Morphology of the Excited Hydroxyl in the Martian Atmosphere: A Model Study—Where to Search for Airglow on Mars?

Abstract

Monitoring excited hydroxyl (OH*) airglow is broadly used for characterizing the state and dynamics of the terrestrial atmosphere. Recently, the existence of excited hydroxyl was confirmed using satellite observations in the Martian atmosphere. The location and timing of its detection on Mars were restricted to a winter season at the north pole. We present three-dimensional global simulations of excited hydroxyl over a Martian year. The predicted spatio-temporal distribution of the OH* can provide guidance for future observations, namely by indicating where and when the airglow is likely to be detected.

1. Introduction

Airglow is a non-thermal emission by planetary atmospheres, which is not related to auroral effects, lightning, and meteor fireballs [1]. Burns [2,3,4] and Yntema [5] were the first to suggest the existence of an irradiating layer in the terrestrial atmosphere based on observations. Rayleigh [6] had shown that it was not associated with auroras. In the 1920s–1930s, the nature of the observed phenomenon was broadly discussed, and Chapman [7] was apparently the first to point out that exothermic chemical reactions are a source of excited atoms and molecules producing airglow. Further studies of all airglow emissions were slowed down by World War II but resumed with triple intensity due to new infrared and other emission-registration technologies developed for military purposes. The full history of this phenomenon is summarized in the overview by Hersé [8].

One of the most used observational methods for obtaining information about dynamics, temperature, and chemical composition in the terrestrial atmosphere is measuring OH* excited states emissions. Meinel [9,10] found lines of vibro-rotational transitions in atmospheric emissions, while Bates and Nicolet [11] suggested the exothermic reaction of ozone and hydrogen as a mechanism for populating vibrational levels of OH*. Krassovsky et al. [12,13,14,15] developed a theory for retrieving temperature and gravity wave parameters at the emissions’ peak altitude. Evans and Llewellyn [16] suggested a method of inferring atomic hydrogen concentrations from OH* emission. Based on Solar Mesosphere Explorer (SME) satellite measurements, Thomas [17] showed how this emission can be used for calculating atomic oxygen concentrations and applied it for studying seasonal variations in this minor component. Currently, observations of OH* emissions in the Earth’s mesopause are used in a range of applications, like quantifying atmospheric variability due to gravity waves (e.g., [18,19,20]), tides [21,22], and planetary waves [23,24]. The OH* emissions were utilized to study sudden stratospheric warming events [25,26] and the quasi-biennial oscillation [27]. Airglow emissions were used for assessing temperature trends and variations induced by the solar cycle (e.g., [28,29,30,31,32]), as well as the chemical composition in the mesopause region [33,34,35].

Hydroxyl emissions are not a strictly terrestrial phenomenon. They have been found on Venus [36,37,38] and Mars [39]. In the Venus atmosphere, OH* emissions were observed for the first time in March of 2007 [36] by the Visible and Infrared Thermal Imaging Spectrometer (VIRTIS) instrument onboard the Venus Express satellite [40]. Six years later, the OH* emissions were detected on Mars with Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) aboard the Mars Reconnaissance Orbiter (MRO) in near-IR limb observations [39]. The OH* emissions for vibrational transitions 1-0, 2-1, and 2-0, with wavelengths 2.81, 2.94, and 1.42 μm, respectively, were detected in winter polar night areas (70–90°N/S) during Martian year 30 (MY30). The presence of OH* on these planets provides an opportunity to apply remote sensing techniques developed for Earth to studies of atmospheric processes there. At the moment, very little is known about spatio-temporal variations in hydroxyl emissions on these planets. This hinders the planning of new space missions and the development of instrumentation for detecting OH* airglow. Theoretical (modeling) investigations of OH* emissions in the Venus atmosphere were already performed in a number of works [41,42]. Our paper addresses this lack of knowledge by predicting where and when observations of OH* are possible in the Martian atmosphere based on model simulations.

