Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
James D. Mayhan* and Arthur B. Weglein
Multiple removaland prerequisite satisfaction:Current status and future plans
May 27‐30, 2014Austin, TX
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 1 / 55
The purpose of this talk1 Show results of Green’s-theorem preprocessing
(required by ISS algorithms)
2 Show results of satisfying the prerequisites for ISSfree-surface-multiple and internal-multiplealgorithms with synthetic data corresponding tooffshore plays— For ISS to reach its full potential and deliver itspromise, its prerequisites must be met
3 Motivate onshore methods for satisfying ISSprerequisites
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 2 / 55
The purpose of this talk1 Show results of Green’s-theorem preprocessing
(required by ISS algorithms)2 Show results of satisfying the prerequisites for ISS
free-surface-multiple and internal-multiplealgorithms with synthetic data corresponding tooffshore plays— For ISS to reach its full potential and deliver itspromise, its prerequisites must be met
3 Motivate onshore methods for satisfying ISSprerequisites
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 2 / 55
The purpose of this talk1 Show results of Green’s-theorem preprocessing
(required by ISS algorithms)2 Show results of satisfying the prerequisites for ISS
free-surface-multiple and internal-multiplealgorithms with synthetic data corresponding tooffshore plays— For ISS to reach its full potential and deliver itspromise, its prerequisites must be met
3 Motivate onshore methods for satisfying ISSprerequisites
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 2 / 55
Outline
Table of Contents I
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 3 / 55
Outline
Table of Contents II5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 4 / 55
Overview
Table of Contents
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 5 / 55
Overview
What’s the problem?
As exploration for hydrocarbons has moved intoareas with increasingly complex geology, there aremore instances in which multiples are proximal to oreven overlap the primaries.
Hence, demultiple algorithms are challenged toremove multiples without damaging proximalprimaries.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 6 / 55
Overview
What’s the problem?
As exploration for hydrocarbons has moved intoareas with increasingly complex geology, there aremore instances in which multiples are proximal to oreven overlap the primaries.
Hence, demultiple algorithms are challenged toremove multiples without damaging proximalprimaries.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 6 / 55
Overview
A solution to add to your seismic toolbox
The ISS can achieve all processing objectivesdirectly and without subsurface information.
In particular, the ISS free-surface-multiple algorithmcan accurately predict the phase (time) andamplitude of free-surface multiples, if itsprerequisites (source signature and deghosted data)are satisfied (Carvalho et al., 1992; Weglein et al.,1997, 2003).
This has been demonstrated on marine field data(Carvalho, 1992; Carvalho et al., 1992; Carvalho andWeglein, 1994; Weglein et al., 2003; Ferreira, 2011).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 7 / 55
Overview
A solution to add to your seismic toolbox
The ISS can achieve all processing objectivesdirectly and without subsurface information.
In particular, the ISS free-surface-multiple algorithmcan accurately predict the phase (time) andamplitude of free-surface multiples, if itsprerequisites (source signature and deghosted data)are satisfied (Carvalho et al., 1992; Weglein et al.,1997, 2003).
This has been demonstrated on marine field data(Carvalho, 1992; Carvalho et al., 1992; Carvalho andWeglein, 1994; Weglein et al., 2003; Ferreira, 2011).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 7 / 55
Overview
A solution to add to your seismic toolbox
The ISS can achieve all processing objectivesdirectly and without subsurface information.
In particular, the ISS free-surface-multiple algorithmcan accurately predict the phase (time) andamplitude of free-surface multiples, if itsprerequisites (source signature and deghosted data)are satisfied (Carvalho et al., 1992; Weglein et al.,1997, 2003).
This has been demonstrated on marine field data(Carvalho, 1992; Carvalho et al., 1992; Carvalho andWeglein, 1994; Weglein et al., 2003; Ferreira, 2011).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 7 / 55
Overview
A solution to add to your seismic toolbox
The current ISS internal-multiple algorithm canpredict the exact phase (time) and an approximateamplitude of all internal multiples, at once,automatically, and without subsurface information(Araujo et al., 1994; Weglein et al., 2003).
This has been demonstrated on marine field data(Matson et al., 1999; Terenghi et al., 2011; Ferreira,2011; Weglein, 2013; Goodway and Mackidd, 2013;Kelamis and Yi Luo, 2013; Ferreira et al., 2013;Dragoset, 2013; Brookes and Jenner, 2013; Griffithset al., 2013; Hegge et al., 2013).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 8 / 55
Overview
A solution to add to your seismic toolbox
The current ISS internal-multiple algorithm canpredict the exact phase (time) and an approximateamplitude of all internal multiples, at once,automatically, and without subsurface information(Araujo et al., 1994; Weglein et al., 2003).
This has been demonstrated on marine field data(Matson et al., 1999; Terenghi et al., 2011; Ferreira,2011; Weglein, 2013; Goodway and Mackidd, 2013;Kelamis and Yi Luo, 2013; Ferreira et al., 2013;Dragoset, 2013; Brookes and Jenner, 2013; Griffithset al., 2013; Hegge et al., 2013).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 8 / 55
Overview
A solution to add to your seismic toolbox
Those ISS properties are what all other currentdemultiple methods (e.g., Feedback-loop methods,modeling and subtraction of multiples, and filtermethods) do not possess and cannot deliver(Weglein, 1999; Weglein and Dragoset, 2005; QiangFu et al., 2010; Yi Luo et al., 2011; Weglein et al.,2011; Ferreira, 2011; Kelamis et al., 2013).
Details concerning the ISS free-surface-multiple andinternal-multiple algorithms were covered in Dr.Weglein’s tutorial.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 9 / 55
Overview
A solution to add to your seismic toolbox
Those ISS properties are what all other currentdemultiple methods (e.g., Feedback-loop methods,modeling and subtraction of multiples, and filtermethods) do not possess and cannot deliver(Weglein, 1999; Weglein and Dragoset, 2005; QiangFu et al., 2010; Yi Luo et al., 2011; Weglein et al.,2011; Ferreira, 2011; Kelamis et al., 2013).
Details concerning the ISS free-surface-multiple andinternal-multiple algorithms were covered in Dr.Weglein’s tutorial.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 9 / 55
Overview
But the ISS demultiple algorithms have prerequisites
The prerequisites for ISS demultiple algorithms canbe met by Green’s-theorem-based algorithms(Weglein and Secrest, 1990; Weglein et al., 2002;Jingfeng Zhang and Weglein, 2005, 2006; JingfengZhang, 2007).
Background on Green’s-theorem-based algorithmswas covered in Dr. Weglein’s tutorial.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 10 / 55
Overview
But the ISS demultiple algorithms have prerequisites
The prerequisites for ISS demultiple algorithms canbe met by Green’s-theorem-based algorithms(Weglein and Secrest, 1990; Weglein et al., 2002;Jingfeng Zhang and Weglein, 2005, 2006; JingfengZhang, 2007).
Background on Green’s-theorem-based algorithmswas covered in Dr. Weglein’s tutorial.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 10 / 55
Overview
But why use Green’s theorem for deghosting insteadof the ISS?
