[Dart-users] updates to DART subversion repository

[Nancy Collins]

hello dart-users,

we've just finished a round of commits to the DART subversion
repository.   i've detailed the most significant updates below,
but you might want to skim the first 2 in particular - one is
a change to how the input_qc_threshold is handled in the filter
code,  and an updated wrf model_mod.f90 file has been committed.

these changes have all been compiled on multiple platforms,
but as always, please contact me if you run into any problems.

have a good weekend,
nancy


1. filter input_qc_threshold has always been documented as
discarding incoming obs which have a data qc *greater than*
the threshold value.  however, the code has always discarded
values *equal to or greater than* the threshold.  the code has
been changed to match the documentation since that's more
consistent with other threshold values in the system.  all the
input.nml files have had their defaults changed from 4 to 3 so
the behavior will be the same as before using the default values.
(this threshold was selected to match the values used in the
NCEP observation files).  the corresponding values in obs_diag
have been changed in the same way to be consistent.

2. updated wrf/model_mod.f90 -- it now can interpolate some
additional hydrometeor fields, as well as the land mask and
a few others.  i've added a file - wrf/wrf_state_variables_table
which has explanations for what values can be put into the
namelist to control which wrf fields are made part of the
state vector.   there were also a couple minor bug fixes from
josh for the single column model, and ryan for outputting
precipitation fields in the diagnostic files.

3. new KINDs added in obs_kind/DEFAULT_obs_kind_mod.F90
to support the mixing ratios for the hydrometeors, and some
additional kinds for radiance forward operators.

4. added support for leap_year() and days_in_year() for
gregorian calendars.

5. updated the World Ocean Database observation converter
for ocean temperature and salinity observations.

6. committed an updated version of the radar obs_def code.  if
you're interested in using it, contact glen romine or david dowell.
they're adding in support for various physics schemes which
affect which fields should be used to compute the various
radar forward observation operators.

7. perfect_model_obs now has the same options for outputting
any failed forward operators to a separate text file.  the output
contains the observation key, and the istatus code from the
model_interpolate() routine.  (the corresponding function in
filter does the same, except it also outputs the ensemble member
number as well, and has separate files for the prior forward
operator and posterior forward operator calculations).

8.  if the dart matlab directories have been added to your matlab
search path, 'help DART' will print out a summary of the various
diagnostic routines supplied in the standard distribution.