Strengthened influence of the East Asian trough on spring extreme precipitation variability over eastern Southwest China after the late 1980s

Yting Nn ,Jinqi Sun ,Mengqi Zhng ,Hixu Hong,c ,Junpeng Yun

a Department of Atmospheric Sciences, Yunnan University, Kunming, China

b Nansen-Zhu International Research Centre, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

c University of Chinese Academy of Sciences, Beijing, China

Keywords:Spring extreme precipitation Southwest China East Asian trough

ABSTRACT The relationship between variations in the East Asian trough (EAT) intensity and spring extreme precipitation over Southwest China (SWC) during 1961–2020 is investigated.The results indicate that there is an interdecadal increase in the relationship between the EAT and spring extreme precipitation over eastern SWC around the late 1980s.During the latter period,the weak (strong) EAT corresponds to a strong and large-scale anomalous anticyclone (cyclone) over the East Asia–Northwest Pacific region.The EAT-related anomalous southerlies (northerlies)dominate eastern SWC,leading to significant upward (downward) motion and moisture convergence (divergence)over the region,providing favorable (unfavorable) dynamic and moisture conditions for extreme precipitation over eastern SWC.In contrast,during the former period,the EAT-related circulation anomalies are weak and cover a relatively smaller region,which cannot significantly affect the moisture and dynamic conditions over eastern SWC;therefore,the response in extreme precipitation over eastern SWC to EAT is weak over the period.The interdecadal change in the relationship between eastern SWC spring extreme precipitation and the EAT could be related to the interdecadal change in the EAT variability.The large (small) variability of the EAT is associated with significant (insignificant) changes in spring extreme precipitation over eastern SWC during the latter(former) period.

1.Introduction

Southwest China (SWC) is located to the north of the Indian Ocean–South China Sea and on the southeastern flank of the Tibetan Plateau.The complex terrain and abundant moisture make SWC vulnerable to extreme precipitation events.The secondary disasters associated with extreme precipitation,such as floods and landslides,bring a tremendous threat to human life and cause great financial losses in the region.Additionally,spring is the critical time for crop growth in SWC,and extreme precipitation events during this time have impacts on agricultural production and further restrict regional economic development.For instance,the floods in Sichuan Province during May 2007 left 23 people killed and 10 people missing and resulted in huge economic losses(http://news.sina.com.cn/c/2007-05-27/052011901209s.shtml).The heavy rainfall struck northeastern Sichuan Province on 18 April 2014,leading to direct economic costs up to 100 million Yuan(http://history.thecover.cn/shtml/hxdsb/20140421/203405.shtml).Therefore,it is of great importance to investigate the variability and influencing factors of spring extreme precipitation over SWC.

Previous studies have mainly focused on the spring seasonal-mean precipitation over SWC.The results of empirical orthogonal function(EOF) and rotated EOF analysis suggest that the interannual variability of spring precipitation features an out-of-phase change over the eastern and western parts of SWC (Fan and Wang,2015 ;Liu et al.,2015 ;Xia et al.,2016 ;Li et al.,2020).Several atmospheric circulation factors have been revealed to influence SWC spring precipitation variability,including the India–Burma trough and western Pacific subtropical high(Fan et al.,2016),the Arctic Oscillation (Huang et al.,2012 ;Yang et al.,2012),North Atlantic Oscillation (Li et al.,2008 ;Feng et al.,2014),and Arabian monsoon (Jing et al.,2021).Additionally,sea surface temperature anomalies over the North Atlantic (Li et al.,2018),Pacific and Indian Oceans (Feng et al.,2014 ;Yu et al.,2015 ;Li et al.,2021) also have significant correlations with SWC spring precipitation.Recently,Wen et al.(2021) reported that the early spring precipitation over the southeastern edge of the Tibetan Plateau experienced an interdecadal decrease in the late 1990s,and they further related it to the covariability of the Pacific Decadal Oscillation,Atlantic Multidecadal Oscillation,and Indian Ocean Basin Mode.

