Photochemistry of Methane and Ethane in the Martian Atmosphere

We develop an existing 1‐D photochemistry model to include a comprehensive description of organic chemistry on Mars that includes the oxidation products of methane (CH4) and ethane (C2H6),a longer‐chain hydrocarbon that can be used to differentiate between abiotic and biotic surface releases of CH4. We find that CH4 is most volatile between 20 and 50 km during Mars' northern summer, where the local atmospheric CH4 lifetime lowers to 25–60 years. We study atmospheric formaldehyde (HCHO) and formic acid (HCOOH), as the two common oxidation products of CH4 and C2H6, and acetaldehyde (CH3CHO) and acetic acid (CH3COOH) as unique products of C2H6. We focus our analysis of these gases at Mars' aphelion and perihelion at latitudes between ‐30° and 30°, altitudes from the surface to 70 km, and from a homogeonous initial condition of 50 pptv of CH4 and C2H6. From this initial condition, CH4 produces HCHO in a latitude‐independent layered structure centered at 20–30 km at aphelion with column‐averaged mixing ratios of 10−4 pptv, and oxidation of C2H6 produces HCHO at 10−2 pptv. Formic acid has an atmospheric lifetime spanning 1–10 sols below 10 km that shows little temporal or zonal variability and is produced in comparable abundances (10−5 pptv) by the oxidation of C2H6 and CH4. We also find that oxidation of 50 pptv of C2H6 results in 10−3 pptv of CH3CHO and 10−4 pptv of CH3COOH. Subsequent UV photolysis of this CH3CHO results in 10−4 pptv of atmospheric CH4, potentially representing a new atmospheric source of Martian CH4.


Introduction
There is considerable debate in the community about the validity and robustness of detections of atmospheric methane (CH 4 ) on Mars. Detections and nondetections of atmospheric CH 4 have been reported using data collected by satellites (Formisano et al., 2004;Fonti & Marzo, 2010;Geminale et al., 2008Geminale et al., , 2011Giuranna et al., 2019), Earth-based telescopes (Krasnopolsky, 2007(Krasnopolsky, , 2011(Krasnopolsky, , 2012Krasnopolsky et al., 1997Krasnopolsky et al., , 2004Mumma et al., 2009;Villanueva et al., 2013), and in situ instruments (Moores et al., 2019;Webster et al., 2015Webster et al., , 2018. This debate highlights the difficulty of measuring atmospheric CH 4 on Mars and the gaps in our current understanding of the production and loss terms that determine atmospheric CH 4 on Mars. We explore how the presence of CH 4 could be determined by its oxidation products and in the process discuss how abiotic CH 4 could be potentially produced by organic chemistry. Detections of atmospheric CH 4 at the Gale crater by the NASA Curiosity Rover were observed episodically rising from mean volume mixing ratios (VMRs) of 0.69 ± 0.25 parts per billion (ppbv) to 7.20 ± 2.10 ppbv across a 60-sol period (Webster et al., 2015). The diurnal variations of CH 4 observations via Curiosity have recently been substantiated by the ExoMars Trace Gas Orbiter's Nadir and Occultation for Mars Discovery (NOMAD) spectrometer and Atmospheric Chemistry Suite (ACS) (Korablev et  This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 2019). These recent measurements by NOMAD and ACS constrain the upper limit of CH 4 in the Martian atmosphere to 0.05 ppbv.
Ground-based telescopes reported a signal of a plume that contained approximately 19,000 tonnes of CH 4 (Mumma et al., 2009) near the Syrtis Major region during Mars' northern hemisphere's midsummer, consistent with an estimated 0.60 kg m −2 s −1 seasonal point source of organic compounds. Other searches for CH 4 on Mars have conversely failed to detect the compound (Webster et al., 2013), including the Trace Gas Orbiter's year and a half long search . This highlights the possibility of methane being significantly more temporally variable than current models predict (Lefèvre & Forget, 2009). This inconsistency with regard to CH 4 's detection has also ignited some polemic against the existance of the gas on Mars (Zahnle et al., 2011). Recent analysis of data from the Planetary Fourier Spectrometer (PFS) aboard the Mars Express orbiter identified the presence of CH 4 in the Martian atmosphere that was confirmed via independent observations by Giuranna et al. (2019). These results appear to confirm Curiosity's detection of CH 4 at the Gale Crater. The team retrieved column integrated VMRs of 15.5 ± 2.50 ppbv above the Gale Crater only 1 Martian day (sol) after the Curiosity rover's measurement of a 5.78 ± 2.27 ppbv (Webster et al., 2015).
To further study these possible emissions of CH 4 and to also provide a more detailed series of observations of the vertical structure and composition of Mars' atmosphere, the ExoMars mission program was established by the European Space Agency (ESA) (Vago et al., 2015) and the Russian Roscosmos State Corporation for Space Activities. The first mission conducted within this program was the launch of the Trace Gas Orbiter (TGO) in 2016 that included two suites of spectrometers, the Nadir and Occultation for MArs Discovery (NOMAD) spectrometer (Vandaele et al., 2018) and the Atmospheric Chemistry Suite (ACS) (Korablev et al., 2017). The TGO underwent 11 months of aerobraking to reduce its orbital speed and altitude, eventually establishing an approximately circular orbit of altitude roughly 400 km allowing scientific observations to start in April 2018 . Over the first year, no successful observations of CH 4 were reported by the TGO instrumentation . We developed our investigation bearing in mind the capabilities of TGO instruments, especially the solar occultation channels of NOMAD (NOMAD-SO) and ACS (ACS-MIR). The NOMAD-SO instrument has been designed to be sensitive to CH 4 abundances as low as 0.025 ppbv when observed in solar occultation mode (Robert et al., 2016). This expected detection limit was compared to the experimental values in . The TGO instruments improved the experimental upper limit of previous instrumentation to reach a limit of roughly 0.05 ppbv for methane. A few profiles reported in Korablev et al. (2019) and obtained via ACS-MIR, measured in clear northern conditions, were able to achieve the most precise detection limits of 0.012 ppbv down to an altitude of roughly 3 km.
Previous studies using 1-D photochemical models (Summers et al., 2002;Wong et al., 2003) calculate the photochemical lifetime of CH 4 to be roughly 300 years below 70 km. In the absence of a strong surface loss process, surface emissions of CH 4 will then become homogeneously distributed across the planet after being introduced into the atmosphere. This inconsistency between models and data means that the available atmospheric data are misinterpreted, and/or there is a loss mechanism that we do no not currently consider in models (Krasnopolsky, 2006). Previous calculations using a global 3-D general circulation model have determined that to reconcile models and data we need a CH 4 loss process that is up to 600 times faster than any known process (Lefèvre & Forget, 2009). That additional sink would lower the atmospheric lifetime of CH 4 from centuries to less than 200 days. With the advent of the TGO, another approach we can take is to analyze observed spatial and temporal variations of the oxidation products of atmospheric CH 4 .
In this study, we describe the development of the 1-D photochemistry submodule from the LMD-UK General Circulation Model (GCM) Lewis et al., 1999) to include organic chemistry and run it as an independent model to study photochemistry on Mars. We report results from a series of numerical experiments that describe how the presence of CH 4 affects photochemistry on Mars. We also report photochemical results when we replace CH 4 with ethane (C 2 H 6 ) to show this longer-chain hydrocarbon produced richer atmospheric chemistry and allows us to consider an abiotic source of atmospheric CH 4 from the oxidation of acetaldehyde. In the next section we describe our developed 1-D model of Mars photochemistry. In section 3 we report results from our numerical experiments. We conclude the paper in section 4 in which we discuss the implications of our results for broadly understanding atmospheric chemistry on Mars but in particular the implications for detecting the presence of CH 4 and C 2 H 6 on Mars using their oxidation products.

