Xingyng Cui ,Boyn Zhu ,Bo Sun,b,c,*
a Collaborative Innovation Center on forecast and Evaluation of Meteorological Disasters/Key Laboratory of Meteorological Disaster, Ministry of Education/Joint International Research Laboratory of Climate and Environment Change, Nanjing University of Information Science and Technology, Nanjing, China
b Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, China
c Nansen-Zhu International Research Centre, Institute of Atmospheric Physics, Chinese Academy of sciences, Beijing, China
Keywords: Boreal summer intraseasonal oscillation Intraseasonal variability Precipitation Yangtze River Valley
ABSTRACT Previous studies have demonstrated that boreal summer intraseasonal oscillation (BSISO) active in the Indian Ocean can affect precipitation over the Yangtze River Valley (YRV).In this study,the impacts of three types of BSISO modes on precipitation over the YRV were investigated and found to vary significantly.First,moisture budget analysis indicated that the process responsible for the precipitation anomalies in all BSISO modes is moisture convergence,while moisture advection is insignificant.In the canonical mode,precipitation anomalies are negative (positive) at pentad -2 (-1 to 3) owing to a BSISO-related cyclone over the northwestern Pacific(anticyclone moving from the subtropical western to northwestern Pacific) accompanied by enhanced (suppressed) convection propagating northeastward.Comparatively,in the eastward expansion mode,precipitation anomalies are positive (negative) at pentad -2 to 0 (1 to 3) via a BSISO-related anticyclonic (cyclonic) circulation over the northwestern Pacific induced by the teleconnection effect of enhanced (suppressed) convection over the Indian Ocean.Different from the previous two BSISO modes,in the northward dipole mode,precipitation anomalies are negative (positive) at pentad -2 to 0 (1 to 3) owing to a BSISO-related cyclonic (anticyclonic) circulation,which moves from the South China Sea to the northwestern Pacific because of the propagation of enhanced (suppressed) convection.
The Yangtze River Valley(YRV)is one of the most densely populated and economically developed regions in China.Summer precipitation over this region shows considerable intraseasonal variability with a period of 10–90 days,which can induce extreme precipitation events and lead to massive economic losses and human casualties(Li and Mao,2019a;Lu et al.,2014;Mao and Wu,2006;Yang et al.,2010).
The tropical intraseasonal oscillation is the most prominent intraseasonal signal across the globe,and it is regarded as an essential predictability source in extended-range forecasts (Hsu et al.,2015;Jones et al.,2000;Lee et al.,2017;Xavier et al.,2014).The boreal summer intraseasonal oscillation (BSISO) is primarily active in the tropical Indian Ocean and western Pacific and exhibits more complicated characteristics than its winter counterpart,the Madden–Julian Oscillation(Madden and Julian,1971,1972).
To better describe the characteristics of the BSISO,Lee et al.(2013)identified two kinds of BSISO,termed BSISO1 and BSISO2,by the leading modes of the multivariate empirical orthogonal functions of atmospheric variables.BSISO1 generally begins in the tropical Indian Ocean and propagates northward/northeastward with a period of 30–60 days,whereas BSISO2 originates from the western Pacific and propagates northwestward with a period of 10–20 days(DeMott et al.,2013;Hsu and Li,2012;Hsu and Weng,2001;Kikuchi et al.,2012).Hereafter,BSISO specifically indicates BSISO1.Previous studies found that BSISO has a high predictability (Ding et al.,2011;Lee et al.,2015).Understanding the impact of BSISO on precipitation over the YRV may therefore have implications for improving our ability to forecast extreme precipitation events in this region.
Although previous definitions of BSISO are useful for scientists to study the connection between it and summer precipitation over the YRV(Chen et al.,2015;Li et al.,2015,2019b),BSISO1 cannot depict all existing BSISO events.In 2020,a BSISO event originated from the Indian Ocean and influenced precipitation over the YRV.However,the convective activity of this BSISO event were persistent in the Indian Ocean and caused heavy and prolonged precipitation over the YRV(Zhang et al.,2021).Furthermore,a recent study identified the diversity in the propagation track of BSISO convection anomalies (Chen and Wang,2021),and they separated BSISO events into three types before then studying their mechanisms of propagation.Although it was found that these three modes exhibit different propagation features and may influence summer precipitation over the YRV in various ways,it is still unclear what effects these three different types of BSISO have on summer precipitation anomalies over the YRV.Thus,we investigated the impacts of the three BSISO types on summer precipitation over the YRV and the associated mechanisms.
Daily mean precipitation data were derived from the gridded CN05.1 dataset(Wu and Gao,2013)with a 0.25?×0.25?horizontal resolution.The daily averaged OLR data from the NCEP/NOAA interpolated OLR dataset(Liebmann and Smith,1996)with a 2.5?×2.5?resolution were used.Daily horizontal wind and specific humidity at eight pressure levels(1000,925,850,700,600,500,400,and 300 hPa)were obtained from the ERA5 reanalysis datasets(Hersbach et al.,2020)with a 0.5?×0.5?resolution.All data covered the period from 1979 to 2018.In this paper,summer is defined as May to August.
