Characteristics and mechanisms of the severe compound cold-wet event in southern China during February 2022

A severe compound cold-wet event occurred in southern China (hereafter referred to as CWESC ) in February 2022, leading to enormous socioeconomic losses. In this study, we proposed a new index to denote the severity of the compound cold-wet event. Based on the multivariate survival method, the CWESC in February 2022 is identified as the severest event during the past six decades. Our results indicate that the CWESC in 2022 is jointly regulated by the La Niña-like SST condition in the tropical Pacific and the warm SST anomalies in the North Atlantic, and a teleconnection in the Northern Hemisphere during winter (hereafter referred to as TNHW) plays the key role. The TNHW pattern originates from the tropical Pacific, and it splits into two routes over the North Atlantic. The northern branch of TNHW propagates via the Arctic and Siberia, causing intensified near-surface northerly wind and partially inducing an anomalous anticyclone over the western North Pacific (WNP). The southern branch of TNHW propagates via the Mediterranean and western Asia, inducing a deepened India–Burma trough and partially inducing the anomalous anticyclone over WNP. The intensified near-surface northerly wind causes enhanced cold advection over southern China, while the deepened India–Burma trough and the anomalous anticyclone over WNP cause increased southerly warm and moist air flow towards southern China, resulting in the CWESC in 2022. Moreover, four groups of numerical experiments forced by tropical Pacific, North Pacific, and North Atlantic SST anomalies are conducted based on the Community Atmosphere Model version 5. The results confirm the important roles of the La Niña-like condition and the warm SST anomalies in the North Atlantic in causing the CWESC in 2022.


Introduction
Southern China is a climate-sensitive region of China, which is vulnerable to extreme climate events due to the rapid development of the economy and high population density.The precipitation-related extreme events over southern China in winter have caused great socioeconomic losses (Zhou et al 2011).In February 2022, a compound cold-wet event occurred in southern China (hereafter referred to as CWESC).
This persistent CWESC caused severe damages to agriculture, transportation, power supply, etc, with more than six million people affected and direct economic losses reaching up to CNK 7.89 billion, ranked one of the severest ten natural disasters in China in 2022 (https://politics.gmw.cn/2023-01/12/content_36297529.htm).Some recent studies have emphasized the influences of SSTs over the tropical Indian Ocean and extratropical Atlantic on the CWESC in February 2022 (Ma et al 2022, Zhang et al 2023), whereas the regulations of SSTs over the tropical Pacific and North Atlantic remain unknown.Given the severity of the CWESC in February 2022, it is of great concern to understand the physical mechanisms and provide early warnings of this kind of event.
Numerous studies have investigated the mechanisms of winter extreme events in southern China focusing on the influences of tropical climate factors (e.g.Cheung et al 2012, Sun et al 2019b, 2022).In particular, the variations of El Niño-Southern Oscillation (ENSO) can significantly influence the variations of precipitation and temperature in southern China via regulating the East Asian atmospheric circulations, where the anomalous anticyclone/cyclone over western North Pacific (WNP) caused by ENSO plays a key role (Zhou et al 2009, Cheung et al 2012, Li et al 2017, Sun et al 2019b, 2022, Liu et al 2020, Wang et al 2022).However, recent studies found that the relationship between ENSO and East Asian winter climate anomalies is unstable under climate change (Zhang et al 2011, Sun et al 2019a).Specifically, the ENSO may affect the East Asian winter climate via three different pathways.The El Nino (La Niña) condition may induce an anomalous anticyclone (cyclone) over WNP via the Pacific-East Asian teleconnection pathway (Wang et al 2000).On the other hand, the El Nino (La Niña) condition may induce an anomalous cyclone (anticyclone) over WNP via the North Pacific pathway and the circumglobal teleconnection pathway (Sun et al 2019b).The influences of the latter two pathways have been strengthened during recent decades (Sun et al 2019b).
In addition, mid-to-high latitude climate factors also have an important impact on the extreme climate events in southern China during winter, such as the Arctic Sea ice (Kim et al 2021, Smith et al 2022) and the land surface processes (Nakamura et al 2019, Zhang et al 2019).Reduced Arctic Sea ice may weaken the meridional temperature gradient between the Eurasian continent and the Arctic, and induce enhanced Siberian high, causing increased northerly cold advection towards southern China (Sun et al 2019a(Sun et al , 2022)).The SST anomalies in the North Atlantic may trigger zonal Rossby waves and induce more active synoptic-scale waves propagating towards East Asia, which exert an impact on the temperature and precipitation in southern China (Sun et al 2019a, Liu et al 2020, Zhang et al 2023).
However, most of the aforementioned studies concentrate on the extreme temperature or precipitation anomalies in southern China during winter.The mechanisms of CWESC are less studied and understood.Note that a compound extreme cold-wet event is not necessarily a combination of an extreme cold event and an extreme precipitation event, but is possibly a combination of cold and wet events that are not themselves extremes, where this combination can cause a severe disaster.Thus, the mechanisms of CWESC are not necessarily a simple combination of the mechanisms of extreme precipitation and temperature events in southern China.
Thus, this study proposes a newly defined probability-based index (PI) to denote the compound cold-wet events.Based on this index, the severity of the CWESC can be well illustrated by the recurrence interval, and the spatiotemporal characteristics and mechanisms of the CWESC are investigated and understood.
The structure of the paper is organized as follows.Section 2 introduces the data and the newly proposed index of cold-wet events based on the survival copula method.In section 3, the spatiotemporal characteristics and underlying mechanisms of the CWESC in February 2022 are explored using statistical analysis and numerical simulations.The conclusion and summary are given in section 4.

