Chloromethanes in the North American Troposphere and Lower Stratosphere Over the Past Two Decades

,


Introduction
Chlorocarbons are emitted from various surface sources, and some have long enough atmospheric lifetimes to be dispersed throughout the troposphere.Through mixing and convective processes, a fraction of the emitted chlorocarbons enters the lower stratosphere (LS), where they contribute to the abundance of ozone-depleting chlorine (Cl) radicals.It is important to understand and quantify the amount of total organic-Cl entering the LS since the ozone layer is situated here and therefore the release of Cl-radicals potentially delays the recovery of the ozone layer (Hossaini et al., 2017;Say et al., 2020).In this work, we focus on chloromethanes, which account for a significant subset of the total organic chlorine burden in the atmosphere.Chloromethanes contribute 25%-30% of the organic Cl at the tropopause, (Figure S1 in Supporting Information S1).We examine measurements of carbon tetrachloride (CCl 4 ), methyl chloride (CH 3 Cl), chloroform (CHCl 3 ), and dichloromethane (CH 2 Cl 2 ).These four compounds exhibit a range of atmospheric lifetimes and include CCl 4 which has been subject to restrictions by the Montreal Protocol.
In 2010 the Montreal Protocol banned the emissive use and production of CCl 4 due to it having a high ozonedepleting potential (ODP).Despite this, the atmospheric abundance of CCl 4 has declined at a slower rate than expected (Hu et al., 2016;Sherry et al., 2018).The main sources of ongoing atmospheric sources of CCl 4 are from the permitted production and use of CCl 4 as a feedstock and unreported emissions from chlor-alkali plants.CCl 4 is used in the production of HFCs (hydrofluorocarbons) and HFOs (hydrofluoroolefins), which replace halocarbons with higher ozone-depleting potentials.Therefore, the global production of HFCs and HFOs is predicted to increase (Laube & Tegtmeier, 2022).Over the last decade, the production of CCl 4 has doubled for its allowed uses (Sherry et al., 2018).According to the WMO 2022 Ozone Assessment, CCl 4 is emitted from the surface at a rate of 44 ± 15 Gg yr 1 and accounts for 3% of total organic chlorine emissions (Laube & Tegtmeier, 2022).There is a well-documented gap between model-derived (top-down) and bottom-up (observations) emission estimates (Happell & Roche, 2003;Sherry et al., 2018;Suntharalingam et al., 2019) indicating that CCl 4 emissions are not entirely accounted for.Major sinks for CCl 4 (total atmospheric lifetime 29 years, (Laube & Tegtmeier, 2022)) include: stratospheric photolysis, soil (Happell & Roche, 2003), and oceanic uptake (Suntharalingam et al., 2019).Recent research suggests lower loss rates for CCl 4 , reducing the discrepancy between the different emission estimates (Valeri et al., 2017).However, after considering these findings, there is still a missing atmospheric source of CCl 4 (Sherry et al., 2018).If the emissions continue as they are now, then in the future CCl 4 is expected to have the largest impact on stratospheric ozone of all the ODS (Laube & Tegtmeier, 2022).
CH 3 Cl (atmospheric lifetime: 0.9 years) emissions (4.7 Tg yr 1 ) make a significant contribution to the lower stratospheric Cl budget (Laube & Tegtmeier, 2022;Umezawa et al., 2014).CH 3 Cl is predominantly emitted from tropical vegetation (Keppler et al., 2005;Yokouchi et al., 2007) therefore CH 3 Cl has a surface latitudinal gradient, which is largest at the tropics (Umezawa et al., 2014).This is a region with rapid vertical transport.For the troposphere, reaction with OH is the largest CH 3 Cl sink, leading to a CH 3 Cl seasonal cycle with a late-summer minimum, augmented by increased soil uptake due to higher soil temperatures at this time of year (Umezawa et al., 2014).Upper tropospheric CH 3 Cl mixing ratios are often higher than in the lower troposphere since the major CH 3 Cl sinks are not as efficient in UT, but they can be more variable than surface data due to increased uplifting of CH 3 Cl-depleted surface air in the Northern Hemisphere (Umezawa et al., 2014).Because of its predominantly natural source, CH 3 Cl is unregulated by the Montreal Protocol and Amendments.However, large uncertainties are associated with the fraction of CH 3 Cl that is not from biogenic or marine sources (Li et al., 2017).The gap between the CH 3 Cl measurements and budget is estimated to be as high as 20% (Engel et al., 2018).CH 3 Cl is used frequently as a chemical feedstock, and it can be emitted from biomass burning, coal combustion, and biofuel use.There are correlations between CH 3 Cl and pollution tracers from China (Li et al., 2017) and Italy (Cristofanelli et al., 2020).
