This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Rick here is the rough history of the official notice of the termination of the A-D grids that comprised the AVN: On Tue, 4 Apr 1995, Peggy Bruehl wrote: > This means that on April 18, all of the AVN global model output > grids will disappear from the HRS data stream. To provide global > model output coverage, the National Weather Service is broadcasting > eight 1.25x1.25 lat/lon grid octants, which when stitched together > provide a global "thinned grid". A description of the > "thinned grid" follows (from the GRIB ed 1 document by John > Stackpole): >NOUS40 KWBC 121446 >FOS NOTICE NO 661 04-12-95 >**************************NOTICE****************************** > >HRS SUBSCRIBERS > >THE SCHEDULED REMOVAL OF THE 2.5 X 5.0 DEGREE FOS BULLETINS >/2440/ FROM THE HRS ON APRIL 18...HAS BEEN CHANGED TO JUNE 5. > >**************************NMC NOTE*************************** ***************************************************************** Since longitudes converge as they approach the pole, thinned grids apply a technique which decreases the number of grid points along a latitude line as the latitudes approach the pole. A process known as thickenning is used to generate values at regular grid points after the grids are received. The current resolution of the AVN model is 1.25x1.25 degrees. The present version of gribtonc supports a command line parameter to specify the resolution to which you wish to "thicken" the grids when stored in the NetCDF file. The thickening resolution must be a multiple of the 1.25x1.25 resolution. That is, you may choose to store the data at full resolution 1.25x1.25 degrees (which takes up more space), or you can store the data at a resolution comparable to the old AVN grids 2.5x5.0 degrees (since 2.5 and 5.0 are multiples of 1.25). Here is the pqact.conf pattern used to store the grids at 2.5x5.0 resolution: # AVN model on thinned grids, interpolated to global 5.0 x 2.5 regular grid HRS ^H.[I-P]... KWB. (..)(..).*(/mAVN|/mSSIAVN) PIPE /usr/local/ldm/decoders/gribtonc -q "lin,dlat=2.5,dlon=5.0" etc/avn-x.cdl data/GRIB/(\1:yy)(\1:mm)\1\2_avn-x.nc The "-q" parameter is used for the resolution option. the "lin" option tells gribtonc to use a linear fit to seam the octants together the dlat and dlon are the delta lat and lons used. As you can see from the pattern above, the quasi-thinned grids are transmitted as grids I-P, where greids I, J, K, and L are in the northern hemispere, and M, N, O, P are in the southern hemisphere. Also note that the distribution center pattern has been changed from KWBC to KWB. since the 4th letter will be used to denote the models in the future. ********************************************************************* To keep apprised of data stream changes you can subscribe to Unidata email lists through the Unidata WWW page. Simply use your forms capable browser to reach: http://www.unidata.ucar.edu Once connected, scroll down to the section: Information of Interest to the Unidata Community Other Services Subscribe to/unsubscribe from Unidata's Mailing Lists Email lists which you may be most interested in are: nws-changes: which provides information on changes in the NWS ldm-users: which procides information about the LDM datastream: which contains information about the data feeds of course you are welcomed to subscribe to any of the lists you find meet your uses. You may also search back through archives of these email lists to find old messages such as I listed at the top of this mail message, or you can search through support messages for responses to users problems which may be germain to a problem you are having. ************************************************************************ So, Now to get you up to speed so that you can receive decode these grids, you will need to upgrade the following: udunits (current version 1.10) netcdf (current version 2.2.4) the above two are needed for the decoders package decoders version 2.2.1 (which contains gribtonc and the cdl files) The old version of gribtonc you have is before the introduction of the thinned grids. and eventually, we will want you to get LDM upgraded to LDM-5.0 All of these packages are available through anonymous ftp to ftp.unidata.ucar.edu. download udunits from: pub/udunits/udunits.tar.Z download netcdf from pub/netcdf/netcdf.tar.Z download decoders from pub/decoders/decoders.tar.Z the cdl file used by gribtonc for the thinned grids is in the decoders tar file as: decoders-2.2.1/src/gribtonc/cdl/avn-x.cdl Each of these packages contains instructions for building, the package- which consists primarily of running the configure script, followed by making and installing the package. Building the decoders requires netcdf and udunits, so you will instructed how to specify the locations of these libraries when building gribtonc. After you get that working, you can download LDM-5.0 a compiled binary is available in pub/binary/sun4-sunos5/ldm.tar.Z Typically, you unpack the ldm distribution under the ldm directory. This will create a directory ldm-5.0 and its sub directories. It is suggested that when you get to this point you move your current ldm4.1.41 distribution that you mentioned is currently called runtime, to ldm-4.1.41 the convention is for runtime to be a link to the directory of ldm that you are currently running. So presently runtime would be linked to the ldm-4.1.41 directory. We adopted this convention to make it easier to install new versions of ldm, install them, then when you are all set up, all you have to do is cange the runtime link to your new directory. in ldm, bin is a link to runtim/bin, lib to runtim/lib, include to runtime/include, man to runtime/man. The only directory that is not linked is usually the etc directory. This is because your cdl files, pqact.conf files etc are usually the same from version to version. When you upgrade to LDM5.0, you may want to print out a copy of the site managers guide in the doc directory. Your pqact.conf file will probably be fine when changing versions, especially since you mentioned it was only 18 lines- so there isn't much that could change. You will create an ldmd.conf file in the etc directory that will soecify your feed sites, permissions etc. This replaces the old ldmadmin.conf, and ldmaccess.conf files in LDM version 4. ******************************************************************* As I mentioned on the phone, it is desireable for you to get to LDM5 since it has several data recovery characteristics for retrieving data if the network is down for a short period. However, it is not critical- yet. So Feel free to work at your pace. The main thing is getting the grids decoded into netcdf. Then when you are comfortable with that, then you can tackle ldm5.0. At that point you can configure your feed pattern to ask for only data that you are needing. Since the thinned grids are in the HDS stream, while WMO is the superset of all the FOS data, you can lessen the network load by specifying the feed type, and the pattern requested from the upstream site. If you have further questions or problems, feel free to send them to address@hidden for quickest response- since I may be out of the office, and others here will be able to answer from the support email address. Steve Chiswell User Support address@hidden