An Interdecadal Change of Summer Atmospheric Circulation over Asian Mid-High Latitudes and Associated Effects

Summer weather conditions and climate in East Asia not only are affected by tropical and subtropical monsoon circulation, but also are subject to the influence of atmospheric circulation over Asian mid-high latitudes ^[1]. The atmospheric circulation near the Ural Mountains and Sea of Okhotsk plays a great role in summer rainfall variations over the middle and lower reaches of the Yangtze River ^[1]. Summer droughts and floods in East Asia are often associated with persistent atmospheric circulation anomalies over Asian mid-high latitudes ^[2-5]. For example, the persistent atmospheric circulation anomalies over Asian mid-high latitudes is found to be related to the great floods in the Yangtze River valley in 1998^{[4, 5]}. Therefore, the atmospheric circulation over Asian mid-high latitudes deserves close attention. At present, most related studies mainly focus on a certain weather system or a short-term process ^[2-9]. A few scholars analyzed climatic characteristics of the atmospheric circulation over Eurasian mid-high latitudes in summer ^{[10, 11]}. Their results revealed the predominance of two typical persistent circulation patterns over the region and showed that the atmospheric circulation over the region experienced obvious interdecadal variations around 1980.

Interdecadal climate variability has received greater attention during recent decades. The East Asian summer monsoon (EASM) underwent a significant weakening in the late 1970s ^[12-15] and affected the decadal variation of precipitation over China, with a significant increase in the precipitation over Yangtze River valley since then ^[16-19]. Tropical forcing, atmospheric heating over the Tibetan Plateau (TP) and air pollution in East Asia are also believed to have played an important role in the weakening of the EASM in the late 1970s ^{[17, 19, 20-24]}. Corresponding to the interdecadal change of summer precipitation in east China around 1980, summer atmospheric circulation over Eurasian mid-high latitudes was found to have experienced obvious interdecadal variation and have a close relationship with the precipitation changes in the Changjiang-Huaihe River valley and north China^[11]. It is worth mentioning that obvious interdecadal changes in the precipitation over China also occurred in the 1990s. For example, summer rainfall of south China experienced a significant increase in the early 1990s ^{[13, 25]} and summer rainfall over northeast and north China decreased significantly in the late 1990s ^[26-28]. Besides, Kwon et al. ^[29] found that the EASM experienced a decadal change in the mid-1990s, associated with a remarkable increase in the rainfall over southeastern China after that. Wu et al. ^[25] suggested the pronounced interdecadal change in summer rainfall over south China around 1992/93 was related to the Tibetan Plateau snow cover and the SST in the equatorial Indian Ocean. The SST in the Pacific Ocean was believed to be an important factor influencing the interdecadal change in summer rainfall over northeast and north China in the late 1990s ^{[27, 28]}.

In summary, previous studies have made significant progresses in understanding the atmospheric circulation over Asian mid-high latitudes and the interdecadal climate shifts in this region. However, the atmospheric circulation over Asian mid-high latitude and its effect on regional climate have not been fully investigated. There are still some questions to be answered. For example, has there been any significant decadal change in the atmospheric circulation over Asian mid-high latitudes since 1980? If any, what is the main mechanism? What kind of role does this change play in the interdecadal changes of precipitation in China during the 1990s and does it have any other impacts on the regional climate? Exploration into these topics will benefit understanding and forecasting of Asian climate variation.

This paper focuses on an interdecadal change of atmospheric circulation over Asian mid-high latitudes in recent decades and its effects on regional temperature and precipitation. The data and analysis methods are introduced in section 2. Section 3 and section 4 examine the interdecadal change of atmospheric circulation over Asian mid-high latitudes and its impacts on regional climate. One possible mechanism behind this change is analyzed in section 5. Finally, a summary and a discussion are given in section 6.

