Jump in Tropospheric Methane Concentrations in 2020–2021 and Slowdown in 2022–2024: New Hypotheses on Causation

Abstract

Earth’s atmospheric methane (CH₄) concentration has risen more than 162% since pre-industrial levels in the mid-18th century, and about 30% of the rise in global temperatures since the pre-industrial era is due to CH₄ The build-up of methane in the atmosphere in 2020–2022 was the largest since systematic measurements started in 1983, more than double the average yearly growth rate measured over the previous 17 years (15.2 ppb yr⁻¹ vs. 5.71 ppb yr⁻¹, respectively). During 2020, with a growth rate of 14.81 ppb yr⁻¹, the level of atmospheric CH₄ broke the previous record (which was set in 1991), and it was broken again immediately the following year, with an increase of 17.64 ppb yr⁻¹ in 2021. For 2022, the final estimate is 13.25 ppb yr⁻¹, the fourth largest annual growth rate. The most recent explanations for this surge in tropospheric CH₄ include increased emissions from tropical wetlands, more floods, and increased temperatures. For 2020 and part of 2021, a reduction in the oxidative capacity of the atmosphere due to COVID-19 lockdowns was also proposed. Our main hypothesis is that this CH₄ surge in 2020–2021 may also be caused by reduced sulfate emissions, which have been shown to decrease methanotrophy and increase methanogenesis rates in wetlands. Then, for the CH₄ slowdown in 2022–2024, our hypotheses are that the emissions from wetlands remained high, but that there was an even higher increase in the oxidative capacity of the atmosphere due to multiple other parameters that are detailed in this article. This perspective review paper is mainly qualitative; it demonstrates that coupled climate–chemistry models will also need to integrate biochemistry, as the evolution of the atmospheric composition is multifactorial and non-linear.

1. Introduction

Since 2007, atmospheric methane concentrations have been continuously increasing, but in 2020, 2021, and 2022, there was a surge, which is still not well explained, although the principal causes have been hypothesized.

1.1. The Year 2020

The atmospheric CH₄ surge in 2020 has been explained by several teams [1,2,3,4]. This growth could be mainly due to increased biogenic CH₄ emissions from wetlands (50%, 66%, or 82%, depending on the studies referenced), in particular from tropical ones, but also from Alaska and Canada. These studies posit the increase in wetland CH₄ emissions is due to an increase in precipitation/inundations, larger wetland areas, higher water tables, and the temperature sensibility of methanogens to a warmer climate.

A smaller part of the CH₄ rise could be due to a decrease in the oxidative capacity of the troposphere (14 to 50%), mainly due to less hydroxyl radical (°OH) generation due to less nitrogen oxide pollution (NOx = NO + NO₂) during the COVID-19 lockdowns [1,2].

The principal natural sink of CH₄ in the troposphere is °OH, also called the detergent of the atmosphere [5], which oxidizes more than 90% of tropospheric methane (tCH₄). Soils and their methanotrophs are the second sink of methane, oxidizing about 5.0 to 6.3% [6], and atmospheric chlorine atoms are the third one, oxidizing about 2.5% [7,8].

The COVID-19 lockdowns significantly reduced the global NOx emissions during 2020 by the aviation industry, which in turn globally reduced tropospheric ozone (O₃) generation [9,10] and, therefore, the °OH sink of CH₄ (even if locally more O₃ was experienced in some cities [11,12,13,14,15,16]).

1.2. 2021 and 2022

In 2021, the yearly growth rate of tCH₄ set a new record as it was higher by more than 2.8 ppb than that in 2020, even though there were almost no COVID-19 lockdowns and likely more NOx emissions, as primary energy consumption and electricity production increased above the 2019 levels and the global economic growth was about 6% (double that in 2022 and 2023). The lower NOx emissions in 2020, which could explain the abnormal increase in CH₄, did not occur in 2021, even though NOx pollution was decreasing globally on the whole.

The reasons for the tCH₄ increases in 2021 and 2022 are not yet fully understood and have so far received little investigation [17], as CH₄ has many sources [18], including anthropogenic sources, such as agriculture (livestock and rice paddies) and oil and gas fields [19], and biogenic sources [20], such as wetlands and other water bodies, aquatic ecosystems [21] (hydroelectric reservoirs, lakes, and ponds), and thawing permafrost peatlands [22],, where the tCH₄ increase is driven by human-made global warming [23].

The model used by one team suggests that the majority of this growth was due to increased CH₄ emissions, in particular from tropical wetlands, and about 10% was due to reduced levels of °OH [11]. Recent models also attribute this surge in 2020–2022 to tropical wetlands [24,25,26].

A more recent study attributes the CH₄ emission surge in 2020–2022 mostly to wetlands and rice paddies from Equatorial Asia (42% of the surge) and Africa (30%) as a result of inundation associated with La Niña conditions, with the decreasing tropospheric °OH sink contributing the remaining 28% [25].

1.3. Changes in Sinks but Also in Sources

As the principal driver of the changes in tCH₄ growth in 2020, 2021, and 2022 is the increased CH₄ emissions from wetlands [1,2,3,11,25,26,27], the reduction in principal tropospheric °OH sinks is a less important factor [25]. Therefore, the recent changes in the tCH₄ concentration are more driven by sources than by sinks, so why did the CH₄ sources increase?

