If you want dissipated power in a TEM line, then P=Irms^2*R+Erms^2*G
(at a particular frequency where R and G have fixed values). Since
Irms and Erms are functions of position along the line,
I would dissagree with this last statement. The root mean square
is a type of averaging (not really just the average though), so how
could it be a function of position??