The next 25 pages present an on-line guide to dynamic neotectonic modeling with Shells.
"Dynamic modeling" (also called "forward modeling") means using physics, assumed rheologies, and boundary conditions to predict what a planet should do, according to our current scientific paradigm. These models can be adjusted to come close to observed geodetic velocities and stress directions (et cetera), but only by trial-and-error. (If you want to use geologic and paleomagnetic data to directly reconstruct what the planet actually did, or if you want to use geologic and geodetic and seismic data to see what the planet is actually doing, then you want to do "kinematic" modeling. For paleotectonic kimetic modeling, you should see Bird (1998) and consider my program Restore. For neotectonic kinematic modeling, you should consider my program NeoKinema.)
"Neotectonic dynamic modeling" means using present topography, heat-flow, and fault networks to compute velocities, fault slip-rates, strain-rates, and stresses. All of these results are valid when considered as long-term averages over several earthquake cycles; the velocities can be easily corrected for temporary fault locking between earthquakes by a little extra computation. Neotectonic models do NOT compute finite strain, modify the topography or lithosphere structure, or generate new faults. (For finite strain and the evolution of topography and lithosphere structure, you could try my continuum dynamic modeling program Laramy. Currently, I do not offer any program to simulate the formation of new faults.)
Shells is the only one of my dynamic modeling programs to use spherical coordinates, and so it is the only code which is accurate for modeling large regions or whole planets. However, for modeling small regions, codes Plates and Faults may have a small advantage because they represent fault shapes more accurately. If you decide to use one of these you should still read through this Guide, because most of the steps will be very similar. (Only a few program names are different, such as DrawGrid instead of OrbWin, FillGrid instead of OrbData5, and Plates2AI or Faults2AI instead of FiniteMap or OrbMapAI.)
If you eventually choose to download any of these programs and use them, I expect you to:
Send me (email@example.com) your name, institution, e-mail address, and a few words about your intentions, so that I can add you to my distribution list for future notices.
Acknowledge the source of these programs in any proposals, seminars, or papers you may produce.