How do scientists study the past atmosphere?

[neginf]

How do scientists determine levels of atmospheric gases in the past ?

 

[Faloren]

Usually ice cores. Antarctic ice can be millions of years old and gases dissolved or simply trapped as bubbles in water are preserved over time. When the ice is bored out and examined, it's possible to look at the traces of different gases present - this is frequently done for carbon dioxide measurements.

Other materials can be sediments, rocks, etc. Depending on the time scale of observation, you need to look at different substrates.

As usual, http://en.wikipedia.org/wiki/Paleoclimatology#Reconstructing_ancient_climates, Wikipedia has a good (but not necessarily credible!) outline to give you some ideas.

 

[Studiot]

As in any discipline it is unwise to rely on a single source of evidence.

Much new work in paleobiology has emerged on this topic in recent years (beware older unsupportable theories).

The Emerald Planet
by
D Beerling

Provides a good modern introduction to this subject.

 

[klimatos]

Usually ice cores. Antarctic ice can be millions of years old and gases dissolved or simply trapped as bubbles in water are preserved over time. When the ice is bored out and examined, it's possible to look at the traces of different gases present - this is frequently done for carbon dioxide measurements.

Yes, but the gas bubble or ice layer today does not necessarily have the same gases in it or in the same proportions as when the bubble or layer was first formed. Ice bubbles continually interchange components with the surrounding ice; and molecules are continually migrating from one layer to another. At normal Antarctic ice temperatures and bubble pressures, the average number of gas molecule collisions with the surrounding ice walls is some 3e+27 times per square meter per second. Some of those molecules "stick" and become part of the surrounding ice. At the same time, vaporization and outgassing from the bubble walls add to the air in the bubble.

Similarly, molecules of both water and dissolved gases migrate from one layer to another over time. We have to be very careful not to assign unrealistic measures of either accuracy or precision to such measurements of temperature proxies.

 

[Studiot]

Analysis of some fossil soils also yield evidence since some minerals interact with the atmosphere (eg rusting) so relative percentages of reaction products can tell us about the atmosphere at the time of fossilisation.

 

[Evo]

This might help.

http://www.ux1.eiu.edu/~cfjps/1400/atmos_origin.html

 

[klimatos]

the average number of gas molecule collisions with the surrounding ice walls is some 3e+27 times per square meter per second.

Oops, jet-lag error. It should read 3 x 10²⁷, not "3e+27"

 

[Studiot]

There are a lot of probablys and perhaps' and might haves in you reference, Evo.

It is still a good reference, but it should be borne in mind that it mixes unproven theory (speculation) with fairly well proven theory.

There are other competing theories about.

 

[Evo]

There are a lot of probablys and perhaps' and might haves in you reference, Evo.

It is still a good reference, but it should be borne in mind that it mixes unproven theory (speculation) with fairly well proven theory.

There are other competing theories about.

Yes Studiot, I agree, I couldn't find two theories that completely agreed, this one seemed to have more items that appeared in other places, but it's overly simplified and incomplete.

Here is another viewpoint from NASA. (yes, I like the stuff for kids)

http://scijinks.nasa.gov/atmosphere-formation