Daily precipitation above 10 mm accelerates CH₄ production, while much higher daily precipitation levels can slow it down [28]. The number of flood disasters in the 2020–2022 period [29] is comparable to the number in 2005–2007, when the growth rate of tCH₄ was close to zero. Indeed, the number of flood disasters does not provide information on their location, season, intensity, duration, area, and the availability of decomposable biomass (which usually would be more important in tropical forests than in recently inundated deserts [30] but is one indicator among many to consider).

Based on field measurements, Bartlett (1993) suggests an increase of five-fold in CH₄ flux for each 10 °C increase in temperature [31], while other studies suggest even more or more [32,33], but Fung (1991) suggests only a two-fold increase [34]. But 2024, 2023, and 2016 are so far the 3 hottest years on record [35,36], and, as seen in Figure 1, the increase in tCH₄ was about 8.39 ppb in 2023 and was estimated to be between 3 and 8 ppb in 2024 [36,37], nearly one half and one quarter of the levels during 2020–2022.

Therefore, from this perspective, we formulate a new hypothesis, helping to answer the following questions: Are there additional parameters to take into consideration, other than inundation, larger wetland areas, and a warming climate, that may explain why the sources of CH₄ increased during 2020–2021 and remained high in 2022? For 2022, 2023, and 2024, which reasons/parameters can explain the decline in the growth rate of CH₄ in the atmosphere?

In the next section, a new additional hypothesis based on a literature review is formulated to explain the high levels of CH₄ emissions in 2020–2022. Then, in Section 3, several parameters are reviewed to attempt to provide explanations for the 2022–2024 decline in the CH₄ growth rate. This is followed by a discussion section on future directions and research needs and then the conclusion.

2. The Background Role of Sulfur Pollution

Here we formulate the hypothesis that the last three years of record growth rates of tropospheric CH₄ (2020–2022) could also be related to, and partially explained by, a reduction in some other atmospheric pollutants, such as sulfur dioxide (SO₂) and the resulting sulfates (SO₄).

2.1. The Sulfur Hypothesis

Aerosols (principally sulfates) have a net climate cooling effect, but they also lead to millions of premature deaths per year as they are deleterious pollutants causing asthma and other pathologies. The reduction in these aerosols started in the 1970s in the US and Europe, as shown in Figure 2 [39], although it has been described as a ’Faustian bargain’ [40,41].

A new pollution reduction strategy involving “sulfur” officially came into force on the 1st of January 2020: the International Maritime Organization 2020 rule (IMO 2020 rule) [42]. It imposes a seven times lower limit on the sulfur content in fuel oil (0.50% instead of the previous limit of 3.5%) and a limit of 0.10% in “designated emission control areas”, mainly near the coasts.

Without going into the chemistry in detail, during the combustion process, the compounds containing sulfur from bunker fuel or other fuels are oxidized and come out from the exhaust chimneys as SO₂, which then is oxidized in the atmosphere to form SO₃ and then to form SO₄ in the form of sulfuric acid (H₂SO₄).

Regional declines in anthropogenic sulfur emissions started much earlier, but globally, emissions continued to increase due to the growing emissions in developing countries [43,44]. Since 2006 and 2012, due to air pollution mitigation policies and taxes in China on coal power plants [45], Chinese SO₂ emissions declined quite rapidly and India became the principal emitter, as seen in Figure 2 [46]. NOx emissions also started declining in similar ways regionally (Figure 3). Consequently, they also started declining globally, though they continue to grow in India and some other smaller countries. Our hypothesis is that the decline in SO₂ emissions has been accelerated due to the “IMO 2020 rule” [42], but also to other factors, such as the economic decline due to the COVID-19 pandemic in 2020 [47], which is also illustrated in Figure 4, showing the abrupt reduction in SO₂ maritime emissions in European maritime waters in 2020 [48,49].

Figure 2. Regional trends of SO₂ emissions (reproduced from [39], based on the global aerosol dataset CEDS [49]): SO₂ emissions have been decreasing since the 1980s in the US and EU, since 2005 in China, and globally since 2012.

Figure 2. Regional trends of SO₂ emissions (reproduced from [39], based on the global aerosol dataset CEDS [49]): SO₂ emissions have been decreasing since the 1980s in the US and EU, since 2005 in China, and globally since 2012.

In order to comply with these new pollution emission rules, some marine vessels have started using cleaner but more expensive fuels with a low sulfur content or diesel. Other marine vessels used exhaust gas cleaning systems (scrubbers) to remove SO₂, NOx, and particulate matter from the exhaust gases of their combustion engines, resulting in an additional reduction in NOx emissions.

The benefits for health are important: before the IMO 2020 rule, SO₂ emissions were estimated to contribute to between 19,000 and 91,000 premature deaths per year in coastal regions [50]. Zhang (2021) estimated that the IMO 2020 rule could prevent premature deaths from PM2.5 attributable to shipping by approximately 30,200 deaths per year [51]. Tier III NOx regulation could prevent an additional 33,000 deaths [51]. Strengthening this regulation by limiting the sulfur content to 0.1% could reduce the annual number of premature deaths by an additional 5070 deaths [51]. Corbett (2016) estimated that delaying the sulfur reduction to 0.5% in maritime fuels until 2025 will cause about 569,000 cumulated additional deaths between 2020 and 2024 [52].

