High frequency recovery via spectral match between surface seismic and crosswell seismic data

Abstract

With the closer location of sources and receivers to the reflectors, avoiding the drastic absorption from the weather layer, crosswell seismic acquisition can provide higher-frequency observations than that of surface acquisition. Making use the spectral relationship of them may help to recover the lost high frequency information in surface seismic data. We present a method to recover the high frequency data of surface seismic via the spectral match relationship between surface seismic data and crosswell seismic data. Considering the excitation frequency of crosswell seismic is much higher than that of surface seismic, we utilize Morlet wavelet decomposition to extract the crosswell seismic data within the recoverable bandwidth of surface seismic data. The match equation has been built in the frequency domain, with adding the regularization term to enable solution more stable. We use seismic wavelets as input data, since they not only have more stable spectrums than seismic traces, but also represent the comprehensive properties of specific acquisition and earth medium. The method has been applied on the real data, the result shows that the high frequency is recovered without loss of low frequency, consequently, the recovered seismic section could be able to better delineate small scale and detailed information.