CH 2 Cl 2 and CHCl 3 are classified as very short-lived substances (VSLS), which are defined as ozone-depleting halogen-containing gases with tropospheric lifetimes <6 months.Their short lifetimes lead to variable tropospheric concentrations that depend on emission source locations and strength.The source of CH 2 Cl 2 is predominantly anthropogenic and the biogenic emissions are not well understood (Say et al., 2020).CH 2 Cl 2 is used in the production of HFC-32 (Chipperfield et al., 2020), as a pharmaceutical processing solvent, and as a foamblowing agent (Feng et al., 2018).Previous studies and ongoing measurements have found that CH 2 Cl 2 emissions are increasing, leading to higher concentrations in the troposphere (An et al., 2021;Elvidge et al., 2015;Hossaini et al., 2017).Models indicate that 80% of the Cl from surface CH 2 Cl 2 emissions reach the stratosphere (Hossaini et al., 2019).Previously, ∼90% of atmospheric CHCl 3 was estimated to be from biogenic sources; however, recent evaluations of CHCl 3 estimate that up to 50% is anthropogenic (Fang et al., 2019;Say et al., 2020).96% of the anthropogenic CHCl 3 emitted in 2016 was due to its use as a feedstock in the production of HCFC-22 (Chipperfield et al., 2020).Other anthropogenic sources include water chlorination and the production of fluoropolymers.NOAA surface data indicates CHCl 3 mole fractions increased in the Northern Hemisphere between 2010 and 2015 (An et al., 2023;Fang et al., 2019), before declining after 2018 (An et al., 2023), Figure S5 in Supporting Information S1.Atmospheric CHCl 3 mixing ratios are lower in the Southern Hemisphere, and are largest in the northern mid-to high latitudes (McCulloch, 2003).Recent studies indicate increasing anthropogenic CH 2 Cl 2 (Fang et al., 2019;Say et al., 2020) and CHCl 3 (An et al., 2023) emissions from eastern Asia.
This study combines a large set of measurements to summarize and describe atmospheric chloromethanes over two decades over North America.By analyzing all the data together we can better assess vertical gradients and trends from the ground to the lower stratosphere.These chloromethane species are both potent ozone-depleting substances in the lower stratosphere and are greenhouse gases, some with high warming potential.Therefore, knowing the vertical distribution is crucial for estimating the impact chloromethanes have upon radiative forcing, stratospheric ozone, and the ozone layer recovery.In addition, aircraft observations contribute to evaluating the effectiveness of the Montreal Protocol regulations, ensuring compliance with current regulations and policy decisions regarding unregulated species.As the abundance of longer-lived chloro-compounds governed by the Montreal Protocol continues to decrease, these four chloromethanes, with no or limited restrictions, will increasingly contribute to the Cl-abundance in the LS.Efforts to continue to reduce stratospheric Cl will need to focus on these unregulated chloromethanes (Villamayor et al., 2023).Airborne measurements are useful for quantifying compound mixing ratios after their distribution throughout the troposphere and LS.This is advantageous for reflecting longer-term trends over larger spatial scales.Variability in distributions and trends derived from the 22-year of airborne measurements can be used to evaluate sources and sinks of the chloromethanes, which are essential for the development of efficient mitigation strategies and for climate modeling improvements and validation.North America is a region of particular interest since the North American Monsoon Anticyclone is an efficient pathway for transporting chloromethanes to the LS.The data in this study can therefore provide information on the mixing ratios, ODP, and trends of chloromethanes that can be used to update emission inventories and improve estimates of ozone hole recovery.