The ISS equation linear in data is D = G0VG0,where D is the measured data (after removing thereference wave), G0 is a known analytic Green’sfunction, and V is the difference between the actualmedium and the reference medium.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 11 / 55
Overview
But why use Green’s theorem for deghosting insteadof the ISS?
The ISS equation linear in data is D = G0VG0,where D is the measured data (after removing thereference wave), G0 is a known analytic Green’sfunction, and V is the difference between the actualmedium and the reference medium.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 11 / 55
Overview
We can deghost by replacing G0 with thedirect-wave Green’s function, G d
0 (Carvalho, 1992):
G d0 G−10 DG−10 G d
0 = G d0 G−10 G0︸ ︷︷ ︸
=I
V G0G−10︸ ︷︷ ︸=I
G d0
D ′ = G d0 VG d
0
However, G−10 can be unstable (near ghost notches)if we don’t know the exact depth of the towed cable.
No similar inverse function appears when usingGreen’s theorem, provides a more stable deghostingalgorithm.
This fact motivates us to find Green’s-theorem-typemethods for onshore.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 12 / 55
Overview
We can deghost by replacing G0 with thedirect-wave Green’s function, G d
0 (Carvalho, 1992):
G d0 G−10 DG−10 G d
0 = G d0 G−10 G0︸ ︷︷ ︸
=I
V G0G−10︸ ︷︷ ︸=I
G d0
D ′ = G d0 VG d
0
However, G−10 can be unstable (near ghost notches)if we don’t know the exact depth of the towed cable.
No similar inverse function appears when usingGreen’s theorem, provides a more stable deghostingalgorithm.
This fact motivates us to find Green’s-theorem-typemethods for onshore.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 12 / 55
Overview
We can deghost by replacing G0 with thedirect-wave Green’s function, G d
0 (Carvalho, 1992):
G d0 G−10 DG−10 G d
0 = G d0 G−10 G0︸ ︷︷ ︸
=I
V G0G−10︸ ︷︷ ︸=I
G d0
D ′ = G d0 VG d
0
However, G−10 can be unstable (near ghost notches)if we don’t know the exact depth of the towed cable.
No similar inverse function appears when usingGreen’s theorem, provides a more stable deghostingalgorithm.
This fact motivates us to find Green’s-theorem-typemethods for onshore.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 12 / 55
Overview
We can deghost by replacing G0 with thedirect-wave Green’s function, G d
0 (Carvalho, 1992):
G d0 G−10 DG−10 G d
0 = G d0 G−10 G0︸ ︷︷ ︸
=I
V G0G−10︸ ︷︷ ︸=I
G d0
D ′ = G d0 VG d
0
However, G−10 can be unstable (near ghost notches)if we don’t know the exact depth of the towed cable.
No similar inverse function appears when usingGreen’s theorem, provides a more stable deghostingalgorithm.
This fact motivates us to find Green’s-theorem-typemethods for onshore.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 12 / 55
Overview
Discussions of Green’s-theorem deghosting inWeglein et al. (2013a,b).
The ability of Green’s theorem to meet prerequisiteshas been tested on SEAM and field data (Mayhanand Weglein, 2013a; Mayhan, 2013).
We show examples in Figures 1 - 4 that will follow.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 13 / 55
Overview
Discussions of Green’s-theorem deghosting inWeglein et al. (2013a,b).
The ability of Green’s theorem to meet prerequisiteshas been tested on SEAM and field data (Mayhanand Weglein, 2013a; Mayhan, 2013).
We show examples in Figures 1 - 4 that will follow.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 13 / 55
Overview
Discussions of Green’s-theorem deghosting inWeglein et al. (2013a,b).
The ability of Green’s theorem to meet prerequisiteshas been tested on SEAM and field data (Mayhanand Weglein, 2013a; Mayhan, 2013).
We show examples in Figures 1 - 4 that will follow.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 13 / 55
Green’s-theorem deghosting
Table of Contents
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 14 / 55
Green’s-theorem deghosting
Green’s-theorem deghosting
Links to the current interest in broadband seismic
Used to satisfy ISS prerequisites
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 15 / 55
Green’s-theorem deghosting
2
10
15
179
13
12
11
14
Basement 1
Mother Salt 2
Cretaceous 3
Oligocene-Paleo 4
Lower Miocene 5
Middle Miocene 6
Upper Miocene 7
Pliocene 8
Pleistocene 9
Water 10
Inv. Lower Mio. 11
Inv. Olig-Paleog. 12
Inv. Cretaceous 13
Salt Suture 14
Salt 15
Hetero Salt 17
SEAM Phase I Model: Indicator Volume
Figure 1: SEAM deep-water Gulf of Mexico model: inline sectionfrom the middle of the model. Layers 3-8 lie sequentially betweenlayers 2 and 9. Layer 1 isn’t shown. Figure courtesy of SEAM.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 16 / 55
Green’s-theorem deghosting
Figure 2: SEAM data, example shot (131373), recorded data at17 m. Tested while an intern at PGS.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 17 / 55
Green’s-theorem deghosting
Figure 3: Green’s-theorem deghosting on SEAM data:blue=input, red=R deghosted, green=S&R deghosted
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 18 / 55
Green’s-theorem deghosting
Caption for Figure 3 on previous slide (from Mayhan andWeglein, 2013b)
SEAM data, shot 131373, frequency spectra:blue=P at 17 m, red=receiver deghosted at 10 musing Green’s theorem, green=source and receiverdeghosted at 10 m using Green’s theorem.
Vertical axis=amplitude (dB), horizontalaxis=frequency (Hz).
The first source notch is at 44 Hz, which lies abovethe source frequency range (1–30 Hz).
Note the shift of the spectrum towards lowerfrequencies (which may be of interest to FWI).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 19 / 55
Green’s-theorem deghosting
Figure 4: Green’s-theorem deghosting on field data:blue=input, red=R deghosted
20 30 40 50 60 70 80 90 100
Frequency (Hz)
-35
-30
-25
-20
-15
-10
-50
Am
plitu
de (
dB)
20 30 40 50 60 70 80 90 100
Frequency (Hz)
-35
-30
-25
-20
-15
-10
-50
Am
plitu
de (
dB)
Inp
ut
hyd
rop
ho
nes
(b
lue)
, rec
eive
r si
de
deg
ho
sted
(re
d)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 20 / 55
Green’s-theorem deghosting
Caption for Figure 4 on previous slide (from Mayhanet al., 2011)
2D field data, shot 841, frequency spectra: blue=Pat 25 m, red=receiver deghosted at the air-waterboundary using Green’s theorem.
Vertical axis=amplitude (dB), horizontalaxis=frequency (Hz).
The receiver notches around 30 Hz, 60 Hz, and90 Hz have been filled in.