The paper is organized as follows: The model for OH* in the Martian atmosphere is outlined in Section 2. The results of the calculations are presented in Section 3. These results are discussed in Section 4, and the conclusions are given in Section 5.

2. Materials and Methods

We assume that the excited hydroxyl is in a photochemical equilibrium during night-time [43]. This assumption enables us to explicitly express the concentration of hydroxyl at all excitation levels [OH_v] in the following form (see [44], Equation (1)):

$$\left[{OH}_v\right]=\frac{\left(\begin{array}{c}{f}_v{r}_1\left[H\right]\left[O_3\right]+\sum _{v^{\prime }=v+1}^9{A}_{v^{\prime }v}\left[{OH}_{v^{\prime }}\right]\left[{CO}_2\right]+\sum _{v^{\prime }=v+1}^9{G}_{v^{\prime }v}\left[{OH}_{v^{\prime }}\right]\left[N_2\right]+\\ +\sum _{v^{\prime }=v+1}^9{B}_{v^{\prime }v}\left[{OH}_{v^{\prime }}\right]\left[O_2\right]+\sum _{v^{\prime }=v+1}^9{D}_{v^{\prime }v}\left[{OH}_{v^{\prime }}\right]\left[O\right]+\\ +\sum _{v^{\prime }=v+1}^9{E}_{v^{\prime }v}\left[{OH}_{v^{\prime }}\right]\end{array}\right)}{\left(\begin{array}{c}\sum _{v^{″}=0}^{v-1}{A}_{v{v}^{″}}\left[C{O}_2\right]+\sum _{v^{″}=0}^{v-1}{G}_{v{v}^{″}}\left[N_2\right]+\sum _{v^{″}=0}^{v-1}{B}_{v{v}^{″}}\left[O_2\right]+\\ +\sum _{v^{″}=0}^{v-1}{D}_{v{v}^{″}}\left[O\right]+\sum _{v^{″}=0}^{v-1}{E}_{v{v}^{″}}+\\ +r_2\left(v\right)\left[O\right]\end{array}\right)},\left(\begin{array}{c}v<v^{\prime }\\ v^{″}<v\end{array}\right),$$

(1)

where v is the vibrational number; $f_{v=9,\dots ,5}=0.47,0.34,0.15,0.03,0.01$ is the nascent distribution [45]; $r_1=1.4·{10}^{-10}·exp\left(\frac{-470}{T}\right)$ is the reaction rate for the reaction of atomic hydrogen with ozone [46]; $r_2\left(v\right)$ is the same for the reaction of excited hydroxyl with atomic oxygen [47]; and A, B, D, and G are the quenching coefficients by carbon dioxide [41], molecular oxygen [45], atomic oxygen [47], and molecular nitrogen [48], respectively. E represents the Einstein (spontaneous emission) coefficients from [49]. The square brackets denote the number densities of chemical species.

In Equation (1), the spontaneous emission, molecular, and atomic oxygen quenching are treated as multi-quantum phenomena; that is, they include relaxation from all the highest vibrational levels to all the lowest ones. Not all multi-quantum quenching coefficients are known. For instance, coefficients for quenching by molecular nitrogen and carbon dioxide have not been fully characterized, but only the so-called collisional cascade rates (where transitions occur to one level below) have been provided [50]. The most recent update on these coefficients has been presented by Krasnopolsky [41] for quenching by carbon dioxide and by Makhlouf et al. [48] for quenching by molecular nitrogen. We adopted these values from Krasnopolsky [41] and Makhlouf et al. [48] in our calculations to construct the diagonal matrix for A_vv′ and G_vv′ for transitions v → v − 1, while we assign zero values to the non-diagonal terms for other transitions. Although argon and carbon monoxide are minor species in the Martian atmosphere, their concentrations are quite large. The quenching rates of excited hydroxyl by these species are not known thus far; therefore, we had to neglect them, just as it is usually performed for the Earth’s atmosphere.

