Tutorial: Hazard Curve Calculator

This tutorial demonstrates how to replicate the hazard curve figures presented in:

Field, E. H, N. Gupta, V. Gupta, M. Blanpied, P. Maechling, and T. H. Jordan (2005), Hazard calculations for the WGCEP-2002 earthquake forecast using OpenSHA and distributed object technologies, Seismological Research Letters, 76, p. 161-167, DOI: 10.1785/gssrl.76.2.161.



Start by opening the required application:

...and follow the instructions below to reproduce various figures:

Figure 2

  1. Choose the IMR (or attenuation relationship) by choosing Campbell and Bozorgnia (2003) from the first list in the "Set IMR" window.
  2. Set the IMT (intensity-measure type) by choosing "PGA" from the first list in the "Set IMT" window.
  3. Set the latitude and longitude in the "Set Site Params" window to be 37.8 and -122.417, respectively. This is a site near downtown San Francisco that happens to be a class "B" site according to the site-conditions map of Wills et al. (2000). An alternative way to set this latitude and longitude is to select "Sites of Interest" from the "Control Panels" list at the bottom of the application, and then choose "San Francisco Class B" from the window that pops up.
  4. We now want to set the "Campbell-2003 Site Type" parameter below the latitude and longitude. We can set this to any of the options listed. However, we'll set this to a value consistent with the Wills et al. (2000) classification map by doing the following: from the "Control Panels" list at the bottom of the application, choose "Set Site Params from Web Services". In the window that pops up, first select "All IMRs" in the drop down box at the bottom, then click "Set IMR Params". Doing this requests the Wills et al. (2000) classification for our location from a Web Service, and the value that is returned ("B") is translated into the appropriate value for the "Campbell-2003 Site Type" (the parameter should now have been automatically set to "Firm Rock").
  5. Choose an ERF (Earthquake Rupture Forecast) by doing the following: click the "ERF & Time Span" tab at the top of the application, and under the "Eqk Rup Forecast" list at the top of the sub-window, choose "WG02 Fortran Wrapped ERF List" (you might need to scroll down to see this option).
  6. In the "Epistemic List Control" that automatically pops up, click the box next to "Plot Average" (the box next to "Plot all curves" should already be selected). Also choose "Plot Fractiles" from the list below, and click "Update Fractile List" at the bottom of the popup window. The window will automatically close when done.
  7. In the main application, set the various ERF's adjustable parameters (listed below the chosen ERF) as follows:

    Rupture Offset = 10.0
    Fault Discretization = 2.0
    Delta Mag = 0.2
    Backgound Seismicity = "Exclude"
    GR Tail Seismicity = "Include"
    Num Realizations = 100

  8. Initiate the computation by clicking "Compute" at the bottom of the application. It will take several minutes to generate all the curves for the 100 realizations. The progress bar should keep you posted.
  9. Convert to a log-log plot by clicking both the "X" and "Y" check-boxes next to "Log scale" at the bottom of the application.
  10. Change the color scheme by clicking "Plot Prefs" at the bottom of the application and in the popup window set the following for each Dataset:
  11. DATASET #1 Curves: Color as Gray and Size as 2
    DATASET #1 Fractiles: Color as Blue and Size as 2
    DATASET #1 Mean: Color as Red and Size as 2

  12. Click "Done". Change the axis limits by clicking "Set Axis" at the bottom of the application; then choose "Custom Scale" from the top of the popup window and set "Min X" = 1.0E-2, "Max X" = 3.0, and "Min Y" = 1.0E-6. Click "OK" to apply the changes and close the window.

    The window you now see represents the screenshot shown in Figure 2.

Figure 3

(assumes completion of steps 1-11 above)

  1. Put the present plot aside by clicking "Peel Off" at the bottom of the application (or click "Clear Plot").
  2. In the main application, make it so we'll only plot the mean curves by doing the following: in the "Control Panels" list at the bottom of the application, choose "Epistemic List Control" (you might need to scroll down to see it). Deselect "Plot all curves", and choose "No Fractiles". Only the "Plot Average" option should remain selected.
  3. Re-compute the average shown in Figure 2 (for Campbell and Bazorgnia (2003) by again clicking "Compute" at the bottom of the application.
  4. Now redo this calculation using a different IMR (attenuation relationship) by doing the following: click the "IMR, IMT & Site" tab at the top of the application (if you haven't already done this), under the "Set IMR" window choose "Boore et al. (1997)" from the list at the top, set the IMT to "PGA in the "Set IMT" window below, and click "Compute" at the bottom. Note that the site-type parameter under the lat and lon was already set in step (4) above.
  5. Repeat step (15) for the "Abrahamson and Silva (1997)" and "Sadigh et al. (1997)" attenuation relationships in that order (and be sure to set the IMT to PGA each time).
  6. After increasing the size of the lines to 2.0 (by clicking "Plot Prefs"), and changing the axis range (by clicking "Set Axis"), you have now reproduced Figure 3. Note that you can save a png version of the image by selecting the "Save" option in the "File" menu at the upper left corner of the application or by clicking on the save icon symbol at the top left of the application.

Figure 4

(assumes completion of steps 1-17 above)

  1. Click either "Peel Off" or "Clear Plot" at the bottom of the application.
  2. In the main application, re-select the Campbell & Bozorgnia (2003) IMR (attenuation relationship), set the IMT to PGA, and click "Compute" at the bottom of the application. This re-plots the mean curve shown in Figure 2.
  3. Click the "ERF & Time Span" tab at the top of the application, and on the right side change the time-span "Duration" from 30 to 20 years. Click "Compute" at the bottom of the application.
  4. Repeat step (20) for durations of 10, 5, and 1 (in that order).
  5. After increasing the size of the lines to 2.0 (by clicking "Plot Prefs"), and changing the axis range (by clicking "Set Axis"), you have now reproduced Figure 4.

Figure 5

(assumes completion of steps 1-22 above)

  1. Click either "Peel Off" or "Clear Plot" at the bottom of the application.
  2. In the main application, re-select the duration of 30 years and click "Compute" (to re-plot the mean curve shown in Figure 2).
  3. For each of the 7 ERF adjustable parameters labeled "[fault name] Prob. Model Wts", click the button and set all values in the popup window to zero, except set that for the "Poisson" model to 1.0. After dong this for the 7 parameters (one for each fault), click "Compute" at the bottom of the application.
  4. Repeat step (25), but rather than giving the "Poisson" model a weight of 1.0, give the "BPT", model a weight of 1.0 in each of the 7 parameters.
  5. After increasing the size of the lines to 2.0 (by clicking "Plot Prefs"), and changing the axis range (by clicking "Set Axis"), you have now reproduced Figure 5.