Extreme Change Events of Stratospheric HCl and N₂O in the Mid-Latitude Region of the Northern Hemisphere

Abstract

Hydrogen chloride (HCl) is the main reservoir species of chlorine and chemical decomposition of nitrous oxide (N₂O) is the primary source of NO_x (=NO + NO₂) in the stratosphere. Changes in stratospheric HCl and N₂O play a critical role in modulating variations in stratospheric ozone. Thus, long-term trends in stratospheric HCl and N₂O have been investigated in many studies, whereas short-term changes have not received enough attention. Here, using satellite observations and a chemical transport model, we found that two extreme change events for HCl and N₂O in the Northern Hemisphere mid-latitude middle and lower stratosphere have occurred over past decades, which are characterized by a sharp increase in HCl and a decrease in N₂O over several months; for example, HCl increased (and N₂O decreased) by 0.135 ppbv (−33.352 ppbv) in 1987/1988 and by 0.196 ppbv (−28.553 ppbv) in 2010/2011. Further analysis shows that the extreme change events of stratospheric HCl and N₂O in these two periods are closely related to anomalous residual circulation caused by the joint effects of the strong easterly phase of the semi-annual oscillation and the strong polar vortex.

1. Introduction

Stratospheric ozone absorbs solar ultraviolet radiation [1,2,3], protecting life on Earth’s surface from biological damage, and via radiative heating, it influences stratospheric temperature and circulation and thus affects global climate change [4,5,6,7,8,9,10,11,12,13,14,15]. Stratospheric ozone levels began dropping from the late 1970s [16], which was mainly a result of human use of compounds called chlorofluorocarbons (CFCs; [17]). Until 1987, with the implementation of the Montreal Protocol, loading of CFCs in the stratosphere has been reducing, and consequently ozone is projected to recover to pre-1980 levels by the middle of the 21st century [17,18,19,20]. However, due to the unpredictable increase or redistribution of substances affecting ozone in the future, such as stratospheric hydrogen chloride (HCl) and nitrous oxide (N₂O), there is still great uncertainty around the recovery of the ozone layer [21,22,23].

HCl, as the largest reservoir of chlorine in the stratosphere, is important to the gas phase chemical reactions of ozone depletion. Consistent with the decline trend in surface chlorine source gases, the negative trend of stratospheric HCl was observed after 1997 [24,25,26]. On the basis of satellite observation data, Jones et al. [25] found that HCl in the mid-latitude middle to upper stratosphere showed a decline of −5%/decade from 1997 to 2008. Subsequently, Carpenter and Reimann [26] summarized the results associated with the recent trend of HCl and concluded a mean decline of 0.6 ± 0.1%/year of HCl in the middle and upper stratosphere for the years 1997–2012. However, in recent years, a significant increasing trend in HCl was observed in the lower stratosphere of the Northern Hemisphere (NH), which is contradictory with continuous monotonic reduction of near-surface source gases and has drawn much attention [27,28,29,30]. Mahieu et al. [28] first proposed that there is an increasing trend of HCl in the NH since 2005, which resulted from a weakening of atmospheric circulation.

As the primary source of stratospheric nitrogen oxides (NO_x = NO + NO₂), which are also known to catalyze ozone depletion [31,32,33], N₂O is expected to remain as the largest emission substance in the stratosphere throughout the 21st century, making it the single most important ozone-depleting emission substance [21,22]. Crutzen and Ehhalt [34] first found the increase in the use of fixed nitrogen as a fertilizer may lead to the reduction of the Earth’s ozone layer by several percent at the beginning of the 21st century. Under the IPCC A2 scenario, Portmann and Solomon [35] pointed out that an increase in N₂O may result in a 2–4% reduction of the total ozone column. Using a coupled chemistry–climate model, Wang et al. [36] indicated that a 50%/100% increase in N₂O between 2001 and 2050 will lead to a 6%/10% decrease in ozone mixing ratios in the middle stratosphere at around 10 hPa.