The main datasets used in this study include the monthly mean European Centre for Medium-Range Weather Forecasts interim (ERA-Interim, 1.5° × 1.5° grid) reanalysis ^[30], the land surface air temperature of the Climatic Research Unit (CRU) Ts3.25 dataset (0.5°× 0.5° grid) ^[31], and the 1979-2015 monthly mean precipitation rate data from the Climate Prediction Center Merged Analysis of Precipitation ^[32] (CMAP, 2.5° × 2.5° grid). All the data are collected from 1979-2015. The monthly rainfall data of 160 stations of the China Meteorological Administration during same period are also used in this study. In addition, the monthly mean SST of the Hadley Centre Sea Ice and Sea Surface Temperature dataset^[33] (HadISST, 1° × 1° grid) and monthly NCEP / DOE Reanalysis Ⅱ ^[34] (2.5° × 2.5° grid) during 1979-2015 are also used in the study.

The vertically integrated water vapor flux (Q) is defined as: $\boldsymbol{Q}=-\frac{1}{g} \int_{p_s}^{p} q \boldsymbol{V} \mathrm{d} p$. Here, Q is the vertically integrated water vapor flux from p_s to p, p_s is the surface pressure, p is set to be 300 hPa, q i_s the specific humidity, V is the horizontal wind vector, and g is the gravity acceleration.

This study focuses on summer. Summer conditions are represented by the averages from June to August. Moreover, a 9-year Gaussian low-pass filter method is used to obtain the interdecadal component of a variable. Empirical orthogonal function (EOF) analysis, composite analysis, Mann-Kendall test, linear regression and correlation are used to examine the interdecadal change of the atmospheric circulation over Asian mid-high latitudes, its effects and influencing mechanism. Statistical significance is assessed by Student's t test or the Monte Carlo simulation.

Figure 1a-1d show the standard deviation distribution maps of summer 850-hPa, 700-hPa, 500-hPa and 200-hPa geopotential height (gpm) during 1979-2015 on the interdecadal time scale. It is found that the fluctuations of summer atmospheric circulation over Asian mid-high latitudes are stronger than that over low latitudes on the interdecadal time scale. One significant change area appears near Lake Baikal, where there is a center exceeding 8 gpm at 850 hPa, 10 gpm at 700 hPa, 12 gpm at 500 hPa and 24 gpm at 200 hPa, respectively. An EOF analysis is applied to 9-yr Gaussian low-pass-filtered 500-hPa geopotential height in summer to extract the dominant interdecadal variation patterns during 1979-2015. Fig. 1e shows the first mode, which accounts for 50.9% of the total variances. Asian mid-high latitudes are dominated by a zonal circulation anomaly, with a positive and negative center over western Siberia and the region south to Lake Baikal, respectively. Similar analyses are also made to the geopotential height at 850 hPa, 700 hPa and 200 hPa and the results are consistent with that in Fig. 1e. This indicates that the region near Lake Baikal is an active center of the dominant interdecadal change pattern over Asia during 1979-2015, which deserves further investigation.

Figure 1.  (a) The standard deviation distribution of summer 850-hPa geopotential height (gpm) during 1979-2015 on the inter‐ decadal time scale. (b) Same as (a), but for 700-hPa geopotential height. (c) Same as (a), but for 500-hPa geopotential height. (d) Same as (a), but for 200-hPa geopotential height. The box with dotted line shows the significant variation center of R1 (41°N-56°N, 92°E-115°E) based on ERA-interim reanalysis data. (e) The first EOF mode of summer 500-hPa geopotential height during 1979-2015 on the interdecadal time scale.

Based on Fig. 1a-1d, the standardized time series of summer geopotential height averaged over the significant change zone near Lake Baikal (R1, 41°N-56° N, 92°E-115°E, shown in Fig. 1d) at 850 hPa, 700 hPa, 500 hPa and 200 hPa are calculated. It is found that the regional height over the region at different levels have similar variation characteristics in recent decades(Fig. 2a). The values of these standardized time series are mainly negative during 1979-1996 and positive during 1997-2015, implying an interdecadal change around 1996 and exhibiting a remarkable increasing trend with statistical significance at the 99.9% confidence level. The standardized time series of summer 200-hPa geopotential height averaged over the area R1 is defined as an index (ERA-I200) to represent the long-term variability of the atmospheric circulation around Lake Baikal in summer. The correlation coefficients between ERA-I200 and the standardized time series of summer geopotential height averaged over the area R1 at 850 hPa, 700 hPa and 500 hPa are 0.54, 0.76 and 0.90 during 1979-2015, which indicates this index can reflect the variation characteristics of the whole troposphere near Lake Baikal in recent decades. The Mann-Kendall test is also used to analyze the interdecadal change-point of ERA-I200. The result also shows it has changed significantly around 1996 (Fig. 2b).

