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

20010529: LDM/IDD: /p and /m portions of headers



Tom,

The /p and /m strings were added to the IDD product identifiers to help 
distinguish
between products which otherwise have identical WMO headers. The characters are
literal, eg not escape characters.

The PIL identifier exists in most US NWS bulletins and is ontained from the 
line following
the WMO identifier. That is, for NWS bulletins having the form:

TTAAII CCCC DDHHNN
NNNXXX

The ingest code in pging/wmo_header.c will determine if the line following the
WMO fits the required format of a PIL, which is 6 characters (space padded to 6 
if less),
and not all numberic. If the criteria is met, the /pNNNXXX is added to the LDM 
product
identifier.

One example of the utility of using the PIL identifiers are the NEXRAD products 
which
do not have unique WMO identifiers, eg:

SDUS55 KBOU 291200 /pFTGN0R
SDUS55 KBOU 291200 /pFTGN0V
SDUS55 KBOU 291200 /pFTGN0S
SDUS55 KBOU 291200 /pFTGVIL

Using the PIL allows 2 things"
1) you can identify whether the product is a base reflectivity (N0R), velocity 
(N0V) etc.
2) Some NWS offices have more than 1 RADAR that they dissiminate. In those 
cases, the CCCC
identifier is the same, but the NNN identifies the actual radar. For example
        SDUS52 KBHM 291200 /pBHXN0R
        SDUS52 KBHM 291200 /pEHXN0R

When trying to identify PILS,
SAUS80 KWBC 291200
METAR^M
.....

SAUS43 KCYS 291200
MTRLND^M

WWUS01 KMKC 291200
WWANC ^M

The word METAR following the SAUS80 line is not a PIL. The carriage return is 
the 6th character on the line,
so the line is not space padded to 6 characters. the MTRLND and WWANC are PILS 
since they are 6 character lines, 
and not all numeric.

For GRIB data, (now most models have migrated to KWBx from all using KWBC, 
where each model has a unique
code in the KWBx, eg RUC is KWBG, ETA is KWBE), the WMO is generally not 
unique. The model ID is obtained
from the PDS block of the grib using the model center and model ID. The ETA84 
tag is created by determining that
the bulletiin contains a GRIB message, then finding that the model center is 
"7" which is NCEP, and the model
id is "84" which is the ETA model. The /mETA84 identifies the model uniquely 
from other versions of the ETA,
such as ETA89 or ETA85, and RUC105 versus RUC86.

The NCEP model numbers are assigned to different modele, eg RUC is 86, but RUC2 
is 105. 84 is now the ETA since
all 4 runs are the same model. Previously, when 2 different ETA models were 
run, eg Early and MESO, 2 different
numbers existed.

The idea of the /p and /m chacaters is just to add additional information to 
the LDM product identifier.
The actual data  carried by the LDM is not changed in any way.

Steve Chiswell
Unidata User Support




>From: "Thomas L. Mote" <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200105291642.f4TGggp09905

>
>Robb?
>
>We just installed a new 3-channel NOAAPort system (PlanetaryData) and 
>are feeding from it to our LDM box. We have an existing 
>single-channel system from the same vendor. We bypass the LDM 
>itself for GINI products, but we feed through the LDM (using a 
>utility the company provided called pqpdinrs) for other products.
>
>I started noticing how my pqact.conf file was set up with a "/p" 
>before the PIL portion of the header (e.g., "/pN0RFFC"). I noticed 
>the "/p" doesn't actually come across NOAAPort. It was simple enough 
>to change PDI's configuration file to add a "/p" before the PIL. 
>Where does this come from in the IDD? It appears it is just plain 
>ASCII text and does not represent and special character.
>
>Now I want to feed across the GRIB products the same way. I notice 
>many of them have a PIL-like "/mETA84" or similar. This does not 
>appear to come from NOAAport. Where does the /m come from? How does 
>the "ETA84" or similar get added? I am trying to figure out how I can 
>recreate this. I would like to be able to seamlessly switch between 
>the IDD and NOAAport as needed, and I certainly don't want to have to 
>create two different pqact.conf files or put it the WMO headers 
>for every model.
>
>I posed some of these questions to the support people at 
>PlanetaryData. They didn't know the answers and thought it would be 
>more appropriate for me to contact Unidata support.
>
>Some of the questions that came up in our discussions were...
>
>Is there a document or web site that spells out all the possible IDD 
>value added information and what is expected in that information?  
>For example, "/m" apparently is used to signify model generation id 
>applied names - such as ETA, NGM, etc.  What model generation id's 
>are matched with what strings for consistency? (i.e. - are all model 
>generation ids #84 and #89 called ETA)?  What is the "/o" and any 
>other value add codes and what should follow them?  Finally, are the 
>"/*" codes always literal, as the "/p" appears to be, or do they have 
>non-ASCII representations at times.
>
>Sorry for all the "slashing" questions. ;-)
>
>Tom
>
>
>**********************************************************
>Thomas L. Mote                               address@hidden
>Associate Professor of Geography         ph:  706-542-2906
>University of Georgia                    lab: 706-542-6060
>Athens, GA 30602-2502 USA                fax: 706-542-2388
>