2.2. Sulfate Deposition over the Continents Reduces Biogenic CH₄ Emissions

The anaerobic oxidation of CH₄ (AOM) to CO₂ by methanotrophs is an important CH₄ sink. In the marine environment, sulfate is the primary electron acceptor of AOM, while in terrestrial water-logged systems, such as wetlands and rice paddies, nitrite- and nitrate-driven AOM is more common [53]. However, studies in North America, Europe, and Asia have shown that increased sulfate deposition also reduces the CH₄ emissions from wetlands [54]. Simulated acid rain using sodium sulfate (Na₂SO₄), a non-acidic sulfate salt, was applied in field experiments to adjust microbial populations [55] (i.e., to out-compete methanogens and favor methanotrophs). This adjustment could reduce CH₄ emissions from natural wetlands by as much as 40% [56,57,58,59]. Similar reductions in CH₄ emissions by sulfur deposition from rice fields were known earlier [60] and probably also occur in lakes [61], dams, and other water bodies. The principal feedback couplings on the atmospheric CH₄ concentration are summarized in Figure 5.

Landfills and waste management are responsible for about 69 Tg CH₄ yr⁻¹ of emissions globally, representing about 19% of total global CH₄ anthropogenic emissions [6]. More and more often, landfills are covered by geotextiles, and mitigation methods are applied, such as biocovers and gas collection [62]. Since their total area is several orders of magnitude smaller than that of wetlands, sulfate deposition and its effects on landfills are not included in this manuscript.

The deposition of SO₄ over the oceans has little or no effect on CH₄ emissions, but as SO₄ has an atmospheric lifetime of about 4 days [62,63], it can be transported over long distances above continents [64,65,66,67]; additionally, it can also be present on soot particles, mineral dust, and other particulate matter (PM) [68,69]. Therefore, it can be argued that the reduction in SO₂ emissions by marine vessels on the high seas can still be important for terrestrial ecosystems, even if near the coasts in “designated emission control areas”, there is a stricter 0.1% sulfur limit. For instance, aircraft sulfate measurements and chemical transport models have estimated that about 56% of the measured sulfate over British Columbia is due to East Asian sources, and that the mean amount of springtime sulfate in Western Canada has increased on the surface by ~30% due to anthropogenic East Asian sulfur emissions [64].

Global sulfate deposition over the continents from past anthropogenic atmospheric SO₂ emissions due to the combustion of coal and fuel has minimized or partially offset the impacts of climate warming by decreasing rates of methanogenesis [55] and reducing global wetland CH₄ emissions since the middle of the 20th century [54].

2.3. The 1991 Mount Pinatubo and Cerro Hudson Eruptions Plus the 1783–1784 Laki Eruption

Volcanic eruptions are another source of SO₂ emissions that result in global cooling.

A computer simulation showed that a large hypothetic Icelandic volcanic eruption can reduce the CH₄ emissions of the northern wetlands by 50% [70].

The Earth has cooled globally by about 0.5 °C during the 15 months following the 1991 eruption of Mount Pinatubo in the Philippines [71]. It is estimated that about 18 (±4) Tg (or million metric tons) of SO₂ were injected into the stratosphere [72].

Though little known, in 1991, another volcano, the Chilean Cerro Hudson, erupted (between August 8th and 15th), injecting between 1.7 and 2.9 Tg of SO₂ into the upper troposphere and lower stratosphere [73].

As can be seen in Figure 1, adapted from [37], the average CH₄ growth rates in the atmosphere are as follows: 11.4 ppb yr⁻¹ during 1984–1990, 6.5 ppb yr⁻¹ during 1991–1998, 0.7 ppb yr⁻¹ during 1999–2006, 6.6 ppb yr⁻¹ during 2007–2014, and 10.8 ppb yr⁻¹ during 2015–2023. Meanwhile, in 1992 and 1993, they were only 2.4 ppb and 3.8 ppb, respectively, with a large amount of variability in the following 4 years. Some have hypothesized a decrease in anthropogenic CH₄ emissions after the collapse of the Soviet Union in December 1991 [74], but this has been contested [75].

Although there are other intervening factors, such as the El Niño–Southern Oscillation (ENSO) (discussed in Section 3), the changes in CH₄ growth after the Pinatubo and Cerro Hudson eruptions have been explained by changes in tropospheric UV levels and due to enhanced stratospheric O₃ depletion caused by the presence of SO₂ and SO₄, as well as by the perturbed climate (cooling due to SO₄ aerosols), which has altered both the sources and sinks of CH₄ [76]. To be more precise, in 1991, the UV flux decreased by ∼12% immediately after the Pinatubo eruption due to direct absorption by SO₂, which reduced °OH generation and led to anomalously large growth rates for CH₄ and CO in late 1991 [77]. The increase in atmospheric CH₄ in 1991 became the biggest on record until 2019.

Then, in 1992, UV and visible light scattering by SO₄ aerosols in the stratosphere reduced the amount of radiation reaching the troposphere, cooling the Earth by about 0.5 °C. This global cooling also reduced the generation of °OH, due to the decrease in water vapor associated with the temperature reduction, but seems to have reduced the CH₄ emission rates from wetlands even more [78]. These effects and feedback mechanisms are summarized in Figure 5.

Taking into consideration the numerous studies [54,55,56,57,58] mentioned in the previous section, the effects on the CH₄ emissions from wetlands due to the additional deposition of sulfates from the Pinatubo and Cerro Hudson eruptions might also be an important factor, although it has not been taken into consideration by many publications. It was only considered a minor factor by Bândă et al. (2015, 2016), who assumed the linearity of the relation between the SO₂ emissions increase and CH₄ emissions reduction [76,79]. The reason for this was that the previous modelling results of the Laki eruption (June 1783–February 1784) [70] showed SO₂ emissions that were six times greater and that led to a relatively modest 8.8 Tg yr⁻¹ decrease in CH₄ emissions from wetlands in the three years following the eruption. It did not take into account the fact that the duration, the location, and the type of the deposition are also important and that data from field experiments have shown that the long-term impact of sulfate deposition on CH₄ emissions lasts 5 to 10 years [57].