1-D Photochemical Model
We use the 1-D photochemistry submodule from the parent 3-D LMD-UK Mars General Circulation Model as the basis for a standalone 1-D model that includes an improved treatment of atmospheric organic photochemistry.
We use this standalone 1-D model to describe time-dependent vertical distributions of trace gases from the surface to an altitude of approximately 70 km, described by 25 vertical layers with a resolution of under 0.5 km below 2 km increasing to a resolution of 10 km above an altitude of 20 km, where 3-D macroscopic processes begin to dominate. We divide a Mars sol into 48 time steps (Δt = 1,800 s), allowing us to describe diurnal variations of trace gases by calculating time-dependent changes in solar zenith angle, taking into consideration changes in solar longitude and axial tilt. To decrease the stiffness of the discretized equations used to compute the photochemical rates of change, determined by a prescribed chemical mechanism described below, we use a chemistry sub-timestep of Δt c = 600 s.
Vertical tracer transport between model layers is described by the classical diffusion equation (Mellor & Yamada, 1982). For details on the vertical diffusion and turbulent mixing routines we refer the reader to Forget et al. (1999). We use a radiative transfer scheme (Madeleine et al., 2011) that uses opacity values from the Mars Climate Database v5.3. We describe the condensation and sublimation of carbon dioxide, (Forget et al., 1998), water ice (Navarro et al., 2014), and hydrogen peroxide (H 2 O 2 ), and an implicit chemistry solver computes production and loss rates from photochemical reactions.
We build on the chemical mechanism by including 41 new organic compounds, guided by previous studies (Summers et al., 2002;Wong et al., 2003) and also the theoretical measurement capabilities of the NOMAD instrument (Robert et al., 2016), to improve understanding of Mars' atmospheric chemistry. We include CH 4 , C 2 H 6 , and its photochemical products. The extended chemical mechanism represents an additional 106 chemical reactions (B3, B4, B5) and 29 photolysis reactions (B6). A complete list of trace gas species within the 1-D model is provided in Table 1. We present the first analysis of C 2 H 6 oxidation products in the Martian atmosphere. We take the organic chemistry rate coefficients from the CAABA/MECCA v4.0 atmospheric box model (Sander et al., 2019), which is used to model organic chemistry within Earth's atmosphere. We have modified the mechanism to include HCO and C 2 H 5 radicals. The CAABA/MECCA v4.0 model neglects these radicals as products and instead includes the products of the radicals with molecular oxygen, O 2 . This approximation is sufficient for Earth, where O 2 is present at 21% mass fraction, but on Mars it is present only at a mass fraction of 10 −3 . Including these radicals allows us to improve the description of organic chemistry. All three-body reaction rate coefficients in the submodule are multiplied by a factor of 2.5, following Nair et al. (1994), to account for the increase in efficiency that CO 2 displays when used as a bath gas in comparison to N 2 or dry air (Kaufman & Kelso, 1967), commonly used in laboratories for the calculation of these coefficients.
To improve the accuracy of the original chemistry routine and to reduce the computational expediency of our chemistry calculation we use a precalculated look-up table to interpolate photolytic frequencies. We calculated these photolysis loss rates using the Tropospheric Ultraviolet and Visible (TUV) Radiation Model (Madronich et al., 2002) that has been adapted for use on Mars (Lefèvre et al., 2004) and using routines to interpolate as function of atmospheric temperature, solar zenith angle, O 3 column density, the total atmospheric column density, the Sun-Mars distance, and the dust opacity. The photolytic reaction j HOCH2OOH of B6 requires the consideration of the abundance of O 2 , as the photolytic product of HOCH 2 OOH is the highly reactive HOCH 2 O radical. HOCH 2 O proceeds to react with molecular oxygen to form HCOOH and OH. To limit the number of compounds that the steady-state approximation has to be applied to, the 1-D model multiplies the photolytic frequency extracted from the TUV look-up table by a temperature independent factor of 3.50×10 −14 (Veyret et al., 1982) and the number density of O 2 at the respective layer.
To improve the accuracy of the original chemistry routine and to reduce the stiffness of the chemistry calculations, we decreased the chemical timestep to 600 seconds. This has the additional benefit of allowing the Semi-Implicit Backward Euler Method (SIBEM) to be used for a larger number of species across timestep Δt c while conserving mass of the studied tracers. For species with photochemical lifetimes shorter than 100 seconds we consider a family of species, for example, odd-hydrogen (HO x ) and odd-oxygen (O x ) families, which have a collective lifetime longer than the timestep. We handle these compounds, H, OH, HO 2 , and O( 1 D), O( 3 P), O 3 , respectively, via the assumption of photochemical equilibrium (Rodrigo et al., 1990). The routine calculates dimensionless partition functions of the loss rates of H/HO 2 , OH/HO 2 , and O( 3 P)/O 3 (s −1 /s −1 ) and sums the abundances of each family, HO x and O x , which possess atmospheric lifetimes greater than that of the chemistry timestep, allowing them to be computed via the SIBEM equation. The partition functions are then applied to compute each individual compounds respective abundance.

