Removing Harmonic Noise
from Geophysical Surface
Nuclear Magnetic
Resonance Measurements
Annette Hein
Mentor: Dr. Andrew Parsekian
Geology and Geophysics
What is Nuclear Magnetic
Resonance?
• Geophysical method
adapted from common
chemistry technique.
• Nothing to do with
nuclear fuels or fission.
• Detects groundwater.
Introduction Methods Results Conclusions
Why do we care?
• NMR produces point soundings (water content
as a function of depth) from a surface
measurement. [5]
• Useful for hydrologic studies, and can replace
expensive wells.
• Not widely used because electromagnetic noise
(generators, powerlines, etc) hinders the
measurement. [1,7]
Introduction Methods Results Conclusions
Signal processing basics
• Signal: what we’re trying to measure; the
groundwater response
• Noise: everything else we measured; lightning,
electric fences, powerlines, ….
• Processing problem:
Signal
+ Signal
+
NOISE Math! NOISE
Introduction Methods Results Conclusions
Removing powerline noise
• Most published approaches use knowledge
about the noise to subtract it. [2,3,4]
• My approach uses knowledge about the signal
to separate the noise.
• Focuses on powerline harmonics, one of the
most common noise sources.
Introduction Methods Results Conclusions
Powerline harmonics
Time domain Frequency domain
Introduction Methods Results Conclusions
Conceptual NMR water signal
Time domain [5] Frequency domain
Introduction Methods Results Conclusions
Removing powerline harmonics
Time domain Frequency domain
Introduction Methods Results Conclusions
Symmetry based noise removal
• Transform NMR data into the frequency
domain.
• Identify and remove corrupted portions of the
spectrum.
• Reconstruct the corrupted portions with
“mirror image” frequencies from the other side
of the spectrum.
• In the absence of other noise, this procedure
recovers an uncorrupted signal.
Introduction Methods Results Conclusions
Symmetry based noise removal
Advantages:
• Simpler than most published methods for
removing powerline harmonics from NMR.
• Does not disrupt the dataset
• Requires no prior knowledge about the
harmonics.
Disadvantages:
• Will not work in certain special cases.
• Will not work on extremely noisy data.
Introduction Methods Results Conclusions
Modeling results
Introduction Methods Results Conclusions
Modeling results
Introduction Methods Results Conclusions
Conclusions
• Symmetry based processing removes the powerline
harmonics without disrupting the signal.
• It is based on the mathematical theory describing
NMR signals and exploits knowledge of the signal.
• When powerline harmonics are removed, the water
content and T2* inversions are more accurate.
Introduction Methods Results Conclusions
Acknowledgments
• Thanks to WyCEHG and Dr Andrew Parsekian.
• Thanks to the Wyoming EPSCoR Community
College Transition Program.
Questions?
References
[1] Dalgaard, E., Christiansen, P., Larsen, J.J., and Auken, E. (2014). A temporal and spatial
analysis of anthropogenic noise sources affecting SNMR. Journal of Applied Geophysics,
110, 34-42.
[2] Ghanati, R., Hafizi, M., and Fallahsafari, M. (2016). Surface nuclear magnetic resonance
signals recovery by integration of a non‐linear decomposition method with statistical
analysis. Geophysical Prospecting, 64(2), 489-504. DOI 10.1111/1365-2478.12296
[3] Jiang, C., Lin, J., Duan, Q., Sun, S., & Tian, B. (2011). Statistical stacking and adaptive
notch filter to remove high-level electromagnetic noise from MRS measurements. Near
Surface Geophysics, 9(5), 459-468. DOI 10.3997/1873-0604.2011026
[4] Larsen, J.J., Dalgaard, E., & Auken, E. (2014). Noise cancelling of MRS signals combining
model-based removal of powerline harmonics and multichannel Wiener filtering.
Geophysical Journal International, 196 (2), 828-836. DOI: 10.1093/gji/ggt422
[5] Legchenko, A. and Shushakov, O. (1998). Inversion of surface NMR data. Geophysics,
63(1), 75-84.
[6] Müller-Petke, M., and Yaramanci, U. (2010). QT inversion—comprehensive use of the
complete surface NMR data set. Geophysics, 75(4), WA199-WA209.
[7] Walsh, D. (2008). Multi-channel surface NMR instrumentation and software for 1D/2D
groundwater investigations. Journal of Applied Geophysics, 66(3-4), 140-150.