geoclaw RP accuracy

[mjberger]

see Kyle's vortex test case for SWE: (https://github.com/mandli/vortex-example)

our riemann solver doesn't igve 2nd order accurcy even with limiters off

[dlgeorge]

for full 2nd order accuracy with any of the f-wave type solvers with source terms, don't we need another 2nd-order correction for the source term?

[mjberger]

This test case has no bathymetry, no source term. If I use the HLLE solver in fwave form, with the tangential and util components of the usual riemann solver, it is second order. So it's only rpn2_geoclaw.f that is an issue. — Marsha

…

On Apr 2, 2025, at 3:58 PM, David George ***@***.***> wrote: dlgeorge left a comment (clawpack/geoclaw#641) for full 2nd order accuracy with any of the f-wave type solvers with source terms, don't we need another 2nd-order correction for the source term? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread. <#641 (comment)> <https://github.com/notifications/unsubscribe-auth/AAGUGC7KOIO7AIEXQ6MXU3D2XQ6PVAVCNFSM6AAAAAB2KLLNM2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDONZTGU4TAMBTGQ> dlgeorge left a comment (clawpack/geoclaw#641) <#641 (comment)> for full 2nd order accuracy with any of the f-wave type solvers with source terms, don't we need another 2nd-order correction for the source term? — Reply to this email directly, view it on GitHub <#641 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGUGC7KOIO7AIEXQ6MXU3D2XQ6PVAVCNFSM6AAAAAB2KLLNM2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDONZTGU4TAMBTGQ>. You are receiving this because you authored the thread.

[dlgeorge]

hmm, that is peculiar...the rpn2_geoclaw.f should be the same as the HLLE solver in fwave form if there is no bathy. delta (hu, mom. flux) is decomposed into (1,s1) and (1,s2) where those are HLLE speeds. something to track down then.