Time-Dependent Meteorological Boundary Conditions
To drive the 1-D photochemistry model, we use time-dependent lateral atmospheric boundary conditions of temperature, wind, water vapor volume mixing ratios, and surface pressure from the Mars Climate Database v5.3 (MCD) (Millour et al., 2017). We interpolate the 3-D meteorological fields from the MCD data set, taking into account latitude, solar longitude, and local time.
All interpolated values from the MCD data set are longitudinal means, calculated independently by ourselves, acknowledging that there are only small longitudinal variations of meteorological parameters that are due primarily to topographical features. We take advantage of this so that the horizonal footprint of our model is representative of a zonal band of 3.75°in latitude, which is the latitudinal resolution of the MCD dataset. Zonal and meridional transport is not accounted for by the 1-D model. We use an interpolation routine for trace gases that have atmospheric chemistry lifetimes greater than the e-folding residence times associated with meridional advection of trace gases out of the zonal band, which is approximately 0.5-2 sols. We use this approach to drive the model with vertical profiles of CO 2 , CO, O 2 , H 2 , and water vapor, ensuring the photochemical environment is consistent with the meteorological fields from the MCD, allowing the 1-D model to be accurately representative of the temporally and spatially variable oxidising environment. Seasonal water vapor profiles from MCD are especially important for the 1-D photochemical calculations as the photolysis of Martian H 2 O is the source of the highly reactive odd-hydrogen species, HO x = H + OH + HO 2 , which help drive the oxidation of organic species and O 3 chemistry (Lefèvre et al., 2004).
The overarching purpose of using these boundary conditions is that we can describe detailed 1-D atmospheric chemistry, including diurnal and seasonal changes associated with Mars' atmosphere, without the computational overhead of solving the 3-D dynamical equations. In particular, the boundary conditions help to maintain realistic values and variations of wind profiles that underpin vertical diffusion calculations that correspond to calculating vertical transport. For all of our calculations, we use a prescribed dust scenario produced from the assimilation of observations of the dust optical depth made by the Mars Global Surveyor's Thermal Emission Spectrometer (Montabone et al., 2015) during Mars Year (MY) 24. This scenario is regarded as a "best guess" of the mean annual dust variability experienced on Mars without the presence of global or significant regional dust storms. The 1-D photochemical model is capable of operating in dust-storm scenarios using data from the MCD v5.3, but these calculations are beyond the scope of the current study.
We limit our calculations to latitudes less than 30°, where most recent observations of CH 4 on Mars have been reported (Mumma et al., 2009;Webster et al., 2015Webster et al., , 2018. At latitudes higher than 40°in both hemispheres, the MCD boundary conditions also allow us to describe the large-scale seasonal changes in atmospheric temperature that result from condensation, deposition, and sublimation of CO 2 , which produce significant changes in surface pressure. However, at these latitudes meridional wind profiles above the planetary boundary layer can become large enough to lower the tracer e-folding timescales to values too small for the 1-D model to neglect while still producing reliable model results. At latitudes less than 30°, the 1-D model in this work can be used to produce time-dependent tracer profiles for periods limited to 2 sols.

Definition of Numerical Experiments
We use the 1-D model to determine vertical profiles of organic compounds that evolve from a hypothetically fixed and vertically uniform 50 pptv profile of CH 4 . We have chosen this value because it is consistent with the upper limit of Martian CH 4 determined by data collected from the ACS/NOMAD instruments aboard the TGO .
For this calculation, we initialize the model 10 sols prior to the point where solar longitude L S = 0°with initial meteorological conditions and tracer profiles of CO 2 , CO, O 2 , water vapor, and H 2 from the previously detailed MCD v5.3 look-up table and 10 ppbv of CH 4 distributed evenly across all 25 atmospheric layers ranging from the surface to approximately 70 km. We then run the 1-D over one Mars year, 668 sols, following a 10-sol spin-up period, taking into account diurnal and seasonal (orbital) variations in solar zenith angle and corresponding changes in solar flux and photolytic frequencies. We report calculations at latitudes of 30°N, 2.5°N, and 30°S to investigate possible spatial variations in organic product profiles.
Following this investigation into the annual variations of the products produced by the steady and homogeneous CH 4 background, we use the 1-D model to investigate time-dependent photochemical processes that result from oxidation of CH 4 and C 2 H 6 . For these experiments, we initialize the 1-D model at the required latitude and solar longitude at 00:00 local time (LT) with 0 pptv CH 4 and other non-methane organic compounds. We use a five-day spin-up period at a constant solar longitude when we use trace gas profiles of CO 2 , CO, O 2 , H 2 , and water vapor from the MCD v5.3 (section 2.2) and diurnally varying atmospheric parameters. This spin-up enables the HO x and O x chemistry to partition to the solar-longitude environment. After five sols, we add a vertically uniform 50 pptv profile of CH 4 at 00:00 LT. We then let the model run freely for one further sol, and present the photochemical products at a local time of 18:00 LT when solar occultation measurements are typically collected. We adopt a similar approach to examine the photochemistry of C 2 H 6 .