As an important circulation system in the midlatitude Northern Hemisphere,the East Asian trough (EAT) modulates the climate variability over East Asia significantly (e.g.,Zhang et al.,1997 ;Leung et al.,2015).Some studies have revealed a close connection between spring precipitation over China and the EAT (Lu,2001 ;Guo et al.,2006 ;Lu et al.,2014 ;Fan et al.,2016 ;Zhang and Sun,2018).For instance,Lu (2001) pointed out that an eastward-shifted and weakened EAT favors the occurrence of spring precipitation in North China.Guo et al.(2006) suggested that increased spring precipitation over northwestern China is associated with a weaker EAT and stronger European trough.Lu et al.(2014) related the decreased spring precipitation over the middle and lower reaches of the Yangtze River basin during spring 2011 to the strengthened EAT.

Compared to the spring-mean precipitation,spring extreme precipitation events give rise to more extensive influences and serious losses.Additionally,heavy rainfall in spring shows an increasing trend over SWC against the background of global warming (Luo et al.,2015).However,few investigations have been made into the variability and physical mechanisms of spring extreme precipitation over SWC from a climate perspective.There are limited case studies in which the spring rainstorm processes have been diagnosed from the perspective of the synoptic scale.For example,two rainstorms that occurred in eastern SWC during May of 2012 and 2013 were closely related to the eastward shift of the southwest vortex and plateau vortex (Hu et al.,2016 ;Zhang et al.,2016).The heavy rainfall over the northeastern Sichuan Basin in April 2019 can be attributed to the interaction between the upper-level and lower-level jet (Gao et al.,2020 ;Zhang et al.,2020).Based on the aforementioned studies,two questions arise: Does the EAT influence the SWC spring extreme precipitation variability? If so,is their relationship stable? The present study seeks to answer these two questions.

2.Data and methods

The observed precipitation data at 699 gauges,provided by the China Meteorological Administration,are employed in this study.The gauge data are removed when the missing values exceed 5% during spring (March–April–May) from 1961–2020,meaning 95 gauges are ultimately selected over SWC.The monthly Japanese 55-year reanalysis dataset (JRA-55;Kobayashi et al.,2015),obtained from the Japan Meteorological Agency,is used to investigate the atmospheric circulation anomalies.The JRA-55 dataset has a horizontal resolution of 1.25° × 1.25° and 37 pressure levels from 1 hPa to 1000 hPa during 1958–2020.

Given the complex terrain and large spatial difference in precipitation variability over SWC,extreme precipitation is defined based on a relative threshold method (Zhai and Pan,2003).The threshold of a rain gauge is defined by the 95th percentile of precipitation on rainy days(daily precipitation more than 0.1 mm) during spring for the period 1961–2020.An extreme precipitation day is identified when the daily precipitation exceeds the threshold at the corresponding gauge.Based on the definition of the Expert Team on Climate Change Detection and Indices,spring R95p index,which denotes the accumulated precipitation amount on the extreme precipitation days,is used to represent the spring extreme precipitation.

Linear trends are removed from all variables to eliminate the effect of long-term variation for the corresponding period.The sliding standard deviation and sliding correlation are calculated to investigate the interdecadal change in EAT interannual variability and the relationship between the EAT and SWC spring extreme precipitation,respectively.The Student’st-test is used to examine the statistical significance.

3.Results

The variation in spring R95p over SWC exhibits a strong regional feature,with large values and variability over eastern SWC (figure not shown).This result indicates that the spring extreme precipitation variability over eastern SWC represents the predominant variability of SWC.Therefore,the gauge-averaged spring R95p over eastern SWC (east of 103.75°E;R95pI) is defined and used in this paper to represent the spring extreme precipitation variability over eastern SWC during 1961–2020.

Fig.1.Regressions of geopotential height (shading;units: gpm) and horizontal wind (vectors;units: m s-1) anomalies at (a) 200 hPa,(b) 500 hPa,and (c) 850 hPa upon spring R95pI during 1961–2020.The dotted areas and vectors are significant at the 95% confidence level.The blue lines indicate the provincial boundaries of SWC.

