Enhanced Velocity Structure from Waveform Tomography of Seismic First-Arrival Data: Application to the Nechako Basin

Key Researcher(s):  R. Clowes

Project ID:  2007-024

Key Research Organization(s):  University of British Columbia

Project Location:  Nechako Basin

Focus Area:  Energy-Oil and Gas


The Nechako Basin in central British Columbia is considered an area prospective for oil and gas resources. The Skeena group, comprising several thousands of meters of Early to mid-Cretaceous non-marine sandstone and conglomerate, is the most prospective sedimentary unit. It is overlain by younger rocks, including recent volcanics and thick glacial deposits, which present problems for seismic exploration. In the late 1970s and early 1980s, Canadian Hunter carried out an exploration program that included a series of seismic reflection lines and a number of exploration wells. However, no follow-on work was carried out.

Project Area In 2008, Geoscience BC recorded a series of modern multichannel reflection profiles, using a vibroseis source, to better determine the subsurface structure and prospectivity of the region. The first-arrival data from such surveys can be used to determine variations in the near-surface velocity structure, which are necessary for timing corrections in the processing sequence and may be used to discriminate rock types (glacial till, sedimentary rocks, volcanics). Typically, such results are obtained from linear inversion or traveltime tomography of picked first-arrival travel times. Recently, a new approach that provides a much enhanced model of velocity structure has been developed. Known as waveform tomography, this technique uses the first-arrival waveform data rather than traveltime picks.

We plan to apply waveform tomography to the first-arrival waveform data from the 2008 seismic survey to generate well-resolved velocity structure for the upper crust. If the data are adequate, seismic attenuation structure also can be derived. However, the method has not yet been applied to data acquired with a vibroseis source. Thus, our first step will be to test the approach on existing vibroseis data. Preferably, we will use selections from the Canadian Hunter data, which with its 2500 m maximum offset will provide structural information to ~ 500 m depth. Otherwise, the method will be tested on existing LITHOPROBE data.

Our second step involves application of waveform tomography to the 2008 data set, which with maximum offsets to ~15 km will enable well-resolved velocity and attenuation structures to depths of 2000-3000 m. In collaboration with geologists knowledgeable of the Nechako Basin and seismologists involved in the main seismic survey, the velocity and attenuation models will be interpreted in terms of subsurface structure, distinguishing glacial deposits, sedimentary rocks and volcanics, and perhaps identifying shallow sub-basins within the Nechako Basin. The enhanced velocity structure also may be used in a second-stage of processing of the main reflection data to provide improved images where warranted.



  • Posters and Presentations

    2010: Improved Near-surface Velocity Models from Waveform Tomography Applied to Vibroseis MCS Reflection First-Arrival Data

    – Mineral Exploration Roundup Poster (pdf, 1.79MB)

  • Technical Articles

    2011: “Improved near-surface velocity models from the Nechako Basin seismic survey, south-central British Columbia, part 2: full-waveform inversion”

    – Summary of Activities 2010, Report 2011-1 p. 239-254 (pdf, 6.0 MB)


    2010: “Improved nearsurface velocity models from the Nechako Basin seismic survey, south-central British Columbia (parts of NTS 093B, C, F, G), part 1: traveltime inversions”

    – Summary of Activities 2009, Report 2010-1 p.227-234 (pdf, 3.8 MB)


    2009: “Enhanced velocity structure from waveform tomography of seismic first-arrival data: application to the Nechako Basin, south-central British Columbia”

    – Summary of Activities 2008, Report 2009-1 p.157-162 (pdf, 4.24MB)

  • Final Deliverables

    Ph.D. Thesis – Brendan Smithyman

    Developments in waveform tomography of land seismic data with applications in south-central British Columbia – University of British Columbia, 20013 – Available digitally through UBC (pdf, 27 MB)

    Abstract: Waveform tomography (WT) is an advanced class of seismic imaging methods that may be applied to generate detailed models of subsurface velocity and attenuation by analysis of recorded field seismograms. WT comprises two steps: 1) initial velocity model building by traveltime inversion (raytracing) and 2) full-waveform inversion to update the model of velocity based on analysis of refracted data waveforms. The nature of this method requires that the initial model and simulated survey parameters reproduce closely those used for the field data acquisition. The application of WT to on-land seismic data is challenged by the geometry irregularities inherent in crooked-line two-dimensional (2D) seismic data acquisition. This thesis develops two methodologies that permit the inversion of on-land crooked-line vibroseis multi-channel seismic (MCS) reflection data by traveltime tomography (TT) and subsequent viscoacoustic full-waveform inversion (FWI) to generate 2D cross sections of P-wave velocity. The first method uses 2.5D TT and applies a TT-derived static correction to enable the subsequent application of 2D FWI. The second method uses 2.5D or three-dimensional (3D) TT followed by 2.5D full-waveform inversion that models 3D survey geometry (a new development). These technical developments are motivated and tested by a multi-part case study, which analyses data from the 2008 Geoscience BC Nechako Basin seismic survey. Both the static-correction method and 2.5D FWI method are applied to a single acquisition line. The results are analysed and compared with the conclusion that 2.5D WT provides a superior result, but at additional computational cost. This cost is still significantly lower than that associated with 3D FWI. The 2.5D WT method is applied to three datasets from the 2008 survey, and TT alone is applied to a fourth. The resulting models constrain the positions and lithology of Eocene volcanic rock units, Cretaceous strata and Jurassic or earlier basement rocks in the Nechako-Chilcotin Plateau region of south-central British Columbia. These results are synthesized and interpreted in conjunction with colocated surficial-geology maps, exploration well logs and geophysical results by other workers. This interpretation enhances the specific knowledge of lithostratigraphy along these acquisition profiles and informs general geological and geophysical analyses of the region.