Results
We use the 1-D model described in the previous section to understand the seasonal chemical composition of the Martian atmosphere as a function of latitude that corresponds to a fixed, uniform vertical distribution of atmospheric CH 4 . We present a similar set of calculations that correspond to a uniform vertical distribution of atmospheric C 2 H 6 . ( 1 D)) and photolysis. The loss rates for atmospheric CH 4 vary as a function of solar longitude. There are a number of high-yield oxidation products that are produced rapidly, potentially allowing us to colocate elevated values with CH 4 emissions and that are observable from orbiting instruments. In this paper we will restrict our analysis to the production of formaldehyde (HCHO) and formic acid (HCOOH). Figure 2 shows an illustrative vertical distribution of CH 4 loss factors (s −1 ) at the solar longitudes of 71°and 271°, times at which Mars' orbit is furthest from and closest to the Sun, respectively. Loss factors, units of s −1 , are distinguished from loss rates, units of molec cm −3 s −1 , as loss factors provide a better insight into the reactivity of the surrounding environment. These values after being multiplied by CH 4 number densities equate to the rate of photochemical loss of CH 4 (molec cm −3 s −1 ), that is, the loss rate.

Methane Oxidation in the Atmosphere of Mars
From the Martian surface to the top of the Martian hygropause, approximately 10 km at L S 71°and 45 km at L S 251°, the dominant loss process for CH 4 is from oxidation by OH. Consequently, at these altitudes the rate of OH production, driven by the photolysis of water vapor, are larger than outside this altitude range. Above these altitudes, the abundance of O( 1 D) increases, reflecting the drop in its loss from reaction with water vapor. Between the top of the hygropause and roughly 60 km, O( 1 D) becomes the dominant loss of atmospheric CH 4 . There are three oxidation channels associated with this reaction (B4), the most efficient producing CH 3 and OH. Photolysis of CH 4 becomes significant only at altitudes higher than 60 km. The decreasing abundance of O( 1 D) atoms in the upper atmosphere results in a small vertical region, 50-60 km, where CH 4 loss rates decline before photolysis becomes important at higher altitudes. Below 60 km, the effect of UV radiation on CH 4 is negligible, and photochemical contributions made by methyl radical isotopologues can be disregarded. This highlights that the primary atmospheric sinks of CH 4 change seasonally, with O ( 1 D) dominating in the aphelion seasons and OH dominating in the perihelion. The stronger reaction rates of O( 1 D) with CH 4 can thus be expected to produce greater abundances of potentially observable organic species within the aphelion seasons of Mars.
The local atmospheric lifetime of CH 4 , incorporating all chemical loss terms (Figure 2), varies as a function of latitude and solar longitude. Differences between latitudes are determined by the solar radiation being received, which is a function of solar longitude and Mars' obliquity. Local atmospheric lifetimes range from 25 to 1,700 years (Appendix A0.1). These values, with a significant and localized surface loss process, would generate a large, slowly varying background value for CH 4 that would be difficult to attribute to individual surface sources. Local atmospheric lifetimes reach a minimum of 25-200 years between the top of the hygropause and approximately 50 km due to the larger abundance of the O( 1 D) atom. Below the hygropause where OH is the dominant sink, local atmospheric lifetimes vary between 200 and 425 years. The longest local lifetimes of 1,000-1,700 years lie between 50 and 60 km during winter (L S = 270-360°), as described above, where the OH and O( 1 D) loss processes decline and before photolysis dominates above 60 km.
These local atmospheric photochemical lifetimes are significantly longer than the corresponding atmospheric transport timescales across all altitudes and solar longitudes. We study the variance of these local photochemical lifetimes with solar longitude, altitude, and Mars latitude to understand where are the most photochemically active regions for CH 4 (and below for C 2 H 6 ) and consequently where we expect the largest production rates for their oxidation products. Calculating the vertically integrated photochemical lifetime of CH 4 with the 1-D model (Appendix A0.1) yields timescales that correspond well to those calculated in previous studies (Krasnopolsky et al., 2004;Lefèvre & Forget, 2009;Summers et al., 2002), with the 1-D model producing values ranging between 250 and 550 terrestrial years across the latitude ranges studied here of 30°s outh to 30°north. Figure 3 shows vertical profiles of the volume mixing ratios of organic compounds, radicals, peroxy-radicals, and long-lived species from left to right that result from the oxidation of a fixed, uniformly distribution 50 pptv of atmospheric CH 4 . For this calculation we initialize the 1-D model with 0 pptv of CH 4 across all vertical layers at L S = 71°and L S = 251°, 00:00 LT, at a latitude of 2.5°, the latitude at which a CH 4 plume has been previously observed (Mumma et al., 2009). We spin-up for five sols using detailed MCD v5.3 tracer profiles (section 2). We then insert a fixed, uniform profile of 50 pptv of CH 4 into the model and run forward for one sol. We sample the model at 18:00 LT on sol 6, allowing sufficient time (18 hours) for products to be produced but a short enough time to ensure these products are not advected out of the column. This allows us to study the resulting CH 4 oxidation product concentrations. We show values for CH 4 , HCHO, HCOOH,
Production rates of methanol are small in our 1-D model, which is driven by a realistic 50 pptv of CH 4 . Previous studies that have used a much larger 100 ppmv profile of CH 4 (e.g., Wong et al., 2003) have still struggled to produce Martian methanol above pptv values. CH 3 OH has prominent features at IR wavelengths, making it a valuable observed species to help constrain our understanding of organic chemistry.
Our 1-D model calculates CH 3 OH column average mixing ratios of 2.09×10 −19 and 1.01×10 −18 molecule/molecule at L S = 71°and 251°, respectively. These small quantities are due to the two main production terms for CH 3 OH (Table B4) involving CH 3 O 2 and CH 3 radicals that are not produced in large quantities by 50 pptv of CH 4 . Consequently, we will only discuss CH 3 OH in context of HCHO and HCOOH.

Production of Formaldehyde from CH 4 Oxidation
HCHO is a high-yield oxidation product of CH 4 on Mars that is observable by the NOMAD and ACS spectrometers aboard the Trace Gas Orbiter. Figure 4 shows vertical distributions of HCHO volume mixing ratios, corresponding to a fixed vertical profile of 50 pptv CH 4 , at 18:00 LT over all solar longitudes to determine when we might expect HCHO to be observable during one Martian year. HCHO production peaks during spring months when there is sufficient production of O( 1 D) in the middle atmosphere to oxidize CH 4 but not water vapor in the lower atmosphere. We find only small variations in the HCHO produced across the three latitudes, where HCHO consistently remains below 10 −3 pptv. This lies well below the expected detection limit of 30-40 pptv for the TGO NOMAD instrument (Robert et al., 2016) and of 0.17-1.70 ppbv for the TGO ACS instrument (Korablev et al., 2017). Figure 5 shows individual and net production and loss rates of HCHO as a function of altitude at 06:30 LT, L S = 71°, and latitude 2.5°N. We find net production peaks at 06:30 LT as the Sun begins to rise above the horizon. Figure 5a shows two distinct regions where HCHO production peaks. Below the hygropause, at roughly 15 km at this L S , reaction between water vapor and excited atomic oxygen result in HCHO production dominated by the reaction of CH 3 O 2 and HO 2 . Above the hygropause at 15 km, the HCHO production is from atomic oxygen reacting with methyl and methoxy radicals. Figure 5b show the HCHO loss rates. We find that HCHO is lost rapidly by photolysis through the atmospheric column. Below the hygropause, HCHO is lost by reaction with HO 2 , which produced the HOCH 2 O 2 peroxy radical. Above the hygropause, the dominant HCHO sink is reaction with atomic oxygen. At solar zenith angles between 60 and 85°, abundances of O( 3 P), CH 3 , and CH 3 O radicals are in large enough at the top of the hygropause to overcome the loss of HCHO from photolysis. At lower solar zenith angles, the rate of formaldehyde photolysis is too great to allow significant net production.
The resulting atmospheric lifetime of HCHO for all Martian seasons is 2-6 hours at latitudes less than 30°(Appendix A0.2), consistent with previously published results (Wong et al., 2003). As a result, a detection of HCHO below 70 km in the Martian atmosphere will require a strong active release of CH 4 in the local vicinity. Our calculations suggest that HCHO will be most likely detected at mid-altitudes (15-30 km) across the tropics during Mars northern spring and summer months. This is due to the lower abundance of water vapor and the increased levels of odd-oxygen species available for reactions with organic radicals and also coincides with the altitude regions where the TGO instruments are expected to display the greatest level of sensitivity (Korablev et al., 2017).

Journal of Geophysical Research: Planets
Using 1-D steady-state model driven by a profile of 100 ppmv CH 4 , (Wong et al., 2003) report an increase in HCHO above 50 km. Our 1-D calculations are time-dependent, driven by a more realistic CH 4 value (published since Wong et al., 2003) and by interpolated atmospheric parameters that include optical opacity and long-lived inorganic tracers from a 3-D data set. As such, we believe our approach provides a more comprehensive and realistic modeling environment to describe atmospheric photochemistry on Mars. The photolysis loss rates for HCHO in our model at latitude 2.5°are comparable to those used by (Wong et al., 2003). Differences between our calculations and those reported by (Wong et al., 2003) are mainly due to smaller quantities of radicals produced by a much lower assumed value of atmospheric CH 4 and our consideration of longitudinal mean variations of atmospheric parameters instead of using global mean parameters.

Production of Formic Acid from CH 4 Oxidation
The oxidation of HCHO via HO 2 radicals leads to pathways that produce HCOOH (Figure 1). HCHO reacts with HO 2 to produce the HOCH 2 O 2 peroxy radical, which can either decay into its original reactants or react with HO 2 to produce HCOOH directly or to produce hydroxymethyl hydroperoxide that subsequently reacts with OH to produce formic acid or HOCH 2 O 2 , with a branching ratio, k HOCH2O2 /k HCOOH , of 0.233 at 298 K. HOCH 2 OOH can also photolyse under UV radiation to produce HOCH 2 O and OH. The HOCH 2 O radical proceeds to react with molecular oxygen to produce formic acid and HO 2 . The only sink of HCOOH in the Martian atmosphere is via oxidation by OH. This reaction is slow and limits loss of HCOOH to regions below the point of the hygropause. In regions where OH concentrations are at a maximum, typically during  Figure 6 shows the net and individual production and loss rates for HCOOH for L S = 71°, latitude 2.5°N, and LT = 18:00 due to a fixed, uniformly distribution of 50 pptv CH 4 . The largest production rate of HCOOH is from HOCH 2 O 2 reacting with HO 2 and from the photolysis of HOCH 2 OOH. The loss of HCOOH from OH oxidation is slow compared to the production rates. Our calculations therefore suggest that a detection of HCOOH would be related to the photochemical loss of HCHO. The resulting atmospheric lifetime of HCOOH is temporally and spatially variant, with lows of 2-5 sols at the tip of the hygropause, increasing to 75-100 sols closer to the surface where the abundances of OH fall.