Input data courtesy of PGS.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 21 / 55
Green’s-theorem deghosting
Zhiqiang (Justin) Wang (now at PGS) testedGreen’s-theorem deghosting while an intern at Total(shown in next slide)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 22 / 55
Green’s-theorem deghosting
Some of Zhiqiang Wang’s deghosting results
Green’s theorem deghosting
and
P(zg,zs,ω) = R
"eik(2zwb−zg−zs)− eik(2zwb−zg+zs)
2ik
+−eik(2zwb+zg−zs) + eik(2zwb+zg+zs)
2ik
#. (4)
represents the water-bottom reflected primary and its source,receiver, and source and receiver ghosts respectively. Herek = ω/c0 is the wave number, R is the water-bottom reflec-tion coefficient, zg is the receiver depth, zs is the source depth,zwb is the water-bottom depth, and we suppose the free surfaceis at depth 0 and the sources and receivers are placed betweenfree surface and water bottom (0 < zs < zg < zwb).
Substitute the Green’s function G0 and the wavefield P intoEquation 1 to perform receiver side deghosting,
Pdeghostedreceiver (z′g,zs,ω)
= [P(z,zs,ω)dG+
0 (z,z′g,ω)dz
−G+0 (z,z′g,ω)
dP(z,zs,ω)dz
]|z=zg
= R
"eik(2zwb−z′g−zs)− eik(2zwb−z′g+zs)
2ik
#. (5)
Here we assume z′g < zg, which means the predicted cable isshallower than the actual cable. The receiver side ghosts areremoved and only the primary and its source side ghost remainin Equation 5. Further, we feed the receiver side deghosteddata into Equation 2 for source side deghosting,
Pdeghosted(z′g,z′s,ω)
= [Pdeghostedreceiver (z′g,z,ω)
dG+0 (z,z′s,ω)
dz
−G+0 (z,z′s,ω)
dPdeghostedreceiver (z′g,z,ω)
dz]|z=zs
=Reik(2zw−z′g−z′s)
2ik. (6)
Again we assume z′s < zs (e.g., the predicted source is shal-lower than the actual source). The result of equation 6 is thewater bottom reflected primary, without source, receiver, orsource and receiver ghosts.
(a) Model one
0 2000 4000 6000 8000 10000
0
500
1000
1500
2000
2500
Depth
(m)
RC
(b) Model two
Figure 1: Models for testing.
0
1
2
3
Tim
e(s)
-2000 -1000 0 1000 2000x(m)
(a) Before deghosting
0
1
2
3
Tim
e(s)
-2000 -1000 0 1000 2000x(m)
(b) Receiver deghosted
0
1
2
3
Tim
e(s)
-2000 -1000 0 1000 2000x(m)
(c) Source & receiver deghosted
0
0.2
0.4
0.6
0.8
1.0
Tim
e(s
)
0x(m)
(d) Before deghosting
0
0.2
0.4
0.6
0.8
1.0
Tim
e(s
)
0x(m)
(e) Receiver deghosted
0
0.2
0.4
0.6
0.8
1.0
Tim
e(s
)
0x(m)
(f) Source & receiver deghosted
Figure 2: Deghosting for model one.
NUMERICAL TESTING
The code used to compute the results in this abstract is the re-ceiver side deghosting code written by J. Mayhan and releasedin 2011 to the M-OSRP consortium. Changes were made toaccommodate both receiver and source side deghosting. Asmentioned above, we will use data with over/under sourcesand over/under receiver cables.
The first tested case (Figure 1a) is a three layer model withsources at 30m and 32m and receivers at 140m and 142m, suchthat the ghosts are not overlapping either with the correspond-ing primaries or among themselves. Figure 2b is the result afterreceiver side deghosting and Figure 2c is source and receiverside deghosting. Figures 2d, 2e, and 2f are the wiggle plotsof the zero-offset traces. We can see the ghosts are mainlyremoved and the algorithm works with good accuracy.
The second tested case (Figure 1b) has 9 layers and is extractedfrom a velocity model provided by TOTAL. In this case, wechoose the sources at 5m and 7m and receivers at 10m and15m, so the events and their ghosts are overlapping. The data,its receiver side deghosted result, and both source and receiverside deghosted result are in Figures 3a, 3b, and 3c, respec-tively. Figures 3d, 3e, and 3f represent the wiggle plots ofthe zero-offset traces and Figures 3g, 3h and 3i are the spec-trum plots of their wavelets. The notch at c0/(2d) = 1500/(2∗12)hz = 62.5hz is removed after receiver side deghosting and
© 2012 SEG DOI http://dx.doi.org/10.1190/segam2012-1246.1SEG Las Vegas 2012 Annual Meeting Page 2
Dow
nloa
ded
03/2
2/13
to 1
29.7
.16.
11. R
edis
trib
utio
n su
bjec
t to
SEG
lice
nse
or c
opyr
ight
; see
Ter
ms
of U
se a
t http
://lib
rary
.seg
.org
/
(a) Velocity model(provided by Total)
Green’s theorem deghosting
0
1
2
3
-2000 0 2000 4000 6000
(a) Before deghosting
0
1
2
3
-4000 -2000 0 2000 4000
(b) Receiver deghosted
0
1
2
3
-4000 -2000 0 2000 4000
(c) Source & receiver deghosted
0
0.2
0.4
0.6
0.8
1.0
0
(d) Before deghosting
0
0.2
0.4
0.6
0.8
1.0
0
(e) Receiver deghosted
0
0.2
0.4
0.6
0.8
1.0
0
(f) Source & receiver deghosted
0
20
40
60
80
0
(g) Before deghosting
0
20
40
60
80
0
(h) Receiver deghosted
0
20
40
60
80
0
(i) Source & receiver deghosted
Figure 3: Deghosting for model two.
both receiver side deghosting and source side deghosting re-cover more low frequency information.
Results are positive and encouraging for both receiver sidedeghosting and source side deghosting when the data with dualsources and dual receiver cables are provided. Below we ex-amine the consequences of two issues associated with the inputdata:
1. One issue is that in common practice, the derivative ofthe field (either through direct measurement or through
0
1
2
3
-4000 -2000 0 2000 4000
(a) Shot plot
0
0.2
0.4
0.6
0.8
1.0
0
(b) Wiggle plot
0
20
40
60
80
0
(c) Spectrum plot
Figure 4: Source deghosting using D f s = 0 for model two.
dual cables) may not be available, especially on thesource side. The industry trend has data with over/understreamers available today, and sometimes over/undersources, as well. However, Zhang (2007) uses Green’stheorem to develop and exemplify a method that com-pletely removes receiver and source ghosts with over/understreamers/receivers or with a single streamer and a sourcesignature. The method does not require over/under sources.Another method that can be applied is based on the no-tion that the pressure field on the free surface D f s iszero. This information can be used as another cable.Figure 4 shows the result when this property is usedfor model two. Comparing with the result using dualsources (Figures 3), the source ghosts are satisfactarilyremoved, although some high frequency informationhas been damaged. For low-frequency, this could besatisfactory.
2. Another issue is the sensitivity to how accurate the depthof the cable is known due to a division if over/under ca-bles are used. The Green’s theorem method is robustto depth sensitivity (Zhang, 2007). Figure 5 comparesthe results when different depth intervals are used forreceiver deghosting. The ghosts are well removed andnot visible after deghosting when the cable intervalsare 2 meters, 5 meters, or 10 meters. When the intervalgets to 20 meters, the ghosts are still largely attenuated(comparing with Figures 2a and 2d).
