Anomalous Atmospheric Circulation Associated with the Extremely Persistent Dense Fog Events Over Eastern China in the Late Autumn of 2018

Home , Extreme weather, Temperature anomaly, Troposphere

Article Anomalous Atmospheric Circulation Associated with the Extremely Persistent Dense Fog Events over Eastern China in the Late Autumn of 2018

Shengjie Chen 1,2,3, Duanyang Liu 2,4,* , Zhiming Kang 1,2, Yan Shi 5 and Mei Liu 1,2

1 Jiangsu Meteorological Observatory, Nanjing 210008, China; [email protected] (S.C.); [email protected] (Z.K.); [email protected] (M.L.) 2 Key Laboratory of Transportation Meteorology, China Meteorological Administration, Nanjing 210008, China 3 School of Atmospheric Sciences, Nanjing University, Nanjing 210023, China 4 Jiangsu Institute of Meteorological Science, Nanjing 210008, China 5 National Meteorological Information Center, Beijing 100081, China; [email protected] * Correspondence: [email protected]; Tel.: +86-25-83287114

Abstract: Under a declining trend of fog days in China, the duration of fog events since the 1990s reached a significant peak in the late autumn of 2018 over Eastern China. The average anomalous fog days were 4.74 d in November 2018 over Jiangsu Province in Eastern China, with a 1.73 standard deviation departure from climatology. Those fogs can thus be identified as a significantly abnormal climatic event with long duration, strong intensity, and extensive coverage. Based on the daily evolutions and correlations of atmospheric parameters, the dense fogs are revealed to be well configured by favorable metrological conditions such as weak dynamic progress, strong inversion in the lower troposphere and saturated air near the surface. If not disturbed, the intensification or duration of these conditions will further promote and maintain the development of fogs. The anomalous atmospheric background associated with those favorable meteorological conditions is revealed by composing the standardized anomalies of circulation fields during the fog days. Over the Citation: Chen, S.; Liu, D.; Kang, Z.; fog areas, vortex activities or cold air invasion is effectively hampered and the atmosphere inclines to Shi, Y.; Liu, M. Anomalous be stable, due to the anomalous circulation pattern composed of the broadened jet stream, weakened Atmospheric Circulation Associated jet core over Eastern China, undermined East Asian trough, declined East Asian winter monsoon, with the Extremely Persistent Dense and enhanced anomalous southerly flows that transport abnormal warm and wet air to Eastern Fog Events over Eastern China in the China. The vapor supplement is intensified by both sustained anomalous northward wind at the Late Autumn of 2018. Atmosphere 2021, 12, 111. https://doi.org/ lower troposphere and anomalous westward wind in the near-surface. Overall, the numbers of 10.3390/atmos12010111 standardized anomalies of 1000–200-hPa height, temperature, wind, and moisture fields during these fog days all significantly depart from climatology for that locale and time of the season, further Received: 23 December 2020 demonstrating that the persistent dense fogs over Eastern China in the late autumn of 2018 is an Accepted: 11 January 2021 unusual weather event with extreme synoptic-scale departures from normal. Published: 14 January 2021 Keywords: persistent dense fog; extreme event; anomalous atmospheric circulation; Eastern China Publisher’s Note: MDPI stays neu- tral with regard to jurisdictional clai- ms in published maps and institutional affiliations. 1. Introduction As a common meteorological phenomenon, fog is defined by horizontal atmospheric visibility of less than 1 km. The low visibility is due to the existence of water or ice Copyright: © 2021 by the authors. Li- droplets with the size from 2 to 65 µm in the atmosphere near the ground [1]. In China’s censee MDPI, Basel, Switzerland. National Standards, fog is categorized into five grades on the basis of visibility (V): thin fog This article is an open access article (1000 m ≤ V < 10,000 m), heavy fog (500 m ≤ V < 1000 m), dense fog (200 m ≤ V < 500 m), distributed under the terms and con- strong dense fog (50 m ≤ V < 200 m) and extreme dense fog (V < 50 m). On fog days with ditions of the Creative Commons At- low visibility, transportation difficulties and even traffic accidents frequently occur. With tribution (CC BY) license (https:// stable meteorological conditions and nearly calm wind, fog can induce a congregation creativecommons.org/licenses/by/ of aerosols, restraining the diffusion of pollutants in the air [2–4]. It thus imposes an 4.0/).

Atmosphere 2021, 12, 111. https://doi.org/10.3390/atmos12010111 https://www.mdpi.com/journal/atmosphere Atmosphere 2021, 12, 111 2 of 17

adverse impact on human activities and health in the locality [5]. On the other hand, the presence of fogs could modify surface energy and moisture budget [6]. Therefore, great attention has been paid to the physical processes [7–9], characteristics [10–12], formation mechanism [3,13–16] and forecasting of fog events [17–22]. Water droplets in fog form as the water vapor condense due to a decrease in air temperature or an increase in absolute humidity. Willett [17] categorized fogs into four major types: radiation fogs, advection fogs, frontal fogs and maritime fogs depending on the formation mechanism. Radiation fogs and advection fogs are most widespread. The radiation fogs are characterized by a marked temperature inversion due to the cooling of the boundary layer during cloudless nights. This type of fog phenomenon is common under the background of anticyclone activities, which favor clear skies. In the clear night, the strong surface long-wave radiation cools the lowest atmospheric layers, and then the saturation is enhanced. Meanwhile, advection fogs are associated with warm and humid air advection over a cold surface or cold air mass moving over a warm water surface. This type has two kinds of scenario [14,17]. In the first scenario, a cold surface decreases the warm air temperature and then increases the relative humidity with the same mechanism by which radiation fogs form. Moreover, the warm and humid air advection can transport water vapor to enhance the absolute humidity. Both forms can enhance saturation near the surface. In the second scenario, when a cold air mass moves over the warm water and the air temperature remains almost constant, the specific humidity increases rapidly due to the rapid evaporation of air over water. It induces a rise of relative humidity. This case is not common in inland areas. Actually, in many cases in China, radiation fogs and advection fogs could co-exist in the formation or enhancement progress of the fog events. To find more signals of the formation and evolution of fogs, the meteorological conditions and atmospheric circulation characteristics associated with fogs have been deeply investigated [3,15,17,23]. Houssos et al. [3] suggested the common favorable meteorological conditions for 1055 fog events over Europe are the surface night cooling under anticyclonic circulation and the advection of warm and humid air masses via a weak southerly flow. In some studies, to find the influential meteorological factors, the relationships have been examined between the long-term trends of fogs [24] or the interannual variation of the number of fog days [6] and atmospheric circulations. Fog formation might have some common metrological prerequisites, including cool air near the surface, warm advection in the lower troposphere, stable atmospheric stratification and affluent moisture. However, fog events could occur under different synoptic circulation patterns. The large-scale fog weather in China can be categorized into the fog under uniform air pressure and that before a front [25]. Ten types of specific atmospheric circulation structure have been found by classifying the 1055 European fog events based on the exact position, the intensity of the synoptic systems and their seasonality. The unique dynamical processes were revealed in detail by examining some cases of the 10 clusters [3]. The revelation of meteorological conditions and circulation patterns favorable to fog formation not only provides crucial signals for fog forecasting, but also promotes the utilization of mathematical models or approaches that incorporate the related physical processes to improve the prediction of fogs. Based on the meteorological factors associated with the dynamic and thermodynamic effects, a multivariate linear regression model was set up by Zhang et al. [15]. The model results show that the meteorological factors well explain more than 2/3 of the variance of the evolution of the fog and haze event in January 2013. Haze, another weather phenomenon with respect to visibility, is defined as dust, smog and water vapor suspended in the atmosphere with horizontal visibility less than 10 km, and relative humidity less than 90% [26,27]. Fog and haze are both low-visibility events, but they have different physical properties [28]. With the advancement of industrialization and intensification of human activities, haze pollution has increased steeply in China since the beginning of the 21st century, and has attracted growing attention [29]. Fog days in China showed noticeable decadal growth in the period of the 1970s to the 1990s, but a declining trend after the 1990s [30]. Liu et al. [24] attributed the reduction of fog days to the ascending Atmosphere 2021, 12, 111 3 of 17

trend of daily minimum temperature and the descending trend of relative humidity in winter. Considering that fog and haze can be interconvertible under certain conditions and that it is difficult to mark a clear and strict boundary between them, when the two appear at the same time they are assessed without any distinction in some studies [15,31]. The most simple and effective way to identify the fog days is just using the criterion of visibility. On this basis, all phenomena with daily visibility less than 1 km are classified as fog. In the late autumn of 2018, i.e., from mid-late November to early December, persistent heavy fog events occurred in the whole of Eastern China, especially in Jiangsu Province, which is a coastal province on the shore of the Yellow Sea. Those heavy fogs, characterized by long duration and extremely low visibilities less than 50 m, exerted unprecedentedly severe effects on road and air transport. Six early warnings of the highest grade were issued by Jiangsu Meteorological Observation. The extreme fog events drew considerable concerns and extensive reports at home and abroad. Contrary to the background of a declining trend in China, fogs days in late autumn of 2018 reached a significant peak since the 1990s. It is widely classified as a rare weather event under the ground of global warming. What abnormal characteristics did the fog events have? How did the meteorological conditions act on the evolution of the fogs? What was the enhancement mechanism of the fog events? When answering the third question, conventional circulation fields are often insufficient to determine whether an event represents a large departure from normal [32]. Here standardized anomaly analysis was applied to dig out some anomalous circulation signals for the formation and duration of fogs, thus providing a new and objective approach for forecasting the intensity of fogs. In this paper, we examined the atmospheric circulation anomalies related to the extremely persistent dense fog events in the late autumn of 2018 over Eastern China. The data and analysis methods used are described in Section2. Section3 presents an analysis of the meteorological conditions, synoptic patterns, and anomalous atmospheric circulations associated with the extremely persistent dense fogs. The conclusions are given in Section4.