Ethane Oxidation
Here, we consider the impact of C 2 H 6 on Mars' photochemistry. The impetus for these calculations is that on Earth, emissions of CH 4 are accompanied by emissions of higher-chain hydrocarbon such as C 2 H 6 (Guenther et al., 2000;Horita & Berndt, 1999). C 2 H 6 is also listed as an observable compound through the NOMAD-SO and ACS-MIR instrument, with detection limits between 0.02 and 0.03 ppbv (Vandaele et al., 2018) and 0.06-6 ppbv (Korablev et al., 2017). Current knowledge puts an upper limit of 0.20 ppbv on C 2 H 6 (Krasnopolsky, 2012;Villanueva et al., 2013). Figure 7 shows that the photochemistry for C 2 H 6 is more complicated than for CH 4 (B4) but follows the same general routes. It is oxidized by OH and O( 1 D) that initiates a series of chemical reactions that result in high yield products of HCHO and HCOOH among other compounds. One of the initial C 2 H 6 oxidation product is the C 2 H 5 radical, which is not described explicitly by the CAABA mechanism. We have used independent sources for reaction rates that involve the C 2 H 5 radical (B5). Figure 8 shows the photochemical loss factors for C 2 H 6 in the Martian atmosphere at noon, L S 71°and 251°a nd at latitude 2.5°N. Our photolysis calculations, based on values from the TUV model (Madronich et al., 2002) and adjusted for the Sun-Mars distance, suggest this loss process is insignificant below 70 km. Similar to CH 4 , below the Martian hygropause OH is the dominant loss process for C 2 H 6 , and above the hygropause O( 3 P) is the dominant loss process for C 2 H 6 , with the C 2 H 5 O 2 peroxy radical being a common oxidation product that results in a cascade of photochemical reactions, as described in Figure 7. The resulting

Journal of Geophysical Research: Planets
TAYSUM AND PALMER photochemical lifetime of C 2 H 6 is typically between 1 and 3.5 years below altitudes of 5 km and much shorter (50-450 sols) above the hygropause. (A1). Figure 9 shows the vertical profiles of organic compounds produced by the oxidation of 50 pptv of C 2 H 6 after one sol at L S 71°and 251°, latitude 2.5°sampled at 18:00 LT, to allow comparisons against the methane investigation of Figure 3. The introduction of longer chain peroxy-radicals via the C 2 H 5 O 2 radical increases the richness of photochemical products that are produced from CH 4 oxidation. In particular, the oxidation of C 2 H 6 results in CH 3 CHO column averaged mixing ratios of 7.24×10 −16 molecule/molecule at aphelion conditions, elevating to 3.18×10 −15 molecule/molecule at perihelion conditions. UV photolysis of CH 3 CHO is a source of CH 4 and carbon monoxide. We find that UV photolysis of 50 pptv homogeneous column of C 2 H 6 produces a CH 4 column averaged mixing ratio of 1.87×10 −17 molecule/molecule at aphelion and 3.97×10 −16 molecule/molecule at perihelion. The ratios of CH 4 /C 2 H 6 at aphelion and perihelion are 3.71×10 −7 and 7.86×10 −6 , respectively, and the ratio of CH 4 /CH 3 CHO at these two orbital points are 2.58×10 −2 and 1.25×10 −1 , respectively. ) of HCHO and HCOOH in our 1-D photochemistry model associated with 50 pptv of C 2 H 6 . Values are for latitude 2.5°N, local time 12:00, and solar longitudes 71°and 251°as a function of altitude.

Journal of Geophysical Research: Planets
Formaldehyde is produced with a much higher yield for C 2 H 6 oxidation than for CH 4 oxidation, resulting in a column averaged mixing ratio of 3.16×10 −15 and 5.74×10 −14 molecule/molecule at aphelion and perihelion, respectively. These yields are ≃6 times and 777 times higher than those made by CH 4 at aphelion and at perihelion, respectively. These higher yields of HCHO from C 2 H 6 oxidation is due to a larger number of production pathways compared to the CH 4 oxidation mechanism. Production of HCOOH from C 2 H 6 photochemistry during perihelion, column mixing ratio 7.30×10 −17 molecule/molecule, is comparable with values from CH 4 oxidation but is much smaller than CH 4 oxidation during aphelion with a column mixing ratio of 1.34×10 −18 molecule/molecule. Lower production of HCOOH from C 2 H 6 oxidation during aphelion is due to lower production pathways for HOCH 2 OOH and HOCH 2 O 2 , which drive subsequent production of HCOOH. Figure 10 shows the production and loss rates for HCHO and HCOOH from the oxidation of 50 pptv of C 2 H 6 . The largest production rate of HCHO is below the hygropause from the reaction of CH 3 O 2 peroxy radical with HO 2 , comparable to values from CH 4 oxidation. However, above the hygropause at 20 km the three-body reaction CH 3 C(O)+O 2 +M and the CH 3 radical reaction with atomic oxygen are the primary sources of HCHO. Combined, the HCHO production rate is an order of magnitude larger than the rate from CH 4 oxidation.

