In the context of global warming, among the major meteorological disasters globally, extreme high temperature and drought events occur frequently in summer. The Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (AR5) points out that the extreme high temperature in most of the world’s land areas has increased significantly, and the occurrence frequency of heat waves in Europe, Australia, and most parts of Asia has shown an increasing trend (IPCC, 2013). Since the beginning of the 21st century, there frequently occurred extreme high temperature events such as the 2003 heat wave in Europe (Beniston, 2004), the 2009 heat wave in Australia (Pezza et al., 2012), the 2010 heat wave in Russia (Barriopedro et al., 2011; Dole et al., 2011; Trenberth and Fasullo, 2012), and the 2011 heat wave in the USA (Hoerling et al., 2013). These high temperature events not only led to huge economic losses, but also caused severe casualties. For example, the 2003 European heat wave resulted in up to 70,000 deaths (Robine et al., 2008).

China is prone to extremely high temperature events (Zhai and Pan, 2003). In recent years, high temperature disasters have caused huge economic losses in China and aroused widespread attention. The 2013 record-breaking extreme high temperature in South China (Sun, 2014; Zou et al., 2015; Peng et al., 2016; Yang and Feng, 2016) and the high temperature episode that abruptly occurred in the middle and lower reaches of Yangtze River in 2016 right after the Meiyu season (Ding et al., 2018) are two good examples of high temperature disasters in China. A previous study (Qin et al., 2015) indicated that among the eight major meteorological disasters in China, the future risk of high temperature events and heat waves is high in all regions.

So far, the studies on summer temperature are mostly concentrated on high temperature and heat waves. There exists a close relationship between summertime mean temperature and high temperature. An overall high temperature is conducive to extreme high temperature and heat waves, which may easily reach their thresholds. Meanwhile, extreme high temperature and heat waves make critical contributions to above normal seasonal mean temperature. Note that high temperature and heat waves only occur during certain periods of the summer. Even for the strong, record-breaking extreme high temperature event in 2013 in China, the number of high temperature days was smaller than half of the number of summer days. Therefore, it has great implications to study the temperature itself and the influences of circulation systems.

Previous studies have revealed three leading modes of summertime temperature change in China (Zhang et al., 2017). The first leading mode displays a uniform cold or warm pattern over entire China; the second leading mode shows a tripolar pattern with out-of-phase changes in northern China, southern China, and central China; and the third leading mode displays a dipole pattern with anti-phase changes between the north and the south with their boundary located along the Yangtze River (Zhang et al., 2017). The uniform cold (warm) pattern over entire China is mainly affected by negative (positive) geopotential height anomalies at the upper levels over East Asia, especially the intensity of the continental high (Zhang et al., 2017). The second leading mode is related to the East Asia–Pacific (EAP) teleconnection pattern and associated wave trains (Zhang et al., 2017). The third leading mode is correlated with the Northeast China cold vortex (NECV) and the western Pacific subtropical high (WPSH) (Zhang et al., 2017). For southern China, the WPSH is the most direct impact factor for high temperature. Descending motions prevail over areas controlled by the WPSH, leading to persistent high temperature and little precipitation (Zhang et al., 2017). Numerous statistical analyses and case studies have confirmed the above results (Zhang et al., 2014; Peng et al., 2016; Yang and Feng, 2016). Due to the opposite zonal oscillation of the South Asian high (SAH) and the WPSH (Ding, 2013), i.e., the eastward stretch of the SAH at the upper level is often accompanied by the westward extension of the WPSH, the SAH also has impacts on the maintenance of high temperature in southern China (Zhang et al., 2014; Li et al., 2015; Zou et al., 2015). In addition, as the upstream synoptic systems of the East Asian summer monsoon, the Somali and the Bay of Bengal cross-equatorial flows can also affect the duration and intensity of high temperature in southern China (Ding et al., 2018). However, due to regional differences in summertime temperature and high temperature distribution, most studies have focused on Huai River valley and its south, whereas few studies have examined the high temperature in the north.

The atmospheric circulation systems that affect summertime temperature and high temperature in China are influenced by various external forcing factors, among which sea surface temperature anomalies (SSTAs) in different oceanic areas play a critical role. As the strongest interannual variability signal in the tropics, the warm phase of ENSO can lead to a stronger than normal WPSH that is located further westward, resulting in higher temperature and less precipitation in China. The high temperature in 2016 is a good example induced by the ENSO warm phase (Ding et al., 2018). Meanwhile, SSTAs in the tropical Indian Ocean can subsequently affect the position and intensity of the WPSH at a later stage (Yuan et al., 2017). In addition, offshore SSTAs can also affect summertime temperature anomalies. For example, the warm pool in the tropical western Pacific is a key region that affects summertime temperature in Northeast China (Zheng and Ni, 1999). In the summer and early autumn when SSTAs are abnormally warm in the Kuroshio, geopotential height at 500 hPa is correspondingly higher than normal from North China to the east of Japan, which weakens the East Asian trough, leading to high temperature in northern China (Yang et al., 2013). The SSTA in North Atlantic also plays an important role in the formation of the 2013 extreme high temperature (Sun, 2014).