2. Data and Methods 2.1. Data (1) The daily data of fog phenomena (visibilities are less than 1000 m) at 2474 national stations are provided by the observation statistical results from the China Meteorological Administration. The weather phenomena of fog are observed four times a day at a special time at a national ground weather station. If fog is observed at any time in a day, the day is deﬁned as a corresponding fog day. The number 1 in the datasets is qualiﬁed for a fog day, while the number 0 indicates that there is no fog phenomenon, and the number 8 refers to a lack of observation data. After removing the missing data in November and December from 1961 to 2017, complete data from 1812 stations in China were left. (2) The visibility data of Jiangsu Province, China is derived from the visual measurements at 70 Jiangsu observation stations eight times per day every 3 h from 1 November to 31 December 2018. If a one-time visual measurement reaches one of the grades of fogs, the occurrence of fog with that grade is presented. The daily mean visibility is calculated by the average of eight observations. (3) The circulation data at four times (00:00, 06:00,12:00 and 18:00 universal time) daily with a horizontal resolution of 1 × 1 is from the European Centre for Medium-Range Weather Forecasts Reanalysis Interim dataset [33]. The daily mean circulation was calculated by averaging the values four times in a day. The time span of data for daily circulation data is from 1980 to 2018. The reanalyzed variables include winds, geopotential height and air temperature at 12 standard pressure levels from 1000 to 100 hPa.

2.2. Methods The correlation coefﬁcients between the daily visibility and different kinds of meteorological conditions were calculated to investigate the impacts of meteorological factors on the daily evolution of visibility. The length of the correlation analysis series is 61 days. Atmosphere 2021, 12, 111 4 of 17

Due to variabilities that exist throughout the year and regionally throughout the world, conventional pressure level circulation fields often are insufficient to determine whether a weather event represents a large departure from normal. In view of that, normalized field departures from local climatology are proposed to provide guidance on the relative rarity of events objectively [32]. The composite analysis of the original values and standard anomalies of all the daily mean circulation fields were conducted to identify the reasons for the formation of meteorological conditions advantageous to the rare persistent dense fog events over Eastern China in the late autumn of 2018. The standard deviations were subtracted from the 21-day running mean values and then divided by their respective 21-day running standard deviations. The calculation formula is as follows:

S = (x − µ)/σ (1)

where x is the original gridpoint value, µ is the 21-day running mean for that grid point, σ is the 21-day running standard deviation, and S is the standardized anomalies for the daily mean circulation ﬁeld at each grid point. The climatological mean is calculated for the period from 1980 to 2017. For example, µ on 1 November is the climatology of the averaged value of x from 22 October to 11 November in each year from 1980 to 2017. This process converts a pseudo-normal into a standard normal distribution [32]. The composite analysis is done from 24 November to 3 December in 2018 with a period of 10 days. Student’s t-test was employed to test the signiﬁcance of the correlation and composite analysis. Although composite analysis only deals with a case with a duration of 10 days, the fog events were featured with rare persistence and intensity so that the robust results of composite circulation should be available for other strong cases to some extent.

3. Results This section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn.

3.1. The Extremely Persistent Dense Fog Events Over Eastern China in the Late Autumn of 2018 In the late autumn of 2018, heavy fog events occurred persistently and frequently over Eastern China. Figure1 shows the daily numbers of stations with the occurrence of fog (visibility less than 1000 m, bars), heavy fog (visibility between 1000 m and 500 m, blue line), dense fog (visibility less than and equal to 500 m and more than 200 m, orange line) and strong dense fog (SDF) (visibility less than 200 m, red line) events over Jiangsu Province in China. From 1 November to 30 December 2018, most of the days witnessed fog events, with at least one station seeing visibility of less than 1000 m for each day. The fog events from 24 November to 3 December are characterized by the strongest intensity and the longest duration in that year, with the daily visibility of at least 24 stations being less than 1000 m. This accounts for about one-third of all the 70 stations over Jiangsu Province. The number of stations with dense fog and SDF both reached the maximum in the late autumn of 2018. In particular, the SDF that covered at least 24 stations and persisted for ﬁve days from 25 to 29 November is rare and impressive. In the middle of November and the end of December, another two peaks of fog coverage appeared, but the durations for the respective fogs were short. Atmosphere 2021, 12, x FOR PEER REVIEW 5 of 17

Atmosphere 2021, 12, 111 5 of 17 November and the end of December, another two peaks of fog coverage appeared, but the durations for the respective fogs were short.

Figure 1. The daily numbers of stations with the occurrence of fog (mean daily visibility less than Figure 1. The daily numbers of stations with the occurrence of fog (mean daily visibility less than 1000 m, bars), heavy fog (visibility less than 1000 m and more than 500 m, blue line), dense fog 1000 m, bars), heavy fog (visibility less than 1000 m and more than 500 m, blue line), dense fog (visibility less than and equal to 500 m and more than 200 m, orange line) and strong dense fog (visibility(visibility less less than than and200 m, equal red to line) 500 events m and over more Jiangsu than 200Province m, orange in China. line) The and dotted strong line dense denotes fog (visibility24 stations, less which than 200are m,about red one-third line) events of al overl the Jiangsu 70 stations Province over inJiangsu China. Province. The dotted line denotes 24 stations, which are about one-third of all the 70 stations over Jiangsu Province. 3.2. The Climate Anomalies of Fog Days in Late Autumn of 2018 3.2. The Climate Anomalies of Fog Days in Late Autumn of 2018 Late autumn is a foggy season. The stations with high numbers of climatological fog Late autumn is a foggy season. The stations with high numbers of climatological fog days during November and December are located in most parts of Eastern China south of days during November and December are located in most parts of Eastern China south of 4040°◦ N, N, and and the the southwestern southwestern and and northwestern northwestern corners corners of of China. China. Standard Standard deviations deviations of of fogfog days days exceeding exceeding 3.0 3.0 days days are are also also located located over over the regions the regions mentioned mentioned above above (Figure (Figure2a,b). It2a,b). suggests It suggests that fog that has fog a high has possibilitya high possibi of occurrencelity of occurrence and large and interannual large interannual variability var- overiability those over areas those during areas late during autumn. late autumn. InIn the the late late autumn autumn of of 2018, 2018, the the fog fog days days were were above above normal normal in in many many areas areas in in Eastern Eastern ChinaChina to to the the east east of of 110 110°◦ E. E. In In particular, particular, the the number number of of anomalous anomalous fog fog days days was was at at least least fourfour days days over over the the downstream downstream area area of the of Yangtzethe Yangtze River River and the and Yellow the Yellow River in River November in No- 2018vember (Figure 20182c,d). (Figure The anomalous 2c,d). The center, anomalous with values center, exceeding with values 8 days, exceeding appeared 8 in days, central ap- northernpeared in Jiangsu central province northern over Jiangsu east province China adjacent over east to theChina Yellow adjacent Sea whereto the anomaliesYellow Sea werewhere above anomalies 3.0 standard were above deviations 3.0 standard from the deviations climatology from (Figure the climatology2c,d). (Figure 2c,d). FigureFigure3a 3a shows shows the the annual annual cycles cycles of fog of occurrence fog occurrence from 1961 from to 1961 2017 averagedto 2017 averaged among 70among stations 70 stations over the over whole the Jiangsu whole Jiangsu Province Province in China. in China. The average The average annual annual value value of fog of daysfog days is 33.9 is d 33.9 at each d at station. each station. The number The number of fog daysof fog varies days with varies month; with itmonth; has a double- it has a peakdouble-peak structure, structure, with the largestwith the value largest of 4.0 value d·year of–1 ·4.0station d·year–1 occurring–1·station–1 in occurring November in and No- anothervember peak and phaseanother of peak 3.0 d ·phaseyear–1 of·station 3.0 d·year–1 appearing–1·station in–1 appearing April. It is in indicated April. It that is indicated the fog eventsthat the in Jiangsufog events occur in mainlyJiangsu in occur boreal mainly autumn, in asboreal the occurrenceautumn, as in the this occurrence season accounts in this forseason approximately accounts for 30.03% approximately of the annual 30.03% fog of occurrence, the annual whilefog occurrence, the number while in summerthe num- isber small, in summer only being is small, 18.20% only of being the total. 18.20% In Novemberof the total. andIn November December and 2018, December the average 2018, numberthe average of fog number days over of fog Jiangsu days over was 8.69Jiangsu d and was 6.47 8.69 d, d respectively, and 6.47 d, respectively, and the anomalies and the wereanomalies 1.0~2.0 were standard 1.0~2.0 deviations standard fromdeviations the grand from mean. the grand The fogmean. day The in November fog day in also No- featuredvember aalso remarkable featured interannuala remarkable variability. interannual In Novembervariability. 2018,In November the average 2018, number the aver- of anomalousage number fog of days anomalous was 4.74 fog d, withdays an was anomaly 4.74 d, of with 1.73 an standard anomaly deviation of 1.73 departurestandard devia- from climatology.tion departure This from was theclimatology. third most This fog days was inthe November third most averaged fog days over in JiangsuNovember Province aver- inaged China over since Jiangsu 1961 Province and the mostin China since since the beginning1961 and the of themost 21st since century the beginning (Figure3b of). the It conﬂicted, to some extent, with the declining trend of fog days after the 1990s [30].

