New evidence for CH4 enhancement in the upper troposphere associated with the Asian summer monsoon

The Asian summer monsoon (ASM) region is a key region transporting air to the upper troposphere (UT), significantly influencing the distribution and concentration of trace gases, including methane (CH4), an important greenhouse gas. We investigate the seasonal enhancement of CH4 in the UT over the ASM region, utilizing retrievals from the Atmospheric Infrared Sounder (AIRS), model simulations and in-situ measurements. Both the AIRS data and model simulation reveal a substantial enhancement in CH4 concentrations within the active monsoon region of up to 3%, referring to the zonal means, and of up to 6% relative to the pre-monsoon season. Notably, the spatial distribution of the CH4 plume demonstrates a southwestward shift in the AIRS retrievals, in contrast to the model simulations, which predict a broader enhancement, including a significant increase to the east. A cross-comparison with in-situ measurements, including AirCore measurements over the Tibetan Plateau and airline sampling across the ASM anticyclone (ASMA), favors the enhancement represented by model simulation. Remarkable CH4 enhancement over the west Pacific is also evidenced by in-situ data and simulation as a dynamical extension of the ASMA. Our findings underscore the necessity for cautious interpretation of satellite-derived CH4 distributions, and highlight the critical role of in-situ data in anchoring the assimilation of CH4.


Introduction
The importance of methane (CH 4 ) as a greenhouse gas lies in its 15%-20% contribution to the global radiative budget since the pre-industrial period: ∼30 times that of the global warming potential of carbon dioxide (CO 2 ) over one hundred years, as well as a relative efficient climate benefit from reducing its emissions (IPCC 2021).The radiative effect of CH 4 depends not only on its atmospheric column integrated amount but also its spatial distribution.The surface temperature is highly sensitive to CH 4 changes in the upper troposphere and lower stratosphere (UTLS, Riese et al 2012).In addition, high CH 4 entering the UTLS critically influences the chemistry in the upper atmosphere globally.Therefore, the evaluation of CH 4 in the UTLS and an understanding of the relevant processes is important for climate change.
The Asian summer monsoon (ASM) has been well recognized as an efficient transport system into the upper troposphere (UT), which rapidly uplifts near-surface polluted air masses through deep convections and confines the air masses through largescale anticyclonic circulation (Randel et al 2010, Bian et al 2020).The remarkable increase in several chemical components with surface emission, including carbon monoxide (CO), water vapor, hydrogen cyanide (HCN), aerosol and CH 4 , confined within the ASM anticyclone (ASMA), has been observed via satellite retrievals and in-situ measurements (Rosenlof et al 1997, Park et al 2009, Xiong et al 2009, Randel et al 2010, Pan et al 2016, Yu et al 2017).The tropospheric components uplifted to the upper-level anticyclone potentially make their way into the stratosphere through slow diabatic ascending (Fan et al 2017, Vogel et al 2019), as well as through the isentropic eddy shedding from east and west directions (Garny and Randel 2013, Pan et al 2016, Honomichl and Pan 2020).
CH 4 plumes over the Asian monsoon region were reported with a similar magnitude at 300 hPa to 150 hPa from May to September by several chemical transport models in comparison with a series of satellite retrievals, including the Halogen Occultation Experiment (HALOE), Atmospheric Infrared Sounder (AIRS) and the Greenhouse gases Observation SATellite (GOSAT) (Park et al 2004, Xiong et al 2009, Belikov et al 2022), and by limited in-situ measurements (Schuck et al 2010, Baker et al 2012).However, the differences in its vertical structure and its spatial distributions were noticed between model simulations and the satellite retrievals (Park et al 2004, Xiong et al 2009, Ni et al 2023), which were not fully discussed due to a lack of insitu measurements for validation, in particular over key regions, e.g. the Tibetan Plateau (TP).The CH 4 enhancement associated with the ASM has received relatively little attention as a transport tracer due to its complicated sources (e.g.rice paddies, wetlands, livestock, fossil fuel extraction) and longer lifetime compared to the widely used monsoon transport tracer CO.In fact, the long lifetime of CH 4 makes it a tracer for ASM long-lasting impacts.Actually, the impact of the ASM after monsoon withdrawal was of interest in a series of recent studies, e.g. the pathways into the tropical tropopause layer, and into the mid-latitude lowermost stratosphere (LMS), as well as into the southern hemisphere (Ploeger et al 2017, Yan et al 2019, Honomichl and Pan 2020, Belikov et al 2022).However, the role of monsoon transport and seasonal elevated emissions from rice paddies in the observed CH 4 total-column enhancement over the ASM region remains debated (Zhang et al 2020, Zeng et al 2021).
A reasonable representation of the CH 4 vertical structure and accurate assessment of the CH 4 regional enhancement is fundamental for understanding the mechanism, analyzing the transport pathways and assessing relevant emissions, as well as its climate impact.Simulations with chemical transport models (CTMs) with inventory emissions optimized via satellite retrievals and/or in-situ datasets are the stateof-art modeling approach to represent the 3D CH 4 field (Monteil et al 2013, Maasakkers et al 2019, Feng et al 2022, 2023).Here, we present a comprehensive assessment of CH 4 increase in the middle to UT within the ASMA and for the plume expanding out of the ASMA toward the northeast and southwest directions using AIRS retrievals, multiple in-situ measurements and assimilated CTM model results.The differences in the spatial pattern, as well as the vertical structure between the satellite products and simulations, will be further discussed in comparison with passenger airline measurements and the newest AirCore measurements over the TP.

