Measuring, imaging and suppressing scattered surface waves Xander Campman*, Kasper van Wijk, John Scales and Gérard Herman** * Dept. of Applied Mathematics Delft University ** Shell Int. E & P and Delft University
Near-receiver scattering
Vs=400 m/s, Vp=1700 m/s Vs=600 m/s, Vp=2500 m/s h d w Example w d < h <<
Vertical velocity
Aligned at first breaks
Stacked
Scattered noise model
Outline of the scheme 1. Estimate velocity of the scattered surface wave u G 2Estimate near-surface scattered energy from one isolated event. 3 Image the scattered energy impedance distribution, 4 Predict and remove scattered surface waves on the entire record
Validation experiment A B
Seismic record, inline Direct Rayleigh wave Sound wave Ghost Rayleigh wave Scattered Rayleigh wave Cavity AB
Step 1: Estimate surface wave velocity c ~ 3000 m/s AB
Step 2: Separate scattered energy Direct Rayleigh wave : d Scattered energy : d 1 ABAB
Step 3: Estimate impedance function Impedance function: cross-section at t=0 for entire data volume A B AB
Scattered noise model
Step 4: Predict scattered energy Ghost Rayleigh wavePredicted scattered field AB AB
Step 4: Subtract predicted scattered energy -= AB AB AB Ghost minus scattered
Transmission experiment Top view Side view AB
Transmission data AB
Step 1: Separate scattered energy First arrival : dScattered energy : d 1 AB AB
Step 2: Estimate impedance distribution Impedance function cross-section at t=0 for entire data volume AB AB
Step 4: Predict scattered field on the rest of the record AB
Input data after dip-filtering AB
Output data AB Rest minus scattered
Latest experiment
Acknowledgments Dutch Technology Foundation (STW) for financial support. Army Research Office and NSF also for financial support Everyone at PAL and CWP Shell Int. E&P and Wim Mulder (SIEP) for the use of FD code Jan Willem van den Dries for the artist impression