Atmosphere 2021, 12, x FOR PEER REVIEW 6 of 17

21st century (Figure 3b). It conflicted, to some extent, with the declining trend of fog days after the 1990s [30]. Atmosphere 2021, 12, x FOR PEER REVIEW 6 of 17

Atmosphere 2021, 12, 111 6 of 17 21st century (Figure 3b). It conflicted, to some extent, with the declining trend of fog days after the 1990s [30].

Figure 2. (a) The climatological fog days in November averaged from 1961 to 2017 (shaded) in China. Standard deviations exceeding 3.0 days are marked by contours; (b) as in (a), but for De- cember; (c) the anomalous fog days (shaded) and the stations that anomalies exceeding 2.0 (black dots)/3.0 (black dots) standard deviations in November 2018 in China; (d) as in (c), but for De- cember. The units of all shaded and contours are days. The two lines on the China map are the Yellow River (line in the northern) and Yangtze River (line in the southern). The boundary of FigureJiangsuFigure2. Province, 2.( a()a) The The climatologicalChina climatological is marked fog fog in days days(c,d in). in November November averaged averaged from from 1961 1961 to 2017to 2017 (shaded) (shaded) in China.in China. Standard deviations exceeding 3.0 days are marked by contours; (b) as in (a), but for De- Standard deviations exceeding 3.0 days are marked by contours; (b) as in (a), but for December; cember;The (heavyc) the anomalous fog in November fog days 2018(shaded) could and be the identified stations that as an anomalies abnormal exceeding climatic 2.0 event. (black (c) the anomalous fog days (shaded) and the stations that anomalies exceeding 2.0 (black dots)/3.0 Whatdots)/3.0 meteorological (black dots) standard conditions deviat areions beneficial in November to the 2018 formation, in China; development, (d) as in (c), but and for dura- De- (black dots) standard deviations in November 2018 in China; (d) as in (c), but for December. The tioncember. of fogs? The unitsWhat of abnormal all shaded circulationand contours anomalies are days. Theand two meteorological lines on the China factors map have are thesig- units of all shaded and contours are days. The two lines on the China map are the Yellow River (line nificantlyYellow River impacted (line in the the occurrencenorthern) and of Yangtzethis extremely River (line persistent, in the southern). dense fog? The These boundary two ofcore inJiangsu the northern) Province, and China Yangtze is marked River (line in ( inc,d the). southern). The boundary of Jiangsu Province, China is questions motivate our research on the meteorological factors and abnormal circulations marked in (c,d). associated with this significant weather event. The heavy fog in November 2018 could be identified as an abnormal climatic event. What meteorological conditions are beneficial to the formation, development, and duration of fogs? What abnormal circulation anomalies and meteorological factors have significantly impacted the occurrence of this extremely persistent, dense fog? These two core questions motivate our research on the meteorological factors and abnormal circulations associated with this significant weather event.

Figure 3. ((aa)) The The climatological climatological annual annual cycles cycles of offog fog days days (dark (dark blue blue line) line) averaged averaged from from 1961 1961to to 2017 over Jiangsu Province Province in in China China (the (the boundary boundary of of is is Jiangsu Jiangsu Province Province marked marked in in Figure Figure 2c).2c). The The ﬁlledfilled lines from from light light orange orange to to dark dark orange orange repr representesent ±1,± ±21, and±2 3 and standard 3 standard deviations deviations among among the the measured sites. The red solid circles mark the fog days of November and December in 2018. (b) The interannual variations of fog days in November averaged over Jiangsu Province in China

from 1961 to 2018. The four dotted lines denote the average fog days and the fog days of one to three standardFigure 3. deviations. (a) The climatological annual cycles of fog days (dark blue line) averaged from 1961 to 2017 over Jiangsu Province in China (the boundary of is Jiangsu Province marked in Figure 2c). The filledThe lines heavy from foglight in orange November to dark 2018orange could represent be identified ±1, ±2 and as 3 anstandard abnormal deviations climatic among event. the What meteorological conditions are beneficial to the formation, development, and duration of fogs? What abnormal circulation anomalies and meteorological factors have significantly impacted the occurrence of this extremely persistent, dense fog? These two core questions

Atmosphere 2021, 12, x FOR PEER REVIEW 7 of 17

Atmosphere 2021, 12, 111 7 of 17

measured sites. The red solid circles mark the fog days of November and December in 2018. (b) The interannual variations of fog days in November averaged over Jiangsu Province in China from 1961 to 2018. The fourmotivate dotted lines our denote research the average on the meteorological fog days and the factors fog days and of one abnormal to three circulationsstandard associated deviations. with this signiﬁcant weather event.

3.3. Meteorological3.3. Conditions Meteorological Conducive Conditions to the Conducive Persistent to Dense the Persistent Fog Events Dense Fog Events The fog in the Thelate autumn fog in the of late 2018, autumn with th ofe 2018,strongest with intensity, the strongest longest intensity, duration longest and duration and most extensivemost coverage extensive that coverageyear, persisted that year, for persisted10 days from for 10 24 days November from 24 to November 3 Decem- to 3 December, ber, which is shownwhich isas shownthe space as the between space betweenthe two vertical the two verticallines in linesFigure in 4a. Figure During4a. During this this period, period, the averagedthe averaged daily dailyvisibility visibility over overJiangsu Jiangsu Province Province in Eastern in Eastern China China shows shows a a persistent persistent lowestlowest value value of about of about 2 km. 2 km. Numerous Numerous research research has has revealed revealed that that the the occur- occurrence of fogs is rence of fogs closelyis closely associated associated with with certain certain meteorological meteorological factors fa [ctors6,15, 27[6,15,27,34].,34]. The meteorological The back- meteorologicalgrounds backgrounds can be can classiﬁed be classified into three into categories: three categories: the dynamic, the dynamic, thermal thermal and vapor conditions. and vapor conditions.The evolution The evolution of each meteorological of each meteorological factor is highly factor consistent is highly withconsistent the development of with the developmentdaily visibility of daily over visibility Jiangsu over Province Jiangsu in Province Figure4b–f in Figure and Table 4b–f1. and It signiﬁcantly Table 1. highlights It significantlythe highlights roles played the roles by some played kinds by some of meteorological kinds of meteorological condition. condition.

Figure 4. (aFigure)The daily 4. (a evolution)The daily ofevolution the number of the of number stations of with statio thens occurrencewith the occurrence of fog and of thefog dailyand the averaged daily visibility over Jiangsuaveraged Province visibility in China over (The Jiangsu boundary Province of is Jiangsu in China Province (The boundary marked in of Figure is Jiangsu2c) and Province ( b–f) the marked daily in evolution of Figure 2c) and (b–f) the daily evolution of different kinds of related meteorological factors. All different kinds of related meteorological factors. All variabilities are averaged over the area of 115◦ E~122◦ E, 30◦ N~35◦ N. variabilities are averaged over the area of 115° E~122° E, 30° N~35° N. The parameters include (1) The parameters include (1) dynamical conditions (b,c): 10 m and 1000 hPa wind speed (units: m·s−1), absolute gradient dynamical conditions (b,c): 10 m and 1000 hPa wind speed (units: m·s−1), absolute gradient of SLP −3 · −1 −2 · −1 −1 of SLP (units:(units: 10 10Pa−3 Pa·mm −)(1) b(b),), 1000 1000 hPa hPa verticalvertical velocityvelocity (units: 10 10−2 Pa·sPa−1s), 700), 700and and 1000 1000 hPa hPaEGR EGR (units: (units: day ), −1 and verticalday shear−1), and of winds vertical between shear of 500 winds and between 850 hPa 500 (units: and m850·s hPa)( c(units:); (2) thermal m·s−1) (c conditions); (2) thermal (d ):conditions vertical difference of temperature(d): vertical (VDT)between difference 1000 of temperature hPa and 2 m, (VDT) between between 925 and1000 1000hPa and hPa, 2 betweenm, between 1000 925 hPa and and 1000 the hPa, skin surface, and betweenbetween 2 m and 1000 the hPa skin and surface the skin (units: surfac◦C),e, and and the between 925 hPa 2 m temperature and the skin advection surface (units: (units: °C), 10− and5 ◦C the·s− 1); (3) vapor conditions (925e,f): hPa 1000 temperature hPa speciﬁc advection humidity (units: (units: 10 g·−kg5 °C·s−1),−1); dew (3) vapor point ofconditions 2 m (units: (e,f):◦C) 1000 (e), hPa the dewspecific point hu- depression of 2 m (units: ◦midityC) and (units: 1000 hPa g·kg relative−1), dew humidity point of 2 (units: m (units: %) ( f°C)). All (e), numbers the dew inpoint the depression brackets in theof 2 legendm (units: of each°C) line-layers are the correlationand 1000 coefﬁcients hPa relative of eachhumidity meteorological (units: %) (f factor). All numbers with the averagedin the bracke dailyts in visibility the legend over ofJiangsu each Province in line-layers are the correlation coefficients of each meteorological factor with the averaged daily China. The period between the two vertical lines is from 24 November to 3 December. visibility over Jiangsu Province in China. The period between the two vertical lines is from 24 No- vember to 3 December.

