Wnjun Ling , Zhn Yng , Jili Luo , * , Hongying Tin , Zhixun Bi , Dn Li , Qin Li ,Jinqing Zhng , Hoyu Wng , Bin B , Yng Yng

a Key Laboratory of Semi-Arid Climate Change and College of Atmospheric Sciences, Lanzhou University, Lanzhou, China

b Key Laboratory of Middle Atmosphere and Global Environment Observation, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

c Department of Atmospheric Science, Yunnan University, Kunming, China

d Atmospheric Observation Centre of Lhasa Meteorological Bureau, Lhasa, China

e Institute of Arid meteorology, China Meteorological Administration, Lanzhou, China

Keywords:Ozone Tibetan Plateau Atmospheric apparent heat source Ascending motion

ABSTRACT The Tibetan Plateau (TP) is an area sensitive to climate change, where the ozone distribution affects the atmospheric environment of the TP and its surrounding regions. The relatively low total column ozone over the TP in boreal summer and its spatiotemporal variations have received extensive attention. In this study, five-year balloon-borne measurements of ozone over Lhasa in boreal summer are used to investigate the influences of the apparent heat source (Q1) on the ozone vertical structure over the plateau. The mechanisms for the above processes are also explored. The results show that the tropospheric ozone mixing ratio over Lhasa decreases when the total atmospheric Q1 in the troposphere over the TP is relatively high. Strengthened ascending motions are accompanied by enhanced Q1 over the main TP region. Consequently, the tropospheric ozone mixing ratio over Lhasa decreases when Q1 is higher in summer, which is attributed to the upward transport of the ozone-poor surface air.

1. Introduction

Ozone is known as an important atmospheric component because of its role in the radiative heat budget ( Ramaswamy et al., 1992 ;Forster and Shine, 1997 ). Atmospheric ozone is concentrated in the stratosphere and peaks at about 20—30 km altitude, while tropospheric ozone only accounts for 10% of the total ozone content. As one of the air pollutants in the troposphere, a “good gas ”in the stratosphere that protects life on Earth from harmful solar ultraviolet radiation ( Kerr and McElroy, 1993 ), and a greenhouse gas that affects global climate change( Wang and Sze, 1980 ; Lacis et al., 1990 ), the distribution and variation of ozone have been investigated widely ( Zhang et al., 2002 ; Angell and Free, 2009 ; WMO, 2011 ; Krzy ? cin, 2012 ; Zhang et al., 2014 ). On the one hand, the observed total ozone column in the middle latitudes demonstrates a significant decline since the 1980s owing to the emissions of anthropogenic chlorofluorocarbons (CFCs) ( WMO, 2011 ), and there is a weak recovery signal in the total ozone column after 2000 because of the control of CFC emissions in the atmosphere in response to the Montreal Protocol signed in 1987 (e.g., Angell and Free, 2009 ; Krzy ? cin, 2012 ;Zhang et al., 2014 ). On the other hand, the tropospheric ozone mixing ratio increased during the 20th century owing to increasing anthropogenic emissions of ozone precursors ( Cooper et al., 2014 ). Fully understanding the vertical distribution of ozone and its variation is important in climate and atmospheric environment assessments.

Often referred to as “the third pole ”, the Tibetan Plateau (TP) is the highest plateau on Earth. It plays an important role in global weather and climate by modulating atmospheric circulation ( Zhou et al., 2009 ;Zhao et al., 2018 ). As a key pathway of global stratosphere—troposphere exchange (STE) owing to its complex terrain and dynamical and thermal features ( Gettelman et al., 2004 ; Fu et al., 2006 ), the TP also influences the atmospheric ozone distribution ( Tian et al., 2008 ). Features of the ozone distribution over the TP have attracted great attention since the summer “ozone valley ”and winter ozone minimum over the TP were discovered ( Zhou et al., 1995 ; Zou, 1996 ; Bian et al., 2006 ; Guo et al.,2017 ). The main mechanisms for the formation of the relatively low total column ozone over the TP include topographic lift, upward transport of surface ozone-poor air to the stratosphere via STE, and chemical processes ( Liu et al., 2003 ; Ye and Xu, 2003 ; Zhou et al., 2006 ; Tobo et al.,2008 ; Tian et al., 2008 ; Bian et al., 2011 ; Zhao et al., 2018 ). Among the above factors, topographic lift and the STE associated with upward transport play key roles in the summertime ozone vertical distribution over the TP. During boreal summer, the TP is recognized as an elevated heat source for the atmosphere ( Reiter and Gao, 1982 ). The accompanying strong ascending motions over the TP can induce upward transport of low-level air in this region, leading to a redistribution of ozone in the upper levels ( Wang, 2006 ; Luo et al., 2018 ). It is still not clear to what extent heating of the TP impacts on the ozone vertical structure.