Methods
Data from multiple aircraft campaigns that sampled the troposphere and LS over the North American continent were collated to analyze the evolution of atmospheric mole fractions over time and altitude (Figure S2 in Supporting Information S1).Details of the data collection such as, aircraft campaign, URLs, and measurement instrumentation can be found in Table S1 in Supporting Information S1.After ensuring uniformity between units the data was averaged over 5-km altitude bins to capture vertical variations in atmospheric layers.To assess vertical profiles over time and to compare values with surface measurements, we calculated for both sets of data the mean over 5 years year intervals to minimize the impact of seasonal and geographic variability while allowing for statistically robust analyses of longer-term trends.Neither aircraft nor ground-site data were filtered to exclude data flagged due to mixing ratios being above a certain threshold above the atmospheric background.Whole air sampling refers to samples collected in stainless steel or glass canisters for later analysis.This involves separation of trace gases by gas chromatography (GC) with a detection technique: mass spectrometry (MS), flame ionization (FID), or electron-capture detection.University of California Irvine (Colman et al., 2001), University of Miami (Schauffler et al., 1998), NOAA iWAS (Lerner et al., 2017) and Programmable Flask Package (PFP) use Whole air sampling to collect samples for offline analysis.Samples can also be measured in situ using a fast online GC-MS, that is, TOGA (Apel et al., 2015).Surface measurements from long-term monitoring sites in North America were made from canisters collected at several locations in North America (https://gml.noaa.gov/dv/data/index.php?category=Halocompounds); CCl 4 , CH 2 Cl 2 and CH 3 Cl surface data were downloaded from the NOAA Halocarbon and other Trace Gas Species (HATS) network.CHCl 3 observations were from the surface PFP measurements (Vimont et al., 2022).Canisters were analyzed using GC-MS techniques on multiple instruments.Surface observations were also collected from the Trinidad Head AGAGE site for all four chloromethanes (Table S1 in Supporting Information S1).Sampling VOCs in the atmosphere is challenging due to a large, dynamic range of mixing ratios, some very low (ppt or lower) concentrations, contamination artifacts, and differences in sample collection and analytical procedures (Andrews et al., 2016;Hall et al., 2014).The data used here comes from different research groups and spans more than two decades, which makes it difficult to have a quantitative measure of uncertainty related to the measurements.Over the study period, the research groups involved have conducted various informal comparisons suggesting reasonable agreement for chloromethanes.No formal comparison has been done between all groups, though there have been several published studies on data comparability among several groups.In the IHALACE study (Hall et al., 2014), multiple laboratories (UCI-WAS and UM-WAS included) measured the same ambient and higher concentration samples contained in highpressure canisters.For CCl 4 , there was a standard deviation of 1.8 ppt (1.9% relative standard deviation) among the measurements.Results for CH 2 Cl 2 , CHCl 3 and CH 3 Cl had RSDs of 9%, 5% and 2.5%, respectively (Hall et al., 2014).In a field-based study, Andrews et al. (2016) compared three instruments, including UM-WAS and NCAR-TOGA, that measured VSLS species from ground-based and airborne platforms.The mean absolute percentage errors of CHCl 3 and CH 2 Cl 2 between 0 and 8 km were 3.7% and 5.1%, respectively, for flights within the same air mass (Andrews et al., 2016).Occasionally, uncertainty for chloromethanes in archived data is reported.For example, for FIREX-AQ, the UCI-WAS reported a precision of 2%, 5%, 10% and 2% for CCl 4 , CHCl 3 , CH 2 Cl 2 and CH 3 Cl respectively, with an accuracy of 10% for CCl 4 , CHCl 3 and CH 3 Cl, and 20% for CH 2 Cl 2 for samples taken from the NASA DC-8.The NCAR-TOGA-TOF was also deployed on the DC-8 during FIREX-AQ and reported overall uncertainties of 15% and 17% for CH 2 Cl 2 and CHCl 3 , respectively.Based on the reported uncertainties and prior informal comparisons, we suggest a conservative estimate of the relative uncertainties for chloromethanes reported here as CCl 4 (5%), CH 3 Cl (10%), CHCl 3 (15%), CH 2 Cl 2 (20%).

