1. Introduction
Extreme high-temperature events (EHEs) have significantly affected economic activity, human health, and natural ecosystems (e.g., Easterling et al. 2000; Perkins 2015; Lu and Chen 2016). Covering the largest area and containing the largest population, Asia suffers from numerous EHEs every year. Under the global warming background, EHEs with stronger intensity occurred more frequently over most areas of Asia (e.g., Zhai and Pan 2003; Meehl and Tebaldi 2004; Alexander et al. 2006; Seneviratne and Hauser 2020; Li et al. 2021). For instance, in 2013, a record-breaking EHE occurred in southeastern China, which persisted for more than 1 month, causing large economic losses and affecting more than half a billion people (Sun 2014; Sun et al. 2014). In 2018, a severe EHE occurred in northeast Asia and broke the daily maximum temperature record at 47 stations over northeast China, which affected more than 50 million hectares of crops in China and took more than 1032 lives (Zhou et al. 2019; Chen et al. 2019; Sun et al. 2020; Ren et al. 2020). Therefore, the increasing threat of EHEs calls for investigations of the physical processes responsible for EHE formation, which are of great importance for EHE forecasting and related disaster prevention.
Some atmospheric patterns have been revealed to play roles in the interannual and interdecadal variations in Asian EHE frequency, such as the western Pacific subtropical high (WPSH; e.g., Wang et al. 2016; Luo and Lau 2017; Ding et al. 2018; Gao et al. 2018), blocking high (e.g., Cassou et al. 2005; Pfahl and Wernli 2012; Li et al. 2020a,b; Fang and Lu 2020), East Asian subtropical jet (EASJ; e.g., Sun 2014; Wang et al. 2016; Li and Sun 2018; Hong et al. 2020), circumglobal teleconnection pattern (CGT; e.g., Wang et al. 2013; Li and Sun 2018; Choi et al. 2020; Luo and Lau 2020), Pacific–Japan pattern or East Asia–Pacific pattern (PJ/EAP; e.g., Lee and Lee 2016; Zhu et al. 2020a; Noh et al. 2021), North Atlantic Oscillation (e.g., Sun 2012; Hong et al. 2022), and so on. These atmospheric patterns can lead to anomalous local highs, which further modulate the heat flux and temperature advection, consequently contributing to EHE occurrence.
In addition to atmospheric patterns, some boundary forcings, such as anomalous sea surface temperature (SST), sea ice, and soil moisture, are also connected to variations in Asian EHE frequency. For example, the anomalous SST patterns in the North Atlantic are connected to the interannual variations in EHE frequency over China by stimulating zonal Rossby wave trains over the North Atlantic–Eurasia sector (Sun 2014; Deng et al. 2019; Zhu et al. 2020a). The El Niño (La Niña) events significantly increase (decrease) the EHE frequency over southern East Asia by inducing anomalous subsidence over the western North Pacific and a stronger South Asian high (Wang et al. 2014; Thirumalai et al. 2017; Luo and Lau 2019), while they significantly decrease (increase) the EHE frequency over central Asia and northeast Asia by modulating the CGT pattern and shifting the EASJ northward (Luo and Lau 2020). The phase shift of the Atlantic multidecadal oscillation (AMO) around the mid-1990s has contributed to the interdecadal increase in EHE frequency over northeastern Asia by exciting a zonal Rossby wave train over Eurasia (Dong et al. 2016; Hong et al. 2020). The Pacific decadal oscillation (PDO) impacts the EHE frequency over southern (northern) China by changing the Walker circulation (PJ/EAP pattern; Zhang et al. 2020; Zhu et al. 2020b). The PDO also modulates the relationship between the WPSH and EHE frequency over southeastern China (Liu et al. 2019). Through land–atmosphere feedbacks, drier soil moisture has been revealed to increase EHE frequency and intensity over eastern China (e.g., Zhang and Wu 2011; Meng and Shen 2014; Wu and Zhang 2015), and the interdecadal decrease in soil moisture over Mongolia after the late 1990s contributes to the interdecadal increase in the local EHE frequency (Erdenebat and Sato 2016). The rapid melting of Arctic sea ice is found to influence the EHE or hot and drought event frequency in East Asia by exciting Rossby wave trains over Eurasia and shifting the EASJ (e.g., Li et al. 2018; Zhang et al. 2018; Wu and Francis 2019; Deng et al. 2020).
The present literatures on EHEs over Asia have mainly focused on mid- to low latitudes. Relatively limited attention has been given to the EHEs in northern Asia (NA). Some studies have indicated that, along with global warming, both the frequency and intensity of NA EHEs show significant increasing trends (Bulygina et al. 2007; Groisman et al. 2013; Degefie et al. 2014; Bardin and Platova 2019; Fang and Lu 2020; Jiang et al. 2022). In addition, recent observations show that NA has suffered from more severe EHEs in recent years. For instance, in 2016, a long-persisting EHE occurred on the Yamal Peninsula (northwestern Siberia), which melted the top layer of the permafrost, leading to the activation of spores and the outbreak of anthrax (Arkhangelskaya 2016; Hueffer et al. 2020; Ezhova et al. 2021). In 2020, an unprecedented EHE hit Siberia, which was the first time that the circumpolar north observed a daily maximum temperature (Tmax) above 38°C (WMO 2020; Overland and Wang 2021; Ciavarella et al. 2021; Xu et al. 2021). Therefore, in addition to the trend, the variability of NA EHEs on other time scales should be investigated. Recently, Hong et al. (2022) explored the interdecadal and interannual variations in NA EHE frequency and related mechanisms from a climate perspective. This study is an extension of Hong et al. (2022) and focuses on the atmospheric circulations responsible for NA EHE formation from a synoptic perspective.
Previous studies have shown that the synoptic circulations responsible for the EHE occurrence are different over different regions (e.g., Chen and Lu 2015; Wang et al. 2016; Hu et al. 2019). The synoptic circulations associated with the EHEs over mid- to low-latitude Asia have been revealed by previous studies (e.g., Chen and Lu 2015; Wang et al. 2016; Luo and Lau 2017; Xu et al. 2019; Li et al. 2020a; Kim et al. 2021). However, currently, we have limited knowledge of the synoptic circulations responsible for the EHEs over NA. This study therefore attempts to explore the specific synoptic circulations responsible for the EHE occurrence over NA and further to investigate the precursors and long-term changes of these synoptic circulation–related EHEs, with aims to deepen our understanding of the NA EHE formation and provide potential signals for the EHE forecast in the future.
The occurrence of EHEs generally exhibits a strong regional feature; therefore, for a large region, regionalization is needed to specifically study EHE formation. In addition, there could be several synoptic circulation patterns influencing the occurrence of EHEs over a certain region (e.g., Cassou et al. 2005; Stefanon et al. 2012; Yeo et al. 2019; Kim et al. 2021; Agel et al. 2021); therefore, clustering the EHE-related atmospheric patterns is helpful for understanding the formation of EHEs. In this study, because NA covers a large region, we first regionalize NA according to the variation in Tmax and then investigate the EHE-related synoptic patterns based on the clustering method. The paper is organized as follows: the data and methods are described in section 2. The features of synoptic patterns responsible for NA EHEs are investigated in section 3. The atmospheric circulation evolution of different synoptic pattern–related NA EHEs is examined in section 4. The long-term changes in the NA EHEs are presented in section 5. The conclusions and discussion are presented in section 6.
2. Data and methods
a. Data
The Tmax data at meteorological stations in NA are mainly derived from Global Historical Climatology Network–Daily (GHCN-D; Menne et al. 2012), version 3.28. In this study, we define NA as the region covering the area at 40°–75°N, 60°–140°E and focus on EHEs occurring in summer [June–August (JJA)], consistent with Hong et al. (2022). Because of the numerous missing Tmax records in the GHCN-D after 2019, the research period is identified as 1960–2018. For data validity, the stations with missing Tmax records greater than 5% in JJA over 1960–2018 were removed, and the remaining missing daily values were replaced by the climatological Tmax values of the given days. Therefore, 224 stations in NA (out of China) are selected. As indicated by Hong et al. (2022), the valid Chinese stations in the GHCN-D rapidly decreased from 53 to 15 after 2013, and consequently, the data at these 53 stations derived from the China Meteorological Administration (CMA; https://data.cma.cn) are employed. The spatial distribution of all 277 meteorological stations is shown in Fig. 1.
Spatial distribution of 277 meteorological stations in NA (40°–75°N, 60°–140°E). The 224 red (53 blue) dots represent the station data derived from GHCN-D (CMA).
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
According to the climatological diurnal temperature cycle, the daily temperature generally reaches its maximum at approximately 1400 local time (e.g., Hughes et al. 2007; Wang et al. 2017). Therefore, investigating the atmospheric circulation and surface heat flux at the hottest hour could be helpful for understanding EHE formation. After regionalization (shown in section 3a), western, central, and southeastern NA cover the UTC + 4 to + 6 h, UTC + 6 to + 9 h, and UTC + 7 to + 9 h time zones, respectively. Therefore, we adopt the UTC + 5 h (UTC + 8 h) time zone for the EHE-related analysis of western (central and southeastern) NA. Hence, in this study, the western (central and southeastern) NA EHE-related analysis is based on the hourly data at 0900 UTC (0600 UTC), which is 1400 local time, and the average or anomalous atmospheric circulations are calculated based on these hourly datasets.
According to the perturbation method (Holton 2004), the meteorological variables can be divided into two parts. One is the climatological mean, which is usually assumed to be independent of time; the other is the anomaly, which is the local deviation of the variable from the basic state. The anomaly can well reflect the changes in synoptical atmospheric circulations. Therefore, to outstand the REHE-related synoptic atmospheric circulation signals, the meteorological variable anomalies are calculated by removing the climatological mean in each grid point, and the climatological mean is calculated by averaging the variable on each calendar day over the 1960–2018 period and smoothing the daily average with a 15-day running mean.
b. Definition
In this study, the EHE days are identified based on the quantile of Tmax, and the process is shown as follows:
-
The 1961–90 period is selected as the reference period for defining the EHE threshold, which is regarded as the standard reference period by the World Meteorological Organization (WMO 2017) and the Expert Team on Climate Change Detection and Indices (https://etccdi.pacificclimate.org).