The anomalous intraseasonal variables were obtained as follows.First,the annual cycles of the original data were removed by subtracting the daily mean and the first three harmonics of the climatology(Nie and Sun,2022).Second,a 30–80-day bandpass Lanczos filter (Duchon,1979) was used on the daily data to extract the intraseasonal signal.Finally,statistical significance was determined based on two-tailed Student’st-tests.
K-means cluster analysis(Kaufman and Rousseeuw,2009)was used to classify the three BSISO modes(Chen and Wang,2021).In this study,BSISO events were identified on the basis of the region-averaged OLR anomalies over the equatorial Indian Ocean (5?S–10?N,80?–100?E)being below their mean minus one standard deviation for at least five successive days.Day 0 was defined as when the minimum was reached in each selected BSISO event.A total of 68 events were identified and then classified into three types usingK-means clustering.Table 1 shows the central dates of the three BSISO modes.There were 27,20,and 21 events identified in each cluster,respectively.

Table 1 Day 0 (year-month-day) of the selected BSISO events in the canonical mode,eastward expansion mode,and northward dipole mode.
Precipitation is closely correlated with the convergence of moisture flux,which can be divided into moisture convergence and moisture advection,and written as follows (Xie et al.,2022):
wherePis precipitation,qis specific humidity,Vis horizontal wind,primes denote the intraseasonal component,and angle brackets represent the vertical integral from 1000 to 300 hPa.
The propagation characteristics of the three BSISO modes,determined through the temporal and spatial evolution of OLR anomalies,are shown in Figs.1–3.In the canonical mode,northwest–southeast-oriented convection anomalies propagate from the Indian Ocean to the northwestern Pacific (contours in Fig.1),which resembles the propagation characteristics of the BSISO1 mode in Lee et al.(2013).In addition,convection anomalies in the eastward expansion mode expand eastward without any northerly propagation and eventually reach the Maritime Continent(contours in Fig.2).In the northward dipole mode,convection anomalies move northward in both the eastern Indian Ocean and the western Pacific;however,there is no eastward propagation in these two regions (contours in Fig.3).

Fig.1.Composite 850-hPa wind(vectors;units:m s-1),precipitation (shading;units: mm d-1),and OLR (contours;interval: 5 W m-2) anomalies from pentad-2 to pentad 3 for the canonical mode.The black frames in (a–f) represent the domain of the YRV(28?–34?N,110?–122?E).Only shading within(15?–40?N,95?–135?E) and vectors within(10?–40?N,90?–140?E) are shown.Stippling in(a–f) denotes where precipitation anomalies are significant at the 90%confidence level.Only those anomalous vectors and contours that exceeded the 90% confidence level are shown.

Fig.2.Composite 850-hPa wind(vectors;units:m s-1),precipitation (shading;units: mm d-1),and OLR (contours;interval: 5 W m-2) anomalies from pentad -2 to pentad 3 for the eastward expansion mode.The black frames in (a–f) represent the domain of the YRV (28?–34?N,110?–122?E).Only shading within(15?–40?N,95?–135?E)and vectors within (10?–40?N,90?–140?E) are shown.Stippling in (a–f) denotes where precipitation anomalies are significant at the 90% confidence level.Only those anomalous vectors and contours that exceeded the 90% confidence level are shown.

Fig.3.Composite 850-hPa wind(vectors;units:m s-1),precipitation (shading;unit: mm d-1),and OLR (contours;interval: 5 W m-2) anomalies from pentad -2 to pentad 3 for the northward dipole mode.The black frames in (a–f) represent the domain of the YRV (28?–34?N,110?–122?E).Only shading within(15?–40?N,95?–135?E)and vectors within (10?–40?N,90?–140?E) are shown.Stippling in (a–f) denotes where precipitation anomalies are significant at the 90% confidence level.Only those anomalous vectors and contours that exceeded the 90% confidence level are shown.
The life cycles of the YRV intraseasonal precipitation anomalies in the three BSISO modes are compared in Fig.4.As can be seen,precipitation anomalies in the canonical mode turn from negative to positive at pentad -1 and reach their peak at pentad 1.In contrast,precipitation anomalies in the eastward expansion mode change from positive to negative at pentad 1,while those in the northward dipole mode change from negative to positive at the same pentad.

Fig.4.(a) Composite evolution of intraseasonal precipitation anomalies (units: mm d-1) averaged over the YRV (28?–34?N,110?–122?E) in the canonical mode(red),eastward expansion mode (green),and northward dipole mode (blue).(b–d) Evolution of moisture convergence (unit: mm d-1) and moisture advection anomalies (units: mm d-1) averaged over the YRV.