Atmosphere 2021, 12, 111 8 of 17

Table 1. The summarization of the correlation coefﬁcients of kinds of meteorological dynamical, thermal and vapor factors with the daily averaged visibility over Jiangsu Province in China. All the correlations exceed the 0.01 signiﬁcance level.

Conditions Parameters The Correlation Coefﬁcient 10 m wind speed 0.51 1000 hPa wind speed 0.51 the absolute gradient of sea level pressure 0.47 dynamical 1000 hPa vertical velocity 0.57 700 hPa Eady growth rate (EGR) 0.49 1000 hPa EGR −0.51 vertical shear of winds between 500 and 850 hPa 0.55 vertical difference of temperature (VDT) −0.68 between 1000 hPa and 2 m VDT between 925 and 1000 hPa −0.64 thermal VDT between 1000 hPa and the skin surface −0.75 VDT between 2 m and the skin surface −0.74 925 hPa temperature advection −0.45 1000 hPa speciﬁc humidity −0.48 − vapor The 2 m dew point 0.54 the 2 m dew point depression 0.56 1000 hPa relative humidity −0.37

3.3.1. Dynamical Conditions Dynamically, fogs hardly occur under the background of strong surface winds or strong motions. Investigating the evolution of the absolute gradient of sea level pressure (SLP) as well as the 10 m and 1000 hPa winds in Figure4b, it is revealed that a stronger (weaker) absolute gradient of SLP and a larger (smaller) surface wind velocity correspond well to higher (lower) visibility, suggesting highly positive correlation coefficients between the three dynamical factors and daily visibility. The correlation coefficients are 0.47, 0.51 and 0.51, respectively, all exceeding the 0.01 significance level (Table1). During the most persistent dense fog events, the absolute gradient of SLP and the surface wind speed hit the minimum, hindering the outward transport of fog and favoring its development and persistence. On the other hand, increased visibility could be induced immediately due to strengthened surface winds. The maintenance and enhancement of lower visibility are also highly related to a weaker vertical motion and the mixing between the lower and middle troposphere (Figure4c). The significant correlation coefficients of 1000 hPa vertical velocity and the vertical shear of horizontal winds between 500 and 850 hPa with daily visibility are 0.57 and 0.55, respectively (Table1). It is indicated that the weak mixing or diffusion of air in the lower and middle troposphere is conducive to the congregation of fog droplets and the decline of visibility near the surface. It is apparent that a more stable and static atmosphere contributes to the formation and concentration of fog. Here, we calculated an index named the Eady growth rate (EGR), which measures the atmospheric baroclinicity [35–37]. The

= f du detailed equation of EGR is EGR 0.31 N dz , f is the Coriolis parameter, N is the Brunt– Väisälä frequency and u is the zonal winds. A higher EGR index means stronger activities of mid-latitude eddies and fluxes of transient eddy heat and momentum in the lower layer. Interestingly, the correlations of the daily 700 hPa and 1000 hPa EGR with visibility are quite opposite, which are 0.49 and −0.51, respectively, both exceeding the 0.01 significance level (Figure4c and Table1). The increased visibility associated with the rise of EGR at 700 hPa can be attributed to the enhancement of atmospheric baroclinicity in the lower troposphere, further intensifies the transport of transient eddy heat and momentum, and destroys the stability. Conversely, increased EGR at 1000 hPa will lead to lower visibility (Figure4c). The underlying physical process might be the intensification of the turbulence in the boundary layer due to the increase of the transient eddy activities near the surface. The turbulent exchange implies a mixing of eddies of different origins, which promotes Atmosphere 2021, 12, 111 9 of 17

air saturation and fog formation [38]. Turbulence in the fog boundary layer (FBL) beneﬁts the formation of fog [39]. Based on the view of Zhou and Ferrier [39] that the deeper a fog is, the stronger a turbulence intensity it can endure, persistent and appropriate turbulent activities are conducive to the long duration of fog in the fog events in the late autumn of 2018. To sum up, weak outward horizontal transport, passive vertical diffusion in the lower troposphere and appropriate turbulence in FBL impose a critical impact on the formation and persistence of low visibility.

3.3.2. Thermal Conditions The unique thermal conditions during the evolution of visibility are a major concern. Here, the daily evolutions of the vertical difference of temperature (VDT) between some layers are shown in Figure4d, such as VDT between 925 and 1000 hPa, between 1000 hPa and 2 m, between 1000 hPa and the skin surface, and between 2 m and the skin surface. The correlation coefficients of the three calculation ways of VDT and the daily visibility are so significantly high (the correlation coefficients are −0.64, −0.68, −0.75, and −0.74 respectively in Table1). This means that the consistent temperature inversion structure captures well the evolution of visibility. A stronger temperature inversion with a longer duration favors lower visibility, resulting in a persistent heavy fog. The formation of the inversion structure is connected with the relative cooling near the surface or the warm advection at the lower tropospheric layer. The two formation methods correspond to the two main types of fog, i.e., radiation fog, and advection fog. Both types were present in the persistent heavy fog events in 2018, as can be seen from the significant correlations of the intensity of inversion and 925-hPa temperature advection (the correlation coefficient is −0.45) with the evolution of visibility.

3.3.3. Vapor Conditions Fog is composed of numerous tiny water droplets. Sufficient and saturated vapor is, as revealed, a prerequisite for the formation and persistence of fog, which means that the absolute and relative humidity conditions are both indispensable to fog events. The significant correlations between the humidity factors (the 1000 hPa specific humidity, 2 m dew point, 1000 hPa relative humidity and the dewpoint deficit at 2 m) and visibility indicate that more vapor and a higher degree of saturation near the surface are favorable for water vapor condensation, thus helping form or enhance the fog phenomena (Figure4e–f and Table1). The above analyses confirm that the persistent dense fog in the late autumn of 2018 was well configured by certain meteorological conditions such as weak dynamic progress, strong inversion and saturated air near the surface. If not disturbed, the duration or intensification of these conditions will further promote the development of fogs.

3.4. The Associated Anomalous Atmospheric Circulations In this section, attention is paid to the anomalous atmospheric background that provides favorable metrological conditions for those extreme fog events with long duration and strong intensity. A deep understanding of the anomalous atmospheric circulations acts as the foundation for the short-term, medium-term and even sub-seasonal forecasts of fog, and is vital to unraveling the underlying physical processes behind the impact of extra-forcings (such as sea surface temperature) on fog at some timescales. The composite analysis and the Student’s t-test analysis of the standardized anomalies [40] were conducted on all the circulation ﬁelds for fog days from 24 November to 3 December in 2018.

3.4.1. Anomalous Dynamic Circulations Figure5a,b depict the distributions of the composite 500 and 925 hPa original and anomalous geopotential heights (GH). During the fog days, a trough with a smaller magni- tude lies over coastal East Asia, which is accompanied by universally positive GH anoma- Atmosphere 2021, 12, 111 10 of 17

lies south to 45◦ N, especially over Eastern China in the middle troposphere (Figure5a ). Meanwhile, in the lower layer, there exist abnormally weaker GH over land and stronger GH over the sea to the east of 105◦ E, resulting in anomalous northward flows along with Eastern China according to the theory of geostrophic wind equilibrium (Figure5b). The East Asian trough in the middle troposphere and the East Asian winter monsoon in the lower troposphere are obviously weakened so that the invasion of cold air to Eastern China is effectively curbed, and the anomalous southerly flows transport more heat and vapor to Eastern China in the lower troposphere. The corresponding anomalous southerly winds are also presented in the troposphere above 900 hPa (the figure is omitted). In the vertical section of composite GH standard deviations across 120◦ E (Figure5c), significant negative GH anomalies only appear at the shallow lower layers under 900 hPa, while significant positive GH anomalies are present almost in the entire troposphere above 900 hPa. It is suggested that the occurrence of the local meteorological conditions closely correlated with the persistent dense fog can be attributed to the certain persistent anomalous large-scale Atmosphere 2021, 12, x FOR PEER REVIEWcirculation pattern. Anomalous circulation signals characterized with the weaker11 of cold17 high

