<P> Some aspects of the Martian atmosphere vary significantly . As carbon dioxide sublimes back into the atmosphere during the Martian summer, it leaves traces of water . Seasonal winds transport large amounts of dust and water vapor giving rise to Earth - like frost and large cirrus clouds . These clouds of water - ice were photographed by the Opportunity rover in 2004 . NASA scientists working on the Phoenix Mars mission confirmed on July 31, 2008 that they had indeed found subsurface water ice at Mars's northern polar region . </P> <P> Trace amounts of methane (CH), at the level of several parts per billion (ppb), were first reported in Mars's atmosphere by a team at the NASA Goddard Space Flight Center in 2003 . In March 2004, the Mars Express Orbiter and ground - based observations by three groups also suggested the presence of methane in the atmosphere at a concentration of about 10 ppb (parts per billion). Large differences in the abundances were measured between observations taken in 2003 and 2006, which suggested that the methane was locally concentrated and probably seasonal . </P> <P> Because methane on Mars would quickly break down due to ultraviolet radiation from the Sun and chemical reactions with other gases, its reported persistent presence in the atmosphere also implies the existence of a source to continually replenish the gas . Current photochemical models alone cannot explain the rapid variability of the methane levels . It had been proposed that the methane might be replenished by meteorites entering the atmosphere of Mars, but researchers from Imperial College London found that the volumes of methane released this way are too low to sustain the measured levels of the gas . </P> <P> Research suggests that the implied methane destruction lifetime is as long as ~ 4 Earth years and as short as ~ 0.6 Earth years . This lifetime is short enough for the atmospheric circulation to yield the observed uneven distribution of methane across the planet . In either case, the destruction lifetime for methane is much shorter than the timescale (~ 350 years) estimated for photochemical (UV radiation) destruction . The rapid destruction (or "sink") of methane suggests that another process must dominate removal of atmospheric methane on Mars, and it must be more efficient than destruction by light by a factor of 100 to 600 . This unexplained fast destruction rate also suggests a very active replenishing source . In 2014 it was concluded that presence of strong methane sinks are not subject to atmospheric oxidation . A possibility is that the methane is not consumed at all, but rather condenses and evaporates seasonally from clathrates . Another possibility is that methane reacts with tumbling surface sand quartz (SiO) and olivine to form covalent Si--CH bonds . </P>

Which gasses from table 3 could mars retain