Response of extreme precipitation to increasing extratropical cyclonic vortex frequency

Jie Zhang,Jiang Liu

Key Laboratory of Meteorological Disaster, Ministry of Education (KLME) /Joint International Research Laboratory of Climate and Environment Change (ILCEC)/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters (CIC-FEMD), Nanjing University of Information Science and Technology,Nanjing, China

Keywords:Extratropical cyclonic vortex Quasi-stationary vortex Synoptic-scale transient eddy Extreme precipitation

ABSTRACT Since the 2000s,extratropical extremes have been more frequent,which are closely related to anomalies of planetary-scale and synoptic-scale systems.This study focuses on a key synoptic system,the extratropical cyclonic vortex (ECV) over land,to investigate its relations with extreme precipitation.It was found that ECVs have been more active post-2000,which has induced more extreme precipitation,and such variation is projected to persist along with increasing temperature within 1.5°C of global warming.An enhanced quasi-stationary vortex(QSV) primarily contributes to the ECV,rather than inactive synoptic-scale transient eddies (STEs).Inactive STEs respond to a decline in baroclinicity due to the tendency of the homogeneous temperature gradient.However,such conditions are helpful to widening the westerly jet belt,favoring strong dynamic processes of quasi-resonant amplification and interaction of STEs with the quasi-stationary wave,and the result favors an increasing frequency and persistence of QSVs,contributing to extreme precipitation.

1.Introduction

Since the beginning of the 21st century,extreme weather and climate events have occurred frequently in boreal summer in the Northern Hemisphere (Coumou and Rahmstorf,2012 ;Zhang et al.,2019).Besides the serious temperature extremes such as heatwaves that have frequently occured in North America (2011,2014,2016,and 2021),Europe (2003,2019,2020,and 2021),Russia,and Japan (2010)(Stott et al.,2004 ;Hong et al.,2011 ;Johnson et al.,2018),serious drought/flood extremes have also occurred in Europe (2016/17)(García-Herrera et al.,2019) and China (2009/10,2012,2016,and 2021).In addition,strong convective thunderstorms,serious gales and hail were frequent in China in 2021.In August 2020,heatwaves with temperatures higher than 44°C occurred in western Europe,while extreme floods resulted in 28 fatalities in Korea,and frequent floods,hailstorms,thunder,and lightning frequently hit northern China.Frequent extremes and strong convection have caused considerable damage worldwide.Thus,the related circulations need to be identified for weather forecasting and hazard prevention,especially in those regions without sufficient adaptation strategies.

Extratropical extremes are directly related to synoptic systems including extratropical vortex anomalies,blocking,etc.,and they overlap and interact with quasi-stationary waves and planetary-scale transient waves.For instance,Eurasian heatwaves are often produced under anomalous circulations characterized by an anticyclonic vortex,persistent blocking (Schneidereit et al.,2012 ;Matsueda,2011),and weakening of the zonal mean jets (Orsolini and Nikulin,2006).Apart from the predominant effect of large-scale zonal winds,the extratropical anticyclonic vortex is another significant contributor to the formation and maintenance of blockings (Shutts,1983).In addition,most extratropical floods are directly related to an extratropical cyclonic vortex (ECV) anomaly,including a quasi-stationary vortex (QSV) and synoptic-scale transient eddy (STE).Recent studies suggest that STEs could drive Rossby wave breaking (Woollings,2011 ;Novak et al.,2015 ;O’Reilly et al.,2016,2017),thereby blocking the flank of the eddy-driven jet (Pelly and Hoskins,2003 ;Woollings et al.,2008).Ren et al.(2008) suggested that a meridional shift in the cyclonic vortex favors a meridional shift in the subtropical jet over East Asia via the convergence of momentum eddy fluxes (Ren and Zhang,2007 ;Xiang and Yang,2012),developing and organizing mesoscale convective systems (Liang and Wang,1998 ;Park et al.,2015),thereby resulting in intense rainfall.The ECVs over northeastern and northern China,called the Northeast Cold Vortex and North Cold Vortex,respectively(Sun et al.,2002 ;Xie and Bueh,2015),are favorable for rainstorms,thunder,lightning,and hailstorms (Sun et al.,2002 ;Wang et al.,2012 ;Zhao and Sun,2007).The ECV in northern China causes 20%–60% of precipitation in the warm season and 53% (22.4%) of hailstorms (floods)(Zhang et al.,2008).The ECV in the lower and middle levels over central Asia,called the Central Asian Vortex,is one of dominent synoptic systems causing rainstorms,snowstorms,and persistent low temperatures.The Central Asian Vortex caused more than 61% of 116 intense precipitation events from 1970 to 1999 (Yang and Zhang,2014) and two of the heaviest precipitation events.