In addition to the above natural systems, human activities and urbanization can also affect high temperature events. Based on analysis of 31 CMIP5 models simulations, Zhou et al. (2014) found that anthropogenic contribution accounts for 47% of the warming during the 2013 heat wave event that occurred in July–August in southern China.Sun et al. (2016) compared the scenarios with and without anthropogenic contribution based on CMIP5, and found that anthropogenic activities made great contributions to the 2015 extreme high temperature event in western China. In addition, a positive feedback loop exists between urban heating and the heat wave intensity, and thus the urbanization effect actually results in more frequent occurrence of heat waves in urban areas (Bian et al., 2015; Wang et al., 2017).

The summer of 2018 is the hottest, based on all meteorological records available after the establishment of the People’s Republic of China. The maximum, mean, and minimum temperatures are all record breaking in 2018. According to the monitoring by the National Climate Center of China, temperature in 2018 is more than 1°C higher than normal in 19 provinces (cities and autonomous regions) of China, while the temperatures in 11 out of the 19 provinces are ranked among top three in their respective historical records. The temperature of 2018 in China has reached the threshold of an extreme event (exceeding the 95th percentile) at 197 stations, among which 55 stations in northern China register record-breaking high temperature. Different to the high temperature events in 2013 and 2016, the 2018 high temperature mainly occurred in northern China, and its influencing systems and external forcing signals are also quite different.

Most of the previous studies of typical high temperature events are focused on the westward and southward shift of the WPSH, which controls more of southern China, leading to the formation and development of high temperature there. Due to the differences in summertime temperature in different regions, high temperature events are fewer in northern China than in southern China. Consequently, studies of high temperature events in northern China are insufficient. Therefore, it is necessary to analyze temperature anomalies in northern China in summer 2018 and explore the circulation systems and external forcing factors that affect the abnormal high temperature. The results will be helpful to understand the mechanism for high temperature in China (especially northern China) and provide information for short-term operational climate prediction.

Daily temperature data are extracted from China National Surface Weather Stations Basic Meteorological Observation Dataset (V3.0), released by the National Meteorological Information Center of China (Ren et al., 2012). Daily maximum, mean, and minimum temperatures over the period of 1951–2018 are utilized for the present study. This dataset has gone through strict data quality check and the inconsistency check between archived datasets at national level and provincial levels. Compared to the previous generation datasets of precipitation observations, both the quality and spatial resolution (number of stations) of the new dataset have been greatly improved. The new dataset has been widely applied to operational meteorological services and scienti-fic studies. Since the number of stations and records in the 1950s are much less than in the later period, the data since 1961 are used in the present study.

The atmospheric circulation data are extracted from the NCEP/NCAR daily reanalysis product, including geopotential height and temperature at various vertical levels and wind fields at 850 and 500 hPa. The horizon-tal resolution of the NCEP/NCAR reanalysis is 2.5° × 2.5° (Kalnay et al., 1996; Kistler et al., 2001). The SST data are from the Hadley Centre monthly mean sea surface temperature (HadISST) dataset, which is on global 1° × 1° grids. The original HadISST data starts from 1870 (Rayner et al., 2003), and the data over 1961–2018 are extracted for the present study. Both air temperature and SST have significantly increased since the 1980s; therefore, unless specifically explained, the linear trends of air temperature and sea surface temperature are removed before calculating the correlation.T-test is applied for significance test of the correlation (Huang, 2004; Wei, 2007). In order to avoid the impact of the special case in 2018, correlation is computed over the period of 1981–2017, and the correlation coefficients corresponding to 0.01, 0.05, and 0.1 significance levels are 0.42, 0.32, and 0.27, respectively.

The present study focuses on summertime temperature, which refers to seasonal averages of daily mean, maximum, and minimum temperatures during June–August. The above variables are calculated to quantify the overall temperature in summer. Heat wave refers to a high temperature event that persists over certain periods, with the temperature reaching a specified threshold. The World Meteorological Organization (WMO) suggests that if daily temperature is 5°C higher than the corresponding climatological value for 5 consecutive days, a heat wave can be identified (Frich et al., 2002). In China, daily maximum temperature exceeding 35°C is used as the threshold to identify high temperature.

In order to highlight the feature of record-breaking temperature in summer 2018, time series of summer maximum, mean, and minimum temperatures weighted by areal size since 1961 are displayed in Fig. 1. A distinct decadal change in the 1990s is observed. During 1961–90, almost all the summertime maximum, mean, and minimum temperatures were below the corresponding climatological values except for a few years. Furthermore, only the minimum temperature exhibited an increasing trend during the above period, while the maxi-mum and mean temperatures only showed interannual variability. However, since the 1990s, all the above three temperatures have shown significant increasing trend. The warming trend of summertime temperature in China started later than that of the wintertime temperature (Ren et al., 2005), and basically synchronized with the global warming in the 1990s (Blunden et al., 2018). In addition to the decadal change, summertime temperature in China also exhibited pronounced interannual variability, which became more distinct since the 1990s. During 1961–90, the largest interannual differences in maximum, mean, and minimum temperatures were 1.62, 1.50, and 1.51°C and the root mean square errors (RMSEs) were 0.31, 0.28, and 0.31°C, respectively. In contrast, during the period 1991–2018, the largest interannual differences of the above temperatures have increased to 1.91, 1.78, and 2.21°C, and the RMSEs have increased to 0.47, 0.44, and 0.50°C, respectively.

