How can the atmosphere rapidly change composition?

Around ~2.3 Billion Years ago, the Atmosphere changed radically: Cyanobacteria started pumping out Oxygen and the atmosphere gradually became Oxygen rich.

~2.3 Billion Years is a long time, even in a geological timescale, I want to see a great De-Oxygenation event take place on a more human timescale, say a thousand years. The change doesn't have to be permanent (though that would be nice) and should occur naturally (eg: reverse-cyanobacteria, all oxygen producing plants dying due to a cross-species virus, etc) but may not be geological (so no volcanoes). The change only has to reduce or remove Oxygen, other gasses don't matter.

To be clear: the change should be caused by something in the biosphere, that includes humans, but it should not be deliberate. The new atmosphere should be unbreathable, or at lease harmful over long periods.

The less time that it takes for the atmosphere to become unbreathable, the better.

  • $\begingroup$ Read Hal Clement's novels Half Life and The Nitrogen Fix. He had it take place over a human lifespan. His idea was that an enzyme emerged to work with a nitogen-based metabolism and this took over the complete biosphere. $\endgroup$ – JDługosz Jan 31 '16 at 22:00

Your timescale is difficult to achieve. If all photosynthesis stopped today but respiration continued as normal, it would take about 4,500 years for the oxygen content of the atmosphere to drop to near zero — see this Wikipedia article on the Earth's oxygen cycle. You need something that will actively remove oxygen from the atmosphere, not just stop it from being replenished.

  • $\begingroup$ That seems right, hence reverse-cyanobacteria (or something). We could consider 4,500 years as an upper bound. $\endgroup$ – Amziraro Jan 31 '16 at 16:58
  • $\begingroup$ You just need the oxygen go down to about 10% to have big problems, at least to humans, so in the case of a linear decrease, you need about half the time before the atmosphere will be harmful to life (not all life anyway) $\endgroup$ – Gianluca Feb 1 '16 at 9:46
  • $\begingroup$ What's the computation? Does it use a differential equation for the change in oxygen consumption as oxygen depletes? $\endgroup$ – can-ned_food Apr 22 '17 at 21:02

One way to speed up the process is to assume that stage two or equivalent (new gas being absorbed by the ocean) is long past and stage 3 (gas absorbed by land surfaces) has just reached its end. Assuming you are talking about a situation similar to the current day, keep in mind that we've only been studying this sort of thing for a relatively short while; any situation that has been relatively stable over the last few thousand years is generally going to be assumed to be inherently stable, even if a few scientists believe otherwise.

Another good way to make the process swift is to have it feed on itself. Just an off the top of my head example: say you decide carbon dioxide is going to be your gas of choice. Rising carbon dioxide levels, in the short term, cause plant growth which raises oxygen levels, which cause more fires, which raise carbon dioxide levels more (as you are freeing more than just the most recent carbon absorbed). If you have a mechanism that is consistently raising the carbon dioxide levels (and I am mostly ignoring global warming here) the O2 levels will first become dangerous for humans (symptoms of carbon dioxide poisoning in humans start at about 2% carbon dioxide in the atmosphere, IIRC), and then start to adversely affect plants.

A little more carbon dioxide can boost plant growth, but too much may be harmful. Research in controlled environments (greenhouses) seems to indicate that levels of carbon dioxide that are too high can stunt plant growth, and changes the chemistry of plants in a way that seems to make them more vulnerable to insects. If those are both are true, then at some point the plants that humans are relying on to consume carbon dioxide will become less effective, causing the process to accelerate.

Again, this is just an off-the-top-of-my-head example (assisted by a couple quick google searches). The basic idea, however, of having a process that feeds on itself, either directly or indirectly, would seem to be a good way to make the process as fast as you are looking for.


Since you are looking for a short term event, then in all probability you will need a geological event with some sort of positive feedback cycle.

The only one which comes to mind for me is to somehow trigger a massive release of methane clathrate hydrates: i.e. methane that is trapped in ice crystals and held in suspension below the permafrost and in deep oceans.

