Relation of Mid-High-Latitude Eurasian ISO to Ural Blocking Frequency and Their Co-Effect on Extreme Hot Events during Boreal Summer

Abstract

Based on NCEP reanalysis daily data during 1979–2018, the spatiotemporal evolution of the 10–30-day atmospheric intraseasonal oscillations (ISO) at mid-high-latitude Eurasia and its effect on the Ural blocking frequency are investigated. The co-effect of the blocking and ISO on extreme hot event frequency is also investigated. The ISO exhibits two modes of eastward and westward propagation. During the eastward (westward) propagating mode, the northwest–southeast tilted quadrupole (east–west dipole) quasi-barotropic geopotential height anomaly coupled with the air temperature anomaly at the troposphere propagates southeastward (westward). The phase composite shows that, during both modes, the mid-high-latitude low-frequency Rossby wave trains significantly affect the frequency of the European blocking during the propagating journey. The most frequent European blocking appears in phase 2 during both the eastward- and westward- propagating mode. Compared with the situation without the Ural blocking, the blocking activity results in the positive geopotential height anomalies throughout Europe and north of 60° N in the Ural Mountains and the negative geopotential height anomalies in the south of 60° N in the Ural Mountains and north of the Japan Sea. The occurrence of Ural blocking is conducive to the occurrence of extreme high-temperature events in Europe and the high latitudes of the Ural Mountains, and a reduced frequency of extreme high-temperature events in the mid-latitudes of the Ural Mountains and north of the Japan Sea. Therefore, the Ural blocking activities significantly regulate the effect of the two propagating ISO modes on the extreme hot events over the middle and high latitudes of Eurasia.

1. Introduction

The atmospheric intraseasonal oscillation (ISO) is an important component of the atmospheric motion, characterized by a periodicity of 10–90 days [1]. It is well known that the tropical ISO dominates a 30–60-day periodicity, which is referred to as the Madden-Julian oscillation (MJO) [2]. Actually, ISO signal exists not only over the tropics, but also over the mid-high-latitude region [3].

The mid-high-latitude ISO is regionally diverse in periodicity and propagation. During boreal winter, the mid-high-latitude surface air temperature shows a southeastward propagation with a 10–30-day time scale [4]. During the boreal summer, the 250-hPa geopotential height anomaly with a periodicity of 30–50 days exhibits both an eastward and a westward propagation [5,6,7], and also exhibits a northwestward propagating feature with a periodicity of 10–30 days [8,9]. There is a close connection between mid-high-latitude ISO and the extreme events. The ISO-related wave train can generate a large-scale extreme precipitation anomaly and heat waves [10,11,12]. The interaction between extratropical and tropical ISOs can lead to a summer rainfall anomaly over Southeastern China [13].

Furthermore, the ISO can significantly affect the blocking activity. The atmospheric blocking, a large-scale circulation pattern in the mid-high-latitude westerly belt, is a result of a meridional-type flow interrupting the normal zonal flow. Blockings can result in extreme weather events [14,15]. For example, the blockings coupled with polar outbreak may induce regional extreme cold events in winter [16,17]. In summer, the heat wave in Russia is closely linked to the European blocking [18]. The five extreme hot event types over northeast China are associated with sustainable anticyclones and the upstream blocking acts as an energy source for downstream extreme hot events anticyclone intensification and leads to extreme hot events formation [19,20]. According to the vertical structure of blocking, the adiabatic warming due to subsidence is the main driver of the positive temperature anomaly, the blockings can lead to regional extreme heat waves and extreme precipitation, and then causes economic losses [21,22].

The teleconnection wave train excited by MJO can regulate the Pacific blocking frequency by changing the geopotential height anomaly [23]. The westward-propagating mid-high-latitude ISO can modulate the blocking activity over the Pacific sector in winter time [24]. The Okhotsk blocking activity is affected by the westward-propagating ISO over mid-high latitude Eurasia during summertime [25]. However, the Ural blocking is also active during summertime. Is the Ural blocking related to the mid-high-latitude ISO during summertime? Does the mid-high-latitude ISO with 10–30-day cycle exist the eastward propagating and westward propagating features in summertime? To answer these questions, the spatiotemporal evolution of the ISO over mid-high-latitude Eurasia and its effect on the Ural blocking frequency are investigated in this study. In addition, the co-effect of Ural blocking and mid-high-latitude ISO on the extreme hot event frequency is also discussed.

