Comparing GYRE with ADIPLS on Model S

[earlbellinger]

Hello,

I am a new GYRE user trying to compare the results of GYRE and ADIPLS on Model S. I'm getting results that differ by about 0.05 uHz. This is certainly a small difference, but I was under the impression that the differences would be even smaller. I've attached a plot showing the differences as well as my GYRE and ADIPLS inlists.

gyre-adipls.png (15.54 KiB) Viewed 2504 times

Is there something that I am doing wrong, or is this the expected result?

Thanks for any assistance,
Earl Bellinger

[rhtownsend]

Hi Earl --

Thanks for the post, esp. for the informative plot.

The differences between GYRE and ADIPLS are certainly significant. I imagine this is because the grids used for the GYRE and ADIPLS calculations are rather coarse. Both GYRE and ADIPLS can resample the model onto finer grids. GYRE does this on the fly using parameters from the &shoot_grid namelist, while ADIPLS has a stand-alone tool (redistrb.d) to do the resampling.

To get very good frequency agreement (sub-picoHertz should be readily achievable), the resampling must be done in the same way for each code. Therefore, I suggest you use the ADIPLS tool to resample modelS.fgong onto a finer grid; and then use this resampled model in both ADIPLS and GYRE.

cheers,

Rich