Hello CROCO community. Past ROMS and new CROCO user here.
I’ve been working on reproducing the SWASH test-case and noticed a possible inconsistency regarding the GLOBEX experiment used as the reference.
Marchesiello et al. (2021) explicitly state that the simulation focuses on experiment B3, but the wave characteristics reported in the paper (a1 = 0.09 m, a2 = 0.01 m, f1 = 0.420 Hz, f2 = 0.462 Hz) actually correspond to experiment B2 in Michallet et al. (2014).
This distinction is also reflected in CROCO’s wave_maker.h, where the two configurations are defined as:
B2: a1 = 0.09 m, a2 = 0.01 m
B3: a1 = 0.07 m, a2 = 0.03 m
Furthermore, the results presented on the CROCO SWASH test-case webpage appear to resemble the B2 experimental observations more closely than B3 (see the figures).
Could you clarify whether the simulation presented in the paper and on the website was actually performed using the B2 configuration, despite being referred to as B3?
This is not a bug/fix issue of the code. Merely need just correct clarification on the actual test. I’m asking because I’m working on reproducing the benchmark and would like to ensure that I’m comparing my results against the correct experimental case.
Also, I would truly appreciate if someone could give me some hints regarding the Fig.2. of the Marchesiello paper. I downloaded the experimental records from Zenodo, but have no clue the figure mentioned, uses which time instant of the records? could it be average?
@MoeinDst
For the B2/B3 issue, I would also suggest check the actual wave_maker.h configuration.
I would not state definitively that Fig. 2 is an average unless you verify it.
Best regards,
Well, this is to say the figure shown under the test-case explanation, is with B2, in contrast to the text explaining the SWASH test-case set to B3 (see the figures attached at the begining of the thread). In fact the CPP options of the case provided in the version [2.0.1] was set to define SWASH_GLOBEX_B2.
So bottom line is, if you want to obtain the same figure, you should use the B2 experiment!
what you mean by it. if someone work on FVCOM then ?
I work on oneAPI Threading Building Blocks (oneTBB)
I will ask patrick marchesiello about it to help you.
Regards,
What I mean is that running the SWASH, if you define SWASH_GLOBEX_B2, you get the same figure as shown in the test-case, and with define SWASH_GLOBEX_B3 you get sth different (as expected cuz the a1,a2,f1, and f2 differ betwwen the experiments). The introduction to Swash, however, states that results are obtained with GLOBEX experiments B3! So either the figure needs to be updated or the text (look at the waves between 50-70 in x-direction).
