Low-frequency, long-offset elastic waveform inversion in the context of velocity model building

Classic imaging approaches consist of splitting the earth into background and reflectivity models. When justified, this separation of scale is quite powerful, although this approach relies on some smoothness and weak contrast assumptions. This approach allows for the imaging methods to be based on acoustic wave propagation after having identified the compressional waves through picking or signal processing. Over the past years, wave-equation tomography and waveform inversion approaches have become routine, complementing the classic approaches to derive background models. They do not rely on high-frequency picks, unlike ray-based traveltime tomography, but on low-frequency cross-correlation to define time shifts and on waveform matching. In the presence of large earth parameter contrasts, time shifts and waveforms of compressional waves may depend on elastic parameters when interferences occur within the Fresnel zones. This challenges the recovery of the background model under an acoustic assumption with low-frequency data. Accounting for an elastic propagation in waveform inversion, even in the context of model building, could help to reduce the artifacts seen in acoustic results. A synthetic and a real data example are presented to illustrate the potential benefit of using an elastic waveform inversion approach when inverting long-offset, low-frequency seismic data.

Interpreter's Corner

The U.S. Gulf of Mexico (GoM) geology is well known for prolific structural hydrocarbon traps created by salt tectonics. In many areas, these structures lie below salt overhangs or thick canopies, requiring advanced seismic imaging to identify prospects and plan exploration wells. Ever-evolving geophysical technologies, such as 3D seismic, wide azimuth, multiwide azimuth, coil, and ocean-bottom node (OBN) acquisition designs, have unlocked the image for some of these structures over the past three decades. Recently, automatic velocity model building methods, particularly full-waveform inversion (FWI), introduced another step change in the subsalt image quality and refocused the acquisition methods on the need to acquire long-offset data. To make such a long-offset program affordable, a new survey geometry was set up with sparse OBN nodes and simultaneous shooting. The actual survey was acquired in 2019 and fully processed within 15 months from the end of the acquisition. Offsets up to 65 km were recorded, enabling FWI velocity updates down to 15 km depth. To provide the reader with a glimpse of the geologic insight that the new technology enabled, we report a few examples of deep geology revealed by this survey in a hydrocarbon- and seismic-data-rich area of the GoM — the Greater Mars-Ursa Basin.

Effects of electrical anisotropy on long-offset transient electromagnetic data

SUMMARY Electrical anisotropy of formations has been long recognized by field and laboratory evidence. However, most interpretations of long-offset transient electromagnetic (LOTEM) data are based on the assumption of an electrical isotropic earth. Neglecting electrical anisotropy of formations may cause severe misleading interpretations in regions with strong electrical anisotropy. During a large scale LOTEM survey in a former mining area in Eastern Germany, data was acquired over black shale formations. These black shales are expected to produce a pronounced bulk anisotropy. Here, we investigate the effects of electrical anisotropy on LOTEM responses through numerical simulation using a finite-volume time-domain (FVTD) algorithm. On the basis of isotropic models obtained from LOTEM field data, various anisotropic models are developed and analysed. Numerical results demonstrate that the presence of electrical anisotropy has a significant influence on LOTEM responses. Based on the numerical modelling results, an isolated deep conductive anomaly presented in the 2-D isotropic LOTEM electric field data inversion result is identified as a possible artifact introduced by using an isotropic inversion scheme. Trial-and-error forward modelling of the LOTEM electric field data using an anisotropic conductivity model can explain the data and results in a reasonable quantitative data fit. The derived anisotropic 2-D model is consistent with the prior geological information.

Definition of the Churchill River Delta and its petroleum potential, offshore Labrador, Canada

Starting in 2011, a multiyear modern 2D long offset broadband seismic survey was acquired offshore Labrador, Canada, by TGS and PGS in partnership with Nalcor Energy. This regional survey covers the slope and deepwater portions of the margin. Three Mesozoic and Cenozoic-aged basins were informally defined from these data, the Chidley, Henley, and Holton; also, the poorly constrained Hawke Basin was remapped. The 2D data set imaged for the first time a very large Cenozoic-aged delta adjacent to the mouth of Lake Melville. We have mapped this delta on a [Formula: see text] 2D seismic grid. The delta is 5–8 km thick, and its aerial extent is [Formula: see text]. The age of this delta has been interpreted to be Eocene to Miocene. Adjacent to this Cenozoic delta on the Labrador shelf, there is a working petroleum system within the proximal Hopedale and Saglek Basins where there are five gas discoveries and one oil discovery. The modern long-offset 2D data set appears to indicate a working petroleum system within the newly mapped Cenozoic delta, and two phases of hydrocarbons may be present.

Reflection intensity waveform inversion

Traditional iteration-based full-waveform inversion (FWI) methods encounter serious challenges if the initial velocity model is far from the true model or if the observed data are lacking low-frequency content. As such, the optimization algorithm may be trapped in local minima and fail to go to a global optimal model. In addition, the traditional FWI method requires long-offset data to update the deep structure of a velocity model with diving waves. To overcome the disadvantages of traditional FWI under these circumstances, we have developed a reflection intensity waveform inversion method. This method aims to minimize the seismic intensity differences between the modeled reflection data and field data. Our method is less dependent on the starting model, and long-offset data are no longer required. The wave intensity, proportional to the square of the original data amplitude, can have a low-frequency band and a higher frequency band, even for waveforms without initial low-frequency content. Our multiscale intensity inversion starts from the low-frequency information in the intensity data, and it can largely avoid the cycle-skipping problem. Synthetic and field data examples demonstrate that our method is able to overcome cycle skipping in handling data with no low-frequency information.