Exploiting night-time averaged spectra from PFS/MEX shortwave channel. Part 2: Near-surface CO retrievals

Because of its important role in the Martian carbon cycle, carbon monoxide (CO) has been the subject of many measurements from ground and from space. Daytime measurements have been mostly exploited to measure the CO abundance because of their good signal-to-noise ratio, but night-time observations have not been documented yet. We demonstrate here the possibility of using PFS (Planetary Fourier Spectrometer) night-time observations to measure the Martian CO abundance. More particularly in this paper, we show that measuring CO during night with PFS can be achieved by averaging a large number of spectra to reach sufficient signal-to-noise. Furthermore, we show that the number of averaged spectra is not the only driving parameter for the detection of CO. High surface temperatures and high thermal contrast (negative in our case) are the two other conditions required for the measurement of the night-time CO abundance. Because of this, the retrievals are especially successful in the Southern Hemisphere during spring and summer when and where these two conditions are met. For night-time spectra with a positive detection, CO vertical profiles are successfully retrieved using the Optimal Estimation method and are characterized in terms of vertical sensitivity. Successful retrievals imply the use of accurate temperature profiles, and in particular an appropriate representation of the thermal inversion in the lower atmosphere. The temperature was obtained by using the relevant information from the CO2 ν3 band (Bauduin et al., 2020, this issue). A complete error budget of the retrieved CO profiles is also performed and includes different sources of uncertainty. Although the retrieved profiles are not resolved vertically, we show in particular that night-time PFS observations carry information mostly on the CO abundance for the 0–10 ​km altitude region, and thus provide a stronger constraint on the near-surface CO abundance compared to daytime observations.