Our OH* model does not include the reaction between hydroperoxy radicals (HO₂) and atomic oxygen. This omission is justified by the limited significance of this reaction as a source for populating vibrationally excited hydroxyl levels [43,50,51,52]. At the initial stages of the study of hydroxyl emissions, the reaction of hydroperoxy radicals with atomic oxygen was introduced in order to reconcile the results of observations and modeling. With the acquisition of new knowledge about the processes of quenching, spontaneous emission, and quantum yield for the main reaction, there was no longer a need for the inclusion of this reaction in the consideration. In addition, there is no laboratory evidence of a significant OH* yield from this reaction. Therefore, we omitted it, following many other authors [53,54,55,56].

The model outlined above has been described in detail and compared with the available observations [39] in our previous studies [44,57]. It reproduces the values and shape of the CRISM observations for transitions 1-0 and 2-0 and shows ~30% lower values near the peak for transition 2-1. The latter is still within the CRISM’s uncertainty (see Figure 1c in [44] and Figure 7 in [39]). For transition 1-0, this result is better than the simulations with the Laboratoire de Météorologie Dynamique General Circulation Model (LMD GCM) and with the OH* models of Krasnopolsky [58] and García-Muñoz [43]. The latter two models overestimate the emission for this transition (Figure 1c in [44] and Figure 7 in [39]). For transition 2-0, all three models show similar values at the OH* peak (~5∙10³ photons∙cm⁻³∙s⁻¹), which are slightly larger than in the observed emission. The main differences between the current version and the previous one are the inclusion of the temperature dependence for quenching by atomic oxygen (D) and the incorporation of the reaction of excited hydroxyl with atomic oxygen ($r_2\left(v\right)$), as introduced in the work of Caridade et al. [47].

In order to calculate [OH*] from (1), we used concentrations of all chemical constituents (O₃, O, H, O₂, N₂, CO₂), air density, and temperature from the Mars Climate Database (MCD), which is based on simulations with the Mars Planetary Climate Model (formerly the LMD GCMl) [59,60]. The latter is a three-dimensional model with a resolution of 64 longitudinal, 48 latitudinal, and 32 vertical grid points covering altitudes from the ground to ~120 km. We calculated OH* distributions at each grid point using Equation (1). In particular, the MCD includes the concentration of ozone [61], which is directly involved in OH* production; water vapor [62], which is the principal source of odd hydrogens (H, OH, HO₂); and other long-lived species (carbon dioxide, atomic oxygen, molecular oxygen, and molecular nitrogen) involved in quenching processes. We utilized the data for Martian year 30, which is characterized by rather low dust [63] and solar activities. Investigation of the impact of dust storms and solar activity on hydroxyl emissions deserves a separate study and is out of the scope of this paper. We calculated the OH* according to Equation (1) at 00:00 LT and 12:00 LT (local midnight and midday, correspondingly) and averaged over all such longitudes during a sol. Thus, night- and daytime values refer to the local time rather than to insolation, as Equation (1) is applicable to all solar zenith angles (insolation conditions).

3. Results

The decline in the concentration of excited hydroxyl with an increase in vibrational number was found using modeling and observations in the Earth’s atmosphere [45,48,50,52,64]. Modeling OH* for the Martian atmosphere shows a similar result, which has been explained theoretically [43,44,57]. Since OH* concentrations increase with decreasing vibrational number, the strongest volume emission is produced by OH* with the lowest vibrational numbers, as is seen from modeling and observations [39,44,57]. Note that observations show a similar volume emission for transitions 2→1 and 1→0. For certainty, we consider only concentrations of OH* with the vibrational number one.

Different instruments have different sensitivities and uncertainties. Therefore, there is no sense in focusing on a particular one. Instead, we introduce a possible detection threshold. The CRISM instrument detected OH* emissions with absolute values above 10³ photons∙cm⁻³∙s⁻¹ with acceptable uncertainties depending on the wavelength. For the first two vibrational numbers, this corresponds to the OH* concentration of ~10² cm⁻³.