IMPACT ON FREE-SURFACE MULTIPLE ELIMINATION
ISS free-surface multiple elimination method has the theoret-ical capability of predicting the exact phase and amplitude ofmultiples if its pre-requisites (source and receiver deghostingin particular) are satisfied. Figure 6 shows the result of apply-ing the deghosted data into ISS free-surface multiple elimina-tion algorithm. The right half is generated by directly subtract-ing the prediction from the input without any adaptive subtrac-tion tool. We can see that all free-surface multiples are wellattenuated and primaries are not touched.
© 2012 SEG DOI http://dx.doi.org/10.1190/segam2012-1246.1SEG Las Vegas 2012 Annual Meeting Page 3
Dow
nloa
ded
03/2
2/13
to 1
29.7
.16.
11. R
edis
trib
utio
n su
bjec
t to
SEG
lice
nse
or c
opyr
ight
; see
Ter
ms
of U
se a
t http
://lib
rary
.seg
.org
/
(b) Spectrum plots of the wavelets
Figure 5: In (b), the left panel is before deghosting, the middle panelis receiver deghosted, and the right panel is source and receiverdeghosted. Both receiver deghosting and source deghosting recovermore low-frequency information. (Zhiqiang Wang et al., 2012)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 23 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Table of Contents
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 24 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
When the prerequisites are satisfied, the predictionof multiples improves.
We show examples in Figures 6-13 that will follow.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 25 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
When the prerequisites are satisfied, the predictionof multiples improves.
We show examples in Figures 6-13 that will follow.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 25 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
The ISS free-surface-multiple algorithm has thefollowing form (in 2D) (Carvalho, 1992):
D ′(kg , ks , ω) =∞∑n=1
D ′n(kg , ks , ω), where (1)
D ′n(kg , ks , ω) =1
iπρ0A(ω)
∫ ∞
−∞dk q exp (iq(εg + εs))
× D ′1(kg , k , ω)D ′n−1(k , ks , ω). (2)
Summing terms D ′1 through D ′n removes free-surfacemultiples of orders 1 through n − 1— and alters higher order multiples.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 26 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
The ISS free-surface-multiple algorithm has thefollowing form (in 2D) (Carvalho, 1992):
D ′(kg , ks , ω) =∞∑n=1
D ′n(kg , ks , ω), where (1)
D ′n(kg , ks , ω) =1
iπρ0A(ω)
∫ ∞
−∞dk q exp (iq(εg + εs))
× D ′1(kg , k , ω)D ′n−1(k , ks , ω). (2)
Summing terms D ′1 through D ′n removes free-surfacemultiples of orders 1 through n − 1— and alters higher order multiples.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 26 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Define variables in equations 1 and 2 (previous slide)
D ′ indicates measured data after removing referencewave and ghosts
kg , ks , ω are Fourier conjugates of receiver xcoordinate, source x coordinate, time
ρ0 is mass density of reference medium (water)
A(ω) is frequency spectrum of point isotropic source
εg , εs are (fixed) receiver and source depths belowthe free surface
Obliquity factor, q, is given byq = sgn(ω)
√ω2/c2
0 − k2
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 27 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Define variables in equations 1 and 2 (previous slide)
D ′ indicates measured data after removing referencewave and ghosts
kg , ks , ω are Fourier conjugates of receiver xcoordinate, source x coordinate, time
ρ0 is mass density of reference medium (water)
A(ω) is frequency spectrum of point isotropic source
εg , εs are (fixed) receiver and source depths belowthe free surface
Obliquity factor, q, is given byq = sgn(ω)
√ω2/c2
0 − k2
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 27 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Define variables in equations 1 and 2 (previous slide)
D ′ indicates measured data after removing referencewave and ghosts
kg , ks , ω are Fourier conjugates of receiver xcoordinate, source x coordinate, time
ρ0 is mass density of reference medium (water)
A(ω) is frequency spectrum of point isotropic source
εg , εs are (fixed) receiver and source depths belowthe free surface
Obliquity factor, q, is given byq = sgn(ω)
√ω2/c2
0 − k2
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 27 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Define variables in equations 1 and 2 (previous slide)
D ′ indicates measured data after removing referencewave and ghosts
kg , ks , ω are Fourier conjugates of receiver xcoordinate, source x coordinate, time
ρ0 is mass density of reference medium (water)
A(ω) is frequency spectrum of point isotropic source
εg , εs are (fixed) receiver and source depths belowthe free surface
Obliquity factor, q, is given byq = sgn(ω)
√ω2/c2
0 − k2
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 27 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Define variables in equations 1 and 2 (previous slide)
D ′ indicates measured data after removing referencewave and ghosts
kg , ks , ω are Fourier conjugates of receiver xcoordinate, source x coordinate, time
ρ0 is mass density of reference medium (water)
A(ω) is frequency spectrum of point isotropic source
εg , εs are (fixed) receiver and source depths belowthe free surface
Obliquity factor, q, is given byq = sgn(ω)
√ω2/c2
0 − k2
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 27 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Define variables in equations 1 and 2 (previous slide)
D ′ indicates measured data after removing referencewave and ghosts
kg , ks , ω are Fourier conjugates of receiver xcoordinate, source x coordinate, time
ρ0 is mass density of reference medium (water)
A(ω) is frequency spectrum of point isotropic source
εg , εs are (fixed) receiver and source depths belowthe free surface
Obliquity factor, q, is given byq = sgn(ω)
√ω2/c2
0 − k2
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 27 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Equation 2 (with the integral) requires only thesource signature A(ω) and data D ′1(kg , k , ω)(deghosted and wavelet deconvolved) as its input.
The free-surface multiples are predictedorder-by-order and are summed (equation 1) to givethe free-surface-demultipled (and deghosted) data.
For a more detailed discussion, please seeSection 5.4 of Weglein et al. (2003).
Equation 2 can be modified to accommodate anextended source with a radiation pattern (JinlongYang et al., 2013).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 28 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Equation 2 (with the integral) requires only thesource signature A(ω) and data D ′1(kg , k , ω)(deghosted and wavelet deconvolved) as its input.
The free-surface multiples are predictedorder-by-order and are summed (equation 1) to givethe free-surface-demultipled (and deghosted) data.
For a more detailed discussion, please seeSection 5.4 of Weglein et al. (2003).
Equation 2 can be modified to accommodate anextended source with a radiation pattern (JinlongYang et al., 2013).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 28 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Equation 2 (with the integral) requires only thesource signature A(ω) and data D ′1(kg , k , ω)(deghosted and wavelet deconvolved) as its input.
The free-surface multiples are predictedorder-by-order and are summed (equation 1) to givethe free-surface-demultipled (and deghosted) data.
For a more detailed discussion, please seeSection 5.4 of Weglein et al. (2003).