-
For each calendar day in JJA, the EHE threshold is defined as the 95th percentile of Tmax for the 450 days (15 × 30 days; 15 days cover 7 days on each side of the day) over the reference period. As revealed by Zhang et al. (2005), much uncertainty still exists in the estimation of the EHE threshold based on a narrow moving window (5 days), and the amplitude and seasonal cycle of the EHE threshold are reduced based on a wide moving window (25 days). In this study, to reduce the uncertainty and retain the seasonal cycle of the EHE threshold, a moderate 15-day moving window is selected following the previous studies (e.g., Della-Marta et al. 2007; Kuglitsch et al. 2010; Fischer and Schär 2010; Perkins et al. 2012; Zhou and Wu 2016; Deng et al. 2019, 2020; Sulikowska and Wypych 2020; Engdaw et al. 2022).
-
A summer day over 1960–2018 with Tmax equal to or greater than the EHE threshold is identified as an EHE day.
In section 2a, we have replaced the missing daily Tmax records with the climatological Tmax values of the given day, and the effect of this procedure on the estimation of the EHE days was further examined here. Specifically, six NA stations with complete records were randomly selected, and the EHE days were identified using the abovementioned method. Furthermore, 95% of the Tmax records at these six stations were randomly sampled (95% × 59 × 92 = 5157 days), and the remaining 5% of records were replaced by the climatological Tmax values of the given days, generating a simulated Tmax. Based on the simulated Tmax, the simulated EHE days were also identified using the abovementioned method and were further compared with the EHE days. If a simulated EHE day is also the EHE day, the day is considered a “robust EHE day,” and the ratio of the robust EHE days relative to the EHE days is defined as the “robust ratio.” The above process was repeated 1000 times, and the average robust ratios of these 1000 times for all the six selected stations are approximately 95%. In addition, 264 of the 277 NA stations have missing Tmax records less than 3% of the total summer days over 1960–2018 (in which 117 stations have no missing Tmax record). The results indicate that the replacement procedure has no significant effect on the estimation of EHE days.
Based on the EHE identified at each station, the regional EHE (REHE) is further defined in each subregion (the whole NA will be divided into three subregions in section 3a), which reflects the days with EHEs occurring at some stations in a certain subregion simultaneously. The REHE is identified by the following steps:
-
For a specific day, the number of stations within a subregion having EHEs is counted, which is referred to as the EHE station number.
-
For each subregion, the 95th percentile of the EHE station number for the 2760 days (92 × 30 days) over the reference period is defined as the REHE threshold.
-
A summer day over 1960–2018 with an EHE station number larger than the REHE threshold is identified as an REHE day.
-
An event containing two or more consecutive REHE days is called a consecutive REHE, and the one containing only an REHE day is called an isolated REHE. Both the consecutive REHE and the isolated REHE belong to the REHEs and are researched in this study.
-
Following Erdenebat and Sato (2016) and Xu et al. (2019), two adjacent REHEs with a gap of 1 day are further merged into one event, and the gap day is also calculated in the REHE days. Furthermore, an REHE lasting for 3 consecutive days or more is referred to as a regional heatwave, and each day of the regional heatwave is referred to as a regional heatwave day.
c. Method
NA covers a large area, and regionalization is essential to identify homogeneous subregions with consistent Tmax variation. Furthermore, the different synoptic patterns responsible for the REHE formation in the same subregion are investigated by subjecting the clustering method to the NA REHE-related atmospheric circulation anomalies. Recently, Tang et al. (2021) introduced a clustering method called spectral clustering into the analysis of extreme precipitation events, which has a better performance than traditional clustering algorithms such as k-means, when the number of clusters is small and the data are sparse (Ng et al. 2002; von Luxburg 2007). In this study, spectral clustering is applied to cluster the synoptic patterns responsible for REHE formation in each subregion of NA, and the Python machine learning package (Pedregosa et al. 2011) is used to perform the spectral clustering analysis on the target field. The target field is the meteorological variable field used for the spectral clustering analysis. Some previous studies clustered the atmospheric circulations associated with an extreme climate event at a single level, such as 500-hPa geopotential height (e.g., Loikith and Broccoli 2012; Chen and Lu 2015; Horton et al. 2015). However, the extreme climate event could be the result of the combination of multilevel atmospheric circulations. Following Tang et al. (2021), we cluster the REHE-related atmospheric circulations using the horizontal wind anomalies in the upper (200 hPa), mid- (500 hPa), and mid- to lower (700 hPa) troposphere. The atmospheric circulation at 850 hPa is generally used to reflect the atmospheric situation in the lower troposphere. However, because of the mountains and the Mongolian Plateau, the climatological surface pressure in summer over some areas of southern NA is less than or around 850 hPa (figure not shown). Therefore, in this study, the horizontal wind anomalies at 700 hPa are used to reflect the atmospheric circulations in the mid- to lower troposphere. Specifically, the target fields are constructed as follows:
-
The horizontal wind (zonal and meridional winds) anomalies at 200, 500, and 700 hPa during the REHEs are selected. If an REHE lasts for 2 days or more, the temporal-average three-level horizontal wind anomalies during the REHE days are selected.
-
To eliminate the difference in the magnitude of wind velocity, the six selected anomalous horizontal wind fields (zonal and meridional winds at three levels) are standardized in the time dimension.
-
The six standardized horizontal wind fields with two dimensions in space are flattened to one dimension in space and are further merged as the target field, sorted by zonal wind after meridional wind and from high level to low level (the order of these variables will not influence the clustering results).
Following Tang et al. (2021), we set some options in the spectral clustering algorithm before clustering: the method to construct the affinity matrix is set to “nearest_neighbors,” and the method to assign the cluster labels is set to “kmeans.” The parameter “n_neighbors” is important for the clustering results in the method of “nearest_neighbors.” Consistent with the 95th percentile threshold for the EHE/REHE definition, we set the parameter “n_neighbors” to 5% × nREHE, where “nREHE” represents the total frequency of the REHE. Based on the spectral clustering method, each REHE is identified to be related to one of the synoptic patterns.
For regionalization, we propose a new hybrid regionalization approach based on the method from Yu et al. (2018, 2021). We replace the k-means clustering analysis in the hybrid regionalization approach from Yu et al. (2018, 2021) by spectral clustering for a better clustering effect. Then, some modifications are made to the hybrid regionalization approach. The specific procedure is as follows:
-
The JJA Tmax data are first standardized at each station. The standardization is calculated by subtracting the climatological values and then dividing by the standard deviation over 1960–2018.
-
Then, the empirical orthogonal function (EOF; Lorenz 1956) is used to analyze the standardized Tmax data. To obtain the EOF modes with more representativeness, the first n EOF modes with larger explained variance are retained.
-
The retained n EOF modes are further rotated by the varimax rotated EOF (REOF; Richman 1986) method, and the first n REOF patterns can be obtained. Our following analysis indicates that the large loadings of the first three REOF patterns can cover the whole NA well and have limited overlapped regions among the patterns. Therefore, the value of n is set to 3 in this study.
-
The loadings of the n REOF patterns at NA stations are further clustered into n clusters by the spectral clustering method. The spectral clustering method has been introduced above, and the parameter of “n_neighbors” is 5% × nstation, where “nstation” represents the station number of 277 over NA. Therefore, the NA stations can be divided into n clusters. According to the spatial distribution of the stations assigned to each cluster, NA can be consequently divided into n subregions with consistent variations in Tmax.
To reflect the propagation direction of Rossby wave energy, the wave activity flux based on the method introduced by Takaya and Nakamura (2001) was employed in this study.
3. Synoptic patterns responsible for NA REHEs
a. Regionalization
According to the hybrid regionalization approach introduced in section 2c, we first apply the EOF analysis to the standardized JJA Tmax data at 277 NA stations. The first three EOF modes are retained and rotated to obtain three REOF patterns, which exhibit large loadings over central Siberia, western Siberia, and eastern Mongolia–northeastern China, respectively (Figs. 2a–c). The first two and first four EOF modes are also retained for the REOF analysis to obtain the REOF1–2 and 1–4 patterns, respectively. The REOF1–2 patterns can hardly cover the whole region of NA (figure not shown); the REOF1–4 patterns can cover the whole NA, but the large loadings of the REOF patterns have some overlap (figure not shown). Relatively, not only do the large loadings of the REOF1–3 patterns cover the whole NA well, but they also have limited overlapped regions among the patterns (Figs. 2a–c), which are more appropriate for regionalization.
(a) REOF1, (b) REOF2, and (c) REOF3 patterns of standardized JJA NA Tmax over the 1960–2018 period. (d) The spatial distribution of the 98 stations over western NA (green dots), 106 stations over central NA (red dots), and 73 stations over southeastern NA (blue dots) by applying spectral clustering to the REOF1–3 patterns. The explained variance of each REOF mode to the total variance is shown in the top-right corner of (a)–(c).
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
To test the robustness of this result, we further examine the REOF1–3 patterns rotated using the EOF1–4, EOF1–5, and EOF1–6 modes, respectively. Although there are some visible differences, the large loadings of the REOF1–3 patterns rotated using the EOF1–4, EOF1–5, and EOF1–6 modes are similar (figure not shown), being consistent with those of the REOF1–3 patterns in Fig. 2. This result indicates that the REOF patterns shown in Figs. 2a–c are robust. Finally, by applying the spectral clustering to these three REOF patterns, the 277 NA stations can be divided into three clusters. The stations assigned to the first/second/third cluster (green/red/blue dots in Fig. 2d) are located in western/central/southeastern NA, and the whole NA is correspondingly divided into these three subregions. After regionalization, each subregion of NA has consistent Tmax variations within, and EHEs tend to simultaneously occur in large areas of the subregion.
