An efficient interpolation approach for insufficient 3D field data

Abstract

Complete three-dimensional (3D) land seismic data are often difficult to acquire due to physical or financial limitations. Therefore, in recent years, interpolation techniques with 3D field data have played an important role in seismic data processing. To improve the efficiency of interpolation for insufficient 3D field data, we developed a new interpolation process that applies two-dimensional curvelet transform-based projection onto convex sets (POCS) to the k(x)-k(y) transformed data of each time slice of 3D data. Additionally, to acquire accurate interpolation results with the proposed new method, we designed the preparation process to render the input data irregularly distributed with small-sized gaps. We found that the interpolation results of our proposed new method were similar to those using a 3D curvelet transform-based POCS method, with reduced computational cost. The quality of the stacked images of sparsely sampled 3D field data was significantly enhanced by applying our interpolation approach.