Equation 2 can be modified to accommodate anextended source with a radiation pattern (JinlongYang et al., 2013).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 28 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Equation 2 (with the integral) requires only thesource signature A(ω) and data D ′1(kg , k , ω)(deghosted and wavelet deconvolved) as its input.
The free-surface multiples are predictedorder-by-order and are summed (equation 1) to givethe free-surface-demultipled (and deghosted) data.
For a more detailed discussion, please seeSection 5.4 of Weglein et al. (2003).
Equation 2 can be modified to accommodate anextended source with a radiation pattern (JinlongYang et al., 2013).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 28 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Model used for testing free-surface-multiple algorithm
Figure 6: Model used for Figures 7 - 10 that will follow (JinlongYang et al., 2012).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 29 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Figure 7: Free-surface-multiple elimination without prior removal ofghosts (Jinlong Yang reported in Lin Tang et al., 2013)
FSM1
FSM2
P1
(a) Input data
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(b) ISS FSMprediction
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(c) After a simple(non-adaptive)subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 30 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Figure 8: Free-surface-multiple elimination with prior removal ofghosts (Jinlong Yang reported in Lin Tang et al., 2013)
FSM1
FSM2
P1
(a) Input data
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(b) ISS FSMprediction
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(c) After a simple(non-adaptive)subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 31 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Figure 9: ISS free-surface-multiple elimination without sourcewavelet deconvolution (Jinlong Yang et al., 2012).
FSM1
FSM2
P1
(a) Input data
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(b) ISS FSMprediction
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(c) After a simple(non-adaptive)subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 32 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Figure 10: ISS free-surface-multiple elimination with source waveletdeconvolution (Jinlong Yang et al., 2012).
FSM1
FSM2
P1
(a) Input data
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(b) ISS FSMprediction
0.5
1.0
1.5
2.0
Tim
e(s)
500 1000 1500Trace Number
(c) After a simple(non-adaptive)subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 33 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
ISS free-surface-multiple algorithm assumes1 Carvalho’s obliquity factor, q, given (in 2D) by
q = sgn(ω)√ω2/c2
0 − k2
— Not available in methods from other consortia
2 Wavelet removed3 Receiver and source ghosts removed
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 34 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
ISS free-surface-multiple algorithm assumes1 Carvalho’s obliquity factor, q, given (in 2D) by
q = sgn(ω)√ω2/c2
0 − k2
— Not available in methods from other consortia2 Wavelet removed
3 Receiver and source ghosts removed
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 34 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
ISS free-surface-multiple algorithm assumes1 Carvalho’s obliquity factor, q, given (in 2D) by
q = sgn(ω)√ω2/c2
0 − k2
— Not available in methods from other consortia2 Wavelet removed3 Receiver and source ghosts removed
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 34 / 55
Effect of prerequisites on ISS free-surface-multiple algorithm
Taking a first order equation and expecting adaptivesubtraction to correct is overuse of adaptive subtraction— Make prediction as strong as possible
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 35 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Table of Contents
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 36 / 55
Effect of prerequisites on ISS internal-multiple algorithm
The ISS internal-multiple algorithm has the following form (in 2D)(Araujo et al., 1994; Weglein et al., 2003; Terenghi et al., 2011):
b3IM(kg , ks , ω)
=1
(2π)2
∫ ∞−∞
dk1 exp (iq1(zg − zs))
∫ ∞−∞
dk2 exp (iq2(zg − zs))
×∫ ∞−∞
dz1b1(kg , k1, z1) exp (i(qg + q1)z1)
×∫ z1−ε
−∞dz2b1(k1, k2, z2) exp (−i(q1 + q2)z2)
×∫ ∞z2+ε
dz3b1(k2, ks , z3) exp (i(q2 + qs)z3) (3)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 37 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Define variables in equation 3 (previous slide)
z1, z2, z3 are the pseudodepths of three genericpoints in the data
Range of integration chosen such that z1 > z2 andz2 < z3Vertical wave numbers are defined byqi = sgn(ω)
√ω2/c2
0 − k2i for i = s, g , 1, 2
b1(kg , ks , z) corresponds to effective incidentplane-wave data in the pseudodepth domain
For a more detailed discussion please see Section 6of Weglein et al. (2003).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 38 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Define variables in equation 3 (previous slide)
z1, z2, z3 are the pseudodepths of three genericpoints in the data
Range of integration chosen such that z1 > z2 andz2 < z3
Vertical wave numbers are defined byqi = sgn(ω)
√ω2/c2
0 − k2i for i = s, g , 1, 2
b1(kg , ks , z) corresponds to effective incidentplane-wave data in the pseudodepth domain
For a more detailed discussion please see Section 6of Weglein et al. (2003).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 38 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Define variables in equation 3 (previous slide)
z1, z2, z3 are the pseudodepths of three genericpoints in the data
Range of integration chosen such that z1 > z2 andz2 < z3Vertical wave numbers are defined byqi = sgn(ω)
√ω2/c2
0 − k2i for i = s, g , 1, 2
b1(kg , ks , z) corresponds to effective incidentplane-wave data in the pseudodepth domain
For a more detailed discussion please see Section 6of Weglein et al. (2003).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 38 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Define variables in equation 3 (previous slide)
z1, z2, z3 are the pseudodepths of three genericpoints in the data
Range of integration chosen such that z1 > z2 andz2 < z3Vertical wave numbers are defined byqi = sgn(ω)
√ω2/c2
0 − k2i for i = s, g , 1, 2
b1(kg , ks , z) corresponds to effective incidentplane-wave data in the pseudodepth domain
For a more detailed discussion please see Section 6of Weglein et al. (2003).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 38 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Define variables in equation 3 (previous slide)
z1, z2, z3 are the pseudodepths of three genericpoints in the data
Range of integration chosen such that z1 > z2 andz2 < z3Vertical wave numbers are defined byqi = sgn(ω)
√ω2/c2
0 − k2i for i = s, g , 1, 2
b1(kg , ks , z) corresponds to effective incidentplane-wave data in the pseudodepth domain
For a more detailed discussion please see Section 6of Weglein et al. (2003).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 38 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Equation 3 can be modified to suppress predictionsthat do not correspond to internal multiples in theseismic data (Chao Ma and Weglein, 2014b,c).
A new ISS algorithm is being developed to eliminate(rather than attenuate) internal multiples (YangleiZou and Weglein, 2014a,b).
As in the free-surface-multiple case, equation 3 canbe modified to accommodate an extended sourcewith a radiation pattern (Jinlong Yang et al., 2014).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 39 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Equation 3 can be modified to suppress predictionsthat do not correspond to internal multiples in theseismic data (Chao Ma and Weglein, 2014b,c).
A new ISS algorithm is being developed to eliminate(rather than attenuate) internal multiples (YangleiZou and Weglein, 2014a,b).
As in the free-surface-multiple case, equation 3 canbe modified to accommodate an extended sourcewith a radiation pattern (Jinlong Yang et al., 2014).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 39 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Equation 3 can be modified to suppress predictionsthat do not correspond to internal multiples in theseismic data (Chao Ma and Weglein, 2014b,c).
