Continuing the seismogram generation

One of the ways of checking the results is to generate 3D plots of rays computed in the model. We used graphic package D3 by Ludek Klimes to generate following plot:

As there is too much rays in the figure, we have generated following four figures, each figure containing just the rays of a single elementary wave:

For the purposes of this sample computation, we are satisfied with the results obtained for first two elementary waves. This is not the case of the fourth and the third wave. We can see, that the rays refracted under the first interface were computed during both the computations of the third and of the fourth elementary waves. These weakly refracted rays satisfy the part 1 1 1 of both the ray codes for the third and for the fourth elementary waves. Thus we changed the parameter IPOINT to -1 for the third wave. The value -1 of IPOINT allows only the rays satisfying the whole ray code to be considered as the two-point rays.

We can also see that the rays of fourth wave, which transmitted through the salt to the bottom interface and then back through the salt, were computed only for receivers at both ends of the profile. We can guess, that these rays do exist also for receivers in the middle of the profile. We thus generate a plot of ray take-off angles in the ray-parameter domain. To do this, we run program RPPLOT and we obtain following plot, where successful rays are coloured and other rays black:

We see, that the rays refracted under the first interface form a large thin blue ellipse. The rays refracted through the salt occupy only very small area located in the middle of this ellipse. There are no successful rays in the left-hand half of the figure. Thus we can reduce the domain of ray take-off parameters to that part displayed as the right-hand half of the figure.

We thus change the data for ray parameters to the form written in file 98-rpa2.dat and we change the CRT input data for ray codes to the form of 98-crt2.dat.

Now we repeat the computation of synthetic seismograms with new input data. Two-point rays are again computed in the first step of the computation. We run the program CRT with the data file 98-rpa2.dat for specification of ray parameters. Then we again generate a plot of ray parameters running RPPLOT with input data 98-plot1.dat:

We can again generate 3D figures of rays:

Finally we can generate plots of the computed synthetic seismograms. Note, that the amplitude scale of following figures is enlarged compared to previous seismogram plots. This enables also some of the later arrivals to be visible.

We can plot both the frequency-domain and time-domain synthetic seismograms into one plot, in order to check, that the results are the same.

Time-domain (in red) and frequency-domain (in black) synthetic seismograms:


First component

Second component

Third component
For detailed view download PostScript files 98-ssb1.ps, 98-ssb2.ps and 98-ssb3.ps with above figures.

If we wish to make also the later arrivals visible, we can use exponential scaling with time. We then have to normalize the maximum amplitude at each trace.

Frequency-domain seismograms, exponential scale with time:


First component

Second component

Third component
For detailed view download PostScript files 98-sse1.ps, 98-sse2.ps and 98-sse3.ps with above figures.

The seismogram computation and the visualization of the results is managed by history file 98-ss.h. Download all the necessary files and type "perl go.pl 98-ss.h" from the command line to run the computation.