HTML
-
The data used in this paper included the ERA-Interim reanalysis dataset (ECMWF [26]; Dee et al. [27]) from the European Centre for Medium-Range Weather Forecasts. The geopotential height field at 850 hPa with T255 Gaussian grid (with a resolution of about 0.7°×0.7°) and 6 hours' interval from 1979 to 2018 was used to identify the NCCV. Monthly data for meteorological parameters, such as temperature, pressure, geopotential height, humidity, wind, and surface heat flux, were also used. The precipitation and temperature data of meteorological stations in China from 1979 to 2018 were derived from the China Meteorological Data Network. This data has been quality-controlled, and the data quality and completeness have been significantly improved relative to previously released products. SST data for the same period from Hadley Center with a resolution of 1°×1° (Hadley Centre Global Sea Ice and Sea Surface Temperature [28]) were also adopted.
-
In this study, a cyclone identification algorithm modified by Lu [9] was used to identify summer (June–August) extratropical cyclones from 1979 to 2018. This algorithm is able to effectively track the extreme values of grids based on fields of different meteorological parameters (e.g., sea level pressure and geopotential height field) and search outward for peripheral closed contours at specific contour intervals. Finally, the extreme grid surrounded by the outer closed contours was considered as the center of the cyclone, and the area within its outermost closed contours was considered as the cyclone influence area. Moreover, Neu et al. [29] compared plenty of methods and found that different identification algorithms usually have uncertainties in identifying cyclones in high terrain; therefore, in order to mitigate the influence of terrain, the 850 hPa geopotential height field at 6 hours' interval was adopted in this paper.
After that, we determined NCCVs based on their influence area. Specifically, if the influence area of a cyclone falls into NEC (42°–55°N, 115°–130°E) at any time frame during its lifetime, it is considered an NCCV. Following this criterion, the composite characteristics of synoptic systems were fully considered, including the cases where the cyclone's center was not located in NEC, but its peripheral area reached and impacted NEC. Furthermore, a composite NCCV intensity index (NCCVI) was defined as a comprehensive measure of NCCV's central intensity and peripheral influence area. The expression is as follows:
$$ \begin{aligned} & \mathrm{NCCVI}=\overline{\left[\frac{\iint \mathrm{CVId} A}{\iint \mathrm{d} A}\right]}, \\ & \mathrm{CVI}=\left\{\begin{array}{c} \operatorname{vor}_{850} \text {, within a cyclone } \\ 0, \text { outside a cyclone } \end{array}\right\} . \\ & \end{aligned} $$ The over line in the expression denotes the summer average, and ∫∫dA indicates the area integration of NEC. vor850 indicates the values of relative vorticity at 850 hPa. The intensity of cold vortex (CVI) only accounts for the area inside NEC. Compared to those considering only the intensity at the center of NCCV, this NCCV intensity definition can provide a more composite representation of the overall NCCV, especially for its peripheral influence area. Then, the summer average NCCVI was calculated and normalized, and when referring to NCCVI in the subsequent text, it denotes the normalized NCCVI in summer.
2.1. Data
2.2. Methods of NCCV identification
-
Statistics on summer cyclones in East Asia found that cyclones occur most frequently in the Mongolian Plateau, Daxinganling and NEC with a maximum frequency of up to 27% (see Fig. 1), suggesting that NCCVs may have potential impacts on the weather and climate of NEC and its surroundings in summer. Therefore, NEC, which is located in the downstream of the Mongolian Plateau, was selected as the main study area.
Figure 1. The 2D frequency (units: %) of extratropical cyclones in East Asia from 1979 to 2018 in summer (contours). The shaded indicates the terrain elevation (units: m), and the black box represents Northeastern China.
From the normalized time series of summer NCCVI (Fig. 2a), it can be seen that NCCV activities exhibited a significant inter-annual variability, but there was no significant trend (–0.22±0.25 per decade). Statistically, the amplitudes of the summer NCCVI in 21 years exceeded one standard deviation (1 STD), i.e., anomalous years (red and blue dots in Fig. 2a). Specifically, the NCCVI indexes were greater than 1 STD in 1981, 1983, 1984, 1987, 1990, 1998, 2003, 2009 and 2013 (i.e., nine strong years), while the NCCVI indexes in 1980, 1982, 1988, 1994, 1999, 2001, 2002, 2004, 2008, 2010, 2011 and 2014 (i.e., 12 weak years) were less than 1 STD, suggesting a large inter-annual difference. Moreover, the results of power spectrum analysis (Fig. 2b) indicate that the summer NCCV activities had an obvious periodicity of 3–4 years and 6–7 years.