The remainder text is structured as follows. Section 2 describes the datasets and methods. The propagation features of mid-high-latitude-Eurasian ISO is revealed in Section 3. The effect of mid-high-latitude ISO on the Ural blocking frequency is studied in Section 4. The co-effect of mid-high-latitude ISO and Ural blocking on extreme hot event frequency is investigated in Section 5. Section 6 is the summary.

2. Materials and Methods

The daily reanalysis data with a horizontal resolution of 2.5° × 2.5° during 1979–2018 provided by National Centers for Environment Prediction-National Center for Atmospheric Research (NECP-NCAR) [26] is used. The variables include geopotential height (Z) and air temperature (T) at 10 pressure levels (1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, and 100 hPa) and 250-hPa horizontal wind (u and v).

A two−dimensional (2D) blocking index, referring to Scherrer et al. [27] and Davini et al. [14], is calculated by the reversal of the meridional gradient of 500-hPa Z. This 2D index is developed from the a 1D blocking index provided by Tibaldi and Molteni [28]. The spatial distribution of the blocking can be better described by this 2D index which can be written as:

$$\mathrm{GHGS}\left(\mathsf{\theta }, \mathsf{\Phi }\right)=\frac{\mathrm{Z}\left(\mathsf{\theta }, \mathsf{\Phi }\right)-\mathrm{Z}\left(\mathsf{\theta },{ \mathsf{\Phi }}_{\mathrm{S}}\right)}{\mathsf{\Phi }-{\mathsf{\Phi }}_{\mathrm{S}}}$$

(1)

$$\mathrm{GHGN}\left(\mathsf{\theta }, \mathsf{\Phi }\right)=\frac{\mathrm{Z}\left(\mathsf{\theta },{ \mathsf{\Phi }}_{\mathrm{N}}\right)-\mathrm{Z}\left(\mathsf{\theta }, \mathsf{\Phi }\right)}{{\mathsf{\Phi }}_{\mathrm{N}}-\mathsf{\Phi }}$$

(2)

where Z is 500-hPa geopotential height; θ is the longitude of each grid point ranged from 0° to 360°, $\mathsf{\Phi }$ is the latitude of each grid point ranged from 30° N to 75° N, and ${\mathsf{\Phi }}_{\mathrm{S}}=\mathsf{\Phi }-15°$, and ${\mathsf{\Phi }}_{\mathrm{N}}=\mathsf{\Phi }+15°.$ Once both $\mathrm{GHGS} > 0$ and $\mathrm{GHGN} <-10 \mathrm{m}/°\mathrm{lat}$ are fulfilled in a given grid, an instantaneous blocking (IB) occurs in this grid. A large-scale blocking (LSB) is defined as when the IB extends for at least 15 continuous longitudes. If a LSB lasts for 5 days within a box area of 10°lon × 5°lat centered in a given grid, a LSB event occurs in this grid. The blocking frequency is defined as the ratio of the total days of LSB events to the total summer days.

In this study, the extreme events are defined according to Alexander et al. [29] and Hsu et al. [30]. For each grid point, the 850-hPa daily mean T of a certain day in each year from 1979 to 2018 are extracted, and then they are ranked from low to high. The temperature at the 95th percentile is selected as the threshold of an extreme hot event on this certain day for each grid point. Therefore, we obtain 365 thresholds per grid point. For a given grid on a given day, if 850-hPa daily mean T exceeds the corresponding threshold, this given day is defined as an extreme hot day grid, which is recorded as 1; otherwise, it is 0. Therefore, it is possible to retrieve when an extreme hot event occurred at each grid point. The frequency of extreme hot events at each grid point is the ratio of the total days of the extreme hot events to the total number of summer days from 1979 to 2018 at that grid point. The summertime is defined as the June–August. For each grid point, the sum of the extreme hot days in each ISO phase during the 40-y analysis period is counted, and then, the frequency of extreme days in each phase is calculated to reveal how the extreme hot events depend on the ISO phase.