Figure 1 Time series of summer maximum (red), mean (black), and minimum (blue) temperatures (°C) averaged over China during 1961–2018, with dashed lines for climate mean and dotted lines for the linear trend during 1961–90 and 1991–2018, respectively.

Figure options

In summer 2018, the averaged maximum temperature over China was 27.9°C, which, together with that in 2006, is ranked the highest since 1961 (it is also the highest since 1951, when meteorological records became available over large areas of China). The mean and minimum temperatures in summer 2018 were 21.9 and 17.2°C, respectively, which are also ranked the highest since 1961 (Fig. 1). The normalized values of the above three temperatures in summer 2018 are 1.88, 2.10, and 2.33. According to the WMO definition of threshold for anomaly (exceeding two standard deviations), the maximum, mean, and minimum temperatures in summer 2018 are all close to or have reached the standard of being abnormally high.

The record-breaking temperature in summer 2018 is also reflected in the spatial distribution of uniformly positive temperature anomalies over entire China (Fig. 2). Except for southern South China, where temperature was slightly lower than normal due to persistent rainfall, more rainy days, and less incident solar radiation, all other areas of China experienced a warmer than normal summer in 2018. Temperature was about 1°C higher than normal over most of the areas to the north of the Yangtze River. For the summertime mean temperature, positive anomaly above 1°C indicates that the summer is significantly warmer than normal. For example, southern China suffered a rare and sustained high temperature event in 2013 (Sun, 2014; Peng et al., 2016), but the summertime mean temperature anomaly in 2013 was only around 1°C over the high temperature area (Gong et al., 2014; Ke et al., 2014). Figure 2 also shows the normalized mean temperature at all stations. In eastern China except Northeast China, temperatures at most of the stations to the north of the Yangtze River were higher than their climatological values by more than two standard deviations. Particularly, in North China, the Yellow River valley, and the Yangtze–Huai River valley, the mean temperature was higher than normal by over three standard deviations. According to the monitoring by the National Climate Center of China, temperature was record breaking at 84 stations in China in summer 2018, and most of these stations are located in southern Northeast China and North China. Among the 84 stations, temperatures at Yuanshi station (station ID 53791, 43°C) and Wuji station (station ID 53699, 41.8°C) in Hebei Province ranked the first and third highest. This indicates that higher than normal temperature in North China made the most important contribution to the record-breaking temperature in whole China. Extreme high temperature is often accompanied with drought, and the two interact with each other. In summer 2018, the high temperature started in middle July (Zhang and Sun, 2018). In early July before the start of the high temperature, precipitation was obviously lower than normal in southern Northeast China, eastern North China, Yangtze–Huai River valley, etc. In summer, severe persistent drought occurred in Liaoning Province, while medium to severe drought occurred over North China and the Yangtze–Huai River valley during various periods in July and August.

Figure 2 Distribution of the mean temperature anomalies (shading; °C) in China in summer 2018, with big and small red dots for the values above 3 and 2–3 times of standard deviations, respectively.

Figure options

Daily mean temperature over entire China and individual subregions in 2018 is displayed in Fig. 3. It is seen that the mean temperature anomaly over China remained positive during almost the entire summer. Temperature anomaly higher than 1°C mainly occurred over two pe-riods: 6–7 and 23–30 June. Starting from 11 July, mean temperature anomaly remained positive over the entire China, and higher than 1°C anomaly persisted over 11 July–15 August. Furthermore, temperature anomaly higher than 2°C occurred in 3 consecutive days during mid July. In order to compare intraseasonal variability of temperature between subregions and entire country, China is divided into seven subregions (denoted by dashed lines inFig. 3), i.e., Xinjiang (XJ), Tibet, west of northern China (WNC), east of northern China (ENC), west of southern China (WSC), east of southern China (ESC), and northeastern China (NEC). Among the seven subregions, the highest correlation coefficient of 0.84 is found between daily mean temperatures in ENC and entire China. Intraseasonal variability of temperature in the above two regions are also quite similar. Temperature was lower than normal only for mid to late June and mid July, whereas it was obviously higher than normal in all other periods. The characteristic temperature variability in WNC is also similar to that over entire China (with a correlation coefficient of 0.72). Persistent positive temperature anomalies maintained from mid and late July to August. Correlations of temperature in Tibet, WSC, and ESC with temperature over entire China are around 0.3, while the periods with lower than normal temperature in the above regions are more than those in WNC and ENC. Although the correlation coefficient between temperatures in NEC and entire China is only 0.3, positive temperature anomalies higher than 4°C occurred in NEC over multiple consecutive days in late July, consistent with the high temperature period over entire China. In contrast, temperature in Xinjiang was basically independent of temperature in entire China, indicating large differences in temperature and associated circulation systems between Xinjiang and eastern China. The high correlation between temperatures in the two subregions in northern China and that over entire China again demonstrates that the high temperature in the north of the Yangtze River made the largest contribution to the positive temperature anomalies in China in summer 2018.