The global CH₄ emissions were reduced by about 23 Tg the first year of the Laki eruption. The model used shows a reduction in CH₄ of about 14 Tg due to the aerosol cooling effects of the SO₄ and about 8.8 Tg due to the sulfur (S) deposition in the northern wetlands. In the following two years, the CH₄ emissions were reduced by about 20 Tg yr⁻¹, with about 10 Tg yr⁻¹ due to each effect (aerosol cooling and S deposition). The CH₄ emission rates continued to be lower than normal for a few more years, probably as the effect of S deposition may have lingered beyond the first 3 years, owing to S recycling [57]. The Laki eruption injected SO₂ into the troposphere, and most of it was emitted during the first two months of the eruption (91 Tg SO₂ out of a total of 122 Tg). As their transport and lifetime are relatively short, the S depositions were localized within the northern latitudes (30°N), exceeding 15–20 kg S ha⁻¹ yr⁻¹, beyond which the SO₄ reduction rates in wetlands showed an asymptote toward a constant value. Therefore, the suppression of CH₄ emission achieved a maximum value of about 40% [54]. Also, the deposition mainly occurred as SO₂ and acid rain (H₂SO₄). Meanwhile, for the 1991 Pinatubo eruption, the SO₂ reached the stratosphere, had a much longer lifetime, and circled the globe before a progressive and wide deposition of the sulfates occurred in both hemispheres and in tropical wetlands. This deposition probably occurred as acid rain, but also as non-acidic Na₂SO₄ salt, after cation exchange with sea salt from sea spray [80].

The global S deposition of the Pinatubo and Cerro Hudson eruptions was in addition to the S deposition from anthropogenic sources due to the large SO₂ emissions from fossil fuel use in the US and the EU and the still limited but growing SO₂ emissions in developing countries (Figure 2) [44]. Therefore, we propose the hypothesis that the S deposition in 1991 and 1992, although about 5–6 times smaller in magnitude than that from the Laki eruption, might have been more effective and may better explain the level of CH₄ emissions from wetlands in the years following (1992–1993) (Figure 1) and perhaps even later, as the effects of sulfur deposition on water bodies last 5 to 10 years [57].

No similar effects on the CH₄ growth rate seem to have occurred after the El Chichón eruption in March–April 1982. The yearly dataset of NOAA only starts in 1984 [37], but another dataset, which is biennial and was established from ice cores in Antarctica and Greenland [81], indicates only slight reductions of about 0.25 ppb yr⁻¹ and 1.55 ppb yr⁻¹ in the CH₄ growth rate in 1982–1984 and 1984–1986, respectively, compared to 1980–1982. One plausible explanation for the El Chichón eruption not having an effect on the SO₂ emissions could be the lower amount of SO₂ released than that during the Pinatubo eruption (i.e., about 7 Tg vs. 18 Tg of SO₂) [82], which is about three times lower when also taking into account the emissions from the Cerro Hudson eruption. It is also possible that somehow the SO₂ emitted by the El Chichón eruption just compensated for the reduction in anthropogenic emissions of SO₂ between 1980 and 1983–1985, as shown in Figure 2 [39].

2.4. Additional Hypothesis: Wildfires, Global Warming, and Decrease in NOx Emissions

Part of the CH₄ growth rate in 1998 (72%) (Figure 1) has been explained by the CO from biomass burning (which increased by more than 40%), which, in turn, decreased global tropospheric °OH concentrations by up to 9% and increased the CH₄ atmospheric lifetime that year by 4% [83].

VOC emissions from plants, bacteria, and fungi exhibit variability in concentrations and compositions due to environmental parameters including temperature and humidity dependence [84]. In the atmosphere, CO is an intermediate compound during the oxidation of VOCs and CH₄ by °OH to CO₂. CO and CH₄ are therefore the principal sinks of °OH. For 2022–2024, if the lower accumulation of tCH4 was due to an increase in °OH, this also means that there will have been a concomitant reduction in global atmospheric CO.

The amount of chemical production of CO in the atmosphere is very similar to the amount of direct CO emissions from the Earth’s surface, mainly due to the incomplete combustion of fossil fuels and biomass. Biomass combustion includes anthropogenic sources as well as forest fires and exhibits high interannual variability [85]. Meanwhile, there is a global tendency towards a decrease in anthropogenic CO emissions [39,86]. Unfortunately, only the data for the latter are available at the time of the writing of this review article, so the evolution of global atmospheric CO will need to be confirmed by further work using coupled global chemistry–climate models. The Australian mega-bushfires (December 2019–January 2020) were unprecedented in their extent and intensity, burning more than eight million hectares of vegetation [87,88]. During the COVID-19 lockdowns in Sub-Saharan Africa, a 17.8% increase in CO was observed [89]. Meanwhile, the 2020 Western US wildfires produced CO emissions that were three times the 2001–2019 average [90], and in 2021, boreal fires reached more than two times the 2001–2019 average [91].

