Chevron U.s.a. Inc.

Dennis George Bones

Kendall Bong Hu Louie

Shoujian Huang

E Fredrick Herkenhoff

Shoujian Jane Huang

A system and method for regularizing irregularly sampled 5D seismic data by assigning each trace to a common midpoint bin and mapping each trace to an offset vector tile (OVT) with a calculated center azimuth; assembling an azimuth sector with all offsets of interest and a narrow range of the center azimuths from the OVTs; rotating the azimuth sector to align with an inline direction and a crossline dimension; selecting a subset of traces with a single crossline value to create an irregular 3D volume; regularizing the irregular 3D volume; and repeating as necessary to generate a regularly sampled seismic dataset. The regularization may include interpolation by an algorithm such as an anti-leakage Fourier interpolator. The regularly sampled seismic dataset may be used to characterize the subsurface by further processing such as tomography.

System and method for regularizing seismic data

A method for enhancing signal-to-noise (S/N) ratio of seismic data is presented. An ensemble of input traces is decomposed into a plurality of frequency bands of traces. Trace and signal power and then a maximum allowable signal-to-noise ratio are estimated for each frequency band. Weights are calculated which are functions of the inverse noise power or rms. The amplitudes in each of the traces are reformed using a fitting function which utilizes the estimated weights to create true relative amplitude signal enhanced traces. The weighting function has data adaptive parameters which can be changed to accommodate noise characteristics such coherency, incoherency, Gaussian and non-Gaussian distributions, etc. The method may be applied across many different coordinate systems. The method may be applied iteratively to seismic data to shape the noise distribution of the seismic data.

Growing community of inventors

Dublin, CA, United States of America

Dennis George Bones