The REHEs are further identified in each subregion of NA, and some features of REHEs in the three subregions are shown in Table 1. There are 98, 106, and 73 stations in the three subregions, respectively, and their REHE thresholds are approximately 20% of the total station number. The maximum number of stations with simultaneous EHEs is applied to illustrate the maximum scope of the REHE. The results show that the EHEs simultaneously occur with a maximum scope at 60%–80% stations in the three subregions, and the highest proportion is 82.2% in southeastern NA. During 1960–2018, the total REHE frequencies are 166 in western NA, 177 in central NA, and 170 in southeastern NA. The total number of REHE days is more than 400 in the three subregions, with an average duration of 2–3 days per REHE.
Features of REHEs over three NA subregions.
b. REHE-related synoptic patterns
Previous studies have indicated that there could be several synoptic circulation patterns influencing the occurrence of EHEs over a certain region (e.g., Cassou et al. 2005; Stefanon et al. 2012; Yeo et al. 2019; Kim et al. 2021; Agel et al. 2021). If we do not divide the REHE-related synoptic circulation patterns, the obtained result is the mixture of these patterns, which is difficult to be physically understood. Therefore, to better understand REHE formation over NA, the associated synoptic circulation patterns are investigated in this subsection using the spectral clustering method introduced in section 2c.
To obtain the proper synoptic patterns, the coverage area of each subregion is expanded and referred to as the classification region. The classification regions (purple boxes in Fig. 3) for clustering the synoptic circulation patterns are larger than the coverage areas of stations in the three NA subregions. The horizontal wind anomalies over the classification region at three levels (200, 500, and 700 hPa) during the REHEs are subjected to the spectral clustering. The cluster number of spectral clustering is determined based on the Calinski–Harabasz score (Caliński and Harabasz 1974). We calculate the Calinski–Harabasz scores for cluster numbers from 2 to 10 for each subregion, and the scores peak in two clusters in all three subregions. Therefore, the REHEs in each subregion are classified into two clusters based on the REHE-related atmospheric circulation patterns, and the temporal distribution of these two-cluster REHEs is shown in Fig. 4.
Composite maps of (a) 500-hPa geopotential height anomalies (shading; units: gpm) and 700-hPa horizontal wind anomalies (vectors; units: m s−1) during the WE-1 pattern–related REHEs. (b)–(f) As in (a), but during the WE-2, CE-1, CE-2, SE-1, and SE-2 pattern–related REHEs, respectively. Stippled areas are significant at the 95% confidence level based on Student’s t test. The purple boxes represent the classification regions for (top) western NA, (middle) central NA, and (bottom) southeastern NA, respectively.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
(a) Temporal distribution of the WE-1 (red rectangles) and WE-2 (blue rectangles) pattern–related REHEs over 1960–2018. The REHEs (not) included in the composite analysis in section 4 are darkened and outlined by black boxes (lightened). The horizontal axis represents the days of summer, and the vertical axis represents each year of 1960–2018. The total numbers of the WE-1 and WE-2 pattern–related REHEs are shown as the first pair of numbers in the parentheses in the top-left corner of (a), and the numbers of the REHEs used in the composite analysis in section 4 are shown as the second pair of numbers. (b),(c) As in (a), but for the CE-1 and CE-2 pattern–related REHEs and the SE-1 and SE-2 pattern–related REHEs, respectively.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
The local atmospheric circulation anomalies during the REHEs assigned to the first/second cluster in each subregion (represented by red/blue rectangles in Fig. 4) are further composited. If an REHE lasts for 2 days or more, the atmospheric circulation anomalies are temporally averaged first. The result shows that the synoptic patterns responsible for the two-cluster REHEs in the same subregion are quite different (Fig. 3). Hereafter, we refer to the synoptic pattern responsible for the first cluster of the western NA REHE as the “WE-1” pattern (“WE-2,” “CE-1,” “CE-2,” “SE-1,” and “SE-2” patterns can be similarly expressed).
The WE-1 and CE-1 patterns are dominated by anomalous quasi-barotropic highs over western NA and central NA, respectively (Figs. 3a,c), which are referred to as “the monopole high pattern.” In contrast, the WE-2, CE-2, SE-1, and SE-2 patterns are dominated by quasi-barotropic dipole patterns over their own subregions, which are referred to as “the synoptic dipole pattern” (Figs. 3b,d–f). The WE-2 (CE-2/SE-2) pattern exhibits a northwest–southeast dipole pattern, with an anomalous high to the west of (around/to the east of) Lake Baikal and an anomalous low around the Ural Mountains (to the north of Lake Balkhash/to the northwest of Lake Baikal; Figs. 3b,d,f), while the SE-1 pattern shows a northeast–southwest dipole pattern, with an anomalous high to the southeast of Lake Baikal and an anomalous low to the north of Sakhalin Island (Fig. 3e).
Furthermore, the spatial differences in EHE frequencies during different synoptic pattern–related REHE days are investigated. For each station, the EHE occurrence probability is defined as the percentage of the EHE days of the REHEs relative to all the REHE days. The EHE occurrence probability in Fig. 5a (Figs. 5b,c) indicates the percentage of the EHE days of the western NA (the WE-1/WE-2 pattern related) REHEs relative to all the western NA (the WE-1/WE-2 pattern related) REHE days. In a similar way, the EHE occurrence probabilities for central and southeastern NA are also calculated. First, the spatial distributions of the EHE occurrence probabilities for western, central, and southeastern NA are generally consistent with those of the regionalization of NA (Figs. 5a,d,g), further confirming the reasonability of the regionalization. The EHE occurrence probabilities associated with different synoptic pattern–related REHE days for each subregion show similarly high probabilities over their own domains (Figs. 5b,c,e,f,h,i). On the other hand, they also exhibit spatial differences. The high probabilities are located more northwestward/southeastward for the WE-1/WE-2 pattern–related REHE days over western NA (Figs. 5b,c), more northward/southward for the CE-1/CE-2 pattern–related REHE days over central NA (Figs. 5e,f), and southwestward/northeastward for the SE-1/SE-2 pattern–related REHE days over southeastern NA (Figs. 5h,i). It is noteworthy that among all synoptic patterns, the REHE days associated with the WE-2 pattern show the highest EHE occurrence probability of over 60% within western NA.
Spatial distribution of EHE occurrence probabilities (units: %) over NA for the (a) western NA REHEs, (d) central NA REHEs, and (g) southeastern NA REHEs. (b) and (c),(e) and (f),(h) and (i) As in (a), but for the WE-1 and WE-2 pattern–related REHEs, the CE-1 and CE-2 pattern–related REHEs, and the SE-1 and SE-2 pattern–related REHEs, respectively. The 98, 106, and 73 stations in (top) western NA, (middle) central NA, and (bottom) southeastern NA are marked by blue dots.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
The persistence of each synoptic pattern–related REHE is evaluated from the perspective of a regional heatwave and is shown in Table 2. Though the heatwave frequencies occupy less than 50% of the REHE frequencies, the regional heatwave days account for more than 69% (50%) of the WE-1, CE-1, CE-2, and SE-2 (WE-2 and SE-1) pattern–related REHE days. The results indicate that NA REHE days mainly occur in the form of regional heatwave days. Moreover, the average duration is more than 4.6 days for the WE-1, CE-1, CE-2, and SE-2 pattern–related regional heatwaves and nearly 4 days for the WE-2 and SE-1 pattern–related regional heatwaves. The longest regional heatwave is related to the WE-1 and SE-2 patterns, lasting 2 weeks (14 days), and the longest duration of the regional heatwave related to other synoptic patterns is 11 days, except that related to the SE-1 pattern, which lasts 7 days.
Persistent features of each synoptic pattern–related REHE.
c. Local mechanisms for REHE formation
The analysis in the last subsection explores the observation that the synoptic circulations responsible for the REHEs in each subregion of NA are different. In this subsection, the influence of the local atmospheric circulation anomalies on the EHE occurrence is diagnosed from both diabatic and adiabatic heating processes. The REHE-related meteorological variables are composited on the synoptic pattern–related REHE days (see the red/blue rectangles in Fig. 4), and the results are shown in Fig. 6.
Composite maps of net surface solar radiation anomalies (shading; units: W m−2) during the (a) WE-1 and (b) WE-2 pattern–related REHE days. The positive (negative) anomalies represent the download (upward) solar radiation anomalies. (e),(f) As in (a) and (b), but during the CE-1 and CE-2 pattern–related REHE days, respectively. (i),(j) As in (a) and (b), but during the SE-1 and SE-2 pattern–related REHE days, respectively. (c) and (d), (g) and (h),(k) and (l) As in (a) and (b), (e) and (f), and (i) and (j), but for vertically integrated horizontal temperature advection (from surface pressure to 700 hPa; shading; units: W m−2) and 700-hPa horizontal wind anomalies (vectors; units: m s−1).
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
The anomalous highs of the synoptic patterns are generally associated with more solar radiation reaching and warming the ground, which favors the EHE formation (Figs. 6a,b,e,f,i,j). In addition to the radiation process, anomalous subsidence motions are induced to the eastern part of the anomalous highs, which heat the air and provide a favorable condition for EHE formation there (figure not shown). In addition, the warm horizontal temperature advection in the lower troposphere plays an important role in the EHE formation and shows relatively stronger signals in the dipole patterns (Figs. 6d,h,k,l) than in the monopole high patterns (Figs. 6c,g). Among the dipole patterns, the warm horizontal temperature advection anomalies are generally smaller than the surface net solar radiation anomalies (Figs. 6b,d,f,h,i–l), except for the WE-2 pattern in the northeastern part of western NA (Figs. 6b,d). The results indicate that the radiation process generally dominates the NA EHE formation for most of the synoptic patterns, but the WE-2 pattern–related horizontal temperature advection contributes more to the EHE formation in the northeastern part of western NA.