Methane clathrate hydrate is suspected to exist in huge quantities (billions of cubic feet of methane if released, the sort of quantities which would make it worthwhile to mine or extract for energy production), but methane is also one of the truly effective greenhouse gasses (far superior to CO2 and not too far behind water vapour; the number one greenhouse gas). Since it is in out of the way places and stabilized by cold and pressure, you will somehow need to remove the cold and pressure keeping it trapped.

Given that much of it is trapped in the oceanic beds, a geological trigger such as an undersea volcano or the action of the Earth's plates disturbing large beds might start the process. An asteroid striking the Earth will likely land in the oceans (70% of the Earth's surface is covered by ocean, so the odds are stacked in favour of an ocean strike), and that too could trigger a massive release of methane into the atmosphere.

The methane will absorb the sunlight and start raising the temperature, creating conditions that will trigger more methane clathrate hydrates melting in shallower waters, pumping more methane into the atmosphere and creating a positive feedback loop, eventually creating a massive methane release and changing the character of the Earth's atmosphere. The huge quantity of methane in the atmosphere will certainly amplify combustion events, and will cause great strain on terrestrial ecosystems, both through the change in temperatures and the new composition of atmospheric gasses. Bacteria which thrive on methane will rapidly expand into many ecological niches (think of the world being overrun with toxic pond scum), and the ecology will change in ways that will tend to stabilize in the new configuration. Less efficient types of photosynthesis would probably be pushed from the stage, resulting in the extinction of may different types of plants (up to 95% if this adversely affects C3 photosynthesis), and massive extinction in the animal kingdom will result as well.

  • $\begingroup$ If you have global warming result in sudden ocean turnover (isolated cold water at the bottom mixes with warmer surface water), you could release a whole lot of the methane without the collateral damage caused by an earthquake or tsunami. $\endgroup$ – Karen Feb 16 '16 at 18:19

So the closest example we have is the Great Oxygenation Event.

It's slightly implausible but lets say that a similar dramatic shift occurs with some new form of bacteria or maybe something interesting like a slime mold evolving. The bacteria consumes oxygen and plant life but uses a brand new metabolic cycle not seen before. Quite possibly the output of that new metabolic cycle would be highly toxic.

The new metabolic cycle of this bacteria consumes large amounts of oxygen but is otherwise hugely more efficient than any previously known and as a result it out-competes everything. It spreads rapidly across the globe simultaneously consuming plants (reducing oxygen production) and oxygen (directly reducing amounts in the atmosphere). at the same time toxic gasses are being produced. Rapid mutations also mean that it adapts to anything we try to throw at it in the same way as bacteria are adapting to antibiotics.

It forms a film over the surface of water and turns plants (and animals although their ability to move and immune systems do give them some defense until they start to starve) to mush, dissolving most of the world into a puddle of goo.

This combined effect means that in as little as a few centuries the upper levels of the ocean and the surface of most land has turned into a sticky sludge and oxygen levels in the atmosphere are severely depleted.


Lots of Sulphur Dioxide and Hydrogen Fluoride belching from a thing that's bigger than a supervolcano?

We do not know much about flood basalt eruptions of the sort that created the Deccan and Siberian traps. Do they build up slowly or does the Earth's crust open in a massive fissure over a mere decade, year, or day? And is what is emitted at the very start a pulse of noxious gasses even nastier than what it will be emitting over the next few thousand years? And was the Siberian Traps event the worst case? That event was at the "great dying", the Permian-Triassic divide, where something like 90% of the then extant species went extinct.

For this story onset is very sudden, very noxious, and the Siberian traps were not the worst case. The largest mass extinction in geological history has started. Over to you.

You might want to research the volcanic eruption in Iceland in 1783 that killed a quarter of Iceland s population and inflicted acid rain that scorched leaves off trees in Ireland and a European harvest failure. Then dial that up to eleven on a global basis with no prospect of the eruption stopping anytime in the next century.

  • $\begingroup$ OP said "may not be geological (so no volcanoes)". $\endgroup$ – WhatRoughBeast Jan 31 '16 at 18:55
  • $\begingroup$ Sorry, missed that. Feel free to vote to delete if you wish. $\endgroup$ – nigel222 Jan 31 '16 at 19:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.