Figure 3 Daily variation of mean temperature anomalies (°C) averaged over China and in different regions during summer 2018. Red (blue) bars are for positive (negative) anomalies. China is divided into seven subregions denoted by dashed lines, i.e., Xinjiang (XJ), Tibet, west of northern China (WNC), east of northern China (ENC), west of southern China (WSC), east of southern China (ESC), and northeastern China (NEC).

Figure options

The abnormally high temperature in northern China made a major contribution to the record-breaking high temperature in summer 2018 in China. In order to statistically verify this conclusion, the linear correlation distribution between summertime mean temperature over China and temperatures at individual stations during 1981–2017 is displayed inFig. 4. It is shown that the correlation coefficient is greater than 0.42 (significant at the 0.01 level) in southern Northeast China and from North China to the middle reaches of the Yellow River as well as from the Huang–Huai River valley to the middle and lower reaches of the Yangtze River; however, temperature changes in northern Northeast China, South China, and Southwest China are less consistent with the mean temperature change over entire China and there basically exists only weak negative correlation or no correlation between them. The EOF analysis of monthly mean temperature in summer reveals that the first leading mode demonstrates a uniformly consistent pattern over entire China, while the spatial distribution indicates that abnormally large values are mainly located in northern China (Zhang et al., 2017). This result further confirms that on seasonal scale, temperature in northern China dominates the temperature anomaly over entire China.

Figure 4 Distribution of the correlation coefficient of summer temperature at each station with the whole China mean during 1981–2017.

Figure options

Positive surface air temperature anomalies in summer are closely related to temperature anomalies in the middle and lower troposphere. Composite temperature of multiple heat waves that have occurred in eastern China in summer shows that the area of positive temperature anomaly could extend upward from the surface to 250 hPa and tilt northward with increasing height (Qian and Ding, 2012). A positive temperature anomaly center with the value of 3°C was located from the surface to 850 hPa above the area affected by the heat wave, while another positive anomaly center with the value of 4°C was located between 400 and 250 hPa (Ding and Qian, 2012; Qian and Ding, 2012). However, large temperature differences can be found between various levels since temperature decreases with increasing height. In order to better represent the vertical distribution feature of the abnormal temperature in summer 2018, the latitude–height cross-section of normalized temperature averaged over 100°–120°E is displayed inFig. 5. It can be seen that tropospheric temperature from the surface up to 200 hPa was abnormally higher than normal, with two centers of large anomaly, over large areas, in summer 2018. One center was located in the area of 12.5°–22.5°N, 400–300 hPa, where the anomaly exceeded two standard devia-tions. However, temperature anomalies in the middle and lower troposphere below 600 hPa over the same zonal band were all negative. The other positive anomaly center appeared in the middle troposphere within 600–400 hPa and 40°–45°N, where the positive anomaly was larger than two and a half standard deviations.

Figure 5 Latitude–height profile of normalized air temperature averaged over 100°–120°E in summer 2018.

Figure options

By comparing the high anomaly centers at various levels, it is found that the positive center tilted obviously northward with increasing height. This result is consistent with results of previous studies (Ding and Qian, 2012; Qian and Ding, 2012). The upward extension and northward tilt of positive temperature anomaly from the surface up to 200 hPa is related to climatological features of the WPSH ridge line. Wen and He (2002) showed that the climatological feature of the WPSH ridge line, which tilts northward with increasing height, is affected by spatially inhomogeneous heating. Solar radiation is the major driving force for multi-year mean WPSH activity, leading to its seasonal north–south move-ment. The spatially inhomogeneous heating can result in non-uniform movement of the WPSH, and unequal velocity of air motion at the upper and lower levels may change the tilt of the ridge line (Wen and He, 2002). The impact of inhomogeneous heating on the WPSH is largely reflected on the vertical wind shear. Although the heating rate on the northern side of the WPSH is smaller than that on the southern side, its impact is more significant, leading to the northward tilt of the WPSH ridge line (Wen and He, 2002). In 2018, for northern area of eastern China, positive temperature anomalies occurred from the surface to the top of the troposphere, with the high anomaly center located in the middle troposphere (see Fig. 5), suggesting that the surface temperature anomalies in summer 2018 were directly linked with circulation anomalies in the middle troposphere.

Previous studies have pointed out that positive temperature anomalies in summer in China (especially in northern China) correspond to positive geopotential height anomalies from the lower troposphere to the lower stratosphere (Qian and Ding, 2012). The positive geopotential height anomalies that control most areas of China correspond to a deep barotropic system, and the continental high-pressure system is significant (Zhang et al., 2017). The 500-hPa circulation pattern corresponding to persistent high temperature in summer in China can be classified into four types: the northwestern warm high pressure, the continental warm high pressure, the northern type of WPSH, and the southern type of WPSH (Zhang, 2011). The WPSH and the continental high pressure are two major systems that affect the summertime high temperature in eastern China (Zhang et al., 2005). In addition to the WPSH, persistent NECV activities can also to a certain degree affect summertime temperature in China. In the summers with frequent cold vortex activities, cen-tral and southern Northeast China and most of North China are prone to abnormally low temperature, leading to the so-called “cold summer” (Hu et al., 2011). The above studies have shown that the WPSH, the continen-tal high-pressure system, and the NECV have critical impacts on summertime temperature in eastern China, especially the northern area of eastern China.

