Spectral Decomposition & RGB(A) blending in OpendTect Spectral decomposition and RGB(A) blending are valuable instruments for studying sub-seismic resolution stratigraphic featu...
Spectral Decomposition & RGB(A) blending in OpendTect
Spectral decomposition and RGB(A) blending are valuable instruments for studying sub-seismic resolution stratigraphic features.
OpendTect’s attribute engine boasts four different attributes for spectral decomposition. Most of these attributes support multiple algorithms and options, and all are released in the free, opensource, part of the system. The “Spectral Decomposition” attribute supports short-window FFT (Fast Fourier Transform) and CWT (Continuous Wavelet Transform) algorithms. CWT supports three wavelets; Morlet, Gaussian, and Mexican Hat. The “CEEMD” attribute supports EMD (Empirical Mode Decomposition), EEMD (Ensemble EMD), and CEEMD (Complete Ensemble EMD) algorithms. “Spectral Decomposition by Local Attribute” is a free attribute developed by Wayne Mogg. Finally, you can decompose seismic data with the “Frequency Filter” attribute using either FFT or Butterworth filters.
Spectral decomposition and RGB(A) blending as a sub-seismic analysis tool works because the amplitude of a frequency component brightens when the layer thickness of a feature is in the tuning range. Colors in an RGB(A) blended image can thus be interpreted in terms of thickness variations. Typically, the user computes a range of components over the full seismic bandwidth. From this range three iso-frequencies are selected. The lowest frequency is displayed in the red channel, the middle frequency in the green channel and the highest frequency in the blue channel. Optionally, to highlight faults in the data, a discontinuity attribute such as similarity, can be displayed in the Alpha (transparency) channel.
In the example shown in this movie we use a dataset from New Zealand. A color blended volume combines 8, 26, and 50 Hz components in the red, green and blue channels, respectively. The display and the sliding through the volume is done in a Wheeler (flattened) scene. The RGB blended volume shows clearly how channels evolved and changed course over time from one single feeding channel to multiple stacked, meandering channels.
Author: Hesham Refayee from dGB Earth Sciences
#OpendTect #seismic #geophysics #geology #spectraldecomposition
