Opened 5 weeks ago

Last modified 5 weeks ago

#2445 new Defect

Minimum roughness when equal to significant wave height

Reported by: jcastill Owned by: ayoung
Priority: normal Milestone: 2020 WP
Component: ZDF Version: 4.0-HEAD
Severity: minor Keywords: wave coupling roughness gls
Cc: mathiot, davestorkey

Description (last modified by jcastill)


When running in wave coupled mode and using roughness equal to wave height (nn_z0_met=3), some times there is a segmentation fault, or a crash when the velocity or SSH are too large.


In the vertical physics equations for the GLS scheme, the minimum roughness appears as a dividing factor. When it is too small or zero, non-physical results or divide by zero errors will be generated.


In the different formulations for surface roughness length in the GLS scheme, a minimum value is imposed (rn_hsro). It would be enough to simply impose a minimum value also when the nn_z0_met=3 scheme is used. This means that, in the zdfgls.F90 file, the line:

         zhsro(:,:) = rn_frac_hs * hsw(:,:)   ! (rn_frac_hs=1.6 see Eq. (5) of Rascle et al. 2008 )

could be substituted by:

         zhsro(:,:) = MAX(rn_frac_hs * hsw(:,:), rn_hsro)   ! (rn_frac_hs=1.6 see Eq. (5) of Rascle et al. 2008 )

I have already tested this modification and it solves the instability and segmentation fault problems, at the same time providing sensible results.

Commit History (0)

(No commits)

Change History (1)

comment:1 Changed 5 weeks ago by jcastill

  • Description modified (diff)
Note: See TracTickets for help on using tickets.