To understand the processes responsible for YRV precipitation anomalies in the canonical mode,we first analyze the related moisture budget,including the moisture convergence and moisture advection.It is found that moisture convergence,instead of moisture advection,plays a key role in providing a favorable environment for precipitation anomalies(Fig.4(b)).
Fig.1 shows the temporal and spatial evolution of precipitation,horizontal wind,and OLR anomalies in the canonical mode.At pentad-2,anticyclonic circulation (vectors in Fig.1(a)) is situated over the southern South China Sea,accompanied by suppressed convection(contours in Fig.1(a)).At the same time,precipitation anomalies are negative because of this anticyclone,which cannot induce sufficient moisture transportation to the YRV (shading in Fig.1(a)).At the following pentad,this anticyclone moves northward and reaches the northern South China Sea(vectors in Fig.1(b))because of the northward propagation of convective anomalies (contours in Fig.1(b)).The anomalous southwesterly wind enhances moisture transportation to the YRV,which causes precipitation anomalies to turn positive on the northwest side of the YRV (shading in Fig.1(b)).Accompanied by the further northeastward propagation of convective anomalies(contours in Fig.1(c,d)),the anticyclone continues to move northeastward and intensify by pentad 0 to pentad 1 (vectors in Fig.1(c,d)),which is conducive to further moisture transportation to the YRV.During these two pentads,precipitation anomalies turn positive and reach their peak at pentad 1(shading in Fig.1(c,d)).By pentad 2 to pentad 3,the positive convection anomalies occurring over the South China Sea (contours in Fig.1(e,f)) induce cyclonic circulation,which causes the precipitation over the YRV to begin to decrease(shading in Fig.1(e,f)).Furthermore,this BSISO-related cyclone will continue to move northward until an anticyclone is triggered and begins to move northward from the South China Sea,denoting a new cycle of negative to positive precipitation anomalies over the YRV.
On the basis of moisture budget diagnosis(Fig.4(c)),we can see that YRV precipitation anomalies are modulated by the moisture convergence in the eastward expansion mode.In contrast,the moisture advection contributes marginally and negatively to the evolution of YRV precipitation anomalies.
Fig.2 shows the composite results for precipitation,850-hPa horizontal wind,and OLR anomalies in the eastward expansion mode.In this BSISO mode,convective anomalies are significant over the Indian Ocean but insignificant over the northwestern Pacific (contours in Fig.2(a)).Zhang et al.(2009) indicated that the energy propagation of Rossby waves triggered by convective anomalies over the Indian Ocean can influence the precipitation over the YRV,suggesting that the changes of circulation may be attributable to the same mechanism.At pentad -2,anticyclonic circulation (vectors in Fig.2(a)) occurs over the northwestern Pacific owing to the Rossby wave induced by enhanced convection over the Indian Ocean (contours in Fig.2(a)).Anomalous southwesterly wind occurs over the YRV,which is favorable for moisture transportation to the YRV.At pentad -1 to 0,enhanced convection(contours in Fig.2(b,c))continues to exist over the Indian Ocean,which induces the existence of the above anticyclone (vectors in Fig.2(b,c)).At the following pentads,suppressed convection (contours in Fig.2(d–f)) gradually replaces the enhanced convection and cyclonic circulation (vectors in Fig.2(d–f)) occurs over the northwestern Pacific because of the opposite effect of convective anomalies over the Indian Ocean.Anomalous northeasterly wind is triggered over the YRV,which inhibits moisture transportation to the region.Thus,precipitation is suppressed over the YRV(shading in Fig.2(d–f)).
The moisture budget diagnosis shows that YRV precipitation anomalies are modulated by the moisture convergence rather than the moisture advection in the northward dipole mode(Fig.4(d)).
The spatial and temporal evolution of precipitation,850-hPa horizontal wind,and OLR anomalies in the northward dipole mode is shown in Fig.3.At pentad -2,an anomalous cyclone (vectors in Fig.3(a)) is located over the western Pacific,accompanied by enhanced convection(contours in Fig.3(a)).Meanwhile,precipitation(shading in Fig.3(a))is suppressed because of anomalous northeasterly winds inhibiting moisture transportation to the YRV.As weakened convective anomalies move northward (contours in Fig.3(b,c)),the associated cyclonic circulation also moves northward and weakens in intensity (vectors in Fig.3(b,c)).At the same time,the inhibitory effect of the above cyclone on moisture transportation to the YRV is weakened,which can decrease the negative anomaly of precipitation over the YRV(shading in Fig.3(b,c)).At the following pentad,anticyclonic circulation (vectors in Fig.3(d)) occurs over the subtropical western Pacific at pentad 1,accompanied by suppressed convection(contours in Fig.3(d)).This anticyclone can provide sufficient moisture over the YRV,which induces the positive anomaly of precipitation over the region (shading in Fig.3(d)).After pentad 1,the associated anticyclonic circulation(vectors in Fig.3(e,f))heads north and intensifies from pentad 1 onward because of suppressed convection (contours in Fig.3(e,f)) continue to move northward.This anticyclone’s movement is responsible for the positive precipitation anomalies because it leads to the establishment of anomalous southwesterly winds,which can enhance moisture transportation to the YRV and provide a convergence environment.In addition,enhanced convection is following the suppressed convection over the Philippine Sea at pentad 2 (contours in Fig.3(e)),which denotes a new cycle of precipitation variation over the YRV.