Fig.2.(a) Time series of normalized spring R95pI (blue line) and EATI (red line).The horizontal dashed lines denote ± 0.8 standard deviation.(b) Sliding correlation between spring R95pI and EATI during 1961–2020,with window lengths of 19,21,and 23 years.Dashed lines represent the 95% confidence level for the corresponding window length.

Fig.1 shows the regressions of geopotential height and wind anomalies at 200 hPa,500 hPa,and 850 hPa,against the spring R95pI during 1961–2020.Corresponding to positive R95pI,significant positive geopotential height anomalies are located over the midlatitude East Asian coast from the lower to upper level.This region is the location of the EAT(Wang and He,2012 ;Song et al.,2016),indicating a close connection between spring extreme precipitation variability over eastern SWC and the EAT during 1961–2020.Based on Fig.1,the area-averaged 500-hPa geopotential height anomalies over the region (25°–45°N,110°–145°E)is defined as an EAT intensity index (EATI),consistent with previous studies (Wang and He,2012 ;Yu and Sun,2020).A positive (negative)EATI indicates a weakened (strengthened) EAT.The correlation coeffi-cient between spring EATI and R95pI is 0.45 during 1961–2020,significant at the 99% confidence level.This result indicates that a weakened(strengthened) EAT is closely related to increased (decreased) spring extreme precipitation over eastern SWC.

Fig.2 (a) depicts the normalized time series of spring EATI and R95pI during 1961–2020.The two time series show different variations over different subperiods,indicating that there could be an unstable relationship between the two.In order to further check their relationship,sliding correlation analysis is applied to the spring EATI and R95pI during 1961–2020.As shown in Fig.2 (b),the sliding correlations with different window lengths consistently display an abrupt increase around the late 1980s,suggesting an interdecadal enhanced relationship between spring extreme precipitation over eastern SWC and the EAT intensity after the late 1980s.According to Fig.2 (b),the whole period is divided into two subperiods: 1972–1988 (P1) and 1989–2020 (P2).The correlation coefficient between spring EATI and R95pI is only 0.02 during P1,while it increases to 0.56 during P2,exceeding the 99% confidence level,further confirming the strengthened relationship between the EAT and extreme precipitation over eastern SWC.

To understand the physical connection between the EAT and extreme precipitation over eastern SWC,the EAT-related atmospheric circulations are studied using composite analysis.Firstly,the EATI is normalized over the two subperiods,separately.Then,a weak (strong) EAT year is selected when the EATI is greater than (less than) 0.8 (-0.8).Fig.3 (a–c) displays the composite vertically integrated water vapor flux and related divergence for weak and strong EAT years and their difference during P2.Corresponding to a weak EAT,a significantly largescale anomalous anticyclone is located over the East Asia–Northwest Pacific region (Fig.3 (a)).Accordingly,strong anomalous northward water vapor fluxes prevail over the region from South China to South Japan,leading to significant moisture convergence over the region.The eastern SWC is also dominated by northward water vapor transport and moisture convergence.Meanwhile,corresponding to a weak EAT,significantly anomalous upward motions are found over eastern SWC(Fig.3 (d)).Such dynamic and moisture conditions are favorable for extreme precipitation over eastern SWC.

The situation is reversed during strong EAT years.There is an anomalous cyclone over the East Asia–Northwest Pacific region,although the strength is weaker relative to that in weak EAT years (Fig.3 (b)).Eastern SWC is dominated by southward water vapor flux and moisture divergence.In addition,there are significantly anomalous downward motions over eastern SWC (Fig.3 (e)).Such circulations provide unfavorable moisture and dynamic conditions for extreme precipitation over the region.

The comparison of the atmospheric circulations between weak and strong EAT years indicates a generally symmetrical impact of the EAT on the extreme precipitation variability over eastern SWC during P2.Therefore,the circulation anomalies between weak and strong EAT years are also significant,with a strong and broad anticyclone over East Asia to the Northwest Pacific (Fig.3 (c)).The southerlies along the western flank of the anticyclone could influence eastern SWC,leading to significant moisture convergence and upward motion over the region (Fig.3 (c,f)).Therefore,during P2,the EAT can significantly influence the extreme precipitation variability over eastern SWC.