Carbon Tetrachloride
Aircraft observations in Figure 1a show that the tropospheric and lower stratospheric concentrations of CCl 4 decreased during 2000-2022 over the potential temperature range 250-500 K. CCl 4 is well-mixed throughout the troposphere and, for any given year, there is a vertically uniform concentration up to approximately 6 km (Figure 1a).This finding is consistent with a previous study that found a "near-zero vertical gradient" of CCl 4 between 0 and 6 km agl (Hu et al., 2016) from ongoing measurements at nearly a dozen profiling locations across the US.HIPPO, ATom, START08, and DCOTSS missions observed lower CCl 4 mixing ratios (<75 ppt) at 310-400 K which are associated with elevated ozone mixing ratios and are likely due to the aircraft sampling stratospheric air parcels in stratospheric intrusions or at higher latitudes over North America (Figure S3b in Supporting Information S1).The enhancements in CCl 4 mixing ratios (>100 ppt) around 300 K (Figure S3a in  Supporting Information S1) are potentially due to sampling air that originated from areas in the United States with previously reported higher CCl 4 emissions, for example, Colorado, Gulf Coast, Bay Area, California (Hu et al., 2016).During CalNex and SARP the aircraft sampled air over California and these samples led to the 2010-2014 average being higher than surface measurements.Above the tropopause, CCl 4 mixing ratios decreased with increasing height as stratospheric conditions lead to a larger photolytic sink of CCl 4 .There were only three aircraft campaigns where samples were collected above 19 km: AVE ( 2004), SEACR 4 S ( 2013), and DCOTSS (2021, 2022).The decreasing change in lower stratospheric CCl 4 mixing ratios is similar to that in the troposphere (Table S2 in Supporting Information S1).Ground-site measurements of CCl 4 were comparable to the airborne measurements in the troposphere, Figure 1a; for 2000-2004 the mean CCl 4 from ground measurements was 94.5 ± 1.3 ppt, and the mean mixing ratio for 0-5 km altitudes was 94.8 ± 1.1 ppt.For 2015-2019, both the mean surface measurement and the mean 0-5 km observation were ∼80.0 ppt (see Table S4 in Supporting Information S1).The mean mixing ratio for each period decreased for all 25 K potential temperature ranges (Table S2 in Supporting Information S1).In the tropopause region (375-400 K) the median CCl 4 mixing ratio decreased from 81 ppt (2005)(2006)(2007)(2008)(2009) to 78.4 ppt (2010-2014) to 69.9 ppt (2020-2022).The Northern Hemisphere surface observations also decreased, and the variability in the surface CCl 4 observations is relatively small, with no significant changes over time (Table S4 in Supporting Information S1).
For each 5 km altitude bin, CCl 4 had a negative trend (Figure S8a in Supporting Information S1, Table 1).
The trend was close to or more than 1 ppt yr 1 for each altitude level, with high Pearson correlation coefficients (r > 0.95 for all except 10-15 km range which included a mix of tropospheric and stratospheric air).NOAA and AGAGE ground measurements from 2019 to 2020 were used to calculate CCl 4 trends for the 2022 WMO Ozone Assessment (Laube & Tegtmeier, 2022).The Assessment reported CCl 4 trends of 1.32 ppt yr 1 (NOAA) and 1.01 ppt yr 1 (AGAGE).The trend calculated from the 0-5 km aircraft observations, was very similar, 0.99 ppt yr 1 , although not a direct comparison.