Data
The gridded monthly mean datasets of precipitation and surface air temperature (CN05.1)developed by (Wu and Gao 2013) are employed in this study, with a spatial resolution of 1

Methods
The univariate return period of extreme temperature and precipitation is calculated as R (x i ) = 1 1−F(x i ) .In our study, the compound cold-wet event is identified based on the PI using the joint survival cumulative distribution of precipitation (X 1 ) and temperature (X 2 ).The traditional cumulative distribution function (CDF) of precipitation (X 1 ) and temperature (X 2 ) is calculated as F 1 (x 1 ) = P (X 1 ⩽ x 1 ) and F 2 (x 2 ) = P (X 2 ⩽ x 2 ), respectively.Here, the survival CDFs of precipitation (F 1 ) and temperature (F 2 ) are further calculated as F i = 1 − F i , respectively.Based on the survival cumulative distribution of precipitation and temperature, the joint survival cumulative distribution is constructed using the t-copula method (Salvadori et al 2013).Therefore, the joint survival CDF is calculated by For a particular probability level of t, the survival layer . Here, µ is the recurrence interval.
The probability level p is expressed as p = 1 − µ/T, where T is the return period (see figure 1(d)).The survival Kendall's quantile of order (q) is obtained by q = q (p) = sup {t ∈ I : K (t) = p} = K−1 (p).

Model description and experiment design
The NCAR's general circulation model CAM5 is used to simulate the response of atmospheric circulation to SST anomalies in the tropical Pacific and North Atlantic, which may exert an impact on the CWESC in February 2022.Here, four sets of numerical experiments are performed (table S1), including a control run (hereafter referred to as CTRL) and three sensitivity experiments.More detailed information can be seen in the supplementary information.

