Time Exposure Acoustics for Imaging Underground Structures

摘要

The coal mine study demonstrates that passive imaging is feasible under realistic conditions, but improvements are certainly possible. For example, some noticeable phase rotation of the seismic traces across the array is evident when the traces are plotted (e.g., as in Figure ii but at higher magnification). The phase rotation is probably due in part to dispersion of the seismic waves as a function of path length (greater path lengths imply greater dispersion). Phase rotation will degrade the resolution of the image. Simulations show that images derived from idealized data in the absence of phase rotation are of higher quality. No phase rotation implies that the signals recorded at different geophones are merely delayed versions of on another, where the delay is proportional to path length differences from the source to the receivers. Phase rotation due to dispersion can be corrected to some extend by inverse filtering, but was not attempted here. The reason for this is that rotation due to pure dispersion due only to path length differences was not clearly evident in the raw data, since systematic phase distortion as a function of path length was not seen across the array. This seems to imply that the phase distortion is probably due primarily to velocity inhomogeneities along the different propagation paths. Partial compensation for this might be achieved by incorporating a layered-earth model into the imaging algorithms, although this would require some a priori knowledge of the local stratigraphy.