Figure 2.  (a) Standardized time series of summer geopotential height averaged over the variation center (R1, shown by the box in Fig. 1d) during 1979-2015. The standardized series of summer 200-hPa geopotential height averaged over the area R1 is defined as an index (ERA-I200) to represent summer atmospheric circulation variation around Lake Baikal. The thick red line shows the inter‐ decadal component of this index. (b) Mann-Kendall test results of summer atmospheric circulation index near Lake Baikal (ERAI200) during 1979-2015. Blue and red straight lines denote the threshold of 95% confidence level.

Based on the variation features of the atmospheric circulation index near Lake Baikal (Fig. 2a and Fig. 2b), 1979-1996 (P1) and 1997-2015 (P2) periods are respectively used to represent the negative and positive phases on the interdecadal scale. Fig. 3a shows the difference of summer 850-hPa geopotential height between the two phases (P2-P1). There are significant positive anomalies near Lake Baikal, with a center of 10 gpm. Negative anomalies mainly appear around the Ural Mountains and western Siberia. The distribution patterns of height anomalies at 700 hPa, 500 hPa and 200 hPa are similar to that at 850 hPa, with significant positive anomaly centers near Lake Baikal. Fig. 3a-3d also indicate that the positive anomaly center near Lake Baikal is deep in troposphere. Corresponding to the height anomaly distribution, an anomalous anticyclonic circulation appears near Lake Baikal and surrounding areas in wind field (Fig. 4). Under its control, the easterly and northeasterly wind anomalies significantly affect most regions north to 30°N in East Asia.

Figure 3.  (a) The difference of summer 850-hPa geopotential height (gpm) between 1997-2015 and 1979-1996 (the 1997-2015 (P2) mean minus 1979-1996 (P1) mean, similarly hereinafter). Shaded regions are significant at the 95% confidence level. (b) Same as (a) but for 700-hPa geopotential height (gpm). (c) Same as (a) but for 500-hPa geopotential height (gpm). (d) Same as (a) but for 200-hPa geopotential height (gpm).

Figure 4.  Same as Fig. 3a but for summer 700-hPa wind (m s^-1). Thick vectors in the figure are significant at the 95% confidence level. Shaded areas in the figure show the topography above 3000 m.

The analyses above show that the atmospheric circulation over Asian mid-high latitudes has strong interdecadal fluctuations and one significant variation center appears near Lake Baikal, where the geopotential height has significantly increased since 1997. Such a remarkable interdecadal change is bound to have important impacts on the regional climate in Asia.