Characteristics of the CWESC in February 2022
Figure 1 shows the spatial distributions of the anomalies of surface air temperature and precipitation in February 2022 with respect to the climatology of 1991-2020.Compound negative temperature anomalies (figure 1(a)) and positive precipitation anomalies (figure 1(b)) occurred in southern China, with significant anomalies of temperature and precipitation exceeding one standard deviation of corresponding variables in this region.Based on the multivariate survival method, the PI is calculated using the precipitation and temperature anomalies.Significant negative PI anomalies occurred in southern China (figure 1(c)), suggesting an occurrence of severe CWESC.
The temporal series of areal mean surface air temperature, precipitation, and PI over southern China in February are computed (figure S1).The negative temperature anomaly in 2022 is ranked as the coldest event during 1961-2022, with the recurrence interval being 38 years (figure S1(a)).Meanwhile, the extreme precipitation anomaly in 2022 is ranked as the third wettest event during 1961-2022, with a recurrence interval of 19 years (figure S1(b)).Since negative temperature anomalies and positive precipitation anomalies do not necessarily co-occur, the combination of them may exacerbate the severity of the event.Based on the multivariate survival method, the CWESC in 2022 is ranked as the severest event during 1961-2022 (figure S1(c)), with the recurrence interval being more than 40 years (figure 1(d)).

Mechanisms of the CWESC in February 2022
Figure 2 shows the anomalous atmospheric circulation patterns relevant to the CWESC in February 2022.There is a notable anomalous anticyclone over the WNP at 700 hPa (figure 2 Hence, the anomalous anticyclone over WNP, the deepened Indo-Burma trough, the anomalous deep cold low over eastern China, and the strengthened Siberian High provided a favorable condition of atmospheric circulation for the severest CWESC during the past 62 years. As shown in figure 3(a), during February 2022, negative 300 hPa geopotential height (Z300) anomalies occurred over tropical Pacific, Greenland and Barents Sea, and East Asia, while positive Z300 anomalies occurred over North Pacific, mid-latitude North Atlantic, West Asia, and Siberia in the troposphere.These alternating positive and negative Z300 anomalies manifest a teleconnection in the Northern Hemisphere during winter (hereafter referred to as TNHW).The TNHW originated from the tropical Pacific, and it split into a northern wave train via the Arctic, Siberia, and East Asia and a southern wave train via mid-latitude North Atlantic, West Asia, and East Asia (figure 3(a)).
Specifically, the northern wave train propagated along the polar front jet route, inducing large-scale teleconnection of atmospheric circulation over the northern hemisphere in February (figures 2(a)-(c)), which is characterized by a '+−+−' pattern of wave train via North Atlantic-Barents Sea-northern Siberia-East Asia (figure 2(c)).This wave train caused an anomalous deep cold low over the middle-andlow latitudes of the East Asian continent (figure 2(c)), which may induce increased cold air mass over the surface of middle-and-low latitudes of East Asian continent, leading to increased cold advection over southern China (figures 2(g) and (h)).
Previous studies suggested that the East Asian climate is affected by wave train propagating along the subtropical westerly jet, where the wave train originates from the west (e.g.Europe, North Atlantic) and the subtropical westerly jet in the upper troposphere is the waveguide (Li and Sun 2015, Ding and Li 2017).In this study, the southern wave train mainly propagates along the subtropical westerly jet, which causes a deepened India-Burma trough along the southern flank of the Tibetan Plateau (figure 2(c)), playing an important role in causing southwesterly wave vapor transport anomalies as well as anomalous water vapor convergence over southern China (figure 2(b)).
Along with the anomalous TNHW pattern, a La Niña-like pattern occurred in tropical Pacific SSTs, while warm SST anomalies occurred in the North Atlantic in February 2022 (figure 4(a)).It is suggested that the ENSO and the North Atlantic airsea interaction may modulate the atmospheric teleconnection as well as the zonally propagating synoptic waves, which have important influences on the East Asian winter climate (Sun et al 2019a, 2019b).To further understand the association between the aforementioned TNHW and ENSO, the dominant empirical orthogonal function modes of Z300 anomalies in the Northern Hemisphere and the associated SST anomalies are computed.The spatial pattern of the second mode of Z300 anomalies (figure 4(b)) well resembles the TNHW pattern in February 2022  (figure 3(a)).The regressed SST anomalies indicate that the interannual variability of this TNHW pattern is associated with ENSO, where a La Niña-like condition (figure 4(c)) corresponds to the positive phase of TNHW (figure 4(b)).In addition, the interannual variability of this TNHW pattern is also associated with warm SST anomalies in the mid-latitude North Atlantic as well as cold SST anomalies in tropical and high-latitude North Atlantic (figure 4(c)), which partially resemble the significant mid-latitude warm SST anomalies in mid-latitude North Atlantic during February 2022 (figure 4(a)).
Figure 3(c) shows the latitude-pressure level section of wave activity flux associated with the geopotential height anomalies over the eastern Pacific during February 2022.A northward propagation of wave activity flux is observed from the tropical eastern Pacific towards the North Pacific, indicating an important role of the La Niña-like condition in modulating the TNHW in the Northern Hemisphere (figure 3(a)).
Moreover, the upward wave activity fluxes occurred over mid-latitude North Atlantic and were split into two routes at 300 hPa pressure level during February 2022, which propagated northward and southward, respectively (figure 3(b)).These two routes correspond to the two wave trains that propagate via the polar front jet route and via the sub-tropical jet route as mentioned before, implying an important role of the warm SST anomalies in the North Atlantic in splitting TNHW into two wave trains.
The above results suggest that the extreme CWESC during February 2022 is closely related to the TNHW pattern modulated by the La Niñalike condition and the warm SST anomalies in the North Atlantic.Particularly, note that the anomalous anticyclone over WNP (figure 2(b)) during February 2022 may result from a joint effect of the northern and southern wave trains that meet over East Asia and the western Pacific (figure 3(a)).

