[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20011204: NMAP Users Guide



>From: "ORAM, TIM (JSC-ZS) (USA)" <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200112050601.fB561xN15671

>I have a distribution of GEMPAK including the NMAP and NMAP2 applications
>from the SOOSAC ftp site.  Unfortunately, there is no user's guide
>describing the configuration of the NMAP tool.  I am having difficulty
>getting model grids to display, although the drawing tools (product
>generation) works fine.  Do you have a users guide or installation guide for
>NMAP?
>
>I had sent this inquiry to HPC who forwarded the question to Mary Desjardins
>group.  Steve Chiswell left a message that my inquiry had been routed to
>him, although I have to admit I am perplexed as to why the NWS cannot answer
>a question from a NWS office about NWS software.
>
>Thanks,
>Tim
>-----------------------------------------------
>Tim Oram
>Meteorologist
>NWS Spaceflight Meteorology Group
>Ph: 281-483-3246
>e-mail: address@hidden
>

Tim,

I package the GEMPAK distribution from NCEP with other programs such as Garp
and create the configurations in order to allow the programs to find
the data etc. this is the distribution that I pass on to Bob Rozumalski
at COMET for the SOO/SAC. Bob makes his own configuration changes for SOO data 
sets.

Bob doesn't use NMAP with the SOOs, so I'll be happy to help you with the 
configuration. Since you aren't getting your data from the same sources as 
our universities, you probably will have to do some localization.

In the case of NMAP, the configuration that Mary desJardins group provides
stores model data in separate directories depending on the model. This
is a very useful method for accessing data. Unfortunately, Garp cannot handle 
this. So I create configurations that allow both NMAP and Garp to find the 
model 
data in a single directory. I was passed your number on since NCEP does not use
the same configurations that I do.

The basic thing to understand when configuring NMAP to find data is the
template file in $GEMTBL/config/datatype.tbl.
http://www.unidata.ucar.edu/packages/gempak/tutorial/filetemplates.html

In NMAP2, when you click on "grid" in the data selection, it presents a list of 
models
it knows how to create products for. These models are those that are found in 
the products 
given in $GEMTBL/nmap/mod_res.tbl (that is, the list of models that pops up
is generated from those models found in all the product lists in the 
mod_res.tbl file).

When you select a model, ETA, for example, it looks for this ALIAS in the 
datatype.tbl
cconfiguration file. The default I provide for ETA is to use the template:
$MODEL/YYYYMMDDHH_eta211.gem. That is the naming that I use as the default for
dcgrib2 decoding (defined in $GEMTBL/grid.gribkey.tbl).

I place all the grids in one singe $MODEL directory since the Garp_defaults
table can only have one grid directory. In reality, NMAP could have a different
directory for every model if you didn't want to use the same directory structure
for Garp.

NMAP2 will then look at the dats of all the model runs it finds that matches 
the 
template, and provide that as a selection in the next popup. After selecting 
a time, you get the list of products for that model as provided in mod_res.tbl.

I define the location of $MODEL in the Gemenviron file. There is also
another variable called $HDS in the Gemenviron file that I genereally set to
the same location and reference it in Garp_defaults. But, I kept them as
separate variables in case anyone wanted to use the package differently.

In Garp_defaults, the model keys have strings like "eta211" which work
very similar to the datatype.tbl configuration file. That provides the
string to search for file names in the gridded data directory. That string
is also used in the $GARPHOME/fdf directory for the pulldown
lists of functions for different models.

The datatype.tbl file is ues by all gempak programs. For instance, in 
GDINFO, you can specify the GDFILE=ETA. That will use the most recent
model run for that template. Or you can use GDFILE=ETA|011205/0000 to
specify a certain ruun time of the ETA (phelp gdfile). You can use this 
to verify that your programs can correctly locate the data sets by the 
templates.
If you can, then NMAP2 should be able to find the data.


Steve Chiswell