Figure 5a shows land surface air temperature difference in summer between 1997-2015 and 1979-1996. In the past 20 years, the summer surface temperature in most Asia has increased significantly and the most prominent warming with centers more than 1.8℃ appears in the region between Lake Baikal and the Tibetan Plateau, which is consistent with the previous studies ^[35-38]. Fig. 5b further shows that the region near Lake Baikal is one of the regions with the fastest warming rate over Asia during 1979-2015, with a center more than 0.7℃ per 10 years. Why has the warming center of Asian appeared in the vicinity of Lake Baikal in recent 40 years, rather than other regions? Further analyses show that apart from the near-surface, temperature has increased in almost the whole troposphere near Lake Baikal (Fig. 5c), indicating a significant warming during 1997-2015. It is noted that the position of this strong warming is consistent with that of the anomalous anticyclone circulation in Fig. 3. Therefore, it is speculated that this significant warming near Lake Baikal may be related to the whole-layer anticyclone circulation overhead. Through analysis, obvious anomalous sinking movements of whole layer are found to dominate over the region near Lake Baikal (R1, 41° N-56° N, 92° E-115° E) in the past 20 years, especially over the intense warming center (47°N-50°N, 92°E-115°E). Except for some areas around 92°E with negative difference (anomalous weak ascending motion) near surface, the intense warming center near Lake Baikal is totally controlled by a whole layer of positive difference (anomalous sinking movement) from 200 hPa to surface(Fig. 5d). The strongest anomalous sinking movement occurs near 100°E with a center of more than 2.5×10^-2 Pa s^-1. This kind of anomalous subsidence in the whole troposphere suppresses local precipitation, benefits regional warming, and corresponds to the intense surface warming over Asian mid-high latitudes between 92°E and 115°E (Fig. 5a). Besides the sinking motion anomalies, the southeasterly wind anomalies to the west flank of the anomalous anticyclone (Fig. 4) may also contribute to the significant surface warming west to 100° E. The anomalous anticyclone can prevent the colder air from higher latitudes or bring warmer air from lower latitudes to this region during 1997-2015. Time series of summer land surface air temperature averaged over the warming region near Lake Baikal (R1) during 1979-2015 is calculated. The correlation coefficient is 0.90 between it and the summer atmospheric circulation index near Lake Baikal (ERA-I200) during 1979-2015, significant at the 99% confidence level. Therefore, it can be suggested the whole layer anomalous anticyclone circulation can contribute to the significant surface warming near Lake Baikal and this warming effect is a response to the atmospheric circulation anomaly overhead. Global warming is also believed to have an effect on the surface warming of this area in recent decades ^[36-37]. Fig. 5e shows the time series of summer land surface air temperature averaged over the warming region near Lake Baikal (R1) during 1951-2016. The warming around Lake Baikal since 1997 has been remarkable even in nearly 70 years. Perhaps it is the combined effects of global change and atmospheric circulation that have led to a significant increase in temperature near Lake Baikal since the mid-late 1990s.

Figure 5.  (a) Summer land surface air temperature difference (℃) between P2 and P1. (b) Linear trend distribution of summer land surface air temperature in Asian during 1979-2015 (℃ 10years^-1). (c) Same as (a) but for the height-longitude cross section of summer air temperature difference (K) along 41°N-56°N. (d) Same as (a) but for the vertical cross-section of summer vertical velocity (10^-2 Pa s^-1) difference along 47°N-50°N. The intense warming center near Lake Baikal is near 47°N-50°N (shown in Fig. 5a and 5b). Shaded regions in light blue are significant at the 95% confidence level and shaded areas in black show the topography. (e) Time series of summer land surface air temperature averaged over R1 during 1951-2016. The red line in (e) shows the average value during 1951-2016.

Figure 6a shows the regressed summer precipitation against the atmospheric circulation index around Lake Baikal (ERA-I200) on the interdecadal scale. It indicates, on the interdecadal scale, the increase of the geopotential height near Lake Baikal is beneficial to a decrease in precipitation between 40°N and 53°N and an increase in precipitation south to 40° N in Asia. The precipitation difference between 1997-2015 and 1979-1996 (Fig. 6b) corresponds to Fig. 6a. From 53° N to 20° N in Asia, the precipitation difference displays a negative-positive pattern on the whole. The precipitation between 40° N and 53° N, compared with that during 1979-1996, significantly decreased in the past 20 years. To China, the precipitation in northeast China and north China is less and the precipitation is more in the central Tibetan Plateau and the region south to the Yangtze River, compared with that during the period 1979-1996.

Figure 6.  (a) The regressed summer rainfall rate (mm d^-1) from CMAP against the atmospheric circulation index around Lake Baikal on the interdecadal scale; (b) the difference of summer rainfall rate (mm d^-1) based on CMAP between P2 and P1. Shaded regions are significant at the 95% confidence level. The boxes in (b) indicates three regions in Asia, including A(40°N-52°N, 110°E-132°E), B (40°N-52°N, 92°E-110°E) and C (21°N-30°N, 110°E-123°E). (c) Same as Fig. 3a but for summer water vapor transport flux (kg m s^-1). The vectors in boldface are statistically significant at the 95% confidence level. (d) Time-latitude section of summer total water vapor transport (kg m s^-1; $\sqrt{( q_u)^2 + ( q_v)^2}$, q_u and q_v are zonal and meridional water vapor transport respectively) over east China (110°E-120°E) from 1979 to 2015.

