Part of a shot record before and after powerline noise removal, with the difference on the right.
Powerline noise, sometimes called highline, harmonic, sinusoidal, monofrequency, or monochromatic noise, appears as a constant-amplitude sine wave running through the entire trace. The usual causes are 50 or 60 Hz electrical currents such as powerlines near geophones or cables. This noise can create the following problems:
- Hard-to-pick first breaks.
- Amplitude and phase distortion in the seismic wavelet after minimum-phase deconvolution.
- Poor residual and trim statics.
- Noisy stack, especially after spectral whitening.
Concerning the second point, powerline noise can create phase distortion in the seismic wavelet that is both large and highly nonlinear, reducing the interpretability of the section. For land data in particular, a good powerline noise remover is essential for optimum resolution.
Phase distortion of the seismic wavelet due to powerline noise.
It was traditional to remove powerline noise with a notch filter, often applied in the field during acquisition. This may have been necessitated by the limited dynamic range of older recording equipment, but it caused serious problems in processing, as a notch in the frequency spectrum can create as much phase distortion as a spike.
Beginning in the 1980s, powerline noise suppressors based on modelling and adaptive subtraction were proposed (see references below). They have the potential to remove powerline noise while doing little or no damage to the underlying signal. Karsli et al. (2016) show the advantages of these methods over notch filters.
Juniper Bay’s powerline remover has the following features:
- It executes quickly.
- Parameters are few and easy to select. Parameter testing is rarely required.
- It searches for the optimal frequency over a given frequency band. The accuracy depends on the trace length, but is typically within .025 Hz. Such accuracy is essential, as even a small error in frequency can degrade the results.
- Little or no damage is done to the underlying signal at the powerline frequency. In particular, there is no spectral notching.
- If no significant powerline noise is found in a trace then it is left untouched. Typically only a fraction of the traces get altered. Some surveys might have no powerline noise removed whatsoever.
- Harmonics of the fundamental frequency are optionally removed if they have significant amplitude. As an example, if a fundamental frequency of 59.91 Hz is detected then harmonics at 119.82 and 179.73 Hz are also searched for.
- Multiple passes can be requested. In other words, if powerline noise is removed from a trace then the altered trace is searched again for a new fundamental frequency, and removed if detected. This is equivalent to running the tool two or more times in sequence, but at only slighter greater computational cost than running it once.
- You can optionally output the difference – that is, the removed noise — rather than the noise-filtered data.
- It writes a report at the end summarizing the results, including the number of traces altered and a histogram showing what powerline frequencies were found within the search band.
It has a limitation:
- It assumes that powerline noise is stationary over the course of the trace recording time. Significant changes in amplitude or frequency over the recording time can result in only partial noise removal.
Amplitude spectrum of the above shot record before and after powerline noise removal.
Ask for a free trial to enhance your ProMAX processing today!
References
Chiu, S., 2006, Interference cancellation by least-squares modeling and adaptive subtraction: SEG Technical Program Expanded Abstracts, 2752-2756.
Karsli, H., and D. Dondurur, 2015, Spectral Domain Local Cancellation Procedure of Harmonic Noise in Seismic Data: 8th Congress of the Balkan Geophysical Society.
Karsli, H., D. Dondurur, and R. Guney, 2016, A Comparison of Post-stack Results after Filtering of Harmonic Noise Using Two Filter Methods: Near Surface Geoscience Conference.
Linville, A. F., and R. A. Meek, 1992, Canceling stationary sinusoidal noise: Geophysics, 57, no. 11, 1493-1501.
Nyman, D. C., and J. E. Gaiser, 1983, Adaptive rejection of high-line contamination, SEG Technical Program Expanded Abstracts, 321-323.
Gao, S., G. Luo, and B. Zhao, 2009, Elimination of the monofrequency interference based on an adaptive subtraction, Beijing International Geophysical Conference and Exposition.
Saucier, A., Marchant, M., Chouteau, M, 2006, A fast and accurate frequency estimation method for canceling harmonic noise in geophysical records: Geophysics, 71, V7-V18.
Xia, J., and R. D. Miller, 2000, Design of a hum filter for suppressing power-line noise in seismic data, J. Environ. Eng. Geophys., No. 5, 31-38.