Escape of light gases from an atmosphere
A Socratic walk-through of Escape of light gases from an atmosphere — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why has Earth held on to its nitrogen while losing almost all of its hydrogen?
Hydrogen is the most abundant element in the universe, and Earth formed out of the same material as everything else in the solar system. Yet the air you are breathing is about seventy-eight percent nitrogen and roughly half a part per million hydrogen. Nitrogen stayed; hydrogen went.
The usual phrasing is that light gases are "blown away" or "boil off," which quietly smuggles in an agent — something doing the blowing. Nothing blows. So let us start by refusing that phrasing and asking what, precisely, removes one gas from a planet and leaves another.
Reasoning it through
REASONING #Begin with the number that seems to settle it. Earth's escape velocity is 11.2 km/s at the surface, and about 10.8 km/s at the exobase, the altitude near 500 km above which an upward-moving atom is unlikely to hit anything else on its way out.
Now the speeds up there. At an exospheric temperature around 1000 K, the root-mean-square speed of atomic hydrogen is close to 5.0 km/s, helium about 2.5, molecular nitrogen about 0.94. Do you see the difficulty? Not one of them is anywhere near 10.8. If typical speed decided the matter, Earth would still have all its hydrogen.
So the mean cannot be what matters. What is the mean an average of?
A gas is not a population of atoms all moving at one speed. It is a distribution — Maxwell and Boltzmann's — with a tail that extends, thinly, to arbitrarily high speeds. There is always some fraction moving at 11 km/s, and always a smaller fraction moving at 15. Escape does not need the gas to be fast. It needs only that the distribution has a tail, and that the tail is above the threshold.
And here is the move the whole phenomenon turns on. Escape is not a force applied to the gas; it is a selection on it. The exobase acts as a threshold that removes exactly the members above a cut and returns nothing. The survivors then collide, re-randomize, and refill the tail from below — so the filter never runs out of candidates. It samples the same distribution again and again, keeping only the extreme.
Why does that turn a modest difference into a total one? Because the tail falls off exponentially in the square of speed. The relevant quantity is the escape parameter, the ratio of gravitational binding energy to thermal energy at the exobase, and the escape flux carries a factor of the exponential of minus that parameter. For atomic hydrogen at 1000 K it is roughly 8; for molecular nitrogen, over two hundred. Hydrogen is only about five times faster on average — but the fraction above the cut differs not by five, but by something like ninety orders of magnitude. Hydrogen leaks steadily. Nitrogen escapes at a rate that is not merely small but physically meaningless.
That is the general lesson hiding in the atmosphere: when a selective threshold sits far out on a distribution's tail, a small shift in the distribution produces an unlimited shift in what gets through.
The analogy
THE ANALOGY #Imagine a very high wall around a crowded square, so high that nobody in the crowd could ever climb it — but with a single narrow chute at the top that anyone moving faster than a sprint can be carried through. Almost nobody in the square ever reaches that speed. Yet the crowd keeps jostling, and every so often somebody does, and is gone. Two squares whose crowds differ only slightly in average briskness will empty at wildly different rates, because what matters is not the average walker but the rare sprinter, and sprinters are exponentially rarer as the required speed rises.
in the square the crowd only thins, whereas Earth's hydrogen is continuously resupplied from below by water vapour and methane, so the escape is a steady-state leak rather than a one-off emptying — and, as it turns out, the leak rate is set by how fast hydrogen gets delivered upward, not by the chute at all.
Clarifying the model
THE MODEL #Three refinements matter, and the third partly overturns the first.
The thermal, tail-driven mechanism just described is Jeans escape, and it is real. But it is not the only channel. Charge exchange with the solar wind, and outflow of ions along open field lines over the poles, both remove atmosphere without any appeal to the thermal tail. On Mars the non-thermal channels appear to dominate.
Second, helium ought to be a puzzle. It is light, its escape parameter is modest, and it should be long gone — yet it sits at about five parts per million. The answer is resupply: radioactive decay of uranium and thorium in the crust manufactures helium continuously, so the atmosphere holds a small standing stock that is constantly refilled and constantly lost.
Third, and most important, Earth's hydrogen loss is not actually limited by the exobase filter. It is limited far below, at the tropopause cold trap, where air chilled to around 190 K freezes out almost all its water before it can rise into the stratosphere. Very little hydrogen ever reaches the exosphere, so the escape rate — of order a few kilograms per second — is set by the delivery bottleneck, a picture usually credited to Hunten's diffusion-limited escape argument. The tail selection is still what does the removing; it simply is not what sets the pace.
There is a lovely confirmation of the mechanism, though. Deuterium is twice the mass of ordinary hydrogen and so sits further from the cut, escaping less readily. Bodies that have lost a lot of hydrogen should therefore be enriched in deuterium relative to their starting material — and Mars and Venus both are, dramatically so.
A picture of it
THE PICTURE #How to readThe bars are typical speeds of four gases at exospheric temperature; the flat line is the escape speed. Every bar falls far short of it, which is the point. Escape happens entirely in the invisible tail above each bar, and how much tail clears the line falls off so steeply with mass that hydrogen leaks while nitrogen effectively never does.
What became clearer
WHAT CLEARED #Atmospheric composition is not decided by what a planet's gravity can hold on average; it is decided by what a threshold, sampling the same distribution over and over, keeps removing from the extreme. Once you see escape as selection on a tail rather than a force on a bulk, the enormous gap between hydrogen and nitrogen stops being surprising and becomes arithmetic.
Where to go next
ONWARD #- Hydrodynamic escape, where a strongly heated upper atmosphere flows outward as a bulk wind and drags heavy species along with the light ones, breaking the selectivity entirely.
- Why Titan, with far weaker gravity than Earth, keeps a thick nitrogen atmosphere — and what its very low temperature does to the escape parameter.
- Isotope ratios as a fossil record of loss: what D/H in Venus's atmosphere implies about how much water it once had.
Key terms
TERMS #| Term | What it means |
|---|---|
| Exobase | the altitude above which collisions are rare enough that an upward-moving atom generally escapes without being deflected. |
| Maxwell-Boltzmann distribution | the equilibrium spread of molecular speeds in a gas, with an exponentially thinning high-speed tail. |
| Jeans escape | thermal escape driven by the fraction of that tail exceeding escape velocity at the exobase. |
| Escape parameter | the ratio of gravitational binding energy to thermal energy at the exobase; the escape rate falls off exponentially in it. |
| Cold trap | the tropopause temperature minimum that freezes water out of rising air, limiting how much hydrogen reaches the upper atmosphere. |
| Diffusion-limited escape | the regime in which loss rate is set by upward transport of the light species, not by the escape mechanism. |
Every term the collection defines is gathered in the glossary.