To abstract the intraseasonal signals, the seasonal cycle and the first three harmonics are removed, and the so-derived data are filtered by Lanczos bandpass filter [31]. In this study, summer is defined as the period from May to September. Referring to Matthews [32], a two-dimensional vector field Z(t) is defined to perform phase analysis for each propagating ISO mode. Taking the westward-propagating mode as an example, Z(t) is written as:

$$\mathrm{Z}\left(\mathrm{t}\right)=\left[\mathrm{PC}1\left(\mathrm{t}\right), \mathrm{PC}2\left(2\right)\right]$$

(3)

$$\mathrm{A}\left(\mathrm{t}\right)={\left[\mathrm{PC}1{\left(\mathrm{t}\right)}^2+\mathrm{PC}2{\left(\mathrm{t}\right)}^2\right]}^{1/2}$$

(4)

$$\mathsf{\alpha }\left(\mathrm{t}\right)=\arctan \left[\mathrm{PC}2\left(\mathrm{t}\right)/\mathrm{PC}1\left(\mathrm{t}\right)\right]$$

(5)

where t denotes the time point, and PC1 and PC2 are the principal components of the first two modes (MVEEOF1 and MVEEOF2) by the multivariable, extended empirical orthogonal function (MVEEOF) analysis of the 250-hPa Z and horizontal wind fields; the geographical domain we use to calculate MVEEOF is 40° N–80° N, 40° E–150° E. A(t) is the amplitude, and α(t) is the phase angle. After transforming α to the range of [0, 2π] on each time point, all the time points could be categorized into eight phases, with the interval between two adjacent phases as π/4. Similarly, the eastward-propagating mode is also divided into eight phases by the principal components (PC3 and PC4) for the third and fourth MVEEOF modes (MVEOF3 and MVEEOF4). The strong ISO events with amplitude greater than 0.8 STD are selected for the phase composite during the eastward and westward propagating modes. The variant field has significant autocorrelation and cross-correlation features on the time dimension, and the method of MVEEOF is based on EOF but shows the propagating features of the variant field [33]. The MVEEOF analysis for the current study with three leading times (0, 3, and 6 days) was employed.

3. Evolution Features of the ISO

Figure 1a shows the horizontal distribution of the standard deviation (STD) of the 250-hPa Z over the mid-high-latitude Eurasia during the boreal summer from 1979 to 2018. As shown, there is a maximum variability center (42.5°–67.5° E, 57.5°–65° N; boxed area in Figure 1a) in the Ural Mountains. Figure 1b depicts the power spectrum of the area mean 250-hPa Z anomaly (with annual cycle and the first three harmonics removed) in the maximum variability center. As shown, there exist two significant cycles (i.e., 10–30 and 30–50 days) for the circulation over the Ural Mountains during the boreal summer, which is consistent with Zhu and Yang [6]. While Zhu and Yang [6] focused on the characteristics of the 30–50-day ISO, and the characteristic of the 10–30-day ISO is not discussed in detail. Although the 10–30-day ISO over the mid-high-latitude Eurasia has been mentioned by previous studies, they either focused on the wintertime ISO [4] or on a single mode of the ISO propagation [5,9]. The previous studies indicate that there may be two propagating ISO modes over mid-high latitudes. The propagating features may differ over different regions [34]. Therefore, this study intends to reveal the different propagating modes of the 10–30-day ISO over the mid-high-latitude Eurasia during the boreal summer and to investigate their impact on the Ural blocking activities.

The MVEEOF method is used to extract the dominant modes of the 10–30-day signal over the mid-high-latitude Eurasia (40°–150° E, 40°–80° N). Figure 2a–d shows the spatial distributions of the first four leading modes. MVEEOF1 shows a dipole pattern (western strong-eastern weak) on day 0. MVEEOF2 exhibits a monopole pattern on day 0; Both MVEEOF3 and MVEEOF4 show a northwest–southeast-orientated wave pattern with alternated positive and negative anomalies on day 0. The variance contributions of the first four modes are 10.2%, 10.1%, 7.9%, and 7.6%, respectively. It is seen that both MVEEOF1 (Figure 2a) and MVEEOF2 (Figure 2b) exhibit a westward propagation, while both MVEEOF3 (Figure 2c) and MVEEOF4 (Figure 2d) present an eastward propagation. Figure 2e shows the variance contributions of the first four modes and their error ranges. As seen, MVEEOF1 and MVEEOF2 (MVEEOF3 and MVEEOF4) are inseparable and statistically independent from higher modes. In addition, the most significant positive/negative correlation can be found when PC1 (PC3) leads/lags PC2 (PC4) by 4 days (Figure 2f). The above analysis indicates that MVEEOF1 and MVEEOF2 (MVEEOF3 and MVEEOF4), which are corresponding to different temporal distributions of the same spatial mode, actually reflect a westward- (an eastward-) propagating mode.

Based on PC1 and PC2, the westward-propagating mode have been divided into eight phases, and strong ISO events whose amplitude is greater than 0.8 STD are selected for a phase composite. Figure 3 shows the evolutions of the 10–30-day-filtered 850-hPa T, 250-hPa Z and wind anomalies with ISO phases during the westward-propagating mode. The 250-hPa Z and the 850-hPa T anomalies are positive over the Ural Mountains in Phase 1 (Figure 3a), and these anomalies intensify and move westward to the East European Pla in in Phase 2 (Figure 3b), which keep propagating in Phase 3 (Figure 3c). Meanwhile, in Phase 3, an upper-level negative Z anomaly appears over the high latitudes to the north of Japan, accompanied by a lower-level negative T anomaly. In Phase 4 (Figure 3d), the positive Z anomaly over the East European Plain weakens and propagates westward to the Western European region, and the high-latitude negative Z anomaly coupled with lower-layer T anomaly strengthens. Due to the westward propagation of the ISO, the anomalies in Phases 5–8 (Figure 3e–h) are opposite to those in Phases 1–4, respectively.

Similarly, based on PC3 and PC4, strong ISO events with amplitude greater than 0.8 STD also are selected for the phase composite during eastward-propagating mode. In Phase 1 (Figure 3i), there is a northwest–southeast-tilted wave train with alternated negative and positive Z anomaly over the mid-high-latitude Eurasia. The East European Plain is controlled by a weak negative Z anomaly coupled with a weak negative lower-level T anomaly. Meanwhile, a strong positive (negative) Z anomaly coupled with a strong positive (negative) lower-level T anomaly is dominated over the Ural Mountains (east of the Ural Mountain). In Phase 2–4 (Figure 3j–l), the negative Z and T anomalies, originated from the East European Plain, gradually strengthen and move eastward to the Ural Mountains. The positive (negative) Z and T anomalies over the Ural Mountains (east of the Ural Mountains) gradually weaken and move eastward to the Central Siberian (Northeast China). Since the ISO circulation propagates eastward, the anomaly patterns in Phases 5–8 (Figure 3m–p) are opposite to those in Phases 1–4, respectively.

To analyze the vertical structure, the longitude height profiles of the 10–30-day Z and T anomalies averaged along 40°–80° N in each phase during the eastward- and westward- propagating modes are shown in Figure 4. It shows that both of the two modes present a quasi-barotropic structure. The Z anomaly center is located at 250 hPa. A positive (negative) Z anomaly is corresponding to a positive (negative) T anomaly in the lower troposphere and to a negative (positive) T anomaly in the upper troposphere, satisfying the hydrostatic equilibrium relationship. The Z anomaly, coupled with the T anomaly, propagates westward (Figure 4a–h) and eastward (Figure 4i–p) throughout the troposphere.

4. Relationships between the ISO and Ural Blocking Frequency

The above analysis suggests that the summer ISO over mid-high-latitude Eurasia exists two modes—that is, the eastward- and westward-propagating modes. Considering that both of the two modes may impact the downstream circulation during the propagating journey, the influence of these two propagating modes on the summer Eurasian blocking frequency is discussed in this section. To obtain the most active summer blocking areas in Eurasia, the horizontal distribution of summer blocking frequency is given in Figure 5a. As seen, the most active blocking area is the Ural Mountains, with a maximum frequency of over 10%. The frequency center area of 55°–90° E, 65°–77.5° N (boxed area in Figure 5a) is selected as a key area of the Ural blocking. Figure 5b shows the summertime blocking frequency averaged along 55°–75° N as the function of the longitude. It shows that a remarkable peak can be found in the Ural Mountains within 50°–90° E, which is consistent with the previous studies [28] by a one-dimensional blocking index.

Figure 6 shows the evolution of the blocking frequency anomaly (shading) and the 10–30-day-filtered 500-hPa Z anomaly (contour) with the phases in westward-propagating mode over the mid-high-latitude Eurasia. Here, the frequency anomaly is the difference between the blocking frequency in each phase and the climatological mean of blocking frequency shown in Figure 5a. The solid black box denotes the key area of Ural blocking. In Phase 1 (Figure 6a), the Z anomaly presents positive over the mid-high-latitude Eurasia, with the center located on the east of the key area. This situation facilitates the blocking activity in the key area, and thus the key area is dominated a positive frequency anomaly. In Phase 2 (Figure 6b), the positive Z anomaly in the Ural Mountains intensifies and moves westward with the anomaly center propagating into the key area, so the Z gradient in the key area intensifies and shows a strong positive frequency anomaly. In Phase 3 (Figure 6c), the positive Z anomaly in the Ural Mountains propagates westward. However, the positive Z anomaly still controls the key area, leading to a positive frequency anomaly in situ. Meanwhile, a negative Z anomaly appears to the east of the Ural Mountains, which intensifies and moves westward in Phase 4 and has an enhanced impact on the key area (Figure 6d). The positive Z anomaly center moves to the west of the Ural Mountains, with little impact on the key area. The eastern part of the key area shows a negative blocking frequency anomaly. In Phase 5 (Figure 6e), due to the westward propagation of the ISO circulation, the entire Ural Mountains are controlled by a negative Z anomaly, with the center located on the east of the key area. The key area shows a negative frequency anomaly. In Phase 6 (Figure 6f), the intensity of the negative Z anomaly in the Ural Mountains increases and the negative anomaly center propagates westward into the key area, thus the key area presents a strong negative frequency anomaly. In Phases 7–8 (Figure 6g,h), the negative Z anomaly in the Ural Mountains weakens and propagates westward, with the impact on the key area gradually weakening. Therefore, the intensity of the negative frequency anomaly in the key area gradually weakens.

Figure 7 shows the evolution of the blocking frequency anomaly and the 10–30-day 500-hPa Z anomaly with the phase of the eastward-propagating mode over the mid-high-latitude Eurasia. In Phase 1 (Figure 7a), there is a northwest–southeast-tilted wave train with negative-positive-negative Z anomalies over the mid-high-latitude Eurasia. A positive Z anomaly controls the key area, and a negative Z anomaly exists to the southeast of the key area, leading to a northwest–southeast-oriented Z gradient in the east of the key area. This causes an easterly anomaly and weakens the westerly in the mid-high latitudes, facilitating the occurrence of blocking events [14]. In Phase 2 (Figure 7b), the negative Z anomaly in the eastern European region strengthens and propagates eastward to the southwestern the key area, which favors the weakening of the Z gradient in the west of the key area. The positive Z anomaly still controls the eastern key area. Therefore, the key area generally shows a positive frequency anomaly. In Phase 3–4 (Figure 7c, d), the negative Z anomaly in the eastern European region strengthens and gradually moves eastward to control the key area. The negative Z center is located to the south of the key area, which is conducive to the weakening of the Z gradient. Therefore, the key area is mainly controlled by a positive frequency anomaly, especially in the northern region. In Phase 5 (Figure 7e), the negative Z anomaly keeps propagating eastward, with the anomaly intensity to the east of the key area increasing, which is conducive to the intensifying of the Z gradient to the east of the key area. Thus, a significant negative frequency anomaly is observed to the east of the key area. In Phase 6–8 (Figure 7f–h), the newly generated Z anomaly over the East European Plain intensifies and moves eastward, which gradually controls the key area and the area to its south. The center of positive Z anomaly is located to the south of the key area, which is favorable to the enhancement of the Z gradient. Therefore, the key area mainly shows the negative frequency anomaly, especially in the northern region.

