Application of wavelet compression algorithm in linear geophysical inversion

摘要

Large-scale linear equation should be solved in geophysical inversion. It needs large memory resources and fast computer speed. For the limitation of computer's storage and computation capability, it's limited for geophysical inversion to solve real problems. In fact, Large redundancy data is included in kernel matrix and can be removed by wavelet compression. We convert kernel matrix to sparse matrix by wave compression method. The matrix operations are carried out in wavelet domain by sparse matrix algorithm. The example of density inversion shows that wavelet compression algorithm can reduce the memory volume and save CUP time obviously.