There are several studies reporting that the ozone in the NH has a decreasing trend after the 2000s, e.g., the negative trend of ozone observed in the midlatitudes lower stratosphere [37,38,39,40] and Arctic lower stratosphere [41]. This may be closely related with an increasing trend of HCl and N₂O in the NH stratosphere after the 2000s mentioned above. Therefore, trends in stratospheric HCl and N₂O are still of great concern, and thus long-term trends in stratospheric HCl and N₂O have received much attention [24,25,26,42,43,44,45,46]. In contrast, short-term changes in stratospheric HCl and N₂O in the mid-latitude region of NH have not received as much attention relative to their trends. However, if extreme variations in stratospheric HCl and N₂O in the mid-latitude region of NH do occur, these could create influences on ozone that cannot be ignored. This study defines the extreme change events for stratospheric HCl and N₂O in the mid-latitude region of NH that have occurred in the past few decades and explains the relevant physical mechanisms. The structure of the paper is as follows: Section 2 provides a brief description of the data and methods used; Section 3 presents the analysis of the extreme change events and the relevant physical mechanisms; and conclusions and a discussion are provided in Section 4.

2. Data and Methods

The primary stratospheric HCl and N₂O dataset used was from version 5 of the Aura Microwave Limb Sounder (MLS) level 2 data (https://acdisc.gesdisc.eosdis.nasa.gov/data/Aura_MLS_Level2/, accessed on 1 January 2021), covering the period from 2005 to 2020. MLS measures atmospheric chemical species daily, with a global coverage from 82°N to 82°S and a vertical resolution of ≈3 km for HCl and ≈5–8 km for N₂O [47]. We used the gridded MLS data at a resolution of 4° longitude × 4° latitude; the quality screening rules for this dataset have been previously described by Livesey et al. [47]. For the three-dimensional (3D) winds, temperature, and potential vorticity fields, ERA5 re-analysis datasets from the European Centre for Medium-Range Weather Forecasts (ECMWF, https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5, accessed on 1 January 2021) for 1979–2020 were used, with a horizontal resolution of 1° longitude × 1° latitude.

To obtain the long-term changes in stratospheric HCl and N₂O, we used a three-dimensional off-line chemical transport model (TOMCAT/SLIMCAT) [48]. This model uses horizontal winds and temperatures from the fifth major global re-analysis produced by ERA5 [49] and uses a hybrid sigma–pressure vertical coordinate with detailed tropospheric and stratospheric chemistry [50]. Vertical advection is calculated from the divergence of the horizontal mass flux [48], and chemical tracers are advected, conserving second-order moments [51]. The model has been extensively evaluated against various ozone satellite and sounding datasets, providing a good representation of stratospheric chemistry (see, for example, Chipperfield [48]; Feng et al. [52]; Dhomse et al. [53]). The TOMCAT/SLIMCAT model used for long-term simulation in this study performs with a horizontal resolution of about 2.8° latitude × 2.8° longitude and with 32 levels, from the surface to 65 km altitude.

The latitudinal average zonal wind in the tropical region (10°N–10°S) at 50 hPa is defined as the quasi-biennial oscillation (QBO) index [54]. The semi-annual oscillation (SAO) index is defined as the zonal-mean wind averaged between 10°N–10°S and 1–5 hPa [55]. The polar vortex (PV) intensity is defined using the 10 hPa zonal mean-zonal wind at 60°N [56]. In addition, the transformed Eulerian mean residual circulation (${\overline{v}}^{*},{\overline{w}}^{*}$) was used to diagnose the residual circulation [57]. In pressure coordinates, these are defined by ${\overline{v}}^{*}=\overline{v}-\frac{1}{{\rho }_0}{\left(\frac{{\rho }_0\overline{v^{\prime }{\theta }^{\prime }}}{\overline{{\theta }_z}}\right)}_z$ and ${\overline{w}}^{*}=\overline{w}+\frac{1}{acos\phi }{\left(\frac{cos\phi \overline{v^{\prime }{\theta }^{\prime }}}{\overline{{\theta }_z}}\right)}_{\phi }$. Here, overbars denote zonal means, and primes are deviations from the zonal mean of a given variable; z is log-pressure height and $\phi$ is the latitude; v and w are the meridional and vertical velocity, respectively; ${\rho }_0$ is density, which is defined as ${\rho }_0={\rho }_s\exp \left(-z/H\right)$ with H = 7.0 km; θ is the potential temperature; and a is the Earth’s radius.

