Inconsistent use of ew in verttransform

[adingwell]

I found that ew is not properly initialized in verttransform.f90

These are all the occurrences of ew in each of the verttransform*-files:

verttransform.f90:

real :: ew,dz1,dz2,dz
...
tvold(ixm,jym)=tt2(ixm,jym,1,n)*(1.+0.378*ew*(td2(ixm,jym,1,n))/ &
     ps(ixm,jym,1,n))
...
tvold(ix,jy)=tt2(ix,jy,1,n)*(1.+0.378*ew(td2(ix,jy,1,n))/ &
     ps(ix,jy,1,n))

verttransform_gfs.f90:

real :: ew,pint,tv,tvold,pold,dz1,dz2,dz,ui,vi
...
tvold=tt2(ixm,jym,1,n)*(1.+0.378*ew(td2(ixm,jym,1,n))/ &
     ps(ixm,jym,1,n))

verttransform_nests.f90:

real :: ew,dz1,dz2,dz
...
tvold(ix,jy)=tt2n(ix,jy,1,n,l)*(1.+0.378*ew*(td2n(ix,jy,1,n,l))/ &
     psn(ix,jy,1,n,l))

Notice that in verttransform.f90 ew is used as a variable, not a function. This variable is defined but not initialized.

[pesei]

I confirm this bug. It is already fixed by v10 developers in the later versions. Please note that v10 is not released yet (cf. ​https://www.flexpart.eu/wiki/FpRoadmap). Probably not a good idea to offer this in the download section, so I propose to remove it.

[pesei]

Download has been delelted. If someone wants to try out v10 (which is beta!) use git!