To further analyze the evolution of the Ural blocking frequency anomaly with the phase of eastward- and westward- propagating modes, the key area-averaged Ural blocking frequency anomaly as a function of the phase is given in Figure 8. As shown, the maximum blocking frequency anomaly (about 10%) in the key area is in Phase 2 during the westward-propagating mode, and the same phase as the eastward-propagating mode with the maximum anomaly reaching 3%. It is worth noting that, here, the frequency anomaly is the difference between the frequency in each phase, and the climatological mean frequency and the regional-mean data of 10% (3%) represents the average state to which extent the blocking frequency is more than the summer climatological mean. Previous studies revealed that both the ISO and the blocking events can cause large-scale extreme hot events in summer [12], but fewer studies have considered their co-effect on the extreme hot events in Eurasia. Therefore, in the next section, we take Phase 2 as a focus to analyze the frequency of summer extreme hot events in Eurasia when the Ural blocking occurs and does not occur during this phase.

5. Co-Effect of ISO-Ural Blocking on Extreme Hot Events

Figure 9a shows the horizontal distribution of the 500-hPa Z anomaly and the extreme hot event frequency anomaly in Phase 2 of the westward-propagating mode when there is no Ural blocking. Here, the extreme hot event frequency anomaly is the difference between the frequency of extreme hot events in Phase 2 and the climatological mean. The solid green box indicates the key area of Ural blocking. As depicted, with no Ural blocking occurring, a dipole-type wave train generally dominates over the Eurasia (negative in the west and positive in the east). Western Europe is controlled by a quasi-barotropic negative Z anomaly, which favors updrafts and leads to a negative T anomaly in the local area. As a result, Western Europe exhibits a significant negative extreme hot event frequency anomaly. A quasi-barotropic positive anomaly dominates over the area from the east of East European Plain to the Ural, with the Ural exhibiting a significant positive extreme hot event frequency anomaly.

Figure 9b shows the horizontal distribution of 500-hPa Z anomaly and extreme hot event frequency anomaly when the Ural blocking happens. The intensity of the negative Z anomaly in Western Europe significantly weakens and exhibits a southward shift, leading to a weak intensity of the negative extreme hot event frequency anomaly in situ. While the intensity of the positive (negative) Z anomaly in the East European Plain region and the Ural Mountains (north of the Japan Sea) increases significantly (Figure 9b). This results in a significant increase in the intensity of the positive (negative) extreme hot event frequency anomaly in the East European Plain and the Ural Mountains (north of the Japan Sea).

To better visualize the role of the Ural blocking in the impact of the mid-high-latitude ISO on the extreme hot events, the difference of each anomaly field between with blocking and without blocking (Figure 9a minus Figure 9b) is given in Figure 9c. When Ural blocking occurs, the Z anomaly is strong and positive for Europe and the high-latitude Ural Mountains and a strong negative north of the Japan Sea and south of the key area. Therefore, the occurrence of Ural blocking favors the extreme hot events in Europe and the high latitude areas of Ural Mountains and is not conducive to the extreme hot events north of the Japan Sea and south of the Ural Mountains. The above analysis suggests that the Ural blocking activity plays an important role in regulating the influence of the westward-propagating ISO mode on the extreme hot events over the Eurasia, especially over the Europe, the Ural Mountains and the region north of the Japan Sea.

