Presentation is loading. Please wait.

Presentation is loading. Please wait.

First Arrival Traveltime and Waveform Inversion of Refraction Data Jianming Sheng and Gerard T. Schuster University of Utah October, 2002.

Similar presentations


Presentation on theme: "First Arrival Traveltime and Waveform Inversion of Refraction Data Jianming Sheng and Gerard T. Schuster University of Utah October, 2002."— Presentation transcript:

1 First Arrival Traveltime and Waveform Inversion of Refraction Data Jianming Sheng and Gerard T. Schuster University of Utah October, 2002

2 Outline MotivationMotivation First arrival traveltime and waveform inversionFirst arrival traveltime and waveform inversion Numerical examplesNumerical examples SummarySummary

3 Motivation Traveltime and waveform of CDP refraction data Given: Goal: High resolution tomogram Problem: Can waveform tomography provide better resolution than provide better resolution than ray-based tomography? ray-based tomography?

4 Ray-based Tomography vs. Full Waveform Inversion Ray-basedtomography Efficient and robust Resolution limited by high-freq. assumption Fullwaveformtomography No high-freq. limitation Slow convergence and local minima problem

5 First-arrival Traveltime and Waveform Inversion Ray-basedtraveltimetomography Efficient and robust First-arrivalwaveforminversion No high-freq. limitation Better convergence and mild nonlinear Initial model

6 Outline MotivationMotivation First arrival traveltime and waveform inversionFirst arrival traveltime and waveform inversion Numerical examplesNumerical examples SummarySummary

7 First Arrival Traveltime and Waveform Inversion Step 1:Step 1: Preprocessing the raw data: band-pass, 3D to 2D transform, trace normalization Step 2:Step 2: Picking first-arrival traveltimes and muting out other waves except first arrivals

8 Step 3:Step 3: First arrival traveltime tomography Minimizes traveltime residual Initial model

9 Step 4: First arrivalStep 4: First arrival waveform inversion waveform inversion Observed Predicted Misfitfunction

10 Multigrid Tomography Traveltime tomography:Traveltime tomography: Dynamic smoothing scheme (to attack local minima problem) (Nemeth, T., Normark, E. and Qin, F., 1992)

11 Outline MotivationMotivation First arrival traveltime and waveform inversionFirst arrival traveltime and waveform inversion Numerical examplesNumerical examples SummarySummary

12 Numerical Examples Synthetic data I: Three-layerSynthetic data I: Three-layer Synthetic data II: WesternGeco (Blind test)Synthetic data II: WesternGeco (Blind test) Redmond mine survey dataRedmond mine survey data

13 Synthetic Model I 020 40 60 Depth (m) 0200100 Distance (m) 25001958 1416 873 331 (m/s) Source Freq. 60 Hz Avg. Velocity 2400 m/s Source Wavelength 40 m Suggested by Konstantin Osypov

14 020 40 60 Depth (m) 0200100 Distance (m) 25001958 1416 873 331 (m/s) 40 m Synthetic Model I

15 Synthetic Data I Synthetic data set was calculated Synthetic data set was calculated by 2-D FD acoustic wave equation by 2-D FD acoustic wave equation solver solver Twenty-one shots and 51 traces Twenty-one shots and 51 traces per shot were used. per shot were used. Computational grid dimension was Computational grid dimension was 401*121. 401*121.

16 Synthetic Shot Gather 0.0 0.1 Time (sec.) -80 -80120 Offset (m) Air Wave

17 Traveltime Tomogram 020 40 60 Depth (m) 0200100 Distance (m) 25001958 1416 873 331 (m/s)

18 Synthetic Model I 020 40 60 Depth (m) 0200100 Distance (m) 25001958 1416 873 331 (m/s)

19 Traveltime Residual 130 Iterations 2.0 1.0 0.0 Traveltime Residual (sec.)

20 Waveform Tomogram 020 40 60 Depth (m) 0200100 Distance (m) 25001958 1416 873 331 (m/s)

21 Synthetic Model I 020 40 60 Depth (m) 0200100 Distance (m) 25001958 1416 873 331 (m/s)

22 Waveform Residual 1 30 Iterations 12,000 0 8,000 4,000

23 Numerical Examples Synthetic data I: Three-layerSynthetic data I: Three-layer Synthetic data II: WesternGeco (Blind test)Synthetic data II: WesternGeco (Blind test) Redmond mine survey dataRedmond mine survey data