Global 3D CH 4 simulation
Three-hourly global CH 4 concentrations for the period 2015-2020 were simulated using the GEOS-Chem atmospheric chemistry transport model at a horizontal resolution of 2 Following the methodology presented in previous studies (Feng et al 2022(Feng et al , 2023)), the CH 4 flux are inferred through inversions of the in-situ network and satellite measurements covering the simulation period using an ensemble Kalman Filter framework.The prior CH 4 emissions consider nature emissions, including wetland emissions (including rice paddies), fire emissions and termite emissions, as well as the anthropogenic emissions based on the EDGAR v4.41 global emission inventory.GOSAT proxy XCH4: XCO2 retrievals (Parker et al 2020) are used for assimilation, which are anchored by a subset of CH 4 mole fraction observations at surface-based sites from the National Oceanic and Atmospheric Administration (NOAA) observation network.

AIRS CH 4 retrievals
AIRS is a nadir cross-track scanning infrared spectrometer, launched in polar orbit (13:30 LST, ascending node) on May 2002 onboard the EOS/Aqua platform.AIRS nominally observes the complete globe twice within one day with spatial resolution ∼13.5 km.The channels near 7.6 µm are used to retrieve CH 4 profiles.The atmospheric temperature profile, water vapor profile, surface temperature and surface emissivity required as inputs are derived from other AIRS channels.The AIRS CH 4 retrievals are sensitive in the range between 800 hPa and the lower stratosphere, with peak sensitivity around 300-400 hPa.
Due to the fact that the AIRS + AMSU products are only available until September 2016, we used the AIRS version 7 IR-only retrievals consistently for the period 2015-2020, downloaded at NASA Goddard Earth Sciences Data and Information Services Center (NASA/GES/DISC) (http://disc.sci.gsfc.nasa.gov/AIRS/data-holdings/by-data-product-V7).
The daily or monthly averaged CH 4 maps are from AIRS level-3 standard products with 1 • × 1 • spatial resolution (daily: AIRS3STD.v007and monthly: AIRS3STM.v007).The level-3 products are generated from the level-2 data that have been quality control (QC) filtered (only the 'best' or 'good' are used) and binned into a 1 • × 1 • grid.The AIRS data used for comparison with AirCore in-situ measurements are from AIRS level-2 version 7 supporting products (AIRX2SUP.v007).

In-situ CH 4 measurements by AirCore and passenger aircrafts
We use nine CH 4 profiles from ground to 25 km measured in the TP by an AirCore air sampling and data processing system.Following the concept proposed by Pieter Tans at NOAA (Karion et al 2010), an AirCore system was developed at the Institute of Atmospheric Physics, Chinese Academy of Sciences (IAP, CAS) and two campaigns were conducted at Da Qaidam (37.5 • N, 95.2 • E) during the ASM in July 2019 and August 2020 (Yi et al 2019).Ambient air masses of different altitudes were sampled into the 150 m long tube, providing a high vertical resolution in the UTLS region (∼300 m at 15 km), and then analyzed by the PICARRO G2401 gas analyzer, with an accuracy of 0.5 ppb for CH 4 .
Besides the AirCore-measured CH 4 profiles with a high vertical resolution, we also use aircraft CH 4 measurements during the period 2015-2020 to investigate the horizonal distribution and seasonal variation.We choose in-situ data from two long-term monitoring programs using commercial airlines: In-Service Aircraft for a Global Observing System (IAGOS), and the Comprehensive Observation Network for Trace gases by Airliners.To avoid QC and data processing, we take advantage of CH 4 GLOBALVIEWplus v5.1 ObsPack data products (Schuldt et al 2023), providing a reformed dataset with consistent frequency (every 15 min) in the UTLS region (300 hPa ∼ 150 hPa).

