Fig. 9.

By employing MVIM, we have recovered $g_1(\tau )$ at r = 2 cm for a two layer model for (a) without prior i.e., $x_0=0$ (from $\kappa (T)$ without noise) (b) with prior information i.e., $x_0=1$ (from $\kappa (T)$ with noise) (c) by iterative scheme (from $\kappa (T)$ with noise) (d) Sensitivity curves indicating that there are two peaks, as there are two different flows.