ECVs also influence and interact with low-frequency circulation.They tend to reinforce monthly mean wave anomalies (Lau and Nath,1991) through vorticity flux divergence and heat transport.STE feedbacks may be modified into an equivalent barotropic response(Magnusdottir et al.,2004),which in turn reinforces the midlatitude wave pattern,thus,STE heat transport dissipates low-frequency temperature variability (Lin and Derome,1995).Conversely,the tropical vortex over the western North Pacific drives the anomalous meridional Pacific–Japan/East Asian Pacific teleconnection (Ren et al.,2008) and thereby enhances the Northeast Vortex,which also interacts with the ECV in the northeast of China.Because of the significant link between ECVs and precipitation,their development and variation are important for weather forecasts.

ECV heat flux results from the enhanced baroclinicity (Novak et al.,2015 ;O’Reilly et al.,2017),which is a response to inhomogeneous diabatic heating (Lee,1997) and the internal variability of the atmosphere.Against the background of global warming,significant surface anomalies have occurred,such as Arctic sea-ice loss,reduced Eurasian snow cover (Zhang et al.,2020 ;Wu et al.,2016),and variation in soil moisture (Chen et al.,2019),which result in large daily to subseasonal variation in diabatic heating over extratropical land areas and atmospheric baroclinicity.Such forcings raise some interesting questions: What is the variability of ECVs besing response to land thermal anomalies? How are ECVs related to the frequency of extreme events? What is the likely future trend of ECVs? Observational and model analyses were conducted in the present study to address these questions.

2.Data and methods

Typically,STEs are defined as the high-pass 2–8-day Lanczos filtered 500-hPa geopotential height (Lau and Nath,1991).For those cyclonic vortexes without an STE,i.e.,QSVs,they are defined as the cyclonic vortex at 500 hPa with a minimum center and at least one closed contour with an interval of 40 gpm from the center within the range of 10°×7°.The total extreme precipitation is defined as the total precipitation rate being equal to or greater than 10 mm d-1in every grid cell.The statistical significance of the linear regression coefficient,anomaly field,and the correlation between the two series was assessed using a Monte Carlo test.

Daily circulation data were obtained from ERA-Interim (http://apps.ecmwf.int/datasets/).Daily precipitation was acquired from NOAA’s Climate Prediction Center (CPC) (http://www.cpc.ncep.noaa.gov/data/teledoc/telecontents.html).

The vortices in the future projections of CMIP6 climate models (https://esgf-node.llnl.gov/search/cmip6/) under the emissions scenario with intermediate greenhouse gas emissions (SSP2-4.5) and very high greenhouse gas emissions (SSP5-8.5) were used in this study.A seven-model ensemble was employed.The individual models were:the Beijing Climate Center Climate System Model;the European Consortium Earth System Model;the Institute for Numerical Mathematics Climate Model;the L’Institute Pierre-Simon Laplace Coupled Model;the Max Planck Institute Earth System Model;the Meteorological Research Institute Earth System Model;and the Norwegian Earth System Model.