3. Results

Figure 1a,b shows the time series for HCl and N₂O anomalies in the NH mid-latitude middle and lower stratosphere from 2005 to 2020 based on MLS data (red line), and from 1979 to 2020 based on the TOMCAT/SLIMCAT simulation (black line). The TOMCAT/SLIMCAT model was employed to reproduce the long-term changes in HCl and N₂O in the stratosphere. It is evident that the changes in stratospheric HCl and N₂O from 2005 to 2020 derived from the TOMCAT/SLIMCAT simulation (black line) are in overall agreement with those from the MLS satellite measurements (red line), providing confidence in the ability of the TOMCAT/SLIMCAT model to reproduce long-term changes in HCl and N₂O in the stratosphere. Here, an extreme change event for stratospheric HCl (N₂O) is defined by two indicators (Figure 1c,d): first, the increase (decrease) magnitude of one event of HCl (N₂O), which is calculated using the maximum value of HCl (minimum value of N₂O) in a year minus the minimum (maximum) value in the past 12 months; second, the increase (decrease) gradient in HCl (N₂O) exceeded the 95th percentile threshold for the whole period, following extreme precipitation events [58]. Note that since the cycle of HCl and N₂O was about 20 months (Figure 2a,b), the difference of extreme value within 12 months was selected to calculate the magnitude changes, and the result was not sensitive to the selected number of months (10 months or 24 months, Figure 2c–f). According to this definition, we found two extreme change events occurred in 1988 and 2011 (Figure 1), with a sharp increase (decrease) of HCl (N₂O) by 0.135 ppbv (−33.352 ppbv) in 1988 and by 0.196 ppbv (−28.553) ppbv in 2011, respectively. Evidently, the anomalies in HCl (N₂O) in 1988 and 2011 are significantly greater than those in other years (Figure 1a,b).

Further analysis showed that both the extreme change events, in 1987/1988 and in 2010/2011, occurred from November to February (Figure 3a,b); therefore, these four months were the main focus of our investigation, as described in the following paragraphs. Compared to N₂O in the troposphere, N₂O in the stratosphere has a much shorter lifetime, but such extreme changes in N₂O within a few months (Figure 1b) should mainly be caused by dynamic transport in the stratosphere [59,60,61]. The residual circulation in the stratosphere transports trace gases from the tropics to polar regions and thus influences the distribution of trace gases in the stratosphere [62,63,64,65,66,67]. Dynamic transport associated with residual circulation was considered in order to examine the cause of these extreme change events for stratospheric HCl and N₂O. There are evident downwelling anomalies of the residual circulation in the mid-latitude upper stratosphere from December to February (Figure 3c–f), and upwelling anomalies in the high-latitude middle and lower stratosphere from January to February (Figure 3e,f), in 1987/1988 and 2010/2011. The downwelling anomalies in 1987/1988 and 2010/2011 were greater than those in other years (Figure 3g), and the upwelling anomalies in 1987/1988 and 2010/2011 were also larger than those in most other years (Figure 3h). Note that the stratospheric residual circulation can be divided into the deep branch in the middle to upper stratosphere and the shallow branch in the middle to lower stratosphere [65,68]. Then, downwelling anomalies of the residual circulation in the mid-latitude upper stratosphere imply an enhancement of the deep branch of the residual circulation in mid-latitude, and upwelling anomalies in the high-latitude middle and lower stratosphere imply a slowdown of the shallow branch of the residual circulation in high latitude compared with the climatology of the residual circulation. Considering that the vertical and meridional gradients of HCl and N₂O distribution are opposite to each other, on the one hand, strengthening of the deep branch of the residual circulation in the mid-latitudes leads to more HCl-rich and N₂O-poor air being transported from the upper stratosphere to the middle and lower stratosphere, resulting in the increase in HCl and decrease in N₂O observed in 1987/1988 and 2010/2011. On the other hand, the weakened shallow branch of the residual circulation in high latitudes weakened the exchange of air between mid-latitudes and high-latitudes. Both of these processes contributed to the sharp increase in HCl and decrease in N₂O in the mid-latitude middle and lower stratosphere in 1987/1988 and 2010/2011.