A new ISS algorithm is being developed to eliminate(rather than attenuate) internal multiples (YangleiZou and Weglein, 2014a,b).
As in the free-surface-multiple case, equation 3 canbe modified to accommodate an extended sourcewith a radiation pattern (Jinlong Yang et al., 2014).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 39 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Model used for testing internal-multiple algorithm
Figure 11: Model used for Figures 12 and 13 that will follow.(Jinlong Yang et al., 2013, 2014b).
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 40 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Figure 12: ISS internal-multiple attenuation without source waveletdeconvolution (Jinlong Yang et al., 2014a,b).
IM1
IM2
P1
P2
(a) Input data
0.4
0.6
0.8
1.0
1.2
1.4
Tim
e(s)
500 1000 1500 2000Trace Number
(b) ISS IMprediction
0.4
0.6
0.8
1.0
1.2
1.4
Tim
e(s)
500 1000 1500 2000Trace Number
(c) After a simple(non-adaptive)subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 41 / 55
Effect of prerequisites on ISS internal-multiple algorithm
Figure 13: ISS internal-multiple attenuation with source waveletdeconvolution (Jinlong Yang et al., 2014a).
IM1
IM2
P1
P2
(a) Input data
0.4
0.6
0.8
1.0
1.2
1.4
Tim
e(s)
500 1000 1500 2000Trace Number
(b) ISS IMprediction
0.4
0.6
0.8
1.0
1.2
1.4
Tim
e(s)
500 1000 1500 2000Trace Number
(c) After a simple(non-adaptive)subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 42 / 55
Effect of prerequisites on ISS internal-multiple algorithm
ISS internal-multiple algorithm assumes1 Same as free-surface-multiple algorithm plus
free-surface multiples removed
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 43 / 55
Onshore
Table of Contents
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 44 / 55
Onshore
Use of Green’s theorem to satisfy ISS prerequisites,as is currently performed for marine seismic data, isdiscussed in Jing Wu et al. (2014b), and a methodfor finding the reference velocity in the near surfaceis discussed in Lin Tang’s presentation.
Use of Green’s theorem to remove ground roll1 isdiscussed in Jing Wu’s first presentation.
1Rayleigh waves, i.e., longitudinal and transverse waves that travelnear the surface
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 45 / 55
Onshore
Use of Green’s theorem to satisfy ISS prerequisites,as is currently performed for marine seismic data, isdiscussed in Jing Wu et al. (2014b), and a methodfor finding the reference velocity in the near surfaceis discussed in Lin Tang’s presentation.
Use of Green’s theorem to remove ground roll1 isdiscussed in Jing Wu’s first presentation.
1Rayleigh waves, i.e., longitudinal and transverse waves that travelnear the surface
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 45 / 55
Summary and Acknowledgements
Table of Contents
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 46 / 55
Summary and Acknowledgements
Revisiting the purposes of this talk:1 We show ISS prerequisites are achievable on
synthetic and field data
2 We show the impact of preprocessing on ISSmultiple removal
3 We want the same for onshore
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 47 / 55
Summary and Acknowledgements
Revisiting the purposes of this talk:1 We show ISS prerequisites are achievable on
synthetic and field data2 We show the impact of preprocessing on ISS
multiple removal
3 We want the same for onshore
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 47 / 55
Summary and Acknowledgements
Revisiting the purposes of this talk:1 We show ISS prerequisites are achievable on
synthetic and field data2 We show the impact of preprocessing on ISS
multiple removal3 We want the same for onshore
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 47 / 55
Summary and Acknowledgements
Revisiting the purposes of this talk:
In complex geology, adaptive subtraction can’t fixeverything we leave out
Prerequisites matter more because we’re aftermultiple elimination
Successes in marine are motivating us to do thesame on land
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 48 / 55
Summary and Acknowledgements
Revisiting the purposes of this talk:
In complex geology, adaptive subtraction can’t fixeverything we leave out
Prerequisites matter more because we’re aftermultiple elimination
Successes in marine are motivating us to do thesame on land
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 48 / 55
Summary and Acknowledgements
Revisiting the purposes of this talk:
In complex geology, adaptive subtraction can’t fixeverything we leave out
Prerequisites matter more because we’re aftermultiple elimination
Successes in marine are motivating us to do thesame on land
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 48 / 55
Summary and Acknowledgements
Code currently available on sponsors-only portion ofmosrp.uh.edu:
Modeling— Cagniard-de Hoop forward modeling (1Dacoustic earth, line source)— Finite-difference forward modeling (2D acousticearth, line source)
Preprocessing— Green’s-theorem marine-data preprocessing (2Dor 3D data sets)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 49 / 55
Summary and Acknowledgements
Code currently available on sponsors-only portion ofmosrp.uh.edu:
Modeling— Cagniard-de Hoop forward modeling (1Dacoustic earth, line source)— Finite-difference forward modeling (2D acousticearth, line source)
Preprocessing— Green’s-theorem marine-data preprocessing (2Dor 3D data sets)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 49 / 55
Summary and Acknowledgements
Code currently available on sponsors-only portion ofmosrp.uh.edu:
Multiple removal— ISS free-surface-multiple elimination (3D datasets)— ISS internal-multiple attenuation (1.5D, 2D, or3D data sets, including limited aperture version)
Adaptive subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 50 / 55
Summary and Acknowledgements
Code currently available on sponsors-only portion ofmosrp.uh.edu:
Multiple removal— ISS free-surface-multiple elimination (3D datasets)— ISS internal-multiple attenuation (1.5D, 2D, or3D data sets, including limited aperture version)
Adaptive subtraction
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 50 / 55
Summary and Acknowledgements
To access the code
Go to mosrp.uh.edu and login as a sponsor
Research > Projects > Coding Projects
Code listed in reverse chronological order (mostrecent code first)
Code has documentation, assumes workingknowledge of Linux, Seismic Unix, MPI, C, Fortran
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 51 / 55
Summary and Acknowledgements
To access the code
Go to mosrp.uh.edu and login as a sponsor
Research > Projects > Coding Projects
Code listed in reverse chronological order (mostrecent code first)
Code has documentation, assumes workingknowledge of Linux, Seismic Unix, MPI, C, Fortran
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 51 / 55
Summary and Acknowledgements
To access the code
Go to mosrp.uh.edu and login as a sponsor
Research > Projects > Coding Projects
Code listed in reverse chronological order (mostrecent code first)
Code has documentation, assumes workingknowledge of Linux, Seismic Unix, MPI, C, Fortran
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 51 / 55
Summary and Acknowledgements
To access the code
Go to mosrp.uh.edu and login as a sponsor
Research > Projects > Coding Projects
Code listed in reverse chronological order (mostrecent code first)
Code has documentation, assumes workingknowledge of Linux, Seismic Unix, MPI, C, Fortran
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 51 / 55
Summary and Acknowledgements
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 52 / 55
Summary and Acknowledgements
Following talks will cover:
How back out onshore reference properties (LinTang)
How predict onshore reference wave (Jing Wu)
The importance of collecting the right data onshore(Qiang Fu)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 53 / 55
Summary and Acknowledgements
Following talks will cover:
How back out onshore reference properties (LinTang)
How predict onshore reference wave (Jing Wu)
The importance of collecting the right data onshore(Qiang Fu)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 53 / 55
Summary and Acknowledgements
Following talks will cover:
How back out onshore reference properties (LinTang)
How predict onshore reference wave (Jing Wu)
The importance of collecting the right data onshore(Qiang Fu)
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 53 / 55
Summary and Acknowledgements
We thank the M-OSRP sponsors for theirencouragement and support.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 54 / 55
Summary and Acknowledgements
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 55 / 55
References
Table of Contents
1 Overview
2 Green’s-theorem deghosting
3 Effect of prerequisites on ISS free-surface-multiplealgorithm
4 Effect of prerequisites on ISS internal-multiplealgorithm
5 Onshore
6 Summary and Acknowledgements
7 References
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 56 / 55
References
References I
Araujo, F. V., A. B. Weglein, P. M. Carvalho, and R. H. Stolt.“Inverse scattering series for multiple attenuation: An examplewith surface and internal multiples.” 64th Annual InternationalMeeting, SEG, Expanded Abstracts. 1994, 1039–1041.
