Approximate imaging without a precise velocity model: Path-summation approach

This paper presents a new method to calculate approximate seismic images without performing velocity analysis or using a priori knowledge about the precise subsurface model. By analogy with the use of Feynman's path integrals in waveform modeling, we develop an algorithm for data-driven (model-independent) imaging by summation of elementary signals over a sample of all possible paths between source and receiver. The constructive and destructive interference of these signals produces an approximate image of the subsurface. Application of the method to synthetic and real data illustrates its feasibility.