[RicardoGoncalves]

Hi everybody,
I have been reading very carefully the book by E. Toro on shock capturing methods for shallow water flows and cannot understand why it uses the criteria h*>hL (or h*>hR) to define the situation when a shock will occur. Shouldn't it be labmda*>lambdaL [where lambda is the eigenvalue of the Jacobian matrix, e.g. u+sqrt(gh)]. Does anyone has read that book and found the same issue?
Thanks,

Ricardo

[Il Prete Rosso]

Lax entropy condition gives
0 > lambdaL - lambda*
0 > uL+sqrt{ghL}-(u*+sqrt{gh*})
0 > uL-2sqrt{ghL}+3sqrt{ghL}-(u*-2sqrt{gh*}+3sqrt{gh*})
(Riemann Invariants, comes from integral curves)
0 > C - C +3(sqrt{ghL}-sqrt{gh*})
h* > hL