Figure 2. (a) The normalized time series of summer NCCVI from 1979 to 2018, and (b) its power spectrum analysis. The gray and red dashed lines in Fig. 2a indicate ±1 standard deviation, and the linear regression of NCCVI fitted by least squares; the red (blue) dots represent strong (weak) anomalous years, respectively. The black line in Fig. 2b indicates the smoothed spectrum estimate values, and the red dashed line represents the red noise standard spectrum is statistically significant at the confidence level of 90%.
-
The atmospheric circulation of summer NCCVs in anomalous years was further investigated by the comparison of several meteorological elements between the strong and weak NCCVI years. Fig. 3 shows the composite differences of the 300 hPa, 500 hPa, and 850 hPa geopotential height fields in the strong and weak NCCVI years. Compared with the weak NCCVI years, the geopotential height fields in the strong NCCVI years showed obvious negative anomalies in the entire troposphere above NEC, suggesting a clear quasi-barotropic structure. In contrast, there were obvious positive anomalies of geopotential height in the entire troposphere over the Ural Mountains. The coupling of geopotential height anomalies between the Ural Mountains and NEC together led to an increase in the zonal gradient of the geopotential height between them. It was prone to trigger anomalous northeasterly winds according to the geostrophic wind principle. On the contrary, the opposite geopotential height anomalies between NEC and the middle and low reaches of the Yangtze River promoted the occurrence of anomalous westerly winds. Then, an anomalous cyclonic circulation was favorable for the maintenance and strengthening of NCCV activities.
Figure 3. The composite differences of geopotential height fields (contours, units: gpm) at (a) 300 hPa, (b) 500 hPa, and (c) 850 hPa between strong and weak NCCVI years. The light (dark) gray shaded area indicates results statistically significant at the confidence level of 90% (95%). The black boxes indicate Northeastern China.
The composite differences of horizontal winds are shown in Fig. 4; it is found that there were obvious northeasterly anomalies between the Ural Mountains and NEC, which was roughly located between the centers of positive and negative anomalies of the geopotential height. An anti-cyclonic circulation anomaly appeared over the Ural Mountains, with a cyclonic circulation anomaly above NEC, which coincided with the geopotential height field. Moreover, the wind anomalies at 300 hPa were significantly stronger than those at 850 hPa, indicating that there was an obvious vertical wind shear near NEC, which was conducive to the enhancement of baroclinicity and provided favorable conditions for the enhancement and maintenance of NCCV activities.
Figure 4. The composite differences of horizontal winds (vectors, units: m s–1) at (a) 850 hPa and (b) 300 hPa between strong NCCVI years and weak NCCVI years. The light (dark) gray shaded area indicates results statistically significant at the confidence level of 90% (95%). The black boxes indicate Northeastern China.
The composite differences of temperature in the lower troposphere are given in Fig. 5. It can be observed that in the strong NCCVI years, there were significant negative temperature anomalies in the lower troposphere in NEC, while there were positive temperature anomalies in the Ural Mountains. The temperature anomaly value of the cold center was < –1.5℃, while the temperature anomaly value of the warm center was > 1.5℃. The temperature anomalies in the lower troposphere satisfied the principle of thermal wind, which was the main reason for the significant vertical wind shear between the Ural Mountains and NEC.
Figure 5. The composite differences of temperature (contours, units: ℃) at (a) 850 hPa and (b) 700 hPa between strong and weak NCCVI years. The light (dark) gray shaded area indicates results statistically significant at the confidence level of 90% (95%). The black boxes indicate Northeastern China.
Moreover, the negative temperature anomalies in NEC promoted the loss of atmospheric column mass due to the cooling contraction of the air column. This was accompanied by negative geopotential height and cyclonic circulation anomalies, providing a favorable local atmospheric condition for the development of NCCV activity. Furthermore, NCCV is generally considered a product of large-scale circulation patterns under specific conditions in NEC. Its activity interacts with blocking highs in a mutually reinforcing manner. Therefore, the circulation of positive geopotential height anomalies over the Ural Mountains also favored the formation of NCCV (Zhang et al. [1]).
-
From the above analyses of the atmospheric circulation in anomalous NCCVI years, we may find that the fields of geopotential height, wind and temperature in NEC and its vicinity cooperated closely with each other and form a circulation background, which was conducive to the maintenance and development of NCCV activities. To explore the connection between the summer NCCV activities and the contemporaneous precipitation in China, the correlation analysis of NCCVI and precipitation was calculated by using data from Chinese meteorological stations as well as the ERA-Interim reanalysis dataset. As shown in Fig. 6a, there was a significant positive correlation between the NCCVI and the contemporaneous precipitation in NEC, and the correlation coefficients in most areas of NEC were higher than 0.4, with the maximum value reaching 0.6. The results calculated with the reanalysis dataset also showed a significant positive correlation there, with relatively higher correlation coefficients (Fig. 6b). This indicates that when the summer NCCV activities were anomalously stronger, the contemporaneous precipitation in NEC increased significantly.