Global warming enhances the biogenic emissions of CH₄ by methanogens and some non-methane VOCs emitted by plants [92]. The fact that the past nine years (2015–2023) have been the nine hottest years on record [93] can explain the latter part of Figure 1, but only partially for 2023 and 2024, which are the hottest years on record so far. The feedback mechanisms between VOCs, CO, CH₄, and other parameters are summarized in Figure 5.

For the year 2020, part of the increase in tropospheric CH₄ was explained by a reduction in the NOx emissions by airplanes during the COVID-19 lockdowns [2]. But there had already been a trend of NOx pollution emission reductions in Europe and the US [94] as well as in China (starting from 2011) [95], as can be seen in Figure 3 (reproduced from [49]).

For the shipping industry, the International Convention for the Prevention of Pollution from Ships Annex VI [96] set a “Tier II” NOx emission limit from 7.7 to 14.4 g/kWh (depending on the engine’s rated speed) for engines installed after 1 January 2011 and a “Tier III” NOx emission limit from 3.4 to 2 g/kWh (depending on the engine’s rated speed) for engines installed after 1 January 2016 and operating in emission control areas. The reduction in NOx emissions required by the IMO Tier III regulations when comparing a Tier III to a Tier II engine are approximately 76%. But due to the enhanced global maritime traffic, the effects of the progressive reductions in NOx emissions from marine diesel engines installed on ships have probably had, at the moment, little effect on the tropospheric O₃ and °OH levels. Nevertheless, the more recent analysis published for 2020–2022 still attributes 28% of the CH₄ surge to a reduction in the °OH sink and the rest to emissions from wetlands, rice paddies, and inundations due to La Niña conditions [25].

3. Can the Significantly Reduced Growth Rates of tCH₄ in 2023 and in 2024 Be Explained?

Although this question is out of the scope of this perspective article and is not included in the title, it is interesting to formulate some hypotheses, as summarized in

Table 1. Summary of hypotheses that explain the higher growth rate of CH₄ accumulation in the troposphere in 2020–2021 and its decline in 2022, 2023, and 2024.

	Years
Hypothesis (Due to)	2020	2021	2022	2023	2024
Observed CH4 growth (ppb) [4]→	14.81	17.64	13.25	8.39	3 to 8 a,b
Inundations, increased area of shallower waters → more CH4 production from wetlands	[2,3,4,24,25]	[4,25]	[25]	--	--
Global temperature increases → more CH4 production from wetlands	[2,3]	[4]	--	This article	This article
Reductions in NOx emissions during COVID-19 lockdowns → less °OH → more CH4	[1,2,4]	--	--	--	--
Lower global NOx emissions → less °OH → more CH4	[2,25]	[4,25]	[25]	--	--
IMO-2020 Rule → less SO4 deposition on water bodies → more CH4 production from wetlands	This article	This article	This article	This article	This article
Global temperature increases → more humidity → more °OH → less CH4	--	--	--	This article	This article
Reduction in cloud cover → more UV → more °OH → less CH4	--	--	--	This article	This article

a: The preliminary estimation from Copernicus for 2024 is 3 ppb [38]; b: The preliminary estimation for Gosat for 2024 is 8 ppb [38].

.

As 2024 and 2023 are the warmest years since preindustrial times, and accounting for the sensitivity of wetlands to temperature [2,97] and the fact that the IMO 2020 rule still applies, our first hypothesis is that the CH₄ emissions from wetlands remained high and might even have continued to grow in 2023 and 2024. Also, we think that it is too early for the Global Methane Pledge, which was signed by about 150 countries in 2021 [98], to have produced effects on CH₄ anthropogenic emissions, as it is a non-binding agreement and several major oil and gas producers, such as those in Russia and Saudi Arabia, have not signed it. However, the growth rate of tCH₄ fell, respectively, by 4.39 ppb in 2022 compared to 2021, by 4.86 ppb in 2023 vs. 2022 [37], and by an estimated 5.39 ppm in 2024 vs. 2023 according to Copernicus’s estimations for 2024 [38]. For 2024, the preliminary estimations of Copernicus [38] are a tCH₄ growth rate of only 3 ppb (±2), the lowest since 2006, and yet 2024 has been the warmest year since preindustrial levels, followed by 2023 and then by 2016.

3.1. What Other Reasons Are There for the Fall of the tCH₄ Growth Rate in 2023 and 2024 Other than Lower CH₄ Emissions from Wetlands?

This hypothesis seems improbable, as 2024 and 2023 are the hottest years since pre-industrial times [35,38] and warmer temperatures result in more CH₄ generation by methanogens and therefore more CH₄ emitted in water bodies such as wetlands, hydroelectric dams, lakes, rivers, and ponds. Also, some sources have suggested the El Niño–Southern Oscillation (ENSO) is a strong force in the CH₄ cycle, but this was not confirmed by Schaefer (2018) [99], even though it was found that wetland CH₄ emissions declined in 2015 during the onset of the 2015–2016 El Niño event but then increased to a record high in 2016 during the later stages of the El Niño event [100]. However, the reverse happened for the 2023–2024 El Niño event, at least for the growth rate of tCH₄. A chemistry–climate model suggests that El Niño events modulate tropical OH (where its levels are highest) by changes in specific humidity, ozone availability, and NOx production through lightning [101], and the conditions in 2023 and 2024 were hotter with more humidity.

Perhaps the area of water bodies decreased? Probably not, as globally, there were almost as many inundations and floods as in previous years [102,103], if not more, and in 2024, 132 floods were recorded [104].

For all these reasons, we think that the fall of the tCH₄ growth rate is not due to lower CH₄ emissions from wetlands.

