Skip to content

Processing Options

Spectral Analysis Tools

SurfQuake provides three main tools for spectral analysis of seismic data: spectrum, spectrogram, and CWT (Continuous Wavelet Transform). These tools enable users to inspect frequency content in different forms—static, time-evolving, and high-resolution wavelet representations.

Spectrum

The spectrum tool computes the frequency-domain representation of a single seismic trace using the multitaper method by default, which is robust against noise and spectral leakage.

Parameters

  • method: Spectral estimation method. Options:
  • multitaper (default)
  • fft or others if extended

Config Example

Analysis:
  process_1:
    name: 'spectrum'
    method: 'multitaper' # 'fft' for conventional spectrum with single taper
    output_path: /Users/surf_test/spectral

Output

  • Amplitude vs. Frequency plot
  • Optional log-log or semi-log axes (loglog, xlog, ylog)
  • Serialized result saved as .sp

Spectrogram

The spectrogram method produces a time-frequency representation of a seismic trace using short-time Fourier Transform (STFT).

Parameters

  • win (float): Window length in seconds (e.g., 5.0)
  • overlap_percent (float): Percent overlap between windows (e.g., 50.0)
  • linf, lsup: Lower and upper frequency limits for plotting

Config Example

Analysis:
  process_1:
    name: 'spectrogram'
    win: 5.0
    overlap_percent: 50.0
    linf: 0 # optional
    lsup: 20 # optional
    output_path: /Users/surf_test/spectral

Output

  • Power over time and frequency in decibels (dB)
  • Plotted as a color-coded matrix
  • Serialized result saved as .spec

Continuous Wavelet Transform (CWT)

CWT provides high-resolution time-frequency analysis using scalable wavelets. Ideal for transient signals and high-precision timing.

Parameters

  • wavelet_type:
  • "cm": Complex Morlet (default and recommended for phase coherence)
  • "mh": Mexican Hat
  • "pa": Paul wavelet
  • param: Wavelet parameter (e.g., 6.0)
  • fmin, fmax: Frequency range
  • nf: Number of frequency bins

Config Example

Analysis:
  process_1:
    name: 'cwt'
    wavelet_type: 'cm'
    param: 6.0
    fmin: 0.5 # optional
    fmax: 20.0 #optional
    nf: 80
    output_path: /Users/surf_test/spectral

Output

  • Scalogram in decibel power
  • Prediction masks showing reliable analysis range
  • Serialized result saved as .cwt

Plotting and CLI Usage

All spectral results can be visualized via the CLI:

# Plot a spectrum
surfquake specplot --file path/to/file.sp

# Plot a spectrogram
surfquake specplot --file path/to/file.spec 

# Plot a wavelet scalogram with optional clipping
surfquake specplot --file path/to/file.cwt --type cwt --clip -120

# Save output to file
surfquake specplot -f file.spec --save_path output.png

CLI Options

  • --file, -f: Path to serialized .sp, .spec, or .cwt file
  • --type, -t: Type of spectral result to plot (spectrum, spectrogram, cwt)
  • --clip, -c: Optional clipping in dB for spectrogram or CWT
  • --save_path: Output path to save the figure

Scientific References

  • Percival, D. B., & Walden, A. T. (1993). Spectral Analysis for Physical Applications: Multitaper and Conventional Univariate Techniques.
  • Torrence, C., & Compo, G. P. (1998). A practical guide to wavelet analysis. Bulletin of the American Meteorological Society, 79(1), 61–78.