In order to better represent the degree of geopotential height anomaly in summer 2018, spatial distribution of normalized 500-hPa geopotential height and the 5880-gpm contours that denote the climatological range of the WPSH and its range in 2018 are displayed in Fig. 6, which are overlapped with 850-hPa wind anomalies and zero vorticity contours in Northeast China. Figure 6 shows that the area of the WPSH was obviously larger than normal, while its westward ridge point was more eastward than normal and the ridge line shifted further northward than normal, in 2018. The entire East Asia to the north of 30°N was under the control of positive 500-hPa geopotential height anomaly, and the anomaly could exceed two standard deviations over large areas of the middle reaches of the Yellow River–North China–southern Northeast China–the Sea of Japan. Meanwhile, the NECV center at 500 hPa was located significantly to the north of its climatological position, while the cyclonic circulation of wind anomalies at 850 hPa was located at around 55°N. The climatological zero vorticity contour is located near 35°N, while it was near 45°N in summer 2018, suggesting that the cold vortex activities were weaker and shifted to the north than normal. As a result, positive geopotential height anomaly prevailed from northern China to the Sea of Japan, which was also the circulation pattern directly conductive to high temperature to the north of the Yangtze River in summer.

Figure 6 Distribution of normalized 500-hPa geopotential height (thin contours), 5880-gpm contours (thick black lines), and 850-hPa wind anomalies (vectors) in summer 2018. The thick red lines denote the southern boundaries (defined as the zero vorticity line) of the Northeast China cold vortex. Note that the thick solid lines are for 2018 and the thick dashed ones are for the climate mean. Letters A and C represent anomalous low-level anticyclone and cyclone, respectively.

Figure options

The circulation feature for the 2018 high temperature is quite different to that for the 2013 and 2016 high temperatures. During the high temperature period in 2013, the WPSH was stronger and shifted more westward than normal, which stably controlled the middle and lower reaches of the Yangtze River; meanwhile, the South Asian high extended eastward. The above two anomalous systems formed a typical circulation pattern conductive to high temperature in southern China (Peng et al., 2016). The weakened cross-equatorial flows and zonal westerlies in the tropical Indian Ocean made great contributions to the intensification and westward expansion of the WPSH in 2016 (Ding et al., 2018). In summer 2018, the WPSH abnormally extended northward by more than three standard deviations, and the positive geopotential height anomalies controlled the area form northern China to the Sea of Japan, leading to abnormally high temperature in northern China.

On the study of the summertime circulation anomaly, the WPSH anomaly should be analyzed first. The above discussion shows that in summer 2018, the westward ridge point of the WPSH was more eastward than normal and the ridge line shifted further northward. On seasonal-scale climatology (figure omitted), the area covered by the 5880-gpm contour never exceeded 35°N, whereas the intraseasonal variations of both the WPSH location and intensity were significant. Figure 7a displays temporal variability of 500-hPa geopotential height in summer 2018 averaged over 110°–130°E. Large variations in the location and intensity of the WPSH could be found in June. In early June, the WPSH was stronger than normal with its ridge line slightly more northward; in mid June, the WPSH weakened and retreated southward; in late June, the WPSH intensified again and advanced northward, with the center intensity greater than 5920 gpm. From the end of June to early July, the WPSH distinctly weakened. However, during 7 July–18 August, the WPSH remained stronger and located further north than normal, while the temperature anomaly in northern China was larger than 2°C during the same pe-riod. The WPSH intensity exceeded 5920 gpm during the periods of 14–21 July and 29 July–5 August. The WPSH remained at the northernmost position from the end of July to early August, and its ridge line was located to the north of 40°N. Correspondingly, temperature anomaly in Northeast China exceeded 4°C and close to 3°C in ENC (Fig. 3). Since late August, the WPSH retreated southward, while it remained stronger than the climatological mean and its ridge line was still located further north than normal. As a result, abnormal high temperature maintained in northern China, although its intensity became weak.

Figure 7 Time–latitude profiles of (a) 500-hPa geopotential height (gpm; shading for 2018, black lines for the climate mean of 5860- and 5880-gpm contours) and (b) vorticity (10−6 m s−2; shading for positive) averaged over 110°–130°E during summer 2018.

