Overcoming Technical Challenges in Shallow-Water 3D SRME
A wide-tow streamer spread introduces a shallow crossline acquisition footprint into 3D seismic volumes at the boundary between adjacent sail lines. Although very shallow seismic events are generally not important from an exploration interpretation perspective, the footprint amplitude striping may be problematic for 3D SRME. The predicted multiple model may be inaccurate in the footprint zones, thereby affecting the success of adaptive subtraction of the multiple model from the input data.
Historical processing efforts have sought to mitigate the problem by building a synthetic representation of the seafloor event, but this can introduce small timing errors into the predicted multiple wavefield, the reflection coefficients may be grossly misrepresented, and resolution is compromised due to incomplete multiple subtraction. The expensive remedy is the use of a narrower streamer spread to mitigate the factors that contribute to the severity of the footprint.
PGS has developed a two-stage workflow that first builds a continuous 3D migrated image of the near-surface, including the seafloor event, relatively unaffected by the aforementioned footprint, and secondly uses 3D Wavefield Extrapolation SRME to predict and then remove the surface multiples. This workflow thereby enables very wide-tow streamer spreads to be used in the most challenging setting to conventional 3D SRME- shallow water.
The caveat is that the acquisition uses dual-sensor streamers (GeoStreamer®), enabling wavefield separation and the subsequent use of Separated Wavefield Imaging (SWIM®) to build the spatially continuous near-surface migrated volume.
Constructing the Continuous Near-Surface Migrated Volume
SWIM enables each receiver location to act as a virtual source in a modified implementation of shot domain one-way wavefield extrapolation migration (WEM). Since the streamer spread provides continuous crossline coverage between adjacent sail-lines, SWIM imaging produces a complete 3D sub-surface image all the way up to and including the seafloor event, even for very shallow water environments.
When several orders of surface related multiples contribute to the final SWIM image, the angular illumination is significantly improved compared to conventional imaging using primary reflections only. This helps to mitigate the historical challenge of using 3D SRME for the outermost streamers that traditionally lacked the required near offset/near angle illumination.
Shallow time slices using conventional migration (left) and SWIM (center) illustrate how SWIM mitigates the crossline acquisition footprint, for shallow water environments. The comparative crossline sections on the right highlight the reflectivity data that can be used in an enhanced wave equation SRME implementation using the output from SWIM.
Predicting a Multiple Model
The shot gathers are injected into the SWIM reflectivity volume, yielding the predicted multiple model using a modified form of 3D Wavefield Extrapolation SRME. Multiple attenuation then proceeds as usual using adaptive subtraction, resulting in less residual surface multiple energy (see figure below) and without the complications associated with the shallow-water acquisition footprint. The output data is then used in a conventional processing and imaging workflow.
Predicted multiple wavefields illustrate how the enhanced SWIM-based implementation of SRME predicts multiple events more successfully in shallow-water.
The result after the subtraction of the multiples contains less residual multiples and has higher resolution.