Figure 11.

We plot a comparison between the model (12)–(15) and the age structured data from Japan by age class. We took $1/\nu =1/\eta =7$ days for each age class. Our best fit is obtained for $f_i$ which depends linearly on the age class until it reaches 90%, with $f_1=0.1$, $f_2=0.2$, $f_3=0.3$, $f_4=0.4$, $f_5=0.5$, $f_6=0.6$, $f_7=0.7$, $f_8=0.8$, $f_9=0.9$, and $f_{10}=0.9$. The values we used for the first day of public intervention are $D_i=13\mathrm{April}$ for the 0–20 years age class $i=1,2$, $D_i=11\mathrm{April}$ for the age class going from $[20,30[$ to $[60,70[$$i=3,4,5,6,7$, and $D_i=16\mathrm{April}$ for the remaining age classes. We fit the data from 30 March to 20 April to derive the value of ${\chi }_1^i$ and ${\chi }_2^i$ for each age class. For the intensity of confinement we use the values ${\mu }_1={\mu }_2=0.4829$, ${\mu }_3={\mu }_4=0.2046$, ${\mu }_5={\mu }_6=0.1474$, ${\mu }_7=0.0744$, ${\mu }_8=0.1736$, ${\mu }_9={\mu }_{10}=0.1358$. By applying the method described in Appendix A, we obtain ${\tau }_1=0.1630$, ${\tau }_2=0.1224$, ${\tau }_3=0.3028$, ${\tau }_4=0.2250$, ${\tau }_5=0.1520$, ${\tau }_6=0.1754$, ${\tau }_7=0.1289$, ${\tau }_8=0.1091$, ${\tau }_9=0.1211$ and ${\tau }_{10}=0.1642$. The matrix $\varphi$ is the one defined in (18).