Figure options

In addition to the anomalous WPSH features in July–August of 2018, the NECV also exhibited weaker than normal features and shifted further northward. Daily evolution of 500-hPa vorticity averaged along 110°–130°E is shown inFig. 7b, which reflects the seasonal variability of the NECV. From early July to the end of August, the vorticity was negative over 30°–45°N, indicating that the NECV was inactive in this period. Especially over 30°–50°N, the vorticity reached the minimum value of −3 × 10⁻⁵ m s⁻² from the end of July to early August. This period corresponded to the period when the WPSH reached its peak intensity and the ridge line was located at its northernmost position (Fig. 7a). The persistently intensifying WPSH and weakening NECV directly resulted in the unprecedent high temperature that occurred in the end of July and early August in northern China, especially in North China and Northeast China. Therefore, the persistent strong and northward shifted WPSH and inactive NECV were two major reasons that directly led to the abnormal high temperature in China in summer 2018.

Are there any possible external forcing factors that affect the abnormal high temperature in summer 2018? To answer this question, the correlation between mean summertime temperature in China and global mean SSTA is calculated. Note that the linear warming trend is excluded in advance from the mean summertime temperature and global SSTA. The result is displayed in Fig. 8, which shows a zonal area of significant positive correlation over the Kuroshio and its extension to Northwest Pacific. This area is also named as the Kuroshio Extension in many previous studies (Li and Sun, 2004; Wang et al., 2012; Qiu et al., 2017). Nonaka et al. (2006) proposed that the SST in the Kuroshio Extension area has been cooling since the 1980s. For the purpose to avoid the impact of decadal temperature change, the time series since 1981 are used for analysis in the present study. The spatial distribution of correlation indicates that when SST is abnormally high in the Kuroshio Extension, summertime mean temperature would also be abnormally high in China. In summer 2018, SST was higher than normal in most areas of North Pacific, especially over the Kuro-shio Extension, where a positive SSTA center was observed, with values larger than 1.5°C (Fig. 9a). This area is basically coincident with the area of significant positive correlation shown in Fig. 8. Based on previous studies, SSTA averaged over the area denoted by the dashed black box in Fig. 8 is defined as the Kuroshio and Kuro-shio Extension (K–KE) index, i.e., the SSTA averaged over 25°–45°N, 120°E–180°. During 1961–2018, the K–KE index demonstrates significant interannual variability with a pronounced linear warming trend since late 1990s (Fig. 9b). Such an interannual variability pattern of the K–KE index is highly consistent with the interannual variability of summertime mean temperature in China (shown inFig. 1). The correlation coefficient between the two is up to 0.75 for the period 1981–2017, and the value can still reach 0.53 after the linear trends of the two series are removed. The correlation is significant at the 0.01 level. Similarly, the correlation coefficient between the K–KE index and summertime mean temperature over China is also calculated for the period of 1961–2017. The correlation coefficients are 0.65 and 0.68 before and after the linear trend is removed. The 11-yr window sliding correlation coefficient is also calculated, and the result indicates that the two are significantly positively correlated for most of the time during 1961–2017. Since the correlations during the two periods lead to similar conclusions for the present study, hereafter only the data over the period 1981–2017 are used for correlation calculation. In summer 2018, the K–KE index was 23.3°C, which exceeded the normal value by about 1.7 positive standard deviation and ranked top of its historical records (together with the value in 1961) (Fig. 9b). By combining the correlation and SSTA, it can be concluded that the anomalous warming over the K–KE area has pronounced impacts on the northward displacement of the WPSH and persistent positive summertime temperature anomaly in China.

Figure 8 Distribution of the correlation coefficient between the time series of summer temperature averaged in China and the global sea surface temperature anomalies (SSTA). The black dashed box denotes the region over which the K–KE index is defined.

Figure options

Figure 9 (a) SSTA (°C) distribution in summer 2018 and (b) time series of normalized K–KE index in summer during 1961–2018. The black dashed box in (a) denotes the region over which the K–KE index is defined.

Figure options

The correlation between the SSTA in K–KE and temperature station observations in China is also calculated. Before the linear trend is removed, the K–KE index is significantly positively correlated with surface air temperature in most areas of central and eastern China except a few stations in western South China. In particular, the correlation coefficient is significant at the 0.05 level in the middle and lower reaches of the Yangtze River and to its north (Fig. 10a). Since the significance can possibly be affected by the decadal warming, the correlation is recalculated after the linear warming trend is removed. The result still shows significant positive correlation between the K–KE index and the temperature over the Yangtze–Huai River valley, North China, and part of Northwest China (Fig. 10b). The above results indicate that the SSTA in the K–KE area exerted distinct impacts on the 2018 high temperature in northern China.

Figure 10 Distributions of the correlation coefficient of summer temperature at each station with the summer K–KE index during 1981–2017: (a) original value and (b) detrended value.