Seasonal CH 4 enhancement in UT over Asian region
Both simulation and satellite observations show a remarkable increase in CH 4 in the middle to UT over the ASMA.Two approaches are widely used to quantify the tracer enhancement associated with monsoons: one is the seasonal increase relative to pre-monsoon conditions (or annual mean), i.e. seasonality of the monsoon region itself; another is the monsoon regional mean relative to the non-monsoon region along the same latitude (or zonal mean), focusing on the spatial signature of the monsoon.Figure 1 shows the CH 4 distribution relative to zonal means (see also figure S1) using the second approach.The corresponding seasonal CH 4 concentration distributions are shown in figure S2, and its anomalies from the first method are shown in figure S3.
As shown in figure 1, the season-averaged increase from 300 hPa to 100 hPa is 1%-3% (20-60 ppb) within the Asian monsoon anticyclone (AMA), referring to the zonal mean according to the simulation.The general amplitude of CH 4 increase relative to the pre-monsoon season within the AMA is 3%-6% (see figure S3(a)) with simulation.The AIRS CH 4 retrieval shows a low bias (2%-4%) in the absolute values (Xiong et al 2008) and also a smaller monsoon enhancement compared with the simulation.The simulated CH 4 enhancement centers at the southern edge of the TP in the mid-to-upper troposphere (300-100 hPa), which is strongly associated with the dynamical center of the ASMA, shown as the geopotential height (GPH) contours.The spatial pattern of CH 4 seasonal enhancement is highly similar to high CO based on microwave limb sounder (MLS) observations (Park et al 2007) or high HCN based on atmospheric chemistry experiment -Fourier transform spectrometer (ACE-FTS) measurements (Randel et al 2010).Compared to the simulated CH 4 enhancement, we found that the enhancement center measured by AIRS generally shifts southwestward.
The model results show that the CH 4 enhancement extends southwest to the Arabian Peninsula and northeast to the Sea of Japan and northwestern Pacific, in particular in the UT, which is marked as orange vectors.The CH 4 plume southwestward and northeastward extension in the UT marks the region with frequent eddy shedding that efficiently detaches air masses inside/outside the ASMA (marked as two orange vectors in figure 1 2)) reported similar spatial mismatch between a series of CTM simulations (MOZART; TM3 and ACTM) and satellite observational datasets (HALOE; AIRS and GOSAT-TIR).One possible explanation is that frequent occurrence of cloud over the ASM center influences the satellite retrieval algorithm as well as moisture-dependent altitude sensitivity in the seasonal variation.It is also suggested that potential inaccurate CH 4 emissions (mainly from rice paddies) exist in the model (Xiong et al 2009).These discrepancies make relevant studies based on simulations or satellite retrievals questionable.Thus, in the next sections, we use several in-situ measurements to assess CH 4 enhancement in detail.

