Evolution of Planetary Atmospheres

In the last few years exciting discoveries about the possibilities of extra-solar planets have been make. There is now good reason to believe that planets are commonly present around other stars. This leads almost immediately to the question of life on those planets. One of the requirements for life as it exists on Earth is an atmosphere. Therefore understanding the processes that create and alter planetary atmospheres is essential to determining the requirements that a planet much meet to be able to sustain life.

This page was made as part of a final project for the Introduction to Planetary Atmospheres course at CU Boulder. The object of this project was to look at the processes of atmospheric creation and evolution and simulate a number of these processes with a simple model that can determine their relative importance over long periods of time. To make this tractable I only consider terrestrial type planets during the time after the majority of the planetary mass has accreted.

Obviously, the problem of atmospheric formation is closely linked to that of planetary formation. Unfortunately there is very little that is definitively known about the relationship between these two processes. Some theories contend that at the late stages of accretion there was still a thick solar nebula present and that a large amount of gas was gravitationally captured directly from it. Others argue that the entire volatile balance of the terrestrial planets was brought in by bodies like comets from further out in the solar system after the majority of the accretion process had ended. As is often the case, the truth probably lies between these two extremes.

Due to the simplicity of this model, however, it is the later of the extreme theories that will be tested because it doesn’t require making assumptions about the initial atmosphere and allows us to focus more completely on the evolutionary processes. The processes that this model includes are Jean’s escape, hydrodynamic escape, sputtering, and impact effects. You can follow the links below to see full explanations of each of these effects and how it is implemented in the model. Many of these pages contain Netscape 3.0 or newer, IE 3.0 or newer, or HotJava. Not all browsers are created equal when it comes to running Java applets, however. While most of my applets seem to work fine on any of these browsers, I have found some odd problems with Netscape 3.01 for Macintosh and I’m sure that many other problems might exist with other browsers.

Jean’s Escape is a process in which particles in the atmosphere gain enough energy to exceed the escape velocity of the planet and leave it far behind.

Hydrodynamic Escape occurs when light particles rise up through the atmosphere due to buoyancy and push larger particles out with them.

Sputtering and other non-thermal loss processes occur when high energy particles from outside of the atmosphere strike atmospheric particles giving them the energy to escape.

Impact erosion occurs when a large impactor hits the planet. If the energy of the impact is sufficient, it can literally blow away a sizable fraction of the atmosphere.

Impact deposition is what occurs when the impactor doesn’t have enough energy to remove a large section of the atmosphere. Instead, the volatile components of the impactor become part of the atmosphere.

To look at these processes and how they interact over time I wrote a program that implements simple versions of each of them. The simulation was written in Java so that viewers can "play" with it as well. Unfortunately it does things that for some reason many Java Virtual Machines that I have seen don’t handle properly. I have found that it doesn’t work with Netscape 3.0x as well as IE 3.02. It does work with the HotJava browser from Sun which can be downloaded for free as well as IE and Netscape 4.0x versions. Even if you have a browser that won’t run the applet properly right now you should click here to read the description of it and some of what I have found with it as well as more information on the parts of the model that aren’t directly linked to the evolutionary processes.

To learn more about the author of this page visit my home page.