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

20010807: 20010628: Gempak<-dss/sfc adp


John,

The old 5.0 programmers guide doesn't have many of the new high level routines.

I create a skeleton routine that you can call with the appropriate
array of decoded data and station information and send that to you
for your use and learning. I generally follow the dc style of using the
file templates and opening and closing routines which didn't exist
back in 1990.

I believe that gddelt will erase the header information, but it doesn't actually
shrink the file size (which would require moving everything in the file
past the deletion point). You should get smaller files if you:
1) created a new file (using gdcfil and CPYFIL the navigation block)
   with the appropriate number of grid entries and then,
2) Use gdmod to copy the forecast times (f000;f006) over to the new file.

Steve Chiswell
Unidata User Support





>From: John Merrill <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200108071953.f77JrC122069

>Dear Steve:
>  We've gotten good at decoding the SCD surface observational archive
>data on our local workstation - it's pretty straightforward. I'm aware
>that there is a Gempak Programmer's Guide, and I've barely looked at it.
>It's dated 1990 - is it still relevant? If it is, and the programming
>you're going to help us with uses these libraries, well, Ruth and I can
>reduce the stupidity of questions we ask by looking through it. I'll
>prepare a sample of what I've done with the surface observations and send
>you a longer email about it.
>  I have another question, related to some nogaps model/forecast files
>which Greg Stossmeister prepared for me (for the ACE-Asia project). There
>are numerous forecast times in the files, out to 144 hours. I don't need
>forecasts after 6 hours (and the 3 hour forecast fields are very limited).
>If I use gddelt to delete forecast times the file size remains unchanged.
>I assume that the only way I can get smaller files is to read and re-write
>the files, presumably using the Gempak library. Is this something which
>ahs been done before? I looked in the support archive and didn't find
>anything I recognized as being connected to this issue. Thanks.   John
>
>John Merrill                                     Telephone:  401-874-6715
>Graduate School of Oceanography, URI             Fax:        401-874-6898
>
>On Fri, 29 Jun 2001, Unidata Support wrote:
>
>> 
>> John,
>> 
>> I'd be happy to help you with this data. I'll give you the parameters
>> of my time constraints. I need to get a new GEMPAK release out shortly 
>> so that I can get all the documentation updated for the GEMPAK workshop 
>> July 30-Aug 1. I will also be updating the workshop materials.
>> 
>> After that, I am fairly free until Aug 18-26 when I'll be out of town.
>> 
>> I can probably help you most readily with the steps of programming
>> to create the GEMPAK file and write out the data etc. once you
>> have your adp stubs acessing the input data.
>> 
>> The primary output to a GEMPAK file will be the fortran REAL data
>> array. GEMPAK is well versed with routines to query the data set to see 
>> if a station has reported data at a time, so that you can merge
>> your archived data into the mix.
>> 
>> Steve Chiswell
>> Unidata User Support
>> 
>
>