Lab 4. Earthquake location To complete this exercise you will need to (a) Complete the table on page 2. (b) Identify phases on the seismograms on pages 3-6 as requested on page 11. (c) Locate the earthquake manually on a globe as requested on page 12-14 and write down your location on page 14. (d) Plot the location of earthquakes in the region and locate the earthquake using tools on the world wide web as described on page 15, print out a map of your result (1 per student) and write down your location. (e) Answer the question on page 15. (f) Complete the exercise on starting on page 16 to locate a second earthquake in South America. For this you will (i) generate a plot showing travel time residual versus location for one choice of earthquake depth, (ii) complete a table showing the travel time residual at different depths, (iii) make a plot of residual versus depth, and (iv) comment briefly on your results. (L. Braile, November, 2004) braile@purdue.edu web.ics.purdue.edu/~braile Introduction: This education module uses real seismograms to demonstrate the S minus P earthquake location method and the determination of magnitudes for earthquakes. The seismograms for the S minus P location exercise were recorded at standard GSN (Global Seismographic Network; www.iris.edu) stations and downloaded from The IRIS DMC (Data Management Center) archive. The seismograms for the magnitude determination exercise were recorded on an AS-1 seismograph. Last modified November 21, 2004 The web page for this document is: http://web.ics.purdue.edu/~braile/edumod/as1lessons/eqlocation/eqlocation.htm. Funding for this development provided by IRIS and the National Science Foundation. Copyright 2004. L. Braile. Permission granted for reproduction for non-commercial uses. Modified by William Wilcock for Ocean/ESS 410. 1
(a) S minus P location exercise, Oaxaca earthquake: Measuring S minus P times of the Seismograms (Paper copies, 20 minute records, Scale: 1 cm = 1 minute): Measure the S minus P times for the 4 seismograms below (Figure 1-4); then use the standard Earth travel time curves (examples are shown in Figures 5 and 6; for this exercise use the S minus P graph, Figure 7, which is plotted with the same time scale as the seismograms) to infer the distance (in degrees and kilometers) from the epicenter to the station; use Table 1 below to record your results. To find the distance corresponding to a given S minus P time, move the metric ruler along the S minus P travel time graph (keeping the ruler parallel with the time axis) until the S minus P time matches the time that you entered in the Table 1. The interpreted distance will then be the position of the ruler on the distance axis (in degrees) as illustrated in Figure 8. Because the S minus P times for the velocity structure that best match the Earth, increase consistently with distance, the observed S minus P time will only match the curves shown in Figure 7 at one position corresponding to a specific distance (the interpreted epicenter-to-station distance). Additional illustrations of this concept are shown in Figure 8 and 9 for a seismogram (in this case the KIP seismogram for the Oaxaca earthquake) plotted using the same time scale as the travel time curve graphs. A map of the S minus P results (from Table 1) showing the calculated location of the earthquake (using triangularization) can be made using a globe (see section 3.1, below) or using the online IRIS DMC Event Search mapping tool (see section 3.2, below). Table 1. Data table for the S minus P earthquake location information. Station (Latitude and Longitude, in degrees): TUC (32.310, -110.785) Tucson, AZ CCM (38.056, -91.245) Cathedral Cave, MO NNA (-11.987, -76.842) Nana, Peru KIP (21.423, -158.015) Kipapa, HI Measured S minus P times (minutes; measure to nearest tenth of a minute = 1 mm on the seismogram): Inferred distance (degrees and kilometers; convert degrees to km by multiplying by 111.19 km/degree): Degrees: Kilometers: 2
Figure 1. Oaxaca earthquake seismogram recorded at station TUC. Seismogram displayed using the AmaSeis software. Time scale (if printed from the MS Word document) is 1 cm = 1 minute). 3
Figure 2. Oaxaca earthquake seismogram recorded at station CCM. Seismogram displayed using the AmaSeis software. Time scale (if printed from the MS Word document) is 1 cm = 1 minute). 4
Figure 3. Oaxaca earthquake seismogram recorded at station NNA. Seismogram displayed using the AmaSeis software. Time scale (if printed from the MS Word document) is 1 cm = 1 minute). 5
Figure 4. Oaxaca earthquake seismogram recorded at station KIP. Seismogram displayed using the AmaSeis software. Time scale (if printed from the MS Word document) is 1 cm = 1 minute). 6
