2D ocean-bottom marine acquisition via jittered sampling: examples and results

Author: Haneet Wason (hwason@eos.ubc.ca)

Date: April, 2013

Contents

Time-jittered acquisition with 1 boat

See the scripts under examples/TimeJitAcq_1boat/.

Load the parameters file

load([resultsdir '/TimeJitAcq_1boat/TimeJitAcq_1boat_params.mat'])

Conventional vs. time-jittered acquisition scenario

flipflop = 'no';
fig = 'yes';
convacq_2arrays(flipflop, tfireint_min, ns, dt, boatspeed, fig);
jitter_airgunarrays(ns, ds, dt, p, nboats, randseed, boatspeed, tfireint_min, tdelay, delayboat, fig);

Boat 1 - airgun array 1
Minimum interval between jittered firing times: 10.228 s
Maximum interval between jittered firing times: 29.7719 s
First jittered shot number: 1
Last jittered shot number: 126

Boat 1 - airgun array 2
Minimum interval between jittered firing times: 10.4796 s
Maximum interval between jittered firing times: 29.5199 s
First jittered shot number: 2
Last jittered shot number: 124

Original data

frame.t = 300;
frame.r = 60;
frame.s = 92;
caxmin = 100;
caxmax = 100;
cmap   = 'gray';
image2DTRSframe([resultsdir '/TimeJitAcq_1boat/Suez_ds25m.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

Time-jittered (or blended) data volume: view 30 seconds of the jittered data volume

t1 = 130.0;
t2 = 160.0;
imageJitteredData([resultsdir '/TimeJitAcq_1boat/TimeJitAcq_1boat_jitdata.rsf'], t1, t2, dt, nr, dr, caxmin, caxmax, cmap);

Recovery by conventional processing: apply the adjoint of the sampling operator

image2DTRSframe([resultsdir '/TimeJitAcq_1boat/TimeJitAcq_1boat_adjrecov.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

NOTE: an empty shot gather image implies that none of the airguns fired at that location.

This step is usually followed by some kind of median filtering on common receiver or CMP gathers. We address the challenge of deblending by a (curvelet-based) sparsity-promoting recovery technique.

Recovery by sparse inversion (via L1 minimization)

image2DTRSframe([resultsdir '/TimeJitAcq_1boat/TimeJitAcq_1boat_L1recov.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

Difference

image2DTRSframe([resultsdir '/TimeJitAcq_1boat/TimeJitAcq_1boat_L1diff.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

Time-jittered acquisition with 2 boats

See the scripts under examples/TimeJitAcq_2boats/.

Load the parameters file

load([resultsdir '/TimeJitAcq_2boats/TimeJitAcq_2boats_params.mat'])

Time-jittered acquisition scenario

fig = 'yes';
jitter_airgunarrays(ns, ds, dt, p, nboats, randseed, boatspeed, tfireint_min, tdelay, delayboat, fig);

Boat 1 - airgun array 1
Minimum interval between jittered firing times: 10.228 s
Maximum interval between jittered firing times: 29.695 s
First jittered shot number: 3
Last jittered shot number: 124

Boat 1 - airgun array 2
Minimum interval between jittered firing times: 10.4861 s
Maximum interval between jittered firing times: 29.2786 s
First jittered shot number: 2
Last jittered shot number: 119

Boat 2 - airgun array 1
Minimum interval between jittered firing times: 10.4796 s
Maximum interval between jittered firing times: 29.5199 s
First jittered shot number: 2
Last jittered shot number: 122

Boat 2 - airgun array 2
Minimum interval between jittered firing times: 10.498 s
Maximum interval between jittered firing times: 29.5598 s
First jittered shot number: 2
Last jittered shot number: 128

Original data

frame.t = 300;
frame.r = 70;
frame.s = 30;
image2DTRSframe([resultsdir '/TimeJitAcq_2boats/Suez_ds12pt5m.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

Time-jittered (or blended) data volume: view 30 seconds of the jittered data volume

t1 = 330.0;
t2 = 360.0;
imageJitteredData([resultsdir '/TimeJitAcq_2boats/TimeJitAcq_2boats_jitdata.rsf'], t1, t2, dt, nr, dr, caxmin, caxmax, cmap);

Recovery by conventional processing: apply the adjoint of the sampling operator

image2DTRSframe([resultsdir '/TimeJitAcq_2boats/TimeJitAcq_2boats_adjrecov.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

This step is usually followed by some kind of median filtering on common receiver or CMP gathers. We address the challenge of deblending by a (curvelet-based) sparsity-promoting recovery technique.

Recovery by sparse inversion (via L1 minimization)

image2DTRSframe([resultsdir '/TimeJitAcq_2boats/TimeJitAcq_2boats_L1recov.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

Difference

image2DTRSframe([resultsdir '/TimeJitAcq_2boats/TimeJitAcq_2boats_L1diff.rsf'], frame, nt, dt, caxmin, caxmax, cmap);

Running the code on your own data

Template scripts are included that can be adapted to run the code on your own data. See the scripts directory.