Similar to #541 this calculates surface fluxes from previous and not current time step for numerical stability with leapfrog scheme. This also seems to make the flux limiters largely unnecessary, so I've just increased them somewhat and it seems to work well! Maybe we could get rid of them entirely, but leaving them in for now.
Similar to #541 this calculates surface fluxes from previous and not current time step for numerical stability with leapfrog scheme. This also seems to make the flux limiters largely unnecessary, so I've just increased them somewhat and it seems to work well! Maybe we could get rid of them entirely, but leaving them in for now.