and the absolute gradient of SLP emerge at the boundary layer. This induces a decline in wind velocity near the surface over Eastern China (Figures5c and6a). Moreover, circulation anomalieshances the alsostability occur of inthe the near-surface higher troposphere. atmosphere,For it contributes example, itto isthe shown formation that and during the extremelymaintenance persistent of fogs. It dense is apparent fog events that the in structure 2018, the of East the inversion Asian jet is stream mainly (EAJS) attributed at 200 hPa wasto the markedly warm advection broadened at 925 and and weakened 850 hPa associated with the with occurrence the southerly of westerly flows. As anomalies seen in north tothe 40 composite◦ N and south925 hPa to winds 20◦ N and and air easterlytemperature anomalies averaged between from 24 November 20~40◦ N. to The 3 De- northern (southern)cember in flank2018 (the(Figure wind 7b), velocity in Eastern is 30 China, m·s− 1especially) shifted northwardover Jiangsu (southward) Province, the from the climatologicalnorthward wind location is nearly of perpendicular EAJS (Figure to6b). the The isotherm, northward-shifted inducing obvious EAJS warm became tem- so ap- parentperature that advection it could anomalies. hinder the The meridional anomaly invasioncenters of ofthe cold standard air from deviations high latitudes of tem- [41,42], perature and temperature advection are above 3.0, representing an unusual and even and the weakened westerly jet between 20~40◦ N might attenuate the baroclinicity belt rare departure from average conditions (Figure 7). Below 850 hPa, the distributions of in the troposphere via wave-zonal flow interactions [36]. Considering the northward and regional and time-averaged temperature advection and the standard deviations at dif- southwardferent heights shifts (Figure of the 8c,d) jet’s flanks,both show and thea distinct weakness inversion. of thejet’s The core, standard the East deviation Asian trough, andanomalies the East of temperature Asian winter advection monsoon, at 850 it hPa might are bemore speculated than 2.5, indicating that the uncommon entire circulation structurethermal circulation is loosened, in the favoring fog days a static compared and stablewith the weather climatology condition at theand lower hampering tropo- any vortexsphere. activities or cold air invasion.

Figure 5. (a) The composite of original 500 hPa geopotential height (GH) (contours, unit: gpm) and Figure 5. (a) The composite of original 500 hPa geopotential height (GH) (contours, unit: gpm) and standard anomalies of GH (shaded) from 24 November to 3 December in 2018. (b) is the same as in standard(a) but for anomalies 925 hPa GH. of ( GHc) The (shaded) vertical fromsection 24 of November composite standard to 3 December anomalies in 2018.of GH ( bacross) is the 120° same as in ◦ (aE.) butThe forcrossed 925 hPaarea GH.in each (c) panel The vertical denotes sectionanomalies of compositeexceeding the standard 0.01 significance anomalies level. of GH across 120 E.

The crossed area in each panel denotes anomalies exceeding the 0.01 signiﬁcance level.

Figure 6. (a) The composite of original sea level pressure (SLP) (contours, unit: hPa) and standard anomalies of the absolute gradient of SLP (shaded) from 24 November to 3 December in 2018. (b)

Atmosphere 2021, 12, x FOR PEER REVIEW 11 of 17

hances the stability of the near-surface atmosphere, it contributes to the formation and maintenance of fogs. It is apparent that the structure of the inversion is mainly attributed to the warm advection at 925 and 850 hPa associated with the southerly flows. As seen in the composite 925 hPa winds and air temperature averaged from 24 November to 3 De- cember in 2018 (Figure 7b), in Eastern China, especially over Jiangsu Province, the northward wind is nearly perpendicular to the isotherm, inducing obvious warm temperature advection anomalies. The anomaly centers of the standard deviations of temperature and temperature advection are above 3.0, representing an unusual and even rare departure from average conditions (Figure 7). Below 850 hPa, the distributions of regional and time-averaged temperature advection and the standard deviations at different heights (Figure 8c,d) both show a distinct inversion. The standard deviation anomalies of temperature advection at 850 hPa are more than 2.5, indicating uncommon thermal circulation in the fog days compared with the climatology at the lower troposphere.

Figure 5. (a) The composite of original 500 hPa geopotential height (GH) (contours, unit: gpm) and Atmosphere 2021, 12, 111 standard anomalies of GH (shaded) from 24 November to 3 December in 2018. (b) is the same as in11 of 17 (a) but for 925 hPa GH. (c) The vertical section of composite standard anomalies of GH across 120° E. The crossed area in each panel denotes anomalies exceeding the 0.01 significance level.

Figure 6. (a) The composite of original sea level pressure (SLP) (contours, unit: hPa) and standard Figureanomalies 6. (a )of The the compositeabsolute gradient of original of SLP sea (shade leveld) pressure from 24 November (SLP) (contours, to 3 December unit: hPa) in 2018. and ( standardb) anomalies of the absolute gradient of SLP (shaded) from 24 November to 3 December in 2018. (b) The composite analysis of original 200 hPa u winds (Only values more than 30 m·s−1 is shown as contours) and standard anomalies of 200 hPa u winds (shaded) from 24 November to 3 December in

2018. The red line in (b) is the climatological u = 30 m·s−1 at 200 hPa. The crossed area in each panel denotes anomalies exceeding the 0.01 signiﬁcance level.

3.4.2. Anomalous Thermal Circulations Matched well with the significant GH anomalies in the vertical section according to the theory of static equilibrium and the air state equation, pronounced warm temperature anomalies fill the whole middle and lower troposphere from 1000 hPa to 400 hPa (Figure7a ). This indicated that the air temperature rose over Eastern China from 24 November to 3 December in 2018. Figure8a,b shows the vertical distributions of regional and time- averaged temperature and standard temperature anomalies. Among the generally positive temperature anomalies with height in the whole troposphere under 400 hPa, the increases of air temperature at 925 and 850 hPa are the most significant with over 3.0 standard deviation anomalies, presenting an obvious inversion in the profiles of regional and time- averaged standard temperature anomaly and a nearly isothermal layer in those of the actual temperature. The vertical structure of inversion below 850 hPa is intimately related to the significant negative GH anomalies under 900 hPa based on the theory of static equilibrium. As the anomalous inversion in the lower troposphere enhances the stability of the near- surface atmosphere, it contributes to the formation and maintenance of fogs. It is apparent that the structure of the inversion is mainly attributed to the warm advection at 925 and 850 hPa associated with the southerly flows. As seen in the composite 925 hPa winds and air temperature averaged from 24 November to 3 December in 2018 (Figure7b ), in Eastern China, especially over Jiangsu Province, the northward wind is nearly perpendicular to the isotherm, inducing obvious warm temperature advection anomalies. The anomaly centers of the standard deviations of temperature and temperature advection are above 3.0, representing an unusual and even rare departure from average conditions (Figure7). Below 850 hPa, the distributions of regional and time-averaged temperature advection and the standard deviations at different heights (Figure8c,d) both show a distinct inversion. The standard deviation anomalies of temperature advection at 850 hPa are more than 2.5, indicating uncommon thermal circulation in the fog days compared with the climatology at the lower troposphere. Atmosphere 2021, 12, x FOR PEER REVIEW 12 of 17

Atmosphere 2021, 12, x FOR PEER REVIEW 12 of 17

The composite analysis of original 200 hPa u winds (Only values more than 30 m·s−1 is shown as Atmosphere 2021, 12, 111 Thecontours) composite and standardanalysis ofanomalies original 200of 200 hP ahPa u winds u winds (Only (shaded) values from more 24 than November 30 m·s− 1to is 3shown December12 as of 17 −1 contours)in 2018. The and red standard line in (anomaliesb) is the climatological of 200 hPa u uwi =nds 30 m·s(shaded)at 200 from hPa. 24 The November crossed areato 3 Decemberin each inpanel 2018. denotes The red anomalies line in (b )exceeding is the climatological the 0.01 significance u = 30 m·s level.−1 at 200 hPa. The crossed area in each panel denotes anomalies exceeding the 0.01 significance level.

Figure 7. (a) The vertical section of composite standard anomalies of air temperature across 120° E. Figure 7. (a) The vertical section of composite standard anomalies of air temperature across 120 ◦ E. The crossed area in each panel denotes anomalies exceeding the 0.01 significance level. (b) The FigureThe crossed 7. (a) The area vertical in each section panel denotesof composite anomalies standard exceeding anomalies the of 0.01 air significancetemperature level.across (120°b) The E. Thecomposite crossed of area original in each winds panel (Only denotes anomalies anoma excelies exceedingeding the 0.01the 0.01significance significance level level. are shown (b) The as composite of original−1 winds (Only anomalies exceeding the 0.01 significance level are shown as shafts; compositeshafts; units: of originalm·s ), original winds (Onlytemperature anomalies (contours, exceeding units: the °C) 0.01 and significance standard anomalies level are shown of temper- as units: m·s−1), original temperature (contours, units: ◦C) and standard anomalies of temperature shafts;ature advection units: m·s (Only−1), original the composite temperature standard (contours, anom units:alies exceeding°C) and standard the 0.10 anomalies significance of leveltemper- are advection (Only the composite standard anomalies exceeding the 0.10 significance level are shaded) atureshaded) advection at 925 hPa. (Only the composite standard anomalies exceeding the 0.10 significance level are shaded)at 925 hPa. at 925 hPa.