Precipitation anomalies are also significant in southern China(shading in Fig.3).At pentad-2,cyclonic circulation(vectors in Fig.3(a)) is situated over the subtropical western Pacific,accompanied by enhanced convection (contours in Fig.3(a)).Anomalous northeasterly winds occur over southern China,which is unfavorable for moisture transportation to the region.Meanwhile,precipitation is suppressed.From pentad -1 to 0 (Fig.3(b,c)),weakened convection moves northward,accompanying weakened cyclonic circulation moving northward to the northwestern Pacific.At the same time,suppressed convection moves to the South China Sea,accompanying an anticyclonic circulation situated over the region.Southerly anomalies occur over the southern side of southern China,which is favorable for moisture transportation to the region.They converge over southern China with northerly anomalies to the western edge of the anomalous cyclone over the northwestern Pacific,which induces positive precipitation anomalies over the region.At the following pentads (Fig.3(d–f)),southerly anomalies strengthen over southern China and the anomalous anticyclone over the tropical western Pacific moves northward.Enhanced southerly anomalies increase moisture transportation to southern China,which is beneficial for precipitation.
In this study,the different impacts of the three BSISO modes defined by Chen and Wang (2021) on precipitation over the YRV were investigated.Using theK-means clustering analysis method,we obtained three types of BSISO based on an analysis of 68 events.First,the canonical mode,which accounted for 27 events,propagates northeastward from the tropical Indian Ocean to the northwestern Pacific.Then,the eastward expansion mode,which accounted for 20 events,expands eastward in the tropical Indian Ocean.And finally,the northward dipole mode,which accounted for 21 events,propagates northward in both the Indian Ocean and the western Pacific.
Results showed that the impacts of the three BSISO modes on precipitation over the YRV exhibit different features.In the canonical mode,YRV precipitation anomalies are negative at pentad -2 and positive during pentad -1 to pentad 3.The moisture budget indicated that moisture convergence,instead of moisture advection,plays a key role in precipitation anomalies.During the dry phase,cyclonic circulation accompanied by enhanced convection is situated over the northwestern Pacific.Anomalous northeasterly winds inhibit moisture transportation to the YRV.During the wet phase,the enhanced effect of the BSISOrelated anticyclone accompanied by suppressed convection propagating northeast on moisture transportation to the YRV is favorable for enhanced precipitation.In contrast,in the eastward expansion mode,YRV precipitation anomalies turn from positive to negative at pentad 1.The moisture budget indicated that moisture convergence is more important than moisture advection,which has a small and negative effect.By pentad-2 to 1,enhanced convection over the Indian Ocean can trigger a Rossby wave and further induces an anticyclone over the northwestern.Anomalous southwesterly wind transports sufficient moisture to the YRV.Thus,precipitation is enhanced.By pentad 2 to 3,suppressed convection over the Indian Ocean plays an opposite role,which induces a cyclone over the northwestern Pacific.This cyclonic circulation is unfavorable for moisture transportation to the YRV.Thus,precipitation is suppressed.Compared with the former two BSISO modes,in the northward dipole mode,YRV precipitation anomalies are negative before pentad 1 and positive after pentad 1.The related moisture budget suggested that moisture convergence is the main process,while moisture advection plays a marginal and negative role.At pentad -2 to 0,the suppressed moisture transportation to the YRV,which is induced by cyclonic circulation accompanied by weakened and northward propagating convection anomalies,leads to suppressed precipitation over the region.By the following pentads,suppressed convection replaces the enhanced convection and moves northward over the western Pacific.At the same time,the associated anticyclonic circulation enhances moisture transportation to the YRV,which can lead to enhanced precipitation.
Authorship Contribution Statement
All authors contributed to the concept and design of the study.The data analysis and first draft of the manuscript were carried out and written by Xiangyang Cui.Baoyan Zhu and Bo Sun helped to edit the paper.All authors read and approved the final manuscript.
Declaration of Competing Interest
The authors declare no competing interests.
Funding
This study was funded by the National Key Research and Development Program of China [grant number 2022YFF0801704].
Atmospheric and Oceanic Science Letters2023年5期