Could the fall of the tCH4 growth rate in 2022-2024 be explained by the reverse hypothesis formulated to explain the increased growth rate of 2020-2021? Is an increase in SO₂ and/or NOx emissions responsible?

More coal power plants became operational and the global coal power plant fleet grew in 2023 by 2% or 48.4 GW [105], with two-thirds of those having been added by China. A similar growth is expected in 2024 with 89 GW from plants already in construction by the end of 2023. But thanks to their internal emissions regulations, the global increases of SO₂ and NOx emissions in 2023 and 2024 have been modest and lower than the 7–8 Tg reduction in 2020 due to the IMO 2020 sulfur rule.

Flare gases contain much more SO₂ than the combustion products of traditional fuels because environmental regulations impose strict standards on the sulfur contents of conventional fuels, while there is no control on the composition of the emitted flare gases. Gas flaring increased by 7% in 2023 [106], with the largest amounts coming from Russia, Iran, Iraq, the United States, and Venezuela. The long-range transport of SO₂ [65,66,67] could reach the principal wetlands, but the increases in SO₂ emitted by gas flaring are lower than the reduction in emissions by shipping since 2020. By putting emissions limits on coal plants and introducing desulphurization technologies that remove SO₂ from smokestacks, China has dramatically reduced local air pollution levels in the last decade [49,107]. There has also not been a major volcanic eruption with significant SO₂ emissions [108,109].

3.2. Our Hypothesis

Our second hypothesis therefore is that in 2023 and 2024, the increase in the principal sink of CH₄ (the generation of °OH) grew and more than offset the increased emissions of CH₄ from wetlands, resulting in lower growth rates of tCH₄ in 2022, 2023, and 2024 compared to 2020 and 2021.

The primary production of °OH occurs principally via O₃ photolysis to the excited state O (¹D) followed by a reaction with water vapor (reactions R1 and R2):

O₃ + hv (λ < 336 nm) → O(¹D) + O₂            (R1)

O(¹D) + H₂O → 2 °OH                                 (R2)

This °OH formation pathway depends on the actinic flux at wavelengths ≤336 nm and on the abundance of H₂O. Consequently, the highest concentrations of °OH are found in the tropical lower and middle troposphere. Also, the stratospheric O₃ column is lower in the tropics than elsewhere, allowing for greater tropospheric penetration of UV radiation [110], and most of the biogenic methane is produced by tropical wetlands [6].

Seventeen global climate–chemistry models have been compared in the context of the ‘Atmospheric Chemistry and Climate Model Intercomparison Project’. There is an agreement among the models indicating that in order to enhance °OH and therefore to reduce CH₄ in the troposphere, the main factors to consider are increased humidity, tropospheric ozone, tropospheric temperature, UV radiation (due to decreases in stratospheric ozone), and NOx emissions [111].

Using box-model inversions and global atmospheric chemistry models, previous studies have calculated that a 4% decrease in °OH is equivalent to an increase of 22 Tg yr⁻¹ in CH₄ emissions [112,113]. Peng (2022) found that a 1.6 ± 0.2 percent decrease in the tropospheric °OH concentration during the COVID-19 pandemic resulted in a global methane imbalance of 7.5 ± 0.8 Tg yr⁻¹ due to a lower CH₄ + °OH reaction [8]. Based on inversions using a 3D chemical transport model, it was estimated that a 1% decrease in the °OH burden leads to an increase of 4 Tg yr⁻¹ in CH₄ [114]. A strategy has been suggested to reduce the tCH₄ concentration by the enhancement of atmospheric water vapor and then increasing °OH sources [115].

In summary, our hypothesis is that in 2023–2024, the production of °OH increased significantly, decreasing the growth rate of tCH₄ globally, even though the emissions of CH₄ from wetlands remained high or even continued to grow.

• Effect of increased humidity

The total amount of atmospheric water vapor reached record values in 2024 and 2023, respectively, which were 4.9% and 3.3% above the 1991–2020 average (and with 3.4% in 2016, which has been the third warmest year so far) [38]. According to [116,117], the most important meteorological factor affecting °OH and CH₄ lifetime is tropospheric humidity. As per the reaction (R2), higher water vapor concentrations in the troposphere mean that more °OH is produced through a reaction with singlet oxygen atoms O(¹D) [118]. In Figure 6, the yearly evolution of atmospheric humidity (the total column of water vapor over the 60°S–60°N domain expressed in % relative to the average for the 1992–2020 reference period [38]) and the surface air temperature are shown comparatively, along with the growth rate of atmospheric CH₄. In Figure 7a,b, an inverse correlation seems to appear for the years 2020–2024 (which is the period considered in this article), but it should be kept in mind that the build-up of tCH₄ is a multifactorial phenomenon and that a single parameter cannot explain its entire evolution. Nicely (2018), Holmes (2018), and other scientists agree on the positive impact of increased water vapor on °OH levels and the reduction in CH₄ lifetime [111,119,120,121]. A 1% increase in global mass-weighted mean water vapor was found to lead to a 2.237% increase in °OH concentration in the total atmosphere. Water vapor can also enhance the production of peroxyl radicals, which drive O₃ production [122].