Evidence of CH 4 enhancement based on in-situ measurements 3.2.1. AirCore measurement at TP
In-situ AirCore measurements at Da Qaidam are used to explore the realistic vertical structure of CH 4 over the TP, the critical AMA region.Figure 2 shows one AirCore-measured CH 4 profile compared to the spatially and temporally nearest satellite-measured (L2) and simulated CH 4 profiles.We can see the enhancement between 170 hPa (∼13 km) and 70 hPa (∼19 km) with strong vertical variations (multi-layer structures) at Da Qaidam indicated by the AirCore measurement.This elevated CH 4 in the UT to LMS can be generally captured by the simulation with reasonable differences (less than ±40 ppbv).It is seen that AIRS significantly underestimates the CH 4 enhancement over the TP (∼100 ppbv for 150-100 hPa).The CTM simulation with assimilation more realistically represents CH 4 enhancement over the TP than AIRS retrieval, suggested by this case of cross-comparison.
Note that CH 4 enhancements between 12-19 km with strong vertical variations were persistently observed over the TP within the ASMA during 2 year summer campaigns (see figure  S4 in the supplementary material).It indicates that the realistic CH 4 enhancement over the TP is not a well-mixed thick layer spanning a few kilometers but the result of multiple thin layers with peaks in CH 4 .The formation of the layer structure might result from interaction between the surface emissions, fast convective uplift and remote transport by large-scale circulation.Figure 4 demonstrates that the relative differences between simulation and all the collected AirCore profiles range from −1% up to +3%.The CH 4 enhancement is underestimated (1%-3%) by the simulation in the UTLS, particularly around the tropopause (tropopause-50 hPa-tropopause + 100 hPa).This is attributable to the underestimation of CH 4 emission, insufficient convective transport and/or representation of eddy mixing.
The highly-varying vertical structure measured by AirCore cannot be captured by either the model with limited vertical resolution or the satellite retrievals using averaging kernels.In particular, this is since AIRS is most sensitive to the CH 4 layer at 50-250 hPa below the tropopause, i.e. 350-150 hPa in this case.AIRS retrievals of CH 4 possibly miss the strongest enhancement at 150 hPa-100 hPa in this case.However, we cannot quantify this impact on the AIRS representation for the overall seasonal pattern due to the lack of in-situ cases covering a larger space.

CH 4 sampled by IAGOS airlines over East Asia
For further validation of the spatial pattern of the CH 4 monsoon plume, we search for the aircraft measurements from the Observation Package (ObsPack) database (Schuldt et al 2023), which is able to validate the horizonal distribution of CH 4 in the UTLS.Here, we show one case on June 2015, measured by intercontinental Civil Aircraft for the Regular observation of the atmosphere Based on an Instrumented Container (CARIBIC) as part of the IAGOS infrastructure.This case is chosen because the air masses were sampled along the flight track over the middle latitudes of Eurasia, the TP and East China, crossing the edge of the ASMA in the early stages of the monsoon.
The comparison in figure 4 illustrates that the simulation provides a reasonable representation of CH 4 distribution in the UT.The elevated CH 4 over east Asia seen in the simulation (east of 90 • E, see figure 4(a)) is confirmed by the corresponding aircraft measurements.The CH 4 volumn mixing ratios (VMRs) agree well between the simulation and in-situ data, in particular within the ASMA (see the solid gray squares in figure 4(b)).However, the daily CH 4 map observed by AIRS clearly underestimates the CH 4 enhancement over east Asia (solid gray squares in figure 4(d)).Note that deep convections and clouds, that is demonstrated by outgoing long-wave radiation (OLR) shown as magenta contours, are highly associated with regions where satellite retrieval of CH 4 is missing (due to QC) or underestimated compared to both simulated map and the aircraft measurements, i.e. over southwestern China and the TP.This case indicates that the mismatch between satellite observations and simulations (as shown in figure 1 Figure 5.The contrasts in CH4 in the UT between the upstream (R1: blue squared region) and downstream (R2: orange squared region) along the westerly jet on the ASMA northern edge, demonstrated by simulation, airline in-situ measurements and AIRS.Panel (a) shows a map of selected regions, flight tracks, the seasonal-mean wind field and the ASMA location (GPH isolines at 200 hPa).The comparison between in-situ measurement and the nearest simulated CH4 is shown in panel (b), which is grouped into upstream (blue) and downstream (orange).The mean seasonal cycles of CH4 in the UT averaged over the two regions are shown in panels (c) and (d).The seasonal cycle in (c) is derived from in-situ measurements (square solid line) and correspondingly sampled CH4 simulations (dashed line).The seasonal cycle in (d) is simulated CH4 (dashed line) and AIRS-observed CH4 (triangle solid line) averaged over the two selected regions.and mentioned in several previous studies) might relate to the satellite retrieval algorithm being influenced by deep convections.However, it is not conclusive that deep convections in the ASM influence the estimate of CH 4 increase on the seasonal average.