4. Evolution of atmospheric circulation patterns associated with NA REHEs
Previous studies have explored precursors of the REHEs occurring over some regions. For instance, the evolution of the REHEs in southern China is associated with a dipole pattern originating from the tropical western Pacific and propagating northwestward (Chen et al. 2016). The REHEs occurring in South Korea and southern–central Japan are linked to two different precursor wave trains, which have an anomalous cyclone and anticyclone in the east of the Aral Sea, respectively (Xu et al. 2019). The different upstream blocking highs over the Europe–Atlantic sector can also be the precursors to the REHEs occurring in different subregions of China in the following few days (Li et al. 2020a). However, we have limited knowledge of the precursor atmospheric circulations associated with the different synoptic pattern–related NA REHEs shown in the last section. Therefore, we further investigate atmospheric circulation evolutions and the precursors of NA REHEs in this section.
Based on the composite analysis, the evolutions of 200-hPa geopotential height anomalies associated with different synoptic pattern–related NA REHEs from day −9 to +3 are exhibited in Figs. 7–9, respectively. Specifically, we refer to the k days before the REHEs as “day −k” and the k days after the REHEs as “day +k,” with “day 0” indicating the REHE days. The period of synoptic processes is approximately 7 days. To prevent contamination between the REHEs close to each other, only the REHEs with no other REHEs occurring over the same subregion in the preceding and following 7 days are selected to perform the composite analysis. According to this criterion, 45, 64, 38, 67, 38, and 55 REHEs are finally selected for the composite analysis of the WE-1, WE-2, CE-1, CE-2, SE-1, and SE-2 pattern–related REHEs, respectively. These selected REHEs can be seen in Fig. 4 (see the dark red/blue rectangles).
Composite maps of the 200-hPa geopotential height anomalies (shading; units: gpm) and the related horizontal wave activity flux (vectors; units: m2 s−2) on (a) day −9, (b) day −6, (c) day −3, (d) day 0, and (e) day +3 of the WE-1 pattern–related REHEs. (f)–(j) As in (a)–(e), but for WE-2 pattern–related REHEs. The stippled areas are significant at the 95% confidence level based on Student’s t test. The three rectangles—A1: 45°–65°N, 30°W–0°; B1: 50°–70°N, 50°–90°E; and C1: 55°–80°N, 5°–45°E in (c) and A2: 55°–75°N, 15°W–15°E; B2: 45°–60°N, 60°–90°E; and C2: 55°–80°N, 35°–60°E in (h)—are the key domains to define PWTI_WE-1 and PWTI_WE-2, respectively.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
As in Fig. 7, but for the (a)–(e) CE-1 and (f)–(j) CE-2 pattern–related REHEs. The three rectangles—A3: 40°–60°N, 10°–40°E; B3: 50°–75°N, 80°–125°E; and C3: 50°–70°N, 45°–70°E in (c) and A4: 65°–85°N, 15°W–35°E; B4: 45°–60°N, 70°–105°E; and C4: 50°–75°N, 40°–65°E in (h)—are the key domains to define PWTI_CE-1 and PWTI_CE-2, respectively.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
As in Fig. 7, but for the (a)–(e) SE-1 and (f)–(j) SE-2 pattern–related REHEs. The areas outlined by the blue boxes are the key regions for the definition of PWTI_SE-1 (left) and PWTI_SE-2 (right), respectively. The two rectangles—A5: 40°–60°N, 95°–125°E and B5: 40°–60°N, 60°–85°E in (c) and A6: 45°–60°N, 95°–130°E and B6: 60°–80°N, 70°–120°E in (h)—are the key domains to define PWTI_SE-1 and PWTI_SE-2, respectively.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
We have plotted the consecutive evolution of 200-hPa geopotential height anomalies from day −10 to +4 (figure not shown) and found that the figures from day −9 to +3 with a 3-day interval can well reflect the evolution of 200-hPa geopotential height anomalies for each synoptic pattern–related REHE. Therefore, to keep the same interval and fewer figures, in this study, we present the 200-hPa geopotential height anomalies on day −9, −6, −3, 0, and +3 of the synoptic pattern–related REHEs to reflect the evolution of atmospheric circulation anomalies in the higher troposphere before and after these REHEs.
To further reflect the evolution of the key atmospheric circulations associated with the NA REHEs, some precursor wave train indices (PWTIs) are defined based on Figs. 7–9. The definition of the PWTIs is shown in Table 3, and the evolutions of these PWTIs before and after the REHEs are shown in Fig. 10.
Definition of precursor wave train indices. Z200 represents the standardized regional-mean 200-hPa geopotential heights.
Temporal evolution of the standardized PWTIs from days −9 to +3 of the (a) WE-1 (red bar) and WE-2 (blue bar) pattern–related REHEs. (b),(c) As in (a), but for the CE-1 and CE-2 patterns and the SE-1 and SE-2 patterns, respectively. Day −k and +k of the horizontal axis represents k day before and after the REHEs, respectively, and day 0 represents the REHE days. The dot inside the bar represents the average index on the day, and the top and bottom of the bar represents the upper and lower quartile value of the indices on the day, respectively. The bar with a dark (light) color and black (white) dot indicates the average index on the day is significant (not significant) at the 95% confidence level based on the Student’s t test.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
a. Western NA
As shown in Figs. 7a,b,f,g, anomalous highs emerge over a similar region of the northeastern North Atlantic 1 week before the WE-1 and WE-2 pattern–related REHEs, but they further develop in different ways, contributing to different precursor wave trains on day −3. The precursor wave train for the WE-1 (WE-2) pattern–related REHEs exhibits two significant anomalous highs to the south (east) of Iceland and the north (northeast) of the Aral Sea and an anomalous low over northern (eastern) Europe (Figs. 7c,h). The activity of the precursor wave train before the WE-1 (WE-2) pattern–related REHEs can be confirmed by the significant positive anomalies of the PWTI_WE-1 (PWTI_WE-2) from day −6 to 0, which begins to show significant signals on day −7 (day −6), peaks on day −1, and weakens afterward (Fig. 10a). From day −3 to 0, the precursor wave trains further propagate eastward to form the WE-1 and WE-2 patterns, respectively, leading to the western NA REHE (Figs. 7d,i). After day 0, the anomalous high of the WE-1 pattern remains over western NA but weakens and disperses wave energy southeastward to excite an anomalous low over southeastern NA (Fig. 7e). In contrast, the WE-2 pattern shows a moving feature, moving eastward and exciting an anomalous low over central East China (Fig. 7j).
b. Central NA
For central NA, the CE-1 and CE-2 pattern–related REHEs are associated with different precursor wave trains originating from western Europe and the Norwegian Sea on day −6, respectively (Figs. 8b,g). From day −6 to 0, the precursor wave train for the CE-1 (CE-2) pattern–related REHEs further propagates northeastward (southeastward) through the Ural Mountains to central NA (Figs. 8b–d,g–i), contributing to the REHE there. Along with the evolution of the precursor wave train, the PWTI_CE-1 significantly increases on day −4, peaks on day 0, and weakens afterward. Differently, the PWTI_CE-2 begins to show significant positive anomalies earlier on day −6, which peak on day −2 and restore quickly on day +1 (Fig. 10b). After day 0, the CE-1 pattern is weakened and disperses wave energy southeastward to excite a tripole pattern over the East Asia–northwestern Pacific sector (Fig. 8e), while the CE-2 pattern moves eastward and disperses wave energy eastward and southeastward, leading to a quadrupole pattern over the East Asia–northwestern Pacific sector (Fig. 8j).
c. Southeastern NA
A common feature for the SE-1 and SE-2 pattern–related REHEs is that anomalous highs are over a similar region to the northeast of Lake Balkhash around a week before (Figs. 9b,g). Differently, an anomalous low is enhanced to the northwest of Lake Balkhash (over northwestern NA) on day −3 of the SE-1 (SE-2) pattern–related REHEs, which combines with the aforementioned anomalous high to form a precursor zonal (meridional) dipole pattern over western–central (northern–central) NA (Figs. 9c,h). From day −3 to 0, the precursor zonal dipole pattern is shifted southeastward and disperses wave energy eastward to excite an anomalous low over northeastern NA, leading to the SE-1 pattern and the REHE over southeastern NA (Fig. 9d); the precursor meridional dipole pattern is strengthened and shifted eastward, leading to the SE-2 pattern and the southeastern NA REHE (Fig. 9i). Correspondingly, the PWTI_SE-1 (PWTI_SE-2) shows significant positive anomalies on day −4, peaks on day −1 (day 0), and weakens rapidly afterward (Fig. 10c). After day 0, the anomalous high (low) of the SE-1 pattern shrinks (remains), dispersing wave energy northeastward to excite an anomalous high around the Bering Strait (Fig. 9e), while the SE-2 pattern still exists but moves eastward, dispersing wave energy southeastward (Fig. 9j).
d. Potential forecast usage
As shown in Fig. 10, the average values of all PWTIs have significant positive anomalies at least 4 days before the REHEs. Meanwhile, all PWTI values are positive on day −1, which indicates that at least 75% of REHEs are related to these precursor wave trains (see the bottom of the bars in Fig. 10) and the PWTIs have potential usage for the forecast of the REHEs over the region.
The potential forecast value of the pronounced precursor wave trains is preliminarily investigated here. We examine the occurrence probability of the NA REHE days associated with the precursor wave trains (shown in Figs. 7–9). Following Xu et al. (2019), we designate a precursor wave train day when the standardized PWTI on the day is greater than 1.5 standard deviations. The occurrence probability of the REHE days associated with the precursor wave train is defined as the ratio of the precursor wave train days accompanied by an REHE within the following 3 days relative to all the precursor wave train days, and the original occurrence probability of the REHE days is defined as the ratio of the REHE days relative to all summer days. Taking the WE-1 pattern as an example, we can see that there are 243 REHE days, accounting for 4.5% of the total 5428 summer days (59 years × 92 days = 5428 days); therefore, the original occurrence probability of the WE-1 pattern–related REHE days is 4.5%. The days with a PWTI_WE-1 greater than 1.5 standard deviations are further designated as the WE-1 precursor wave train days, and 40.1% of these WE-1 precursor wave train days have REHEs in their following 3 days. Therefore, the occurrence probability of the REHE days associated with the WE-1 precursor wave train is 40.1%, which is 8.9 times the original occurrence probability (4.5%). As shown in Table 4, the occurrence probability of the REHE days constrained by the precursor wave train for each synoptic pattern is 30%–45%, more than five times the original occurrence probability. The results indicate that the precursor wave trains revealed in this study could be useful for the forecast of the NA REHE occurrence.
