search

fmihpc / vlasiator

Vlasiator - ten letters you can count on
https://www.helsinki.fi/en/researchgroups/vlasiator
Other
43 stars 35 forks source link

AMR refinement parameters are actually half-widths/heights/thicknesses #792

Open ykempf opened 1 year ago

ykempf commented 1 year ago 
[Magnetosphere]
refine_L1tailthick = 6.371e7 # 10 RE
refine_L2tailthick = 2.5e7 # 3.92 RE
refine_L3tailheight = 1.0e7
refine_L3tailwidth = 8.0e7

These parameters are actually half-widths, heights, thicknesses, which got me confused in a test. Should they be renamed? And while at it, height is the same concept as thick in these.

lkotipal commented 1 year ago

Thick is y and z, height and width are single dimensional. They should really be renamed though, with maybe explicit dimensions in height and width as well, eg.: refine_LN_tail_yz_max refine_L3_tail_y_max Or something to that effect.

markusbattarbee commented 11 months ago

refine_L1tailthick and refine_L2tailthick are not evaluated in the y-direction, only in z. The y-directional width of the tail sheet is dictated by the L1 or L2 ionospheric refinement region radius.

Thus the L3 tail sheet uses height and width, which dictate the box extents in the y-z-plane. So I kinda like having a different name for that (thick vs height).

ykempf commented 10 months ago

Well, OK then, close this and make a mental note that these are halves?

markusbattarbee commented 10 months ago

or learn to read the comments in Magnetosphere.cpp :)