https://github.com/HYCOM/HYCOM-src/commit/00dfbdf06eb6573cfff424367808c028fe6ca1c6
bugfix to dp0cum in hybgenaj - "If you have encountered single column model blow ups in HYCOM that were resistant to halving the time step, they might have been due to this bug. In practice HYCOM's baroclinic time step, baclin, is often set to less than its theoretical maxium to maintain stability. It is possible that the bugfix will allow a longer baclin time step in practice, i.e. the instabilites that limited baclin might have been from this bug."
HYCOM codes are here - https://github.com/HYCOM/HYCOM-src
Bring in codes from the following commits -
https://github.com/HYCOM/HYCOM-src/commit/00dfbdf06eb6573cfff424367808c028fe6ca1c6 bugfix to dp0cum in hybgenaj - "If you have encountered single column model blow ups in HYCOM that were resistant to halving the time step, they might have been due to this bug. In practice HYCOM's baroclinic time step, baclin, is often set to less than its theoretical maxium to maintain stability. It is possible that the bugfix will allow a longer baclin time step in practice, i.e. the instabilites that limited baclin might have been from this bug."
https://github.com/HYCOM/HYCOM-src/commit/de5f57e093d31060047bf992ff3d2011730a8ebc Added the macro MASSLESS_1MM - "In mxkprf.f the "lowest substantial mass-containing layer" is at least 10cm thick by default. The compile time macro MASSLESS_1MM changes this to 1mm, which is the minimum possible in this context."