first of all, the GOTM team is happy to present a new member, Georg
Umgiesser from ISDGM in Venice, Italy. Georg did and will do useful
work in the area of linking GOTM to 3D models, structure of the model etc...
The GOTM team is furthermore glad to announce the release of GOTM v.
1.2.0, which contains two basically new features:
a separation of the mean flow and the turbulence part in order to
allow for easy inclusion of GOTM turbulence closure schemes into three-dimensional
a number of new stability functions.
For the details, see below.
This new GOTM version needs separate namelists for general settings
and for turbulence specifications,
such that the old namelists have to be replaced. The existing GOTM
scenarios (test cases with forcing and
validation data) will be equipped with these new namelists. For those
who would like to continue working
with the old version GOTM v. 1.1.3, the old namelist can still be downloaded
from the GOTM web site.
GOTM v. 1.2.0 is still written in FORTRAN77, since many three-dimensional
models are not yet on
FORTRAN90 standard. Before end of this month, we however intend to
release GOTM v. 2.0.0 which will
be written in FORTRAN90. It will include all innovations of GOTM v.
1.1.3. Moreover, GOTM v.2.0.0 will have many more innovations and will
be accompanied with numerous new test cases. This will in detail be discussed
in a further round mail by end of this month.
For those who are confused about our version numbering:
GOTM v. 1.1.3: F77 version, released Dec 28, 1999.
GOTM v. 1.2.0: F77 version, released today, Feb 18, 2000. New structure,
new stability functions.
GOTM v. 2.0.0: F90 version, hopefully released late Feb 2000. Complete
rewrite to F90.
The innovations of GOTM v. 1.2.0 in detail:
Separation of mean flow and turbulence models. GOTM is rewritten such that
all turbulence models are called by only one subroutine 'gotmturb' in the
main loop. All mean flow properties needed for the forcing of the turbulence
part are included in the parameter list of 'gotmturb'. These are shear
and buoyancy frequency, layer thicknesses, and surface and bottom friction
velocity. There are two include files now, one for general and mean flow
settings (const.i) and one for turbulence settings (gotmturb.i). The turbulence
subroutines are by means of this completely separated from the rest of
GOTM. The turbulence subroutines together with the turbulence common block
and a turbulence namelist (for specifying turbulence model settings) may
now be incorporated into a three-dimensional ocean model such as MOM (Modular
Ocean Model). On the side of the 3D model, an interface to GOTM extracting
water column vectors in GOTM format from the 3D model format would have
to be created. GOTM would then have to be called at each water column when
the 3D model is looping over the horizontal grid. The GOTM team will work
on a prototype for such an interface between GOTM and a 3D model. All present
or future GOTM users are invited to participate in this effort. Which will
be principally more comfortable with the F90 version. We would like
to acknowledge here Encho Demirov
(for mailing address, see GOTM
user group) for reviving our interest in separating turbulence and
mean flow during a visit at IfM Hamburg, Germany. He suggested to make
a coupling between GOTM and MOM.
New stability functions. Prof. Vittoro
Canuto (for mailing address, see GOTM
user group) made a manuscript (submitted to J. Phys. Oceanogr.) about
new stability functions available to the GOTM team. These functions prove
to allow for mixing even at higher Richardson numbers and therefore solve
a problem always associated with differential turbulence closure models:
that they predict a mixed layer depth too low and therefore a sea surface
temperature too high. Hans Burchard and Karsten Bolding submitted a manuscript
to J. Phys. Oceanogr., in which they show how these new stability functions
may be properly incorporated into a k-epsilon model. A file of this manuscript
in electronic format (ps or pdf) may be downloaded from the GOTM web site,
see publications. A number of other new, more simple stability functions
are implemented as well.
For the k-kL model, the buoyancy parameter E3 may now be read in from the
namelist. In a short manuscript, Hans Burchard shows how this parameter
can be properly estimated, see the electronic file for download in 'publications'.
The turbulence closure model by Bougeault & Andre (1986) has been completely
rewritten. Some bugs were removed. It does now not any more require extra
evaluation of the equation of state. We would like to acknowledge Henrique
Coelho (for mailing address, see GOTM
user group) for assistance in improving the code. This subroutine is
still in a preliminary state, for comments, please contact Manuel
A small bug in the flux boundary condition for the dissipation rate epsilon
had already been removed for version 1.1.3. It was Patrick
Luyten who found it and won a beer ...
Any questions/comments/suggestions are warmly welcome.
Finally, we would like to inform you about the possibility to make your
turbulence related and freshly submitted manuscripts available to
the GOTM public by sending them in electronic form (ps and pdf formats
preferred) to our webmaster Pierre-Philippe Mathieu. He would place your
submitted manuscript in the Publications
section. Since, for some technical reasons, we do not yet provide an automatic
mailing list, you may send an accompanying text to Hans
Burchard, which would then be sent to the complete GOTM mailing list.
A copy of that text will as well be sent for further dissemination to Ilse
Hamann, the administrator of the CARTUM