Seasonal and latitudinal variations of the HDO/H2O ratio in the martian atmosphere

Abstract

We observed latitudinal variations of the HDO/H2O ratio in the martian northern midsummer at LS ​= ​128° using CSHELL at NASA IRTF. This is the last observation in our decadal monitoring of HDO/H2O on Mars (Krasnopolsky, 2015, Icarus 257, 377–386). The observed latitudinal variations agree with those predicted for that season by the GCM model for HDO/H2O by Montmessin et al. (2005, J. Geophys. Res. 110, E03006).The observation shows very high correlation between HDO/H2O and the mean temperature in the lowest scale height with correlation coefficient of 0.955. The Arrhenius plot for this function is well fitted by activation energy of 1072 ​± ​90 ​K. However, similar plots for our observations at LS ​= ​42° and 70° with correlation coefficients of 0.89 and 0.72 gave activation energies of 2123 ​± ​350 ​K and 819 ​± ​300 ​K, respectively. The observed correlation between HDO/H2O and the mean temperature is caused by the temperature-dependent isotope fractionation. However, the quantitative assessment of the correlation and its variations remain uncertain, and activation energy is not a proper tool to explain latitudinal variations of HDO/H2O. Fractionation of HDO/H2O along the martian globe is too complicated to be described by unique activation energy.The observational data for seven seasonal points are collected into a map of the seasonal-latitudinal variations of HDO/H2O. The map makes it possible to predict HDO/H2O at given season and latitude. A similar map in the GCM by Montmessin et al. (2005) is the only analog of our map. HDO/H2O is more variable in our map than in the GCM map, and this is partly caused by the observational uncertainties in our map and zonal averaging in the GCM map. However, some features in our map look real. For example, the GCM predicts for LS ​= ​60° to 150° rather constant HDO/H2O ​≈ ​4.5 from 60N to 40S, while our map shows 5 ​± ​1 from 80N to the equator and then the decrease to ≈2 ​at 40S. The GCM for HDO/H2O by Montmessin et al. (2005) neglects the isotope fractionation by photolysis of water vapor. Here we compare times of vertical and global mixing with those of condensation and photolysis and conclude that this neglect is valid up to ≈50 ​km.