Exactly. Well, it's not an easy task, but I'm sure that there are some artifices where the fluxes can be combined and use them in relation to the bottom heat flux, Cp and rho. Again, the problem is when dealing with the temperature difference because it's a dynamic one. I was thinking that...