Methyl Chloride
CH 3 Cl mixing ratios are centered around 550 ppt in the troposphere (<14 km), with no clear temporal trend between 2000 and 2022 (Figure 1b and S4a in Supporting Information S1).Between 300 and 350 K the FIREX-AQ and SEAC 4 RS aircraft campaigns reported more variable and elevated (up to 14.3 ppb) mixing ratios of CH 3 Cl, Figure S4b in Supporting Information S1).The high and variable CH 3 Cl from these campaigns is associated with target sampling of biomass burning emissions which is a source of CH 3 Cl.Above 15 km (400 K) the CH 3 Cl mixing ratio decreases with increasing height (Figure 1b, Table S2 in Supporting Information S1) at approximately 24 ppt/10 K.In the stratosphere, CH 3 Cl is still lost via reaction with OH but also has additional sinks such as photolysis and reactions by O 1 D and Cl (Umezawa et al., 2014).The mean surface measurements are consistent for every 5-year period, with a range between 540 and 557 ppt.The mean mixing ratios from the aircraft data combined over 0-12 km are within 1 standard deviation of the mean surface observations (548.1 ± 36.5 ppt).Typically, CH 3 Cl exhibits a late summer minimum due to high photochemical oxidation rates (Cristofanelli et al., 2020;Umezawa et al., 2014), but in the aircraft measurements, elevated mixing ratios were observed during summer and autumn, primarily attributed to biomass burning emissions.The trends for CH 3 Cl are all positive, indicating increasing CH 3 Cl mixing ratios, though there is an indication that the increase in CH 3 Cl Note.There was insufficient data to reliably calculate the trend for data above 20 km.

Geophysical Research Letters
10.1029/2024GL108710 is slowing down with increasing altitude (Table 1).For example, the 0-5 km and 5-10 km altitude levels have trends of 2.40 ± 1.65 and 2.01 ± 1.31 ppt yr 1 whereas 10-15 km and 15-20 km have lower trends (0.36 and 0.85 ppt yr 1 , respectively (Figure S8b in Supporting Information S1).The 0-5 km trend (+2.40 ± 1.65 ppt yr 1 ) closely aligns with trends reported in the 2019-2020 WMO Ozone Assessment, as calculated by AGAGE (3.3 yr 1 ) and NOAA (3.0 yr 1 ) global surface data (Laube & Tegtmeier, 2022).Despite these consistent trends, annual mean fluctuations in mixing ratios exhibit variability, resulting in concentrations centered around 550 ppt.Notably, while these trends indicate an overall increase, they represent modest changes compared to the mean CH 3 Cl mixing ratios, with substantial fluctuations observed in ground data occurring on similar timescales to aircraft measurements.

Dichloromethane
Between 2000 and 2022, CH 2 Cl 2 mixing ratios have been increasing at the surface, within the free troposphere, and into the LS (Figure 1c, Figure S5a in Supporting Information S1) over North America.The aircraft data shows increasing CH 2 Cl 2 concentrations at all altitude ranges between 0 and 20 km (Figure S5a and Table S5 in Supporting Information S1).For data reported after 2020, the mean CH 2 Cl 2 mixing ratio was 84.3 ± 10 ppt.The mixing ratio increase in CH 2 Cl 2 is less pronounced in the LS than in the troposphere.At 325 K, the median CH 2 Cl 2 mixing ratio was 40.4 ppt for 2010-2015 and 73.5 ppt for samples collected after 2020, an increase of 33.1 ppt (Table S3 in Supporting Information S1).At 425 K however, the CH 2 Cl 2 increase over the same period was only 6.8 ppt, from 16.8 to 23.6 ppt.As altitude increases, mole fractions of CH 2 Cl 2 decrease in the atmosphere (Figure 1c, Tables S3 and S5 in Supporting Information S1).For data collected after 2020, the mean mixing ratios decreased from 51.0 ppt at 5-10 km, to 33.7 ppt at 10-15 km, to 10.1 ppt at 15-20 km.In the LS (>15 km) the CH 2 Cl 2 abundance decreases rapidly with increasing altitude, to approximately 10 ppt at >20 km in the most recent time interval.Consistent with the shorter lifetime and different emission sources of CH 2 Cl 2 , the surface observations are more variable than the longer-lived CCl 4 and CH 3 Cl.Field campaigns investigating deep convection (e.g., SEAC 4 RS, 2013 and DCOTSS, 2022) have higher than mean mixing ratios of CH 2 Cl 2 in the 10-15 km and 5-10 km ranges (Table S5 in Supporting Information S1).At the highest altitude range (>20 km, 500 K), the CH 2 Cl 2 mixing ratios are consistent with stratospheric background levels (<20 ppt).The 2022 WMO Ozone Assessment reported positive trends of 1.2 ppt yr 1 (AGAGE) and 1.3 ppt yr 1 (NOAA) for CH 2 Cl 2 between 2019 and 2020 (Laube & Tegtmeier, 2022).The Assessment indicates the growth in atmospheric CH 2 Cl 2 is slowing down; the peak CH 2 Cl 2 growth rate between 2012 and 2016 was 6.7 ppt yr 1 (NOAA) or 6.6 ppt yr 1 (AGAGE).For 0-5 km the aircraft observations between 2000 and 2022 indicate a positive trend of 2.40 ppt yr 1 and, between 0 and 20 km, CH 2 Cl 2 trends were all increasing (Figure S9a in Supporting Information S1), with the largest increase in the 5-10 km level (3.56 ± 0.68 ppt yr 1 ).