Figure options

In terms of atmospheric circulation, earlier analyses in this paper have shown that the abnormal northward shift of the WPSH and the weaker than normal NCEV are two reasons that directly resulted in higher than normal geopotential height from northern China to the Sea of Japan and anomalously high temperature in northern China. In terms of external forcing factors, we now further analyze the mechanism behind the influences of positive SSTA over the K–KE area on crucial circulation systems and high temperature anomaly in northern China. Due to the large heat capacity of the sea water in the Kuroshio area and the east–west oriented oceanic front over the Kuro-shio Extension, the K–KE area is the key area of ocean–atmosphere interaction (Wang et al., 2016). The SSTA over the Kuroshio in the previous winter can simultaneously affect the NECV and the WPSH, leading to cyclonic circulation anomaly at upper levels (Wang et al., 2016). The easterly wind anomalies to WPSH’s north-west are favorable for the formation of anticyclonic anomaly advection, which weakens the NECV; meanwhile, the southwesterly wind anomalies to WPSH’s southwest can promote intensification of the anticyclonic circulation in the upper troposphere, leading to strengthening of the WPSH (Gao and Gao, 2014). Numerical experiments have shown that when the Kuroshio Extension becomes abnormally warm in winter, the anomalous anticyclonic circulation over Northeast China would be stronger than normal and the WPSH overall would intensify, particularly in late July and middle and late August (Wang et al., 2011). Correspondingly, except for a small area in the upper troposphere and lower stratosphere, where temperature decreases, all other areas to the south of 45°N experience uniform temperature increases, with large increases concentrated over 35°–45°N from North China to Northeast China (Wang et al., 2011). The case study and numerical experiment of Wang et al. (2016) revealed that the warm SSTA in summer 2010 was located over the Kuroshio Extension. They also found that upward propagating wave activity fluxes induced by pronounced heating in the lower troposphere led to abnormally strong WPSH and its northward displacement along the coast of East Asia.

The above analyses indicate that the WPSH being persistently shifted northward and the weaker than normal NECV are the two major circulation systems responsible for the high temperature in northern China in summer 2018. Correlations of the mean summertime temperature in China and that in northern China with 500-hPa geopotential height both show significant positive values over most areas to the north of the Yangtze River to Japan (Figs. 11a, b), suggesting that positive geopotential height anomalies in the above areas would lead to higher than normal temperature in most areas of China, especially northern China. Meanwhile, significant positive correlation is also found between the K–KE index and 500-hPa geopotential height over the region of North China–Northeast China–Japan (Fig. 11c), which is slightly eastward than that in Figs. 11a, b. This indicates that the anomalously high temperature over K–KE is favorable for formation of the positive 500-hPa geopotential height anomaly over North China–Northeast China–Japan, which subsequently leads to higher than normal temperature in northern China (Fig. 10b). The time–latitude cross-section of the correlation between the K–KE index and summertime 500-hPa geopotential height (Fig. 12) shows that the most significant correlation occurs in summer and the correlation coefficient is significant at the 0.01 level. However, positive correlation can also be found between the SSTA in K–KE and summertime 500-hPa geopotential height over 110°–130°E in April–May. The above result suggests that the preceding summer positive SSTA over the K–KE may have leading impacts on 500-hPa geopotential height over the midlatitude region of East Asia. This is worth further investigations.

Figure 11 Distributions of the correlation coefficient between 500-hPa geopotential height and time series of summer temperature in (a) China, (b) northern China (north of 30°N and east of 100°E), and (c) summer K–KE index, respectively. Shading indicates the area with the correlation above the 0.05 significant level.

Figure options

Figure 12 Time–latitude profile of the correlation coefficient between the monthly K–KE index and 500-hPa geopotential height averaged over 110°–130°E during summer 1981–2017.

Figure options

Figure 13 presents a schematic diagram of the mechanism for the warmest summer of 2018. In summer 2018, SST was abnormally high over the K–KE area, which resulted in strong anticyclonic circulation anomaly over Northeast China. The NECV was weaker than normal and shifted to the north of its normal position; and convections over Northwest Pacific were active, favorable for the intensification of the WPSH and its northward displacement. Northern China was under the control of descending motions, resulting in abnormally high temperature there.

Figure 13 Sketch map for the possible causes of the hottest summer 2018 in China. The blue thin dashed line (marked as WPSHclim) and the thick red line (marked as WPSH2018) denote the WPSH position of the climate mean and that of 2018, respectively. The thick blue lines indicate the southern boundaries (defined as the zero vorticity line) of the Northeast China cold vortex, with solid line for 2018 and dashed line for the climate mean.

Figure options

Based on the temperature observations collected at 2400 weather stations in China, NCEP/NCAR global reanalysis product, and Hadley Centre SST data, the present study analyzes the temperature characteristics of China in summer 2018 and explores the possible mechanisms behind. The summer 2018 is the warmest summer in China since historical records are available. Particularly, temperature anomalies in northern China made the greatest contribution to the overall record-breaking high temperature over entire China. Different from the high temperature in southern China induced by intensified and westward shifting WPSH, the WPSH area was larger than normal and its ridge line remained abnormally and persistently northward during summer 2018. Meanwhile, the NECV was weaker than normal and located further north, while positive geopotential height anomalies prevailed over the area from northern China to the Sea of Japan. The above circulation anomalies directly led to higher than normal temperature in northern China. Further analysis suggests that positive SSTA in the K–KE region played a critical role for the positive geopotential height anomalies above northern East Asia and high temperature in northern China in summer 2018. Major conclusions are as follows.

