[Wrf-users] landuse, tmn, changing to land points

Michael John Shaw michaejs at cires.colorado.edu
Wed Nov 15 15:10:27 MST 2006


Hello all.

Just another try at getting some response to:

Any ideas as to why water points are converted to silty clay loam  soil
type (thus seaice takes silty clay loam soiltype, which matters in Noah
LSM)?  This seems to only happen when using wrf_real_input from
mm5pconvert (in namelist for real.exe, surface_input_source = 2), even
though, e.g., the LANDMASK, VEGCAT, and LU_INDEX look ok (other than for
the seaice, which is 0 rather than 24 in wrfinput_d01 for the mm5pconvert
case, which was an initial tip of these questions).

Any ideas as to why Noah LSM appears to use veg tables for land use
categories where this is printed, "LANDUSE TYPE = USGS FOUND           27
CATEGORIES", below, while usgs landuse tables are actually 24 category?

Has anyone else experienced either or both of these?

Incidentally, has anyone seen issues with wrfsi grid shift for
non-nested wrfsi processing?  (I think I remember seeing that the shift
that people saw was in nested cases, at least, but I seem to be seeing
this problem with non-nested cases.)

Thanks.

> Hi.
>
> I have a few questions.
>
> First, lu_index seems to only indicate seaice versus non-seaice in my
> output, and like seaice is incorporated as "permanent ice".  It does not
> appear that there is another reasonable landuse variable to check this
> (xlus e.g.?) in my output or input files.  It does appear that vegcat
> could serve as landuse, though from a quick look into the code, it looks
> more like lu_index is important to have correct.  Incidentally, the
> output file indicates that mminlus is "USGS" (if I ncdump, e.g.), at
> least when I've tried using ncep1 data (which was processed with
> mm5pconvert), but not with ncep2 data (which was processed
> with wrfsi all the way through).  Perhaps relatedly, this is echoed by
> wrf:
>
>
>
> WRF V2.1.1 MODEL
>  DYNAMICS OPTION: Eulerian Mass Coordinate
>  med_initialdata_input: calling input_model_input
>  INPUT LANDUSE =
>  LANDUSE IN INPUT FILE DOES NOT MATCH LUTABLE: TABLE NOT USED
>  INITIALIZE THREE Noah LSM RELATED TABLES
>   INPUT LANDUSE = USGS
>   LANDUSE TYPE = USGS FOUND           27  CATEGORIES
>   INPUT SOIL TEXTURE CLASSIFICAION = STAS
>
>
>
> which seems to indicate that there is a landuse mismatch, yet USGS landuse
> IS being obtained.  Any thoughts on this?
>
> I also get lots of tmn errors echoed by real.exe, yet my tmn looks ok in
> output.  In the input files (wrf_real...), tmn looks very different
> from the output values - it's a nearly uniform field that gets
> slightly cooler toward the pole.  The errors are, e.g.,
>
>
>
> i,j=            5           11
>  landmask=    1.000000
>  tsk, sst, tmn=    273.1546        273.1154        0.000000
>  error in the TMN
>  is the echo by real.exe.
>
>
>
> Any thoughts?
>
> Finally, real.exe echoes a bunch of this:
>
>
>
>  changing to land at point             6           10
>   1   2   3   4   5   6   7   8   9  10  11  12  13  14  15  16  17  18 19
> 20  21  22  23  24
>   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0  0
> 0   0   0   0   0
>
>
>
> followed immediately by this:
>
>
>
> forcing artificial silty clay loam at 2035 points, out of   3726
>
>
>
> Any thoughts?  I'd love to get some feedback before digging much more into
> the code, e.g..
>
> It is hard to tell from any other output whether landuse is being properly
> utilized (though lu_index being such a mess does not look like a good
> sign!).
>
> Thanks.
>
> _______________________________________________
> Wrf-users mailing list
> Wrf-users at ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/wrf-users
>




More information about the Wrf-users mailing list