2 m Temperature diagnosis wrong with 5-layer soil model (V2.0.2)
(posted 8/18/04)
Problem: The 2 m temperature,
a diagnosed variable, is wrong if sf_sfclay_physics = 1 AND sf_surface_physics
= 1 options are selected.
Solution: coming in 2.0.3.
More on specified boundary condition problem (V2.0.2) (posted
8/18/04)
Problem: An additional problem
was found with the specified boundary condition option. This has
to do with the geopotential variable. Again small effect in most
cases. Potential large impact on longer simulations.
Solution: Will be corrected
in 2.0.3.
Error Computing I/O Intervals and Fractional Time Steps (V2.0.2)
(posted
6/17/04)
Problem: Code in the ESMF Time
Manager library provided with WRF
contains an error in computing the fractional part when a time
interval
is divided by an integer. Also, the code to set a ringing ESMF
alarm to
ring at the next time interval was incorrect.
Solution: Download new ESMF_Clock.F90
and ESMF_TimeInterval.F90
files and replace the versions in the external/esmf_time_f90 directory.
clean -a, then reconfigure and recompile.
Two-way Nesting with the Nest Starting at a Later Time (V2.0.2)
(posted 6/14/04)
Problem: A time counter is
not properly set, and coarse domain fields are reinitialized.
Reported by Ryon Torn of UW.
Solution: coming in 2.0.3.
For now, do not start the nest at a later time than the coarse
domain starting time.
Specified Boundary Condition Problem (V2.0.2) (posted 6/14/04)
Problem: specified boundary
conditions did not apply to u and v correctly. Small effect in
most cases. Potential large impact on longer simulations.
Solution: download a new solve_em.F
and replace the one in dyn_em/. Recompile.
Compiling WRF 2.0.x on a Linux computer using PGI compiler version
3.x
Problem: The compiler appears
to have trouble with some 'equivalence' statements in module phys/module_ra_gfdl.F,
and the compile log shows errors like 'Conflicting
equivalence between t4 and table 3'.
Solution: download a modified
version of module_ra_gfdleta.F
and recompile.