3.Results

3.1. A case of ECV distribution

In 2020,the cutofflow and blocking high were active,indicating a persistent vortex disturbance,which was accompanied by weather extremes.Fig.1 shows the geopotential height over 500 hPa at 1200 UTC 9 August 2020 and the precipitation larger than 10 mm d-1on 8–9 August 2020 retrieved from CPC daily data.This result reflects extreme and intense precipitation over northern America (NA),southern Europe to the Mediterranean (Eu-Med),Central Asia (CA),and East Asia (EA) on 8–9 August 2020 being similar to the observed result.Reports show that extreme floods resulted in 28 deaths in Korea on 8 August 2020.Subsequently,strong convective weather occurred frequently in northern China on 8–9 August 2020,accompanyied by lasting floods,hailstorms,thunder,and lightning.The 500-hPa geopotential height exhibits significant meridional anomalies such as a cutofflow and blocking high.The ECVs exhibited by the cutofflow have four significant centers over land,and corresponding to those four centers,the precipitation is larger than 10 mm d-1,revealing the close relationship between extreme precipitation,the ECV,and quasi-stationary wave.In addition,the anticyclonic vortex corresponds to a blocking high.The geopotential height distribution exhibits the ECV and blocking high distribution,which strengthens the likelihood of weather extremes.

This study defines four centers,named NA (35°–60°N,110°–60°W),Eu-Med (35°–45°N,10°–45°E),CA (35°–60°N,50°–80°E),and EA (35°–60°N,110°–150°E),to analyze the statistical characteristics of the ECV.The EA region covers the most active position of the Northeast Cold Vortex (Sun et al.,2002),and the CA region covers the most active position of the Central Asia Vortex (Yang and Zhang,2017),most of which generally result from the development of planetary-scale waves and QSV.However,a highly frequent ECV indicates the significant effect of STEs on the quasi-stationary wave.QSVs and STEs are the significant parts as for the Northeast Cold Vortex and the Central Asia Vortex.The non-uniform QSV distribution possibly relates to the topographic effect,strong surface friction,land–ocean distribution and internal variability(Son et al.,2009).Fig.1 shows the monthly variations of the ECV frequencies over the four defined regions averaged in every 1°×1°grid cell in 1979–2018.The ECV mainly occurs from mid-spring to mid-summer(April to July) in NA and EU-Med;however,it mainly occurs in late spring and summer (May to September) in CA and EA,which follows the active period of the Northeast Cold Vortex (Xie and Bueh,2015).Therefore,the monthly ECV is frequent in spring and summer and is worthy of exploration.

3.2. Spatial distribution of ECVs in the Northern Hemisphere

The most frequent positions of ECVs are over the North Atlantic and North Pacific;however,they are also active over land.The latitude–time and time-mean distributions of ECVs over land in JA (July–August) are shown in Fig.2 (a,b),respectively.It is active between 48°N and 65°N,and the ECV frequency is larger than 10 d in JA in 1979–2018;however,it is less than 5 d at latitudes lower than 48°N.It is between 5–10 d at latitudes higher than 65°N.The ECV distribution with time reveals highfrequency ECVs concentrated in the midlatitude belt between 50°N and 60°N,with the largest variability.Given the significant abrupt occurrence of ECVs in 1998,by comparing the ECVs in 1979–1998 with 1999–2018,we find that the ECV frequency increases,which should garner more attention.Fig.2 (c) shows the spatial distribution of the 40-yr vortex frequency for one grid cell,revealing active vortex over the Eurasian than the North American continent.There are four highly frequent ECV regions.Compared with the climatological state,the ECV distribution is active in the wave trough of the climatological quasi-stationary wave,revealing that the meridional development of quasi-stationary waves interacts with the formation of a low cutoffand ECV.Previous studies have suggested that blocking circulation over the Ural Mountains and the Sea of Okhotsk are favorable to wave troughs around Lake Baikal and the Northeast Cold Vortex (Wang and Lupo,2009),a type of mixed STE and QSV.According to upstream effects,blocking circulation over the Ural Mountains favors a wave trough around Balkhash Lake and the Central Asia Vortex (Yang and Zhang,2017).These results show that the formation of the ECV is very similar to the formation of the QSV over the wave trough position.Conversely,when the ECV is imposed on the wave trough,it will deepen the wave trough of the quasi-stationary wave.Previous studies have shown that a negative-phase west Pacific teleconnection pattern (Qu and Huang,2012) favors ECV disturbance and the Northeast Cold Vortex (Xie and Bueh,2015) over the EA region.Tibetan Plateau heating and the Indian monsoon also stimulate a tripole pattern disturbing the ECV over northern China and the Northeast Cold Vortex.