Due to their temporal and spatial continuity, satellite observations are widely used in studying the ozone distribution study over the TP( Bian et al., 2006 , 2011 ; Zhou et al., 2006 ; Guo et al., 2012 ; Wang et al.,2016 ). However, the vertical resolution of satellite datasets is not fine enough for examining the vertical structure of atmospheric elements.

In-situ

measurements, especially balloon-borne measurements, collected over the TP, provide high-resolution and highly accurate data,which can be used to explore the ozone structure and variation under the impacts of different thermal and dynamical factors during the Asian summer monsoon season. In this study, ozone vertical profiles obtained from the Sounding Water vapor, Ozone, and Particle (SWOP) campaign( Bian et al., 2012 ; Li et al., 2017 , 2018 ) at Lhasa (30°N, 91°E; elevation: 3650 m) are used to investigate the vertical distribution features of ozone and its relationship with the atmosphere apparent heat source over the TP. Note that Lhasa is located within the main body of the TP,which provides ideal settings for the present study to reveal the influences of the TP’s heat on the overlying vertical structure of ozone.

This study aims to identify the impact of heating over the TP on the distribution of ozone. Section 2 describes the data and analysis method.Ozone distributions over Lhasa associated with heat source variations,as well as the underlying mechanism, are discussed in Section 3 . A summary and conclusions are presented in Section 4 .

2. Data and methods

Ozone profiles were obtained from the SWOP campaign during the Asian summer monsoon season from 2010 to 2019. Balloons were launched from Lhasa in 2010 (7 soundings in August), 2013 (22 soundings in August), 2016 (2 soundings in July and 12 soundings in August),2018 (1 sounding in July and 12 soundings in August), and 2019 (1 sounding in July and 8 soundings in August). More detailed information on the campaign and payloads can be found in Bian et al. (2012) and Li et al. (2017 , 2018 , 2020 ). All the observations were made at nighttime, so that the influence of photochemical reactions on tropospheric ozone was negligible.

ERA-Interim is a reanalysis dataset produced by the European center for Medium-Range Weather Forecasts ( Dee et al., 2011 ). Its horizontal resolution is 1°×1°(latitude ×longitude). In this study, daily data from 2010 to 2019 are extracted from ERA-Interim to calculate the atmospheric apparent heat source (Q1). The vertical velocity, wind field, air temperature, and ozone mixing ratio at 37 pressure levels are used in this study.

The calculation of Q1 is based on the temperature field, wind field,vertical velocity, and pressure. The equations of Q1 at each level, Eq. (1) ,and the tropospheric integrated 〈Q1 〉, Eq. (2) , are as follows ( Yanai et al.,1973 ; Zhang et al., 2016 ):

Where

T

is the atmosphere temperature of different levels,

t

refers to the time integration interval (here we adopted 6 h interval according to the given reanalysis datasets),

θ

is the potential temperature,

V

is the horizontal winds,

ω

is the vertical velocity under the P-coordinate system,

k

=

R

/

C

p

, and

R

and

C

p

represent the specific gas constant and specific isobaric heat capacity, respectively.

P

= 1000 hPa,

P

is the ground pressure, and

P

1 = 100 hPa here. The Q1 of the large-scale motion system consists of the heating due to radiation, the release of latent heat by net condensation, and vertical convergence of the vertical eddy transport of sensible heat. In this study, Q1 is calculated in the region (20°—50°N, 60°—110°E), and we pay more attention to the Q1 over the major region of the TP (29°—40°N, 80°—104°E). Q1 in our study indicates the tropospheric integrated atmospheric apparent heating.

3. Results

We first evaluate the seasonal mean ozone vertical structures over Lhasa and their relationships with Q1 over the major region of the TP.Fig. 1 (a) shows the interannual variabilities of JJA (June—July—August)mean Q1 from 2010 to 2019. What we found is that, during this study period, Q1 fluctuates irregularly. 2010 and 2018 are identified to be relatively high Q1 years (based on the mean Q1 in JJA for all 10 years).2013, 2016, and 2019 are relatively low Q1 years.

Seasonal mean ozone vertical profiles at Lhasa are shown in Fig. 1 (b).The ozone mixing ratios at Lhasa differ from year to year in the troposphere but remain almost the same in the stratosphere. Tropospheric ozone mixing ratios over Lhasa are lower in the relatively high Q1 years than in other years, which is attributed to the stronger ascending motions in higher Q1 years that induce upward transport of the ozone-poor air from lower levels ( Wang, 2006 ; Luo et al., 2018 ).

