hsubp.f

HSUBP computes the local pressure scale height at one point, the edge of a convective region, for the overshoot calculation (comment: "PD Dec. 1984, revised for YREC/Mark6 by MHP 2/91").

• Who calls it: only OVERSH, with JEDGE the point at the edge of the convection zone.
• Inputs: the model's composition HCOMP, and $\log\rho$, $\log P$, $\log r$, $\log m$, $\log T$ (HD, HP, HR, HS, HT). Output: PSCAHE, the scale height in cm.
• Calls: EQSTAT, the equation of state.
• Modules: the constants of CONSTANTS_MOD (for CLN and $\log G$, CGL); DEBHU_MOD, where it sets the composition used by the Debye–Hückel correction (XXDH, YYDH, ZZDH, ZDH(1..3)) before calling the EOS, shared with EQRELV, PHYSIC, COEFFT and others; and the flag LDH from NMLPAR_MOD (formerly in /DEBHU/).

When LDH is true, HSUBP overwrites XXDH, YYDH, ZZDH, ZDH(1..3) in DEBHU_MOD with the composition at JEDGE (lines 60–67↗). The interior's EOS callers set them again first, shell by shell (coefft.f, lines 207–209↗; physic.f, lines 100–102↗), as CRRECT does before SURFBC (crrect.f, lines 136–143↗), so the overwrite does not reach them.

The scale height returned is the hydrostatic one,

$$H_p = \frac{P}{\rho g} = \frac{P\,r^2}{\rho\,G\,m},$$

computed in logs: PSCAHE $= 10^{\log P + 2\log r - \log\rho - \log G - \log m}$. Here $\log\rho$ is not the model's HD(JEDGE): DL is an output argument of EQSTAT, which recomputes the density from $P$, $T$, $X$, $Z$, and PSCAHE uses that value. The EOS call also returns RMU (probably $\mathcal{R}/\mu$), which goes only into a second, ideal-gas estimate PSCAP $= (\mathcal{R}/\mu)\,T r^2/(G m)$, which is never returned (unused?).