• Effects of increased temperature

Higher tropospheric temperatures result in increased oxidation rates and decreased CH₄ [118,123]. The CH₄ + °OH reaction rate increases significantly with temperature [124]. Through its positive effect on tropospheric humidity, as well as its effect on the CH₄ + °OH reaction rate and the absorption cross-section of O₃ (which is important for photolysis to produce O(¹D)), temperature can be directly and indirectly linked to the lifetime of °OH and methane [118]. The lifetime of CH₄ could be reduced by 4.7 ± 1.7% per degree of tropospheric temperature increase, due to the combined effect of a higher oxidation rate and a higher water vapor concentration [125].

• Effects of clouds and of UV radiation from sunlight

Recent work [126] suggests that the unusually high global temperatures of 2023 are due to the record-low planetary albedo level, which is the consequence of reduced low-cloud cover in the northern mid-latitudes and the tropics, resulting in more sunlight reaching the Earth’s surface. As already mentioned, the highest concentrations of °OH occur in the tropics, where the amount of incoming sunlight is at its greatest and O₃ is readily photolyzed [127]. The radiative influence of clouds is a strong contributor to the variability in the tropospheric oxidizing capacity, as cloudiness affects shortwave radiation. Clouds directly affect gas-phase tropospheric chemistry by decreasing shortwave radiation below them and increasing it above them. This difference leads to important effects on photolysis and, through O(¹D) modifications, on °OH concentrations above and below clouds (especially in the boundary layer), with a variety of subsequent effects on ozone and its precursors [122,128,129].

Several studies have found that with clouds, the global annual average net chemical production of ozone decreases by 8–15% at the surface compared to clear-sky conditions [122,128,129]. It is known that clouds decrease mass-weighted global °OH concentrations with a higher level of CH₄ loss above them [130,131]. During periods with clouds, the O₃ concentrations are reduced as a result of decreased formation [132]. Additionally, NOx, which acts as a catalyst in the photochemical formation of O₃, is less available in clouds. At night, NOx is depleted by clouds via the scavenging of N₂O₅, which significantly reduces NOx concentrations during subsequent daylight hours. Consequently, clouds directly reduce the concentrations of NOx, O₃, and therefore °OH [132]. Of course, a reduction in the number of clouds has the reverse effect.

These are only some hypotheses to try to explain the 2023–2024 reduction in the rate of the CH₄ build-up in the atmosphere, and the principal ones are summarized in Table 1. Nonetheless, it should be kept in mind that not a single parameter could account for this multifactorial process and these hypotheses need to be confirmed by global climate–chemistry models, which is outside the scope of this perspective article.

4. Discussion, Research Gaps, and Prospects

In their article “Global warming in the pipeline”, Hansen et al. (2024) demonstrate that humanity is driving an acceleration of global warming, with the main factor of this acceleration being the reduction in human-made aerosols in the atmosphere [133]. Both SO₂ and NOx have cooling effects [134], and between 2019 and 2022, there was a reduction of 9 Tg in anthropogenic SO₂ emissions (from 85.9 to 76.9 Tg yr⁻¹, respectively) and also a significant reduction in NOx emissions (from 142.9 to 131.8 Tg yr⁻¹, respectively) [135].

The reduction in harmful and deadly pollution has many benefits for health as seen in Section 2.1 but, as part of the ’Faustian bargain’ [41], risks becoming more and more complex when accounting for additional effects, such as the fact that lightning, one of the natural sources of biogenic NOx, decreased over shipping lanes by about 40% after 2020 [136]. Also, S is an essential nutrient for plants, and between 2000 and 2020, the amount of plant-available S in soil has decreased by 34–86% [137]. Therefore, the reductions in the atmospheric deposition of S to agricultural soils might lead to future nutrient deficiencies [138,139].

The addition of Na₂SO₄ or other neutral sulfate salts to rice paddies and wetlands has been experimentally proven to reduce CH₄ emissions [54,55,56,57,58,59]. Ammonium sulfate (NH₄)₂SO₄, as well as iron, magnesium, zinc, potassium, and calcium sulfates, are commonly used as fertilizers to increase agricultural yields [140], and FeSO₄, which is authorized for organic farming, is also used to fight plant chlorosis [141]. Therefore, the addition of neutral sulfate salts to freshwater wetlands might be a possible mitigation method to reduce CH₄ emissions, but environmental impact assessments should be performed before the large-scale use of neutral sulfate salts to reduce CH₄ emissions.

Climate–chemistry transport coupled global models already take into account changes in the emissions of short-lived pollutants and, in particular, tropospheric O₃ precursors, which interact in many ways to influence the atmospheric chemical composition and affect its oxidizing capacity, thereby influencing the lifetime of CH₄ [142]. A decline in °OH has been used to explain the growth in tropospheric methane due to biogenic sources [2,25,112], which also explains the decline in the atmospheric isotopic ¹³CH₄/¹²CH₄ ratio [143].

Due to their short lifetimes, SO₂ and NOx are not well-mixed gases and have mainly regional effects, but they have a global cooling effect. The climate simulations related to the effects of large volcanic eruptions should use climate–chemistry and transport global models integrating the effects of sulfur deposition on CH₄ biogenic emissions from continental freshwater bodies, as well as the cumulative sulfate regional or local deposition due to anthropogenic sources. It should be noted that coastal and ocean sources of methane already contain SO₄ and emit much lower amounts of CH₄ (12 Tg) than freshwater bodies such as lakes, ponds, dams, and rivers (112 Tg) [6].

According to a methanotrophy model, from 1900 to 2015, the global CH₄ uptake by soils doubled and could increase further until 2100 [144], but this model does not take into account the global reduction in SO₄ deposition. From 1980 to 2011, the CH₄ uptake in global forest soils increased slightly [145], but some field measurements show that, more recently, the CH₄ uptake in forest soils is declining [146].

