Download & Support

Download Coulomb Software and User Guide

Coulomb FAQ

  • What is Coulomb?
  • Coulomb 3.0 is designed to investigate Coulomb stress changes on mapped faults and earthquake nodal planes, and is intended both for publication-directed research and for university teaching and instruction
  • What is Coulomb used for?
  • One can calculate static displacements (on any surface or at GPS stations), strains, and stresses caused by fault slip, magmatic intrusion or dike expansion. Problems such as how an earthquake promotes or inhibits failure on nearby faults, or how fault slip or dike expansion will compress a nearby magma chamber, are germane to Coulomb. Geologic deformation associated with strike-slip faults, normal faults, or fault-bend folds is also a useful application. Calculations are made in an elastic halfspace with uniform isotropic elastic properties following Okada [1992].
  • How can I run Coulomb 3.0 on my computer?
    1. Install a latest version (recommended) Matlab.
    2. Download and unzip Coulomb software.
    3. Set current directory of Matlab as Coulomb directory
    4. Type coulomb in Command line of the Matlab.
  • Where can I get a username and password?
  • You need to send an email to vsevilgen@usgs.gov with the subject as coulomb registration. You will receive username and password.
  • How do I change from Cartesian (x, y) to Lon/lat coordinates, or the other way around?
  • First, lon/lat information must be included in the input file for this to be possible. If so, there are two methods to toggle between Cartesian and lon/lat: (1) Input > Preferences > ‘Coordinates’ button. Then reload the input file by Input > Open existing input file. Or, (2) in the 2D grid view, click on the map frame using the right mouse button. A window pops up, choose ‘change coordinates’ and make a left mouse click.
  • Is the regional stress used in 'specified fault' or ?calculate stress on faults? calculations?
  • No, it is only used on the optimally-oriented stress calculations (opt Strike-slip, opt Thrusts, opt Normal, or Opt Fault). It is ignored in all other calculations, except for the ?calculate stress on nodal planes? option, after loading a focal mechanism catalog. In this case, you can see the impact of the regional stress only by examining the resulting Excel file.
  • Can one make a dipping cross-section (a cross-section that is not vertical)?
  • You can do so for stress cross-sections, but not for displacement or strain cross-sections. For stress sections, the ?Cross section control panel? has a field for dip that you can change (the default is 90). This permits you to calculate the stress change on the source fault surface or a receiver fault surface.
  • I am having trouble creating input files in my text editor that run properly in Coulomb; perhaps I don?t have the formats and spaces exactly correct. What should I do?
  • Either build your file by modifying an existing input file by Input > Open and edit existing input file, and then save it with a new name, or choose Input > Build input file from x & y map, or Build input file from lon. & lat. map, and create a new one. Then all formats will be automatically correct and all parameters will be properly labeled.
  • To get the fault to dip in the correct direction, I entered a negative dip, but Coulomb objects.
  • The dip must be positive. All you need to do is make the dip positive, and swap the (start-x, y) and and (finish-x, y) values in your input file. This follows from the ?right-hand rule? explained and diagrammed in the User Guide.
  • Do the input file slip values always correspond to strike-slip and dip-slip?
  • No, they do not. Kode 100 shear dislocation sources also have a (net slip, rake) representation so that one can input a receiver fault with a rake. For sources that are not Kode 100 (such as dikes, sills, or point sources of inflation), these two parameters can represent opening or potency. To see the proper labels for any of the tutorial files, Input > Open and edit existing file.
  • Since the fault dip is parameterized in the input file by the top and bottom vertical depth of the fault, how do I input horizontal (decollement) faults or volcanic sills?
  • If you make the bottom depth slightly greater than the top depth, you can input near-horizontal sources, with a dip less than 1i. But it is not possible to input a 0i dip except in "calculate stress on a point", or "calculate stress on nodal planes"
  • The default friction seems to be set to 0.4. What is the basis for this, and shouldn't it vary for different types of faults?
  • We use 0.4 simply because it is midway between the likely extremes of 0.0-0.8. Our work has led us to expect that , in general, smooth faults with a large amount of cumulative slip (tens to hundreds of kilometers) exhibit low friction, where as young faults or those only kilometers of cumulative slip exhibit high friction, since they are more highly fractured and thus less smooth (Parsons, JGR 1999; Lin & Stein, JGR 2004). This is an inference, not a certainty. Also, remember that you are specifying the friction of the receiver fault(s), not the source fault.
  • I am running a variable-slip source fault and a tiled (subdivided) receiver fault, but the calculations take very long. Am I doing something wrong?
  • Variable-slip sources with numerous patches greatly increase the calculation time, and the Stress > ?Calc. Stress on fault? option takes even longer; together they tax the computer very heavily. Our advice is to start simple and only increase the complexity of the source and receiver after you have first run more idealized models. Once the receiver fault is more than several patch or fault lengths from the source, the source variability adds no additional accuracy to the result, and so unnecessarily slows the process.
  • I am getting a string of error messages in the MATLAB command window. What should I do?
  • First, re-run the calculation. If the messages persist, make sure that one of the tutorial input files runs properly. If it does, you have an error in your input file. If it does not, quit and relaunch Coulomb. If the messages still persist, quit and relaunch MATLAB and then relaunch Coulomb. If all else fails, send us a screen shot of the error message and the input file and we will test it.
  • How do I make left and right mouse clicks if I am using a one-button mouse?
  • Hold the ?control? key down when you click the mouse button. But we recommend that you buy a three-button mouse.
  • If I have lat/lon coord's (rather than Cartesian coord's) for my source or receiver fault, how do I enter it into Coulomb?
  • Choose Input > Build input file from lon. & lat. map . You can enter either an Aki-Richards fault endpoint or the fault center. Then you enter the fault length, width, strike, dip, and rake. Coulomb will generate the fault coord's. This is explained in the User Guide section 4.11.
Coulomb participated institutions AIST USGS WHOI USAID OFDA