The anomalous anticyclone near Lake Baikal has played a great role in the changes of Asia precipitation in the past 20 years. Fig. 6c is the composite difference of summer Q between P2 and P1 and a significant anticyclonic transport anomaly appears near Lake Baikal. Under its influence, the northeasterly and easterly water vapor flux anomalies significantly affect most parts south to 50°N in East Asia. In addition, there is a weak anomalous cyclonic transport circulation over the region south to 30°N in east China. Corresponding to the water vapor transport flux anomaly over Asia (Fig. 6c), the region southwest to Lake Baikal, northeast China and most parts of north China are water vapor divergence regions, which is not beneficial to the occurrence of local precipitation. Most parts of the Tibetan Plateau and the region south to the Yangtze River of east China are water vapor convergence regions, which is favorable for more precipitation. It can be found that the divergence distribution of water vapor transport anomaly and precipitation anomaly correspond to each other.

Similar analyses are performed for the precipitation gauge observations in China. The precipitation difference of east China also shows less precipitation in north and more in south between two phases. Compared with the climate mean state of water vapor transport in summer, the water vapor transport anomalies over northeast and north China indicate the decrease of water vapor transport with summer monsoon (Fig. 6c). Fig. 6d clearly shows the obvious reduction in total water vapor transport between 40°N and 50°N over east China (110° E-120° E) since the mid-late 1990s. A large amount of water vapor can only be transported to the region south to 30°N in the recent 20 years. To the region south to 30° N in east China, sufficient water vapor and regional water vapor convergence are favorable for more precipitation in this region. To northeast China and north China, the decrease of the water vapor transports and regional water vapor divergence are not beneficial to the occurrence of local precipitation. Fig. 7a and 7b show the time series of precipitation in three regions in Asia. These regions are shown in Fig. 6b, including A (40°N-52°N, 110°E-132°E), B(40°N-52°N, 92°E-110°E) and C (21° N-30° N, 110° E-123° E). The precipitation in A and B both have an obvious interdecadal change around 1997(Fig. 7a), which further shows the effects from the changes of atmospheric circulation near Lake Baikal. Based on Fig. 6b and 7a, the precipitation in northeast China and the north part of north China (Region B) has decreased significantly since the late 1990s, which is consistent with previous studies ^[26-28]. It can be seen that the interdecadal change in the atmospheric circulation near Lake Baikal has played an important role in the change of precipitation in northeast China and north China around 1997. To the region south to 30°N in east China (Region C), the precipitation has increased significantly around 1993, which is earlier than that in northeast China and north China. Wu et al. ^[25] suggested that the pronounced interdecadal change in summer rainfall over south China (22.5°N-27.5°N, 105°E-120° E) during 1992-1993 was related to the Tibetan Plateau snow cover and the SST in equatorial Indian Ocean. Thus, there may be multiple influencing factors leading to the obvious increase in precipitation in the region south to 30° N in east China in 1990s and associated mechanisms are also complicated.

Figure 7.  Standardized time series of precipitation in three regions in Asia during 1979-2015. Three regions are shown in Fig. 6(b), including A(40°N-52°N, 110°E-132°E), B(40°N-52°N, 92°E-110°E) and C (21°N-30°N, 110°E-123°E). The thick lines in the figures show the interdecadal components of the standardized time series of regional mean precipitation. A_CMAP and A_station show the time series of precipitation from CMAP and observation station.