If the hypothesis developed here is confirmed, a number of earlier findings should probably be revised, as illustrated in the next examples.

Numerous publications have tried to explain the drivers of “the pause” or “the slowdown” in CH₄ growth from 1999 to 2006 (Figure 1) and its renewed growth onward [112,147,148,149] (and references cited therein), but to our knowledge, none have taken into consideration the effects of regional or local sulfate deposition on wetlands, hydroelectric dams, rice paddies, and other freshwater bodies or of nitrate and nitrite deposition, which can affect methanotrophy in synergy with SO₄ [61].

Under the marine boundary layer, sulfuric acid (and nitric acid) displacement from NaCl sea salt aerosols [80,81,82] produces sodium sulfate (and sodium nitrate) and releases hydrochloric acid (HCl), a precursor of chlorine atoms, which is another natural sink for CH₄ [7]. Consequently, the projected reduction of 8 Tg in SO₂ emissions from global shipping [150] due to the IMO-2020 rule has also possible implications for other tropospheric oxidants [151,152] and for CH₄ by reducing tropospheric HCl and, therefore, the generation of Cl atoms.

Keith (2017) suggested that climate engineering by stratospheric sulfate injection (SSI), which is no substitute for cutting GHG emissions, can help reduce the atmospheric carbon burden by up to 100 gigatons of carbon (367 gigatons CO₂) in an RCP8.5 scenario by its cooling effect [153], but the author did not account for the additional effects on the reduced CH₄ emissions due to continued global sulfate deposition. This effect was not taken into account by at least two other teams, who concluded that this type of SSI would increase the lifetime of CH₄ (by approximately 16%) and its concentration by 200 ppb [154] and that the sulfate deposition would not be enough to impact most ecosystems [155]. If the hypothesis of S deposition reducing CH₄ emissions is correct, the conclusions drawn as a result might be quite different.

A global warming effect from 0.03 °C up to 0.2 °C due to the IMO 2020 sulfur rule is ‘anticipated’ [156,157,158,159,160,161,162] but is difficult to detect and might not be detectable immediately. Two recent papers suggest that the IMO 2020 rule might have contributed to a greater level of warming, leading to the recent record-breaking warming years [163,164]. Taking into account the increase in CH₄ emissions could improve these estimates.

We suggest that new simulations should be performed that take into account the effect of lower sulfur deposition on wetlands and the resulting increased emissions of CH4, to better evaluate the effects of the IMO 2020 rule.

5. Conclusions

The IMO 2020 rule has reduced the sulfur content of marine fuels by 86% and the SO₂ emissions of the shipping industry by about 75% [42]. The reduction in SO₂ emissions in 2020 was about 7–8 Tg and cumulated with the results of other air pollution control policies for health purposes. These policies include the IMO Tier III regulations on NOx emissions reductions (a Tier III engine represents a −76% reduction when compared to a Tier II engine) [96] and all the long-term trends for global SO₂ and NOx emissions reductions on land from power plants and vehicles (Figure 2 and Figure 3).

In 1990, Lawrence and Crutzen [165] showed the influence of NOx emissions from ships on marine photochemistry by surface O₃, °OH, CH₄, and its climate cooling effects. For the year 2020, part of the increase in tropospheric CH₄ was explained by reduced amounts of NOx emissions during the COVID-19 lockdowns [2,3,27]. As NOx emissions are linked to surface ozone production, which is a GHG and is harmful for all living organisms, from the climate change and air pollution perspective, its reduction seems logical but will not necessarily improve air quality (which also depends on the amount of NMVOC present) and can increase the lifetime of CH₄ [166] (Figure 5), adding to the paradox that cleaner air can increase global warming [167].

The new maritime regulations [42] and the reduced activities [168] during the COVID-19 pandemic lockdowns also accelerated the decline in SO₂ emissions and SO₄ deposition, which started in the 1970s in the US and Europe and in 2005–2015 in Asia (Figure 2). Based on the numerous experimental studies about the effects of SO₄ deposition on wetlands and rice fields [54,55,56,57,58], we formulate the hypothesis that the decline in SO₂ emissions might also partly explain the sudden jump in CH₄ concentrations during 2020–2022 due to an increase in biogenic CH₄ emissions, in addition to global warming [31,32,33]. A recent article confirms our hypotheses and finds that an extra 20–34 Tg (million tons) of CH₄ could be released annually due to reduced SO₂ emissions [169]. Our retrospective analysis of the S emissions in 1991 after the Pinatuno and Cerro Hudson eruptions could explain the small rates of increase in tCH₄ in 1992–1993.

This new hypothesis still has to be verified using global chemical transport and deposition model simulations, as well as climate–chemistry coupled global models. Additional hypotheses include the reduction in nitrate and nitrite deposition over the continents and increased biogenic NMVOC emissions due to global warming and CO due to wildfires.

The explanations for the 2020–2021 CH₄ surge, as well as the 2022–2024 decline, are multifactorial. Concerning the observed decline in the growth rate of tCH4 in 2022, 2023, and 2024 by more than 4 ppb yr⁻¹, our hypotheses are that the emissions from wetlands continued to be higher than before 2020, but that the amount of °OH, the principal sink of CH₄, increased even more, due to increased humidity and UV radiation, which compensated for the increased CH₄ emissions. These hypotheses also need to be verified using global chemical climate models, which is out of the scope of this perspective article.