Figure 6. The correlation coefficient between NCCVI and monthly total precipitation was calculated from (a) Chinese meteorological station data and (b) the ERA-Interim reanalysis dataset. (c, d) are similar to (a, b), but for the correlation between NCCVI and monthly mean temperature at 2m. (c) is based on data from Chinese meteorological stations, and (d) is based on the ERA-Interim reanalysis dataset. The light (dark) gray shaded area indicates results statistically significant at the confidence level of 90% (95%). The black boxes indicate Northeastern China.
Moreover, Fig. 6c shows that there was also a good negative correlation between the NCCVI and summer surface temperature in NEC, either in the Chinese meteorological station data or the ERA-Interim reanalysis dataset. The correlation coefficients in most areas of NEC were lower than –0.4, and the minimum value could reach –0.6. This indicates that the summer surface temperature in NEC was significantly lower during strong NCCVI years. Furthermore, the results of Chinese meteorological station data and the ERA-Interim reanalysis dataset consistently show that there was a significant positive correlation between NCCVI and surface temperature in South China and the Yangtze River.
Meanwhile, the scatter diagram of the NCCVI and regional average precipitation (surface temperature) in NEC (Fig. 7) also shows a significant positive (negative) correlation between NCCVI and precipitation (surface temperature). The correlation coefficient between NCCVI and the regional average of precipitation (surface temperature) in NEC was 0.52 (–0.33), which was statistically significant at the confidence level of 95%. That is, in strong NCCVI years, the summer precipitation increased in NEC with lower surface temperatures. In weak NCCVI years, the situation was the opposite.
Figure 7. (a) The scatter diagram of summer NCCVI (x-axis) and regional average monthly total precipitation (y-axis) in Northeastern China. (b) is similar to (a), but for the scatter diagram of summer NCCVI (x-axis) and regional average monthly surface temperature at 2 m (y-axis). The red lines indicate linear regressions of NCCVI on precipitation and surface temperature, respectively. The correlation coefficients between NCCVI and the regional average of precipitation or surface temperature are 0.52 and –0.33, respectively; the results are both statistically significant at the confidence level of 95%.
Based on the statistics of total precipitation contribution in summer, it is found that precipitation associated with NCCV activities constituted approximately 58.5% of the total precipitation in NEC, which emphasized the important role of NCCV as one of the major precipitation systems in NEC. Moreover, the composite difference in surface temperature between periods of NCCV presence and absence reveals that when NCCV activity occurred, the surface temperature in NEC was significantly lower, with a regional average temperature difference of –0.95 K (statistically significant at the confidence level of 90%).
To further show the climatic effects of NCCV, the integrated water vapor flux and its divergence are given in Fig. 8a. From which it can be seen that during the strong NCCVI years, a large amount of water vapor converged in NEC due to the anomalous NCCV activities, which was mainly from two water vapor transportation belts. One of the belts was located on the southeast side of NEC, and water vapor from the downstream of the Yangtze River flowed into NEC through this belt associated with the East Asian summer monsoon. Another belt was located on the east side of NEC, transporting the water vapor from the Northwestern Pacific into NEC. Furthermore, under the influence of NCCV activities, there was a significant anomalous upward motion in the lower troposphere (Fig. 8b). Increased water vapor and upward motion both provided favorable conditions for potential precipitation in NEC.
Figure 8. The composite differences of (a) integrated water vapor flux (vectors, units: 102 kg m–1 s–1) and its divergence (shaded, units: 10–5 kg m–2 s–1), as well as (b) vertical velocity (shaded, units: 10–2 Pa s–1) at 850 hPa between strong and weak NCCVI years. The gray (black) dots indicate results statistically significant at the confidence level of 90% (95%). The black boxes indicate Northeastern China.
