Forcing Integration Height Consistency in LOS Delay Calculation

[mohseniaref]

Hi all,

When using RAiDER for LOS atmospheric delay calculations, I noticed that for raytracing, the code allows you to specify a maximum integration height (zref), but for Zenith/Projected LOS calculations, it defaults to the vertical grid provided by the weather model file.

My concern is that weather model heights are often provided as geopotential heights, while zref (and many geodetic tools) use ellipsoidal height. This could lead to inconsistencies if I want to restrict LOS calculations to a specific altitude (e.g., 18 km ellipsoidal).

Is there a recommended way to ensure the integration is done up to a consistent physical height (e.g., 18 km ellipsoidal) regardless of height type in the weather model?

Should I convert all weather model levels to ellipsoidal height before slicing?
Is there a built-in utility or config option in RAiDER to handle this?
Is there a risk of mixing reference systems if I use wm_levels <= zref directly?
Any advice or best practice for this scenario would be greatly appreciated!

Thanks!