Original occurrence probability of REHE days associated with each synoptic pattern, and occurrence probability of REHE days associated with each synoptic pattern precursor wave train.
5. Long-term changes in NA REHE days and synoptic pattern–related REHE days
Under the background of global warming, the long-term trends in different atmospheric circulation patterns have different contributions to the long-term trend of temperature extremes (Horton et al. 2015). Therefore, the long-term changes in NA REHE days and synoptic pattern–related REHE days are further investigated in this section.
a. Long-term trends
As shown in Figs. 11a–c, the REHE days in all three subregions show increasing trends over the 1960–2018 period. Among the three subregions, the REHE days in central NA experience the strongest increasing trend, while those in western NA and southeastern NA have comparable increasing trends (Figs. 11a–c). Except for the WE-2 and SE-1 patterns, the REHE days associated with all other patterns exhibit significant increasing trends over the 1960–2018 period (Figs. 11d–i). The results indicate that the significant increasing trend of the western NA (southeastern NA) REHE days mainly derives from the WE-1 (SE-2) pattern–related REHE days. Although both the increasing trends of the CE-1 and CE-2 pattern–related REHE days are significant, the CE-2 pattern–related REHE days show a larger increasing trend and contribute more to the increase in the REHE days in central NA than the CE-1 pattern–related REHE days.
Time series of REHE days (solid line) in (a) western NA, (b) central NA, and (c) southeastern NA over the 1960–2018 period; the black dashed lines are linear trends. (d)–(i) As in (a), but for the REHE days associated with the WE-1, WE-2, CE-1, CE-2, SE-1, and SE-2 patterns, respectively. The histograms in (a)–(i) represent the 5-yr total REHE days (the last bar for 4 years from 2015 to 2018). The histograms with different colors represent the REHE days occurring in different months. The long-term trends and the corresponding significance levels are shown in the top left of the panels, and the trends significant at the 95% confidence level based on Student’s t test are shown in red. (j) Difference between the percentages of the 5-yr total WE-1 and WE-2 pattern–related REHE days relative to the 5-yr total REHE days in western NA (the last bar for the 4-yr total from 2015 to 2018). The red and blue bars indicate that the percentage of the WE-1 pattern–related REHE days is higher and lower than that of the WE-2 pattern, respectively. (k),(l) As in (j), but for the CE-1 and CE-2 patterns in central NA and the SE-1 and SE-2 patterns in southeastern NA, respectively.
Citation: Journal of Climate 36, 15; 10.1175/JCLI-D-22-0205.1
In addition, the relative contribution of the two synoptic pattern–related REHE days to the REHE days in each subregion over the 1960–2018 period is further measured (Figs. 11j–l). The WE-2 pattern–related REHE days dominate the western NA REHE days before 1980, and the WE-1 pattern–related REHE days begin to play the leading role afterward (Fig. 11j). For central NA, the relative contribution of the CE-1 and CE-2 pattern–related REHE days to the REHE days exhibits an interdecadal variation feature; the CE-2 (CE-1) pattern–related REHE days contribute more to the central NA REHE days before 1985 and after 2000 (over the period of 1985–99; Fig. 11k). For southeastern NA, the SE-2 pattern–related REHE days generally dominate the REHE days, except for the periods of 1965–74 and 1990–94 (Fig. 11l).
b. Monthly distribution and monthly trend
According to the bars with different colors in Fig. 11, the synoptic pattern–related REHE days are different in June, July, and August. Therefore, we further analyze the monthly distribution and the monthly trend of the REHE days associated with these synoptic patterns, which are shown in Table 5. The WE-1 pattern–related REHE days show comparable proportions in the 3 months, while more than 80% of the WE-2 pattern–related REHE days occur in mid- to late summer, with nearly 50% of the REHE days occurring in August. Similar to the WE-2 pattern, approximately 75% of the REHE days associated with the CE-2 pattern (the CE-1 and SE-1 patterns) occur in mid- to late (early to mid) summer, while the REHE days associated with the SE-2 pattern occur in the 3 months with similar possibility.
Monthly distribution and monthly trend of synoptic pattern–related REHE days. Trends significant at 95% significant level based on Student’s t test are in boldface.
For the monthly trend, more than half of the increasing trend in the WE-1 pattern–related REHE days [1.11 days (10 years)−1; Fig. 11d] derives from the significant increasing trend in August [0.57 days (10 years)−1]. The long-term trends of both the WE-2 and SE-1 pattern–related REHE days in the 3 months of summer are not significant at the 95% confidence level, which further contributes to their insignificant increasing trends in summer (Figs. 11f,g). For central NA, the CE-1 pattern–related REHE days show a significant increasing trend only in June [0.56 days (10 years)−1], while the significant increasing trend of the CE-2 pattern–related REHE days can be examined in the mid- to late summer, with trend values larger than 0.6 days (10 years)−1. Similar to the CE-2 pattern, the REHE days associated with the SE-2 pattern show a significant increasing trend in mid- to late summer but with lower values.
c. Interdecadal variation
Hong et al. (2022) explored an interdecadal increase in the NA EHE frequency around the mid-1990s. Do the synoptic pattern–related NA REHE days have similar interdecadal increases? To answer this question, the interdecadal variation in synoptic pattern–related NA REHE days is investigated.
According to the bars in Figs. 11a–c, the REHE days in the three subregions experience similar interdecadal increases around the mid-1990s. The REHE days in western (central/southeastern) NA increase from 5.1 (5.3/4.8) days yr−1 over the 1960–94 period to 10.6 (14.8/11.6) days yr−1 over the 1997–2018 period, which are all significant at the 95% confidence level. These results indicate that the REHE days over the three NA subregions have a consistently significant interdecadal increase around the mid-1990s. Moreover, all REHE days associated with the synoptic patterns also experienced significant interdecadal increases around the mid-1990s (Figs. 11d–i), which can be confirmed by the significant interdecadal difference between REHE days over the 1960–94 and 1997–2018 periods. The results indicate that both synoptic pattern–related REHE days contribute to the interdecadal shift to more REHE days in each subregion of NA.
6. Conclusions and discussion
In this study, the synoptic atmospheric patterns responsible for REHEs over NA are investigated. First, based on the method from Yu et al. (2018, 2021), we propose a new hybrid regionalization approach using REOF and spectral clustering analyses. By applying this hybrid regionalization approach to the standardized JJA Tmax at 277 NA meteorological stations, NA can be divided into three subregions: western NA, central NA, and southeastern NA. Moreover, the atmospheric circulation patterns associated with the REHEs in each subregion are further divided into two types by spectral clustering analysis, namely, the WE-1, WE-2, CE-1, CE-2, SE-1, and SE-2 patterns.
The WE-1 and CE-1 patterns are dominated by an anomalous quasi-barotropic high over western NA and central NA, respectively, while the WE-2, CE-2, and SE-2 (SE-1) patterns are dominated by an anomalous northwest–southeast (northeast–southwest) dipole pattern over their own subregions. Therefore, the WE-1 and CE-1 patterns (the WE-2, CE-2, SE-1, and SE-2 patterns) are referred to as the monopole high pattern (the synoptic dipole pattern). In each subregion, the higher EHE occurrence probabilities associated with the two synoptic patterns also exhibit spatial differences. A higher EHE occurrence probability is located more northwestward/southeastward (northward/southward; southwestward/northeastward) over western NA (central NA; southeastern NA) for the WE-1/WE-2 (CE-1/CE-2; SE-1/SE-2) pattern–related REHE days. Furthermore, the persistence of REHEs associated with each synoptic pattern is evaluated based on regional heatwaves. The results show that regional heatwave days account for more than 50% of the REHE days for all synoptic patterns, indicating that NA REHE days mainly occur in the form of regional heatwave days. The WE-1, CE-1, CE-2, and SE-2 pattern–related REHE days exhibit the strongest persistence, with an average regional heatwave duration over 4.6 days, and the WE-2 and SE-1 pattern–related REHE days have similar persistence, with an average regional heatwave duration of approximately 4 days.
The influences of the synoptic patterns on the NA EHE frequency are further investigated. The results show that the anomalous high in each pattern is associated with more surface solar radiation reaching and warming the ground, which favors the EHE formation. In addition to the radiation process, the anomalous subsidence motions in the eastern part of the anomalous highs of the synoptic patterns heat the air and lead to EHE formation there. The horizontal temperature advection anomalies in the synoptic dipole patterns play a more important role in the EHE formation than those in the monopole patterns. Generally, the radiation process dominates the NA EHE formation for most of the synoptic patterns, but the WE-2 pattern–related horizontal temperature advection has a larger contribution to the EHE formation in the northeastern part of western NA.