To further verify the effect of the residual circulation on the extreme change events of stratospheric HCl and N₂O, Figure 4 and Figure 5 show the month-to-month evolution of HCl + residual circulation anomalies and N₂O + residual circulation anomalies, respectively, from November 2010 to April 2011. Since the changes in stratospheric HCl and N₂O and the relevant residual circulation in 1987/1988 were similar to those in 2010/2011 (Figure 1 and Figure 3), and observations were available for 2010/2011, the case for 2010/2011 is mainly examined here. From November 2010 to February 2011 (Figure 4a–d and Figure 5a–d), there was an evident clockwise cell in the low–mid-latitude middle and upper stratosphere (arrows), accompanied by the strong sinking motion in the mid-latitude upper stratosphere. This implies an enhancement of the deep branch of the residual circulation in the mid-latitude upper stratosphere. The strong downwelling branch leads to HCl-rich and N₂O-poor air being transported from the upper stratosphere to the lower and middle stratosphere, corresponding to an increase in HCl (contours in Figure 4) and decrease in N₂O (contours in Figure 5) just below the position of the strengthening deep branch of the residual circulation. From January 2011 to March 2011 (Figure 4c–e and Figure 5c–e), upwelling anomalies of the residual circulation were found in the high-latitude middle and lower stratosphere, implying a weakened shallow branch of the residual circulation (in Figure 4c–e and Figure 5c–e). Weakened residual circulation slowed down the exchange of air between low–mid latitudes and high latitudes, leading to an accumulation of HCl-rich and N₂O-poor air in the low–mid latitudes (contours in Figure 4c–e and Figure 5c–e). In addition, in the high-latitude middle and lower stratosphere, weakened residual circulation was favorable for chemical decomposition of HCl and N₂O, which is consistent with the evident decrease in HCl and N₂O in high latitudes of the middle and lower stratosphere from January 2011 to March 2011 (contours Figure 4c–e and Figure 5c–e). By April 2011, upwelling anomalies of the residual circulation were small (arrows in Figure 4f and Figure 5f), speeding up the exchange of air between low–mid latitudes and high latitudes. This explains why the center of positive HCl and negative N₂O anomalies moved to high latitudes of the lower stratosphere (contours in Figure 4f and Figure 5f). The case for 1987/1988 (not shown) agrees with that for 2010/2011 (Figure 4 and Figure 5).

The reasons for the strengthening deep branch of the residual circulation in mid-latitudes, as well as the weakening shallow branch of the residual circulation in high latitudes in 1987/1988 and 2010/2011, require further investigation. The tropical QBO, the SAO, and the PV are the dominant factors influencing short-term changes of circulation in the upper stratosphere and high latitudes [69,70,71,72]. Figure 6 shows QBO, SAO, and PV indices from 1980 to 2020 (see Section 2 for definitions of QBO, SAO, and PV indices). There were strong negative SAO indices and positive PV indices in 1987/1988 and 2010/2011, but the QBO was not in a strong phase. This indicates the simultaneous occurrence of a strong easterly phase of the SAO (ESAO) and a strong polar vortex (SPV). Under the ESAO, clockwise circulation was observed in the NH upper stratosphere (Figure 3c–f), which is consistent with previous studies [55,73]. The circulation induced intense downward transport in the mid-latitude upper stratosphere, coinciding with the location of an enhanced deep branch of the residual circulation. In addition, the SPV corresponded to the weakened wave activity in the NH high-latitude lower stratosphere, which was further consistent with a weakened shallow branch of the residual circulation in NH high latitudes.

