The scripts in this directory reproduce the Figures shown in the paper "Finite-difference modeling experiments for seismic interferometry". Some of those scripts take a very long time to run. In this file some guidelines and runtime indications are given of the scripts to reproduce the Figures. The runtimes are given for a 2.5 Ghz Intel Core 2 Duo. OpenMP parallelization is not used. All scripts use some commands from Seismic Unix (SU) to plot results or post-process the data. ==> Please make sure that SU is compiled without XDR (In $CWPROOT/Makefile.config make sure that XDRFLAG is NOT set). The SU output files of fdelmodc are base on local IEEE data. Figure2.scr ------------- starts fdelmodc only to compute the source positions, 1 s. runtime: 5 seconds Figure3.scr ------------- First the shots along the box-shaped contour are modeled for each shot separately. This takes a very long time. To model these shots two other script are called: Simple_model_base.scr, to model 900 shots at level z=3600 Simple_model_sides.scr, to model 2x360 shots at the sides of the model at x=1000 and x=9000 runtime one shot: 270 seconds. total runtime modeling: 122 hours. Figure3_ref.scr ------------- direct modeled reference result Figure 3d runtime: 500 seconds. Figure4.scr ------------- 5 different source signature lengths Runtime: 3.5 hours per shot of 120 s. modeling Figure5.scr ------------- simulates 8000 short (2.5 s) sources Runtime: 3.5 hours Figure6.scr ------------- 5 different number of random sources Runtime: 3.5 hours per shot of 120 s. modeling Figure6f.scr ------------- make postscript file of middle trace after Figure6.scr, 1 s. Runtime: a few seconds to produce the figure. The script Figure6.scr must be run first Figure6length.scr ------------- NOT used in paper: The same as script Figure6.scr, but a random source signature of a fixed length. Figure7.scr ------------- as Figure 6, but with 1000 deep sources Runtime: 3.5 hours Figure7length.scr ------------- NOT used in paper: The same as script Figure7.scr, but a random source signature of a fixed length. Figure7fmax.scr ------------- NOT used in paper: The same as script Figure7.scr, but with varying maximum frequency in source signature. Figure8-9.scr ------------- for random and ricker wavelet deep, volume and plane sources Runtime: 6 x 3.5 hours Figure8-9Hom.scr ------------- for reviewer, same as Fig. 8-9 in homogeneous medium NOT used in paper: to answers reviewer, same as fig 8-9 in homogeneous medium, Runtime: 6 x 3.5 hours Figure10.scr ------------- reference and 2 SI results for visco-acoustic media Runtime reference: 2 x 3 minutes Runtime deep and volume: 2 x 1.5 hours Figure11.scr ----------- Uses the result of fdelmodc_long.scr and produces the postscript figures. The program to compute the correlation in a continuous way (corrsmp) is missing and the correlation result can not be reproduced. The program corrsmp is still in a development state and not yet suited for release. Figure12.scr ------------- The program to compute the correlation in a continuous way (corrsmp) is missing and the correlation result can not be reproduced. The program corrsmp is still in a development state and not yet suited for release. Figure13.scr ------------- amplitude variations on source strength Runtime: 3 x 1500 s. Figure14-15.scr ------------- The reference shot can be reproduced, together with the long 3600 s modeling. The SI retrieved results can not be reproduced. runtime: 161 hours for the 3600 seconds passive measurement 61 output files will be created each containing 65535 time samples. The program to compute the correlation in a continuous way (corrsmp) is missing and the correlation result can not be reproduced. The program corrsmp is still in a development state and not yet suited for release. FigureSourcesAppendixA.scr ------------- source construction shown in Figure A2-A3-A4 runtime: 150 seconds FigureGreenAppendixA.scr ------------- compares FD result with analytical result runtime 80 seconds. SIrand.scr ---------- Script used to generate the retrieved reflection from passive measurements. fdelmodc_long.scr ------------ This script is used to model the passive measurements and takes ~100 hours. The program to compute the correlation in a continuous way (corrsmp) is missing and the correlation result can not be reproduced. The program corrsmp is still in a development state and not yet suited for release. cross.scr and FigureCCsources.scr ---------- used to answer a question of reviewer 3 FigurePres.scr and MakeGifMovie.scr ------------- Generates snapshots which can be used in a presentation to explain SI