The evolutions of atmospheric circulation associated with different synoptic pattern–related NA REHEs are analyzed, and some PWTIs are defined to better capture the precursory signals. The WE-1 and WE-2 pattern–related REHEs are associated with different precursor wave trains over the North Atlantic–Eurasia sector. The precursor wave train for the WE-1 (WE-2) pattern–related REHEs exhibits two anomalous highs to the south (east) of Iceland and the north (northeast) of the Aral Sea and an anomalous low over northern (eastern) Europe. The precursor wave trains further propagate eastward to western NA, leading to the WE-1 or WE-2 pattern and a western NA REHE. The precursor wave train for the CE-1 (CE-2) pattern–related REHEs originates from western Europe (the Norwegian Sea) and propagates northeastward (southeastward) through the Ural Mountains to central NA, contributing to the CE-1 (CE-2) pattern and a central NA REHE. For the SE-1 and SE-2 pattern–related REHEs, similar anomalous highs occur to the northeast of Lake Balkhash a week earlier. In the following few days, an anomalous low is enhanced to the northwest of Lake Balkhash (over northwestern NA) and combines with the aforementioned anomalous high to form the precursor zonal (meridional) dipole pattern; the precursor zonal (meridional) dipole pattern is further shifted southeastward (eastward), leading to the formation of the SE-1 (SE-2) pattern and an REHE over southeastern NA. The potential usage of these precursor wave trains in the forecast of NA REHE days is also preliminarily investigated. The results show that the occurrence probability of REHE days constrained by the precursor wave train for each synoptic pattern is 30%–45%, more than five times the original occurrence probability, indicating the possible forecast usage of the precursor wave trains for NA REHE occurrence.
Furthermore, the long-term changes in REHE days over the 1960–2018 period are studied. The REHE days in all three subregions show significant increasing trends over the whole period, and the four synoptic pattern–related REHE days (except for the WE-2 and SE-1 patterns) also exhibit significant increasing trends. Horton et al. (2015) revealed that the different long-term trends of synoptic patterns have different contributions to the long-term trends of temperature extremes. In this study, we find that the WE-1 (CE-2/SE-2) pattern–related REHE days contribute more to the increase in western (central/southeastern) NA REHE days. The relative contribution of the two synoptic pattern–related REHE days to the REHE days in each subregion over the 1960–2018 period is also measured. The results indicate that the WE-2 (WE-1) pattern–related REHE days dominate the western NA REHE days before (after) 1980; for central NA, the CE-2 (CE-1) pattern–related REHE days contribute more to the REHE days before 1985 and after 2000 (over the period of 1985–99); for southeastern NA, the SE-2 pattern–related REHE days generally dominate the REHE days, except for the periods of 1965–74 and 1990–94.
We further investigate the monthly distribution and monthly trend of the REHE days associated with the synoptic patterns. The monthly distributions of the synoptic pattern–related REHE days are different. The WE-1 and SE-2 pattern–related REHE days show a relatively uniform distribution in the 3 months of summer, while approximately 75% of the REHE days associated with the WE-2 and CE-2 (CE-1 and SE-1) patterns occur in mid- to late summer (early to midsummer). Furthermore, the increasing trend of the WE-1 (CE-1) pattern–related REHE days over the whole period mainly derives from the significant increasing trend in August (June), while that of the CE-2 and SE-2 pattern–related REHE days is contributed by the significant increasing trend in mid- to late summer. The insignificant increasing trends in the 3 months of summer lead to the insignificant increasing trend of the WE-2 and SE-1 pattern–related REHE days in the whole summer. In addition, significant interdecadal increases in REHE days of the three subregions are exhibited around the mid-1990s, and all synoptic pattern–related REHE days also experience an interdecadal shift to higher frequencies at the same time.
On the synoptic time scale, the NA REHE-related atmospheric circulations are explored in this study. Previous studies have indicated that the preceding spring precipitation deficit could also be connected to the summer heatwave intensity across Europe through a change in the soil moisture (e.g., Fischer et al. 2007a,b; Vautard et al. 2007; Quesada et al. 2012). In the future, the influence of preceding spring precipitation on the NA REHEs should also be investigated. The associated work may deepen our understanding of NA EHE formation and provide more useful information for NA EHE forecasting.
In this study, an EHE is defined only using Tmax. The compound extreme climate event defined using Tmax and daily minimum temperature (e.g., Chen and Zhai 2017; Wang et al. 2021; Xie and Zhou 2023) or temperature and humidity (e.g., Sherwood and Huber 2010; Chen et al. 2022) may bring more severe consequences and deserves to be investigated in the future.
Acknowledgments.
This study was supported by the National Natural Science Foundation of China (41991281) and Dragon 5 Cooperation 2020–24 (59376).
Data availability statement.
The datasets used in this study are freely available on the following websites: GHCN-D, https://www.ncei.noaa.gov/products/land-based-station/global-historical-climatology-network-daily; CMA, https://data.cma.cn/; and ERA5, https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5.
REFERENCES
Agel, L., M. Barlow, C. Skinner, F. Colby, and J. Cohen, 2021: Four distinct northeast US heat wave circulation patterns and associated mechanisms, trends, and electric usage. npj Climate Atmos. Sci., 4, 31, https://doi.org/10.1038/s41612-021-00186-7.
Alexander, L. V., and Coauthors, 2006: Global observed changes in daily climate extremes of temperature and precipitation. J. Geophys. Res., 111, D05109, https://doi.org/10.1029/2005JD006290.
Arkhangelskaya, S., 2016: The Siberian plague. Accessed 22 March 2022, https://www.rbth.com/longreads/siberian_plague/.
Bardin, M. Y., and T. V. Platova, 2019: Long-period variations in extreme temperature statistics in Russia as linked to the changes in large-scale atmospheric circulation and global warming. Russ. Meteor. Hydrol., 44, 791–801, https://doi.org/10.3103/S106837391912001X.
Bulygina, O. N., V. N. Razuvaev, N. N. Korshunova, and P. Y. Groisman, 2007: Climate variations and changes in extreme climate events in Russia. Environ. Res. Lett., 2, 045020, https://doi.org/10.1088/1748-9326/2/4/045020.
Caliński, T., and J. Harabasz, 1974: A dendrite method for cluster analysis. Commun. Stat., 3, 1–27, https://doi.org/10.1080/03610927408827101.
Cassou, C., L. Terray, and A. S. Phillips, 2005: Tropical Atlantic influence on European heat waves. J. Climate, 18, 2805–2811, https://doi.org/10.1175/JCLI3506.1.
Chen, H., W. He, J. Sun, and L. Chen, 2022: Increases of extreme heat–humidity days endanger future populations living in China. Environ. Res. Lett., 17, 064013, https://doi.org/10.1088/1748-9326/ac69fc.
Chen, R., and R. Lu, 2015: Comparisons of the circulation anomalies associated with extreme heat in different regions of eastern China. J. Climate, 28, 5830–5844, https://doi.org/10.1175/JCLI-D-14-00818.1.
Chen, R., Z. Wen, and R. Lu, 2016: Evolution of the circulation anomalies and the quasi-biweekly oscillations associated with extreme heat events in southern China. J. Climate, 29, 6909–6921, https://doi.org/10.1175/JCLI-D-16-0160.1.
Chen, R., Z. Wen, and R. Lu, 2019: Influences of tropical circulation and sea surface temperature anomalies on extreme heat over northeast Asia in the midsummer of 2018. Atmos. Oceanic Sci. Lett., 12, 238–245, https://doi.org/10.1080/16742834.2019.1611170.
Chen, Y., and P. Zhai, 2017: Revisiting summertime hot extremes in China during 1961–2015: Overlooked compound extremes and significant changes. Geophys. Res. Lett., 44, 5096–5103, https://doi.org/10.1002/2016GL072281.
Choi, N., M.-I. Lee, D.-H. Cha, Y.-K. Lim, and K.-M. Kim, 2020: Decadal changes in the interannual variability of heat waves in East Asia caused by atmospheric teleconnection changes. J. Climate, 33, 1505–1522, https://doi.org/10.1175/JCLI-D-19-0222.1.
Ciavarella, A., and Coauthors, 2021: Prolonged Siberian heat of 2020 almost impossible without human influence. Climatic Change, 166, 9, https://doi.org/10.1007/s10584-021-03052-w.
Degefie, D. T., E. Fleischer, O. Klemm, A. V. Soromotin, O. V. Soromotina, A. V. Tolstikov, and N. V. Abramov, 2014: Climate extremes in south western Siberia: Past and future. Stochastic Environ. Res. Risk Assess., 28, 2161–2173, https://doi.org/10.1007/s00477-014-0872-9.
Della-Marta, P., M. R. Haylock, J. Luterbacher, and H. Wanner, 2007: Doubled length of western European summer heat waves since 1880. J. Geophys. Res., 112, D15103, https://doi.org/10.1029/2007JD008510.
Deng, K., S. Yang, M. Ting, P. Zhao, and Z. Wang, 2019: Dominant modes of China summer heat waves driven by global sea surface temperature and atmospheric internal variability. J. Climate, 32, 3761–3775, https://doi.org/10.1175/JCLI-D-18-0256.1.
Deng, K., X. Jiang, C. Hu, and D. Chen, 2020: More frequent summer heat waves in southwestern China linked to the recent declining of Arctic sea ice. Environ. Res. Lett., 15, 074011, https://doi.org/10.1088/1748-9326/ab8335.
Ding, T., H. Gao, and W. Li, 2018: Extreme high-temperature event in southern China in 2016 and the possible role of cross-equatorial flows. Int. J. Climatol., 38, 3579–3594, https://doi.org/10.1002/joc.5518.
Dong, B., R. T. Sutton, W. Chen, X. Liu, R. Lu, and Y. Sun, 2016: Abrupt summer warming and changes in temperature extremes over northeast Asia since the mid-1990s: Drivers and physical processes. Adv. Atmos. Sci., 33, 1005–1023, https://doi.org/10.1007/s00376-016-5247-3.
Easterling, D. R., G. A. Meehl, C. Parmesan, S. A. Changnon, T. R. Karl, and L. O. Mearns, 2000: Climate extremes: Observations, modeling, and impacts. Science, 289, 2068–2074, https://doi.org/10.1126/science.289.5487.2068.
Engdaw, M. M., A. P. Ballinger, G. C. Hegerl, and A. K. Steiner, 2022: Changes in temperature and heat waves over Africa using observational and reanalysis data sets. Int. J. Climatol., 42, 1165–1180, https://doi.org/10.1002/joc.7295.
Erdenebat, E., and T. Sato, 2016: Recent increase in heat wave frequency around Mongolia: Role of atmospheric forcing and possible influence of soil moisture deficit. Atmos. Sci. Lett., 17, 135–140, https://doi.org/10.1002/asl.616.