We consider the concentration $\left[{OH}_1\right]=100$ cm⁻³ as a detectability threshold because the corresponding volume emission for the 1-0 transition is $V_{10}=1760$ photons∙cm⁻³∙s⁻¹ (E₁₀ = 17.6 s⁻¹, [49]). This emission is practically guaranteed to be measured with an acceptable accuracy.

Figure 1 presents the calculated seasonal variations in the night-time mean (averaged over a sol) $\left[{OH}_1\right]$ at several latitudes, while Figure 2 illustrates its latitudinal structure during different seasons. The red solid lines in the figures indicate the peak of the layer, while the threshold value $\left[{OH}_1\right]=100$ cm⁻³ is shown with the white dashed line. It is seen that the distributions are not fully periodic functions of Ls. This is a consequence of using the MCD data for a specific year (MY30), which are affected by the inter-annual variability. It is also seen from Figure 1a,e and Figure 2 that the largest concentrations occur at high northern latitudes (≥60°N) in the second half of the Martian year (L_S = 180–360°) and at high southern latitudes (≥60°S) in the first half (L_S = 0–180°). The concentration reaches more than 10³ cm⁻³ and is located at altitudes of ~45–55 km. The hydroxyl peak extends to the maximum height of 75 km in northern mid-latitudes (~40°N) and in high southern latitudes (≥60°S), both around the perihelion season L_S = 270° with concentrations of several tens of cm⁻³. The lowest height of the peak is ~42 km, which occurs in high southern (≥60°S) and equatorial latitudes near the aphelion (L_S = 90°) with values of more than 10³ cm⁻³ and at ~30°S during the northern spring equinox (L_S = 0°) with values of a few hundred cm⁻³.

Since OH* concentrations depend on atomic oxygen and temperature, we plotted Figure 3 for illustrative purposes to show the temperature (left column) and atomic oxygen (right column) for spring (first row), summer (second row), fall (third row), and winter (last row).

Figure 4 presents the computed seasonal-latitudinal cross-sections of [OH₁] concentrations at the height of the peak values in the Martian atmosphere for (a) night-time and (b) daytime conditions, along with the peak height itself at (c) night and (d) daytime. The white dashed line denotes regions where concentrations exceed the threshold required for detection.

4. Discussion

There are similarities and differences in the variations in the nocturnal hydroxyl on Mars and in the terrestrial mesosphere. The figures demonstrate that the peak on Mars varies by ~30 km between ~45 and 75 km. On Earth, the variations are much narrower and amount to only about 10 km (e.g., [65] and references therein). On Mars, the largest vertical variation is predicted near northern mid-latitudes (~40°N) and southern high latitudes (70°S) (Figure 1b,e). The terrestrial OH* airglow layer varies annually and semiannually [22,27,66,67], whereas on Mars, only the annual cycle is seen. Unlike in the Earth’s mesopause, the annual variations on Mars demonstrate no latitudinal symmetry with respect to the equator, except at high latitudes (Figure 1a,e). Providing an explanation for the fluctuations in OH* is beyond the scope of this paper. Such a study could be carried out using the approach published in the previous works [44,57], where it was shown that if ozone is in chemical equilibrium, the concentration of OH* is proportional to that of atomic oxygen and air density and inversely proportional to the power of temperature (~1/T^2.4). Thus, variations in the hydroxyl layer are ultimately determined by variations in these three components. Moreover, variations in air density and atomic oxygen concentrations were found to be primary factors, while variations in temperature contribute less. In case of strong deviations in ozone from chemical equilibrium, variations in OH* can be considered only in terms of components directly involved in its formation, i.e., of ozone and atomic hydrogen. Therefore, it is important to establish with certainty the times and locations where the photochemical equilibrium of ozone holds before considering factors contributing to the OH* variations.

Near the terrestrial mesopause, the concentration of excited hydroxyl at the peak anticorrelates with its altitude [27,68,69]. Figure 1 demonstrates a similar inverse correlation on Mars. This relationship was theoretically predicted under the assumption of ozone being in photochemical equilibrium [44,57]. The expression for the peak altitude was derived in [44,57]. It shows that the altitude depends on the amount of atomic oxygen, temperature, and their vertical gradients (see Equation (9) in [44,57]). However, it is important to note that such an inverse correlation can occur without photochemical equilibrium as well.

