Interannual variability in hydrogen escape from Mars and implications for water loss: Insights from ExoMars Trace Gas Orbiter data assimilation into a global climate model

Investigations into martian paleoclimatology have shown Mars in the past (~ 4 billion years ago) used to contain significant amounts of surface liquid water in contrast to present day conditions [1,2]. Given the importance of liquid water for life on Earth, understanding the evolution of the martian climate from a planet containing surface liquid water to a cold and dry planet today will improve our understanding of the conditions that influence terrestrial planetary habitable environments. Hydrogen escape is proposed as a major process in the depletion of water from Mars. This is where water molecules disassociated into hydrogen and oxygen after which hydrogen escapes the martian atmosphere through atmospheric escape processes such as Jeans escape and sputtering [3].

Previous investigations of martian hydrogen escape using a 1-D upper atmospheric model have shown that water loss rates on Mars are driven by the availability of water vapour near the boundary between the middle and the upper atmosphere (~ 80 km altitude), with water loss on Mars seasonally driven during perihelion and enhanced atmospheric dust activity such as regional and global dust storms (GDS) [4,5]. GDS events on Mars strongly influence interannual variations in the martian climate, occurring approximately every 5–10 Earth years, with the most recent GDS occurring in 2018 (hereafter referred to as martian Year 34 or MY34) [6]. The MY34 GDS also coincided with the start of the ExoMars Trace Gas Orbiter (TGO) observations, with TGO providing the greatest level of spatial coverage of water vapour profiles to date. These observations indicate enhanced transport of water vapour to the upper atmosphere [7,8,9].

Constraining the seasonal and interannual availability of water vapour in the upper atmosphere is therefore integral to determine the amount of water escaping from Mars. One such method of constraining is through data assimilation. This methodology combines the data of accurate but incomplete spatiotemporal observations of the martian atmosphere by orbiters with the complete spatiotemporal coverage albeit parameterised simulation of the martian atmosphere provided by models to reconstruct the martian atmosphere most in line with observations ensuring physical consistency. Recent work on data assimilation of TGO water vapour column and profiles into the Mars Planetary Climate Model UK-Spectral version (MPCM) has shown there is improved representation in the MPCM of water vapour profiles in line with orbiter observations around the upper atmosphere in contrast to free run simulations [10]. Therefore in this presentation, we explore the hydrogen escape flux using MPCM water cycle simulations based on data assimilation of TGO data.

The martian water cycle is simulated from L_S = 160˚ in MY34 to the end of MY35 using the MPCM with data assimilation of ExoMars Trace Gas Orbiter water vapour column and profile measurements, alongside temperature profiles from Mars Reconnaissance Orbiter and dust optical depths from Mars Reconnaissance Orbiter and Mars Odyssey, to constrain the model’s atmospheric state and reproduce the martian water cycle in the most representative state. The outputs from the MPCM water cycle are coupled to the 1D upper atmosphere photochemical model from which the hydrogen escape is calculated from L_S = 160˚ in MY34 to the end of MY35. The hydrogen flux during the MY34 GDS is compared to MY35 which is a martian year with nominal atmospheric dust levels.

 

Figure 1: MY34/MY35 Hydrogen escape flux ratio vs L_S from MPCM simulations using data assimilation of ExoMars Trace Gas Orbiter water vapour profiles and columns along with Mars Reconnaissance Orbiter temperature profiles

 

We find that the hydrogen escape during the MY34 GDS is larger by a factor of 6 with respect to MY35 around L_S = 195 – 215˚. There is also enhanced hydrogen escape by around 4 times in magnitude with respect to MY35 between L_S = 330 – 345˚ in MY34 which correlates with the annual C – Type [6] regional dust storm activity which was considerably stronger in MY34 compared to MY35 (Figure 1). These results show there is enhanced water loss during increased atmospheric dust activity such as GDS and C-Type regional dust events, suggesting that interannual dust activity needs to be considered for long term martian water loss calculations. We present the reanalysis of the water cycle during the MY34 GDS and findings of the atmospheric processes influencing variations in the hydrogen escape flux. We will also present estimates for the interannual water loss from Mars based on MPCM water cycle simulations constrained by multi-orbiter assimilation.

Reference: [1] Ramirez, R.M. & Craddock, R.A. (2018) Nat. Geosci., 11, 230–237. [2] Pollack, J. et al. (1987), Icarus, 71(2), pp. 203–224. [3] Kasting, J.F. and Catling, D. (2003), Annu. Rev. Astron. Astrophys., 41(1), pp. 429–463. [4] Chaffin, M., Deighan, J., Schneider, N. et al. (2017), Nature Geosci 10, 174–178 [5] Holmes, J.A. et al. (2021) Earth Planet. Sci. Lett., 571, p. 117109. [6] Kass, D. et al. (2016), GRL, 43 (12), 6111-6118. [7] Vandaele, A. C. et al. (2018) Space Sci. Rev. 214(5). [8] Korablev, O. et al. (2018), Space Sci. Rev. 214(1), 7 (2018). [9] Aoki, S. et al. (2019), JGR (Planets) 124(12), 3482–3497. [10] Holmes, J.A. et al. (2022) JGR (Planets), 127(10).