Stages of the LFS processing and interpretation
1. Filtering of strong rolling wave noise from non-moving surface sources using a linear
2. Filtering of an ambient background surface noise from everywhere using a crosscorrelation
function between Z-components of two records.
3. Wells data analysis (oil depths, oil-filled thickness (net pay))
4. Model preparation using vertical seismic profiling (VSP) and other conventional seismic
results (structural and isochronous map).
5. Numerical simulation of the vertical component of media responses.
6. QC of the filtering and simulation stages by matching filtered spectra nearby wells with
7. Construction of correlation maps.
8. Hydrocarbon presence probability map construction.
The goal of the LFS Technology data processing is to get the spectra of vertically
directed P-waves from ambient background microseismic noise. A major problem is that rolling
waves (Scholte waves (on seabed) or Rayleigh waves (onshore)) and other types of waves are
mixed with useful background signal (vertically directed P-waves). Let us consider rolling waves
of two classes:
1) Strong rolling waves from non-moving surface sources (pumps, drilling equipment,
2) Ambient background rolling waves from everywhere – omnidirectional rolling waves.
Stage 1. The first type of rolling waves (generated by non-moving surface sources)
effectively filters out using the linear prediction algorithm, which takes into account the
correlation dependence between components of rolling waves in the microseismic record. This
algorithm subtracts from the vertical component of the record the parts that are correlated with
horizontal components of the noise. Fig. 1.1 presents the result of filtering of o.p.1 using the
linear prediction algorithm. Fig. 1.2 presents the result of filtering of o.p.2 using the linear