24 True Velocity Model 0.0 1.0 Depth (km) 0.0 26 Horizontal distance (km) 1000 m/s 2050~2500 m/s

25 True Density Model 0.0 26 Horizontal distance (km) 0.0 1.0 Depth (km)

26 Recorded CSG # 150 -30003000 Offset (m) 0.0 2.0 Time (sec.)

27 Guessed Density Model 1000 5000 Velocity (m/s) 3400 1400 Density (kg/m 3 )

28 Source Wavelet 400 Amplitude 0 -600 0.0 0.25 Time (sec.)

29 Waveform Matching Amplitude 0.00.2 Time (sec.) Offset(m) -50 25 50 0 -25

30 Traveltime Tomogram 2712 2284 1856 1428 1000 m/s 0.0 26 Horizontal distance (km) 0.0 1.0 Depth (km)

31 Traveltime Tomogram 2409 2057 1705 1352 1000 m/s 5.0 8.75 Horizontal distance (km) 0.0 0.4 Depth (km) 0.1 0.2 0.3

32 Waveform Tomogram 2700 2275 1850 1425 1000 m/s 5.0 8.75 Horizontal distance (km) 0.0 0.4 Depth (km) 0.1 0.2 0.3

33 Migration section 5.08.75 Horizontal distance (km) 0.0 0.4 Depth (km) 0.1 0.2 0.3

34 Predicted CSG #150 0.0 2.0 Time (sec.) -30003000 Offset (m)

35 Recorded CSG # 150 -30003000 Offset (m) 0.0 2.0 Time (sec.)

36 Numerical Examples Synthetic data I: Three-layerSynthetic data I: Three-layer Synthetic data II: WesternGeco (Blind test)Synthetic data II: WesternGeco (Blind test) Redmond mine survey dataRedmond mine survey data

37

38 Salt Diapir Data Thirty-one shots and 120 traces Thirty-one shots and 120 traces total 3188 traveltimes picked. total 3188 traveltimes picked. Shot interval: 20 m Shot interval: 20 m geophone interval 5 m geophone interval 5 m Source frequency 40 Hz. Source frequency 40 Hz. Record length 1 sec. Record length 1 sec. sample interval 0.5 millisecond. sample interval 0.5 millisecond.

39 CSG for Field Data After Preprocessing 1120 Geophone # 00.2 Time (sec.)

40 CSG for Field Data After Muting 1120 Geophone # 00.2 Time (sec.)

41 Wavelet Extracted 0 0.1 Time (sec.)

42 Traveltime Tomogram 0130 Depth (m) 0590 Distance (m) 55004500 3500 2500 1500 500 (m/s) Tunnel 20 m 55 m SALT

43 Traveltime Residual 1 30 Iterations 2.0 1.0 0.0 Traveltime Residual (sec.)

44 0130 Depth (m) Waveform Tomogram 0590 Distance (m) 55004500 3500 2500 1500 500 (m/s) Tunnel 55 m 20 m SALT

45 Traveltime Tomogram 0130 Depth (m) 0590 Distance (m) 55004500 3500 2500 1500 500 (m/s) Tunnel 20 m 55 m SALT

46 Waveform Residual 1 30 Iterations 6,000 0 4,000 2,000

47 Predicted CSG 1120 Geophone # 00.2 Time (sec.)

48 CSG for Salt Data After Muting 1120 Geophone # 00.2 Time (sec.)

49 2 Log10 Amplitude 0 -2 -4 0400 Offset (m) Logarithmic Amplitude Vs. Offset Synthetic Observed

50 Problems Seismic attenuation Surface wave noise Source wavelet inversion & objective function

51 Outline MotivationMotivation First arrival traveltime and waveform inversionFirst arrival traveltime and waveform inversion Numerical examplesNumerical examples SummarySummary

52 Summary Synthetic results show that the waveform tomogram is much more resolved;Synthetic results show that the waveform tomogram is much more resolved; The preliminary results for the field data are not as good as expected, and further work is needed.The preliminary results for the field data are not as good as expected, and further work is needed.

53 Acknowledgment I thank the sponsors of the 2002 University of Utah Tomography and Modeling /Migration (UTAM) Consortium for their financial support. I thank Konstantin Osypov for providing the data set.


Download ppt "First Arrival Traveltime and Waveform Inversion of Refraction Data Jianming Sheng and Gerard T. Schuster University of Utah October, 2002."

Similar presentations


Ads by Google