Introduction

The near surface presents many problems for seismologists, but it should also present many opportunities. While there are often severe velocity and absorption anomalies in the near surface, the recorded wavefield samples this region densely. Kjartansson Claerbout (1985); Kjartansson (1979) developed a method of locating lateral velocity variations. His method is based on taking the trace power in some time window and performing slant stacks of the power maps in order to determine the anomaly depth. The idea is that energy aligned along the midpoint axis is due to anomalies at depth and energy aligned closer to the shot and geophone axes is due to shallow anomalies. Claerbout 1993 repeats Kjartansson work and proposes a method of performing reflection tomography without picking.

I outline a method which incorporates Kjartansson's 1979 method and wavefield extrapolation to estimate near surface velocity distribution. The methodology is not yet complete, but the building blocks are being put in place. It is these building blocks that I present in this report.

I begin by modeling shallow velocity anomalies which are shorter in lateral extent than the spread length. Synthetic data are generated by upward continuing a synthetic wavefield through a v(x,z) model. The location of the velocity anomaly is determined by analyzing a plot of trace power in midpoint-offset coordinates. The trace power plots are then input to an iterative slant-stack algorithm to generate images of the velocity anomalies. I downward continue the data through the v(x,z) model to demonstrate that distortions due to the anomaly can be removed. This method can be used as an iterative inversion process to determine near surface velocity.