> why CrankNicolson over backward? @User I use semi-implicit schemes for describing the non-linear term. Besides, there is no point in going fully implicit if you have Co<1