Fig.3.Composite anomalies of (a–c) vertically integrated water vapor flux (vectors;units: kg m-1 s-1) and divergence (shading;units: 10-6 kg m-2 s-1) and (d–f)latitude–pressure cross section of vertical velocity (shading;units: 10-2 Pa s-1) and meridional circulation (vectors;meridional wind units: m s-1) along 103.75°–110°E during 1989–2020.(g–l) as in (a–f) but for 1972–1988.The left (middle) column is for weak (strong) EAT years,and the right column is the difference between weak and strong EAT years.The dotted areas are significant at the 90% confidence level.The purple (gray) vectors in (a–c) and (g–i) are significant (insignificant)at the 90% confidence level.The vectors in (d–f) and (j–l) are significant at the 90% confidence level.The red dashed lines indicate the region of eastern SWC.

In contrast,during P1,the EAT-related anomalous anticyclone and cyclone over the East Asia–Northwest Pacific region are relatively weaker and narrower,which cannot influence eastern SWC (Fig.3 (g–i)).Corresponding to an anomalous EAT,the vertical motion and moisture anomalies are weak over eastern SWC (Fig.3 (g–l)).Therefore,during P1,there is a weak relationship between the EAT intensity and extreme precipitation variability over eastern SWC.

The aforementioned analysis implies a weak (strong) influence of the EAT on the dynamic and moisture conditions over eastern SWC during P1 (P2).Furthermore,the composite differences in the average moisture and dynamic conditions over eastern SWC between weak and strong EATI years during P1 and P2 are calculated.The results show that the magnitude of the anomalies of all variables increases in P2 compared with P1: the meridional component of vertically integrated water vapor flux anomalies increase from 8.36 kg m-1 s-1 to 23.03 kg m-1 s-1 ;the divergence of vertically integrated water vapor flux anomalies increase from 0.2 ×10-6kg m-2s-1to -7.5 ×10-6kg m-2s-1;the precipitable water anomalies increase from 0.07 kg m-2 to 1.66 kg m-2 ;and the 500-hPa vertical velocity anomalies increase from -0.09 ×10-2Pa s-1to -0.55 ×10-2 Pa s-1.These results further indicate that the changes in the variability of the EAT can significantly change the magnitude of the moisture transport,moisture amount,and vertical motion over eastern SWC.

4.Discussion

The results presented above indicate a strengthened influence of the EAT intensity on spring extreme precipitation variability over eastern SWC after the late 1980s.In this section,we further discuss the possible reason for the interdecadal change.From Fig.2 (a),we can see that the EAT shows a large (small) variability during P2 (P1),consistent with high (low) correlation between the EATI and R95pI.This result indicates that the EAT variability could be related to its relationship with extreme precipitation variability over eastern SWC.

To confirm this suggestion,Fig.4 further shows a scatterplot of the sliding correlation between R95pI and EATI against the sliding standard deviation of EATI,with a 21-year window length during 1961–2020.The figure suggests that,over the subperiods with a strong (weak) EAT variability,the EAT generally shows significant (insignificant) correlations with extreme precipitation over eastern SWC.In addition,we calculate the same sign rate between EATI and R95pI over the period 1961–2020.Consistent with the composite analysis,the value of 0.8 is used to select anomalous EATI years.In the years with EATI greater than (less than) 0.8 (-0.8),there is 83.3% (85.7%) with more (less) extreme precipitation over eastern SWC,and notably,this kind of situation mainly appears over the period 1989–2020.Meanwhile,when the EATI is between 0.8 and -0.8,the same sign rate between the EATI and R95pI is about 50%.This result further implies that the enhanced relationship between the EATI and R95pI around the late 1980s could be related to the interdecadal increase in the EAT variability.

Fig.4.Scatterplot of the sliding correlation between R95pI and EATI against the sliding standard deviation of EATI.The window length of the sliding correlation and sliding standard deviation is 21 years.The horizontal dashed line indicates the 90% confidence level.The vertical dashed line indicates the mean of 21-year sliding standard deviations of EATI during 1961–2020.The corresponding year for each dot is marked in the figure.