Therefore, what is the possible mechanism behind the interdecadal change in the atmospheric circulation near Lake Baikal around 1996? Fig. 8a shows the regressed summer 200-hPa geopotential height against the atmospheric circulation index around Lake Baikal (ERA-I200) on the interdecadal scale. There is a zonal wave train over the mid-high latitudes of the Northern hemisphere. The positive anomalies centers mainly appear over northwestern Atlantic, Europe and Lake Baikal, while the negative anomalies centers appear over eastern North Atlantic and western Asia. Similar anomalous wave train is found in the composite difference of summer 850-hPa, 700-hPa, 500-hPa and 200-hPa geopotential height between 1997-2015 and 1979-1996. Fig. 8b show the composite difference at 200-hPa geopotential height. It can be seen that the significant change center near Lake Baikal is closely linked to a teleconnection wave train and is its main part on the interdecadal time scale. Fig. 8c shows the regressed summer SST against the atmospheric circulation index around Lake Baikal (ERA-I200) on the interdecadal scale. The most significant related region appears in northwestern Atlantic with a center more than 1.4℃, which indicates the anticyclonic circulation near Lake Baikal is closely linked to the warming in northwestern Atlantic in recent 20 years. The difference of summer SST between P2 and P1 is also analyzed in this study (Fig. 8d). The distribution of summer SST difference in North Atlantic is similar to that in Fig. 8b and northwestern Atlantic is the most significant variation area, with a center more than 1.7℃. Zhou et al. ^[39] found the anomalous wave train in Fig. 8a and Fig. 8b could be triggered by SST anomaly in northwestern Atlantic based on climate model simulations (the key SST change area in northwestern Atlantic given by Zhou et al. ^[39] is almost identical to the most significant related region in Fig. 8c), and an increase of summer northwestern SST corresponded to the anomalous wave train with positive anomalies centers mainly appear over northwestern Atlantic, Europe and Lake Baikal, which is similar to that in Fig. 8a and Fig. 8b. Fig. 8e shows the regressed 200-hPa geopotential height against the standardized time series of summer Atlantic SST over the significant positive anomalies (65°W-39.5°W, 42°N-55° N) on the interdecadal scale. Similar anomalous wave train appears with positive anomalies center near Lake Baikal. Thus, the anomalous anticyclonic circulation near Lake Baikal is closely related to the interdecadal warming of northwestern Atlantic during 1997-2015, compared with that during 1979-1996. As for Asian climate change in the recent 40 years, it can be further concluded the northwestern Atlantic SST is an important influencing factor.

Figure 8.  (a) The regressed summer 200-hPa geopotential height (gpm) against the atmospheric circulation index around Lake Baikal on the interdecadal scale; (b) The difference of summer 200-hPa geopotential height (gpm) between 1997-2015 and 1979-1996. (c) Same as (a), but for regressed summer SST(℃); (d) Same as (b), but for summer SST(℃). (e) The regressed summer 200-hPa geopotential height (gpm) against the standardized time series of summer Atlantic SST over the significant positive anomalies in (65° W-39.5°W, 42°N-55°N) on the interdecadal scale. Shaded regions are significant at the 95% confidence level.

Previous studies suggest that the geopotential height in the midlatitudes of the Northern Hemisphere has trended upward due to global warming since the late 1970s (He et al. ^{[40, 42]}; Wu and Wang ^[41]). In order to eliminate the impact of global warming, the eddy geopotential height is used to further examine the interdecadal change of summer atmospheric circulation over Asian mid-high latitudes around 1996, which is defined as a difference of geopotential height from its zonal mean. Fig. 9a and 9b show the composite difference of summer 700-hPa and 200-hPa eddy geopotential height between 1997-2015 and 1979-1995. A significant positive anomalous center appears near Lake Baikal, similar to that in Fig. 3b and Fig. 3d. This center near Lake Baikal has the same position as the significant change zone in Fig. 1d (R1, 92°E-115°E, 41° N-56° N). An anomalous wave train also appears over the mid-high latitudes of the Northern Hemisphere, similar to that in Fig. 8b. The standardized time series of summer eddy geopotential height averaged over the significant change zone near Lake Baikal (R1) are calculated. Fig. 9c shows the standardized time series of at 700 hPa and 200 hPa and they also display an interdecadal change around 1996. The correlation coefficients between the standardized time series of summer 700-hPa (200-hPa) geopotential height and 700-hPa (200-hPa) eddy geopotential height averaged over the area R1 is 0.84 (0.77) during 1979-2015, significant at 99.9% confidence level. The above results further show that the atmosphere circulation near Lake Baikal has undergone a significant interdecadal change around 1996.

Figure 9.  (a) Same as Fig. 8b but for summer 700-hPa eddy geopotential height (the contours; gpm). The box shows the significant variation center of R1 (41°N-56°N, 92°E-115°E) in Fig. 1(d). (b) Same as (a) but for 200-hPa eddy geopotential height (the contours; gpm). Shaded regions are significant at the 95% confidence level. (c) Standardized time series of summer eddy geopotential height averaged over R1 during 1979-2015.