The main characteristic of the meridional circulation associated with the SAO is the existence of tropical areas of convergence and divergence, which are produced by a negative and positive anomaly in the meridional wind in the NH tropical zone [55,74]. To confirm the effect of the SAO on the strengthening deep branch of the residual circulation, month-to-month evolution of the meridional wind anomalies from November 2010 to April 2011 are shown in Figure 7. From November 2010 to February 2011, significant positive meridional wind anomalies were found in the NH tropical upper stratosphere, and negative meridional wind anomalies were found in the NH tropical lower stratosphere (Figure 7a–d), corresponding to the convergence and divergence zones, respectively. The existence of these convergence and divergence zones corresponded to the rising and sinking motions. The rising motion over the tropical wind shear zones and sinking motions over the subtropics and mid-latitudes, added to the meridional divergence and convergence motions, formed a clockwise circulation cell (vectors in Figure 4a–d and Figure 5a–d). Until March 2011, the positive meridional wind anomalies in the tropical upper stratosphere became weak; thus, the corresponding clockwise circulation cell disappeared (Figure 4e, Figure 5e and Figure 7e).

Figure 8 further shows the month-to-month evolution of the zonal-mean wind anomalies from November 2010 to April 2011, in order to confirm the effect of a SPV on the weakened shallow branch of the residual circulation. Positive zonal-mean wind anomalies from January 2010 to March 2011 imply an anomalous SPV (Figure 8c–e), corresponding to the weakened residual circulation (Figure 4c–e). In April 2011, the negative zonal-mean wind anomalies in the high latitudes imply the collapse of the polar vortex (Figure 8f), and the corresponding weakened residual circulation disappeared (Figure 4f).

The above analysis points out the effect of significant anomalous residual circulation on the extreme change events for stratospheric HCl and N₂O in 1987/1988 and 2010/2011, and furthermore indicates that the anomalous residual circulation resulted from the joint effects of ESAO and a SPV.

Table 1. Simultaneous occurrence of the easterly phase of the SAO (ESAO) and the strong polar vortex (SPV) from 1979 to 2020.

Year	ESAO	SPV	Year	ESAO	SPV
1979–1980			2000–2001
1980–1981			2001–2002
1981–1982			2002–2003
1982–1983			2003–2004
1983–1984			2004–2005
1984–1985			2005–2006
1985–1986			2006–2007
1986–1987			2007–2008
1987–1988	✓	✓	2008–2009
1988–1989			2009–2010
1989–1990			2010–2011	✓	✓
1990–1991			2011–2012
1991–1992			2012–2013
1992–1993			2013–2014
1993–1994			2014–2015
1994–1995			2015–2016
1995–1996			2016–2017
1996–1997			2017–2018
1997–1998			2018–2019
1998–1999			2019–2020
1999–2000

SAO index lower than a negative standard deviation and PV index larger than a positive standard deviation were selected as the ESAO and SPV events, respectively.

shows the simultaneous occurrence of ESAO and SPV from 1979 to 2020 according to the data in Figure 6. It can be seen that the ESAO and SPV only occurred simultaneously in 1987/1988 and 2010/2011. This may explain why the extreme change events for stratospheric HCl and N₂O only occurred in 1987/1988 and 2010/2011.

