A brief guide to using reflectivity seismogram simulator adapted to
the needs of receiver function analysis.
General ideology is outlined in Levin & Park, GJI 131, pp 253-266 1997
The prototype code anirec was written by Jeffrey Park.
Miscellaneous I/O functions and spectra division/smoothing patch
were hacked in by Vadim Levin.
To understand what it does: read the paper, then read the code.
Else - use at your own risk.
rfgen-asc
takes a model and outputs receiver functions obtained by division of smoothed
spectra of respective components. Takes into account phase velocity and BAZ of
the incident phase. Output goes to files RRF.table and TRF.table (names hardwired),
in the form of ASCII tables with 3 columns (time, BAZ, RF_value).
velocity model used by all codes looks as follows:
AAK model <--- any string
5 <--- number of layers not counting last
0 0 <--- tilt from vertical and azimuth from north of the axis
700 3500 0.00 0.00 2000 0.00 2000 <--- bottom of the layer, Vp(m/s), B, C, Vs(m/s), E, rho (density, kg/m^3)
50 200
2000 4500 -0.10 0.00 2600 -0.10 2100
00 00
9000 5800 0.00 0.00 3350 0.00 2100
70 10
23000 6500 0.00 0.00 3400 0.00 2600
0 0
55000 6900 0.00 0.00 3900 0.00 2800
0 0
90000 8100 0.00 0.00 4500 0.00 3300 <---- LAST "LAYER" simulates the halfspace,
HAS TO BE ISOTROPIC!!!!
The code prompts the user for input.
Running it
interactively is good for learning
purposes. In a production mode using shellscripts like
the one below is recommended.
#take model name, baz and phase velocity as input
#
echo $2 > runfile
echo $1 >> runfile
echo $3 >> runfile
echo 1 >> runfile <--- this means 'smooth the spectra"
echo 450 >> runfile <--- Spectra is smoothed with an expression exp(-(i*i/(4*aa*aa)),
where i is frequency and aa is smoothing factor.
rfgen_asc < runfile > /dev/null
/bin/mv RRF.table $1.$2.$3.rrftab <--- doing this is essential since
/bin/mv TRF.table $1.$2.$3.trftab <--- files get written over