two fields reaches 0.31 when the jet leads SAT by 1 d, confirming the strong modulating effect by jet stream.

In addition to the modulation effect by jet stream, the formation of Rossby wave packets is triggered by a wave source (strong anticyclonic anomaly) in the subpolar North Atlantic. As suggested by previous studies (Nie et al 2019), the anticyclonic anomaly over subpolar North Atlantic may result from an active two-way coupled ocean-atmosphere interaction (e.g. North Atlantic sub-polar gyre). Figure 4(c) shows the composite maps of detrended SST anomalies preceding the heat extremes by 10 d. The SST shows a significant warm anomalies centered at 20^∘ W, 55^∘ N, well matching the strong anticyclonic anomaly over subpolar North Atlantic. On the one hand, warm subpolar North Atlantic SSTs can be driven by the high pressure via a reduction in clouds and an increase in downwelling shortwave radiation. On the other hand, warm SST anomalies in subpolar North Atlantic could strengthen the meridional temperature gradient to the north of 55^∘N and decrease the temperature gradient to the south of 55^∘N, which may help to form a anticyclonic anomaly. Such two-way interaction between warm SST and atmospheric anticyclonic anomaly may provide a persistent wave sources for the Eurasian wave packets. Note that the North Atlantic ocean experienced record-breaking high temperatures during 2023 spring, and strong marine heatwaves in early summer (Madeleine Cuff 2023). This led to a maintenance of strong wave sources, which may partially explain the strongest Eurasian wave packets during early summer 2023.

The above observational diagnostic results suggest that the Northeast Asian heat extremes are formed through midlatitude Rossby wave packets that originate from subpolar North Atlantic and propagate along the Eurasian jet. In this subsection, we carry out numerical experiments to understand the wave packet response to a heating forcing in the subpolar North Atlantic and try to explain why the wave packets are significant only in early summer. As suggested by previous studies, changes to the jet stream waveguide structure for planetary waves are hypothesized to modulate the frequency of quasi-resonance amplification of planetary waves (Kornhuber et al 2017, Mann et al 2017). Figures 5(a) and (b) show the horizontal and vertical distribution of heating forcing in the heating experiments. We add a gaussian-shaped shallow heating centered at 20^∘ W, 55^∘ N, which is roughly consistent with the observed SST anomalies shown in figure 4(c). The vertical profile of heat source is determined based on the observed vertical structure of temperature anomalies above the warm SST anomalies, where the temperature anomalies are largest in the boundary layer (figure not shown). For simplicity, we maximize the heat source at the surface.

Zoom In Zoom Out Reset image size

Figures 5(c) and (d) show the 275 hPa zonal wind distributions in early-summer and late-summer control experiments. In the early-summer experiment, the zonal wind displays a split-jet structure in the Eurasian continent. The Northeast Asia is located to the east of the midlatitude jet exit. The weak zonal wind here is in favor of energy accumulation (Nie et al 2022). The convergent wind in the jet exit region also facilitates the conversion of kinetic energy from mean flow to anomalies, thereby fostering the development of stationary waves (Lau and Lau 1992, Wang et al 2023). In the late-summer control experiment, the zonal wind shows a single subtropical jet structure in the Eurasian continent and the jet center moves westward. figures 5(e) and (f) display the corresponding quasi-stationary response of 275 hPa geopotential height anomalies and wave activity flux. Here the quasi-stationary component of geopotential height anomalies is calculated as the departures from the zonal-mean component. In the early-summer heating experiment (figure 5(e)), the geopotential height shows strong positive anomalies in the subpolar North Atlantic, weak negative anomalies in the Ural region and weak positive anomalies in the central and East Asia. The wave activity flux shows evident wave packets from subpolar North Atlantic to Asia. This is consistent with the observed circulation pattern associated with historical Northeast Asian heat extremes shown in figure 3(a). In the late-summer heating experiment (figure 5(f)), the geopotential height anomalies show much weaker amplitudes and the wave activity flux anomalies almost disappear. These results suggest that the Eurasian jet structure plays a key role in modulating the wave packet.