Corresponding to an active ECV,the total precipitation with a rate greater than or equal to 10 mm d-1is high over the three regions(Fig.2 (d)).The most significant extreme precipitation occurs in EA,followed by NA and EU-Med.The results reveal a close relationship between ECVs with extreme precipitation and other extreme events.Apart from these three regions,there are active ECVs over CA;however,the response of extreme precipitation in CA to ECVs is the lowest.The possible reason is that the 10 mm d-1precipitation threshold is slightly higher for CA,a typical Asian dryland region.Our statistical results show that the 10 mm d-1precipitation threshold is close to the 90th percentile of precipitation.However,for the 6 mm d-1precipitation threshold,being close to the 80th percentile of precipitation,the extreme precipitation shows a good correlation with ECVs in CA.

3.3. Increasing trend of ECVs and response of extreme precipitation

To further explore the response of extreme precipitation to ECVs and their variations,Fig.3 shows the temporal distribution of standardized ECV frequency,total precipitation at the 10 mm d-1threshold,and the occurence numbers of QSVs and STEs over NA,EU-Med,and EA.It shows an increasing trend of ECVs over the three regions,with statistical significance at the 95% confidence level.

An 11°E ×7°N range around the ECV center is suggested as an effective region to explore extreme precipitation,as it covers more than 80%of extreme events over Asia accompanied by an ECV (Hu et al.,2010).Corresponding to an ECV,the total precipitation at the 10 mm d-1threshold (hereafter called total P10) also shows an increasing trend over the three regions (Fig.3 (a2–c2)).In particular,it shows a decadal increase since the 2000s.There is a statistically significant correlation between the ECVs and total P10 at the 95% confidence level over the three regions,which reveals that the more ECVs there are,the more extreme precipitation and other weather extremes occur.In addition,we found that the decreases in ECVs over northern Europe and total P10 are consistent with decreasing baroclinicity,corresponding to increasing anticyclonic vortex and heatwaves over northern Europe (Zhang et al.,2020).There are low correlations between the ECVs and total P10 over CA before 2000;however,the correlation is significant after 2000 along with increasing meridional circulation and meridional water vapor flux,revealing that water vapor flux is another dominant factor besides ECVs for extreme precipitation in Asian dryland regions (Zhang et al.,2021).Fig.3 (a3–c3) shows the ECV event numbers,STE decreases,however,increasing QSVs are the dominant contributor to the increase in ECVs and total P10.

Fig.2.(a) Latitude–time cross-section of total ECV frequency in 1979–2018 over land,expressed by the ECV center position (units: 40 yr-1).(b) Time-mean of total ECVs with latitude and ± standard deviation,as well as the contrast of ECV with latitude between 1979–1998 and 1999–2018 (units: 40 yr-1).(c) Spatial distribution of 40-yr vortex frequency (units: 40 yr-1) and (d) the related total P10 (units: × 100 mm 40 yr-1) in JA over land.The yellow boxes are northern America (NA),southern Europe to the Mediterranean (EU-Med) and East Asia (EA).

