Determination of a time-dependent diffusivity in a nonlinear parabolic problem

Figures & data

Figure 1. The domain used in the numerical experiments.

Figure 2. Noise $e=0\%$: numerical value of ${\tilde{k}}_i$ using the first solution method (a) and the second solution method (b); $i=1,\dots ,50.$

Figure 3. Noise $e=0.5\%$: numerical value of ${\tilde{k}}_i$ using the first solution method (a) and the second solution method (b); $i=1,\dots ,50.$

Figure 4. Noise $e=1\%$: numerical value of ${\tilde{k}}_i$ using the first solution method (a) and the second solution method (b); $i=1,\dots ,50.$

Figure 5. The absolute ${\tilde{k}}_i$-error using the first solution method (a) and using the second solution method (b) for the different noise levels; $i=1,\dots ,50.$