3D Tomography using Efficient Wavefront Picking of Traveltimes
Abdullah AlTheyab and G. T. Schuster King Abdullah University of
Science and Technology (KAUST) 1
Slide 2
Outline Introduction Areal Picking 3D Tomography using Areal
Picks Conclusion 2
Slide 3
Introduction For conventional acquisition geometry, receiver
lines are sparse. Picking is done on time- offset sections.
first-arrivals x y t 3
Slide 4
Field Data Example 4 3D OBS data parameters: 234 OBS stations
129 source-lines 50m inline spacing 400m OBS spacing 40-50m water
depth Source boat sail lines Receiver stations
Slide 5
Human picking time 30,186 sections to pick, each with 360
receivers. Estimated picking time: 2 section/minute 251hrs 8
hr/day: 31 days 5 12 km 0 3 Time [sec] CRG Shingling Low SNR
Slide 6
Quality Control and Cycle-skipping 12 km 0 3 Time [sec]
Shingling Traveltime [sec] Traveltime Map 14 2 Distance [km] 1 3
Shot 73Shot 74Shot 75 6
Slide 7
Memory Footprint The size of the data is 80 GB (at 4ms
sampling, after windowing). Interactive picking software require:
Large memory, Swapping to hard drives. Memory access pattern for QC
is complex. 7
Slide 8
Conventional Picking Approach Disadvantages: 1.Large human
piking time (31 days) 2.Laborious to QC and correct picks 3.Large
memory footprint (80 GB) 8
Slide 9
Outline Introduction Areal Picking 3D Tomography using Areal
Picks Conclusion 9
Slide 10
Areal picking For conventional acquisition geometry, receiver
lines are sparse. Picking is done on time- offset sections.
first-arrivals x y t 10
Slide 11
x y t Areal picking For dense-receiver acquisition geometry We
propose picking on time- slices (Areal Picking). 11
Slide 12
Areal picking For dense-receiver acquisition geometry We
propose picking on time- slices (Areal Picking). y t x 12
Slide 13
Areal picking For dense-receiver acquisition geometry We
propose picking on time- slices (Areal Picking). y t x 13
Slide 14
Areal picking For dense-receiver acquisition geometry We
propose picking on time- slices (Areal Picking). y t x 14
Slide 15
Areal picking: Interpolation We implemented a program that does
real-time interpolation. 15 Cartesian picks Polar interpolation
Continuous Polygon Picks are interpolated in
polar-coordinates.
Slide 16
Field Data Example 2 4 y[km] 14 x [km] 19 4 Time slice @ 0.8
sec 16
Slide 17
Field Data Example 2 4 y[km] 14 x [km] 19 4 Time slice @ 0.8
sec 17
Slide 18
Field Data Example y[km] 14 x [km] 19 4 Time slice @ 2.4 sec
18
Slide 19
Field Data Example y[km] 14 x [km] 19 4 Time slice @ 2.4 sec
19
Slide 20
Field Data Example: Human picking time 20 200 ms time-slice
spacing for 5 Hz FWI. 234 shots x 15 slices/shot= 3,510 slices (vs.
30,186 sections) to pick. Estimated picking-time: @2 slices/minute:
30 hrs @8 hr/day: 4 days (vs. 31 days)
Slide 21
Field Data Example: Quality Control Polygon must not cross.
y[km] 14 x [km] 19 4 Time slice @ 2.4 sec 21
Slide 22
Field Data Example: Quality Control Min Apparent velocity Max
22 Detect mispicks. Apparent Velocity Map Explore regional
trend
Slide 23
Field Data Example: Memory footprint 80 GB Slicing for 5Hz FWI
2 GB Slices are spaced at of the shortest period. 23
Slide 24
Outline Introduction Areal Picking 3D Tomography using Areal
Picks Conclusion 24
Final Traveltime Tomogram 28 0 3.5 10 depth slice x [km] y [km]
10 inline xline 0 z [km] y [km] 0 15004500 Velocity [m/s] 018
Structural cross-section
Slide 29
Field Data Example: Waveform Comparison 29 12 km 0 3 Time [sec]
Observed
Slide 30
Field Data Example: Waveform Comparison 30 12 km 0 3 Time [sec]
Calculated
Slide 31
Field Data Example: Waveform Comparison 31 12 km 0 3 Time [sec]
Observed
Slide 32
Field Data Example: Waveform Comparison 32 12 km 0 3 Time [sec]
Calculated
Slide 33
Field Data Example: Waveform Comparison 33 12 km 0 3 Time [sec]
Observed
Slide 34
Field Data Example: Waveform comparison 34 12 km 0 3 Time [sec]
Calculated
Slide 35
Field Data Example: Waveform Comparison 35 12 km 0 3 Time [sec]
Observed
Slide 36
Field Data Example: Waveform Comparison 36 12 km 0 3 Time [sec]
Calculated
Slide 37
Outline Introduction Areal Picking 3D Tomography using Areal
Picks Conclusion 37
Slide 38
Conclusions Areal picking allows for building 3D tomograms in
reasonable time. Advantages of areal picking: About 70-90%
reduction in human picking time (31 vs. 4 days) Easier QC and
correct mispicks Much lower memory footprint (80 GB vs. 2 GB)
38
Slide 39
Thank you Acknowledgments: Pemex for providing the data.
Sponsors of CSIM Saudi Aramco for supporting the FWI project.
Research Computing at KAUST. 39