Journal of Geophysical Research: Planets
The higher reactivity of C 2 H 6 produces a larger quantity of HOCH 2 O 2 peroxy radicals and hydroperoxy radicals in the regions just above the hygropause compared to CH 4 oxidation. With no additional loss processes, the increased net production rate results in larger concentrations of HCOOH, with production falling off rapidly towards the surface. Figure 11 shows production and loss rates of CH 4 associated with a 50 pptv homogeneous column of CH 3 CHO at L S = 251°, latitude 2.5°, and at local noon to study the importance of UV photolysis of CH 3 CHO. We calculate the photolysis loss rate using the TUV model (section 2) that penetrate efficiently to the surface. Net CH 4 production rates peak at 31.13 molec cm −3 s −1 , corresponding to an increase in CH 4 mixing ratio of 1.69×10 −16 molecule/molecule s −1 , at ≃1.30 km above the Martian surface at this time of day. To our knowledge, this photolytic source of atmospheric CH 4 is the only proposed source that may be possible within the CO 2 -dominated atmosphere of Mars.

Discussion and Concluding Remarks
We find that the oxidation of atmospheric CH 4 in the Martian atmosphere, at magnitudes similar to reported observations, produces formaldehyde and formic acid as photochemical products but at volume mixing ratios too low for successful detection via remote sensing by the NOMAD and ACS spectrometers aboard the ExoMars Trace Gas Orbiter. Our photochemical lifetimes of formaldehyde are consistent with previous studies (e.g., Wong & Atreya, 2004), but we have also reported variations as a function of altitude and solar longitude. Our use of the MCD v5.3 atmospheric parameters and tracer profiles enabled us to deduce that the largest atmospheric lifetimes of HCHO at latitudes less than 30°are typically around 4 hours during the northern spring and summer above the point of H 2 O vapor saturation. Our model expands upon the findings from the steady state model by revealing the layered seasonal structure of HCHO that can develop, while refining the modeled lifetimes and revealing the seasonal variability of the compound.
Our 1-D model refines the vertical structure and seasonal variability of the CH 4 within the equatorial regions of Mars, revealing that altitudes with low water vapor content can provide O( 1 D) abundances large enough to suppress the lifetime of CH 4 to between 50 and 100 years, shorter than previous estimates (Krasnopolsky et al., 2004;Summers et al., 2002;Wong & Atreya, 2004). We reveal a region in the upper section of the 1-D model, corresponding to 50-70 km, where the absence of OH, O( 1 D), and lower UV photolysis increases CH 4 lifetimes to 400-800 years during the northern spring and to 800-1,300 years during the mid-northern summer to northern winter, significantly higher than previous model estimations. Lifetimes close to the surface and below the hygropause (0-30 km) are invariant with latitude during the northern spring but display variations between the north and southern hemisphere throughout the rest of the year, controlled by the sublimation of the respective hemispheres polar caps water ice content.
We find that the photochemical lifetime of C 2 H 6 is correlated to the atmospheric water vapor content on Mars. Within the hygropause, lifetimes are found to be within 3-4 years, where OH is the dominant photochemical sink. At higher altitudes, these lifetimes are reduced to 100-400 Mars sols (0.15-0.6 years), due to the higher abundance of atomic oxygen. The lifetimes of acetaldehyde reach values of approximately one sol close to the Martian surface across all equatorial latitudes during the northern spring that lower to roughly 0.5 sols in the northern hemisphere summer to winter periods. Acetic acid displays a similar seasonal trend, with lifetimes in the northern spring being between 3 and 4 sols below the hygropause, and lowering to between 0.2-1 sol with the elevation of the water saturation point. Oxidation of C 2 H 6 in the Martian atmosphere results in a distinct profile of acetaldehyde, as well as greater yields of HCHO and HCOOH, compared to CH 4 oxidation. The 1-D model predicts atmospheric lifetimes of between 15 and 32 hours below altitudes of 25 km during L S = 0-135°, lowering to 4-12 hours outside of this time frame across all equatorial latitudes, for CH 3 CHO. We therefore propose that any instrument detection of CH 3 CHO can be attributed to a surface release of C 2 H 6 within the immediate local environment of the site of observation.
Our more comprehensive description of atmospheric chemistry, involving 135 organic reactions, significantly expands on the schemes used by Wong and Atreya (2004) and Summers et al. (2002). No formic acid was reported by the model of Wong and Atreya (2004) for their 1-D steady state model with 100 ppmv of CH 4 , whereas we report volume mixing ratios of similar magnitude to HCHO below the hygropause with CH 4 abundances of 50 pptv, the current upper limit on Mars derived in Korablev et al. (2019). This discrepancy can be explained by our more detailed description of peroxy radical chemistry that is taken from the CAABA/MECCA v4.0 chemistry scheme. In our 1-D model, photochemical lifetimes of HCOOH are inversely proportional to the abundance of OH available. At the top of the hygropause, lifetimes have magnitudes of 1-10 sols. Below, the lifetimes vary substantially depending on the water vapor availability. Below 5 km, photochemical lifetimes in the drier southern latitudes can reach values exceeding 10 3 sols, whereas in the north, values between 10 and 100 are commonly found. These long photochemical lifetimes of HCOOH, in comparison to HCHO, make it the most likely photochemical product of CH 4 oxidation that could provide independent verification of CH 4 . Spectral features of HCOOH lie within the IR wavelengths measured by the ACS aboard the TGO, and the long wave channel of the PFS covers the wavenumber range of 250-1,700 cm −1 (Formisano et al., 2005) that includes two absorption features of the HCOOH molecule (Gordon et al., 2017).
We find that the introduction of CH 4 at magnitudes similar to an empirically derived upper limit (<50 pptv) fails to produce significant perturbations to CO, O 3 , or H 2 O vapor (not shown) that will observable by instruments aboard the TGO. CH 4 reactions with O( 1 D) above the hygropause and below 50 km have the net effect of increasing OH concentrations by magnitudes of 10 −3 % after one sol of exposure, which results in catalyzing the conversion of CO to CO 2 via reaction e 1 of B1. The increase in OH arises from the previously described reactions of O( 1 D) atoms reacting with CH 4 in the drier altitudes, most notably reaction b 7 . CO experiences relative perturbations of -10 −6 % in the mid-altitudes, which will be lost within instrumental noise. This rise in OH, and the loss of O( 1 D) to CH 4 interactions and O( 3 P) with organic radical interactions, results in a drop in O 3 production in this region and increased O 3 loss. These perturbations are small, however, with drops of 10 −3 % after one sol, an amount that is below instrument signal-to-noise values. This highlights that source