The full width at half maximum layer on Mars can be as large as 20–25 km, while on Earth, it is approximately 8–10 km [70]. This implies that the Martian OH* layer is broader and that retrieving minor chemical species like atomic oxygen and atomic hydrogen would be possible over a wider range of altitudes.

Another similarity between the excited hydroxyl in the Martian and terrestrial atmospheres is the formation of double-maxima structures, as seen in Figure 1c and Figure 2b,d. This phenomenon has been reported in the terrestrial mesopause in several studies [71,72]. Such structures are formed because [OH*] is directly proportional to the concentration of atomic oxygen and inversely proportional to the power of temperature. Consequently, one maximum of OH* can occur near the peak of O and the other forms near the temperature minimum. Similar reasons can explain the double structure on Mars.

By comparing Figure 2 and Figure 3, one can see that large values of OH* concentrations occur when atomic oxygen concentrations are large, even though the temperature is also large in these regions (high and middle latitudes). In previous works, it was shown that atomic oxygen plays a primary role in OH* formation [44,57]. Regions of high hydroxyl concentration can be formed when the amount of atomic oxygen is not the largest, but the temperature is low, as, for example, in the equatorial and low latitudes in spring and summer (Figure 3a–d). We do not present a more extensive analysis of the impacts of temperature and atomic oxygen because this question is out of the scope of the current study. Nevertheless, we note that it can be achieved either by utilizing the decomposition approach from [44,57] or by conducting numerical experiments with constant (averaged) temperature and variable [O], and vice versa.

It is seen in Figure 4 that, during night-time, the detection of excited hydroxyl on Mars is feasible at middle and low latitudes in the first half of the year, up to L_S = 160°, and at high latitudes during almost all seasons, except for a one-month period in the summer hemispheres around L_S = 70° and L_S = 270°. The areas near L_S = 70° and L_S = 270° at high latitudes correspond to the polar day, where ozone (directly involved in OH* formation) is reduced via dissociation. During daytime, detection is only possible at high latitudes, where OH* concentrations closely resemble night-time values, and diurnal variability is either nonexistent or very weak. This fact suggests that ozone may not be in a photochemical equilibrium in this region but rather behaves as a passive tracer, similar to what occurs in the stratosphere of Earth. However, this aspect requires further in-depth exploration and is out of the scope of this paper.

5. Conclusions

In this paper, we characterized the main features of the excited hydroxyl (OH*) morphology in the Martian atmosphere obtained from modeling and discussed the similarities and differences in the latitudinal-seasonal variations in OH* in the terrestrial atmosphere. The three-dimensional distributions of OH* were calculated using concentrations of chemical species, air density, and temperature from the Mars Planetary Climate Model (formerly the Laboratoire de Météorologie Dynamique General Circulation Model) for MY30. The comparison of the model’s results with observations has been presented in past works [44,57].

Measurements of the excited hydroxyl airglow can be used for obtaining information on minor chemical species, for studying variations in temperature and dynamics. The main conclusion of our work is that the night-time detection of hydroxyl is possible at high latitudes (above ~60°) throughout the year, with the exception of a short period of about one Martian month in the summer hemispheres with L_S = 70° and L_S = 270°, which correspond to the polar day regions. At middle and equatorial latitudes (from ~60°S to ~60°N, see Figure 4), hydroxyl can be detected from L_S = 0° to L_S = 160°. The detection of hydroxyl in the daytime is likely to be very difficult at middle and equatorial latitudes through the entire year and is more feasible at high latitudes, except for a short period around L_S = 70° and L_S = 270° (under the polar day conditions).

Finally, we note that because the atmospheres of Venus and Mars are both CO₂-dominated and processes of formation and deactivation of OH* are identical on both planets, the presented approach can be directly applied to the Venusian atmosphere as well.