Ezhova, E., and Coauthors, 2021: Climatic factors influencing the anthrax outbreak of 2016 in Siberia, Russia. EcoHealth, 18, 217–228, https://doi.org/10.1007/s10393-021-01549-5.
Fang, B., and M. Lu, 2020: Heatwave and blocking in the northeastern Asia: Occurrence, variability, and association. J. Geophys. Res. Atmos., 125, e2019JD031627, https://doi.org/10.1029/2019JD031627.
Fischer, E. M., and C. Schär, 2010: Consistent geographical patterns of changes in high-impact European heatwaves. Nat. Geosci., 3, 398–403, https://doi.org/10.1038/ngeo866.
Fischer, E. M., S. I. Seneviratne, D. Lüthi, and C. Schär, 2007a: Contribution of land–atmosphere coupling to recent European summer heat waves. Geophys. Res. Lett., 34, L06707, https://doi.org/10.1029/2006GL029068.
Fischer, E. M., S. I. Seneviratne, P. L. Vidale, D. Lüthi, and C. Schär, 2007b: Soil moisture–atmosphere interactions during the 2003 European summer heat wave. J. Climate, 20, 5081–5099, https://doi.org/10.1175/JCLI4288.1.
Gao, M., J. Yang, B. Wang, S. Zhou, D. Gong, and S.-J. Kim, 2018: How are heat waves over Yangtze River valley associated with atmospheric quasi-biweekly oscillation? Climate Dyn., 51, 4421–4437, https://doi.org/10.1007/s00382-017-3526-z.
Groisman, P. Y., and Coauthors, 2013: Climate changes in Siberia. Regional Environmental Changes in Siberia and their Global Consequences, P. Ya. Groisman and G. Gutman, Eds., Springer Environmental Science and Engineering Book Series, Springer, 57–109.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Academic Press, 535 pp.
Hong, H., J. Sun, and H. Wang, 2020: Interdecadal variation in the frequency of extreme hot events in northeast China and the possible mechanism. Atmos. Res., 244, 105065, https://doi.org/10.1016/j.atmosres.2020.105065.
Hong, H., J. Sun, and H. Wang, 2022: Variations in summer extreme high-temperature events over northern Asia and the possible mechanisms. J. Climate, 35, 335–357, https://doi.org/10.1175/JCLI-D-21-0043.1.
Horton, D. E., N. C. Johnson, D. Singh, D. L. Swain, B. Rajaratnam, and N. S. Diffenbaugh, 2015: Contribution of changes in atmospheric circulation patterns to extreme temperature trends. Nature, 522, 465–469, https://doi.org/10.1038/nature14550.
Hu, L., J.-J. Luo, G. Huang, and M. C. Wheeler, 2019: Synoptic features responsible for heat waves in central Africa, a region with strong multidecadal trends. J. Climate, 32, 7951–7970, https://doi.org/10.1175/JCLI-D-18-0807.1.
Hueffer, K., D. Drown, V. Romanovsky, and T. Hennessy, 2020: Factors contributing to anthrax outbreaks in the circumpolar north. EcoHealth, 17, 174–180, https://doi.org/10.1007/s10393-020-01474-z.
Hughes, M., A. Hall, and R. G. Fovell, 2007: Dynamical controls on the diurnal cycle of temperature in complex topography. Climate Dyn., 29, 277–292, https://doi.org/10.1007/s00382-007-0239-8.
Jiang, W., H. Chen, and Z. Shi, 2022: Anthropogenic influence on extreme temperature changes over the mid–high latitudes of Asia. Int. J. Climatol., 42, 8619–8631, https://doi.org/10.1002/joc.7753.
Kim, H.-K., B.-K. Moon, M.-K. Kim, J.-Y. Park, and Y.-K. Hyun, 2021: Three distinct atmospheric circulation patterns associated with high temperature extremes in South Korea. Sci. Rep., 11, 12911, https://doi.org/10.1038/s41598-021-92368-9.
Kuglitsch, F. G., A. Toreti, E. Xoplaki, P. M. Della-Marta, C. S. Zerefos, M. Türkeş, and J. Luterbacher, 2010: Heat wave changes in the eastern Mediterranean since 1960. Geophys. Res. Lett., 37, L04802, https://doi.org/10.1029/2009GL041841.
Lee, W.-S., and M.-I. Lee, 2016: Interannual variability of heat waves in South Korea and their connection with large-scale atmospheric circulation patterns. Int. J. Climatol., 36, 4815–4830, https://doi.org/10.1002/joc.4671.
Li, C., F. Zwiers, X. Zhang, G. Li, Y. Sun, and M. Wehner, 2021: Changes in annual extremes of daily temperature and precipitation in CMIP6 models. J. Climate, 34, 3441–3460, https://doi.org/10.1175/JCLI-D-19-1013.1.
Li, H., H. Chen, H. Wang, J. Sun, and J. Ma, 2018: Can Barents Sea ice decline in spring enhance summer hot drought events over northeastern China? J. Climate, 31, 4705–4725, https://doi.org/10.1175/JCLI-D-17-0429.1.
Li, M., D. Luo, Y. Yao, and L. Zhong, 2020a: Large-scale atmospheric circulation control of summer extreme hot events over China. Int. J. Climatol., 40, 1456–1476, https://doi.org/10.1002/joc.6279.
Li, M., Y. Yao, I. Simmonds, D. Luo, L. Zhong, and X. Chen, 2020b: Collaborative impact of the NAO and atmospheric blocking on European heatwaves, with a focus on the hot summer of 2018. Environ. Res. Lett., 15, 114003, https://doi.org/10.1088/1748-9326/aba6ad.
Li, R.-X., and J.-Q. Sun, 2018: Interdecadal variability of the large-scale extreme hot event frequency over the middle and lower reaches of the Yangtze River basin and its related atmospheric patterns. Atmos. Oceanic Sci. Lett., 11, 63–70, https://doi.org/10.1080/16742834.2017.1335580.
Liu, Q., T. Zhou, H. Mao, and C. Fu, 2019: Decadal variations in the relationship between the western Pacific subtropical high and summer heat waves in East China. J. Climate, 32, 1627–1640, https://doi.org/10.1175/JCLI-D-18-0093.1.
Loikith, P. C., and A. J. Broccoli, 2012: Characteristics of observed atmospheric circulation patterns associated with temperature extremes over North America. J. Climate, 25, 7266–7281, https://doi.org/10.1175/JCLI-D-11-00709.1.
Lorenz, E. N., 1956: Empirical orthogonal functions and statistical weather prediction. Scientific Rep. 1, 52 pp., https://eapsweb.mit.edu/sites/default/files/Empirical_Orthogonal_Functions_1956.pdf.
Lu, R.-Y., and R.-D. Chen, 2016: A review of recent studies on extreme heat in China. Atmos. Oceanic Sci. Lett., 9, 114–121, https://doi.org/10.1080/16742834.2016.1133071.
Luo, M., and N.-C. Lau, 2017: Heat waves in southern China: Synoptic behavior, long-term change, and urbanization effects. J. Climate, 30, 703–720, https://doi.org/10.1175/JCLI-D-16-0269.1.
Luo, M., and N.-C. Lau, 2019: Amplifying effect of ENSO on heat waves in China. Climate Dyn., 52, 3277–3289, https://doi.org/10.1007/s00382-018-4322-0.
Luo, M., and N.-C. Lau, 2020: Summer heat extremes in northern continents linked to developing ENSO events. Environ. Res. Lett., 15, 074042, https://doi.org/10.1088/1748-9326/ab7d07.
Meehl, G. A., and C. Tebaldi, 2004: More intense, more frequent, and longer lasting heat waves in the 21st century. Science, 305, 994–997, https://doi.org/10.1126/science.1098704.
Meng, L., and Y. Shen, 2014: On the relationship of soil moisture and extreme temperatures in East China. Earth Interact., 18, https://doi.org/10.1175/2013EI000551.1.
Menne, M. J., I. Durre, R. S. Vose, B. E. Gleason, and T. G. Houston, 2012: An overview of the Global Historical Climatology Network–Daily database. J. Atmos. Oceanic Technol., 29, 897–910, https://doi.org/10.1175/JTECH-D-11-00103.1.
Ng, A. Y., M. I. Jordan, and Y. Weiss, 2002: On spectral clustering: Analysis and an algorithm. Proc. 14th Int. Conf. on Neural Information Processing Systems: Natural and Synthetic, Vancouver, BC, Canada, Association for Computing Machinery, 849–856, https://dl.acm.org/doi/10.5555/2980539.2980649.
Noh, E., J. Kim, S.-Y. Jun, D.-H. Cha, M.-S. Park, J.-H. Kim, and H.-G. Kim, 2021: The role of the Pacific–Japan pattern in extreme heatwaves over Korea and Japan. Geophys. Res. Lett., 48, e2021GL093990, https://doi.org/10.1029/2021GL093990.
Overland, J. E., and M. Wang, 2021: The 2020 Siberian heat wave. Int. J. Climatol., 41, E2341–E2346, https://doi.org/10.1002/joc.6850.
Pedregosa, F., and Coauthors, 2011: Scikit-learn: Machine learning in Python. J. Mach. Learn. Res., 12, 2825–2830.
Perkins, S. E., 2015: A review on the scientific understanding of heatwaves—Their measurement, driving mechanisms, and changes at the global scale. Atmos. Res., 164–165, 242–267, https://doi.org/10.1016/j.atmosres.2015.05.014.
Perkins, S. E., L. V. Alexander, and J. R. Nairn, 2012: Increasing frequency, intensity and duration of observed global heatwaves and warm spells. Geophys. Res. Lett., 39, L20714, https://doi.org/10.1029/2012GL053361.
Pfahl, S., and H. Wernli, 2012: Quantifying the relevance of atmospheric blocking for co-located temperature extremes in the Northern Hemisphere on (sub-) daily time scales. Geophys. Res. Lett., 39, L12807, https://doi.org/10.1029/2012GL052261.