Validation of CAM5
To examine the roles of La Niña-like conditions and warm SST anomalies in the North Atlantic in affecting the extreme CWESC during February 2022, five sets of numerical experiments are conducted (see table S1).The results of EXP_LA minus CTRL (figures S2 and S3) indicate that the La Niñalike condition may trigger an atmospheric teleconnection pattern characterized by anomalous high over North Pacific, mid-latitude North Atlantic, and Siberia, and anomalous low over North America and East Asia (figure S3), which resembles the northern wave train in the TNHW during February 2022 (see supplementary text).Similar results can be found in the results of EXP_NA minus CTRL (figures S4 and S5).Meanwhile, the warm SST anomalies in the North Atlantic may also modulate a zonal wave train along the sub-tropical jet route (figure S5), resembling the southern wave train in the TNHW, which may cause a deepened Indo-Burma trough and an anomalous anticyclone over WNP (figure S4, see supplementary text).
Figure 5 shows the composite differences in atmospheric circulation between EXP_LA_NA and CTRL, which represent the combined effect of La Niña-like conditions and warm SST anomalies in the North Atlantic.A TNHW pattern resembling that observed during February 2022 occurs in response to the SST forcings in EXP_LA_NA, which is split into a northern wave train via the Arctic and Siberia, and a southern wave train via West Asia and India (figures 5 and S7).These wave trains induce an anomalous deep cold low over the middle-and-low latitudes of the East Asian continent (figure 5