Figure 8. The profiles of the vertical distribution of composite (a) temperature, (b) standard anom- Figurealies of 8. temperature, The profiles ( ofc) temperaturethe vertical distribution advection, andof composite (d) standard (a) temperature,anomalies of (temperatureb) standard anom-advec- Figure 8. The profiles of the vertical distribution of composite (a) temperature, (b) standard anomalies aliestion averagedof temperature, over the (c )area temperature of 115–122° advection, E, 30–35° and N in(d )Eastern standard China anomalies from 24 of November temperature to 3advec- De- ofcember temperature, in 2018. (c) temperature advection, and (d) standard anomalies of temperature advection tion averaged over the area of 115–122°◦ E, 30–35°◦ N in Eastern China from 24 November to 3 De- cemberaveraged in over2018. the area of 115–122 E, 30–35 N in Eastern China from 24 November to 3 December in3.4.3. 2018. Anomalous Vapor Circulations 3.4.3.In Anomalous Anomalous the boundary Vapor Vapor air, Circulations weakening of the surface wind velocity during fog days (Figure 9a) is closely related to the southerly wind anomalies in the lower troposphere going InIn the boundaryboundary air,air, weakening weakening of of the the surface surface wind wind velocity velocity during during fog fog days days (Figure (Figure9a ) along with the weakened East Asian winter monsoon (Figures 5 and 6a). The weakened 9a)is closely is closely related related to the to southerlythe southerly wind wind anomalies anomalies in the in lower the lower troposphere troposphere going alonggoing velocity at the surface further hampers the export of fogs. Meanwhile, east of Eastern alongwith the with weakened the weakened East Asian East winterAsian winter monsoon monsoon (Figures (Figures5 and6a). 5 Theand weakened6a). The weakened velocity China lies a vast ocean and the central place of these fog events, Jiangsu Province, is ad- velocityat the surface at the further surface hampers further thehampers export the of fogs.export Meanwhile, of fogs. Meanwh east of Easternile, east China of Eastern lies a jacent to the Yellow Sea. The easterly wind can transport abundant vapor from the sea. Chinavast ocean lies a and vast the ocean central and placethe central of these place fog of events, these Jiangsufog events, Province, Jiangsu is Province, adjacent tois thead- jacentYellowThe composite to Sea. the The Yellow specific easterly Sea. humidity wind The easterly can advection transport wind abundantiscan over transport 3.0 vapor and abundant exceeds from the thevapor sea. 0.01 The from significance composite the sea. Thespecificlevel composite over humidity most specific regions advection humidity in East is over China advection 3.0 (Figure and exceedsis over9b). It 3.0 thecontributes and 0.01 exceeds significance to an the anomalous 0.01 level significance over persis- most levelregionstent supplementover in Eastmost China regions for high (Figure in humidityEast9b). China It contributes at (Figure the near-surface to9b). an It anomalous contributes layer. The persistent to ananomalous anomalous supplement saturated persis- for tenthighand humidsupplement humidity conditions at for the high near-surface over humidity Eastern layer. Chinaat the The arene anomalous ar-surfacealso presented layer. saturated significantly. The andanomalous humid The conditions dewsaturated point andoverdepression humid Eastern conditionsof China2 m is less are over alsothan Eastern presented3 °C. The China specific significantly. are also humidity presented The is dewmore significantly. point than 6 depression g/kg The with dew a of pointwave 2 m depressionisridge less of than the of 3specific 2◦ C.m is The lesshumidity specific than 3 line°C. humidity The along specific isth moree coast humidity than in the 6 is g/kg boundary more with than aduring 6 wave g/kg thewith ridge fog a ofwave days the ridgespecific(Figure of humidity9a,b).the specific The linesustained humidity along theweak line coast alongeasterly in thethe winds, boundarycoast inabundant the during boundary vapor the fog duringtransport, days (theFigure andfog9 daysa,bcon-). (FigureThedensation sustained 9a,b). in theThe weak boundary sustained easterly are winds,weak the easterly abundantindispensable winds, vapor circulationabundant transport, vaporfactors and condensationtransport, for the formation, and in con- the densationboundary in are the the boundary indispensable are the circulation indispensable factors circulation for the formation, factors enhancement,for the formation, and maintenance of fogs over Eastern China. Moreover, due to the anomalous southerly winds, the positive humidity anomalies at the middle latitudes are distributed to the whole troposphere (Figure9c ). Figure9c shows the center of composite original specific humidity

Atmosphere 2021, 12, x FOR PEER REVIEW 13 of 17

Atmosphere 2021, 12, 111 enhancement, and maintenance of fogs over Eastern China. Moreover, due to the13 anom- of 17 alous southerly winds, the positive humidity anomalies at the middle latitudes are distributed to the whole troposphere (Figure 9c). Figure 9c shows the center of composite original specific humidity across 120° E at the whole lower troposphere is more than 6 ◦ ◦ ◦ acrossg/kg 120and Ejust at theover whole 32° N~35 lower° troposphereN in the middle is more latitudes. than 6 g/kg The andcomposite just over standard 32 N~35 devia-N intions the middle of specific latitudes. humidity The over composite that region standard are mostly deviations over of 3.0, speciﬁc indicating humidity the rarity over thatof the regionvapor are during mostly the over fog 3.0,events. indicating the rarity of the vapor during the fog events.

FigureFigure 9. 9.(a ()a The) The composite composite 10-m 10-m winds winds (only (only anomalies anomalies exceeding exceeding thethe 0.010.01 significancesignificance levellevel are −1 shownshown as as shafts, shafts, units: units: m m·s·s−1),), the the standardstandard deviation deviation ofof 1010 mm windwind velocityvelocity standardstandard deviationdeviation ofof 10 m wind velocity (only anomalies exceeding the 0.01 significance level are shaded), and the 2 m 10 m wind velocity (only anomalies exceeding the 0.01 significance level are shaded), and the 2 m dew point depression (contours are values less than 4 °C). The green oblique line area denotes the dew point depression (contours are values less than 4 ◦C). The green oblique line area denotes the anomalies of 2-m dew point depression exceeding the 0.01 significance level. (b) The composite anomalies1000 hPa ofspecific 2-m dew humidity point depression(contours are exceeding values more the 0.01than significance6g·kg−1), 1000 level. hPa vapor (b) The flux composite (only −1 1000anomalies hPa specific exceeding humidity the 0.01 (contours significance are values level are more shown than as 6g vectors,·kg ), units: 1000 hPa10−2g·kg vapor−1·m·s flux−1), (only and − − − anomaliesstandard exceeding deviation theof specific 0.01 significance humidity level advectio are shownn (shaded). as vectors, The crossed units: 10area2g denotes·kg 1·m anomalies·s 1), and of standardspecific deviation humidity ofadvection specific humidityexceeding advection the 0.01 significance (shaded). The level. crossed (c) The area vertical denotes section anomalies of com- of specificposite humidity original specific advection humidity exceeding (con thetours 0.01 are significance the values level. more (c than) The 2g·kg vertical−1) and section standard of composite devia- originaltion ofspecific specific humidityhumidity (contours(shaded) across are the 120° values E. The more crossed than 2garea·kg denotes−1) and anomalies standard deviation of specific of specifichumidity humidity exceeding (shaded) the 0.01 across significance 120◦ E. The level. crossed area denotes anomalies of specific humidity exceeding the 0.01 significance level. Overall, the anomalous atmospheric circulation pattern in the entire troposphere duringOverall, the extremely the anomalous persistent atmospheric dense fog circulation events over pattern Eastern in China the entire in the troposphere late autumn duringof 2018 the can extremely be concluded persistent in the dense schematic fog events diagram over shown Eastern in China Figure in 10. the During late autumn the fog ofdays, 2018 canthe anticyclonic be concluded circulation in the schematic anomaly diagram in the north shown of inEAJS Figure and 10 the. During cyclonic the circula- fog days,tionthe anomaly anticyclonic in the circulation south of EAJS anomaly broaden in the and north weaken of EAJS the and core the of cyclonic the jet over circulation Eastern anomalyChina. The in the East south Asian of EAJStrough broaden with the and anticyclonic weaken the circulation core of the anomaly jet over Easternat 500 hPa China. is ef- Thefectively East Asian weakened. trough withMeanwhile, the anticyclonic in the lower circulation layer in anomaly Eastern at China, 500 hPa the is cyclonic effectively and weakened.anticyclonic Meanwhile, circulation in anomalies the lower layeroccur in in Eastern the west China, and the the east, cyclonic respectively, and anticyclonic resulting circulationin the weakening anomalies of occur East inAsian the west winter and monsoon the east, respectively,and the enhancement resulting in of the the weakening anomalous ofsoutherly East Asian flows winter that monsoon transport and abnormal the enhancement warm and of the wet anomalous air to Eastern southerly China. flows The that ab- transportnormal abnormalcirculation warm pattern and in wet the air tropospher to Easterne China.hampers The the abnormal occurrence circulation of any vortex pattern ac- in the troposphere hampers the occurrence of any vortex activities or cold air invasion. The atmosphere is inclined to be stable, and the vapor condensation and adequate vapor amount provide the necessary metrological conditions for fog as well. Correspondingly, the

Atmosphere 2021, 12, x FOR PEER REVIEW 14 of 17

tivities or cold air invasion. The atmosphere is inclined to be stable, and the vapor condensation and adequate vapor amount provide the necessary metrological conditions for fog as well. Correspondingly, the sustained anomalous northward wind at the lower troposphere and anomalous westward wind in the near-surface both intensify the vapor supplement. Atmosphere 2021, 12, 111 It is noteworthy that the numbers of the standardized anomalies of 1000–20014 ofhPa 17 height, temperature, wind, and moisture fields during these fog days all significantly depart from climatology for that locale and time of the season. It has been further demonstrated that the persistent dense fogs over Eastern China in the late autumn of sustained2018 are a anomalous significant northward unusual weather wind at theevent lower with troposphere extreme synoptic-scale and anomalous departures westward windfrom normal in the near-surface [32]. both intensify the vapor supplement.