The composite analysis in Fig.3 indicates that,corresponding to a large (small) EAT variability during P2 (P1),the influencing area of the EAT is large (small).During P2,the EAT-related anomalous anticyclone and cyclone over the East Asia–Northwest Pacific region are stronger and wider.The anomalous southerlies (northerlies) along the western flank of the EAT-related anomalous anticyclone (cyclone) extend westward,leading to significant changes in vertical motion and moisture conditions over eastern SWC and consequently exerting significant influence on the extreme precipitation over the region.In contrast,during P1,the EAT-related anomalous anticyclone and cyclone are relatively weak and narrow,which is only confined to the eastern China–Northwest Pacific region and cannot influence the extreme precipitation over eastern SWC.

To reflect the changes more clearly in the influencing area of the EAT,Fig.5 (a) shows a longitude–time cross section of 21-year sliding regressions of 700-hPa meridional wind averaged over the eastern SWC zone (22°–35°N),against the spring EATI during 1961–2020.For comparison,Fig.5 (b) further displays the 21-year sliding standard deviation of spring EATI.It can be seen that a large (small) variability of the EAT is consistent with a large (small) influencing area,manifested by strong(weak) and westward-extended (eastward-retreated) low-level meridional wind anomalies.Corresponding to large EAT variability,the significant signals can extend westward to 103.75°E,covering the region of eastern SWC,which consequently results in a significant relationship between the EAT and R95pI.Meanwhile,corresponding to small EAT variability,the significant signals are generally confined to the east of 107.5°E,with weak anomalies over most of eastern SWC,and thus the EAT has a weak relationship with eastern SWC R95pI.

The results indicate that the EAT variability could play an important role in its relationship with the extreme precipitation variability over eastern SWC,through modulating the size of the influencing area.However,the reason for the decadal amplification of the EAT variability is still not clear and needs to be investigated further in the future.

Fig.5.(a) Longitude–time cross section (22°–35°N averaged) of 21-year sliding regressions of 700-hPa meridional wind (shading;units: m s-1),against the spring EATI during 1961–2020.The dotted areas are significant at the 95% confidence level.The horizontal dash lines indicate the region of eastern SWC.(b)The 21-year sliding standard deviation of spring EATI.The horizontal dash line indicates the mean of 21-year sliding standard deviations of EATI during 1961–2020.

5.Conclusion

Extreme precipitation over SWC often cause landslides,which poses severe threats to human life and causes great economic losses.In addition,extreme precipitation in spring also significantly affects agricultural activity and production over SWC.Therefore,it is of significance to explore the mechanism for the variability of extreme precipitation over SWC.

Our analysis in this study indicates that the EAT has an enhanced influence on spring extreme precipitation variability over eastern SWC after the late 1980s.After the late 1980s,corresponding to a weak (strong)EAT,there is a strong and broad anomalous anticyclone (cyclone) over the East Asia–Northwest Pacific region.The southerlies (northerlies)along western flank of the anticyclone (cyclone) dominate eastern SWC,leading to significant upward (downward) motion and moisture convergence (divergence) over the region.Such anomalous atmospheric circulations provide favorable (unfavorable) moisture and dynamic conditions for extreme precipitation over eastern SWC.In contrast,before the late 1980s,the EAT-related anomalous anticyclone/cyclone is weaker and narrower,which cannot exert any significant influence on moisture and dynamic conditions over eastern SWC.Consequently,the EAT shows a weak relationship with extreme precipitation variability over the region.

The interdecadal change in the relationship between the EAT and spring extreme precipitation over eastern SWC could be related to the interdecadal change in the EAT variability.Large (small) EAT variability corresponds to strong (weak) and large-scale (small-scale) atmospheric circulation anomalies over the East Asia–Northwest Pacific region,consequently leading to significant (insignificant) impacts on spring extreme precipitation variability over eastern SWC.

Funding

This research was jointly supported by the National Natural Science Foundation of China [grant number 41825010] and the Strategic Priority Research Program of the Chinese Academy of Sciences [grant number XDA23090102].