Journal of Geophysical Research: Planets
TAYSUM AND PALMER regions of CH 4 will not be identifiable by perturbations made to inorganic trace gas species observable to the ExoMars Trace Gas Orbiter.
Finally, our model highlights the possible existence of an atmospheric source of CH 4 from the photolysis of CH 3 CHO, an oxidation product of C 2 H 6 . Photolysis of CH 3 CHO in the Martian atmosphere is capable of producing trace amounts of CH 4 at all altitudes above the surface. We find that 50 pptv of CH 3 CHO is capable of producing CH 4 at a volume mixing ratio rate of 4.54×10 −16 s −1 .
This opens up the possibility of further studying CH 3 CHO as a possible source for CH 4 detection on Mars. Although no evidence exists for CH 3 CHO on Mars, one hypothetical source could be due to a subsurface acetylene hydratase (Rosner & Schink, 1995) enzyme that converts acetylene (C 2 H 2 ) to CH 3 CHO through reactions with liquid water.
Appendix A: Martian Atmospheric Lifetimes of Methane, Formaldehyde, Formic Acid, and Ethane A1 Methane Figure A1 shows the net atmospheric lifetime of CH 4 , incorporating all loss terms (Figure 2), as a function of latitude and solar longitude. Atmospheric lifetimes reach a minimum of 25-400 years between the top of the hygropause and approximately 50 km due to the larger abundance of the O( 1 D) atom. Below the hygropause where OH is the dominant sink, atmospheric lifetimes vary between 400 and 800 years. The longest lifetimes of 800-1,600 years lie between 50 and 60 km during winter (L S = 270-360°), as described above, where the OH and O( 1 D) loss processes decline and before photolysis dominates above 60 km.
Variations in atmospheric lifetime are driven by the position of the Martian hygropause and the water vapor content beneath, which is determined by 3-D model output from the MCD (section 2). Water vapor columns reach their maximum during hemispheric summer months when polar water ice sublimates with rising atmospheric temperatures. SPICAM water vapor column measurements at L S = 50° (Fedorova et al., 2006) show only small variations at latitudes less than 30°, with values ranging from 2-10 pr μm, with saturation values under 25 km. This is consistent with our model loss processes that determine the variation in atmospheric lifetime. Above 50 km, the stronger flux of solar radiation in the northern hemisphere due to the planets axial tilt results in a larger abundance of O( 1 D) and a subsequent lower photochemical lifetime of CH 4 than in the southern hemisphere. As water vapor column abundances and saturation altitudes increase with L S , the low latitude atmospheric lifetime of CH 4 increases below 50 km due to H 2 O reacting with O( 1 D) and decreases below roughly 25 km. The decrease in solar flux during northern autumn and winter months (L S = 180°-360°) reduces the production rate of O( 1 D) relative to the southern hemisphere, which explains the variation in lifetimes between the hemispheres. Figure A2 shows the resultant seasonal variability of HCHO lifetime at latitudes 30°N, 2.5°N, and 30°S. We find the longest lifetimes, between 4 and 5 hours, are found below 5 km in the southern hemisphere during the northern summer/southern winter, when this region contains low levels of water vapor.   Figure A3 shows the resultant seasonal variability of HCOOH lifetime at latitudes 30°N, 2.5°N, and 30°S. The lifetime of HCOOH is anticorrelated with water vapor, as expected. The longest lifetime (of magnitude's greater than 10 4 sols) is during the northern spring at altitudes greater than 60 km where OH is lowest. As OH values increase with the supply of water vapor from northern polar ice sublimation, atmospheric lifetimes falls to 1-10 sols close to the hygropause and 10-200 sols at lower altitudes closer to the surface (below 5 km).

A2 Formaldehyde
With atmospheric lifetimes of this length, describing seasonal changes in HCOOH profiles using a 1-D photochemistry model becomes problematic due to the photochemical lifetimes exceeding the venting timescales associated with our 3.75°thick zonal band that our model represents. Without horizontal transport equations to describe the loss of HCOOH to the zonal band edges, the resultant profiles after a one-year 1-D model run will be subject to gross inaccuracies. As previously discussed in section 2, solving the 1-D model equation for steady state conditions will not yield useful information due to the observed temporal and localized nature of CH 4 emissions, invalidating the notion of a steady-state environment. Figure A4 shows that C 2 H 6 has longer photochemical lifetimes below the points of H 2 O saturation, where we find typical values that range from 3 to 5 years. In the middle atmosphere, the large abundance of O ( 3 P) reduced the lifetime to 30-450 sols. Based on these calculations, we suggest that any detection of C 2 H 6 made by the TGO instruments will due to an active release.   (Tsang & Hampson, 1986) Note. Bimolecular rate coefficient units are cm 3 molec −1 s −1 . Values denoted by superscript a are three-body reactions with values taken with atmospheric number density complying to temperatures of T = 298 K and pressures of 660 Pa.