Brookes, D. and Ed Jenner. An Onshore example of 3D InterbedMultiple Attenuation. Presentation given at the Post SEGConvention Workshop on Internal Multiples, Houston, Texas,September 2013.
Carvalho, P. M. and A. B. Weglein. “Wavelet estimation for surfacemultiple attenuation using a simulated annealing algorithm.”64th Annual International Meeting, SEG, Expanded Abstracts.1994, 1481–1484.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 57 / 55
References
References II
Carvalho, P. M., A. B. Weglein, and R. H. Stolt. “Nonlinear inversescattering for multiple suppression: Application to real data. PartI.” 62nd Annual International Meeting, SEG, ExpandedAbstracts. 1992, 1093–1095.
Carvalho, Paulo Marcos. Free-surface multiple reflection eliminationmethod based on nonlinear inversion of seismic data. PhD thesis,Universidade Federal da Bahia, 1992. In Portuguese.
Dragoset, W. Internal Multiple Attenuation for Land Data:Requirements for Success. Presentation given at the Post SEGConvention Workshop on Internal Multiples, Houston, Texas,September 2013.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 58 / 55
References
References III
Ferreira, A., P. Terenghi, A. Weglein, and A. Oliveira. InternalMultiple Removal in Offshore Brazil Seismic Data Using theInverse Scattering Series. Presentation given at the Post SEGConvention Workshop on Internal Multiples, Houston, Texas,September 2013.
Ferreira, Andre. Internal multiple removal in offshore Brazil seismicdata using the inverse scattering series. Master’s thesis,University of Houston, 2011.
Fu, Qiang, Yi Luo, Panos G. Kelamis, ShouDong Huo, Ghada Sindi,Shih-Ying Hsu, and Arthur B. Weglein. “The inverse scatteringseries approach towards the elimination of land internalmultiples.” 80th Annual International Meeting, SEG, ExpandedAbstracts. 2010, 3456–3461.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 59 / 55
References
References IV
Goodway, W. and D. Mackidd. Multiples. . . The Elephant in theRoom. Presentation given at the Post SEG Convention Workshopon Internal Multiples, Houston, Texas, September 2013.
Griffiths, M., A. Pica, and B. Hung. Internal Multiple removalsolutions for multiple environments. Presentation given at thePost SEG Convention Workshop on Internal Multiples, Houston,Texas, September 2013.
Hegge, R., A. Klaedtke, R. van Borselen, and E. Otnes. Aspects ofstandard practice and possible enhancements to 3D IME.Presentation given at the Post SEG Convention Workshop onInternal Multiples, Houston, Texas, September 2013.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 60 / 55
References
References V
Kelamis, P. and Yi Luo. Myths & Realities of Land Internal MultipleElimination Technologies. Presentation given at the Post SEGConvention Workshop on Internal Multiples, Houston, Texas,September 2013.
Kelamis, Panos G., Yi Luo, and Arthur B. Weglein. “Strategies ofLand Internal Multiple Elimination based on Inverse ScatteringSeries.” Technical Session 21: Recent Development in SeismicImaging/Processing, International Petroleum TechnologyConference, Beijing, China, 26-28 March 2013. 2013, 1–4.
Luo, Yi, Panos G. Kelamis, Qiang Fu, Shoudong Huo, Ghada Sindi,Shih-Ying Hsu, and Arthur B. Weglein. “Elimination of landinternal multiples based on the inverse scattering series.” TheLeading Edge 30 (August 2011): 884–889.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 61 / 55
References
References VI
Ma, Chao and Arthur B. Weglein. “Inverse Scattering Series (ISS)leading-order internal-multiple-attenuation algorithm andhigher-order modification to accommodate primaries and internalmultiples as input: 1-D normal incident test on interfering events,and extension to multi-D.” M-OSRP 2013-2014 Annual Report.2014.
Ma, Chao and Arthur B. Weglein. “Short note: Lower-order internalmultiples act as subevents in the ISSinternal-multiple-attenuation algorithm in a three-reflectorexample.” M-OSRP 2013-2014 Annual Report. 2014.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 62 / 55
References
References VII
Matson, Ken, Dennis Corrigan, Arthur Weglein, Chi-Yuh Young, andPaulo Carvalho. “Inverse scattering internal multiple attenuation:Results from complex synthetic and field data examples.” 69thAnnual International Meeting, SEG, Expanded Abstracts. 1999,1060–1063.
Mayhan, James D. Wave theoretic preprocessing to allow the InverseScattering Series methods for multiple removal and depth imagingto realize their potential and impact: Methods, examples, andadded value. PhD thesis, University of Houston, December 2013.
Mayhan, James D., Paolo Terenghi, Arthur B. Weglein, and NizarChemingui. “Green’s theorem derived methods for preprocessingseismic data when the pressure P and its normal derivative aremeasured.” 81st Annual International Meeting, SEG, ExpandedAbstracts. 2011, 2722–2726.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 63 / 55
References
References VIII
Mayhan, James D. and Arthur B. Weglein. “First application ofGreen’s theorem-derived source and receiver deghosting ondeep-water Gulf of Mexico synthetic (SEAM) and field data.”Geophysics 78 (March 2013): WA77–WA89.
Mayhan, James D. and Arthur B. Weglein. Source and ReceiverDe-Ghosting of SEAM Dual Sensor Data. Invited presentationgiven at the SEG SEAM Workshop, San Antonio, Texas, May2013.
Tang, Lin, James D. Mayhan, Jinlong Yang, and Arthur B. Weglein.“Using Green’s theorem to satisfy data requirements of multipleremoval methods: The impact of acquisition design.” 83rdAnnual International Meeting, SEG, Expanded Abstracts. 2013,4392–4396.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 64 / 55
References
References IX
Terenghi, P., X. Li, Shih-Ying Hsu, and Arthur B. Weglein. “1Dpreprocessing of Kristin data.” M-OSRP 2010-2011 AnnualReport. 2011, 35–49.
