Is it possible that a planet originally full of water, and the solar winds decompose the water and hydrogen escapes out, then the oxygen combines with carbon at crust, just like Venus but the whole planets are mostly carbon dioxide and hence become a gas giant?


1 Answer 1


While not absolutely impossible this is not a likely situation. The reason is that for a planet to be a "giant", then it will have enough gravity to hold on to its hydrogen and helium. Conversely, if you remove the H an He from a planet, it won't be "giant". There are only two gasses that are common enough in the universe to form a massive gas mantle, and those are Hydrogen and Helium.

  • $\begingroup$ Based on surface temperature escape velocity, it might be possible. en.wikipedia.org/wiki/Atmospheric_escape#/media/… about 2000 K for loss of hydrogen and helium. 4000 K for loss of most gasses other than CO2. Give or take. I'd say it's still highly unlikely, but not impossible under the right circumstances. $\endgroup$
    – userLTK
    Nov 1, 2016 at 22:02
  • 1
    $\begingroup$ That sounds right, but it still leaves the second objection: If you heat a gas giant to 3000K, and remove the H and He, what remains is not a gas giant made of CO2, it is a terrestrial planet. $\endgroup$
    – James K
    Nov 1, 2016 at 23:44
  • $\begingroup$ Fair enough. :-) Or, another possibility, It could form outside the frost line to obtain lots of frozen gases/ices, then migrate inwards and it would need to be maybe 10 Jupiter masses initially to retain about 1 Jupiter mass after losing hydrogen and helium. A difference between objects like Ceres/Ganymede and Mercury, Ceres and Ganymede have retained much of their ices from formation. Mercury never had any. Then, add heat, the ices melt, the planet loses it's lighter gases and you have a CO2 atmosphere, potentially a massive one. That said, the improbabilities make my brain hurt. $\endgroup$
    – userLTK
    Nov 1, 2016 at 23:53
  • $\begingroup$ And (sorry), looks like he took your check away. I think your answer is still correct - hence my comment, not a new answer. I was just throwing long shot ideas out there. $\endgroup$
    – userLTK
    Nov 1, 2016 at 23:59

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .