ON THE USE OF THE CHECKERBOARD TEST TO ASSESS THE RESOLUTION OF TOMOGRAPHIC INVERSIONS

This note is devoted to the confrontation of intuitive ideas in the field of inverse problems, especially in tomographic seismological studies, with the results of a more rigorous approach. With the help of a simple example, we show that tests commonly used to illustrate the quality of inversion results can be misleading. Based on a classical mathematical analysis, we explain the origin of the problems that we have seen. Our main conclusion is that, in circumstances not so unrealistic, and in contradiction to a generally accepted idea, small-size structures like in the 'checker-board test' can be well retrieved while larger structures are poorly retrieved.