Remarks on viscosity solutions for evolution equations (Q1185147)

The following degenerate parabolic equation is considered: \[ {{\partial u} \over {\partial t}}+ F(t,x,u, \nabla u, \nabla^ 2 u)=0. \tag{1} \] It is proved that a viscosity subsolution of (1) is left accessible at each (parabolic) interior point of the domain of definition for a wide class of \(F\). Also the relation between viscosity subsolutions defined on \((0,T)\) and those on \((0,T]\) is clarified.