3.4. Projection of ECVs and extremes

How will ECVs vary in the future? And is their variation related to global warming? To address these questions,Fig.4 shows the projection of ECV variation and related precipitation extremes from seven CMIP6 coupled models for the period 2015–2065 aimed at the goals of 1.5°C and 2°C of the Paris Agreement.According to the ensemble simulations of the seven models,the 1.5°C goal under SSP2-4.5 and SSP5-8.5 emission scenarios occurs in 2046 and 2037,respectively,and the 2°C goal in 2061 and 2047.Before the goal of 1.5°C,the ECV frequency in 1°×1°grid cells is higher than the averaged ECV frequency in 2015–2065;however,it is lower than average beyond the goal of 1.5°C under SSP2-4.5 in NA and EA.The variation in ECV frequency within the goal of 1.5°C under SSP5-8.5 is similar to that under SSP2-4.5,but the higher value lasts until the goal of 2°C in NA.The decreased frequency is significant beyond the goal of 2°C.Such results indicate that the higher ECV frequently lasts for more than 25 years under SSP2-4.5 and for more than 19–26 years under SSP5-8.5,closely related to the relatively high local baroclinicity,if CMIP6 models are credible.These results also reveal the nonlinear variation of ECVs with global warming,especially within the goal of 1.5°C,which is possibly related to the dynamic contribution,besides baroclinicity variation.

Fig.3.Temporal variation of (a1-c1) standardized ECV frequency,(a2-c2) total P10,(c1-c3) STE and QSV process numbers over NA,EU-Med,and EA.(a1-a3,b1-B3) Thick lines and dashed lines are 5-yr filtered and linear trends,respectively.The r_trd is the correlation of linear trend,r_eddy_P is correlation between vortex frequency and total precipitation larger than 10mm,rSTE,rQSV are the correlation of linear trend of STE and QSV,respectively.

4.Conclusions and discussion

An ECV is a prominent factor for extratropical weather extremes,and it is the interaction of a planetary-scale transient vortex and synoptic-scale transient eddy with quasi-stationary waves (Sheng and Derome,1991).The energy transferred from the ECV is three times higher than the seasonal mean flow.Since the 2000s,ECVs have increased in frequency over some land areas,with a significant increasing trend in NA,Eu-Med,and EA.Future projections indicate ECVs will remain high in frequency until the goal of 1.5°C in 2046 under SSP2-4.5,and in 2037 under SSP5-8.5.

An increase in ECV frequency corresponds to enhanced inhomogeneous baroclinicity and instability over land;however,this cannot support a high ECV frequency after 2000.With global warming,high baroclinicity on the transient timescale directly excites STEs and QSVs;however,on average,the weakening surface temperature gradient decreases the baroclinicity in the midlatitude zone,but widens the meridional activity range of the westerly jet,which is favorable for meridional circulation development.This strengthens the resonance of the quasistationary wave (Mann et al.,2018),which is favorable to QSVs and cutofflows,such as the Central Asian Vortex,Northeast Cold Vortex,and North American Vortex.

Precipitation extremes occur along with QSV and STE disturbances.Corresponding to an ECV,the total P10 increases over NA,Eu-Med and EA.As one type of synoptic circulation,the ECV frequency increases due to inhomogeneous warming.The land–atmosphere interaction with inhomogeneous warming is partially responsible for the transient and subseasonal baroclinicity and impacts on the ECV,which explains increasing extremes.Furthermore,the ECV disturbance responses to land process and land cover change are worthy of explanation;In addition,related simulations should be further explored to reveal the variations and physical processes in future work.

Funding

This research was supported by the National Natural Science Foundation of China (Grant No.41975083).

Fig.4.Temporal variation of ECV numbers (green line) and its 75/25 percentile range (shaded) in JA from 2015 to 2065 under SSP2-4.5 and SSP5-8.5 in (a1,a2)NA and (b1,b2) EA and (a3,b3) scatter relationship between standardized ECV and total P10 in NA and EA within the goal of 1.5°.