SeisImager Instructions rev.4_08
SeisImager comes in two parts – Pickwin95, and Plotrefa. Pickwin95 is the program you will use to load in the .dat files taken in the field, pick first-breaks (the start time of the signal from the shot), and store the results.
PickWin Instructions
1) double click on Pickwin95: after a few seconds, a window comes up that says “No waveform data”
2) Go to the “File” menu: SEG2 file: Open a SEG2 file (ends in .dat). These are the files stored when you saved the files on the seismograph. Fred has edited them as necessary to be sure they have the correct shot and geophone locations and start times. You should have these files on a CD.
3) go to the folder containing the data and double click on the file you are interested in. Use the “new” data option. The data will appear in the window. Look at it, and check to see that the shot location and geophones are at the locations you think they should be in.
4) Go to the “File” menu again and load the data from the other spread for the same shot location on the same profile. Use the “add” data option. You will now see the complete data set for that shot point – as though we had a 24-geophone system instead of a 12-geophone system.
5) Check and change any bad header information (shot location, geophone location, etc.) by going to the “Edit” menu: Show/edit source/receiver locations, etc. A table of parameters is displayed. Be sure the numbers under Rec. Distance and “Source” are the expected values of feet.
6) Hit the “OK” button when all numbers are correct. The program will return to the travel time window.
7) If any of the traces have poor data or no data, you can click on the white arrow, and then on the trace to turn it red and then hit “Edit/Select”, “selected traces”, “delete” to make it go away. Don’t pick bad data.
8) Now pick the first arrivals: Go to the menu: “Pick First Arrivals”: and click “Pick first arrival time automatically”, alternatively, click on the red “P” in the menu bar. Vertical red lines appear where the program thinks the signal starts. While it’s pretty good, it often takes some fudging.
9) Use the red-over-green UP and Down arrows to the left side of the menu bar at the top to adjust the trace sizes until you can see why the program picked the arrivals where it did. Move any pick by just clicking where you think it should be (often based on the surrounding traces). You can also adjust the distance and time scales using the red-over-dark blue arrows. Adjust the traces until you can see the traces well for picking where the signal starts. It’s usually best to adjust the trace you are working on until you can see noise on the trace at times before the signal arrives.
10) Once you are happy with the picks, click on the red star in the menu bar, and a line will connect all your picks. This is NECESSARY before you do the next step.
11) If you’re still happy with the picks, you need to store your results. Menu - “File”:: and click on “Save SEG2 file”. Name the file something like “KN_N-S1_120”, standing for Kualoa Nancy profile N-S1 shots at 120 feet.
12) Now save your travel times: Menu – “File”:and click on “ Save first break pick file”. Use the same file name as above, except this will be a .vs file, rather than a .dat file. Once saved, the data plot and travel times can be re-opened and edited in PickWin at any time.
13) Before moving on, PRINT your data window with the travel time picks.
14) Now you’re ready for the next shot location.
15) Since we have three profiles with shots at 5 locations each, you should have 15 files when finished. This task should take about 3 hours.
DO THIS FOR ALL SHOT LOCATIONS FOR TUESDAY, April 15.
PLEASE let me know if you find errors or other problems with these instructions – or in the data files.
[the next section will be updated next week]
PlotRefa Instructions
Once you have picked all your first breaks and corrected the shot location and geophone locations, if necessary, you are ready to run the analysis software to determine the structure under each line.
This is done with PlotRefa, which uses an algorithm called the “delay time” method to fit a model to your data.
1) Generate an elevation file for each profile. Generate a text file of your elevations vs. distance along the line, where the first number is the distance from zero and the 2nd is the elevation in feet. You should probably change your elevations so that they are all positive. Only the relative elevations are needed, not the elevation above sea level.
for example:
5 2.210.0 3.4
20.0 3.6
….
2) start PlotRefa
3) in the file menu, open the travel time pick file (.vs) containing ALL your PickWin results for one profile. (there should be five files per profile) Looking at the traveltime curve, you should see the time picks you made in PickWin for each shot location.
3) in the “file” menu, read in the elevation file that goes with this line.
4) Use the “Time Term Inversion” method to generate a model for your data
Use only 2 layers.
a) in the travel time menu, click on “connect common source travel time curves”. This will connect data that were shot at different times, but belong on the same profile.
b) Use the two “longest” shots (at 0 feet and 245 feet) to calculate traveltime difference curves to help find the inflection points where rays enter the deeper layer at the crossover point.
1] in “traveltime” menu, choose “traveltime difference” curve”
2] choose a shot (245 for example) and a 2nd shot (125 for example)
3] a ‘reasonably’ flat blue line appears where the curves are shot in the same direction. Where this curve makes a significant change in slope is where travel times are beyond the crossover point.
4] Write down these distances for each shot in both directions where applicable.
c) in the “time-term inversion” menu, choose “assign layer 2 arrivals”
d) click on the dot where each gather changes from layer 1 to layer 2 (the cross-over distance) as determined above in both directions. The layer 1 arrivals should turn red and the layer 2 arrivals green.
e) in the “Time-term inversion” menu, click “do time-term inversion”
f) go to “ray tracing” hit execute, to generate rays for the model
g) adjust the scale of the plot to show the velocity model at appropriate scale.
h) go to “view” menu: show travel time curves: PRINT
show ray paths: PRINT
show time term: PRINT
5) the “Reciprocal Method” takes a lot of work, and I don’t think it improved the model much – at least for my picks. You can try this method if you like, or ignore it.
6) The tomographic method seems to work well.
A) In the “tomography” menu, choose: generate initial model
B) choose a depth velocities about like what you found above
C) choose “inversion”
D) choose “raytracing: execute”
E) plot everything as above. – Change parameters as you like and try it again.
Your write-up should include geological evaluations of the models obtained and some estimate of the validity of the models.
You should also answer the following questions:
1) The velocity of sound in air is about 1083 feet/sec. Is it possible that you chose this velocity as your direct arrival? If so, it would be extremely unlikely that the true layer velocity is that of air. What does this imply about the real velocity of the upper layer. What should you do to get a better estimate of the upper layer velocity?
2) From what you have determined about the history of this region, what do you think is the origin of the refracting layer?
3) Does the magnetometer data shed any light on what the origin of the fill might be?