Figure 10. Schematic diagram of the anomalous atmospheric circulations during the extremely Figure 10. Schematic diagram of the anomalous atmospheric circulations during the extremely persistent dense fog events over Eastern China in the late autumn of 2018. persistent dense fog events over Eastern China in the late autumn of 2018.

4. ConclusionsIt is noteworthy and Discussion that the numbers of the standardized anomalies of 1000–200 hPa height, temperature, wind, and moisture fields during these fog days all significantly depart fromGlobal climatology warming for thatis generally locale and expected time of to the intensify season. extreme It has been weather further events demonstrated [43]. Me- thatteorological the persistent events dense directly fogs impact over Eastern processes China taking in the place late autumnat the Earth’s of 2018 surface. are a significant Fog with unusuallow visibility weather is a eventkind of with hazardous extreme weather. synoptic-scale Fog impacts departures transportation from normal difficulties, [32]. air quality and public health [4,44]. 4. ConclusionsFrom mid–late and DiscussionNovember to early December in 2018, extremely persistent dense fog eventsGlobal with warminglong duration, is generally low expectedvisibility, toand intensify extensive extreme coverage weather occurred events [in43 ].Eastern Mete- orologicalChina, especially events directlyin Jiangsu impact Province, processes a coasta takingl province place aton the the Earth’s shore of surface. the Yellow Fog with Sea. lowThe visibilityheavy fog is weather a kind ofhad hazardous an unprecedented weather. Fogsevere impacts effect transportation on the road and difficulties, air transport air qualityand was and widely public reported health [at4, 44home]. and abroad. From mid–late24 November November to 3 December to early Decemberin 2018, the in fog 2018, events extremely were characterized persistent dense by fogthe eventsstrongest with intensity long duration, and the low longest visibility, duration and extensive for that coverage year. At occurred least 24 in stations, Eastern China,about especiallyone-third inof Jiangsuall the 70 Province, stations a coastalover Jiangsu province Province, on the shore observed of the visibilities Yellow Sea. of The less heavy than fog1000 weather m during had this an period. unprecedented In particular, severe the effect SDF (visibility on the road less and than air 200 transport m) that and covered was widelyat least reported24 stations at homeand persisted and abroad. for five days from 25 November to 29 November was rare Fromand impressive. 24 November The tofog 3 Decemberdays averaged in 2018, overthe Jiangsu fog events in November were characterized and December by the2018 strongest were 8.69 intensity d and 6.47 and d, the respectively, longest duration and compared for that year. with Atthe least climatology 24 stations, of fogs about in one-thirdlate autumn, of allthe the anomalies 70 stations were over 1.0~2.0 Jiangsu standard Province, deviations observed from visibilities the grand of mean. less than The 1000 m during this period. In particular, the SDF (visibility less than 200 m) that covered at least 24 stations and persisted for five days from 25 November to 29 November was rare and impressive. The fog days averaged over Jiangsu in November and December

2018 were 8.69 d and 6.47 d, respectively, and compared with the climatology of fogs in late autumn, the anomalies were 1.0~2.0 standard deviations from the grand mean. The number of fog days in November 2018 ranked the third-highest since 1961 and the ﬁrst since the beginning of the 21st century despite a declining trend after the 1990s. Undoubtedly, the evolution of each meteorological factor conducive to fog is consistent with that of the daily visibility over Jiangsu Province. To sum up, the common favorable Atmosphere 2021, 12, 111 15 of 17

metrological conditions include weak dynamic progress, stable atmospheric stratification, strong inversion in the lower troposphere, and affluent and saturated air near the surface. For instance, dynamic processes such as weaker outward horizontal transport, more passive vertical diffusion in the lower troposphere and suitable turbulent conditions in the fog boundary layer impose a critical impact on the development and persistence of lower visibility. During the fog days, the formation and maintenance of the inversion structure, an important factor for the stability of the atmosphere, are connected with the relative cooling near the surface and the warm advection at the lower tropospheric layer. It should also be emphasized that the absolute and relative humidity conditions are both indispensable for such persistent dense fogs with extensive coverage. The persistent dense fogs over Eastern China in the late autumn of 2018 are not only an unusual weather event with synoptic-scale departures from normal but also a significantly abnormal climatic event with the third most duration days since 1961. The anomalous atmospheric circulation pattern in the entire troposphere, which matches the extreme fogs well, is analyzed in detail as follows. During the fog days, the East Asian jet stream broadened and the core of the jet over Eastern China weakened associated with the anticyclonic circulation anomaly in the north of EAJS and the cyclonic circulation anomaly in its south. In the middle troposphere, the East Asian trough with the anticyclonic circulation anomaly at 500 hPa was present in a weaker amplitude. In the lower layer in Eastern China, the cyclonic and anticyclonic circulation anomalies in the west and the east resulted in weakened East Asian winter monsoon. All the abnormal circulation conditions in the troposphere were inclined to hamper vortex activities or cold air invasion and thus intensify the atmospheric stability. By confining the affluent and saturated air near the surface, the sustained anomalous northward wind at the lower troposphere and anomalous westward wind in the near-surface play a great part in the vapor supplement. The revelation of anomalous circulation and abnormal features of synoptic systems that provide favorable meteorological conditions for fog effectively enhances the capability to forecast extreme fog in the short-term, medium-term and even sub-seasonal timescales. Although this research only deals with a case study, our findings of the abnormal features of fog and the associated circulation situation based on synoptic-scale departures from normal lay the foundation for unraveling the underlying physical processes behind the impact of extra-forcing on fog at interannual or even longer timescales. In future work, more historical rare fog cases will be researched to further verify our robust schematic diagram. From a synoptic point of view, the dense fog events over Eastern China in the late autumn of 2018 could be described as an extreme or a rare event. There are still many difficulties in producing accurate and effective forecasts for similar severe fog events that are characterized by long duration, strong intensity and extensive coverage on the medium-term and even sub-seasonal timescales. It is necessary to find objective methods or standards to rank such fog events, and their predictability needs to be further investigated in future studies.

Author Contributions: The study was completed with cooperation among all the authors: Concep- tualization, S.C., D.L. and Z.K.; data and methodology, Y.S.; writing—review and editing, S.C. and D.L.; visualization, M.L.; project administration, Z.K. and D.L. All authors have read and agreed to the published version of the manuscript. Funding: This work was jointly supported by the National Key Project of MOST (2016YFC0203303 and 2016YFC0201903), the open fund by the Key Laboratory for Aerosol–Cloud–Precipitation of CMA–NUIST in China (KDW1801), the National Natural Science Foundation of China (91544231 and 41575135), and the Jiangsu meteorological bureau key fund (KZ201801). Institutional Review Board Statement: Not applicable. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: The circulation data is publicly available datasets analyzed in this study. The data can be found here: https://apps.ecmwf.int/datasets/data/interim-full-daily/levtype=sfc/. Atmosphere 2021, 12, 111 16 of 17