In summer 2018, the average maximum, mean, and minimum temperatures over entire China were 27.9, 21.9, and 17.2°C, respectively, which are all record-breaking high since 1961. Temperatures were higher than normal over most areas of China, and the temperature anomaly in northern China was larger than 1°C, exceeding two standard deviations. The intraseasonal and interan-nual temperature variations over northern China were consistent with those over entire China, indicating that temperature anomalies in northern China made the largest contribution to the record-breaking high temperature of entire China. Positive temperature anomalies in northern China are directly linked with circulation anomalies in the middle troposphere, among which the WPSH and NECV are the two most important systems. In summer 2018, the area of WPSH was larger than normal and the ridge line of WPSH remained persistently abnormally northward. Meanwhile, the NECV was weaker than normal and located further north, while positive geopotential height anomalies prevailed over the area from northern China to the Sea of Japan. The above circulation anomalies directly caused the record-breaking high temperature in China. Particularly, from late July to mid August, the WPSH was abnormally strong and its ridge line was located north of 40°N while the NECV was weaker and shifted northward than normal, leading to persistent high temperature in northern China. In addition, the SST anomaly in the K–KE area was up to 1.7 standard deviation, which is the largest since 1961. Statistical analysis suggests that the positive SST anomaly in this region was associated with positive 500-hPa geopotential height anomaly over the midlatitude region of East Asia, which promoted the northward displacement of the WPSH and weakening of the NECV, and thus played a critical role for the abnormal high temperature in northern China.

The persistent circulation anomalies in East Asia, especially the WPSH anomaly, is undoubtedly the direct reason for the record-breaking high temperature over China in summer 2018. Further analysis shows that the circulation anomalies are linked with the East Asia–Pacific (EAP) teleconnection pattern. According to the definition byHuang and Wu (1989), during positive (negative) EAP phase, negative (positive) convection anomalies develop in the tropics, positive (negative) convection anomalies develop in the midlatitudes, and the main body of the WPSH shifts northward (southward). The EAP index in mid summer 2018 ranked the second highest since 1961 with the value exceeding two standard deviations, and the spatial distribution of convection anomalies in the tropics–subtropics actually reflected the impact of the EAP. However, in summer 1994 when the EAP index was then the largest since 1961, the WPSH position was not shifting further north than that in 2018. Similarly, in those years (such as 2015) when the EAP index was abnormally lower than normal, the WPSH position was not really shifting southward than normal. Additional analysis in the present study reveals that the Somali cross-equatorial flow, the Bay of Bengal cross-equatorial flow, and the phase anomaly of the Madden–Julian Oscillation (MJO) may have all contributed to the abnormal high temperature in China in summer 2018. Furthermore, convections in the tropical western Pacific were anomalously active in summer 2018 (ranked top since 1980); the MJO in its 6th phase (over the western Pacific area) remained stronger than normal persistently; and 9 out of 14 tropical cyclones were originated to the north of 18°N and most of them followed the north or northwest tracks. All these factors played an important role for the WPSH ridge line to reach the northernmost position in 2018 (detailed analysis will be presented in another paper).

It is worth noting that the abnormal high temperature in summer 2018 is not an isolated event. In fact, higher than normal temperature occurred over the entire midlatitude region of the Northern Hemisphere. Corresponding to surface observations, positive geopotential height anomalies dominated the middle and lower troposphere, indicating that the abnormal hot summer of 2018 in China occurred in the background of large-scale warm anomalies. Therefore, further studies are needed to investigate the impact of other circulation systems.

In addition to the signals from internal atmospheric circulation and external oceanic forcing, additional external forcing factors such as the upstream Arctic sea ice and the Eurasian snow cover may also have significant impacts on the East Asian summer monsoon circulation system, which directly links with the summertime temperature (high temperature). For example, Wu et al. (2013) proposed that the sea ice concentration anomaly to the west of Greenland could affect springtime circulation anomaly in southern Newfoundland, and subsequently affect the circulation in northern Eurasia in summer. Compared with those years with negative sea ice anomalies, the 500-hPa geopotential height in the years with positive sea ice anomalies exhibited “positive–negative–positive” wave train pattern in East Asia. Similarly, the thermal effect of snow cover in Eurasia can also lead to summer monsoon circulation and temperature anomalies. Particularly, when the springtime snow equivalent water content is lower than normal in Eurasia, the lagged snow (soil moisture) effect will result in higher than normal geopotential height in the mid and high latitude regions of East Asia. As a result, temperature will be abnormally high to the north of 40°N (Yasunari et al., 1991; Wu et al., 2009). The sea ice to the west of Greenland was lower than normal in winter 2017/18, which was favorable for the formation of positive geopotential height anomaly in summer from south of Lake Baikal to northeastern China. Descending motion prevailed in northern China, and temperature was abnormally high. Meanwhile, the number of days with snow cover in Eurasia in spring 2018 was smaller than normal, which may also contribute to the positive geopotential height anomaly in northern East Asia in the subsequent summer. These results indicate that the Arctic sea ice and the Eurasian snow cover might have potential impact on the abnormal high temperature in summer 2018. However, no specific study has been conducted to investigate the impact of sea ice and snow cover on summertime high temperature, and further in-depth analysis is needed to explore the underlying associations.