Supervirtual Refraction Interferometry on Noisy Land Seismic Reflection Data

Abstract

Supervirtual refraction interferometry (SVI) uses cross-correlation to obtain virtual traces and then convolves them with the real ones resulting supervirtual refraction traces. The method can be used to enhance the first arrivals in seismic refraction data in order to easily pick them for the static corrections. In this paper SVI was applied on common-source gathers (CSGs) recorded in a land reflection seismic survey from an area with relatively large elevation variation. The paper analyzes how the level of noise of the CSGs used in SVI influence the quality of the supervirtual first arrivals. The dataset has 160 CSGs and the ones with the weakest first arrivals were selected. After removing all signal and noise 100 ms above and 200 ms below the first arrivals SVI was applied in two cases: 1 – nine out of ten CSGs used have high signal-to-noise ratio (SNR) and continuous and high amplitude supervirtual first arrivals were obtained; 2 – nine out of ten CSGs used have low SNR and even though the initial first arrivals improved they are still discontinuous and cannot be picked for accurate static corrections. The SNR of the best supervirtual first arrival is considerably higher than the initial one.