Restrictions apply to the availability of the fog observation data. Data was obtained from China Meteorological Administration and are available from the authors of this paper with the permission of China Meteorological Administration. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Jiusto, J.E. Fog structure. In Clouds, Their Formation, Optical Properties and Effect; Hobbs, P.V., Deepak, A., Eds.; Academic Press: New York, NY, USA, 1981; pp. 187–239. 2. Nemery, B.; Hoet, P.H.; Nemmar, A. The Meuse Valley fog of 1930, an air pollution disaster. Lancet 2001, 357, 704–708. [PubMed] 3. Houssos, E.E.; Lolis, C.J.; Bartzokas, A. The main characteristics of atmospheric circulation associated with fog in Greece. Nat. Hazard. Earth Syst. 2009, 9, 1857–1869. 4. Gautam, R.; Singh, M.K. Urban heat island over Delhi Punches Holes in widespread fog in the Indo–Gangetic Plains. Geophys. Res. Lett. 2018, 45, 1114–1121. 5. Li, Z.H.; Yang, J.; Shi, C.E. The Physics of Regional Dense Fog; China Meteor Press: Beijing, China, 2008; p. 160. (In Chinese) 6. Ye, H. The influence of air temperature and atmospheric circulation on winter fog frequency over Northern Eurasia. Int. J. Climatol. 2009, 29, 729–734. 7. Emmons, G.; Montgomery, R.B. Notes on the physics of fog formation. J. Meteorol. 1947, 4, 206. 8. Choularton, T.W.; Fullarton, G.; Latham, J.; Mill, C.S. A field study of radiation fog in Meppen, West Germany. Q. J. R. Meteorol. Soc. 1981, 107, 381–394. 9. Brown, R.; Roach, W.T. The physics of radiation fog, II–a numerical study. Q. J. R. Meteorol. Soc. 1976, 102, 335–354. 10. Herckes, P.; Marcotte, A.R.; Wang, Y.; Collett, J. Fog composition in the Central Valley of California over three decades. Atmos. Res. 2015, 151, 20–30. 11. Degefie, D.T.; El–Madany, T.S.; Hejkal, J.; Held, J. Microphysics and energy and water fluxes of various fog types at SIRTA, France. Atmos. Res. 2015, 151, 162–175. 12. Stolaki, S.; Haeffelin, M.; Lac, C.; Dupont, J.C.; Elias, T.; Masson, V. Influence of aerosols on the life cycle of a radiation fog event. A numerical and observational study. Atmos. Res. 2015, 151, 146–161. 13. Guedalia, D.; Bergot, T. Numerical Forecasting of Radiation Fog. Part II, A Comparison of Model Simulation with Several Observed Fog Events. Mon. Weather Rev. 1994, 122, 1231–1246. [CrossRef] 14. Houssos, E.E.; Lolis, C.J.; Gkikas, A.; Hatzianastassiou, N.; Bartzokas, A. On the atmospheric circulation characteristics associated with fog in Ioannina, north–western Greece. Int. J. Climatol. 2011, 32, 1847–1862. [CrossRef] 15. Zhang, R.H.; Li, Q.; Zhang, R.N. Meteorological conditions for the persistent severe fog and haze event over eastern China in January 2013. Sci. China Earth Sci. 2014, 57, 26–35. 16. Telišeli–Prtenjak, M.; Klaić,M.; Cuxart, J.; Jerirević,A. The interaction of the downslope winds and fog formation over the Zagreb area. Atmos. Res. 2017, 214, 213–227. [CrossRef] 17. Willett, H.C. Fog and Haze, Their Causes, Distribution, and Forecasting. Mon. Weather Rev. 1928, 56, 435–468. [CrossRef] 18. Meyer, M.B.; Lala, G.G. Climatological aspects of Radiation Fog Occurrence at Albany, New York. J. Clim. 1990, 3, 577–586. [CrossRef] 19. Bergot, T.; Guedalia, D. Evaluation of the success of fog forecasting by a numerical model. Meteorologie 1996, 14, 27–35. 20. Van der Velde, I.R.; Steeneveld, G.J.; Wichers Schreur, B.G.J.; Holtslag, A.A.M. Modeling and Forecasting the Onset and Duration of Severe Radiation Fog under Frost Conditions. Mon. Weather Rev. 2010, 138, 4237–4253. [CrossRef] 21. Steeneveld, G.J.; Ronda, R.J.; Holtslag, A.A.M. The Challenge of Forecasting the Onset and Development of Radiation Fog Using Mesoscale Atmospheric Models. Bound. Lay. Meteorol. 2015, 154, 265–289. [CrossRef] 22. Herman, G.R.; Schumacher, R.S. Using Reforecasts to Improve Forecasting of Fog and Visibility for Aviation. Weather Forecast. 2016, 31, 467–482. [CrossRef] 23. Meng, Y.J.; Wang, S.Y.; Zhao, X.F. An analysis of air pollution and weather conditions during heavy fog days in Beijing area. Meteorl. Mon. 2000, 26, 40–42. 24. Liu, X.N.; Zhang, H.Z.; Li, Q.X.; Zhu, Y.J. Preliminary research on the climatic characteristics and change of fog in China (in Chinese). J. Appl. Meteorol. Sci. 2005, 16, 220–229. 25. Lin, J.; Yang, G.M.; Mao, D.Y. Spatial and temporal characteristics of fog in China and associated circulation patterns (in Chinese). Clim. Environ. Res. 2008, 13, 171–181. 26. Wu, D.; Bi, X.Y.; Deng, X.J.; Li, F.; Tan, H.B.; Liao, G.L.; Huang, J. Effect of atmospheric haze on the deterioration of visibility over the Pearl River Delta. Acta Meteorol. Sin. 2007, 21, 215–223. 27. Ding, Y.H.; Liu, Y.J. Analysis of long–term variations of fog and haze in China in recent 50 years and their relations with atmospheric humidity. Sci. China Earth Sci. 2014, 57, 36–46. [CrossRef] 28. Ma, N.; Zhao, C.S.; Chen, J.; Xu, W.Y.; Yan, P.; Zhou, X.J.A. novel method for distinguishing fog and haze based on pm2.5, visibility, and relative humidity. Sci. China Earth Sci. 2014, 57, 2156–2164. [CrossRef] 29. Zhang, X.Y.; Wang, Y.Q.; Niu, T.; Zhang, X.C. Atmospheric aerosol compositions in China, spatial/temporal variability, chemical signature, regional haze distribution and comparisons with global aerosols. Atmos. Chem. Phys. 2012, 12, 779–799. [CrossRef] Atmosphere 2021, 12, 111 17 of 17

30. Sun, Y.; Ma, Z.F.; Niu, T.; Fu, R.Y.; Hu, J.F. Characteristics of climate change with respect to fog days and haze days in China in the past 40 years(in Chinese). Clim. Environ. Res. 2013, 18, 397–406. 31. Wu, D. More discussions on the differences between haze and fog in city (in Chinese). Meteorl. Mon. 2006, 32, 9–15. 32. Hart, R.E.; Grumm, R.H. Using Normalized Climatological Anomalies to Rank Synoptic–Scale Events Objectively. Mon. Weather Rev. 2001, 129, 2426–2442. [CrossRef] 33. Dee, D.P.; Uppala, S.M.; Simmons, A.J.; Berrisford, P.; Poli, P.; Kobayashi, S.; Andrae, U.; Balmaseda, M.A.; Balsamo, G.; Bauer, P.; et al. The ERA-interim reanalysis, configuration and performance of the data assimilation system. Q. J. R. Meteorol. Soc. 2011, 137, 553–597. [CrossRef] 34. Liu, D.Y.; Yang, J.; Niu, S.J.; Li, Z.H. On the evolution and structure of a radiation fog event in Nanjing. Adv. Atmos. Sci. 2011, 28, 223–237. [CrossRef] 35. Eady, E.T. Long waves and cyclone waves. Tellus 1949, 1, 33–52. [CrossRef] 36. Hoskins, B.J.; Valdes, P.J. On the existence of storm–tracks. J. Atmos. Sci. 1990, 47, 1854–1864. [CrossRef] 37. Lindzen, R.S.; Farrell, B. A simple approximate result for the maximum growth rate of baroclinic instabilities. J. Atmos. Sci. 1980, 37, 1648–1654. [CrossRef] 38. Rodhe, B. The effect of turbulence on fog formation. Tellus 1962, 14, 49–86. [CrossRef] 39. Zhou, B.; Ferrier, B.S. Asymptotic Analysis of Equilibrium in Radiation Fog. J. Appl. Meteorol. Clim. 2008, 47, 1704–1722. [CrossRef] 40. Grumm, R.H.; Hart, R.E. Standardized Anomalies Applied to Significant Cold Season Weather Events, Preliminary Findings. Weather Forecast. 2001, 16, 736–754. [CrossRef] 41. Wang, J.; Zhao, Q.H.; Zhu, Z.W.; Qi, L.; Wang, J.X.L.; He, J.H. Interannual variation in the number and severity of autumnal haze days in the Beijing–Tianjin–Hebei region and associated atmospheric circulation anomalies. Dyn. Atmos. Ocean. 2018, 84, 1–9. [CrossRef] 42. Yin, Z.C.; Wang, H.J. Role of atmospheric circulations in haze pollution in December 2016. Atmos. Chem. Phys. 2017, 17, 11673–11681. [CrossRef] 43. Diffenbaugh, N.S. Verification of extreme event attribution: Using out-of-sample observations to assess changes in probabilities of unprecedented events. Sci. Adv. 2020, 6, eaay2368. [CrossRef][PubMed] 44. Kim, H.; Collier, S.; Ge, X.; Xu, J.; Sun, Y.; Jiang, W.; Wang, Y.; Herckes, P.; Zhang, Q. Chemical processing of water-soluble species and formation of secondary organic aerosol in fogs. Atmos. Environ. 2019, 200, 158–166. [CrossRef]