Summary
The aircraft observational data collected over North America was used to calculate trends of chloromethanes in the atmosphere.Our findings highlight the usefulness and need for continued monitoring and research at both the surface and using aircraft platforms.Our analysis confirmed a consistent decrease of atmospheric CCl 4 across all altitudes of the vertical profiles studied here.The average trend of 1.01 ppt yr 1 is similar to the trend reported by the WHO Ozone Assessment made using surface-level measurements.This is still a slower rate than anticipated from models and our current knowledge of CCl 4 sources and sinks.If the CCl 4 decline continues to slow it will slow the recovery of the ozone layer.CH 3 Cl had an observed positive average trend of 1.12 yr 1 , which is insignificant compared to the average abundance of 550 ppt.This reflects relatively stable levels throughout the troposphere and lower stratosphere.The CH 3 Cl surface mixing ratios are more variable (±50 ppt) than the CCl 4 (±5 ppt).This variability reflects the seasonal changes in photochemical destruction of CH 3 Cl and the spatial/ temporal variability of emissions from vegetation and fires (Cristofanelli et al., 2020).The surface data in Figure 1b is data collected year-round so the spread of observations is also augmented relative to CCl 4 because of seasonal variability of photochemical destruction of CH 3 Cl and emissions from vegetation and fires.Most of the airborne observations were collected in summer and spring, with only a few missions occurring in winter.CH 2 Cl 2 exhibited an increase throughout the troposphere, with an average tropospheric trend of 2.41 ppt yr 1 over the 22 years studied here.The stratospheric measurements indicated increasing levels of CH 2 Cl 2 , albeit at a slower rate than the troposphere.The large increasing emissions of CH 2 Cl 2 highlights the urgency of investigating its sources and potentially mitigating atmospheric emissions.The observations indicated a rapid decrease of CHCl 3 mixing ratios from the surface to higher altitudes.Both the surface level and lower atmospheric observations indicated a growth in CHCl 3 emissions between 2012 and 2018, with subsequent plateauing in more recent years.
It is of increasing importance to continue to monitor the state of the atmosphere concerning ozone-depleting substances, and their impact on the Earth's radiative budget.The growth of CH 2 Cl 2 and potentially CHCl 3 abundance throughout the lower atmosphere is concerning if their levels in the atmosphere continue to increase.
An increase in CCl 4 and CH 2 Cl 2 emissions is anticipated due the increase demand for HFOs and HFCs, so their atmospheric mixing ratios and trends in the lower stratosphere must continue to be closely monitored.The data is used to examine the impact of international policy decisions and will be important for understanding future climate and environmental impacts of chloromethanes in the atmosphere.

Figure 1 .
Figure 1.Altitude versus.the mean and standard deviation of (a) CCl 4 , (b) CH 3 Cl, (c) CH 2 Cl 2 and CHCl 3 .The data was binned into 5-year periods and 5 km altitude bins.Observations from the North American ground sites in the NOAA HATS and AGAGE surface networks (stars on blue dashed line) are shown for comparison.The bar plot underneath shows the median, interquartile range (box-edges), and the rest of the distribution (whiskers), including outliers (diamonds) of the surface measurements.

Table 1
The Atmospheric Mole Fraction Trends and Standard Errors of the Trends for 2000-2022, Calculated Using Linear Regression Over the Annual Mean CCl 4 , CH 3 Cl, CH 2 Cl 2 and CHCl 3 Mixing Ratios for Each 5 km Bin