Figure 5. Standard Earth travel time curves for a source depth of 0 km (can be used for shallow earthquakes at distances of ~20 to 120 degrees). Travel times for many different phases (types of seismic waves and paths through the Earth) are shown. The first (or direct) P and S arrival times are shown by heavier lines. Note that the difference between the S and the P times increases smoothly with distance. Therefore, a seismogram with a given S minus P time will only match the travel time data at one specific distance. 7
Figure 6. Simplified standard Earth travel time curves showing only the P and S times. 8
Figure 7. Simplified standard Earth travel time curves showing only the P and S times (the difference between the P and S times shown in Figure 6; time scale: 1 cm = 1 minute). To find the distance corresponding to a given S minus P time, move the metric ruler along the S minus P travel time graph (keeping the ruler parallel with the time axis) until the S minus P time matches the time that you entered in the table for a given seismogram. The interpreted distance will then be the position of the ruler on the distance axis (in degrees). 9
Figure 8. Illustration of ruler placed on S minus P travel times graph (Figure 7) showing position corresponding to 8 minutes (8 cm on the scaled seismograms and S minus P graph shown above) indicating that the epicenter-to-station distance for this seismogram is about 58 degrees. Figure 9. Overlaying a seismogram (the KIP seismogram, Figure 4; plotted using the same time scale as the underlying graph) on the standard Earth model travel time curves. Similar to the measurement illustrated in Figure9, this diagram shows that the S minus P arrival times indicate an epicenter-to-station distance of about 58 degrees. The AmaSeis travel time curve tool provides a similar display (Figure 10) although the graph is rotated so that the seismograms are plotted horizontally. 10
Figure 10. KIP seismogram for the Oaxaca earthquake displayed in the AmaSeis travel time curves window. The seismogram is moved (by dragging with the mouse cursor) until the P and S arrival times match the travel time curves. The epicenter-to-station distance that corresponds to the interpreted S minus P times is displayed to the left of the seismogram. (b) Identifying other phases Using the travel time curves on Figure 5, identify phases other than the P and S waves on the other seismograms - this is not entirely straightforward because sometimes phases are not visible and there are a couple of phases visible on the seismograms which do not clearly correspond to anything plotted on Figure 5. 11
(c) Mapping the S minus P information: S minus P mapping on a globe: The epicenter location can be determined from the S minus P data by drawing distance circles on a globe as illustrated in Figures 13-15. In class we have a globe that you can mark with chalk Use a line of longitude (Figure 13) to provide a distance scale (in degrees) to determine the length of the string corresponding to the epicenter-to-station distance for each seismogram. The distance in degrees (or angular distance or geocentric angle) is illustrated in Figure 12. The distance in km (along the surface) can be found by multiplying the distance in degrees by 111.19 km/degree. Draw a circle or part of a circle around each station (Figure 14) using the appropriate string length from your Table 1 data and the degree scale (Figure 13). The results should be similar to the map display shown in Figure 15. The circles should approximately intersect at a point. One can compare the S minus P determined epicenter with the official epicenter (calculated from arrival times from over a hundred seismograph stations) by plotting the official epicenter location on the globe (Figure 15). For the Oaxaca earthquake, the epicenter is reported as: 16.01 degrees North latitude, and 96.93 degrees W longitude. An additional example of mapping an S minus P earthquake location on a globe is illustrated at: http://www.eas.purdue.edu/~braile/edumod/as1mag/as1mag2.htm. Figure 12. Cross section through the Earth showing the major spherical shells (layers; crust, mantle, outer core and inner core), selected raypaths of various seismic arrivals (phases), and the epicenter-to-station distance. 12
OCEAN/ESS 410 Figure 13. Determining distance in degrees on a globe using the lines of latitude. The string is measured for a distance of 40 degrees in this illustration. 13
OCEAN/ESS 410 Figure 14. Plotting an S minus P distance circle from a station. The length of the string is measured to the interpreted epicenter-to-station distance (in this case about 35 degrees) for station NNA (yellow dot) and an arc (circle or part of a circle) is drawn around the station. Figure 15. Results of S minus P location. The S minus P estimated earthquake epicenter is indicated by the intersections of the circles. Station locations are indicated by the yellow dots. The actual location is indicated by the red dot. Write down your location Latitude = Longitude = 14