Figure 10a is the same as Figure 9a, but for the eastward-propagating mode. As shown, without blocking, the Eurasian continent is under a control of a northwest–southeast-oriented wave train with alternated positive and negative Z anomalies. The Southwestern Europe and the Ural Mountains are controlled by a positive anomaly, which stimulates downward motion and hot events in the local areas. As a result, these areas are dominated by a significant positive extreme hot event frequency anomaly. The high-latitude Western Europe, the East European Plain, and the Northeastern China Plain are mainly controlled by a negative Z anomaly, corresponding to a significant negative extreme hot event frequency anomaly over these regions.

Figure 10b is the same as Figure 9b but for the eastward-propagating mode. With Ural blocking occurring, the positive Z anomaly in Western Europe and the Ural Mountains is significantly stronger and more northerly than in Figure 10a. Meanwhile, the negative Z anomaly in the East European Plain is weaker and more southerly. The distribution of these anomalies favors the positive extreme hot event frequency anomaly in Europe and the high-latitude areas of the Ural Mountains.

Figure 10c is the same as Figure 9c, but for the eastward-propagating mode. As shown, the occurrence of the Ural blocking is conducive to the extreme hot event over Europe and the high latitudes of the Ural Mountains and not conducive to the extreme hot event in the mid-latitudes of the Ural Mountains and the area north of the Japan Sea. Thus, the Ural blocking activity also regulates the influence of mid-high-latitude eastward-propagating ISO on the extreme hot events in Europe, the Ural Mountains, and north of the Japan Sea. In addition, we also calculate the SAT anomalies with the situation of no blocking occurring, blocking occurring in Phase2 during the two propagating modes, the results are similar to the above conclusion (not shown). The occurrence of Ural blocking is conducive to the the positive temperature anomalies in Europe and the high latitudes of the Ural Mountains and a passive temperature anomaly in the mid-latitudes of the Ural Mountains and north of the Japan Sea. Therefore, the Ural blocking activities significantly regulate the effect of the two propagating ISO modes on the surface temperature anomalies over the middle and high latitudes of Eurasia.

6. Conclusions

Based on the NCEP/NCAR daily reanalysis data from 1979 to 2018, the three-dimensional structure and evolution characteristics of two ISO modes over mid-high-latitude Eurasia in summer and their influence on the frequency of the Ural blocking are investigated. Then, the co-effect of the mid-high-latitude ISO and the Ural blocking on the extreme hot events in Eurasia are further discussed.

The results show that the maximum variability of geopotential height over mid-high latitude Eurasia occurs in the Ural Mountains (45°–67.5° E, 57.5°–62.5° N). The power spectrum analysis of the 250-hPa geopotential height field at the center of maximum variability reveals that there are two significant oscillation periods (10–30 and 30–50 days) in the atmosphere over mid-high latitudes. This study focuses on the 10–30-day ISO and uses the multivariate EEOF method to reveal that it has two modes over mid-high latitude Eurasia, namely the eastward- and the westward-propagating modes. The composite analysis shows that in the westward-propagating mode, the 250-hPa geopotential height anomaly over mid-high latitude Eurasia is mainly presented as an east-west-orientated dipole wave train, which shows a quasi-barotropic structure in the troposphere. The quasi-barotropic geopotential height anomaly is coupled with the temperature anomaly, and both propagate westward. The geopotential height and temperature anomalies satisfy the hydrostatic equilibrium relationship. That is, a quasi-barotropic positive (negative) geopotential height anomaly in the troposphere is corresponding to a positive (negative) temperature anomaly in the lower troposphere and to a negative (positive) temperature anomaly in the upper troposphere. In the eastward-propagating mode, the 250-hPa geopotential height anomaly over mid-high latitude Eurasia mainly presents a northwest–southeast-orientated wave train with alternated positive and negative anomalies. This wave train has a quasi-barotropic structure in the troposphere. The quasi-barotropic geopotential height anomaly, coupled with the temperature anomaly field, propagates southeastward. The geopotential height the temperature anomalies also satisfy the hydrostatic equilibrium relationship.

