Assume we have $F \in C^2(\mathbb{R})$ and $u_0 \in L^\infty(\mathbb{R})$ with sompact support. I am wondering whether the entropy solution to the Cauchy Problem on $(0,T) \times \mathbb{R}$: $$ \partial_tu(t,x)+\partial_x F(u(t,x))=0, \quad u(0,\cdot)=u_0 $$ is compactly supported as well. Intuitively, one might expect that the initial datum ist transported, i.e. support indeed remains compact. Does anyone have a reference/proof/counterexample?
Thank you!
Likely yes. I wonder, does the following help you to find a reference?
[Andreianov...] say:
That talking about finite speed of propagation is promising. (Thought the term might be misleading.)
They talk more about the pure hyperbolic local case (A = 0) later and I think they circle around your topic :
I hope that the works they cite can you lead to something like that if $F$ has (a certain degree of Holder) continuity, you have a positive result.
Andreianov, Boris; Brassart, Matthieu, Uniqueness of entropy solutions to fractional conservation laws with “Fully infinite” speed of propagation, J. Differ. Equations 268, No. 7, 3903-3935 (2020). ZBL1473.35619. (HAL)