-
As mentioned above, favorable atmospheric circulation occurred in NEC during strong NCCVI years, accompanied by coupled circulation in the Ural Mountains, and there was a significant positive correlation between NCCVI and precipitation in NEC. To further understand the intensification mechanisms of NCCV activities in summer, we explored the linkage between NCCVI and the contemporaneous SST of the North Atlantic (Fig. 9). It can be observed that there was a significant positive correlation between them. The correlation coefficient can reach up to about 0.3 in the mid-latitude Atlantic. Furthermore, the regression of geopotential height fields at different levels on the NCCVI are given in Fig. 10; it is found that there existed a wave-train-like anomalous geopotential height in the troposphere from the mid-latitude Atlantic to NEC via western Europe and Ural Mountains. The circulation characteristics near the Ural Mountains and NEC were similar to the composite analysis results of geopotential height (Fig. 3). This implies that anomalous Rossby waves were excited in the Atlantic and propagate eastward, which finally affected NEC. Many previous studies have also emphasized the importance of the similar Rossby wave train propagating from the Northern Atlantic to East Asia, such as the East Atlantic/West Russia teleconnection pattern (Barnston and Livezey [30]; Lim [31]), Scandinavia teleconnection pattern (Wallace and Gutzler [32]; Bueh and Nakamura [33]), British-Baikal Corridor teleconnection pattern (Xu et al. [34]), as well as Silk Road teleconnection pattern (Lu et al. [35]; Wang et al. [36]).
Figure 9. The correlation of NCCVI and SST of the Atlantic. The gray (black) dots indicate results statistically significant at the confidence level of 95%.
Figure 10. The regression of geopotential height fields (contours, units: gpm) at (a) 300 hPa, (b) 500 hPa, and (c) 850 hPa on NCCVI. The "+" symbol in (c) indicates the location of the SST warm anomaly center. The light (dark) gray shaded area indicates results statistically significant at the confidence level of 90% (95%), and the black boxes indicate Northeastern China.
The regressions of the surface sensible heat and latent heat flux fields on NCCVI are given in Fig. 11. As a result of the Atlantic SST heat anomalies, large amounts of surface sensible and latent heat flux were transported into the lower troposphere. Based on the composite analysis of the vertical circulation (Fig. 12) and the surface thermal characteristics over the Atlantic, it is found that as the Atlantic SST rose, the geopotential height increased in its vicinity and a positive geopotential height anomaly occurred in the east side of its heat source. Moreover, it was accompanied by an anomalous upward motion, which formed an anomalous thermal circulation (Fig. 12). As a result, an eastward diverging flow was formed in the upper troposphere and propagates downstream.
Figure 11. The regressions of (a) surface sensible heat flux (units: J m–2) and (b) surface latent heat flux (units: J m–2) on NCCVI. The gray (black) dots indicate results statistically significant at the confidence level of 90% (95%).
Figure 12. The regression of 50°–60°N meridional averaged profile of u-w winds (streamline) and geopotential height (shaded, units: gpm) on NCCVI.
This phenomenon is inseparably linked to the propagation of the Rossby wave train. The horizontal T-N wave activity fluxes (Takaya and Nakamura [37]) are an effective diagnostic of Rossby wave activity, and their expression is as follows:
$$ \begin{aligned} & W=\frac{p \cos \varphi}{2|\boldsymbol{U}|} . \\ & \binom{\frac{U}{a^2 \cos ^2 \varphi}\left[\left(\frac{\partial \psi^{\prime}}{\partial \lambda}\right)^2-\psi^{\prime} \frac{\partial^2 \psi^{\prime}}{\partial \lambda^2}\right]+\frac{V}{a^2 \cos \varphi}\left[\frac{\partial \psi^{\prime}}{\partial \lambda} \frac{\partial \psi^{\prime}}{\partial \varphi}-\psi^{\prime} \frac{\partial^2 \psi^{\prime}}{\partial \lambda \partial \varphi}\right]}{\frac{U}{a^2 \cos \varphi}\left[\frac{\partial \psi^{\prime}}{\partial \lambda} \frac{\partial \psi^{\prime}}{\partial \varphi}-\psi^{\prime} \frac{\partial^2 \psi^{\prime}}{\partial \lambda \partial \varphi}\right]+\frac{V}{a^2}\left[\left(\frac{\partial \psi^{\prime}}{\partial \varphi}\right)^2-\psi^{\prime} \frac{\partial^2 \psi^{\prime}}{\partial \varphi^2}\right]} \\ & \end{aligned} $$ where U indicates the climatological horizontal wind vectors in summer, including zonal component U and meridional component V, $\psi^{\prime}=\frac{\varphi^{\prime}}{f}$indicates stream function anomalies, and p, φ, λ, and a indicate basic meteorological parameters respectively. Based on the regression of horizontal T-N wave activity fluxes on NCCVI (Fig. 13), it is evident that a clear Rossby wave train propagated eastward from the Atlantic via western Europe and the Ural Mountains, eventually affecting NEC.