In both the eastward- and westward-propagating modes, the propagation of the atmospheric ISO can significantly affect the downstream circulation systems. Statistics reveal that, in summer, the Ural Mountains region has a large value of blocking frequency in Eurasia in area of (55°–90° E, 65°–76° N) region. which is defined as a Ural blocking key area. The phase composite analysis shows that the strongest positive blocking frequency anomaly in the key area appears in Phase 2 during both the westward- and the eastward-propagating modes, reaching 10% and 3%, respectively. In Phase 2 during the westward-propagating mode, a strong positive geopotential height anomaly controls the Ural blocking key area. In Phase 2 during the eastward propagating mode, a strong negative geopotential height anomaly appears on the southwest of the key area. The weakening of the geopotential height gradient in the key area leads to an easterly wind anomaly. There is a stronger positive geopotential height anomaly in the eastern key area, which is conducive to the positive blocking frequency anomaly in the key area.

Phase 2 with the most frequent blocking events in the key area was focused on analyzing the co-effect of the mid-high-latitude ISO and the Ural blocking on extreme hot events. The results showed that the Ural blocking activity can regulate the influence of each ISO mode on the extreme hot events. In Phase 2 during the westward propagating mode, when blocking occurs, the negative geopotential height anomalies in Western Europe are significantly weaker and more northerly. In contrast, the intensity of the positive (negative) geopotential height anomalies in the East European Plain and the Ural Mountains (north of the Japan Sea) are significantly stronger, leading to a significant increase in the strength of the positive (negative) anomalies of extreme high-temperature frequency in these regions. In Phase 2 during the eastward propagating mode, when blocking occurs, the positive geopotential height anomalies are significantly stronger and more northerly in Western Europe and the Ural Mountains. In contrast, the negative geopotential height anomalies are weaker and more southerly in the East European Plain. This distribution of geopotential height anomaly is more conducive to the positive anomalies of extreme high-temperature frequency in Europe and the high latitudes of the Ural Mountains.

Compared with the situation without the Ural blocking, the blocking activity results in the positive geopotential height anomalies throughout Europe and north of 60° N in the Ural Mountains and the negative geopotential height anomalies south of 60° N in the Ural Mountains and north of the Japan Sea. The occurrence of Ural blocking is conducive to the occurrence of extreme high-temperature events in Europe and the high latitudes of the Ural Mountains and a reduced frequency of extreme high-temperature events in the mid-latitudes of the Ural Mountains and north of the Japan Sea. Therefore, the Ural blocking activity plays an important role in regulating the impact of the eastward and westward propagating modes of mid-high-latitude atmospheric ISO on extreme high-temperature events in Europe, the Ural Mountains, and north of the Japan Sea.

It is worth noting that the definition of extreme events does not refer to the maintaining of the extreme hot events; the extreme hot events that happened in Phase 2 may be also affected by the previous ISO signal, which can make the extreme hot events happened in Phase 2 last much longer. According to the previous studies, the sources of the mid-high-latitude ISO are linked variously with tropical forcing [35], local air–sea interactions [36], and topography effects [37]. ENSO as a strong tropical signal can truly affect the mid-high-latitude atmospheric variation. To understand the effect of ENSO on mid-high-latitude ISO, we also test our conclusions during the EP and CP El Nino years (not shown). The general propagation characteristics of the eastward and westward propagating modes are similar between the EP and CP years, but the intensity of the tropospheric height anomaly modes differ. In addition, previous studies have shown that the heat waves over European continent are affected by the impact of the North Atlantic, such as the NAO and teleconnection wave trains from the North Atlantic to Asia [22]. Do the NAO and large-scale tropical signals such as ENSO affect the mid-high-latitude atmospheric ISO? The above contents will be discussed in future works.