4. Conclusions and Discussion

In this paper, we investigated extreme change events of stratospheric HCl and N₂O concentrations in the mid-latitude region of the NH middle and lower stratosphere by using the latest satellite observations and a chemical transport model. From 1979 to 2020, there were two extreme change events of stratospheric HCl and N₂O in the NH middle and lower stratosphere, one in 1987/1988 and one in 2010/2011, which were characterized by a sharp increase in HCl and decrease in N₂O over several months. These extreme change events were closely related to significant anomalous residual circulation; that is, a stronger deep branch of the residual circulation in the mid-latitude upper stratosphere brought about more HCl-rich and N₂O-poor air from the upper stratosphere down to the lower stratosphere; meanwhile, a slowdown of the shallow branch of the residual circulation in the high-latitude middle and lower stratosphere weakened the transport of air between mid-latitudes and high latitudes. Both of these processes contributed to the sharp increase in HCl and a decrease in N₂O recorded in the mid-latitude lower stratosphere in 1987/1988 and in 2010/2011. Further analysis suggests that during the period from 1979 to 2020, strong ESAO and SPV occurred simultaneously in 1987/1988 and 2010/2011 and were responsible for the significant anomalous residual circulation.

The question arises here as to whether the westerly phase of the SAO (WSAO) and a weakened polar vortex (WPV) would induce the opposite effect on stratospheric HCl and N₂O changes. The simultaneous occurrence of the WSAO and WPV from 1979 to 2020 is also presented in

Table 2. Simultaneous occurrence of the westerly phase of the SAO (WSAO) and weakened polar vortex (WPV) in winter from 1979 to 2020.

Year	WSAO	WPV	Year	WSAO	WPV
1979–1980			2000–2001
1980–1981			2001–2002
1981–1982			2002–2003
1982–1983			2003–2004
1983–1984			2004–2005
1984–1985			2005–2006
1985–1986			2006–2007
1986–1987	✓	✓	2007–2008
1987–1988			2008–2009
1988–1989			2009–2010
1989–1990			2010–2011
1990–1991			2011–2012
1991–1992			2012–2013
1992–1993			2013–2014
1993–1994			2014–2015
1994–1995			2015–2016
1995–1996			2016–2017
1996–1997			2017–2018
1997–1998			2018–2019
1998–1999			2019–2020
1999–2000

SAO index larger than a positive standard deviation and NAM index lower than a negative standard deviation were selected as the WSAO and WPV events, respectively.

. It was found that only in 1986/1987 did the WSAO and WPV occur simultaneously, corresponding to a decrease in HCl and an increase in N₂O (Figure 9a), but the magnitude of the change was smaller than those observed in 1987/1988 and 2010/2011 (Figure 3a,b) and not evident from the data for the variation in HCl and N₂O (Figure 1). Figure 9b,c further shows the anomalies in the residual circulation in 1986/1987. Under the influence of the WSAO and WPV, anti-clockwise circulation in the low–mid-latitude upper stratosphere and downwelling mass flux anomalies in the high-latitude middle and lower stratosphere were evident (Figure 9b), which is opposite from what was observed in 1987/1988 and 2010/2011 (Figure 3c–f). The strength of the ascending motion of the anti-clockwise circulation and downwelling mass flux anomalies in 1986/1987 was significantly smaller than that in 1987/1988 and 2010/2011 (Figure 3c–f and Figure 9c), and this may explain why the changes in stratospheric HCl and N₂O caused by the joint effects of the WSAO and WPV (1986/1987) were not evident.

In addition, the extreme change events above mainly represent the extreme decrease in HCl and increase in N₂O, and then whether there were reversed extreme HCl and N₂O events from 1979 to 2020. Figure 10 shows the variations of the HCl (N₂O) gradient that was calculated using the minimum value of HCl (maximum value of N₂O) in a year minus the maximum (minimum) value in the past 12 months. According to the definition of an extreme change event, extreme decease events of HCl were found in 2010 and 2019, and extreme increase events of N₂O were found in 2018 and 2020. Unlike in 1988 and 2011, reversed extreme change events of HCl and N₂O did not occur at the same time. This may be related to the climatology distribution of HCl and N₂O or to the changes of the residual circulation having different effects on the chemical reactions of HCl and N₂O.

There are still issues worthy of further study with regard to changes in HCl and N₂O. For example, how much impact would these extreme change events for HCl and N₂O have on ozone? Moreover, will there be more and more extreme change events occurring in the future?