In Treeby et al. (2012), the numerical results are compared with the analytical solution in Fig. 3(c).

However, the time axis of these two data may not accord with each other due to dispersion.

Thus, can anyone tell me how to calculate `the least squares error`

?