Fig. 1. (a) Interannual variabilities of the vertically integrated Q1 from the surface to 100 hPa (units: W m?2 ) over the major region of the TP (29°—40°N,80°—104°E) during boreal summer (JJA mean) from 2010 to 2019. The red dots indicate years of relatively high Q1 (based on the mean Q1 in JJA for all 10 years). Blue dots denote relatively low-value years. (b) JJA mean vertical profiles of ozone mixing ratio over Lhasa for individual years. The red lines are for years of relatively high seasonal mean Q1, and blues ones are for years of lower Q1.

The daily Q1 over the major region of the TP for each summer also fluctuates irregularly ( Fig. 2 ). Taking the 0.5 σ(standard deviation) as a criterion in every year, the subseasonal Q1 variations differ from day to day and high Q1 days mainly occur in July. Whether the daily ozone distribution is related to Q1 variation is worthy of investigation.

Fig. 2. Subseasonal variabilities of the vertically integrated Q1 from the surface to 100 hPa (units: W m?2 ) over the major region of the TP during boreal summer of year 2010, 2013, 2016, 2018, and 2019. The red dots indicate days of relatively high Q1 (based on + 0.5 σ(standard deviation) for each year). Blue dots denote relatively low-value days. 6_01 represents 1 June and so on.

Fig. 1 (b) indicates that the ozone mixing ratio over Lhasa decreases when Q1 is relatively high. However, the results are not so robust since the samples are limited. Therefore, we further examine the Q1 horizontal distribution under relatively high and low Q1 days, and the correlation between the ozone mixing ratio and Q1 in different high levels, to verify the relationships discussed above.

Fig. 3 (a—c) show the horizontal distributions of the summer mean Q1, the mean Q1 for all the relatively high Q1 days (125 days in total),and that for all the relatively low Q1 days (127 days in total), retrieved during the selected years. It is evident that, although the Q1 over the main region of the TP is significantly higher than normal on the relatively high Q1 days, it remains similar to the normal value over the surrounding areas of the TP in summer. Specifically, the Q1 over Lhasa can be much higher than its seasonal mean on some days. This is because Lhasa is located within the main body of the TP. The enhanced integrated tropospheric Q1 over the TP is closely associated with the vertical motions over Lhasa.

Fig. 4. Vertical profiles of correlation coefficients between daily ozone mixing ratio and Q1 over Lhasa. The red lines represent the correlation coefficients of ozone with Q1 on the same day. Blue, green, and black lines represent 1-day,2-day, and 3-day lagged correlation coefficients between ozone mixing ratio(lagged) and Q1. The results are based on all the observations during the years in Fig. 1 (b). The thick lines are for values statistically significant at the 0.1 level.

Fig. 3. Horizontal distributions of JJA (a) mean Q1, (b) mean Q1 for all relatively high Q1 days (based on + 0.5 σ(standard deviation) for each year), and (c) mean Q1 for all relatively low Q1 days (based on ? 0.5 σ(standard deviation) for each year) during the years in Fig. 1 (b). Color shading represents Q1. The red contour indicates the location of the TP (isoline of 3 km elevation). The black solid frame is the major region of the TP. The red star shows the location of Lhasa (30°N, 91°E).Dotted areas are for values statistically significant at the 0.1 level, based on the Student’s t -test.

Previous studies have shown that deep convection has a significant effect on the Asian summer monsoon anticyclone ( Hoskins and Rodwell, 1995 ; Fujinami and Yasunar, 2004 ; Randel and Park, 2006 ).Randel and Park (2006) pointed out that there is an apparent time lag between deep convections and the circulation variations. Thus, the correlations between the daily Q1 and ozone mixing ratio over the entire vertical range of observations at Lhasa are examined, and time lags between Q1 and ozone content are considered ( Fig. 4 ). Consistent with the seasonal mean shown in Fig. 1 (b), Q1 is anti-correlated with the tropospheric ozone vertical distribution at Lhasa. In addition, the correlation coefficient between the Q1 and ozone mixing ratio at Lhasa is about ? 0.5 below 400 hPa, and it decreases with increasing altitude.Note that the time lag also has an impact on it, which shows that a oneday delay has the best anti-correlation below 200 hPa, and reaches the peak value at nearly 500 hPa.

