next up previous print clean
Next: Conclusions Up: Brown: LSJIMP field data Previous: Extension to non-seabed peglegs

Results

In 1997, WesternGeco distributed a 2-D test dataset, acquired in the Mississippi Canyon region of the Gulf of Mexico, for the testing of multiple suppression algorithms. The data contain a variety of strong surface-related multiples, and enough geologic complexity to render one-dimensional methods ineffectual. Figure [*] shows the stack of the raw data. The geologic setting of the area is a fairly well-behaved deepwater sedimentary basin, interrupted between 4000 and 16000m by a shallow tabular salt body.

In addition to the seabed peglegs, peglegs from the top of salt and strong reflector R1 are included in the inversion. Rather than including each of these surfaces separately, the R1 and top salt events are assumed to arise from a single reflector; from midpoint 0 m to roughly 6000 m, it is R1, while from 6000 m to 20000 m it is the top of salt. Only first order multiples are included in the inversion. Thus in equation (9), nsurf=2 and p=1.

I tested LSJIMP on 750 CMP gathers of the Mississippi Canyon dataset. Figure [*] illustrates the stack of the LSJIMP result. From the difference panel, note that important peglegs (TSPL and WBM) are almost entirely removed. Primaries (like PR) are not damaged. R1PM, and to a lesser extent TSPM, are removed effectively. Salt rugosity contributes negatively to the separation, by forming diffractions which are not modeled by HEMNO, and by violating HEMNO's small reflector dip assumption. Much deep multiple energy remains. Some of it is likely internal salt multiples. Additionally, time imaging operators (like HEMNO) are notoriously poor at imaging subsalt events.

Figures [*] and [*] show the LSJIMP results at two midpoint locations (0 m and 14400 m, respectively). From the top rows of these figures, notice that most important peglegs are separated from $\bold m_0$, while the primaries are preserved. The estimated seabed and R1/top of salt peglegs seem to match the events in the data, both in term of kinematics and amplitudes.

Figures [*] and [*] illustrate the superior performance of the HEMNO operator versus 1-D NMO when used in LSJIMP. The strong top of salt seabed pegleg in Figure [*], dipping over this midpoint range, is clearly better removed when HEMNO is used. Figure [*] illustrates a more insidious problem: because the 1-D NMO operator did not focus a pegleg at the correct time, it caused a spurious event to be manufactured at a slightly smaller time.

 
gulf.stackraw
gulf.stackraw
Figure 3
Stacked Mississippi Canyon 2-D dataset (750 midpoints), annotated with important horizons and multiples. Labeled events: R1 - strong reflection; TSR - top of salt; BSR - bottom of salt; WBM - first seabed multiple; R1PL - seabed pegleg of R1; R1PM - R1 pure surface multiple; TSPL - seabed pegleg of TSR; BSPL - seabed pegleg of BSR; TSPM - TSR pure surface multiple.


view burn build edit restore

 
stackcomp.gulf
stackcomp.gulf
Figure 4
Stack comparison of Mississippi Canyon data before and after LSJIMP. All panels windowed in time from 3.5 to 5.5 seconds. Top: Raw data stack. Center: Stack of estimated primary image, $\bold m_0$. Bottom: Stack of the subtracted multiples. Labeled events: PR - underlying primary; WBM - first seabed multiple; R1PL - seabed pegleg of reflector R1; R1PM - R1 pure surface multiple; TSPL - seabed pegleg of TSR (top of salt); BSPL - seabed pegleg of BSR (bottom of salt); TSPM - TSR pure surface multiple. BSTSPL - TSR pegleg of BSR.


[*] view burn build edit restore

 
comp1.lsrow.gulf.0
comp1.lsrow.gulf.0
Figure 5
Mississippi Canyon CMP 1 (y=0 m). All panels NMO'ed with stacking velocity and windowed in time from 3.5 to 5.5 seconds. Top row (L to R): Raw data; Estimated primaries ($\bold m_0$); Estimated non-primaries (difference). Center row (L to R): Raw data; Estimated total first order seabed multiple ($\sum_{k=0}^1 \bold R_{1,1} \bold N_{1,k,1} \bold S_{1,1} 
 \bold G_{1,1} \bold m_{1,k,1}$); difference. Bottom row (L to R): Raw data; Estimated total first order ``salt'' (R1 at this location) multiple ($\sum_{k=0}^1 \bold R_{1,2} \bold N_{1,k,2} \bold S_{1,2} \bold G_{1,2} 
 \bold m_{1,k,2}$); difference.


[*] view burn build edit restore

 
comp1.lsrow.gulf.540
comp1.lsrow.gulf.540
Figure 6
Mississippi Canyon CMP 540 (y=14400m). All panels NMO'ed with stacking velocity and windowed in time from 3.5 to 5.5 seconds. Top row (L to R): Raw data; Estimated primaries ($\bold m_0$); Estimated non-primaries (difference). Center row (L to R): Raw data; Estimated total first order seabed multiple ($\sum_{k=0}^1 \bold R_{1,1} \bold N_{1,k,1} \bold S_{1,1} 
 \bold G_{1,1} \bold m_{1,k,1}$); difference. Bottom row (L to R): Raw data; Estimated total first order ``salt'' multiple ($\sum_{k=0}^1 \bold R_{1,2} \bold N_{1,k,2} \bold S_{1,2} \bold G_{1,2} 
 \bold m_{1,k,2}$); difference.


view burn build edit restore

 
stackcomp-dipcomp.1.gulf
stackcomp-dipcomp.1.gulf
Figure 7
LSJIMP stack comparison of HEMNO versus 1-D NMO operator. All panels windowed from 4.4 to 4.8 seconds in time; 14000 to 15000 meters in midpoint. From L to R: Raw data stack; Stack of estimated $\bold m_0$ using HEMNO; Stack of estimated $\bold m_0$ using 1-D NMO operator; HEMNO difference; 1-D NMO difference. Seabed pegleg from top of salt reflection is outlined in all panels.


view burn build edit restore

 
stackcomp-dipcomp.2.gulf
stackcomp-dipcomp.2.gulf
Figure 8
LSJIMP stack comparison of HEMNO versus 1-D NMO operator. All panels windowed from 4.8 to 5.2 seconds in time; 9200 to 10200 meters in midpoint. From L to R: Raw data stack; Stack of estimated $\bold m_0$ using HEMNO; Stack of estimated $\bold m_0$ using 1-D NMO operator; HEMNO difference; 1-D NMO difference. The ovals highlight a nonexistent event ``removed'' from data due to 1-D NMO's inferior performance over nonflat structure.


view burn build edit restore


next up previous print clean
Next: Conclusions Up: Brown: LSJIMP field data Previous: Extension to non-seabed peglegs
Stanford Exploration Project
7/8/2003