Conclusions and discussion
In this study, the severest CWESC that occurred in February 2022 during the past 62 years is identified by a newly proposed multivariate survival method.The severe CWESC in February 2022 is associated with a TNHW pattern in the troposphere that is split into two wave trains via a northern route and a southern route, which causes strengthened Siberian High, a deepened India-Burma trough, an anomalous anticyclone over WNP, and an anomalous deep cold low over the middle-and-low latitudes of East Asian continent.These atmospheric circulation anomalies are key to the anomalous low temperature and increased precipitation in southern China during February 2022.The La Niña-like condition and warm SST anomalies in the North Atlantic play important roles in modulating the above TNHW pattern, which is confirmed by numerical experiments.
The results of this study demonstrate the importance of the combined effect of SST anomalies in the tropical Pacific and North Atlantic to the extreme climate events in East Asia.However, it should be noted that the warm SST anomalies were evident in the central North Pacific during February 2022.The results of CAM5 experiment of prescribed North Pacific SSTs (table S1) indicate that the warm SST anomalies in North Pacific may modulate the northern branch of the wave train (figure S7), which causes an anomalous deep cold low over the mid-latitudes of East Asian continent (figure S6(d)), being a factor in causing the increased near-surface cold advection (figures S6(g) and (f)) and cooling (figure S6(a)) in southern China during February 2022.
During past decades, the amplified arctic warming has led to a weakened temperature gradient between the mid-latitude Eurasian continent and the Arctic region (Rantanen et al 2022).This weakened temperature gradient results in increased meanders in the westerlies over the Eurasian continent during winter, which provides favorable conditions for increased blockings, strengthened Siberian High, and severe cold events in the East Asian continent (Mori et al 2014, Matsumura and Kosaka 2019, Kim et al 2021).As the Arctic continues warming at a rate of nearly four times of the globe during recent decades (Rantanen et al 2022), the atmospheric circulation has a background favoring severe cold events in southern China.Hence, the extreme cold event in southern China during February 2022 may result from the combined effects of interannual variability driven by SST anomalies and atmospheric circulation background associated with amplified Arctic warming trends.
Meanwhile, under global warming, the global hydrological cycle has been notably intensified, which is mainly characterized by increased water vapor in the atmosphere as well as increased water vapor flux and convergence/divergence (Held and Soden 2006, Allan et al 2020, 2022).This intensified hydrological cycle provides a favorable condition for the occurrence of extreme precipitation events.In addition, it has been suggested that the vertical motions in the atmosphere associated with precipitation extremes are strengthened under global warming, which is also favorable for the occurrence of extreme precipitation events.Hence, the changes in thermodynamic and dynamic processes under global warming are conducive to more extreme precipitation events, being an important background for the large precipitation anomalies in southern China during February 2022.
Thus, we suggest that the severe cold-wet event in southern China during February 2022 results from the combined effects of interannual variability driven by SST anomalies and favorable climate background associated with amplified Arctic warming, intensified hydrological cycle, and changed dynamic processes under global warming.

Figure 1 .
Figure 1.Spatial distribution of anomalous (a) surface air temperature (unit: o C), (b) precipitation (unit: mm), and (c) probability-based index (PI; unit: 1) in February 2022 relative to the climatology of 1991-2020.Stippling denotes where the anomalies exceed one standard deviation.The black rectangle represents the region of southern China.(d) Scatter plot of areal mean surface temperature (Y-axis, unit: o C) and precipitation (X-axis, unit: mm) anomalies in southern China during February during 1961-2022.The red star denotes the surface temperature and precipitation anomalies in February 2022.The colored isolines denote return period levels from 5 years to 45 years at an interval of 5 years.The yellow isoline denotes the 5-year return period, and so on.Results are based on the detrended temporal series.
(b)), along with the anomalous southwesterly wind that is dominant over southern China from 200 hPa (figure 2(a)) to 700 hPa (figure 2(b)).The anomalous southwesterly wind along the western flank of the anticyclone over the WNP may bring sufficient moisture from the ocean to southern China.Meanwhile, the deepened India-Burma trough (figure 2(c)) may lead to more water vapor transportation from the Indian Ocean to southern China (figure 2(b)).Moreover, anomalous upward motion is evident over southern China at 500 hPa (figure 2(e)), inducing decreased outgoing longwave radiation (figure 2(f)), which suggests increased convection over southern China.The increased moisture transportation and enhanced convection provide a favorable condition for precipitation in southern China, being an important factor for the extreme precipitation events in southern China during February.During February 2022, a deep cold low occurred over the middle-and-low latitudes of the East Asian continent (figure 2(c)), which may cause increased cloudy weather that leads to less solar radiation reaching the ground, providing a favorable condition for cold events.Meanwhile, positive SLP anomalies extend from Siberia towards southern China (figure 2(e)), suggesting a strengthened Siberian High.The strengthened Siberian High may induce increased northerly near-surface cold advection over eastern China flowing towards southern China (figure 2(g)), resulting in significant cooling over southern China during February (figure 2(h)).