Fig. 5. Scatterplots of ozone mixing ratio at 500 hPa versus Q1 at Lhasa during the years in Fig. 1 (b). Each point represents a pair of observations of ozone and daily Q1. The black lines show linear fits. The correlation coefficient and confidence level for each individual linear fit are given at the top of the corresponding panel.The values (a) without any time lag, (b) with 1-day lag, (c) with 2-day lag, and (d) with 3-day lag of ozone mixing ratio all are shown.

Fig. 5 shows scatterplots of all the observations of daily ozone mixing ratio around 500 hPa versus Q1 at Lhasa during 2010—19. As shown in Fig. 5 , the correlations of the daily 500-hPa ozone mixing ratio over Lhasa and the Q1 are all greater than or equal to ? 0.5. When the time lag is considered, the peak value ( ? 0.57) is found. This indicates that the ozone mixing ratio at 500 hPa (lower troposphere) over Lhasa decreases when Q1 over the TP is relatively high, and the influence of Q1 lasts several days.

To verify the stronger ascending motions associated with higher Q1 result in the difference in ozone vertical distributions between different Q1 days or different years, the vertical velocity over the TP is examined in Fig. 6 . Fig. 6 (a) shows the latitude—height cross section of the difference in vertical velocity along 91°E between the days with relatively high Q1 and low Q1 for JJA during the years in Fig. 1 (b). It is evident that significant upward anomaly flow is found over the plateau.The upward motions can reach the upper troposphere. The ozone-poor air near the surface could be transported to upper levels by the stronger ascending motions over the TP when Q1 is relatively high. Fig. 6 (b)complements the horizontal features of the difference in vertical velocity between the days with relatively high Q1 and low Q1 for JJA during the years in Fig. 1 (b). Stronger ascending motions (or weaker descending motions) are found over the whole plateau and its southern edge. As Lhasa is located in the upward anomaly regions, more ozone-poor air will be observed during the higher Q1 days.

From the above analysis, we can see that the decreasing of the tropospheric ozone mixing ratio with altitude (see also Fig. 1 (b)) is associated with stronger upward motions during the relatively high Q1 days and years. The correlation coefficients are about ? 0.5 around 400 hPa and decrease with altitude. Thus, we suppose that the effects of the Q1 on the distribution of ozone can be attributed to airmass transport.

Fig. 6. (a) Latitude—height cross section of the difference in vertical velocity (units: Pa s? 1 ) along 91°E between the days with relatively high Q1 and low Q1 for JJA (the same as Fig. 3 ). Solid lines indicate descending motions and dashed lines represent ascending motions. Shaded areas are for values statistically significant at the 0.1 level. (b) Horizontal distribution of the difference in vertical velocity between the days with relatively high Q1 and low Q1 for JJA during the years in Fig. 1 (b). Solid lines, dashed lines, and shaded areas carry the same meaning as in (a). The blue lines indicate the regions where vertical velocity equals zero. The gray contour indicates the location of the TP (isoline of 3 km elevation). The red star shows the location of Lhasa (30°N, 91°E).

4. Conclusions

In this study, ozone vertical profiles from balloon-borne observations over Lhasa in five years (2010, 2013, 2016, 2018, and 2019) are combined with ERA-Interim reanalysis to examine the influences of Q1 on the vertical structure ozone at Lhasa. The mechanisms of the influences of Q1 on ozone —that is, the tropospheric ozone over Lhasa is anti-correlated with Q1 —are also explored.

The anti-correlation features between Q1 and the ozone mixing ratio over Lhasa are found in both seasonal means and daily values. As Q1 shows both interannual and subseasonal variations, seasonal mean and daily tropospheric ozone vertical structures over Lhasa also differ from year to year and day to day. The tropospheric ozone over Lhasa is relatively low when Q1 is relatively high. The correlations of lower tropospheric ozone over Lhasa and Q1 are about ? 0.5 and the peak value( ? 0.57) also occurs at nearly 500 hPa. The influence lasts for several days. The relation is due to the transport of low-ozone-concentration air to upper levels over Lhasa when Q1 is relatively high. Significant upward anomaly motions exist over the whole plateau when Q1 is relatively high.

Overall, this study reveals the potential impact of Q1 on ozone vertical structures over Lhasa. The underlying mechanisms have also been proposed. Note that this study is based on observations at only one site and the samples are limited. More observations combined with numerical simulation are needed to obtain a comprehensive understanding of how the thermal and dynamical factors influence ozone distributions over the TP.

Declaration of Competing Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

This research was supported by the second Tibetan Plateau Scientific Expedition and Research Program (STEP) [grant number 2019QZKK0604] and the National Natural Science Foundation of China[grant numbers 91837311 , 41705025 , and 41705021 ].