CH 4 enhancement due to ASMA outflow
The air masses inside the ASMA were occasionally split off through the quasi-isentropic eddy shedding.These events commonly occur over two preferred regions, where the transport barrier is weak.The two regions are marked by two orange vectors in figure 1(c): one is westward eddy shedding, together with the subtropical westerly jet over west Pacific (Tomsche et al 2019); another one is eastward eddy shedding; together with the tropical easterly jet over northern Africa.(Honomichl and Pan 2020, Pan et al 2022).To further quantify and validate the CH 4 enhancement extending out of the ASM domain, here we quantify the CH 4 concentration differences between the air masses entering (blue vectors in figure 1(c)) and exiting (orange vectors in figure 1(c)) the ASM region, i.e. air masses influenced by the ASMA and not-yet-influenced by the ASMA, on the northern edge (westerly jet) and the southern edge (easterly jet), respectively.
Figure 5 shows the upstream and downstream regional differences in CH 4 on the northern edge of the ASMA (R1: upstream region; R2: downstream region) along the westerly jet.The aircraft measurements from the ObsPack dataset within the two regions confirm the fact that the CH 4 concentration over the west Pacific in the UT is increased remarkably (by 50-80 ppbv) compared to the observed air masses before flowing into the ASMA.The increase in CH 4 due to the ASMA quantitively agrees reasonably between the simulation and in-situ measurements (see figures 5(b) and (c)).However, the contrast in CH 4 concentration between R1 and R2 is oppositely represented by the AIRS satellite data (see figure 5(d), as well as figures 1(b) and (d)).This is consistent with the seasonal pattern shift shown in figure 1.It might relate to a fact proposed in a previous study (Xiong et al 2015) that a pressure-dependent bias near the UT and the spatiotemporal variations by AIRS include some artificial impacts from the retrieval sensitivity.
When comparing the aircraft measurements within R1 and R2 in figure 6, a contrast in CH 4 concentrations between the air masses flowing into (blue region, R1) and out of the ASMA (orange region, R2) along the tropical easterly jet also exists.Note that since only a few airlines were collected for region R2, the CH 4 enhancement might be insufficiently represented by those measurements.Despite the differences in climatological means, the simulation and satellite show a similar seasonal variability of CH 4 in R2 (see

Discussion and conclusions
We presented evidence of CH 4 enhancement associated with the ASM based on model simulation, satellite observations and in-situ measurements.Quantitively, the multiple-year averaged CH 4 enhancement relative to the zonal means by AIRS observation and model simulation is up to 3%, and the monsoon-region averaged seasonal increase is up to 6% relative to the pre-monsoon season.In particular, a mismatch of the UT enhancement pattern between simulated and AIRS-observed CH 4 is identified, i.e. a southwestward shift of the enhancement by AIRS and a remarkably low bias in the eastern part of the anticyclone compared with simulation, which was also proposed in a series of previous comparisons between several operational infrared sounders and CTMs (Park et al 2004, Xiong et al 2009, Belikov et al 2022).
The comparison among the CH 4 profiles measured in the TP by AirCore, AIRS satellite and model simulation indicates that enhancements in the UT (∼100 ppbv in our case) are underestimated by AIRS.
The model results demonstrate a reasonable enhancement in the UT, although the remarkable vertical variabilities shown by the AirCore-measured CH 4 profiles are largely smoothed.The aircraft-sampled CH 4 concentration crossing the edge of the ASMA also shows reasonable agreement with the simulation.The satellite CH 4 retrievals, meanwhile, underestimate the CH 4 concentrations, in particular, when clouds are present.
We further quantify the ASM influence extending out of the main anticyclone region in the UT.The contrasts in CH 4 concentrations between the 'entrance' (flow-in) and 'exit' (outflow) of the anticyclone at the southern and northern side of the ASMA, represented by our simulation and aircraft insitu measurements, confirmed our theoretical expectations, i.e. the CH 4 concentration at the 'exit' is clearly elevated compared to that at the 'entrance' , whereas the satellite observations show opposite results.
The CH 4 enhancement in the UT can be underestimated by AIRS retrievals, in particular in the north and east of the ASM region.This is possibly related to the retrieval process influenced by convective cloud, as illustrated by the case study shown in figure 4, which was also suggested by previous studies (Susskind et al 2003, Xiong et al 2009, 2010).It is also possibly due to the vertical sensitivity (retrieval with averaging kernels) interacting with the complex vertical structure in reality, as shown in figure 2.
Our evidence confirms the efficient transport to the UT by the ASM system and indicates that the CTM simulation with optimized assimilation emission flux reasonably represents CH 4 enhancement in the UT, in terms of the seasonal mean horizonal distribution inside and around the ASMA, which is largely controlled by the large-scale circulation.It is challenging for the model to reproduce small-scale horizonal and vertical variability.Therefore, in-situ measurements play an important role in validating the simulations and anchoring the assimilation emission flux.
(c)).Easterly flow on the southern edge of the ASMA transports ASM air toward the equator, which is possibly further being merged into a tropical pipe or crossing the equator to the southern hemisphere (SH) (Ploeger et al 2017, Yan et al 2019, Belikov et al 2022).The CH 4 plume northeastward extension results from the frequent northeastward eddy shedding over the northwestern Pacific (Honomichl and Pan 2020, Pan et al 2022), which strongly contributes to the transport pathway toward the mid-latitude LMS.We noticed that the northeastern CH 4 enhancement is clearly represented by the simulation but not by AIRS, whereas AIRS measures higher CH 4 southwestern enhancement than the simulation.The discrepancies in enhancement magnitude and pattern clearly exist.Previous studies (e.g. Park et al 2004 (figure 3), Xiong et al 2008 (figure 1), Belikov et al 2022 (figure