It is worth mentioning that the above interdecadal variation features of the atmospheric circulation over Asian mid-high latitudes are obtained based on the ERA-Interim reanalysis. NCEP/DOE reanalysis data are used to carry out the same analyses. The atmospheric circulation over Asian mid-high latitudes based on NECP / DOE reanalysis shows similar interdecadal variation features to that from ERA-interim. In Fig. 10a, R1 and R2 (44° N-57° N, 96E° -115° E) are the significant interdecadal variation centers based on ERA-interim and NECP / DOE reanalysis data, respectively. They almost include the same areas. The standardized time series of summer 200-hPa geopotential height over R1 and R2 of the two reanalysis data are shown in Fig. 10b, which show consistent variation features during 1979-2015. The circulation index over Lake Baikal based on NCEP / DOE reanalysis is defined as the standardized time series of summer 200-hPa geopotential height over the R2 (NCEP-I200). During 1979-2015, it has a coefficient of 0.98 with the index (ERA-I200) based on ERA-interim. The result of Mann-Kendall test also shows that NCEP-I200 changed significantly around 1996. Thus, the interdecadal change in mid-late 1990s can also be captured by NECP /DOE reanalysis data. In addition, the anticyclonic circulation anomaly near Lake Baikal in recent 20 years (Fig. 10c) and the related anomalous wave train are also captured by NECP/DOE reanalysis data.

Figure 10.  (a) Same as Fig. 1(d), but for NECP/DOE reanalysis data. The box with dotted line is for the region R1 and the box with solid line shows the region R2 (44°N-57°N, 96°E-115°E). (b) Standardized time series of summer 200-hPa geopotential height over R1 and R2 based on NCEP/DOE and ERA-Interim. ERA-R1 stands for the time series over R1 from ERA-interim reanalysis data. NCEP-R1 (R2) represents the time series over R1 (R2) from NECP/DOE reanalysis data. (c) Same as Fig. 3(c), but for NECP/DOE reanalysis data.

In this study, an interdecadal change of atmospheric circulation over Asian mid-high latitudes in recent decades and the associated effects are investigated. The results show that the atmospheric circulation over mid-high latitudes of Asia has stronger interdecadal fluctuations than that over low latitudes and one intense change center is located near Lake Baikal. Further analysis shows that the atmospheric circulation near Lake Baikal had a significant interdecadal change in around 1996 and a deep anomalous anticyclonic circulation controls this region in recent 20 years. This anomalous anticyclone circulation has been contributing to the significant surface warming near Lake Baikal since 1997 and has made the region a remarkable warming center in Asia during recent 40 years. The summer precipitation pattern of less in north and more in south in east China in recent 20 years is also related to the anomalous anticyclonic circulation near Lake Baikal. Especially, the anomalous circulation near Lake Baikal has been found to contribute to the interdecadal change of the precipitation in northeast China and north China around late 1990s.

On the interdecadal time scale, the atmospheric circulation variation near Lake Baikal is closely related to a large-scale anomalous wave train over mid-high latitudes of the Northern Hemisphere in summer. The sea surface temperature (SST) of northwestern Atlantic is an important influence factor to the interdecadal change in the atmospheric circulation near Lake Baikal around 1996.

The warming of northwestern Atlantic in recent 20 years is suggested to be related to the Atlantic multidecadal oscillation (AMO) ^[39-43], which has an obvious phase transition from negative to positive around 1996 ^[43-44]. Therefore, the anomalous wave train in Fig. 8a may be triggered by the AMO-related SST anomaly in mid-high latitudes of the North Atlantic. It should be noted that the Atlantic SST is just one of the factors influencing the interdecadal change of summer atmospheric circulation over Asian mid-high latitudes in 1990s. Based on recent studies ^[45-51], the Silk Road Pattern (SRP), the Eurasian snow cover, the Arctic Oscillation and the Arctic sea ice also can affect the atmospheric circulation over Asian mid-high latitudes. The influences of these factors on the interdecadal change in the atmospheric circulation over Asian mid-high latitudes require further analysis in the future. In addition, it is worth mentioning that the anomalous wave train in Fig. 8a has significant anomaly centers over Europe, West Asia and Central Asia, which may affect the temperature, precipitation and dust in these regions ^[52-53]. Further analysis needs to be carried out regarding these regions.