open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I've been reading more in-depth on the "SCHEME OPTION FOR ADVECTION..." I want to use a second-order scheme compatible with dry zones/tidal flats. Currently I've been using PSI with the LIPS scheme-option (ADVECTION SCHEME: 5, ADVECTION OPTION: 4).

PSI is second-order in space. However, reading "Distributive advection schemes and dry zones, new solutions (Hervouet et al., 2015) I notice that it mentions LIPS isn't a second-order scheme. Reading further, in "Residual distribution advection schemes in Telemac" (Hervouet et al., 2017) it mentions that LIPS is not second order in time. Am I right in thinking that LIPS is second-order only in space?

Also, is the ERIA scheme usable with tidal flats and is it available to use? I know it is second order in space but not in time, but is a lot less diffusive. Using advection option 15 and treatment of negative depths 3 I get the message "SCHEME OPTION FOR ADVECTION UNKNOWN" (version v7p2r3).

Thanks!

Hello,

I am not a specialist about advection schemes, but I will try to answer your questions.
For more accurate answers I recommend you check the PhD thesis by Sara Pavan: www.theses.fr/2016PESC1011.

- it seems to me that the classical PSI scheme implemented in TELEMAC is only 1st order in space and time, maybe the best is to check this in the abovementionned work

- if you want to use less diffusive schemes with tidal flats in TELEMAC-3D, you need to choose the PSI advection scheme (or N) and then specify the value of the option for the advection scheme: SCHEME OPTION FOR ADVECTION OF TRACERS = 4. Without tidal flats, the scheme option 1 is the classical explicit PSI scheme, option 2 is the first-order predictor-corrector and option 3 is the second-order predictor-corrector. I think the PSI scheme with option 4 is second-order but I'm not sure. Again, I recommend you check this in S. Pavan's PhD thesis.

Maybe Sara will improve my answer when she's back from holidays :)

Best regards,
Agnès

Hello,

LIPS scheme is second order in space only in steady cases (like classical PSI scheme). For unsteady problems it remains a first order scheme but however it shows very low numerical diffusion compared to other first order schemes.

As you say, to choose the LIPS, you have to set ADVECTION SCHEME: 5 (4 is also ok) and SCHEME OPTION FOR ADVECTION OF TRACERS: 4. I think is one of the best option if you have tidal flats and you want low numerical diffusion.

ERIA scheme is not yet available in TELEMAC-3D (but it is available in 2D).

Best regards,
Sara