(d) Locating an earthquake with S minus P times using online mapping tools: To determine the event location, go to the following link: http://www.iris.edu/quakes/make_event_search_map.pl?user_def=y This will give you an option to set map limits and to enter information in the User Defined Station box that you can fill in with your own data. Here is an example with the correct station locations but with the calculated radii and the earthquake location you calculated manually (User Defined Event) left blank. NOTE that the IRIS website accepts diameter, not radius! Figure 16. User-defined data entered into the IRIS DMC Event Search mapping tool. Now click the Make Map button which should be to the right of the map (if you cannot find this button, click on the very center of the map to get it to redraw). The map will redraw with circles at the specified radii from each station. The location is where they intersect. Choose the option to Download the Postscript Map and print a copy using GSView Write down your location: Latitude = Longitude = How well does your location agree with the one you calculated manually and with the official one (16.01 N, 96.93 W)? 15
(e) You are now going to locate another earthquake using the program EqLocate, which triangulates the earthquake location using just P-waves arrival times. To do this you need at least 4 stations (three to constrain X, Y and Z coordinates and the other to constrain the earthquake time). To run the software, which is a little clunky, follow these instructions: 1. On the class web site the Labs tab of the class web site right click on link Zip file for Lab Computers for Lab04, select Save Target As and save to c:\courses\ocn410\yourname. 2. With Windows Explorer navigate to the directory with the zip file (c:\courses\ocn410\yourname), right click on the lab04.zip file and select Extract All. 3. Click on EqLocateSetup.exe to install the software (chose Run not Extract All to do this) 4. Start EqLocate from the Desktop 5. Select the menu item File Open Event, navigate to c:\courses\ocn410\yourname\lab04\southamerica directory, select all the files, and click Open 6. A window should open with five waveforms. Pick the P waves using the left mouse button. There are zoom arrows which allow you to expand the vertical scale for each earthquake and a zoom arrow and scroll bar at the bottom which allows you to expand the time scale you will need this for accurate picks. Picks will appear as brown vertical lines. 7. Select the menu item Controls Set Depth and select a value of 0 km. (sometimes this window gets lost behind the help window but you can move it to one side to keep it visible). 8. Select the menu item Controls Set Maximum RMS Value and enter a large value of 200 s. 9. Use the arrow and zoom buttons on the map of the globe to zoom in on South America. (Also, sometimes the screen does not update when you overlay windows but you can solve this by successively clicking the zoom in and zoom out button). 10. Click on the screen at a trial location (hint: start near the station with the earliest P-wave). The upper left hand corner will display the latitude and longitude of this location together with the root mean square (RMS) travel time misfit (a measure of the average difference between the predicted time of the P waves and the time you selected. If this is less than 200 s a small splotch of color will appear representing the RMS travel time misfit. Blue and green vertical lines will appear on the seismograms showing the predicted P wave and S wave times, respectively. 11. Repeat the last step many times building up a color contour map that should define a bulls eye pattern with the minimum RMS travel time misfit at the middle which defines the best fitting location. You will need to change the Maximum RMS value (say to 20 s) to get a tighter bull s eye. 12. Record the minimum RMS on the table below together with the latitude and longitude (you can get this by clicking in the middle of the bull s eye and looking in the yellow box at the top of the window the yellow box may be hidden behind the hints dialog box). 13. The next step is to print your solution but do not use the Print option in the Menu it crashes the program. Use the Alt-Print Screen on the keyboard to copy the EqLocate window to the clipboard. Now select Start All Programs Accessories Paint. In the application (1) choose Paste from the Edit menu; (2) chose Print Page Setup from the Paint menu (top left), set up the printing to Fit to 1 x 1 and select OK and then finally select Print Print from the Paint menu. 14. Repeat steps 8-12 for Depths of 200, 400, 600, 800 and 1000 km to complete the table on the following page (There is no need to print these plots) 16
Depth Latitude Longitude RMS travel time misfit 0 km 200 km 400 km 600 km 800 km 1000 km 15. Make a plot of residual versus depth on a piece of graph paper and hand-draw a smooth curve through your observations. 16. Comment on your results. What is the most likely depth of the earthquake and why? 17