Quesada, B., R. Vautard, P. Yiou, M. Hirschi, and S. I. Seneviratne, 2012: Asymmetric European summer heat predictability from wet and dry southern winters and springs. Nat. Climate Change, 2, 736–741, https://doi.org/10.1038/nclimate1536.
Ren, L., T. Zhou, and W. Zhang, 2020: Attribution of the record-breaking heat event over northeast Asia in summer 2018: The role of circulation. Environ. Res. Lett., 15, 054018, https://doi.org/10.1088/1748-9326/ab8032.
Richman, M. B., 1986: Rotation of principal components. J. Climatol., 6, 293–335, https://doi.org/10.1002/joc.3370060305.
Seneviratne, S. I., and M. Hauser, 2020: Regional climate sensitivity of climate extremes in CMIP6 versus CMIP5 multimodel ensembles. Earth’s Future., 8, e2019EF001474, https://doi.org/10.1029/2019EF001474.
Sherwood, S. C., and M. Huber, 2010: An adaptability limit to climate change due to heat stress. Proc. Natl. Acad. Sci. USA, 107, 9552–9555, https://doi.org/10.1073/pnas.0913352107.
Stefanon, M., F. D’Andrea, and P. Drobinski, 2012: Heatwave classification over Europe and the Mediterranean region. Environ. Res. Lett., 7, 014023, https://doi.org/10.1088/1748-9326/7/1/014023.
Sulikowska, A., and A. Wypych, 2020: Summer temperature extremes in Europe: How does the definition affect the results? Theor. Appl. Climatol., 141, 19–30, https://doi.org/10.1007/s00704-020-03166-8.
Sun, J.-Q., 2012: Possible impact of the summer North Atlantic Oscillation on extreme hot events in China. Atmos. Oceanic Sci. Lett., 5, 231–234, https://doi.org/10.1080/16742834.2012.11446996.
Sun, J.-Q., 2014: Record-breaking SST over mid-North Atlantic and extreme high temperature over the Jianghuai–Jiangnan region of China in 2013. Chin. Sci. Bull., 59, 3465–3470, https://doi.org/10.1007/s11434-014-0425-0.
Sun, Y., X. Zhang, F. W. Zwiers, L. Song, H. Wang, T. Hu, H. Yin, and G. Ren, 2014: Rapid increase in the risk of extreme summer heat in eastern China. Nat. Climate Change, 4, 1082–1085, https://doi.org/10.1038/nclimate2410.
Sun, Y., S. Dong, T. Hu, X. Zhang, and P. Stott, 2020: Attribution of the warmest spring of 2018 in northeastern Asia using simulations of a coupled and an atmospheric model. Bull. Amer. Meteor. Soc., 101, S129–S134, https://doi.org/10.1175/BAMS-D-19-0264.1.
Takaya, K., and H. Nakamura, 2001: A formulation of a phase-independent wave-activity flux for stationary and migratory quasigeostrophic eddies on a zonally varying basic flow. J. Atmos. Sci., 58, 608–627, https://doi.org/10.1175/1520-0469(2001)058<0608:AFOAPI>2.0.CO;2.
Tang, Y., A. Huang, P. Wu, D. Huang, D. Xue, and Y. Wu, 2021: Drivers of summer extreme precipitation events over East China. Geophys. Res. Lett., 48, e2021GL093670, https://doi.org/10.1029/2021GL093670.
Thirumalai, K., P. N. DiNezio, Y. Okumura, and C. Deser, 2017: Extreme temperatures in Southeast Asia caused by El Niño and worsened by global warming. Nat. Commun., 8, 15531, https://doi.org/10.1038/ncomms15531.
Vautard, R., and Coauthors, 2007: Summertime European heat and drought waves induced by wintertime Mediterranean rainfall deficit. Geophys. Res. Lett., 34, L07711, https://doi.org/10.1029/2006GL028001.
von Luxburg, U., 2007: A tutorial on spectral clustering. Stat. Comput., 17, 395–416, https://doi.org/10.1007/s11222-007-9033-z.
Wang, J., and Coauthors, 2021: Anthropogenic emissions and urbanization increase risk of compound hot extremes in cities. Nat. Climate Change, 11, 1084–1089, https://doi.org/10.1038/s41558-021-01196-2.
Wang, K., Y. Li, Y. Wang, and X. Yang, 2017: On the asymmetry of the urban daily air temperature cycle. J. Geophys. Res. Atmos., 122, 5625–5635, https://doi.org/10.1002/2017JD026589.
Wang, W., W. Zhou, X. Wang, S. K. Fong, and K. C. Leong, 2013: Summer high temperature extremes in southeast China associated with the East Asian jet stream and circumglobal teleconnection. J. Geophys. Res. Atmos., 118, 8306–8319, https://doi.org/10.1002/jgrd.50633.
Wang, W., W. Zhou, and D. Chen, 2014: Summer high temperature extremes in southeast China: Bonding with the El Niño–Southern Oscillation and East Asian summer monsoon coupled system. J. Climate, 27, 4122–4138, https://doi.org/10.1175/JCLI-D-13-00545.1.
Wang, W., W. Zhou, X. Li, X. Wang, and D. Wang, 2016: Synoptic-scale characteristics and atmospheric controls of summer heat waves in China. Climate Dyn., 46, 2923–2941, https://doi.org/10.1007/s00382-015-2741-8.
WMO, 2017: WMO guidelines on the calculation of climate normals. WMO-1203, 29 pp., https://library.wmo.int/doc_num.php?explnum_id=4166.
WMO, 2020: Reported new record temperature of 38°C north of Arctic Circle. WMO, accessed 22 March 2022, https://public.wmo.int/en/media/news/reported-new-record-temperature-of-38°C-north-of-arctic-circle.
Wu, B., and J. A. Francis, 2019: Summer Arctic cold anomaly dynamically linked to East Asian heat waves. J. Climate, 32, 1137–1150, https://doi.org/10.1175/JCLI-D-18-0370.1.
Wu, L., and J. Zhang, 2015: The relationship between spring soil moisture and summer hot extremes over North China. Adv. Atmos. Sci., 32, 1660–1668, https://doi.org/10.1007/s00376-015-5003-0.
Xie, W., and B. Zhou, 2023: On the atmospheric background for the occurrence of three heat wave types in East China. Wea. Climate Extremes, 39, 100539, https://doi.org/10.1016/j.wace.2022.100539.
Xu, K., R. Lu, B.-J. Kim, J.-K. Park, J. Mao, J.-Y. Byon, R. Chen, and E.-B. Kim, 2019: Large-scale circulation anomalies associated with extreme heat in South Korea and southern–central Japan. J. Climate, 32, 2747–2759, https://doi.org/10.1175/JCLI-D-18-0485.1.
Xu, P., and Coauthors, 2021: Amplified waveguide teleconnections along the polar front jet favor summer temperature extremes over northern Eurasia. Geophys. Res. Lett., 48, e2021GL093735, https://doi.org/10.1029/2021GL093735.
Yeo, S.-R., S.-W. Yeh, and W.-S. Lee, 2019: Two types of heat wave in Korea associated with atmospheric circulation pattern. J. Geophys. Res. Atmos., 124, 7498–7511, https://doi.org/10.1029/2018JD030170.
Yu, Y., Q. Shao, and Z. Lin, 2018: Regionalization study of maximum daily temperature based on grid data by an objective hybrid clustering approach. J. Hydrol., 564, 149–163, https://doi.org/10.1016/j.jhydrol.2018.07.007.
Yu, Y., Q. Shao, Z. Lin, and I.-S. Kang, 2021: Characteristics analysis and synoptic features of event-based regional heatwaves over China. J. Geophys. Res. Atmos., 126, e2020JD033865, https://doi.org/10.1029/2020JD033865.
Zhai, P., and X. Pan, 2003: Trends in temperature extremes during 1951–1999 in China. Geophys. Res. Lett., 30, 1923, https://doi.org/10.1029/2003GL018004.
Zhang, G., G. Zeng, C. Li, and X. Yang, 2020: Impact of PDO and AMO on interdecadal variability in extreme high temperatures in North China over the most recent 40-year period. Climate Dyn., 54, 3003–3020, https://doi.org/10.1007/s00382-020-05155-z.
Zhang, J., and L. Wu, 2011: Land–atmosphere coupling amplifies hot extremes over China. Chin. Sci. Bull., 56, 3328–3332, https://doi.org/10.1007/s11434-011-4628-3.
Zhang, R., C. Sun, and W. Li, 2018: Relationship between the interannual variations of Arctic sea ice and summer Eurasian teleconnection and associated influence on summer precipitation over China (in Chinese). Chin. J. Geophys., 61, 91–105, https://doi.org/10.6038/cjg2018K0755.
Zhang, X., G. Hegerl, F. W. Zwiers, and J. Kenyon, 2005: Avoiding inhomogeneity in percentile-based indices of temperature extremes. J. Climate, 18, 1641–1651, https://doi.org/10.1175/JCLI3366.1.
Zhou, X., and Coauthors, 2019: Climatic characteristics and major meteorological events over China in 2018 (in Chinese). Meteor. Mon., 45, 543–552.
Zhou, Y., and Z. Wu, 2016: Possible impacts of mega–El Niño/Southern Oscillation and Atlantic multidecadal oscillation on Eurasian heatwave frequency variability. Quart. J. Roy. Meteor. Soc., 142, 1647–1661, https://doi.org/10.1002/qj.2759.
Zhu, B., B. Sun, and H. Wang, 2020a: Dominant modes of interannual variability of extreme high-temperature events in eastern China during summer and associated mechanisms. Int. J. Climatol., 40, 841–857, https://doi.org/10.1002/joc.6242.
Zhu, B., B. Sun, H. Li, and H. Wang, 2020b: Interdecadal variations in extreme high–temperature events over southern China in the early 2000s and the influence of the Pacific decadal oscillation. Atmosphere, 11, 829, https://doi.org/10.3390/atmos11080829.