Terenghi, Paolo, Shih-Ying Hsu, Arthur B. Weglein, and Xu Li.“Exemplifying the specific properties of the inverse scatteringseries internal-multiple attenuation method that reside behind itscapability for complex onshore and marine multiples.” TheLeading Edge 30 (August 2011): 876–882.
Wang, Zhiqiang, Arthur B. Weglein, James D. Mayhan, PaoloTerenghi, and Christian Rivera. “Green’s theorem deriveddeghosting: fundamental analysis, numerical test results, andimpact on ISS free-surface multiple elimination.” 82nd AnnualInternational Meeting, SEG, Expanded Abstracts. 2012, 1–5.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 65 / 55
References
References X
Weglein, A. B. and W. Dragoset, editors. Multiple Attenuation.SEG Geophysics Reprint Series. Society of ExplorationGeophysicists, 2005.
Weglein, A. B., S. A. Shaw, K. H. Matson, J. L. Sheiman, R. H.Stolt, T. H. Tan, A. Osen, G. P. Correa, K. A. Innanen, Z. Guo,and J. Zhang. “New approaches to deghosting towed-streamerand ocean-bottom pressure measurements.” 72nd AnnualInternational Meeting, SEG, Expanded Abstracts. 2002,2114–2117.
Weglein, Arthur B. “Multiple attenuation: an overview of recentadvances and the road ahead (1999).” The Leading Edge 18(January 1999): 40–44.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 66 / 55
References
References XI
Weglein, Arthur B. Multiple Attenuation: Recent Advances and theRoad Ahead (2013). Keynote presentation given at the Post SEGConvention Workshop on Internal Multiples, Houston, Texas,September 2013.
Weglein, Arthur B., Fernanda V. Araujo, Paulo M. Carvalho,Robert H. Stolt, Kenneth H. Matson, Richard T. Coates, DennisCorrigan, Douglas J. Foster, Simon A. Shaw, and Haiyan Zhang.“Inverse Scattering Series and Seismic Exploration.” InverseProblems 19 (October 2003): R27–R83.
Weglein, Arthur B., Fernanda Araujo Gasparotto, Paulo M.Carvalho, and Robert H. Stolt. “An Inverse-Scattering SeriesMethod for Attenuating Multiples in Seismic Reflection Data.”Geophysics 62 (November-December 1997): 1975–1989.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 67 / 55
References
References XII
Weglein, Arthur B., Shih-Ying Hsu, Paolo Terenghi, Xu Li, andRobert H. Stolt. “Multiple attenuation: Recent advances and theroad ahead (2011).” The Leading Edge 30 (August 2011):864–875.
Weglein, Arthur B., Hong Liang, Jing Wu, James D. Mayhan, LinTang, and Lasse Amundsen. “A new Green’s theorem de-ghostingmethod that simultaneously: (1) avoids a finite differenceapproximation for the normal derivative of the pressure and, (2)avoids the need for replacing the normal derivative of pressurewith the vertical component of particle velocity, thereby avoidingissues that can arise within each of those twoassumptions/approaches: Theory and analytic and numericexamples.” Journal of Seismic Exploration 22 (November 2013):413–426.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 68 / 55
References
References XIII
Weglein, Arthur B., James D. Mayhan, Lasse Amundsen, HongLiang, Jing Wu, Lin Tang, Yi Luo, and Qiang Fu. “Green’stheorem de-ghosting algorithms in k , ω (e.g., P − Vz
de-ghosting) as a special case of x , ω algorithms (based onGreen’s theorem) with: (1) significant practical advantages anddisadvantages of algorithms in each domain, and (2) a newmessage, implication and opportunity for marine towed streamer,ocean bottom and on-shore acquisition and applications.”Journal of Seismic Exploration 22 (September 2013): 389–412.
Weglein, Arthur B. and Bruce G. Secrest. “Wavelet estimation for amultidimensional acoustic or elastic earth.” Geophysics 55 (July1990): 902–913.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 69 / 55
References
References XIV
Wu, Jing and Arthur B. Weglein. “Green’s Theorem BasedWavefield Separation Application on Elastic/Land.” 84th AnnualInternational Meeting, SEG, Expanded Abstracts. Submitted,2014.
Yang, J. and A. B. Weglein. “Incorporating source and receiverarrays in the Inverse Scattering Series free-surface multipleelimination algorithm: Theory and examples that demonstrateimpact.” M-OSRP 2011-2012 Annual Report. 2012, 114–132.
Yang, J. and A. B. Weglein. “(1) Incorporating the source array inthe ISS free surface multiple elimination algorithm: the impact onremoving a multiple that interferes with a primary; and (2) thefirst test of wavelet deconvolution on the internal multiplealgorithm.” M-OSRP 2012-2013 Annual Report. 2013, 105–119.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 70 / 55
References
References XV
Yang, Jinlong, James D. Mayhan, Lin Tang, and Arthur B. Weglein.“Accommodating the source (and receiver) array in free-surfacemultiple elimination algorithm: Impact on interfering or proximalprimaries and multiples.” 83rd Annual International Meeting,SEG, Expanded Abstracts. 2013, 4184–4189.
Yang, Jinlong and Arthur B. Weglein. “Incorporating source waveletand radiation pattern into the ISS internal multiple attenuationalgorithm.” 84th Annual International Meeting, SEG, ExpandedAbstracts. Submitted, 2014.
Yang, Jinlong and Arthur B. Weglein. “Incorporating the sourcewavelet and radiation pattern into the ISS internal multipleattenuation algorithm: theory and examples.” M-OSRP2013-2014 Annual Report. 2014, 63–78.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 71 / 55
References
References XVI
Zhang, Jingfeng. Wave theory based data preparation for inversescattering multiple removal, depth imaging and parameterestimation: analysis and numerical tests of Green’s theoremdeghosting theory. PhD thesis, University of Houston, 2007.
Zhang, Jingfeng and Arthur B. Weglein. “Extinction theoremdeghosting method using towed streamer pressure data: analysisof the receiver array effect on deghosting and subsequent freesurface multiple removal.” 75th Annual International Meeting,SEG, Expanded Abstracts. 2005, 2095–2098.
Zhang, Jingfeng and Arthur B. Weglein. “Application of extinctiontheorem deghosting method on ocean bottom data.” 76thAnnual International Meeting, SEG, Expanded Abstracts. 2006,2674–2678.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 72 / 55
References
References XVII
Zou, Yanglei and Arthur B. Weglein. “The pre-stack 1D ISS internalmultiple elimination algorithm for all reflectors Part I: strengthsand limitations.” 84th Annual International Meeting, SEG,Expanded Abstracts. Submitted, 2014.
Zou, Yanglei and Arthur B. Weglein. “The pre-stack 1D ISS internalmultiple elimination algorithm for all reflectors Part II: addressingthe limitations.” 84th Annual International Meeting, SEG,Expanded Abstracts. Submitted, 2014.
Mayhan & Weglein () Multiples and prerequisites May 27-30, 2014 73 / 55