Aquifer properties can be obtained from envelopes of surface nuclear magnetic resonance (NMR) signals, but this demands high-quality data. To retrieve reliable envelopes using synchronous detection from the intrinsically low signal-to-noise ratio (SNR) surface NMR recordings, a variety of signal processing techniques are employed to mitigate noise. We present a different approach to retrieve complex envelopes using spectral analysis and a sliding window, which can potentially improve SNR significantly. The complex envelope is composed of the spectral values at the Larmor frequency found through the Fourier transform of surface NMR data using a sliding window. We discuss how to maximize the SNR of envelope by selecting the optimum length and shape of the sliding window. An accompanying method for determining the Larmor frequency is presented and we address how noise can deteriorate the envelope retrieval in spectral analysis. Results obtained from synthetic models and field measurements in low and high noise environments reveal that the proposed method not only improves the accuracy and efficiency of envelope retrieval, but also eliminates the transient distortion of early-time signal caused by the filtering procedure.
- Instrumental noise
- Time-series analysis
- Programme Area 2: Water Resources