Figure 1 .
Figure 1.The mean CH4 distribution for the monsoon season (JJAS) from 2015 to 2020 averaged between 300-200 hPa (a) and (b), and between 200-100 hPa (c) and (d) based on Geos-Chem simulation (a) and (c), and AIRS observation (b) and (d).The zonal mean values are subtracted (see the zonal mean values in figure S1).The yellow contours show the AMA center (demonstrated by selected GPH isolines) at 200 hPa and 100 hPa.The magenta contours indicate thermal tropopause.

Figure 2 .
Figure 2. (a) Cross-comparison among GEOS-Chem (GC) simulated (black), AIRS satellite observed (blue) and AirCore-measured CH4 profiles (red) close to the site Da Qaidam (95.3E/37.8N,marked by a triangle) on 8 August 2020.The dots represent the means; lines show the standard deviation (±σ) and the shades indicate the ranges spanning from the minimum to the maximum.(b) The spatial matching among the three datasets.The grid points from simulation inside the inner circle (r = 500 km) within 3 h from the conducting of in-situ measurement (color circles) and the satellite footprints inside the outer circle (r = 1000 km) within 6 h from the conducting of in-situ measurement (colored squares) are considered.Due to the large tracer gradience across the edge of AMA, only the points within a similar GPH range (14.4-14.46km at 150 hPa) with the location of Da Qaidam are selected for comparison and are shown in the left panel.

Figure 3 .
Figure 3. GC simulated CH4 (blue) compared to AirCore measurements (red) conducted in August 2019 and 2020 at Da Qaidam, compared to the simulated zonal mean (gray).Four CH4 profiles observed by AirCore within the AMA are used in the comparison.The left panel shows relative differences to the zonal mean profile.The dots represent the means; lines show the standard deviation (±σ), and the shades (if shown) indicate the ranges spanning from the minimum to the maximum.

Figure 4 .
Figure 4.The CH4 measured by IAGOS aircraft on 18 June 2015 (colored squares) compared to the daily mean XCH4 (column averaged mixing ratios from 300 hPa to 200 hPa) from GC simulation (color shades in (a)) and AIRS data (color shades in (c)).The comparisons of in-situ measurements to the corresponding simulation and satellite data are shown in panels (c) and (d), where the data points inside(/outside) the ASMA are shown as white(/gray) squares.The location of the ASMA is shown as black contours (GPH isolines at 150 hPa).The wind field is shown as the arrows.The magenta contours of low outgoing long-wave radiation (200 W m −2 and 170 W m −2 OLR isolines) are shown as a proxy of deep convections.

Figure 6 .
Figure 6.The same as figure 5 but for the contrast between two regions along the easterly jet on the southern edge of the ASMA: blue squared region (R1) and orange squared region (R2).The panel (a) shows map of selected regions, flight tracks, seasonal-mean wind field and the ASMA location (GPH isolines at 200hPa).The comparison between in-situ measurement and nearest simulated CH4 is shown panel (b), which is grouped into R1(blue) and R2 (orange).The mean seasonal cycles of CH4 in the UT averaged over the two regions are shown in panel (c) and (d).The seasonal cycle in (c) is derived from in-situ measurements (square solid line) and correspondingly sampled CH4 simulations (dash line).The seasonal cycle in (d) is simulated CH4 (dashed line) and AIRS-observed CH4 (triangle solid line) averaged over the two selected regions.